INTERNATIONAL MARKET SELECTION –

SCREENING TECHNIQUE:

Replacing intuition with a multidimensional framework

to select a short-list of countries

A doctoral dissertation

Richard R. Gould

R M I T University
Faculty of Constructed Environment
School of Social Science & Planning
GPO Box 2476V
Melbourne Vic 3001
Australia

Email: [email protected]
[email protected]

August, 2002
 ii

DECLARATION

This is to certify that this work:

1) is that of the candidate alone, except where acknowledged;

2) has not been submitted previously, in whole or part, to qualify for any
other academic award;
3) is the result of work which has been carried out since the official
commencement date of the doctoral research programme, initially at
Deakin University then at RMIT University, and
4) has not benefited from paid or unpaid editorial contributions other than
those of my supervisor.

...............................
R. R. Gould
 iii

ACKNOWLEDGMENTS

The topic originated from managing export consultants in the Australian

Trade Commission. It then developed into a strong, long-term personal
interest when enquiries to other Trade Commissions, then a formal review of
the literature, showed that this very practical question had not been
adequately answered.

An addiction such as this embroils both willing and unwitting assistance

from many people and organisations.

In the academic area, my supervisor Associate Professor Mark

McGillivray has been my mentor, collaborator and guide. His intellectual input
on conceptual matters, his advice about methodology and statistics, and his
help interpreting unexpected findings have been substantial. I deeply
appreciate his assistance.

Associate Professor Lloyd Russow of Philadelphia University requires

mention for his important advice at distance, he being the previous contributor
to this area of knowledge.

At a personal level, my marital partner, Magda, has been a most visible

contributor with advice, support and an untiring sense of humour. Our
children, Jacqui and Ian, have been both bemused onlookers and supportive
participants. In the same way that rigorous MBA courses include a
valedictory appreciation to the family, this doctorate justifies the same
attribution.

There have been a number of organisations and individuals that have

generously provided data, sometimes at some inconvenience because it has
been either atypical or difficult to provide to an outsider. These have been:
 iv

Mr Laza Kekic, Regional Director, Central & Eastern Europe

Economist Intelligence Unit
15 Regent St
London, SW1Y 4LR
United Kingdom
for country data on legal arbitrariness

Mr Orio Tampieri, Statistics Division

Food & Agricultural Organisation (FAO)
Viale delle Terme di Caracalla
Rome
Italy
for country data on production and trade of beef and wool

Ms Jocelyn Troussard, Energy Statistics Division

International Energy Agency (IEA)
2, rue André-Pascal
75775 Paris CEDEX 16
France
for country data on production and trade of coking coal

Dr Niels Noorderhaven
The Institute for Training in Intercultural Management (ITIM)
1190 AB Ouderkerk aan de Amstel
The Hague
Netherlands
for cultural distance data for countries additional to those in
Hofstede (1980)
 v

Mr Brent Schulz, Vice President

Power Systems Research
1365 Corporate Center Curve
St. Paul, Minnesota 55121
U. S. A.
for country data on production and trade of petrol engines

Adrian Shute, Marketing Manager

The PRS Group
6320 Fly Road
East Syracuse, NY 13057
U. S. A.
for country data on market risk

Vasyl Romanovsky and

Joseph Habr
United Nations Statistics Division
United Nations
New York, NY
U. S. A.
for country data on production and trade in fixed telephone sets

Ms F. Tandart, Division of Statistics

UNESCO
7 Place de Fontenoy
17353 Paris 07SP
France
for country data on production and trade in undergraduate
degrees

To all these visible, and to the unsung, supporters,

my sincere and heartfelt ‘Thanks’.
 vi

TABLE OF CONTENTS
Page
DECLARATION ii
ACKNOWLEDGMENTS iii
TABLE OF CONTENTS v
LIST OF FIGURES v
LIST OF TABLES v
SUMMARY 5

CHAPTER 1 INTRODUCTION 5
1.1 Introduction 5
1.2 Statement of the Problem 5
1.3 Objectives of This Research 5
1.4 Summary of Methodology 5
1.5 Hypotheses 5
1.6 Overview of Chapters One to Seven 5

CHAPTER 2 INTERNATIONAL MARKET SELECTION 5

2.1 Introduction 5
2.2 Market Screening in Context 5
2.3 Definition of Screening 5
2.4 The Role of Screening 5
2.5 How Many Markets Should Screening Deliver? 5
2.6 Components of a Screening Methodology 5

CHAPTER 3 LITERATURE REVIEW 5

3.1 Introduction 5
3.2 Do Organisations Presently Screen Markets? 5
3.3 Theoretical Foundations for Screening 5
3.4 Previous Studies of Screening 5
3.4.1 Intuitive Approaches 5
3.4.2 More Structured Approaches 5
3.5 Screen First, or Select Mode First? 5
 vii

(Chapter 3 continued)
3.6 Conclusions about Gaps in the Literature 5
3.7 Specific Research Questions 5

CHAPTER 4 SCREENING: A CONCEPTUAL FRAMEWORK 5

4.1 Introduction 5
4.2 Description of Proposed Framework & Variables 5
4.2.1 Dependent Variable 5
4.2.2 Independent Variables 5
4.2.3 Intervening Variables 5
4.2.4 Moderating Variables 5
4.3 Underlying Theory 5
4.3.1 Dependent Variable 5
4.3.2 Independent Variables 5
4.3.3 Intervening Variables 5
4.3.4 Moderating Variables 5
4.4 Measurement of Dimensions 5
4.5 Conclusion 5

CHAPTER 5 METHODOLOGY AND DATA 5

5.1 Introduction 5
5.2 Research Design and Validity 5
5.3 Instrument Selection, Development, Validity and
Reliability 5
5.4 Sampling: 5
5.4.1 Product Sample Selection Rationale 5
5.4.2 Products Selected 5
5.5 Data Collection Method 5
5.5.1 Secondary Data 5
5.5.2 Primary Data 5
5.6 Variables Tested 5
5.7 Variable Weights 5
5.8 Statistical Analysis 5
5.8.1 Principal Components Analysis (PCA) 5
 viii

(Chapter 5 continued)
5.8.2 Cluster Analysis (CA) 5
5.8.3 Multiple Discriminant Analysis (MDA) 5

CHAPTER 6 RESULTS 5
6.1 Introduction 5
6.2 Principal Components Analysis (PCA) 5
6.3 Cluster Analysis (CA) 5
6.4 Multiple Discriminant Analysis (MDA) 5
6.4.1 Preliminary Matters 5
6.4.2 The Null Hypothesis 5
6.4.3 Relative Dimension Importance 5

CHAPTER 7 INTERPRETATIONS AND CONCLUSIONS 5

7.1 Introduction 5
7.2 Interpretations and Conclusions 5
7.2.1 Principal Components Analysis Results 5
7.2.2 Cluster Analyses Results 5
7.2.3 Multiple Discriminant Analyses Results 5

CHAPTER 8 MAIN FINDINGS, LIMITATIONS AND

RECOMMENDATIONS 5
8.1 Introduction 5
8.2 Summary of Chapters One to Seven 5
8.3 Limitations 5
8.4 Recommendations for Future Research 5

APPENDICES
APPENDIX A Management Questionnaire 5
B Measurement of Variables 5
C Researcher Questionnaire 5
D Product Definitions 5
E List of Variables 5
F Values of Variables – Summary 5
 ix

(Appendices continued)
G Language Scores by Country and Organisation
Competence Level 5
H Selecting the Optimal Number of Clusters 5
I Countries in Each Cluster, by Product and Level of
Organisational Capability 5
J Plots – Education: Country Scores By Variable 5
K Relative Importance of Variables 5

BIBLIOGRAPHY 5

REFERENCES 5

LIST OF FIGURES

Figure No. Title

3.1 Inventory and Taxonomy of Statistical Approaches to
International Market Screening 33
3.2 Selection of Foreign Markets: Six Steps 5
4.1 Summary of A Contingent Eclectic Approach to Screening
International Markets 5
4.2 A Contingent Eclectic Approach to Screening International
Markets (the full framework) 5
4.3 An Alternative Perspective of A Contingent Eclectic
Approach to Screening International Markets 5
4.4 Porter-Lawler Motivation Model 5
4.5 Proposed Cultural-related Influences on International
Market Screening 5
4.6 Locating Screening in the Personality-Learning-
Competence-Performance Chain 5
7.1 Two Types of Discrimination Undertaken by Each
Relevant Predictor Variable 5
 x

LIST OF TABLES

Table No. Title Page

5.1 Four Types of Independent, Moderating and Intervening
Variables Used to Generate Three Kinds of Market
Selection 5
5.2 Test of the Impact of Each Generic Variable, Using Three
Types of Independent, Moderating and Intervening
Variables 5
5.3 Demonstration Organisational Objectives Weights 5
5.4 Summary of Weights for Each Independent, Moderating
and Intervening Variable Used to Generate the Seven
Kinds of Market Selection 5
6.1 Total Variance Explained 5
6.2 Communalities 5
6.3 Rotated Component Matrix 5
6.4 Final Cluster Centre Values in K-Means CA 5
6.5 Numbers of Markets in Each Cluster – All Products 5
6.6 High Potential Markets Proposed by Screening – Each
Product 5
6.7 Collinearity: Competition Variable with Market risk Variable
– Bivariate Correlation 5
6.8 Collinearity: Competition Variable with First Economic PC
Variable – Bivariate Correlation 5
6.9 Multicollinearity: Competition Variable – Multiple Correlation,
as Indicated by Tolerance 5
6.10 Statistical Significance of the Canonical Discriminant
Functions, Using Wilks' Lambda 5
6.11 Classification Results (Hit Ratios) Compared with Chance 5
6.12 Tests of Equality of Group Means for Statistical Significance 5
6.13 Variable Importance: Hierarchy Calculated Using Three
Methods 5
7.1 Commonality Among Markets Selected 5
 1

SUMMARY

The aim of this research was to develop an international market

selection methodology which selects highly attractive markets, however
defined, and which allows for the diversity in organisations, markets1 and
products.

Conventional business thought is that, every two to five years, dynamic

organisations which conduct business internationally should decide which
additional foreign market or markets to next enter. If they are internationally
inexperienced, this will be their first market; if they are experienced, it might
be, say, their 100th market. How should each select their next international
market?

This is not an easy decision. There are some 230 countries in the
world. Each has its own characteristics in terms of market attractiveness.
Moreover, judging from the relevant literature, a well-informed selection
decision could consider over 150 variables that measure aspects of each
foreign market’s economic, political, legal, cultural, technical and physical
environments (Root 1994; Russow 1989), along with the organisation’s
resources, managerial abilities and preferences. Considering even a fraction
of these variables for a reasonable number of countries would, in the
overwhelming majority of cases, be beyond the resources of the decision-
maker in terms of time, money and expertise. The relevant literature
suggests, therefore, that international market selection ought to be broken up
into two or more stages (see, for example, Ball & McCulloch 1982; Douglas,
Le Maire & Wind 1972; Douglas, Craig & Keegan 1982; Liander et al. 1967;
Root 1987; Russow 1989; Russow & Okoroafo 1996; Samli 1972; Toyne &
Walters 1993; Walvoord 1980; and, for comprehensive surveys,
Papadopoulos & Denis 1988 and Samli 1977).

1
Whilst markets commonly flow across national borders, aggregate statistics rarely do so.
For the purpose of this study and consistent with the literature, markets are defined at the
level of countries.
 2

The first stage involves screening a relatively large number of countries

on the basis of a select, yet limited, number of criteria. The outcome of this
(screening) stage is the selection of a much smaller sub-sample of countries.
The screening stage ought to not only result in the selection of the right group
of markets. As Root (1994, p.55) observes, it should also minimise two
errors: (1) ignoring countries that offer good prospects for the organisation’s
product, and: (2) spending too much time investigating countries that are poor
prospects. Root argues that the first of these errors can be avoided by
including all countries in the initial sample – a significant matter in terms of
resources and computing sophistication. A more intensive, in-depth
assessment of the market potential of the sub-sample of countries selected in
the completion of the screening stage is then recommended during the
second and later stages of market selection. This in-depth assessment
involves consideration of a larger number of variables, possibly measuring
such factors as industry-specific and/or company-specific sales potential,
which may include data gained from primary sources. The outcome of this
stage is the selection of an even smaller sub-sample (usually, but not
necessarily, one country), to which the organisation typically exports.

However, the screening decision can not presently be made properly

because there is insufficient knowledge about which of the abovementioned
150 or more decision variables to use. An unsolicited order (found by Bilkey
(1978) to be the market selection mechanism used by 67 per cent of firms),
the language spoken by a key manager, or even where a manager had
holidayed overseas, are just a few of the common practical geneses. The
consequence is that, in Australia alone, an estimated $A225,000,000 per
annum (see section 1.1) is wasted every year as a result of inefficient and
ineffective foreign market selections.

Empirically determining which variables are important requires

developing a large database. Sufficiently accurate data about meaningfully
defined products to enter into the database only began to become available
during the 1980s. Cross-disciplinary knowledge of economics and psychology
is required. And even a powerful personal computer takes several minutes to
 3

perform the 3,000,000 calculations required for most of the market selection
cluster analyses. The result is that Russow (1989) is the only previous
researcher to have attempted to empirically determine which variables are key
when selecting foreign markets.

This study began by collecting all market selection variables suggested

by 350 items in the literature. It then rationalised them into a comprehensive,
eclectic, theory-driven framework and tested them empirically using principal
components analysis, cluster analysis and multiple discrimination analysis. Of
the 46 potential variables, 30 were used as representative of the 38 that can
be quantified.

Of the framework’s 30 operationalised dimensions:

• seven variables captured attributes of each market (e.g. product

consumption; their economies);
• four variables captured the organisation’s abilities in each market
(e.g. languages; contacts);
• nine variables captured organisation abilities and psychological
values (e.g. general business competencies; values about further
internationalising);
• four variables captured organisation goals (e.g. sales growth; risk
minimisation); whilst
• six variables are required for technical purposes (e.g. to assess
data quality and availability; to set the criteria and method of market
evaluation).

There are eight additional variables which this study was not able to
operationalise as quantifiable criteria (e.g. strategy, marketing mix). A further
eight variables were not required for the particular products and organisations
assessed (e.g. product physical attributes; organisation objective of other
business benefits).
 4

A database of some 10,000 statistics was developed covering six

products and approximately 230 countries. Scores were constructed for two
hypothetical organisations having respectively the minimum necessary, and
the maximum possible, abilities and attitudes. The data were then treated
using the statistical techniques of principal components analysis, cluster
analysis and multiple discriminant analysis. The objective of principal
components analysis was to compress four economic sub-variables into a
single component which was then used to indicate the state of the economy in
each country. Cluster analysis grouped markets into sub-sets having differing
attractiveness. Discriminant analysis tested the cluster analyses, assessed
relative importance of variables, and added shades of understanding about
the solutions developed. Three sets of results were computed for each
product – for a best markets only, for a low capability and values organisation,
and for a high capability and values organisation2. These 18 solutions were
supplemented with two solutions for one product to demonstrate how to add
organisational objectives to the market screening procedures.

The approach herein is prescriptive, being that of the maximising

rational economic person or Allison’s unitary actor (Allison 1971; Schoemaker
1993). It was originally suggested by Liander et al. (1967). Subsequent
decision sub-optimisations due to organisational politics, strategy, human
limits and so on are not considered.

Using the 30 variables produced results which typically accounted for

95 per cent of variance at cross-classification rates of 95 per cent. Significant
findings of the research are as follows:

1) The procedures succeed at allocating markets to groups having

differing levels of attractiveness,
2) Different groups of markets are appropriate (“best”) for
organisations having different abilities and values,

2
As detailed later, these 3 situations consider attributes of the market only, attributes of the
market and a low capability organisation, and attributes of the market and a high capability
organisation.
 5

3) Findings (1) and (2) mean that a “One size fits all” approach to
international market selection is not correct. The “best” (most
attractive and appropriate) markets for organisations of high
international capability will usually be different from those for low
capability organisations. The specific attributes of the
organisation (competencies and values) help determine the
market in which it will harvest the largest profit. Governments
are thus incorrect when they indicate that particular markets are
the most appropriate for all organisations in a particular industry
to target,
4) All seven variables describing the markets, and all four variables
describing the organisation’s abilities in each market, were found
to have a role in producing solutions for one or more of the 20
product-organisation combinations tested,
5) The strength of their importance differed by product and the
organisation’s level of competence,
6) There is no universal rule about which variables are most
important in all situations. However, there are indications about
which variable will dominate market selections for a specific
product and level of organisational competence,
7) The other 19 variables used by the framework describe the
organisation and prescribe the market selection system. They
also all appear to play roles in determining which markets are
most attractive and appropriate but the level of certainty about
this is not as high as for the above 11 variables.

The contributions to knowledge made by this work are that it has:

• collected data on arguably all the influential variables mentioned in

the literature,
• developed a predictive framework within which to house those
variables (see Figure 4.3),
 6

• extended Russow’s (1989) demand-side only market assessment to

include firm-specific and business environment considerations
which affect the market screening decision,
• undertaken the second quantitative test of market screening,
producing results which typically accounted for 95 per cent of
variance at cross-classification rates of 95 per cent, and
• uncovered two significant weaknesses in the cluster analysis
methodology recommended by the lineage descended from Liander
et al. (1967).
 7

CHAPTER 1

INTRODUCTION

Despite the significance of the initial entry decision,

little is known about the actual process by which firms
come to identify both foreign markets and specific buyers
within those markets.
(Ellis 2000, p. 443)

1.1 Introduction

This chapter introduces the topic. The commercial dilemma, which led
to the research being conducted, is described, then the problem is stated, the
research objectives are set, the methodology followed is outlined, and formal
hypotheses are stated. The chapter closes with a summary of each of the
dissertation’s eight chapters.

How should an organisation attempt to select one country from the 230
possibilities around the world to market its product in? Academe treats the
problem as solved: we offer a list of several hundred variables, and tell the
organisation to pick the relevant ones. However, we do not advise how to
pick the correct variables because we do not know which are most important.
Further, it would typically take an organisation considerable research to learn
what selection processes it is supposed to apply using the variables which
determine the best markets. That answer is in the literature – as shall be
demonstrated shortly – but is only mentioned in a small number of the
teaching texts.
 8

Businesses are thus effectively on their own when dealing with the
problem. Yet they must solve it to become, remain or expand internationally.
Their usual response is to short-cut the process by: (1) not attempting to
assess all markets, and (2) not using formal decision criteria to produce a
short-list of markets. They generate an intuitively developed short-list of likely
markets using irrational or incomplete criteria (Bilkey 1978; Russow 1989),
then apply appropriate and intensive effort to picking the best. However, an
even better market could easily have been amongst those which they did not
consider. The result for the organisation and society is an un-researched, but
probably high, aggregate opportunity cost.

That cost comprises the four possible components of opportunities

foregone, excess effort applied through entering a more difficult market than
necessary, market entry failures, and constrained strategic options. The
wastage from being unable to screen markets is a probably significant
amount. If one assumes that these four matters produce a loss of 10 per cent
in efficiency and effectiveness on typical market entry costs of $A 150,000
(from factoring up to today’s cost the 1987 figures in Downey, Watson and
Neumann (1988c:4)), and that there are 15,000 attempted market entries per
annum in Australia3, the cost to the country is $A 225,000,000 per annum.

The literature recommends that market selection comprise the two

steps of screening markets then conducting in-depth assessment on the
short-list of preferred markets. This dissertation sees screening as requiring
an assessment of all markets (including ones in which the organisation
presently sells), principally using secondary data, and using variables which
do not presume a particular mode of entry. Research has yet to determine
which variables to use, and whether we can assign typical weights to any.

1.2 Statement of the Problem

3
From 22,000 businesses known to be exporting (Australian Bureau of Statistics and
Australian Trade Commission 2000:9) plus estimates for businesses not included in the
survey (details available from the author) bringing the number up to approximately 45,000
businesses, divided by 3 to allow for a new market entry each three years.
 9

There is little agreement in the literature about how to screen markets.

Russow and Okoroafo (1996) found that the 40 different authors referred to by
Samli (1977) and Papadopoulos and Denis (1988) all proposed different
screening criteria and measures. A well-informed decision could take account
of at least 150 measures of each market’s competitive, economic, political,
legal, cultural, technical and physical environments (e.g. Root 1994; Russow
1989), as well as organisational resources, managerial abilities and
preferences. There is also a wide range of differing methodologies proposed
to evaluate the measures and criteria.

Practitioners disregard the researchers’ advice about how to select

markets. Research by Bilkey (1978) produced the astounding finding that
approximately 67 per cent of firms actually decide which foreign market to
enter as a result of receipt of an unsolicited foreign order. Other researchers
to find that most businesses select their international markets in that and other
irrational ways are Pavord and Bogart (1975); Perkett (1963); Papadopoulos
(personal communication 18/7/1997); Simmonds and Smith (1968); Simpson
and Kujawa (1974); Sinai (1970); Snavely et al. (1964); Tesar (1975); and
others cited in Bilkey (1978). What commonly appears to extract the short-list
of markets is the evoked set process whereby market salience arises from
holidays, trade activity, cultural similarity, language spokes, and so on
(Kramer 1964; Papadopoulos 1983; Robinson 1978; Russow & Solocha 1993;
Samli 1977). The country screening decision is thus made on opportunistic,
not on rational or on systematic, grounds. It produces a solution that is
unlikely to be optimal4.

The US government’s advice to intending exporters about the first step

is to “Select ten countries of interest … (previous) knowledge of these
countries is not required, although it may be helpful” (National Trade Data

4
The optimal screening solution is the one which selects countries having the highest country
attractiveness, as defined by the organisation’s objectives, subject to motivation, commitment,
capabilities and preferences. Country attractiveness typically results from maximising
potential customer aspects, country competition and other environmental matters whilst
minimising industry competition and market risk.
 10

Bank, 1997). The author’s experience as an Australian Trade Commissioner

is that no government trade facilitation body has a satisfactory system for
selecting international markets. The sole exception, but still only a partial and
non-optimal system, is the French approach of using trade data for 100 large
countries – so excluding production, market risk and all other variables, and
also excluding more than half the countries of the world (personal discussions,
Mr Didier Bourguignon, Centre Francais Du Commerce Exterieur, Paris, 31
March, 2000).

The absence of a satisfactory screening procedure may relate to the

decision information requirements being huge (200 countries times 150
variables equals 30,000 datum needed, and these change annually). The
size and complexity of the problem presumably overwhelm the business
analyst per Simon’s (1997) concept of bounded rationality. Decision
satisficing, rather than maximising, then occurs (Johanson & Vahlne 1977, p.
26; Simon 1997, p. 118). A final contribution is made by the absence of
feedback about the opportunity cost paid by the organisation for entering the
market chosen rather than the optimal one.

Academic research contributes by providing conceptual methodology

which is incomplete at the operationalisation stage. Reading each of the 40
approaches to screening cited by Samli (1977) and Papadopoulos and Denis
(1988) leaves the reader uncertain about which conceptual and statistical
approaches should be used. Further, there is no consensus on which
variables to use, with each author proposing their preferred but untested list.
Russow (1989) is the only prior research to this dissertation which attempts to
empirically assess variable importance.

Whatever methodology is used to screen international markets should

ideally be suitable for use by big organisations and small, by new-to-
internationalisation firms and those with great experience, by born globals and
those primarily domestic, by highly motivated and by weakly motivated
organisations, and by organisations with high ability and those with little
ability. The method needs to be able to select the best from 230 countries, be
 11

able to handle either goods or services, and allow for differences in personal
preferences (e.g. towards taking risk). It’s a big ask. One may judge whether
the work described here meets those requirements.

1.3 Objectives of this Research

The question addressed by this dissertation is how organisations

should select their next international market at the market screening stage.
The literature is used to propose what is believed to be a complete list of
influential variables, then empirical tests are conducted on many of those
variables. The work particularly endeavours to determine which variables are
critically important to organisations when selecting a short-list of the most
desirable markets.

1.4 Summary of Methodology

This research followed the usual hypothetico-deductive stages of

literature review, hypotheses development, research design, data collection,
statistical analysis and dissertation writing.

The literature review accumulated some 200 independent, moderating

and intervening variables which were said to be influential upon the
dependent variable – the preferred short-list of markets. These variables
were reviewed to eliminate overlaps (e.g. Gross National Product, Gross
National Product per capita, Gross Domestic Product, etc.) and those not
fundamental to all products (e.g. agricultural production, agricultural wages,
etc.), producing a list of 46. A framework was constructed which used these
46 predictor variables (see Figure 4.2). Each variable is individually
supported in the literature as well as being part of at least one of several
overlapping, underpinning theoretical approaches. The framework is thus
contingent upon the literature’s variables and is supported by eclectic theory –
a contingent, eclectic approach to screening international markets.
 12

The framework was then tested empirically in ways that will be

described below using a one shot case study comprising six cases. The null
hypothesis and three research hypotheses were developed. Six diverse
products were chosen using a two-step heterogeneous-sample selection
process based on strategic criteria and strata. The products thus selected
were coking coal, beef, wool, fixed telephones handsets, undergraduate
degrees and internal combustion engines. Scores were estimated for
organisations at the low and high ends of the continuum of internationally
relevant capabilities and psychological values. Actual country data for the
other variables were collected for the years 1990 and 1995.

Statistical analysis was undertaken using SPSS software. Principal

components analysis reduced four economic sub-components to a single
component which represented multiple economic facets of each market.
Cluster analysis sorted markets into groups having different levels of business
potential, especially markets of high potential. Multiple discriminant analysis
tested the number and composition of clusters indicated by cluster analysis,
considering whether the cluster centroids were statistically different from each
other and whether the assignments of countries to clusters could have
occurred by chance. Multiple discriminant analysis also assessed the relative
importance of each screening variable – a major interest of this study – using
structure matrices, potency indices and the size of the univariate F ratio. The
variables were brought into play in three steps to assess their impact in
determining three situations: the best markets per se, the best markets for
organisations with low and high competencies and values, and the best
markets for organisations with low and high competencies and values when
their organisational objectives are included. This produced results for 20
situations to then be assessed, and from which to draw conclusions.

1.5 Hypotheses

Statistical testing of the proposed screening framework occurred using

the following hypotheses:
 13

H0 There are no differences among the clusters of countries

produced using cluster analysis and the selection variables
proposed,
H1 Statistically significant differences exist among the centroids of
the country groups,
H2 The discriminating power of the multiple discriminant analysis
(MDA) functions exceeds that of chance group assignment,
H3 The product-specific consumption and consumption growth
variables are the most important characteristics for
discriminating among country groups.

The null hypothesis asserts that use of the screening framework will
produce clusters of countries having no differences in market attractiveness.
The first two research hypotheses provide the statistical tests which would
lead to lack of support for the null hypothesis, whilst the last research
hypothesis indicates expectancies about the relative importance of two of the
independent variables which produce the country clusters.

1.6 Overview of Chapters One to Seven

The dissertation is organised as follows:

Chapter one introduces the topic. The commercial dilemma which led
to the research being conducted is described, then the problem is
stated, the research objectives are set, the methodology followed is
outlined, and formal hypotheses are stated. The chapter closes with a
summary of each of the dissertation’s seven chapters;

Chapter two begins by locating screening in the internationalisation

process which is followed by organisations, then proceeds on to define
market screening, describe the role that screening plays, question the
number of markets that should be delivered by screening, and detail
the components which should be included in a screening methodology;
 14

Chapter three presents the relevant literature. It shows whether

organisations do presently screen markets, considers the theoretical
foundations for screening, summarises the numerous previous studies
of screening, and puts a position on whether to screen markets first or
to select the mode of distribution first. Those matters allow conclusions
to be drawn about gaps in the existing literature, and lead to the
research questions addressed by this work;

Chapter four presents the conceptual framework which, it is suggested,

an organisation should use when making a foreign market screening
decision. The chapter has two sections: firstly, a description of each of
the 46 selection variables and the key impacts that each is expected to
make on the choice of a group of highly attractive markets; secondly,
the underlying reasoning which caused the use of each variable is set
out. The first section uses plain English to explain the concepts,
leaving to the second section the more complex terminology associated
with explication of theory and the bulk of the citations from the
literature;

Chapter five describes the research design chosen and discusses the
related validity issues. It then traverses instrumentation matters,
sampling of the chosen products, data collection methodology, the
variables which were tested, calculates variable weights, and discusses
the three statistical procedures used;

Chapter six presents the results, which are sequenced in the order that
the statistical routines needed to be carried out: principal components
analysis, cluster analysis and multiple discriminant analysis;

Chapter seven provides interpretations and conclusions for the three

statistical procedures – principal components analysis, cluster analysis,
and multiple discriminant analysis. Of particular import are the country
assignments for the 20 situations examined, the tests of the statistical
 15

significance of the assignments and the assessments of the relative

importance of each dimension used;

Chapter 8 commences with a summary of the seven previous chapters

in the dissertation, provides a discussion of the limitations of this
research, and then closes with recommendations for future research.
 16

CHAPTER 2

INTERNATIONAL MARKET SELECTION

An Ancient Chinese Classification of Animals:

(a) those that belong to the Emperor, (b) embalmed ones,
(c) those that are trained, (d) suckling pigs, (e) mermaids,
(f) fabulous ones, (g) stray dogs, (h) those that are included
in this classification, (i) those that tremble as if they were
mad, (j) innumerable ones, (k) those drawn with a very fine
camel’s hair brush, (l) others, (m) those that have just
broken a flower vase, and (n) those that resemble flies from
a distance.
Jorge Luis Borges, Other Inquisitions: 1937-1952
(in Aldenderfer & Blashfield 1984, p. 7)

2.1 Introduction

The previous chapter introduced the topic from both the practical and
theoretical perspectives, then overviewed the subsequent parts of the
dissertation. International market screening is part of the international market
selection process, and international market selection is part of the
internationalisation process. An understanding of screening requires it to be
placed in relation to these two matters.

Chapter two begins by locating screening in the internationalisation

process that is followed by organisations, then defines market screening,
describes the role that screening plays, questions the number of markets that
 17

should be delivered by screening, and details the components which should

be included a screening methodology.

2.2 Market Screening in Context

Market screening is one of many steps in the internationalisation

process. Analysis of the literature suggests to this researcher that
internationalisation is a process that ideally involves:

1) perceiving a positive or negative stimuli to internationalise;

2) a decision to explore, or accept impelled, internationalisation;
3) undertaking screening of markets to produce a short-list;
4) performing in-depth research on and analysis of the short-listed
markets;
5) final selection of the market or markets to enter;
6) selection of the mode of entry;
7) a decision to proceed with market entry;
8) preparation for and entering the chosen market;
9) consolidation in the new market;
10) repeating the process;
11) expansion in some of the already entered foreign markets;
12) global rationalisation.

The above is an original synthesis of the literature. There is not

universal agreement about either the steps comprising a firm’s
internationalisation nor the sequence because, in practice, steps are
sometimes omitted entirely whilst others are sometimes undertaken in a
different order (see, for example, Leonidou and Katsikeas 1996 for a review of
the literature about the numerous models comprising different numbers of
stages of internationalisation). Debate about those matters relevant to
screening is provided later in the dissertation.

Steps (3), (4) and (5) comprise international market selection whilst
steps (3) and (4) comprise international market assessment. Although market
 18

assessment is often conveniently simplified as involving ‘screening’ and ‘in-

depth assessment’, there are differing views about where to draw the line
between components of one part of a sub-task and another. For example,
should assessment of economic aspects be undertaken during the screening
step, the in-depth step, or during both? The purpose for the market
assessment relates, as will be seen shortly.

Toyne and Walters (1993, pp. 294-6) suggest three purposes for
foreign market assessment: (1) market entry assessment, (2) marketplace
assessment, and (3) non-economic assessment. Market entry assessment
identifies and selects among opportunities in new markets, and comprises
screening and in-depth assessment. New markets are the focus of that
research and this dissertation. Marketplace assessment covers changes
within an existing market; non-economic assessment evaluates social and
political environments in existing or prospective markets. Neither
marketplace, nor non-economic, assessment were the subject of this
research.

Toyne and Walters (1993) recommend breaking a market entry

assessment into four further steps. These are: (1) a preliminary stage, (2) an
initial assessment, (3) an advanced assessment and, finally, (4) an internal
trade-off analysis. They see the preliminary stage as eliminating markets
which are unattractive due to domestic regulations and management
preferences, whilst the initial market entry assessment eliminates markets due
to economic and competitive factors. An advanced assessment is an in-depth
evaluation of the few remaining markets for competition, market barriers, and
market demand. Their step four compares the remaining market/s with
company objectives, current activities and resources to determine their validity
for expansion. Screening thus comprises the ‘preliminary’ and ‘initial’ steps of
Toyne and Walters’ market entry assessment.
 19

A slightly different paradigm is described by Douglas and Craig (1995).

They distinguish between: (1) market entry, (2) market expansion, and (3)
global rationalisation. Market entry is the first and subsequent entries to a
foreign market. Local market expansion is the seeking of growth in markets
where operations have already been established. Global rationalisation
involves developing and implementing strategies which improve efficiency of
the firm’s existing worldwide operations.

Douglas and Craig’s (1995) description of the initial market entry stage
includes both the screening stage and the detailed market research stage.
However, they see screening as only one of several possible ways of data
handling:

“Approaches (can) range from qualitative evaluation and/or ranking of

data, to the development of elaborate simulation models.” … “One
approach ... is a sequential screening approach (in which) ... a number
of sequential screens of relevant variables can be established, and
countries or markets assessed successively on each screen. At each
stage, those that do not pass certain minimal cut-off points or are
ranked lowest on a given screen can be eliminated” (p 64).

A final matter is the ambiguity in the literature about one of the key
selection criteria, the assessment of product potential. Aside from
distinguishing between existing, unmet and latent demand, this can be:
(1) worldwide demand, (2) industry demand, or (3) company demand. Root
(1987) and Bradley (1991) are proponents of collecting these latter three data
when screening.

‘Market selection’ thus means different things to different researchers.

In consequence, the activity of screening requires definition.
 20

2.3 Definition of Screening

Foreign market selection comprises between two and four steps (see,
for example, Albaum et al. 1994; Bradley 1995; Hassan & Blackwell 1994;
Jane 1993; Root 1994; Walvoord 1980). Key ingredients are quickly and
cost-effectively reducing a large number of potential markets to a small
number (screening), then undertaking in-depth investigation of the short-list to
select the preferable one or ones to enter.

International market screening has a number of names, including basic

needs assessment, filtering, go or no-go analysis, initial foreign market
analysis, initial foreign market assessment, initial opportunity analysis, market
scanning and preliminary market selection.

The following definition of screening concentrates on endeavouring to

answer the question of where the organisation can sell its products:

“Screening is an initial step in a selection process. In the context of

international marketing, it is a preliminary stage of the in-depth global
assessment of opportunities. The objective of screening is to identify
potential markets quickly and inexpensively without regard to method of
entry.” (Russow & Solocha 1993, p. 67)

In contrast to screening, market selection is:

“… the decision-making activities which are employed in the selection

of one or more suitable foreign markets, from at least two potential
ones. The salient elements of the decision are the criteria on which the
decision is based, the sources from which the information is gathered,
and the methods of analysis used.” (Sheridan 1988, p. 15)

Sheridan claims to be the first to produce an operational definition of

the overall international market selection process, and makes clear that he
 21

believes that these words encompass the two tasks of screening markets and
the in-depth analysis of markets.

2.4 The Role of Screening

Screening needs to efficiently and effectively short-list opportunity

markets, resolving the tensions of, on the one hand, minimising the size of the
large data collection and assessment task, whilst on the other, maximising the
risk-reward trade-off in the Bayesian way5.

There are approximately 60,000 individual products and services

(Downey et al. 1988). These can be aggregated into like groups such as
consumer or industrial, goods or services, capital or consumption, then into
categories such as convenience, shopping, specialty and unsought consumer
goods. Screening must be able to handle this product diversity.

Many academics (for example, see Ball & McCulloch 1982;

Papadopoulos 1987; Root 1987; Russow & Solocha 1993) propose assessing
all markets, although some do not explicitly state a position (e.g. Ayal & Zif
1987; Connolly 1987). Not including all countries can lead to not discovering
the highest potential markets, so incurring an opportunity cost, yet the
workload implication of including all markets is significant (200 countries by
150 measures requires processing 30,000 datum). An optimistic scenario
would be for the research of Russow 1989 to be replicated, so affirming that
screening requires data input on ten criteria, plus growth in three of them
(production, imports and exports) in some 200 countries. This involves
collecting and processing some 3,200 statistics per product screened ([10 by
one by 200] plus [three by two by 200]) – still a sizeable task. A fundamental
assumption is that screening uses secondary data, not primary data, for cost
and speed.

5
See, for example, Kinnear et al. 1993, pp. 94-99 or McCloskey 1982, p 54 for discussion of
risk-reward methodology.
 22

The literature’s inference is that the most important variables should be

used to screen markets, then a larger number of less important variables
should be applied during in-depth assessment to discriminate between the
short-listed markets. This presumes that the differing criteria, which determine
demand for each of the 60,000 products, are known, that the same criteria
apply to both domestic and foreign markets, and that there are not differences
in weights of criteria between markets. None of the literature surveyed
suggested that this is the case, so that screening is usually carried out based
upon intuitively selected criteria.

Where there is a choice between considering a matter in either the

screening stage or the in-depth stage, where should that aspect be dealt with?
For example, if well-documented demand data by segment are available,
should it be considered during screening, in-depth assessment, or both? The
critical matter of demand probably should be considered during both screening
and in-depth, but how do we then decide where to treat less significant matters,
such as market risk levels and psychic distance?

Consider also how different markets get proposed if different criteria get
used during screening and in-depth assessment. To give a simple example:
say screening only uses country population; China will be the top market. Then
assume that in-depth assessment is determined using only Gross National
Product per capita; the USA will be the top market. However, if Gross National
Product per capita and population were both used during in-depth assessment,
some third market might be suggested. The point is that, if screening and in-
depth assessment do not use identical criteria, differing assessments may be
made. Market selections can be distorted, intentionally or unintentionally, by
the sequence in which the variables are brought into play. None of the
literature raise this clearly important issue. The troubling practical result is that,
since much of the literature recommends sequential staging of variables,
organisations are being directed to ineffectively distort their market selections.

Should screening propose for in-depth assessment markets which are

beyond the resources, marketing competence and psychic distance of the
 23

organisation? Surely markets which are clearly infeasible should not be

presented for in-depth assessment. Yet, that would involve injecting primary
data about the organisation into the screening process to match the
organisation’s specific attributes with aspects of the various markets to
determine which are feasible. Whilst the literature clearly specifies that
screening should use secondary data about the target market, it is silent on the
matter of whether primary information about the organisation can be collected
and used during screening. The elimination of markets which are beyond an
organisation’s competence is cost-effective. Market screening principally uses
cheap secondary data, whilst in-depth assessment uses much more expensive
primary data. The smaller the pool of markets to assess during in-depth
assessment, the cheaper the process6.

Screening must also be able to respond to the differing stages of

organisations’ international evolution. The International Product Life Cycle
(IPLC) suggests that organisations sometimes move from being an
international novitiate through (up to four fundamental) subsequent stages –
typically comprising: exporting; beginning foreign production; facing foreign
competition in their export markets; then facing import competition in their
original market (see, for example, Asheghian & Ebrahimi 1990; Ball &
McCulloch 1990; Vernon 1966; Wells 1968). These different modes of entry
require different organisational resources, including differing skills from
managers.

Countries also move through different stages of development, having

product and service needs that vary with their level of economic development.
One of the tasks of screening is to approximately determine the match
between these underlying levels of development of buyer and seller,

6
Each market assessed in-depth in a rapid but effective way by an experienced organisation
typically costs in excess of $A10,000. This allows for desk research time, a trip to the market
(requiring executive time, airfares, on-ground travel, accommodation, meals and
refreshments), communication costs, and a small amount of purchased data. It does not
include the sensible cost for input by an economical yet effective international market
research consultant, such as by a Trade Commission. Typical in-depth research costs are
likely to be higher than this minimum. Further details are available from the author.
 24

irrespective of the infinite degrees of difference along each selection

dimension.

An added level of sophistication expected of screening is that it can

handle an organisation which needs to concurrently use many modes of entry
– perhaps exporting and countertrading in some markets, having sales offices
in others, joint venturing in others, and having its own production sites in yet
others.

Screening thus needs to aspire to be: simple, cost effective, flexible

enough to handle the variety of organisations, products and markets, and not
be likely to forgo opportunities. These are daunting requirements of a single
method.

2.5 How Many Markets Should Screening Deliver?

The background to this question is that screening principally uses

relatively cheap secondary data, whilst in-depth assessment uses much more
expensive primary data. Screening needs to assess all markets then deliver a
high potential sub-set for in-depth research. On the other hand, every
unnecessary market evaluated during the in-depth stage is a wasteful
consumption of resources, a potential cause of delay whilst the data are
collected, and adds additional complexity to making the ultimate decision.

There are two answers to the original question, and they are
psychological and practical in orientation. Psychological research has found
that there is a limit of around seven on the size of an evoked set – in this case,
the number of markets – that can be held in most human minds (e.g. Schiffman
& Kanuk 1991, p. 175). An international market selection practitioner would
typically see this as the maximum number of markets to be considered by in-
depth assessment – additionally justified by the cost, timeliness and complexity
issues mentioned above. Most practitioners would consider screening to have
failed if it delivered, say, ten markets for in-depth assessment. However,
statistical logic suggests that the number of markets should not be limited
 25

arbitrarily, and that the data going into a theory-driven screening process
should determine the number of markets to be assessed at the in-depth stage.
Hence, techniques such as cluster analysis and switching regression are
employed in similar situations to this to non-subjectively determine cut-off
points on set sizes. These two competing concepts determine the answer to
this question.

2.6 Components of a Screening Methodology

The components of a screening methodology flow from the matters

previously raised in this chapter. Screening needs to be able to cope with the
following:

• differing markets (of which there are over 200);

• different products (of which there are some 60,000). This implies
the method needs to cope with different customer needs, buying
decision processes, adoption rates, segments and customer
targets;
• different organisation sizes and competence levels. This implies
the method needs to suit different levels of organisation resources,
varying organisation objectives, different competencies at
researching international markets, varying organisation strategies
and, therefore, differences in the derived marketing mixes and
product positions;
• different modes of entry (subject to when the mode is determined);
• minimising the errors of: (1) ignoring countries that offer good
prospects, and (2) spending too much time investigating countries
with poor prospects (Root 1994, p. 55). The consequent
implications are that: (1) all markets must be screened, and
(2) principally secondary data are used during screening;
• somehow trading off the various influences (e.g. demand against
competition) to be left with a few optimal markets.
 26

The abstract components of a screening methodology are as follows:

1) a decision about the number of countries to assess (all or some),

2) determining which selection criteria to use (demand, political
conditions, etc.),
3) a decisions about when to consider the various criteria (e.g.
consider mode of entry before, or after, screening?),
4) weighting of each criteria,
5) the country selection process (mathematical and psychological
selection methods such as multivariate techniques and evoked
set which trade off differing scores and weights for all selection
criteria).

The literature review indicates that few empirical answers to

components two, four and five (above) have so far been determined through
research.
 27

CHAPTER 3

LITERATURE REVIEW

Market entry decisions are among the most critical made

by a firm relative to international markets. The choice of
which country to enter commits a firm to operating on a
given terrain and lays the foundation for its future
international expansion. It signals the firm’s intent to key
competitors and determines the basis for future battles.
(Douglas & Craig 1992, p. 302)

3.1 Introduction

The previous chapter located screening in relation to both international

market selection and internationalisation by organisations. It bypassed
providing the formal literature review.

This next chapter therefore presents the relevant literature. It shows

whether organisations do presently screen markets, considers the theoretical
foundations for screening, summarises the numerous previous studies of
screening, and puts a position on whether to screen markets first or to select
the mode of distribution first. Those matters allow conclusions to be drawn
about gaps in the existing literature, and lead to the research questions
addressed by this work.

Additional literature is presented in chapter four to argue the relevance

of each of the 46 variables used in the market screening framework.
 28

3.2 Do Organisations Presently Screen Markets?

We have to approach the answer to this obliquely, but can be quite

certain about the position. Bilkey’s (1978, p. 33) description of the export
behaviours of organisations in several countries includes the stunning finding
that the initiator of exporting was unsolicited orders in some 67 per cent of
cases (with a range from 40 to 83 per cent). Ellis’ (2000, p. 454) study of
internationalised Hong Kong toy manufacturers found that an even lower level
of foreign market entries were seller-initiated – 11 per cent. If organisations
rarely proactively initiate new foreign business, they must rarely screen
markets. Whilst it is possible for organisations to screen markets following
receiving an unsolicited approach, anecdotal evidence clearly indicates that
this does not happen.

Why is this so? This chapter will use the literature to show that a
fundamental cause is that we do not presently fully know how to screen. This
is partly because we do not know which variables are important, and partly
because there are questions about which statistical method is appropriate.

3.3 Theoretical Foundations for Screening

The large literature review undertaken for this dissertation (350 items)
led to the conclusion that there is negligible theory applied specifically to
market screening. The cause is probably the present incomplete
understanding of the reasons for international trade (e.g. Asheghian &
Ebrahimi 1990; Ball & McCulloch 1990, p. 95). Whilst a number of areas of
thought would probably claim market screening to be part of the process they
examine, they do not specifically articulate the theoretical or practical
connections. Examples are international performance, international product
life cycle, learning, mode of entry, motivation, networks, strategic advantage,
stages of internationalisation and transaction cost.
 29

Numerous writers make assertions about components of the screening

process, and some of these use the literature to argue their position, but
almost none have conducted empirical research. This dissertation uses the
literature and further theorising to develop a screening framework, then
empirically tests it.

3.4 Previous Studies of Screening

The main body of screening literature appears over the 30 years

between 1960 and 1990.

Bartels (1963) wrote with sponsorship from the American Marketing

Association. He discussed numerous criteria he thought important when
comparing markets, these being general criteria such as ‘The Nation’, ‘The
Society’, ‘The Economy’, sometimes with further sub-divisions. He does not
include product-specific data, nor suggest a methodology to aggregate the
criteria, then compare one country against another. Conners (1960) reported
how he, as 3M’s then manager of international market research, assessed
country potential. The method comprised multiplying market size by market
quality. Population (as a percentage of U.S. population) was the indicator of
market size, whilst market quality was calculated from the average of six
indicators of economic activity (national income, steel consumption, kilowatt
hours produced, motor vehicles in operation, telephones in use and radios in
use).

The pioneering academic work by Liander et al. (1967) (again, strongly

supported by industry) suggested three techniques for assisting to identify
which international markets have potential. Their first methodology involved
assessing industrial development by grouping ‘similar’ countries together
following assessment of numerous economic, technological, marketing and
cultural indicators, and the application of factor and cluster analysis. The
second method suggested was a regional typology approach, whereby
numerous secondary data indicators were ranked, converted to indices, then
countries ranked. The third approach was a normalised two dimension direct
 30

score method whereby secondary indices were produced, normalised and

scored. Goodnow and Hansz (1972), Litvak and Banting (1968), Russow
(1989), Samli (1977) and Sethi (1971) are examples of subsequent authors
who elaborated and applied the Liander et al. (1967) conceptual and
methodological work.

The number of steps in international market selection had slowly been

increasing, from Alexandrides’ (1973) one stage model and Deschampsneufs’
(1967) three-stage evaluation of markets. Walvoord’s (1980) articulation of
the idea of ‘filters’ or screens seems to have been an important, although
perhaps coincidental, turning point. He was not the first with the idea – for
example Douglas, Le Maire and Wind (1972) suggested screening in their
model – but the notion of market assessment having a fast, preliminary step
to screen many markets and a later in-depth assessment of a few high
potential markets appear much more frequently in the literature following
Walvoord’s article.

The market selection methods of these and subsequent writers can be

categorised as either intuitive, or more structured, approaches.

3.4.1 Intuitive Approaches

Intuitive approaches include a component that allows managers to

apply their unreasoned beliefs about cause and effect, but still partly
systemise the market selection process. Douglas and Craig (1983) and
Cavusgil (1985) thus produced frameworks which allow managers to benefit
from an existing structure which these managers customise to their perceived
needs. Decision-makers are required to decide which criteria to use, and to
weight them. Solloway and Ashill (1995) are a published example of such an
application in a New Zealand agribusiness setting.

Sheridan (1988) claims to be the first to publish empirical research

specifically on the international market selection process. His exploratory
research findings relate to Canadian technology organisations of small and
 31

medium size, and describe how they select international markets. Sheridan
found that these organisations initially select international markets on the
basis of intuition, usually as a response to some type of external stimulus
(p. 104). His research suggests that, with experience, some organisations
become more systematic in their decision-making.

The main intuitive methods employed by the Canadian firms in

Sheridan’s research were: word of mouth unsolicited sales; choosing the
U.S.A. due to size, proximity or similarity; unsolicited sales resulting from
advertisements in magazines with international coverage; unsolicited sales
resulting from presence at trade shows; a generalised “we chase
opportunities”; the ‘old boy network’ leading to sales; servicing Canadian firms
abroad; and firms going where other technology firms were successful (p. 71).

Sheridan’s contribution was to empirically confirm that market

selections can be produced by an intuitive decision-making process. Its
implied challenge to research was to further refine methodology so as to
reduce the volume of intuition which organisations use and to increase the
volume of researched insight to improve the quality of the decision made.

3.4.2 More Structured Approaches

All market selection frameworks have a degree of structure. When

compared with the previous approaches, the following substantially reduce the
amount of managerial customisation required and increase the quantitative
complexity.

Market Grouping vs Market Estimation. One way of grouping the many

different quantitative approaches to market selection is to separate those
techniques which endeavour to estimate the product’s potential (Market
Estimation) from the techniques which group countries together on the basis
of similarity of language, culture, economic development, etc. (Market
Grouping). Figure 3.1 presents the helpful taxonomy developed by
Papadopoulos and Denis (1988).
 32

Figure 1 INVENTORY AND TAXONOMY OF STATISTICAL APPROACHES TO IMS

Statistical approaches

Market grouping Market estimation

Macro-segmentation Total demand potential Import demand potential

Multiple Econometric Multiple Econometric Shift-share

factor methods criteria methods analysis
Micro-segmentation indices

Macro-
criteria

Bartels, 1963 Micro-

criteria
Liander, et al.,
1967
Hodgson and Connors, Douglass, Moyer, 1968 UNCTAD/GATT, Alexandrides, Green and
Litvak and Uyterhoeven, 1962 1960 Lemaire, 1968 1973 Allaway
Banting, 1968 and Wind, 1972 Armstrong, 1985
Wind and Douglas, Dickensheets, 1970 CFCE, 1979 Alexandrides
Sethi, 1971 1972 1963 Douglas, Craig and Moschis,
and Keegan, 1982 Singh and 1977
Sethi and Douglas and Craig Liander et al., Kumar, 1971
Curry, 1973 1982 1967 Douglas and Craig,
1983 Ferguson,
Sheth and Papadopoulos, Beckerman, 1979
Lutz, 1973 1983 1966
Lindberg
Ramond, Moyer, 1968 1982
1974 Source: Papadopoulos, Nicolas, and Denis, Jean-Emile
Samli, 1977 (Autumn 1988) "Inventory, Taxonomy and Assessment of
Doyle and
Methods for International Market Selection" International
Gidengil,
1977 Marketing Review p 40
 33

Market grouping methods can then be divided into the sub-groups of

micro-segmentation and macro-segmentation. Market estimating methods
can also be separated into two sub-groups: those that attempt only to assess
import demand, and those which attempt to assess demand via any mode of
entry (total demand) – a matter which will be revisited later.

Macro-segmentation and macro-criteria use such general factors as

economic, cultural, political, legal and technology which influence the milieu
within the target market, and which in turn affect product potential. Micro-
segmentation and micro-criteria use product-specific factors such as
demographic, psychographic and behavioural matters which drive product-
specific demand. Micro- and macro- segmentation will also reappear later
herein.

Although shift-share analysis was categorised as a method of

assessing import demand, later work by Russow (1989) and Kumar, Stam and
Joachimsthaler (1994) has used it to assess total demand, so not pre-
specifying the mode of entry. The techniques suggested by many of the
authors in the inventory by Papadopoulos and Denis (1988) are appropriate to
both screening markets and in-depth market assessment, although the
article’s focus is on screening.

A completely different way of breaking the structured market selection

approaches into sub-categories is to use the concepts of evoked set, export
assessment, and basic need assessment (Russow & Solocha 1993).

Evoked set techniques assess a preselected, limited number of

markets that has entered the organisation’s perceptual set of feasible,
potential markets. A market’s insertion into the assessors’ set is as a result of
ethnic origins, cultural similarity, geographic proximity, mentions in the daily or
trade news, or other reasons. The evoked set short-listing of potential
markets is often a result of intuitive and sub-conscious processes, and does
not follow a systematic assessment of product potential in a large number of
markets. However, once a market has entered the evoked set, a systematic
 34

assessment of it using conscious thought processes does occur. This

detailed assessment then affirms whether (or not) the market is likely to have
potential. Papadopoulos (1983, 1987), Reid (1981) and Russow and Solocha
(1993) have written about the concept, whilst Kramer (1964), Papadopoulos
(1983), Robinson (1978) and Samli (1977) have suggested evoked set
assessment processes.

Organisations generally use their evoked set to screen markets, just as

most people select their next automobile in the same way. It is irrational,
incomplete, and almost certainly sub-optimal, but is one way to quickly reduce
the large number of markets to a manageable set. In-depth research is then
relied upon to ensure that a degree of market profitability is likely.

Methods which Russow and Solocha (1993) categorise as assessing

export potential are also called ‘import demand potential’ methods by
Papadopoulos and Denis (1988). An export demand potential assessment
method does not consider alternative modes of entry (e.g. joint venture,
investment, etc.), and aims to assess market potential on the basis that the
goods will be exported.

Important contributions to export assessment have been made as

follows: Deschampsneufs (1967) suggested a three-stage in-depth evaluation
of an evoked set of markets; UNCTAD/GATT’s (1968) method of analysing
import and export (but not consumption) statistics; Alexandrides (1973) and
Alexandrides and Moschis (1977) proposed a multiple regression in-depth
approach; Walvoord (1980) proposed a four stage evaluation that articulated
and included the concept of filtering (i.e., screening); Cuyvers et al. (1995)
operationalised Walvoord’s process to suit government decision-making;
Douglas and Craig (1983) proposed customised screening and in-depth
export market assessment; Cavusgil (1985) interviewed 70 exporting
executives, proposed an export market selection process, then went on to
develop the CORE software which assesses COmpany Readiness to Export;
Jain (1990) proposed a three-stage process which includes screening; Green
 35

and Allaway (1985) took the shift-share mathematical methodology of Huff

and Sherr (1967) and applied it to export market assessment.

Whilst some organisations may wish to limit their risk or capital by

using exporting, they unwisely limit their profit potential by presupposing
export mode (Ball & McCulloch 1982, 1990; Connolly 1987; Papadopoulos
1987; Root 1987; Russow & Solocha 1993; Russow & Okoroafo 1996).

The basic need assessment cluster of methods of screening and in-

depth market assessment presently appear to hold the most promise for
effective screening of markets. The term originated with Ball and McCulloch
(1982) as “basic need potential”, and has since been developed substantially
by writers such as Root (1987), Russow (1989), Russow and Okoroafo (1996)
and Russow and Solocha (1993), and to some extent by Connolly (1987),
Cundiff and Hilger (1984), Douglas and Craig (1995) and Papadopoulos
(1987). Papadopoulos and Denis’ (1988) taxonomy would be likely to place
basic needs assessment into their ‘Market Estimation of Total Demand
Potential’ category.

Ball and McCulloch (1982) saw market selection as commencing with

an assessment of the factors which cause demand – their “basic needs
potential”. Physical forces (climate, topography and natural resources) are
included. They cite the example of air conditioners, where a market analyst
would exclude countries with cold climates because of apparent lack of
significant demand. They presented a conceptual approach which they did
not elaborate upon nor operationalise. Their six-step model for market
screening and in-depth assessment is at Figure 3.2.
 36

Figure 3.2
Selection Of Foreign Markets:
Six Steps

Initial screening:
Basic Need Potential and/or
Foreign trade and investment

Second screening:
Economic and financial forces

Third screening:
Political and legal forces

Fourth screening:
Sociocultural forces

Fifth screening:
Competitive forces

Final selection:
Personal visits plus, in some cases,
research in the local market

Source: Ball & McCulloch (1982) p. 311

Root (1987) suggested a three-stage model for selecting a target

country, comprising: (1) preliminary screening, (2) estimating industry market
potentials, and (3) estimating company sales potentials. Each of these has
between two and four sub-components. He states that screening should
assess all countries, endeavouring to identify country markets whose sales
potential warrants further investigation. Root also offers the mathematical
concepts for quantification of estimated market potential and says that, ideally,
 37

screening should comprise determining the product-specific consumption in

terms of production, plus imports minus exports, in all countries. Root’s useful
screening suggestions cover demand by market but do not consider: (1)
whether the organisation is capable of filling that demand, nor (2) whether
there are strategic or personal preferences that favour some markets.

Russow (1989) became the first to operationalise and empirically test

the Root concepts. He took 1970 and 1980 data about six randomly selected
SITC products (Standard International Trade Classification, Revision 2, United
Nations 1981) in 173 countries. To estimate the product growth variable,
Russow applied the shift-share mathematics7 of Huff and Sherr (1967) which
Yandle (1978) had used to identify brand performance and Green and Allaway
(1985) used in their domestic market demand screening. Russow found that
the four most important screening factors were those that he named using the
Delphi technique:

• indirect market size (comprising GDP, population, manufacturing as

a percentage of GDP, money supply, domestic production of the
product, and international reserves) (16.4 per cent),
• level of economic development (comprising GDP per capita,
agriculture as a percentage of GDP, and average hourly wages in
manufacturing) (13.9 per cent),
• product-specific market size growth (based on production, imports
and exports, and using shift-share methodology to calculate this
factor) (10.1 per cent), and
• product-specific trade (comprising imports and exports) (9.5 per
cent).

Figures in parenthesis indicate the amount of variance explained by the

criteria during principal components analysis.

7
Shift-share computes expected growth estimates based on the average growths of all
markets under consideration. Each market’s growth is then compared with its actual growth,
the difference being net shift.
 38

Confirmation that these four factors (which require input of data about
ten sub-factors) are of dominant importance would be a most helpful step
forward. It would mean reducing the present number of 150 factors and so
considerably reduce the size of the screening task. As the size of the data
collection and processing task reduces, more organisations can be expected
to attempt systematic screening.

Root’s (1987) arguments in favour of assessing product-specific

demand is implicitly an argument for basic needs assessment, his argument
for screening for both exports and investment modes is an implicit argument
against exports assessment, whilst his argument for screening all countries is
an implicit argument against evoked set screening. Russow and Okoroafo
(1996) convert these implicit arguments into explicit arguments in favour of
screening by basic needs assessment.

Russow (1989) was the first to empirically test and operationalise the
technique, coining the title “basic needs assessment”. Russow and Solocha
(1993) and Russow and Okoroafo (1996) have also strongly argued the logic
of acting rationally and selecting the mode of entry after screening countries
for demand.

The only known surveys of exporters for their views about relevant
variables and their significance are Wood and Robertson (2000), a series of
articles by Brewer starting in 2000 and Rahman (2000).

Wood and Robertson (2000) found (from a sample of only 137, all
USA, SME respondents, and with results qualified by other cautionary
sample-biasing matters) that market potential, legal and political were the
three most important variables. However, it must be noted that Wood and
Robertson’s definitions of variables are different from many interpretations.
Two of many anomalies are: (1) “market potential” includes competition, and
(2) “legal” includes barriers to entry, and excludes likelihood of receiving a fair
trial). Brewer’s doctoral work studied six Australian cases to describe and
analyse the way they selected international markets, and found little evidence
 39

of quantitative methods being applied to the selections. The 195 respondent

sample in Rahman (2000) was analysed by factor analysis and structural
equation modelling to determine the variables used in market selection.
Strongly important variables were found to be GNP, currency reserves,
stability of exchange rate, rate of inflation, population size, and product
significant demographics.

The reader of these three studies is left with the consistent impressions
that: (1) neither the researchers nor their subject organisations clearly
separate between the two stages of screening and in-depth at either the
conceptual or operational levels, and (2) screening is treated superficially. For
example, screening amounts to eliminating some countries for “practical and
corporate issues” (Brewer 2000, p. 20). The three authors describe what
exporters actually do when selecting markets; yet, that may or may not relate
to what they should do – since (1) we do not yet know what the optimum
market selection behaviour is, and (2) they get no feedback on the
effectiveness (or otherwise) of their market selections. Whilst the screening
methods described are incomplete and sub-optimal, the market selection
criteria used by the organisations at the in-depth stage are generally
systematic ones. In summary, the three works reinforce a screening
advocate’s view that there is considerable potential for obtaining productivity
gains8 from improving the screening technology.

3.5 Screen First, or Select Mode First?

The matter of when during the screening process to decide the mode of
entry is important. Deciding the mode before screening appears illogical
because it may cause foregoing a more profitable entry mode. Deciding the
mode also influences some of the screening factors to be used. For example,
presupposing export mode would exclude considering host government
incentives to investors but include searching for low tariff protection markets.

8
Productivity gains arising from selecting markets which have higher sales potential, lower
market risk, and which are appropriate to the organisation’s abilities, values, objectives,
motivation and commitment.
 40

These mode-specific variables may then bias the selection away from the
highest potential market. The mode of entry also affects production location
and production capacity – a decision to export is a vote for keeping the current
configuration of production plant.

Proponents of selecting the mode of entry after screening include Ball

and McCulloch (1982), Connolly (1987), Papadopoulos (1987), Root (1987),
Russow and Solocha (1993) and Russow and Okoroafo (1996). However, the
impelling logic of the mode neutral approach may require more research to
finalise and that research is beyond the scope of this dissertation. The
proposed screening framework uses variables which are mode neutral.

3.6 Conclusions about Gaps in the Literature

There are major gaps in our understandings about screening:

• The full set of criteria which are influential in screening decisions

have not been determined.

Russow (1989) began the work of empirically exploring which

criteria are significant. He took 21 of Root’s (1982, 1987) 178
suggested criteria, to three of which he applied Green and Allaway’s
(1985) relative gain calculation. Russow then calculated the relative
significance of each of those 21 criteria to the screening decision.
However, additional criteria can be argued to be relevant to the
screening decision, but these have not been explored in detail
either theoretically or empirically. The literature subsequent to 1989
has suggested further dimensions, and we remain unsure about a
number of matters (e.g. desirable age and education levels of
management in high performing international organisations).

If a screening decision is made after not taking into account all

important influences, we make an incomplete and potentially faulty
 41

decision. Some authors bypass this difficulty by leaving selecting

screening dimensions completely up to the organisation, for
example “... the onus is on management to identify the set of
relevant criteria” (Douglas & Craig 1983, p. 112). This assumes that
organisations have the motivation and competence to determine the
influences on demand in their markets, and to determine the
probably different levels of importance of different variables in
different markets. Few have. Screening decisions are thus made
on the basis of only partial information about causation and
moderation;

• Some of the suggest dimensions have not been fully

operationalised. For example, what are the variables involved in
the “cultural forces” which Cundiff and Hilger (1984) suggest be
used when selecting markets? If we accept the Hofstede (1980,
1984, 1991, 1994, 2001) five dimension answer to this question for
the 60 markets he researched, what do we do for data for the other
150 markets which Hofstede did not research?

• The relative importance of each criterion to each other has not been
determined. We have yet to learn the relative importance of say,
demand to country difficulty;

• A related matter is interaction effect between moderating variables

– where, for example, a low level of industry competition may offset
a low level of international experience to allow an inexperienced
organisation to successfully enter an otherwise difficult market;

• Almost none of the authors offer any underlying theory behind their
suggested screening criteria. Criteria selection is usually intuitive
with no theoretical or statistical rationale for their use. Theoretical
principles can be imputed for many of the dimensions, but the
authors rarely argue the relationship;
 42

• There are no accepted decision rules for accepting or rejecting

countries. For example, how do you trade off high potential
demand against high political risk and high competition? Root
(1982) proposed a set of rules, but there has been negligible debate
about the validity of these, nor alternative rules, nor the advantages
and disadvantages of whatever rules are used;

• How should the screening process make allowance for differences

among organisations – small versus large sized ones having
different capabilities and preferences to enter markets having
differing cultural, competitive and other settings.

In summary, we have yet to determine empirically or theoretically the

criteria, their weights, their interaction effects and the best decision rules to
screen markets so that individual organisations having varying attributes are
matched with a probably suitable market for their specific product.

3.7 Specific Research Questions

This dissertation aims to add to our understanding in several ways.

Using theory and empiricism, it aspires to:

• provide a complete conceptual framework;

• operationalise as many variables as possible;
• determine criteria weights, and so assess which criteria are
influential when screening international markets, and
• assess differences in markets selected between organisations of
different abilities, values, motivation and commitment, and different
managerial preferences.
 43

CHAPTER 4

SCREENING: A CONCEPTUAL FRAMEWORK

In the beginning there was information.

The word came later.
(Dretske 1981, p. vii)

4.1 Introduction

The previous chapter presented the literature on screening. Since it

argued that existing screening methods are flawed, an alternative paradigm
needs to be presented to solve the problem.

Chapter four therefore presents the conceptual framework which, it is

suggested, an organisation should use when making a foreign market
screening decision. The chapter has two sections: firstly, a description of
each of the 46 decision variables and the key impacts that each is expected to
make on the choice of a group of highly attractive markets; secondly, the
underlying reasoning which caused the selection of each variable is set out.
The chapter’s first section provides a non-technical explanation of the
concepts, leaving to the second section the more complex terminology
associated with explication of theory and citations from the literature.

4.2 Description of Proposed Framework and Variables

Let us commence by suspending for a moment the practical constraints

about data availability for screening markets. What information should firms
ideally have about each possible new foreign market to be able to make the
 44

optimal selection amongst those markets? The guiding principle should surely
be to obtain data that allow a particular organisation to hierarchy all markets
for its specific product by the estimated levels of risk-adjusted reward and
effort. Figure 4.1 provides the overall concepts involved in the decision based
on this principle. It categorises each of the variables into dependent,
independent, moderating or intervening9. Note the feedback loops.

9
“The dependent variable is the variable of primary interest … The researcher’s goal is to
predict or explain the variability in the dependent variable” (Sekaran 1992 p. 65).
“An independent variable is one that influences the dependent variable in either a positive or
a negative way. ... with each unit of increase in the independent variable, there is an increase
or decrease in the dependent variable also” (op. cit., p. 66).
“(A) moderating variable is one that has a strong contingent affect on the independent
variable-dependent variable relationship” (op. cit., p 67).
“An intervening variable is one that surfaces between the time the independent variables
operate to influence the dependent variable and their impact on the dependent variable.
There is thus a temporal quality or time dimension to the intervening variable. The intervening
variable surfaces as a function of the independent variable(s) operating in any situation, and
helps to conceptualise and explain the influence of the independent variable(s) on the
dependent variable.” (op. cit., p 70).
 45
Figure 4.1
Summary of a Contingent Eclectic Approach to
Screening International Markets

Causal Theories
(economic, marketing, management & behavioural)

Independent variables
.probable causes .controllable
Moderating variables
. uncontrollable Organisation’s product-specific,
expected-achievable,
Each market’s attributes: objectives in each market
- Potential customer aspects {Targets} ↔ {Strategy decisions}
- Competition & substitutes (Size of assessment task)
- Other environmental matters (Criteria & method of evaluation)
- Market risk levels
(includes new & existing markets) ↓
---------------------------------------------------------------------------------------------------------

Organisation’s capabilities & Intervening variables

preferences: . behavioural responses
- Organisation’s abilities Management motivation &
(generic & market-specific)
commitment to further
- Managerial values internationalise
---------------------------------------------------------------------------------------------------------
{Marketing mix}
Data availability & quality Customer adoptions & usage
regarding both the markets & the organisation
Organisation’s product-specific
demand
(Selection process)

↓
Dependent variable
.results
Organisation’s short-list of
attractive & preferred of markets
 46

Figure 4.1 showed how an organisation’s level of product-specific sales

in each market is partly driven by the organisation’s objectives. For example,
are low volume sales at high profit margins desired, or are high volume sales
at lower margins sought? What level of risk-taking to acquire those sales is
acceptable to management? These and other objectives in turn determine
strategy. Strategy and target customers get developed interactively. They
lead to development of an appropriate marketing mix, which in turn needs
thought about adaptation to the particular market chosen.

The organisation presently has 330 markets to consider, each of which

varies in terms of customers, competitors, and other environmental aspects
(such as physical, economic, political, cultural, etc.). Each market also has a
particular level of risk whereby economic downturn may occur, the
government of the day may be overturned, etc.. These characteristics of each
market are necessary, but incomplete, determinants of which market is most
desirable to the organisation.

The organisation which is undertaking the market screening brings

varying abilities to the situation. Examples include its competencies at
domestically marketing and managing its business, the human and financial
resources it has available to apply to entering another foreign market, and the
level of experience it has at internationalisation. Screening and in-depth
market assessments are tasks which require skills at collecting, processing
and interpreting data. These matters affect which markets the organisation
will be competent at assessing and later entering.

Managerial values about internationalisation will further help or hinder

the market selection and entry process. Some managements derive pleasure
from further internationalising their business whilst others find the work
burdensome. Each management has its own risk preferences that mandate
which markets will be viewed as desirable or undesirable. Managements
have strategic preferences, for example to enter market X before competitor A
does so, to go to markets requiring little product alteration, etc.. These
matters affect which markets will be preferred for entry.
 47

When considering all these factors, the organisation has choices to

make about the criteria it uses and the methods of evaluation of the markets.
For example, do high scores on a market’s sales volume get adjusted
downward by high scores on competition, or alternatively, are some factors
go/no-go matters which completely eliminate that market?

All the aforementioned variables interact to determine the likely

strategy, the segment-target-position chain, the consequent customer
adoption levels and thus the likely sales levels and other outcomes, and the
likely level of management motivation and commitment. These are calculated
by the selection process used – the mechanical, mathematical methods used
to weigh and trade off the many variables. The final outcome is a hierarchy of
markets based on their likely sales volumes and the other matters which
determine their attractiveness to the organisation.

Management’s motivation and commitment to further internationalise

deserves elaboration. Management’s assessment of the probability that the
expenditure of effort will lead to the particular rewards they desire will
determine their motivation and commitment. If motivation is weak,
management will be unlikely to have the enthusiasm to pursue difficult
markets. If not committed, management will not have the stamina to
persevere over the necessary period of time before becoming profitable in the
new market. Motivation and commitment will be a function of market
attractiveness (market attributes perceived positively), organisation capability
and organisation preferences in conjunction with the organisation’s objectives.

An example of how the above variables interact is the case where a

highly attractive market, being also a market which is easily within the
organisation’s capabilities to succeed in and which meets its preferences, and
which seems likely to meet the organisation’s objectives, will be a market that
requires a low level of motivation and commitment.
 48

The next diagram, Figure 4.2, portrays expanded details of the

concepts in Figure 4.1, again categorising each of the variables as dependent,
independent, moderating or intervening. Forty two causal variables are
predicted to determine the dependent variable.

An alternative layout of the same variables in a possibly more

meaningful assemblage is provided in Figure 4.3.
 Figure 4.2
A Contingent Eclectic Approach to
Screening International Markets
Causal Theories
(economic, marketing, management & behavioural)
especially about product-specific demand, country
attractiveness & managerial behaviour

Moderating variables
. uncontrollable Independent variables
.probable causes .controllable
Each market’s attributes:
- Potential customer aspects: Organisation’s perceived effort →
reward probability & value (product-
(needs) & segment population size specific, expected achievable,
PS
→ {segments }, {intrinsic objectives in each new market):
adoption - highest probable sales &
rates PS}, segment PS (or industry) sales growth (thus profits),
demand volumes & growth rates. - highest personal benefits
- Competition & substitutes - highest other benefits
- Other environmental matters: - chosen risk level and
physical PS , - lowest effort applied.
economic, {Targets} ↔ {Strategy decisions}
technological PS, (Size of assessment task)
legal, (Criteria & method of evaluation)
political, cultural,
ecological/environmental PS. ↓
- Market risk levels
(includes new & existing markets) Intervening variables
--------------------------------------------------------------------------------------------------------

Organisation’s capabilities & .behavioural responses

preferences: Management motivation &
- Organisation’s abilities: commitment to further
SCA (generic & market-specific), internationalise
general business competencies, {Marketing mix}
further internationalisation {Customer adoptions & usage}
competencies, (generic & market-specific) {Organisation’s product-specific
including international market research,
language & network contacts, demand}
key stakeholder support, (Selection process)
other resources.
- Managerial values: ↓
further internationalisation,
{strategic}, & Dependent variable
personal (including cultural .results
difference, risk & market research) Organisation’s short-list of
--------------------------------------------------------------------------------------------------------

Data availability & quality attractive & preferred markets

regarding both organisation & markets
 51
Figure 4.3 An Alternative Perspective of
A Contingent Eclectic Approach To Screening International Markets

Each market’s attributes: Organisation’s capabilities &

(includes new & existing markets)
Organisation’s: preferences:
Perceived effort → reward probability &
- Potential customer aspects: - Organisation’s abilities:
value (= Product-specific, consistent,
(needs) & segment population objectives in each new market): SCA (generic & market-specific),
sizes → {segments PS }, sales & sales growth (thus profits), general business competencies,
intrinsic adoption rates PS, {personal benefits}, other business further internationalisation
segment PS (or industry) demand benefits, risk level & lowest effort competencies (generic & specific),
volumes and growth rates. (i.e. rewards, risks & effort) including international market research,
language & network contacts,
- Competition & substitutes {Targets} ↔ {Strategy decisions}
key stakeholder support
- Other environmental matters: (Size of assessment task) (generic & specific),
physical ,PS (Criteria & method of evaluation) other resources.
(Independent variables = controllable)
economic, - Managerial values:
↓
technological PS, further internationalisation,
legal, Management motivation & commitment {strategic}, &
political, to further internationalise personal (including
cultural, {Marketing mix} cultural difference, risk,
Customer adoptions & usage rate market research & {other} values).
ecological/environmental PS,
- Market risk levels
Organisation’s product-specific demand - Data availability & quality
(Selection process) (Moderating variables = uncontrollable)
- Data availability & quality (Intervening variables = behavioural responses)
(Moderating variables = uncontrollable) PS
= Product-specific variable
↓
XXX = a heading
Market clusters {XXX} = descriptive (non-quantitative) variable
(Dependent variable) XXX = quantified variable
XXX = quantifiable variable
Organisation-specific short-list of (XXX) = other irregularity
attractive & preferred markets (As in Figures 4.1 & 4.2, feedback between
↓ variables occurs, but can not satisfactorily
be fully shown in this configuration)
 52

Multiple interactions between variables occurs. For example, culture

affects many of the other moderating variables, then flows through each to
affect independent and intervening variables, and so the screened list of
markets. Another example of multiple interaction among variables is where
low market attractiveness is offset by high organisation capability, then those
two areas flow through to affect objective setting and ultimately market
selection.

Feedback loops are shown in Figures 4.1, 4.2 and 4.3. If the
organisation iterates through a full screening decision cycle, it learns the
impact of the moderating and independent variables upon the intervening
variables, then the dependent variable. Feedback has occurred. That
knowledge will affect motivation to re-screen and conduct in-depth research,
and could stimulate a new round of screening assessment with altered
objectives.

Descriptions of the detailed concepts indicated in Figures 4.2 and 4.3

follow, along with the key interrelationships amongst each. Reading the rest
of this chapter will be made easier by continual reference to one of those two
figures.

For the sake of clarity, the arguments from the literature which support
the framework and its components, and most of the literature’s citations, are
given in a separate, following section titled “Underlying Theory”.

4.2.1 Dependent Variable

The dependent variable in a screening decision is the organisation-

specific short-list of attractive and preferred markets10. This short-list is later

10
Underlined and italicised words are the aggregate variables and sub variables in Figures
4.1, 4.2 and 4.3 (e.g. Potential customer aspects and needs).
 53

subjected to an in-depth assessment process when the organisation selects

the best one, or few, markets11.

4.2.2 Independent Variables

4.2.2.1 Organisation’s Product-specific, Expected-achievable,

Objectives

A major determinant of the screened short-list of markets is the

organisation’s product-specific, expected-achievable, objectives from entering
an additional foreign market – different objectives cause selection of different
markets. Objectives are a combination of the following:

• highest probable sales and highest probable sales growth,

• highest probable personal benefits,
• highest probable other business benefits,
• chosen probable risk level, and
• lowest probable effort.

Objectives summarise the expected product-specific sales to, and

attractiveness of, each market – what the organisation wants, constrained by:
(1) the external limitation of what the market will permit and (2) the internal
limitations of its ability to harvest the potential rewards. Managers express
how important each sub-component of the objectives is to them when they
complete a questionnaire (Appendix A). Different managers place different
priorities (weights) on each sub-component, reflecting differing individual
preferences and situations of their organisations. Research suggests that
managers usually think of the value for each of the five sub-components of the
objectives as a range, not a hard and fast number. Perhaps this is
unsurprising because of the uncertainty about the probability of achieving
each and because one sub-component affects another. For example, a
probably high risk, high effort, high profit market may be judged by a manager

11
The literature and experienced practitioners advise that very few firms have the resources,
 54

to be equally attractive to another market which is probably low risk, low effort
and low profit.

4.2.2.1.1 Sales and Growth in Sales.

Sales and Growth in sales are very common indicators of market

attractiveness in both domestic and international settings12. They also roughly
proxy profits, although the link between sales and profits is not strict because
sales are a necessary, but not sufficient, requirement for an organisation to
make a profit. Maximising segment sales and sales growth, constrained by
achieving individually varying modest levels of the other objectives, is likely to
be the norm for most organisations.

4.2.2.1.2 Personal Benefits

Personal benefits are the rewards (psychological and material) set as

objectives to flow to individuals (not the organisation) as a result of the
organisation entering the additional market. The management literature has
for several decades recognised that personal goals have an important
influence on work goal setting (see, for example, Koontz & O’Donnell 1968).
Work-related consequences from the new market entry on the key individuals
are that it will be likely to affect their competences, responsibility, pay and
promotion. There will also be the personal emotional pain and pleasure from
travel to foreign markets and from striving to succeed there, which in turn are
likely to produce changed levels of self-direction, stimulation, power, security,
benevolence, and so on. These psychological drives can subtly influence the
market selection. An example is a hedonistic manager who enjoys travel will
(all other things being equal) prefer to select a market requiring travel to it
ahead of a market to which travel is not required. These personal objectives

motivation and competence to enter numerous markets simultaneously.

12
Most of the firms in the Shoham & Kropp (1998:120) research into the causes of successful
international performance did not use profits as a measure of performance. Those authors
conjectured that the cause is that costs are built into the firm’s prices such that achieving
sales budgets would result in acceptable levels of profit.
 55

are likely to be covert, and perhaps partly unknown, to those involved –

particularly if they are inexperienced at foreign market entry.

Different foreign markets will yield differing kinds and levels of personal
rewards. Economic utility theory and psychological needs theory argue that
different individuals may perceive the same benefit to have a different level of
significance to them. However, whilst recognising that these personal goals
are likely to influence which markets are preferred, the screening framework
does not attempt to take account of them. Data are regrettably not presently
available for each country to allow this variable to be assessed via the
questionnaire to management, weighted as one of the other four sub-
components of the objectives, then used to influence the market selection.
That will eventually change, so allowing this matter to be brought to account.

4.2.2.1.3 Other Business-related Benefits

Other business-related benefits include seeking to: increase market

share; increase corporate image; increase organisational and political clout;
dispose of excess capacity; dispose of products no longer attractive in other
markets; derive learning (production, marketing, international business, etc.);
reduce counter cyclicality; and obtain resources or technology. Different
markets will have differing capabilities to meet these objectives. Screening for
these benefits requires specific secondary data and a weighting of the
variable in the organisation’s objectives. Both are presently feasible. For
example, each country could be rated on a zero to one scale on its attribute of
counter cyclical demand for specific fruit, clothes, or whatever the product;
management’s weighting for this objective might be, say, 10 per cent of the
100 per cent weighting for all objectives. Since this variable’s operation is
quite straightforward, data has not been synthesised to demonstrate
implementing the variable in the research presented herein.

4.2.2.1.4 Risk Level

 56

Risk level chosen is the amount of risk the manager wishes to accept in
the foreign market that is selected by screening. It is a global, intuitive
assessment which takes into account the possibilities of both positive and
negative outcomes to both organisation and person.

Risk can be estimated as the likelihood of ‘good’ and ‘bad’ events, such
as achieving sales, currency fluctuations, profit variations, loss of capital, loss
of reputation, competitive retaliation and so on. There are several dozen
common, specific risks which can be aggregated to be categorised as those
relating to economic, political, financial, competitive, commercial and personal
risks. Significance of failure also needs considering – a big organisation
which risks only a small amount of finance and reputation is in a different
situation from a sole trading individual who may be bankrupted and lose their
private dwelling through failure. A contrary aspect of risk is that, whilst the
notional rational economic person should take the lowest possible risk, in
contradiction, many people enjoy the thrill of taking risk (a psychic reward).
There is also the positive effect from diversification of markets if the
organisation is not already operating in numerous countries.

4.2.2.1.5 Effort

Effort must be expended to fulfil managements’ expectancies about

goal fulfilment. Effort here means the quantity, quality and duration of
conventional resources plus mental and emotional energy voluntarily applied
to achieve the task of market entry and reach break-even profitability. Each
market will require a different quantity and quality of personnel, finances,
mental energy and knowledge to be injected for different lengths of time to
reach target objectives and cease making a loss. In general, ‘easier’ markets
will require less effort whilst ‘difficult’ markets will take more effort to enter. In
general, the organisation will wish to minimise effort expended because it is
costly, but will be willing to expend more when there is a higher risk-adjusted
reward likely from doing so. The effort perceived necessary for entry into a
new market is a function of the interaction of the other objectives, resources,
 57

motivation, organisation capability, market difficulty and personalities of the

people involved.

The levels requested in each of these six sub-components of an

organisation’s objectives will vary from organisation to organisation in some
largely predictable patterns. Examples are:

• Organisations with low resources are unable to expend

considerable effort on new market entries, so will be wise to seek
profitable sales, to not seek many supplemental personal benefits
beyond security of existence, and not take major risks. On the
other hand, organisations with many resources have the freedom of
being able to accept losses on their market selection for an
individually determined but longer period of time in return for
personal benefits (e.g. hedonism from travel by the owner-manager
of a profitable firm), strategic benefits (e.g. market share growth),
and can take higher on levels of risk and/or apply more effort to the
market entry,

• Predictors of the objectives which are likely to be chosen are

management’s preferences toward further internationalisation,
strategy and their personal values (including risk preferences).
Irrational managerial behaviour arising from incompetence,
desperation, intense thrill seeking, etc. introduces an unpredictable
element which this dissertation does not concentrate upon but
which is allowed for through the managerial values component of
the proposed framework,

• The organisation’s stage of internationalisation should have an

influence. Novices at internationalisation need to experiment and
so develop knowledge. These organisations would be wise to set
moderate sales and personal benefits goals, limiting risk and effort
by seeking to enter low psychic distance, low resource-consuming
 58

markets whilst they accumulate experience at internationalisation.

In contrast, experienced internationalisers may set more taxing
objectives, such as high sales growth and high other benefits (e.g.
market rationalisation) because they have a higher level of
expertise to support the more demanding behaviour. The number
of markets already entered will be likely to also affect the level of
risk and effort of the next market entered. For example, an
experienced firm will, on the one hand, need to expend increased
effort because the lower risk and lower effort markets should have
been entered during the early stages of internationalisation. On the
other, additional knowledge should increase competence and lower
risk, so reducing effort,

• How large a part is played by personal benefits will especially be

determined by managerial personality, organisation resources and
managerial decision autonomy.

4.2.2.2 Targets

Customer targets, the second of the ‘segment, target and position’

trilogy, require selection by the organisation from among the segments – in
conjunction with the organisation’s strategy and objectives. This variable
represents the choice of the potential customer types which the organisation
wishes to attract, and is chosen from among the previously determined
descriptions of the segments13. It is a non-quantitative, descriptive variable.

The strategy variable is divided into two parts, strategy decisions (an
independent variable) and strategic values (a moderating variable).

4.2.2.3 Strategy Decisions

13
e.g. geographic, demographic, psychographic and behavioural if a consumer product, or
demographic, operating variables, purchasing approaches, situational factors and personal
characteristics if an industrial product.
 59

Making strategy decisions involves the organisation in selecting from

the basket of strategic options after considering such matters as its desired
objectives, possible segments and desired targets, the likely advantages and
disadvantages of foreign markets, its capabilities and preferences and its
motivation to succeed. The preferred international marketing strategy options
are grounded in the organisation’s sustainable competitive advantage,
whether it is a market leader, challenger, follower or nicher, and whether the
product being screened will produce synergy with other products or markets.
Strategy decisions influence development of the marketing mix, which in turn
influences the important matters of customer adoptions rate and customer
usage rate.

Strategy options do not get assigned numeric values nor get

statistically manipulated in the framework. However, strategy options affect
the goals set by the organisation – which are quantified here – so that strategy
directions are implicit in screening. The literature shows that strategic
planning directly impacts performance (e.g. Cavusgil & Zou 1994; Evangelista
1994; Ramanujam & Venkatraman 1987; Shoham & Kropp 1998), so wether
or not strategic planning is undertaken by the firm is therefore one of the
aspects included for assessment in this work. The question asked of the
organisation in Appendix A seeks to cover both the strategic orientation (the
content) and the planning sophistication (process) aspects by asking whether
the organisation annually undertakes in writing the sub-set of international
market planning.

4.2.2.4 Size of Assessment Task

The size of assessment task undertaken by the organisation is

determined by management – all research involves a decision by the research
decision-maker about how thorough a job they wish to undertake. The upper
limit on the size of the screening task is to collect and analyse some 30,000
pieces of datum but that figure may reduce if this research is successful (see
section 2.4). When management believe that that amount of research is not
worthwhile (for example, believe that research will produce ambiguous or
 60

unreliable results) or when it is not known how to screen markets, the

organisation will minimise or completely eliminate screening. Management’s
level of motivation to enter an additional market, and their value about using
market research, will both impact on the quantity and quality of research
undertaken. Assessment is further limited by the two moderating sub-
variables which measure how capable the organisation is at undertaking
research and the availability of information.

Although in practice managements currently reduce the number of

variables or markets assessed, it was argued in chapter two (e.g. section 2.4)
that that behaviour yields sub-optimal results. This variable is thus inserted
here for conceptual completeness but the proposed screening framework
does not provide the option of decreasing the size of the screening task.

4.2.2.5 Criteria and Method of Evaluation

The criteria and method of evaluation of the desirable markets are

generated through eclectic theory. This produces a list of 46 causal variables
which describe the particular organisation, the world’s markets, and the
assessment system. Some variables are switched off on a product-specific
basis whilst others assess the various matters on a zero to one ratio scale.
Shift-share analysis is used to estimate sales and growth in sales by market,
and Bayesian weights value all variables. Principal components analysis
compresses several economic sub-components into a single component.
Cluster analysis then trades the various factors off against each other to
suggest the short-list of markets. Low scores in some criteria may be
compensated for by high scores in others (but there is no elimination of
markets using ‘go/no-go’ variables). Multiple discriminant analysis tests the
cluster analysis result, and indicates the weight of each predictive variable.

The screening framework thus uses a hybrid of theories, through

market estimating and market grouping methods, to systematically select the
likely best markets for the particular organisation and product. Whilst not
recommended, altering the settings for the components of this variable could
 61

occur and would be expected to significantly change the markets that are
recommended for in-depth assessment. The screening framework therefore
does not provide the option.

4.2.3 Intervening Variables

Akin to a line of dominoes knocking each other down in turn, the

independent variables determine the intervening variables; the intervening
variables then determine the dependent variable. The moderating variables
affect the process.

The screening framework comprises five intervening variables:

management motivation and commitment to further internationalise, the
marketing mix, customer adoptions and usage rate, the organisation’s
product-specific demand, and the selection process. They typically vary over
time, reflecting, and affecting, the continuous choices made about further
internationalising the organisation.

4.2.3.1 Management motivation and commitment to further

internationalise

The item management motivation and commitment to further

internationalise details the level of management motivation to add an
additional market, and to persist with the decision whilst receiving a mixture of
positive and negative rewards. It is not the more generalised motivation to
internationalise – a precursor to adding a market. Thus, motivation to remain
at the present level of internationalisation is not included, but motivation to
enter an additional foreign market, and motivation to expand in an existing
foreign market, are both included.

The levels of motivation and commitment are determined by the

moderating variables in the framework and by the independent variables.
Management commences the screening task with a particular level of
enthusiasm (motivation) for entering an additional market, that level having
 62

been developed over time by the person’s attitudes, values and perceptions
about the organisation and its environment. That level changes as a result of
comparing the organisation’s abilities (SCA, etc.) and managerial values (e.g.
to further internationalise, etc.) with new market attributes (demand,
competition, etc.), then determining the perceived effort – reward probability
(setting product-specific, consistent objectives for the new market). Those
objectives then interact with the customer kind to be target and the
organisation’s strategy.

Motivation and commitment are attitude dimensions, and are normally

related to a specific market (e.g. country X). Prior to screening, the particular
market is not known, so the framework substitutes an abstracted
representation of the preferred market kind (e.g. rewards likely, level of
competition, economic milieu, etc.). The similar but subtly different
managerial values about further internationalisation is a more generic
psychological dimension which is used to guide the assessment of specific
situations (see section 4.3.4.2.2 for the definition of values).

Management’s level of motivation will affect the support given (or not
given) to the new market entry via the marketing mix. The horsepower
available to push the organisation into internationalising another increment is
thus determined by this level of motivation. High motivation and commitment
will at times compensate for low competence or high market difficulty to allow
the organisation to achieve a successful new market entry.

Whatever the initial motivation level, it is later modified by information

gained through screening. Feed back is shown in Figures 4.1 and 4.2 as
research develops knowledge about the many aspects of the possible
expansion (for example, whether foreign demand seems likely to be large,
whether that demand is located in low cultural distance markets, what the
levels of competition and country risk are, etc.).

Assessing the level of commitment (not motivation) to further

internationalise will be more accurately undertaken with more internationalised
 63

organisations than with novice internationalisers. This is due to commitment

being partly dependent on attitudes and knowledge, but less internationally
experienced organisations do not have the knowledge to truly know what they
are getting into. When they attempt market entry, we can predict that their
probably less accurate perceptions may be more at variance with reality than
is likely with their more experienced counterparts who have reached a higher
stage of internationalisation. Thus, less experienced organisations may
commence with the same level of motivation, but change their minds and
subsequently withdraw their commitment. Market screening would thus seem
relatively more valuable to less internationally experienced organisations if it is
able to eliminate markets in which their commitment will flag.

This independent variable, managerial motivation and commitment, is

assessed as management’s pre-screening level of interest in expansion into
an additional market. So, if screening produces very different results from
those expected by management, the levels of motivation and commitment
prior to screening will probably be markedly different from their levels after
screening. If the organisation is also an international novice, there will be
wider variance around the subsequent level of commitment.

4.2.3.2 Marketing Mix

The marketing mix is not directly manipulated in this research. It is a

necessary inclusion for the assumptions that need to be made about its
impact on the organisation-specific customer adoption and usage rate, and so
the organisation’s estimated product-specific demand, and so the objectives
set for the screening process to meet. For example, having the necessary
funds to implement the marketing mix, the willingness to make product
alterations, willingness to meet competitor prices, etc., all influence the
customer adoption and usage rates, and thus the organisation’s objectives.

4.2.3.3 Customer Adoptions and Usage Rate

 64

Guided by its strategy and the other influences, the organisation

attempts to position its marketing mix so as to produce a specific customer
adoptions and usage rate. It needs to influence targeted potential new
customers to begin consuming (a new product) or to switch suppliers (an
existing product) by moving those potential customers through the buying
decision process (the hierarchy of effects’ steps of awareness, knowledge,
liking, preference, intention-to-buy and purchase). The next variable, the
organisation’s product-specific demand, is an aggregation of demand from all
the potential buyers who have reached the last step of the adoption process
and so become customers.

Irrespective of the organisation offering a better product, potential

customers may not be willing or able to switch suppliers, and if they do, there
is commonly a time lag. Organisations’ and individuals’ supply sources are
often determined by more than the best price, delivery and quality – such as a
firm’s parent’s directives to buy from related companies, nationalism, network
contacts, corruption, incapacity to pay, and so on. A period of time also
usually elapses before buyers switch suppliers because existing supply
contracts may have several years to run, current item may not have worn out,
the new supplier may need to prove themselves, etc.. The arguments in this
paragraph mean that sales normally take several years to peak after entry to
the chosen market – so break-even cannot be expected immediately after
market entry.

Customer adoptions and rate of consumption are constrained or

enhanced by the numerous moderating variables in the framework. Notable
among those is sufficient product-specific segment demand (if there is no
sales potential, there is no need to consider any other matters) and the
organisation’s further internationalisation competencies (the organisation
needs to have the abilities to harvest the sales). Another key influence is
management’s motivation and commitment because that affects how hard and
long the organisation strives to achieve its desired objectives. Receiving
unsolicited enquiries or orders from a market may indicate customer
 65

displeasure with or under-supply by existing suppliers, making that market

additionally attractive for entry.

A fundamental gamble by the organisation is expending the minimum

effort through the marketing mix to push a sufficient volume of new customers
all the way through the hierarchy of effects. Spending too little will wastefully
leave substantial numbers of potential customers marooned part way through
the process. Spending too much on the marketing mix will wastefully over-
saturate potential customers, so draining profit. Adoption of most products
also requires on-going repurchases, not just a trial purchase, to occur, and in
sufficient numbers to achieve long term profitability (initial entry costs need to
be recouped as well as ongoing costs). The gamble about the size and
distribution of the marketing mix spend is made more hazardous when user
needs and segments must necessarily be presumed to be the same in all
markets as in the home, or other marker, markets. Sometimes that will be
right, at others, not.

4.2.3.4 Organisation’s Product-specific Demand

The organisation’s product-specific demand quantifies the

organisation’s estimated sales volume in each market. It can be produced by
taking the adjusted segment demand then applying a customer adoption rate
(if this were known or calculable) in each market. However, adoption rates of
the products researched herein are not known, so demand has not been
calculated.

4.2.3.5 Selection Process

The screening selection process is the final intervening variable to

consider. It is the moment at which all data in the influence paths is
instantaneously brought together by the predetermined method of evaluation
to judge which markets, if any, have the necessary characteristics to meet the
organisation’s objectives, and ideally to hierarchy markets. The mathematical
tools used were described in criteria and method of evaluation.
 66

4.2.4 Moderating Variables

These variables are beyond the ability of the organisation or its

individuals to effectively control in the short term. This includes inability to
substantially and rapidly alter factors such as:

• the external environment (the physical, economic, technological,

legal, political, cultural and ecological milieu in each market is
essentially beyond the organisation’s ability to manipulate in any
major way in the short term, particularly in a new market),
• the customers and competitors who operate in these aspects of the
environment,
• managerial attitudes (achieving attitude change in people is slow
and uncertain),
• resources (quality changes to the quantum of funds and trained
people normally requires a medium-term time frame),
• or any of the other moderating variables.

4.2.4.1 Each Market’s Attributes

Each market’s attributes are a function of the quantity and nature of its
product-specific demand (potential customer aspects) and the innate difficulty
of that market to be harvested by the organisation (competition and
substitutes, other environment matters and market risk levels).

4.2.4.1.1 Potential Customer Aspects

Potential customer aspects are comprised of the sub-components of

customer needs, segment population sizes, segments, intrinsic adoption
rates, and segment or industry demand.
 67

4.2.4.1.1.1 Customer Needs

Customer needs explores the causes of, and constraints on, demand,
by market. These are the base for later segmenting, targeting and positioning
via the marketing mix. There are thousands of differing human needs that are
summarised differently by different authors. Just three examples of
paradigms are clusters of needs related to: (1) existence, relatedness and
growth (Alderfer 1972, p. 142), (2) physiological, safety, love, esteem and self-
actualisation needs (Maslow 1965), and (3) the 45 items that Schwartz (1992)
aggregates into ten clusters. Customer needs are met by products, which
commonly vary by sizes, colours, units (imperial/metric), styles, quality,
accessories, service, one-stop-shopping, packaging, labelling language/s,
usage instructions, health warnings, warranties, and so on. There are then
the other three dimensions of the marketing mix, together with the
environment, which both respond to and shape customer needs, so they are
potentially different in each market. The end result is that exogenous and
endogenous stimuli to potential customers – from the physical, economic,
technological, legal, political, cultural and ecological aspects of the
environment – respond to, produce and shape individual needs and wants.

Primary research into customer needs in each market can not be

conducted during screening due to the cost of obtaining that information, and
the knowledge is rarely already available as secondary data, so customer
needs are often assumed to be similar to those in either the home or a marker
market. That assumption will probably be correct in some markets and
incorrect in others. This critical assumption can limit the efficacy of the entire
screening process by causing subsequent segmenting, targeting and
positioning to be wrong. There is no way around this dilemma until the causal
forces for each of the 60,000 products and services which exist14 has been
researched and become available as secondary data. Where there is
considerable doubt about which needs are met by the product, screening

14
SITC Revision 3 (United Nations Statistical Office 1986) defines 35,000 tangible products,
some of which are composites, and Downey et al. (1988) of the Australian Trade Commission
estimated there were a further 20,000 service products.
 68

should be repeated using different assumed needs and segment bases i.e.
selecting different target customers.

A world of perfect knowledge would allow a regression function to be

inserted into the framework – using the causes of customer needs and the
constraints on need fulfilment – to predict industry demand. For example,
cement sales in each market might be a function of domestic and commercial
building commencement statistics, lagged a month, constrained by any
rainfall. There ought to be a different equation for each of the 60,000
products and services, and the causal variables and weights could differ by
market. The proposed screening framework conceptualises this but it is rare
for an organisation to already have undertaken the research necessary to
provide the equation for its product at home, let alone abroad. In the absence
of the organisation supplying a regression equation, this variable is a
descriptive one which is not mathematically manipulated.

4.2.4.1.1.2 Segment Population Size

Segment population size in each market drives the maximum possible

size of sales, operations and profits. As in statistics, the unit for population
may be people, firms, cars, or whatever is the ultimate product user. Financial
capacity to pay (e.g. income level of individuals or organisations) is included
either here or in intrinsic adoption rate, depending upon how the segments are
defined. This population figure is reduced to the likely number of purchasers
when multiplied by the intrinsic adoption rate then adjusted for rate of
consumption.

4.2.4.1.1.3 Segments

Segments is a non-quantitative description of the feasible customers

which is used to obtain related data.

An individual’s needs can be unique so there might ideally be a

different product for each user in each country. That is usually not profit
 69

optimal for the organisation. It thus determines (in targets and strategy) the
degree to which it will meet individually differing needs by aggregating
potential customers into larger clusters having roughly similar needs to obtain
economies of scale and scope for itself. Segments may number only one (as
in mass marketing) through to several (segmental marketing) to many (as in
niche marketing).

A top down view of the operation of needs, segments and segment

population size sees statistics on industry demand in each market being
subdivided into customer segments, each segment having differing needs,
characteristics, demand causes, demand levels and propensities to switch
suppliers. Customer segments affect the strategy chosen by the organisation,
drive the customers targeted and the marketing mix developed to attract them,
and the adoption (or not) by a sufficient number of customers to cause
profitable sales.

4.2.4.1.1.4 Intrinsic Adoption Rates

Different kinds of potential customers (segments) will be more likely to

use a product (have different intrinsic adoption rates) in differing volumes
(usage rate) to determine segment demand.

Potential customers become customers once they have both needs

and have adopted the product. A small, or a large, number of the potential
customer population may have need to use, or be highly likely to be
influenced to use, the product form. Different segments have different usage
rates. Adoption occurs after potential customers move through the buying
decision process (i.e. the hierarchy of effects’ steps of awareness, knowledge,
liking, preference, intention-to-buy and purchase). A world of perfect
information would allow collection of these data by primary research, but that
is impossible for cost reasons. ‘Intrinsic’ is used here to distinguish the native
adoption rate from the altered rate induced by the organisation’s impact on
potential customers through application of its marketing mix. Capacity to pay
for the product should be considered prior to estimating intrinsic adoption and
 70

usage rate, or in segment population size, depending on how the segments

are defined.

This variable is the base for later forecasting (in customer adoptions
and usage rate, an intervening variable) how quickly, how many, and how
frequently, potential customers are expected to change to the new entrant.
However, in the research presented here, sales by market is not calculated
because of the absence of adoption rates by product and country. Instead,
the cluster of the attractive markets is provided.

4.2.4.1.1.5 Segment Demand

Segment demand is the most fundamental of all screening data about

each market and is the engine which powers the results from the entire
screening process. The organisation should not be active in a market which
has less than some minimum threshold level of demand that justifies
overcoming the costs of market entry15. However, low demand should not
exclude a market if other highly attractive features, say high profitability,
compensate it.

Industry demand. Statistics for the segment’s demand are often

unavailable, requiring the organisation to consider data for the entire industry
then reduce them to the estimated segment volume. Similar segments (if
any) which are served by the organisation and its competitors aggregate to
become the industry’s demand. As the size of industry demand increases, so
does the range of potential strategies possible for the organisation to
operationalise, for example, mass market versus niche, use a wider range of
modes of entry, etc..

15
This assumes that loss-making but rational behaviours do not occur (just two examples of
which are to block competition or to cross-subsidise another product), and that managerial
psychological needs do not distort the process (e.g. deriving status and hedonism from
travel). Matters such as these are allowed for in the framework under ‘other benefits’ and
‘psychic benefits’.
 71

A specific society’s industry demand for most products is unchangeable

in the short to medium term, although organisation strategy can affect the
share obtained. Industry demand and segment demand are thus treated as
uncontrollable, moderating variables. The separate aspect of the
organisation’s product-specific demand by market is able to be influenced by
the organisation (through its strategy and marketing mix), so is an intervening
variable which is derived during the screening selection process.

Adjustments. Industry and segment demand statistics usually under-

represent demand by not normally being able to include unmet demand and
latent demand. The framework partly accounts for this by including an
allowance for unsolicited orders and enquiries received by the organisation.
These are symptomatic of inadequate supply, customer discontent with the
product, relationship discontent between existing supplier and buyer, etc..
Past demand is not necessarily representative of future demand, so long- and
short- term growth and discontinuities in demand also have to be estimated
and allowed for.

Some theorists and practitioners suggest that some markets should be

eliminated prior to conducting screening (a Go / No-go pre-screening
elimination of some markets). The reasons commonly advanced to disbar a
market are:

• domestic regulations prohibit sale to specific countries, including

foreign trade barriers,
• foreign market regulations prohibit purchase from the country of the
potential supplier,
• restraints on trading (sales or purchasing) due to territory
allocations, intra group arrangements or sourcing policies.

Considering these matters at this stage is a potentially sub-optimising

behaviour because it can reject high potential markets which, in the short to
medium term, could be highly profitable. Regulations can frequently be
 72

overcome by lobbying, networking, making minor product changes or finding

legal loopholes. Large sales can be made at national borders. Joint ventures
with a local firm can be entered into. Exceptions can often be found to
prohibitions (for example, it would be an unwise alcohol manufacturer which
rejected at the screening stage all Muslim countries, because alcohol can
usually legally be exported to these countries for consumption by the
diplomatic community, and sometimes also by the ex patriot population, and
sometimes also by the often large non-Muslim populace). Insufficient is
usually known at the screening stage to safely exclude markets for these
reasons, so their possible exclusion should occur later during the in-depth
assessment step.

Growth in demand. Demand growth is seen as desirable by

organisations because it generally provides increasing returns of scale and
scope, subject to competitive actions.

The above section of the framework on Potential customer aspects

may have seemed overly detailed. However, that detail is necessary both for:
(1) conceptual completeness and (2) because demand statistics need to be
able to be calculated in a variety of ways to compensate for the absence of
statistics for 60,000 products in the over 200 countries. Thus, the various
Potential customer aspects can be operationalised to estimate (1) product-
specific market size, and (2) product-specific growth figures by either:

1. (i) using an organisation-specific customer needs regression, or

(ii) (a) multiplying each market’s segment population by each
segment’s estimated intrinsic adoption rate, then adjusting for
usage rate, any unsolicited foreign orders and enquiries
received and any estimated temporary growth or decline in
segment demand, or
(b) taking local production, adding imports, deducting exports,
adding or subtracting inventory change then making
appropriate adjustments (above), and
 73

2. taking the shift-share of each market’s product-specific production,

imports and exports (with adjustments, above).

New and existing markets. It is necessary to include both new and

existing markets when comparing where to expand. The proposed screening
framework is grounded on the organisation having previously decided that it
wishes to internationalise further. However, it may be that the organisation
should instead expand in its existing markets. Existing markets are implicitly
involved in the screening selection decision because they provide part of the
information base for setting realistic levels of achievable objectives in the new
market. Existing markets should be explicitly included in screening so that, if
they emerge in the short-list of preferable markets, they are further assessed
for market expansion in place of a foreign market entry. The higher level of
existing knowledge about them and the investment already in them could
provide the organisation with a source of lower cost, lower risk, additional
sales volume and profit than those gained from entering unknown new
markets.

4.2.4.1.2 Competition and Substitutes

Competition and substitutes is a major aspect of country attractiveness

which varies between markets at two levels – the economy wide and specific
industry levels. Competition can be thought of as the degree of competitive
behaviour of organisations, and the industry/market structure within which
organisations operate. Competition is important because it affects each
organisation’s ability to influence the price and terms of sale, and so ultimately
the profit earned.

At the industry level, each market may have different amounts of

competitive intensity. Economists such as Lipsey, Langley, and Mahoney
(1986) tell us that this is determined by:
 74

• the number, size, degree of knowledge and political influence of

competitors in each market,
• the number, size and knowledge of customers in each market,
• how homogeneous the product is, and the degree of substitution
possible between different products,
• the barriers to entry and exit, and
• supplier ability to influence demand by promotion.

Marketing strategists add:

• the likelihood of and impacts from retaliation,

• any first mover advantage, and
• the levels at which the industry is presently meeting customer
needs.

A new entrant ideally seeks competitive industry markets (with low

entry barriers, fractured competitors, etc.), then after entry to monopolise the
market (to prevent others entering, charge maximum prices, etc.).

At the economy-wide level, the legal/political/cultural encouragement of

competition in each market will affect the competitiveness of the economic
environment within which the industry, suppliers, customers and competitors
make transactions. The factors of production in each market will thus be
relatively more or less competitively priced, generic entry and exit will be
easier or harder, etc.. A new entrant with good sustainable competitive
advantage seeks competitive markets because these have low levels of costs
and minimise constraints on supply or demand.

These micro and macro competitive stresses affect the resources and
effort likely to need to be expended by a new entrant to achieve market entry
and subsequent profitability. These stresses also place limits on the
organisation’s feasible strategies. For example, the degree to which
competitors are meeting the needs of a particular market segment (assuming
 75

the new entrant has experience with and a desire to service it) will affect the
kinds of products offered and the level of sales likely to ensue. In turn, likely
volume will affect strategy, for example, mass marketing or niche. The new
entrant’s level of international experience and their innate strategic
competitive advantages may be strengths which it can use to ultimately
dominate the particular competitive circumstances. An appropriate level of
resources must also be applied by the new entrant to pursue its chosen
strategy. The organisation’s level of sustainable competitive advantage is
available from the organisation by primary research and can be used as a
general indication of its ability to compete at the industry level (see section
4.2.4.2.1.1). However, that does not indicate the opposing strength or
behaviour of industry competitors.

This doctoral research used an estimation of the economy-wide level of

competition based on the WEF16 assessment of competitiveness in countries.
If, in future, data become available on the industry-specific or product-specific
competitiveness, this should be included.

16
The World Economic Forum.
 76

4.2.4.1.3 Other Environmental Matters

Aside from competitive and customer matters, the other environmental

matters which influence aspects of each country’s marketing attractiveness
are driven by the physical, economic, technological, legal, political, cultural
and ecological dimensions17, and the availability and quality of data. These
each interact with other moderating, independent and intervening variables to
enhance or retard many of these other variables. Some of these variables
affect all products (economic, cultural, political and legal influences) whilst
others affect only some products (physical, technological and ecological
variables). Each of these eight sub-components is now briefly discussed.

4.2.4.1.3.1 Physical Influences

Physical matters often differ in each market and include climate

(temperature, rainfall, snowfall, humidity, wind, sunshine), topography (hills,
altitude, other geographic features) and population density. These sometimes
affect aspects of product-specific demand (e.g. weather may affect demand
for sunburn protection products, but be irrelevant to demand for legal
services) or the marketing mix (e.g. product features required to cope with
climate). This is thus a product-specific variable which applies to some
products, but not all.

4.2.4.1.3.2 Economic Influences

Economic market attractiveness flows from the plethora of economic

criteria (e.g. size or growth of population, level of wealth, infrastructure, etc.)
whose values vary, so crudely indicating to a firm that some markets are more
desirable than others. The economy has a direct influence on the volume and

17
It could be argued that these ‘Other environmental matters’ should not be considered. This
line of argument would suggest that the firm’s analysis of its product’s customer needs,
segments, targets and positioning would already have compelled consideration of the impacts
of all environmental influences on the marketing mix. However, that assumes an unchanging
marketing mix, irrespective of country – an increasingly rare event. Be the product mass
marketed or micro-segmented, these environmental influences differ by market, so producing
a differential attractiveness of one market over another.
 77

growth of product-specific demand, and an indirect influence as derived

demand. The general economic environment further indirectly affects the
marketing mix via funds availability and cost, tax levels, cost and availability of
infrastructure, range of promotional options, etc.. The economy’s volatility and
nature can affect the other moderating variables of competition, technology
and risk. Data on economic outcomes are a function of the discipline’s
predetermined decision logic, resource endowments and each market’s
discretionary political decisions.

4.2.4.1.3.3 Technological Influences

Technological environment refers to the match between market and

product in terms of the sophistication or complexity of science and technology
(e.g. the necessary presence or absence of computers, bioengineering,
aerospace and other attributes or industries). The technology levels
demanded by customers affect the volume and kind of product to be supplied.
Some products are technology sensitive whilst others are not, so this is a
product-specific variable which applies to some products, but not all.
Customers may bypass technological shortcomings of a market (e.g. industrial
customers, but less frequently consumers, use generators to supplement
erratic electricity supply) so the technology needs to be essential, not just
desirable, for product use by the key segment.

4.2.4.1.3.4 Legal Influences

The legal environment in each market consists of two matters. Firstly,

there are aspects such as the product-related controls, the trade practices
laws and regulations, restrictions on ownership, and contractual relationships
with customers, employees and suppliers. Some of these matters will
differentially affect some organisations in some markets , for example, trade
practices laws may help small or new entrants whilst hindering large
multinational firms. The screening framework does not allow for product-
specific assessment of these matters. It is argued that it is unwise to screen
markets out on product-specific legal grounds because, given an ethical
 78

product and a strong potential demand in these times of increasing

internationalisation, entrepreneurs will in the long term usually find a way to
negotiate around legal barriers. How quickly and easily that can occur is more
desirably made a matter for more detailed research at the in-depth
assessment stage.

The second legal matter arises out of the popular term “The rule of
law”. That notion summarises – without specifically articulating – how open,
predictable and fair the country’s judicial system is. Concrete examples are
whether the nation’s rules are codified or not and whether there is an
independent judiciary. Openness, fairness and predictability are universal
legal influences on all organisations seeking to operate in that market. Their
enforcement makes it much harder for individuals or organisations to by-pass
or change laws to suit self-serving needs. This dissertation thus weights
market selection for these universal, non product-specific, legal influences.

4.2.4.1.3.5 Political Influences

Political attractiveness of each marketplace includes components such

as government stability, degree of influence on organisations and markets,
national priorities, attitude toward foreign firms, and efficiency of the
bureaucracy. Each country’s political environment greatly affects the
economic and legal milieus. No overt judgment is made about the desired
kind of political organisation (e.g. capitalist – communist), although some
organisations in some countries may wish to add this to their selection criteria.

4.2.4.1.3.6 Cultural Influences

Cultural aspects include such matters as the market’s religious, ethnic

and linguistic variation and tolerance. These then affect screening framework
aspects such as potential customer needs, the organisation’s segment
preferences, demand volume and kind, and the marketing mix. Different
cultures also frequently have different business practices, so the ways in
which the screening organisation’s home market employees and management
 79

communicate and conflict with, manage, etc. their opposite number in the
eventual foreign market will partly be dependent upon cultural distance. Thus,
the higher the cultural distance between home and host market (moderated by
the organisation’s level of experience at internationalisation), the more difficult
it is for the organisation to adapt to the new market. More details are to be
found in and following Figure 4.5.

Cultural distance is calculated in this screening framework using

Hofstede’s 1984 schema, with rescaling from the home market per Kogut and
Sing (1988). The result is the comparison of the manager’s home culture’s
score with the scores of those in each market. The distances between the
markets are then included as one of the screening framework’s variables to
influence the calculation of each market’s attractiveness. The Hofstede
questionnaire is included in Appendix A as an example of the instrument used
to measure cultural difference, but is presently unable to be used in practice
for technical reasons that are detailed in the theory section (4.3.4.1.3.6).

4.2.4.1.3.7 Ecological/environmental Influences

Ecological or ‘green’ or ‘environmental’ aspects include the particular

market’s pollution levels and attitudes which differentially affect the quantity
and kind of market demand, the marketing mix and the other operations of the
organisation. For example, the adequacy of sewerage treatment, land fill
sites, transfer stations, safe drinking water and pollution monitoring stations,
together with the cultural values re the various kinds of pollution, should affect
the quantity and nature of demand for related goods and services and the way
in which goods are domestically produced. The objective is to isolate
significant positive or negative: (1) demand and (2) sentiment for or against
the product. Ecological concerns may not affect customer perceptions in
some or all markets, so this is a product-specific variable which applies to
some products, but not all.
 80

4.2.4.1.4 Market Risk Levels

Returning to Figures 4.2 and 4.3, risk levels in each market vary in
numerous ways, including economic, political, financial, competitive,
commercial and personal. Examples include high inflation, high debt service
payments, invasion, civil war, government loan default, kidnapping,
assassination, customer payment default and delayed customer payments.
Offsetting market risk slightly is the reduction in risk, which arises from
diversification – not having ‘all the eggs in the one basket’. This Market risk
levels variable quantifies the probable levels of risk in each market, not the
organisation’s attitude toward risk – this latter matter being considered later
under Risk Preferences.

Commercial risks associated with particular customers, and the

personal risks of travelling to or living in the market, will also not be
considered during screening – that level of detail being a subject for in-depth
assessment. However, customer and personal risks will partly covary with
economic, political and financial risks and so are indirectly included to a
limited extent during screening. Country-specific risk can be reduced by
mode of entry choices, but that reduction should not be considered until in-
depth assessment for optimisation reasons detailed in section 3.5.

4.2.4.2 Organisation’s Capabilities and Preferences

The specific organisation’s capabilities and preferences comprise:

• the organisation’s abilities, being in turn composed of the

organisation’s sustainable competitive advantages, its general
business competencies, further internationalisation competencies,
resources, and whether stakeholders are supportive further
internationalisation, and
 81

• managerial values, which include values about further

internationalisation, preferences about strategy, and personal
values.

These matters affect how capable a specific organisation will be at

developing the market potential. In contrast, the earlier country attractiveness
matters relate to the size of market potential and how difficult it would be for
any organisation to develop that potential. Primary data about the
organisation’s capabilities and values are collected from the organisation to be
used to hone the market selection to the organisation’s unique traits. The
international performance literature and the psychological literature are the
predominant influences on this section of this dissertation.

The discerning reader will have noted the different terminology

between the headings and sub-headings in this area of the framework
(‘Organisation’s abilities’ and ‘Managerial values’ become ‘Organisation’s
capabilities and preferences’). The sub-variables which comprise
‘Organisation’s abilities’ measure existing abilities (competencies and
resources). These are used as indicators of the organisation’s likely success
at doing something new in the future – internationalising a further increment.
Since the organisation has not yet performed the task, a decision has to be
made about whether the organisation has the potential to succeed. The
psychological literature similarly distinguishes between ‘ability’ and ‘aptitude’.
A second matter is the difference between the ability to do something, and the
willingness to do it. An organisation will at times have the ability, but not the
values which lead to desire, to further internationalise. It may alternatively
have the values, but insufficient ability. If the organisation has both,
preconditions have been met. In other words, both necessary and sufficient
conditions have been met for further internationalisation. Additionally, when
the two combine, synergy is involved, and the organisation has an even
higher potential to perform. The screening framework’s change of terminology
in the cluster heading to ‘Organisation’s capabilities and preferences’ signals
that it is considering whether both these preconditions have been met, and
 82

that synergy may be operating. This semantic subtlety (ability versus

capability) is maintained throughout this dissertation18.

There is a body of literature which speculates that some industries may

have peculiarities in behaviours or assets which produce unique performance
result. For example, Holzmüller and Stöttinger (1996, p. 44) found industries
varied in their support of export activities, consequently affecting the position
they held in international markets, and so their organisation’s export
performance. Consideration was therefore given in the current research to
inserting a nominally scaled ISIC (International Standard Industry
Classification) identification code as a dummy variable to allow isolation of the
product, and any peculiarities it may have, from blended results for all
products. However, the SITC code fulfils this need, and is translatable to each
of the other coding systems, such as ISIC, HS and ANZIC (the Harmonised
System, and the Australian and New Zealand Standard Industrial
Classification system, respectively).

4.2.4.2.1.1 Sustainable Competitive Advantage (SCA)

The first of the organisation’s abilities, its sustainable competitive

advantage (SCA), is assessed because it can be an inherent limitation – an
organisation with a ‘me-too’ product or a product which can not be
internationalised will be at a disadvantage that is probably insurmountable.
SCA thus is one of the determinants of how forcefully an organisation may
compete within its industry.

Sustainable competitive advantage stems from a unique asset or

assets among the factors of production19. If the organisation is already
operating internationally, in comparison with domestic-only organisations, it

18
Common language use often blurs the distinction between ‘ability’ and ‘capability’, but
dictionaries show a clear difference. See, for example, Collins (1979) pp 3 and 223;
Macquarie (1981) pp 49 and 292; Oxford (1962) pp 2 and 112.
19
i.e. land; a location advantage such as proximity to market or resources; labour cost or skill;
government legislation and policies; technology; monopoly rights over patents, brand names,
trade marks, and/or a particular raw material or market structure (e.g. position in distribution
chain as producer, wholesaler &/or retailer).
 83

may have certain additional advantages20. These various advantages lead to

economies of scale or scope (Porter’s 1980 cost, differentiation or focus)
which can in turn produce customer perceptions about absolute, and
competitor-relative, value which competitors may not be able to match. The
screening framework assesses the organisation’s sustainable competitive
advantages using a modified version of the questionnaire developed by Aaker
(1995, p. 80) which is based on economic theory – details are in the Theory
section. Aaker’s questionnaire includes negligible assessment of the abilities
of the organisation’s existing employees, including management and
functional specialists. It therefore requires supplementation in that area (see
additional sub-headings in Organisation’s abilities and Managerial values).
People aspects are increasingly important as economies move from
producing tangibles to services, and in eras of high change (instanced by
concepts such as the knowledge organisation, continuous change and the
communications revolution). SCA may be generic (e.g. where an organisation
has a leading-edge production technology) or market-specific (e.g. where an
organisation has the monopoly supply of a resource in a particular market).

4.2.4.2.1.2 General Business Competencies

General business competencies are one of two kinds of relevant

competencies which are developed through learning. Figure 4.6 summarises
how competencies relates to screening.

People are exposed to intentional and unintentional learning as they

live out their lives. Relevant learning may lead to increases in performance
competence. Psychologists see learning arising cognitively, affectively and
behaviourally (roughly equivalent to learning from thinking, feeling and doing),
and expect that which of these learning styles is optimum to be determined by
individual preferences and the situation.

20
i.e. to conduct transfer pricing, shift liquid assets, diversify investments to reduce risks,
reduce the impact of industrial unrest by parallel production, and to engage in international
product- or process- specialisation (Dunning 1980:9).
 84

General business competencies are the abilities needed in an

organisation (other than the international business abilities) to be successful
as a result of marketing their specific product in the home market.
Competence involves having both suitable and sufficient skills for success at
the tasks. Incompetence in managing the home market will normally
contaminate the organisation’s performance in foreign markets.

Component functional skills of general competence relate to

management, marketing, financial (and perhaps planning, communication,
human resource, networking, operations, research and development, and
quality control) skills, all in differing quantities and kinds that probably partly
depend on the demands of the industry-specific business environment. The
organisation’s individuals have competency types and levels which are
determined by their innate capabilities, their education, training and
experience, and to some extent by their attitudes and temperament.
Interaction between competence attributes and effort levels of the
organisation’s employees and the environment determines organisational
performance. Top management requires different skills from the
organisation’s functional operatives. The framework therefore assesses these
general competencies in both management and the international specialists.

4.2.4.2.1.3 Further International Competencies

The organisation’s further internationalisation competencies may be

market-specific (e.g. deep language skills, network contacts, market-specific
research) or generic (e.g. processing letters of credit, assessing and insuring
against market-specific risks, generic international market research
capability). Again, see Figure 4.6. ‘Further’ internationalisation distinguishes
the lower level maintenance competencies from the higher level competencies
needed to expand international involvement.

Internationalisation competence has far reaching interactions with

many other variables (e.g. such matters as the psychological ability of the
organisation’s employees to cope with differing cultural demands, their ability
 85

to make correct decisions despite information shortages, and the ability to

cost-effectively minimise marketing mix effort yet achieve foreign customer
adoptions). This dimension is similar to the previous variable’s generic skills
but the organisation’s product is now set in an international environment (i.e.
the previously domestic-only organisation needs additional abilities to
manage, market and finance its business internationally). It is perhaps again
industry-specific and probably again requires different capabilities in top
management from functional operatives. International competence begins
with whatever education has been undertaken. It is enlarged by training and
practical experience as a result of the duration, intensity and range of
situations experienced – different cultures, economic conditions and foreign
competitive intensities encountered by different staff seniorities and functions.

The conceptual path from pre-export to experienced internationalised

organisation requires employees to learn to use a range of ideas and
behaviours that maximise organisation and employee utility. The stage of
internationalisation of the organisation is highly likely to indicate the level of
learning and the probable behaviours (such as whether the organisation will
be undertaking market expansion or global rationalisation) and probable
outcomes (such as how difficult a market the organisation can profitably
enter). Organisations at different stages of internationalisation usually have
different motives, psychic distances, market risks profiles, and perceive
different obstacles to and advantages of further internationalisation.

Whether it is an individual or an organisation which has experience at

internationalisation requires comment. Organisations are inanimate
aggregations of individuals; their accumulated experience and wisdom are
largely held in the minds of the employees. It is thus the level of international
competence of the organisation’s key individuals which determines the level of
internationalisation of the organisation. An excellent individual with great
experience may typically change the aggregate level of capability of the whole
only marginally. That individual’s impact will depend upon their hierarchical
power to influence decisions, the amount of experience flowing in to other
organisation employees from having to deal with existing foreign customers,
 86

the levels of general and international competencies of other significant

employees, and so on. There is thus an interaction between individuals and
the rest of the organisation which determines the stage of internationalisation.
All key international decision-makers and functional specialists need
assessing when measuring this dimension.

Whilst measuring stage of internationalisation is controversial, it is later

argued that it is intuitively logical, consistent with the literature and
theoretically appropriate to use the following indicators of the stage of
internationalisation achieved by the organisation so as to indicate the
organisation’s level of competence to undertake further internationalisation:

• number of markets regularly sold in,

• range of cultural distances of markets regularly sold in,
• number of modes of entry regularly employed,
• number of recent market entries, expansions and rationalisations,
• language competencies of international marketing staff,
• formal relevant education and training accreditation obtained in
international business matters,
• the amount of practical experience in such functional areas as
international marketing, international finance, international
management, etc.,
• whether the top decision-makers were born abroad, have lived
abroad or worked abroad,
• a comparison of the degree of internationalisation of the
organisation with the degree of internationalisation of the industry,
• turnover of the international marketing staff,
• satisfaction of the key target customer segment with the
organisation, and
• management’s assessment of performance.

The higher the total score across the variables, the more internationally
competent the organisation.
 87

Adding:
• last year’s international training hours per international employee,
and
• last year’s international market research per international employee,
signals whether the organisation is continuing to learn, although very high
figures here would indicate insufficient current productive activity which would
then retard short and medium term performance.

An important sub-group of international competencies are those which

relate to specific markets, because they indicate an existing capacity to enter
those markets more efficiently and effectively than it the organisation did not
have them. Significant language skills, recent market research on a particular
market, and relationships with buyers (or other important network contacts)
are possibilities which can justifiably sway deciding in favour of one market
over another, all other things being equal. The Questionnaire to Management
thus asks about these matters then the framework includes considering them
during the market selection process.

The sub-dimension international market research competence

assesses the organisation’s ability to conduct international market research,
both screening and in-depth. It is actually a sub-component of the umbrella
variable further internationalisation competencies, but its significance in this
topic is felt sufficient to separately emphasise it.

Different organisations have different levels and kinds of market

research skills. The separate variable, market research values, measures
how enthusiastic the organisation is to use whatever research competences
they possess. What minimum level of research competence is necessary?
Within limits, do increases in screening proficiency add value to the final
screening decision?
 88

Research competence is driven by the quality of research resources

available (of time, software and hardware, staff and accumulated knowledge).
The quantity of each of these which is available is a matter for the resources
variable. Employee training, experience and past performance at using
international marketing research are expected to be important indicators of
research competence.

Stage of the organisation’s internationalisation will affect the

organisation’s accumulated experience, including perhaps its international
market research capability. Presumably different screening research skills are
needed at different stages of internationalisation. (For example, Australian
and New Zealand firms will find researching each other’s markets a much
easier task than researching China, where secondary data often need
translating into English, has unfamiliar hidden assumptions buried in the data,
and the country has a less developed infrastructure which precludes delivery
of the elements of the marketing mix in the usual manner.)

However, since there is not yet an agreed, effective screening

methodology, no organisation at any of the stages of internationalisation can
yet have built up great expertise at screening markets. The screening
methodology needs to be discovered before the competencies necessary for
its use can be determined with certainty. This variable, international market
research competence, therefore can at this time safely consider broad
international market research competencies. It can also, from the literature,
presume certain indicators of the organisation’s employees’ potential to
conduct and use screening, but the set of competencies believed to be
involved in international market screening may change in the future as a result
of further research in this area.

If the premises of this dissertation are borne out, it seems that the
marketing and mathematical competencies needed to screen international
markets should be able to be assimilated by someone who has passed
reputable undergraduate degree courses in both marketing and statistics.
However, the volume of data needed to be collected, the unusual sources for
 89

much of that data, and the normally infrequent use of the specialist statistical
procedures, may make the task onerous for a non-specialist. Foreign market
screening may thus be a cost-effective but boutique market research
specialty.

Important influences on the level of research competence reached by

the organisation include: (1) management’s views about the importance of
research (their situational motivation to use, and their long-term values about
using, it), (2) data availability and quality (to permit research practice), and (3)
management’s ability to interpret the data. The market research values
aspects appear in the framework as a sub-set of personal values whilst an
organisation’s ability to use research is implied by general and international
business competencies.

4.2.4.2.1.4 Key Stakeholder Support

Few occasions arise where management is truly autonomous because

resources and employee support are needed to face competition and other
restraints and to harvest opportunities. Support from various directions makes
it less likely that additional internationalisation is neither vetoed, slowed, nor
halted part way through. There are differing amounts of support, so there are
ultimately different degrees of managerial decision-making autonomy. The
extent of this support is assessed by the variable Key stakeholder support.

Key stakeholders may be internal or external, and comprises the

further two sub-divisions of support needed and support available. Most
organisations have several internal people involved in decisions as
gatekeepers and/or facilitators, such as managers, key functional specialists
and shareholders. Banks, governments and others external to the
organisation sometimes have strong views about aspects of an organisation’s
behaviour, along with a capacity to impose their views. These internal and
external views can affect the other moderating variables through such
mechanisms as influencing what is seen by the organisation as an ‘attractive’
or a ‘difficult’ potential market, affecting other aspects of managerial attitudes
 90

and motivation, and altering resource allocations to internationalisation. A low

level of stakeholder support for internationalisation will cause management to
attempt to enter less difficult and less risky markets. Since entering new
markets and reaching profitability is rarely achieved in less than three years,
and since aborting a market entry part way through is a gross waste of
investment, it is desirable that the organisation be confident that key
stakeholder support will be ongoing. An allowance for the level of stakeholder
support is thus included when determining the difficulty of the market able to
be handled by the organisation.

4.2.4.2.1.5 Other Resources

Other resources is the quantity, quality and duration of fiscal, physical

and human input available to the market expansion. Quality of human
resources has already been covered in the competence and key stakeholder
support variables so are not included here.

The additional market’s demand for resources can not exceed supply if
organisational performance is to be maximised. Although capitalistic
philosophy argues that fiscal resources are theoretically available to any
‘good’ scheme, the practical matters of the level of the organisation’s
accumulated reserves, the project’s profit and risk, and the supply of
management time, ability and motivation to find and sell investors, all do
actually limit supply and timing of finance. Human resource supply is that
required for continued operation in existing markets plus a small addition
available from hiring new employees. Only small increments in staff can be
added when limited supply, product-specific, specialist skills need to be
attracted and inducted, and when maintenance of the organisation’s existing
culture is desired. Financial and human resources are thus considered by the
framework as being fixed in the short and medium terms.

Note that an actual resource figure is not calculated in the proposed

framework because profit is not being calculated. This is partly because, as
previously argued in section 3.5, the cost-determining mode of entry is not
 91

selected until the later, in-depth stage. All that is required at the time of
screening is a go/no-go determination that the organisation probably has the
resources to both (a) undertake initial market entry in an additional market,
and (b) to support consequent, medium term expansion in that market. Whilst
the exact amount of resources needed can not be predicted accurately at the
screening stage, internationally experienced organisations can be expected to
have an idea of whether or not they have the resources to attack an additional
market, and if so, whether or not they can, for example, handle extremes such
as a large, difficult, long pay-back market. Internationally inexperienced
organisations are unlikely to be able to as accurately answer this question, but
will nevertheless have some ideas from their domestic activities.

This forecasting difficulty is partly assisted by the fact that theoretical

optimums of each resource are really ranges, outside of which significant
diseconomies of scale operate to retard efficient outcomes from the attempted
new market entry. Below some product-specific resource level, the
organisation has insufficient wherewithal to adequately expand, particularly
into long pay-back, risky projects such as internationalisation; above another
resource level, there are reducing returns to scale which also retard
performance. The need to exactly calculate resources is thus reduced.

The amounts required will partly depend on most of the other variables
in the framework. The modes of entry which could be chosen (e.g. export
versus foreign manufacture) require different amounts and kinds of resources.
Since mode is determined after screening, an implied assumption is
sometimes made by management during screening about which mode or
modes can be afforded. Amount of resources is also partly driven by strategic
competitive advantage and strategy – for example, cost leadership and mass
production versus small volume, customised production – as is kind of
resource – differentiation requires strong creative and marketing people skills
which are different from those needed in staff implementing a cost based
strategy. Higher levels of competition and low levels of organisational
competence increase the amount of resources needed, whilst risk level
chosen requires deeper pockets to ride out the potentially greater volatility in
 92

outcomes. Objectives chosen and management motivation influence internal

allocation of scarce resources to competing opportunities available to the
organisation.

There are two issues of concern to management in relation to resource

duration: the start-up demands and the later steady-state operating demands
for funds, physical assets and people. Several years are typically required to
reach break-even in a new market, and significant resources may be needed
for several more years, as success will require market expansion to follow the
initial entry. Management’s motivation for the particular new market entry thus
needs to be sustained for several years – during which circumstances,
priorities and even managements commonly change. Gross wastage of
assets occurs when the organisation funds a market entry, then has to
withdraw because of resource over-commitment.

Resource availability for the envisioned task is the constraint being

considered here, not the dimension ‘firm size’ for which there are empirically
mixed findings (e.g. the eight studies reported in Bilkey 1978). The resources
needed to achieve the task, not the size of the organisation, is surely what is
important. Using ‘firm size’ should confound research results because it
amalgamates resources, general competence and international competence
and ignores the resources required for accomplishment of the specific task.

Each organisation ultimately thus has a different ability and willingness

to redeploy and/or increase its level of funds, equipment, intangibles and
people for use in the additional market entry.

4.2.4.2.2 Managerial Values

Managerial values underpin the organisation’s behaviours, and

behaviours affect performance quality. For example, management’s
education and experiences in the international sphere affect their views about
the worth of internationalisation, affect their interest in it vis-a-vis domestic
activity, and influence their belief about their abilities to actualise international
 93

potential; these then affect international performance. Thus, because

managers and some other employees have discretionary decision-making
power to affect market entry outcomes, their individual and unique
characteristics need to be considered. Each individual has different views and
competency quotients which, despite the same decision circumstances, can
cause a different decision.

This cluster comprises further internationalisation values, strategic

values, and personal values, including cultural values, risk preferences and
market research values. Some of these items reappear in the independent
variables section in a slightly different form (i.e. strategy decisions, personal
benefits and chosen risk level). The subtle difference is that those shown
among the moderating variables are ingrained and largely uncontrollable
values which are unchanging in the short or medium term, whilst those among
the independent variables are controllable matters which have been elected.
Another way of viewing this is that these moderating variables are preferences
whilst their independent variable equivalents have become expectancies.
 94

4.2.4.2.2.1 Further Internationalisation Values

Further internationalisation values relate to the two areas of how good,

and how difficult, it is likely to be to further internationalise. Alternative terms
are rewards and difficulty, or the cost/benefit, of entering another foreign
market. Note that these are values (generic and long-term psychological
dimensions), whereas motivation and commitment comprise attitudes
(application of the values to a specific situation).

Management has values about whether further internationalisation will

be ‘good’ or ‘bad’ for their organisation; this obviously influences the amount
of internationalisation they undertake. These values are partly produced by
the manager’s international training and experience. The stage of
internationalisation (a reflection of the accumulated experience of the
organisation) will be a helpful predictor of what obstacles or benefits are
perceived likely to be experienced by the organisation and its management if
it internationalises further. Personal values are related – international learning
will affect expectancies such as what hedonistic pleasures, achievement
needs and security needs will flow from entering the next market. Further
internationalisation values can also be expected to interact with objectives,
strategy decisions and motivation, and so affect the market preferred.

4.2.4.2.2.2 Strategic Values

Strategic values include ceteris paribus preferences about such

matters as standardisation or adaptation, incremental or simultaneous market
entries, and so on (more components are detailed in section 4.3.2.3). Those
decisions are driven by more fundamental values, including the manager’s
preference for long term versus short term solutions, premium versus mass
customer targets, economic value adding versus other forms of measuring the
effectiveness of debt and equity use, the desire to operate globally versus
locally, and whether the organisation aspires to become the market leader,
challenger, follower or nicher.
 95

The predisposition to select specific strategic options is thus rooted in

the manager’s values, a personality matter which is largely unchangeable in
the short or medium term, then tempered by education, training and
experience. Judged against a person’s strategic values preference set,
different markets will be viewed as having different opportunities and
difficulties for goal achievement when implementing the preferred strategy
options. For example, if their personality or circumstances produce a high risk
aversion, survival may dominate profit making when that manager makes their
strategic choices, so excluding such markets as the presently risky and
volatile former USSR states. New facts about the characteristics of each
market may cause the manager to alter his/her decision about what is
apparently ‘best’ in each strategic aspect. For example, the forecast product-
specific demand level should be one of the influences on the choice between
product standardisation and product adaptation.

Strategic values are obliquely taken into account via the objectives
chosen, the weights given each objective and the embedded strategic
assumptions leading through the marketing mix to the customer adoptions
and usage rate. However, the screening framework does not directly
integrate strategic values into the selection mechanisms.

4.2.4.2.2.3 Personal Values

An individual’s personal values determine the pleasures and

discomforts they will derive from the specific situations they experience.
Values also shape the person’s decision choices when handling various
business situations, such as predisposing them to take a long-term or a short-
term perspective, be aggressive or cooperative, proactive or reactive, need a
high or low degree of control, enjoy ambiguity or not, and so on. These
values are enduring, individually variable, personality components which
guide their attitudes and behaviour before, during and after entering another
market.
 96

Positive and negative personal satisfactions can flow as a result of the

market entry tasks of:

1) interaction with foreign locations, particularly travel to and living

in them. Travel and residence yield possibilities for satisfying
personal needs, and a person from a western country could be
expected to experience various levels of stimulation, hedonism,
self-direction and security (Schwartz Value System 1992 items).
These needs are satisfied varyingly by different markets, and
different managers have different interest and satisfaction
patterns, so the screening framework should ideally cause a
matching process to occur between individual and market by
both dimension and strength;

2) attempting to, and perhaps succeeding at, accomplishing the

other managerial and technical processes of further
internationalisation. The degree of success of these
accomplishments is measured by the level of (hopefully
increased) sales, market share, profits, ROI, EPS, organisational
value, image, influence, achievement, or whatever the desired
outcomes. Attempting and/or achieving these outcomes can be
privately satisfying, and so meet psychological needs such as
achievement, security, and probably power, hedonism,
stimulation, self-direction, benevolence, universalism, conformity
and tradition. Extrinsic benefits, such as pay or promotion, may
also follow success. Thirdly, there will also be knock-on affects,
because publicly achieving the outcomes may enhance personal
and organisational prestige and capability, leading to employees
obtaining further satisfaction of their psychic and material needs.

For both of matters one and two, people have differing capabilities to
cope with different cultures’ approaches to their deep-seated (and so largely
unchangeable) attitudes and values, such as toward bribery, sexuality,
punishment, etc.. The experiences gained during coping then feed back to
 97

affect capability, and so ultimate performance of the individual and

organisation.

To some extent, there is an interaction between what we humans like

doing and what we become capable at. This is the triangular linking between
a person’s abilities and aptitudes, their attitudes and values, and learning. It
can thus be argued that the individual’s (1) personal values, together with
(2) their abilities and aptitudes, and (3) their formal and experiential training
and education, combine to affect competency-type dimensions in the
screening framework. Competence or not at internationalisation follows, and
is reinforced by the resulting success or otherwise. Competence then
influences which kind of market should be selected by screening.

Note the difference between personal values and personal benefits.

Personal benefits are items that the manager requests as screening
objectives. The manager’s personal values will affect the quantum and kind of
objectives desired, but personal values are optimums or ideals which normally
need to be compromised (because of the influences of the moderating
variables) before becoming consciously desired personal benefits (objectives).

4.2.4.2.2.3 (a) Cultural Distance

Cultural distance. One definition of culture is that it “... is the collective

programming of the mind which distinguishes the members of one ... society
from those of another” (Hofstede 1984, p. 82). Each individual expresses
aspects of their mind via personality. One definition of personality (Byrt 1971,
p. 32) is that it comprises: (1) abilities and aptitudes (2) attitudes and values
and (3) temperament. The collective programming of each individual’s mind
come about from common enculturation and acculturation processes which
produce some common values in the society’s people. These values place a
boundary around the person’s intercultural competence through limiting their
ability and motivation to perceive, learn and respond to the needs and wishes
of other cultures. As mentioned, attitudes and values additionally affect what
the person sees as satisfying and dissatisfying, and so which cultures they
 98

prefer to deal with. Values therefore affect the further internationalisation

objectives that management selects when undertaking markets screening.
These various matters result in the difference between the individual and the
cultures with which they interact being a partial indicator of the likelihood that
the individual will cope and succeed.

There are several methods for calculating the differences between one
culture and another, and reading the international marketing literature of the
last 10 years causes two names to stand out – Hofstede and Schwartz. Both
approaches have conceptual strengths and weaknesses and practical
advantages and disadvantages. Hofstede’s appeals include having over
116,000 respondents from 60 countries, and the way that his approach has
withstood the tests of time and other research. The disadvantages include
that his approach originates from data, rather than theoretical concepts, and
that individuals are unable to be tested and then be compared with groups21.
An alternative method of quantitatively assessing cultural difference is the
Schwartz Value System’s (1992) ten individual-level cultural values. These
dimensions are self-direction, stimulation, hedonism, achievement, power,
security, conformity, tradition, benevolence and universalism. The Schwartz
method does not have the same inability of the test instrument to be used to
assess individual cultural differences. However, the Schwartz value data are
available on only some 50 countries and have yet to prove themselves as
enduring as Hofstede’s. The Hofstede approach of measuring cultural
difference between countries was used in the screening framework.

In an increasingly multicultural world, people are born in one country,

may grow up in another, perhaps undertake part of their education in yet
another, then work in yet another country. It would therefore be preferable to
measure the cultural values of the organisation’s key managers, determine

21
Comparison is impossible because IDV and MAS, two of Hofstede’s five cultural sub-
scores, are quantified by being based on the first and second factors extracted from pooled
group responses. Since a sample size of even 100 is poor (Tabachnick & Fidell 1996:640),
Hofstede’s questionnaire cannot directly be used at the individual level. More importantly,
using factor analysis, country-level scores are usually different from individual-level factors
(Hofstede 2000 p.32).
 99

the mean values, then compare those means against cultural values in each
country. That presently appears infeasible. It does not rest with this
dissertation to resolve the matter, so Hofstede’s questionnaire is included in
Appendix A for the sake of completeness and as an indication only of the kind
of instrument necessary to assess cultural distance at the individual level.

In conclusion, minimising cultural distance between individual and

country is desirable to maximise performance – although it may not
necessarily maximise personal satisfaction in cases where the individual
prefers higher or lower levels of difference.

4.2.4.2.2.3 (b) Risk Preferences

Risk preferences is a sufficiently important psychological variable that it

has also been isolated for special treatment. Risk preferences of individual
managers vary, as do their levels of capability at handling situations with
differing amounts of economic, political and financial risk. There is also
interaction between the two – for example, a high risk market might be offset
by a manager with high competence in the market-specific and risk-specific
matters. The screening framework thus needs to ensure reasonable
congruence between this psychological risk preference, the risk objective set,
and the market risk level.
 100

4.2.4.2.2.3 (c) Market Research Values

Different managers have different predispositions towards conducting

research, hence the sub-variable Market research values. By a combination
of their personal values, capabilities and experience, people develop an
individually variable value about the helpfulness of marketing research.
Different occupations within the same national culture, and different cultures,
each can be expected to place different valuations on the usefulness of
market research. The stage of internationalisation of the organisation can be
expected to influence the kinds of matters researched, because organisations
perceive different obstacles and objectives at different stages of
internationalisation. A further dimension of stage of internationalisation comes
from the EPRG perspective (ethnocentric, polycentric, regiocentric and
geocentric) where the organisation’s evolution is commonly accompanied by
changing values about the importance of meeting customer needs, and so to
research which reveals those needs.

Motivation models tell us that an interaction can be expected between

the levels of this variable and management’s international market research
competence, constrained by data availability and quality, which then
determines the resources and effort put into market screening. The higher the
levels of organisation resources and effort put into screening, the higher the
likely capacity to conduct quality research that produces recommendations
which maximally meet the organisation’s objectives.

This variable presently reflects a preference by management toward or

away from market research in general, not screening research. Should a
satisfactory screening methodology be developed, the variable could be
honed to express management’s preference for the more pertinent market
screening research.
 101

4.2.4.2.2.3 (d) Other Personal Values

Other personal values comprises the remainder of a person’s values

after extracting cultural distance, risk values and market research values
(above). Included are such dimensions as needs for achievement, affiliation
and power. There is, regrettably, insufficient country data to allow the
comparison of these matters with individual managers’ psychological
preferences to weight market selection for matters in this area. Future data
availability is likely to eventually permit these influences to be included (e.g.
the Schwartz Value System 1992).

Only values related to differentiating among markets are considered

during screening – not those related to internationalising in general. This is
because the screening framework has been predefined to apply only to
managements which have already made the decision to further
internationalise their organisation and who have indicated being motivated to
do so.

4.2.4.3 Data Availability and Quality

Data availability and quality constrains the whole screening process.

Once the screening decision variables have been determined by research
such as this, the information used in each of them needs to be timely,
accurate, comparable and complete – “garbage-in, garbage-out” is the old
maxim. Timeliness and completeness are factual assessments whilst
accuracy and comparability require evaluation. Market-related data are
limited to secondary data because of the cost and impracticality of collecting
primary data in all 230 markets, and the 230 different data systems are not all
of like standard. There are often shortages of market data – data are missing
for specific countries or variables or cells. The screening framework also uses
some primary data about the organisation, some of which may be inaccurate
due to (intentional or unintentional) response bias causes such as veracity,
extremity bias, social desirability bias, etc..
 102

Distortions of data from these two sources can lead to asymmetric

information about markets or the client, which can cause a biased selection
decision. Data shortages also lead the researcher to needing rules about how
to treat cells which have data missing, for example, does the researcher insert
an average value, a value perceived as ‘typical’ for that particular cell,
eliminate the variable or market entirely, etc.. This dissertation used
researcher’s estimates to provide missing market data and to determine
scores for the minimally and maximally capable organisations.

High data availability and quality is the situation where quality data are
available for all variables in all markets – an improbable circumstance – and
when the organisation is willing and able to accurately assess its capabilities
and preferences. That situation would maximise the quality of the screening
selection process, minimise subsequent effort during market entry, and would
not detract from management motivation to further internationalise.

A significant screening data deficiency presently includes the absence

of product-specific data on most of the 20,000 service products – the high
growth area of international trade. All countries have only recently agreed on
the statistical categories and boundaries which they will use, but most have
not begun collecting that data for their country’s imports, exports and local
production. It will be about a decade from now before many countries have
collected several years of services product data and the United Nations have
cleaned it up to make it internationally comparable. Another variable with a
noticeable deficiency of data is cultural difference, where data from the two
doyen of empirical cultural difference research (Hofstede and Schwartz) have
only been published for some 60 countries. Finally, data of most kinds sought
for screening are not available for, or are many years late in becoming
available from, some 70 countries – mainly the smaller, less developed ones.

The screening decision reaches a crescendo during the screening

selection process step, when all the above variables are assembled, traded
off, and the preferred short-list of markets is produced.
 103

A motivated manager can overcome many hurdles when striving to

succeed in business. The proposed screening framework does not intend to
imply that specific hurdles such as resources, competition, cultural distance,
bureaucratic red tape, etc. cannot individually be overcome. Nor does it imply
that any particular matter dominates the decision – product-specific demand
perhaps excepted. However, accumulating the ease or difficulty encapsulated
in each of the variables is expected to result in a momentum that makes some
markets very much harder or easier to enter than others. Organisation
performance will then be retarded or advanced.

4.3 Underlying Theory

The first part of this section discusses the theoretical approaches which
underpin the overall framework, then the literature’s support for each variable
is examined.

Success in a domestic business requires operating in a complex, multi-

variable environment. Succeeding in international business involves
performing in an even more complicated environment. No single economic
theory presently explains international trade (e.g. Ball & McCulloch 1990, p.
95) and perhaps the international business environment will always be too
complex for a single theoretical approach to successfully handle it.

The proposed screening framework thus draws eclectically on multiple

theoretical bases in a contingency approach to short-list preferred markets.
Bilkey (1978), Leonidou and Katsikeas (1996, p. 517), Madsen (1994, p. 25),
Reid (1981, 1986), Sarkar and Cavusgil (1996, p. 826), Zou and Stan (1998),
and other international marketing performance researchers do likewise, as do
Cromley, Hempel and Hillyer (1993) in a domestic market selection.

The proposed screening framework is consistent with the Upsalla

School’s stages of internationalisation. It recognises market knowledge,
market commitment, commitment decisions and current activities (Johanson &
 104

Vahlne 1977) as influential components, and measures the amount of

gradually acquired and used knowledge about foreign activities to then assess
how difficult a market the organisation can approach and be likely to succeed
in.

The proposed framework is also consistent with the Reid (1981, p. 101)
view that:

“In spite of the evidence which suggests that foreign entry and
expansion behavior may be the product of complex interactions
between organisation and decision maker variables (my emphasis), few
researchers have attempted to investigate export expansion behaviour
bearing this in mind.”

The proposed framework takes account of both of Reid’s areas and

extends on to include interaction with both the macro and micro environments.

The proposed framework is consistent with the Bilkey and Tesar (1977)
(repeated in Bilkey 1978, p. 40) formulation of the stages of
internationalisation depicted in microeconomic style as a multiple regression
equation:

A = a + bE - cI + dF + eM

where A is the organisation’s export activity for the stage in question;

E is management’s expectations regarding the benefits of exporting
after it has been developed;
I is the inhibitors that management perceives to initiating exporting;
F is the facilitators (unsolicited orders, information, subsidies, etc.)
management perceives to initiating exporting;
M is the quality and dynamism of the organisation’s management plus
the organisation’s organisational characteristics that affect exporting.
 105

Small case letters are coefficients which tend to differ from one stage of the
internationalisation process to another. Causation of increased
internationalisation is from right to left then back from left to right in dynamic,
feedback loops as the organisation reaches higher stages of international
activity. This equation is reflected in the moderating variables in the proposed
screening framework.

Bilkey (1978, p. 44) states that this equation is consistent with

Marshallian theory of the organisation, and so with classical economic theory.

Aaby and Slater (1989) reviewed 55 studies of export performance

covering 9,000 cases during the 10 years ending 1988. Their synthesis of the
organisation controllable, causal criteria into a model (p. 9) led them to
conclude that organisation competencies, characteristics and strategy
determine performance. All sub-components of these aspects (except for
mode of entry and profit, which this dissertation elsewhere argues against
using in the screening situation for pragmatic reasons) are contained in the
proposed screening framework. Aaby and Slater specifically excluded all
external matters. They therefore: (1) do not integrate external matters such
as the economic, technological, legal, cultural and ecological milieus, nor
isolate the aspects related to customers, such as needs, segments, adoption
process and rates, nor fully consider competition, (2) do not consider the
organisation’s capability at research and the availability of market selection
information. These additional matters are included in the proposed screening
framework.

Leonidou and Katsikeas (1996) critiqued 11 different, accepted,

empirically developed models of the stages of export development. They
concluded that the key concepts identified22 could provide the nucleus for the
conceptual framework of future empirical models (p 542). All of these
concepts are included in the proposed screening framework although the two
categories of ‘facilitators and inhibitors’ and ‘stimuli and barriers’ have been
 106

dismantled and spread among the functional areas. For example, the
facilitative or inhibiting affects of ‘managerial characteristics’ have been
located in the proposed framework as ‘general competencies’,
’internationalisation competencies’ and ‘managerial values’, whilst
‘organisational resources’ are badged as ‘other resources’.

International market screening has noticeable commonality with several

areas of international marketing theory and practice. It overlaps:

• Theory about the causes of organisations’ successful international

performance. Many of the variables are the same (e.g. competition
and organisational capabilities) because screening of markets
affects the organisation’s subsequent performance,

• The theory of and variables used in determining preferable mode of

entry. Variables such as market risk levels and strategic
necessities affect both the screening decision and the mode of entry
decision. A number of authors argue the irrelevance of this
commonality, believing that mode of entry should not be determined
until after screening is complete (see section 3.5). However, if
screening is to eliminate the many unsuitable markets, it must
discard markets which are infeasible for rational mode of entry
reasons. This does not mean screening should eliminate markets
due to management preference,

• Stage of internationalisation reached by the organisation, which has

far reaching affects on how much information and experience the
organisation has accumulated, and so its capability at handling
more stressful and demanding competition, cultures, political
situations, etc., and so the markets which should be selected by
screening,

22
of facilitators and inhibitors, export information needs and acquisition, stimuli and barriers,
 107

• International marketing strategy, so that ill-fitting markets are not

recommended.

Theory used in the proposed framework is drawn eclectically from

economics, marketing, international marketing, psychology, management,
finance and sociology. Some matters are considered by several disciplines,
for example, customer needs by economics, marketing, international
marketing, management and psychology. Economics provides the largest
number of relevant theories whilst psychology is the second most frequent
underpinning.

Psychology contains some six main philosophies; there is acrimonious

debate among proponents of each about concepts and treatment. The
presently dominant school of cognitive social psychologists (Pervin 1993,
p. 425) sees the individual’s world in very similar ways to the proposed
screening framework (e.g. Pervin 1993 chapters 12 and 13). They
concentrate on the complex interaction between people and their environment
(reciprocal determinism) – reflected in the proposed framework’s
consideration of the organisation, the potential customers, and their host
environments. The cognitive social theory elements of ‘competencies’,
‘goals’, and ‘self-efficacy’ are visible in the proposed framework as
‘competencies’ (general and international), ‘objectives’, and ‘achievable
objectives’. Goals and standards chosen imply prior risk assessment, another
item in the proposed framework. The cognitive theorists believe that learning
is prompted by ‘observational learning’ and ‘self-regulation’; the proposed
framework attributes learning through the organisation’s post-screening
market entry providing its employees with experience and feedback about
performance against their individualised, pre-specified, perceived-achievable
objectives. Cognitive social theorists also emphasise premeditated thought,
proactivity, standards for the goals, choice amongst goals, competencies
being situation specific, self-perception of competency affecting goal selection

foreign market selection, entry and expansion, and marketing strategy.

 108

and behaviour, and self-reports of self-efficacy. All these concepts and

components are consistent with the proposed screening framework.

An earlier name for cognitive social psychologists was learning theory

psychologists. The previously mentioned Johanson and Vahlne experiential
approach is also a learning theory / cognitive social approach.

Two other psychological concepts from organisational behaviour also

underlie the framework: performance and motivation. Performance is a
function of ability, motivation and environmental conditions (Bartol et al. 1998,
p. 491). The screening framework labels these as organisational abilities,
management motivation and commitment to further internationalise and the
attributes of each market.

The Porter and Lawler (1968, p. 165) motivational model is also

relevant and reproduced below as Figure 4.4. Its components have the
following equivalents in the screening framework:

1) ‘Value of reward’ is the organisation’s assessment of the values

(utilities) of each of its screening objectives (the value of sales,
sales growth, etc.);

2) ‘Perceived effort leading to reward probability’ arises from the

framework’s objectives being expected and achievable after
management assesses the expected effort necessary and
rewards likely (sales, sales growth, personal benefits, other
business benefits and risk levels);

3) ‘Effort’ is the quantity, quality and duration of mental and

physical resources applied to achieve market entry then reach
break-even profitability;
 109

4) ‘Abilities and traits’ are the organisation’s abilities and values

clusters (internationalisation competencies, personal values,
etc.);

5) ‘Role perceptions’ is the manager’s value of the importance of

each of the various internationalisation duties of organisation
members;

6) ‘Performance (accomplishment)’ is achievement of the eventual

market entry;

7) ‘Intrinsic rewards’ are the psychic components of the personal

benefits set as objectives; ‘Extrinsic rewards’ are the other
objectives, including any material personal benefits, set as
objectives;

8) ‘Perceived equitable rewards’ is management’s perception that

the objectives set are a fair and equitable reward for whatever
effort will be expended in achieving market entry;

9) ‘Satisfaction’ in the future is determined by the degree of fit

between the manager’s values and the other variables in the
framework.

Having the Porter and Lawler model underlying the screening

framework means that screening outcomes are likely to contribute toward
positive organisational motivation, and to not detract from employee
performance.
 110
Figure 4.4
The Porter-Lawler Motivation Model

1 4 8
Value of reward Abilities & traits Perceived equitable
rewards

7A
Intrinsic rewards
3 6
Effort Performance 9
(Accomplishment) Satisfaction

7B
Extrinsic rewards

2 5
Perceived effort Æ Role
reward probability perceptions

Porter, Lyman D. & Lawler, Edward E. (1968) Managerial Attitudes & Performance, Irwin, Homewood, IL p. 165
 111

The following arguments from the literature are argued to have the
explanatory power which justifies selecting specific variables for the proposed
screening framework. It is the complex interaction of the variables, many with
feedback loops, that determines firm performance (Madsen 1989, p. 43), not a
simple, mechanical relationship.

4.3.1 Dependent Variable

Multivariate statistical techniques manipulate data on the moderating

and independent variables to produce the dependent variable – the
organisation-specific short-list of attractive and preferred markets. Due to the
cost and impracticality of not using data originating from national statistical
agencies, markets are here substituted with countries. There are 261 nations
and dependencies (CIA World Factbook 1996, p. ix), of which some 230 are
nations and 185 are members of the United Nations. UN members generally
comprise most of the world’s activity (however measured), although the
important non-members of Hong Kong, Switzerland, Taiwan, and the former
Yugoslavia usually need adding (and were in this research). Complete data
(other than product-specific data) were collected on 220 of the 230 countries.
Product-specific data then reduced that number to between 42 (petrol
engines) and 205 (beef) countries which produced and/or consumed the
product. See section 5.4 for more.

4.3.2 Independent Variables

4.3.2.1 Objectives

The economist’s rational economic person is supposed to make

decisions which maximise a particular variable or variables (e.g. profit,
revenue or utility). Bilkey and Tesar (1977) found that a firm’s stage of
internationalising was influenced by its expected advantage from
 112

internationalising. Organisations have different motives at different stages of

internationalisation (Bilkey & Tesar 1977; Czinkota 1982; Root 1987).

The cost/benefit assessment of whether further internationalisation is

likely to be worthwhile to the firm can be expected to affect its level of
motivation and commitment to further internationalise, including whether or
not it decides to endure the pains and pleasures of dealing with new cultures.
Figure 4.5 suggests that a more informed view about costs and benefits of
internationalising will be associated with higher levels of specific market
knowledge, higher levels of general internationalisation knowledge, an open,
flexible home culture programming (enculturation) and low psychic distance.
A high level of commitment to internationalisation will also cause the
organisation to minimise perceived disadvantages and/or make attempts to
overcome the difficulties. The stage of internationalisation has been shown to
predict what organisations see as advantages and disadvantages of
internationalising and to affect costs and benefits perceived (see, for example,
Aaby & Slater 1989, p. 17; Leonidou & Katsikeas 1996, p. 537). Psychic
distance is known to affect the optimal market selection (Papadopoulos &
Jansen 1994, pp. 38-39). The psychological processes of rationalisation and
cognitive dissonance presumably operate to justify whatever attitude is taken
toward the costs and benefits of further internationalising the firm.

Managements have multiple objectives (see, for example, Cavusgil &

Zou 1994; Diamantopoulos & Schlegelmilch 1994, p. 162; Evangelista 1994,
p. 209; Green & Allaway 1985; Huff & Sherr 1967; Johnson & Scholes 1993,
p. 157; Madsen 1989, p. 41; Reid 1981, p. 104; Simon 1997). The world is a
complex place and it is unsurprising that numerous dimensions are necessary
to capture and assess the range and complexity of managerial aspirations.
The six areas which have been chosen result from an extensive literature
review to provide a full range of options. Weighting of each alternative then
provides for individual preferences. Note that motives for increasing the
organisation’s level of internationalisation are not included unless they help
discriminate between markets (e.g. reduce dependence on local market, or
 113

escape a shrinking or low growth market are arguments for increasing

international sales but do not help screening).

Objectives at the screening stage are expected – they are not finally
locked in by management until after in-depth research confirms their
attainability. Objectives must be perceived to be achievable (Lawler 1973;
Locke 1968; Porter & Lawler 1968; Vroom 1964). Achievability is a key part of
expectance theory (ibid.) which is a cognitive psychological approach that
derives from the choice behaviour and utility concepts of economics. It was
mentioned in the introductory pages of this dissertation’s theory section and is
fundamental to the screening framework’s components and relationships.

Each of the six objectives will now be discussed.

4.3.2.1.1 Sales and Sales Growth.

Bounded rationality theory (e.g. Simon 1947, 1997) says that, subject
to the moderating constraints, an organisation will be likely to wish to
maximise its sales so as to maximise its profit. The mechanics of accounting
determine that, in an organisation which trades or produces a good or service,
sales are a necessary but not sufficient pre-condition to profit. Whilst it would
be preferable for screening to proceed beyond estimating sales to estimating
profit, the estimating by market of profits would require collecting very much
more data than to estimate sales. Additionally, profit estimates are highly
dependent on the accuracy of the sales estimate, and accuracy of the sales
estimate is likely to be much lower at the screening stage than at the in-depth
stage. The considerable additional work would thus return information of
dubious accuracy.

In the absence of profit forecasts, organisations can generally be

expected to seek markets likely to provide both sales and sales growth.
Markets that are growing are normally increasingly profitable (because of
increasing economies of scale and scope). Markets are more desirable if they
have both volume and growth, and generally, the higher the volume and
 114

growth, the better. That presents a problem: How does one determine
growth? An absolute measurement of dollars or units increased/decreased
tends to overstate the size of the change in large markets and tends to
understate the size of the change in small markets. Conversely, percentage
change tends to overstate the size of the change in the smaller markets, and
understate the size of the change in larger markets. Net shift technique (also
called ‘shift-share’) (Green & Allaway 1985; Huff & Sherr 1967) satisfactorily
avoids the distortions inherent in both methods. It measures the relative gains
or losses by individual market areas compared with the total market’s growth,
so providing a perspective of growth that both of the other two methods lack.
The net shift technique eliminates:

1) low-growth, large-sized markets (suggesting mature or saturated

market conditions, and likely existing supplier-purchaser buyer
relations, all of which are associated with higher levels of
competition), and
2) high-growth but small-sized markets (suggesting low total
volume).

4.3.2.1.2 Personal Benefits

Personal benefits may be extrinsic or intrinsic (e.g. Bartol et al. 1998, p.

501; Luthans 1989, p. 249). Extrinsic rewards such as bonuses and
promotion are familiar mainstream ideas to most managers. Less familiar are
the intrinsic rewards, which are generated by the interaction between task and
the individual’s psychological makeup.

Beginning with the Hawthorn experiments (Roethlisberger & Dickson

1939), managers discovered that psychological benefits, not money alone,
had an important role in motivating employees. Job performance involves
transacting with other people, so potentially satisfying the psychological need
for affiliation; work on particular tasks may provide experience which makes
the person promotable; tasks may allow the employee to self-actualise; and
so on. Unintended personal benefits thus came to be seen as having a role in
 115

motivating employees. A further step is to accept that, at times, some of an

organisation’s profits intentionally get traded off for non-economic rewards
(personal goals). This is accepted by the behavioural and human relations
schools of management and by bounded rationality theory (e.g. Barney &
Griffin 1992, pp. 59-60; Simon 1997, p. 158). The framework does not
presently operationalise this dimension because country data are yet to
become available.

4.3.2.1.3 Other Business Benefits

Other business benefits captures motivators that discriminate between

markets that are not included elsewhere in the framework. Sample citations
for each include: increase market share (this is a common reformulation of
increased sales or growth); increase corporate image (-); increase
organisational political clout (Papadopoulos & Denis 1988); dispose of excess
capacity (Bradley 1995, p. 263); dispose of products no longer attractive in
other markets (Bradley 1995, p. 263); derive learning (production, marketing,
international business, etc.) (Albaum et al. 1994, p. 18); reduce counter
cyclicality (ibid, p. 35); and obtain resources or technology (ibid, p. 23).
Different markets will have differing abilities to meet these objectives. Note
that only organisational objectives are included (i.e. personal objectives are
located in the personal benefits section above) and that strategic objectives
are captured as a moderating variable (in strategic values). This variable was
not operationalised by synthesising data for the organisations, but could have
been. The Management questionnaire (Appendix A) determines whether any
of the other business benefits matters apply to the real life firm, in which case
a market-discriminating variable is tailored to suit the screening.
 116

4.3.2.1.4 Risk level chosen

Cognitive social psychologists see perceived achievability of goals as

highly influential on goals and standards selection (Pervin 1993). Perceived
risk levels influence goal desirability in two ways: the absolute likelihood of
achieving the goal and the degree to which the person is dependent upon
factors beyond their control. Bayesian theory allows us to adjust expected
returns against our objectives with a risk probability that reduces the average
expected payoff (Kinnear et al. 1993, pp. 94-99). Mathematics’ probability
theory also allows risk adjustments (e.g. Kinnear et al. 1993, pp. 99-100).
Finance theory derived from economics the capital assets pricing model23,
which similarly calculates expected payoff after considering diversifiable
unique risk and undiversifiable market risk (Brealey & Myers 1984, p. 129).
Johanson and Vahlne (1977, p. 30) see the organisation’s maximum tolerable
risk limiting the existing market risk situation. However, bounded rationality
research by Simon 1997 clearly indicates that people make global judgments,
not detailed calculations, about probable outcomes.

4.3.2.1.5 Lowest Effort

Effort is “a physical or mental exertion; an applied force acting against

inertia; a determined attempt” (Collins dictionary 1979, p. 468). Psychologists
see effort as a matter which determines the outcome of goal directed
behaviour (Pervin 1993, p. 390). Porter and Lawler (1968, p. 165) developed
a nine component model of motivation that involves ‘effort’ as a key causative
variable. Porter and Lawler’s diagram, and the equivalents between their
terms and those used in this dissertation, have already been covered in
section 4.3. Examination of the Porter and Lawler term ‘effort’ reveals that it
comprises both the motivation and the effortful behaviour involved with task

23
Capital Asset Pricing Model attributed to Jack Treynor, William Sharpe, and John Lintner.
See Sharpe, W. F. (1964) Capital Asset Prices: A Theory of Market Equilibrium Under
Conditions of Risk, Journal of Finance 19 (September) pp 425-442, and Lintner, J. (1965) The
Valuation of Risk Assets and the Selection of Risky Investments in Stock Portfolios and
Capital Budgets, Review of Economics and Statistics, 47 (February) pp 13-37. Treynor’s
article was not published.
 117

accomplishment. This dissertation separates the two into ‘effort’ (discussed

here) and ‘motivation’ (discussed elsewhere).

Porter and Lawler describe ‘effort’ as “... the amount of energy an

individual expends in a given situation” (p 21). They say the energy expended
involves motivation and physical and mental effortful behaviours (pp 21-22),
moderated or enhanced by abilities and traits (p 23) and role perceptions (p
24) (see Figure 4.4). Thus, effort expended during entry into a new market is
a function of the organisation’s motivation and its effortful behaviour,
moderated or enhanced by the organisation’s capabilities and preferences
and the market’s attributes. Market attributes are added here because they
are a situational variable which Porter and Lawler did not need to include for
their research situation. Motivation, abilities and traits are treated elsewhere
in the dissertation whilst role perceptions largely are constant for each market.
‘Effort’ should thus be measured by the quantity, quality and duration of both
physical and mental resources voluntarily applied to achieve the task of
market entry and reach break-even profitability.

Whilst ‘effort’ and ‘resources’ are similar in meaning, effort is preferred

because resources are not normally thought of as including the quantity and
quality of mental energy consumed by a task. As in engineering, where
efficiency is calculated by dividing outputs by inputs, market entry efficiency
can be measured by dividing the sum of intrinsic and extrinsic rewards by the
effort expended.

4.3.2.2 Targets

Targets get chosen by developing measures of segment attractiveness

then selecting the target segments (e.g. Dibb & Simkin 1996, p. 15; Kotler et
al. 1994, p. 121). As segmentation data for each market are usually
incomplete, screening’s targeting usually aims at similar segments to those
targeted in the home or another market. Not only is this necessary but it is
rational. Just as network theorists predict that internationally inexperienced
organisations will optimise performance by targeting nearby markets having
 118

similar conditions to their domestic markets (e.g. Madsen, 1994, p. 37), it can
be expected that inexperienced organisations will be more likely to wish to,
and be wise to, target apparently similar segments to those with which they
have experience.

4.3.2.3 Strategy Decisions

The strategy variable needs to be divided into two parts. Decisions

about strategic options are made by the organisation’s manager – and
controllable variables are independent variables. On the other hand, the
manager’s strategic values are psychological components of personality – and
values are very difficult to change in the short term because of their
interconnectedness with each value sub-component. Strategy decisions is
thus an independent variable whilst strategic values is one of the moderating
variables.

Sarkar and Cavusgil (1996, p. 835) believe the arguments in favour of

incorporating global strategic considerations for explaining international
performance are compelling, and cite Harrigan (1985), Kogut and Singh
(1988) and Porter (1985) as examples of other researchers who contend
likewise. Madsen (1989, p. 54) found that export policy (essentially strategy
plus marketing mix) had great affect upon export sales and profit results,
either directly or after including interaction effects. Cavusgil and Zou (1994 p.
1) likewise found strategy formulation and execution (essentially strategy plus
marketing mix) to be a key determinant of export performance.

Strategy development for a given product in an additional market

involves the following choices: between product standardisation and
adaptation; whether the organisation will compete on cost, differentiation or
focus; whether to collaborate (directly or indirectly) or compete (by defence,
offence, flank or niche via pre-emption, confrontation or build up); whether
they undertake country concentration or diversification, and whether there will
be incremental or simultaneous market entries (e.g. Douglas & Craig 1995;
Kotler et al. 1994, Sarkar & Cavusgil 1996).
 119

4.3.2.4 Size of Assessment Task

Whether or not market research should be conducted depends on: (1)

there being sufficient time available, (2) having access to sufficient data that
are accurate, timely, complete and comparable, (3) for the decision to be of
sufficient strategic or tactical importance, and (4) the cost/benefit of research
justifying the task. If all four of these matters are not favourable, research is
not justified (Zikmund 1994, p. 19). A limiting constraint has been the
absence of an effective screening methodology that ensures that useful
results are produced to justify the cost of comprehensive screening. If this
dissertation’s research is confirmed, it can then be argued that Zikmund’s four
matters are all answered favourably.

Hypothetically, an organisation could undertake partial screening (i.e.

eliminate some of either the 46 decision variables and/or the 230 markets).
However, as argued from section 2.4 and on, that would be illogical and be
likely to produce sub-optimal results. The framework therefore only permits
either screening to be undertaken, or for it to not be undertaken.
Mathematical treatment of this item is thus either with as zero (do not screen)
or one (undertake screening). Intermediate values are treated as 0, when
screening decision-processing ceases.

4.3.2.5 Criteria and Method of Evaluation

The criteria and method of evaluation of the desirable markets

comprise the following matters:

• The criteria which are considered. The rationales for each of these
are argued individually throughout the pages of this section and
comprise the following:
∗ objectives (product-specific sales, product-specific sales
growth, personal benefits, other benefits, risk and effort)
 120

∗ customer characteristics
∗ competition
∗ other aspects of country attractiveness, for example, political
stability
∗ organisation’s capabilities
∗ management preferences
∗ management motivation
∗ data availability and quality
∗ strategic options.
It has been separately argued that these should be assessed in all
markets and that mode of entry be determined during in-depth
assessment;

• Criteria weights. These will be assessed later in this research by

statistical manipulation of the data;

• What interactions are recognised and considered. For example,

whether there will be synergy between existing markets and the
newly entered markets. This will not be assessed during this
exploratory research;

• Selection rules. Papadopoulos and Denis (1988, p. 40) use the

categories ‘quantitative’ and ‘qualitative’, although all the 31
screening methodologies they reviewed were categorised as
quantitative. Qualitative methods can be ‘systematic’, but are non-
statistical, they believe. Quantitative methods are systematic by
definition. When discussing organisational decisions in general,
Gray and Starke (1988, p. 297) use the categories of ‘rational’ or
‘irrational’. Psychologists would challenge the certainty of those
categories, due to the variability of individual perception and values.
Gray and Starke also categorise decision-making into: (1) the
rational economic model, (2) the administrative model, (3) the
political model and (4) the garbage can model (p 303). Aguilar-
 121

Manjarrez, Thwaites and Maule (1997, p. 13) use a hybrid scheme

of ‘rational’, ‘bounded rational’, ‘political’ and ‘garbage can’ for their
research into sports sponsorship decision-making. Simon (e.g.
1947, 1997) expounded the notion of bounded rationality, the
practical constraints on the ability of humans to make rational
economic decisions. These constraints are information
incompleteness, human selfishness, emotion, human computational
ability and differences in goals between individuals and their
organisation. In an interesting application of decision-making rules
to the obverse side of international market selection, international
supplier selection, Liang and Stump (1996) use rationality, bounded
rationality, satisficing, heuristics and compensatory decision
processes (p 788).

Applying the literature to market screening, it was decided to use

the following scheme:
∗ unsystematic rules:
- intuition
- opportunistic (e.g. unsolicited foreign order)
- irrationality
- haphazard
These approaches were rejected as less effective than the
systematic, mathematical approaches because of the
number of variables involved and the complexity of trading
them off against each other;
∗ systematic selection rules:
- mathematical methods
~ market grouping, for example, cluster analysis
~ market estimation: multi factor, econometric, shift-
share
~ Bayesian (probability multiplied by payoff)
- whether compensatory or non-compensatory rules are
used. If non-compensatory, rules can be:
 122

~ disjunctive
~ conjunctive
~ lexicographic, or
~ sequential elimination
(Gray & Starke 1988, p. 301; Loudon & Bitta 1993, p. 521)
Considering the alternative market selection methods,
systematic selection using mathematical methods is logically
the soundest approach for optimising the organisation’s
objectives because of the size and complexity of the
selection task. Shift-share is used to assess size and growth
of market sales for reasons argued in section 4.3.2.1.1.
There is probably justification for some variables to be non-
compensatory (e.g. extreme risks above a specified level
justifying rejection of that market, irrespective of positive
scores for other variables, when screening markets for a firm
with low resources). However, the early stage of
development of the approach being presently researched
means there is insufficient information to determine the
thresholds, so compensatory rules have been used;
∗ mixed rules:
- heuristics (using a logical ad hoc method developed from a
mixture of experience and intuition. However, experience
is known to include incomplete knowledge).
This approach was rejected because one of the reasons for
carrying out the present research is that practitioners are
currently using ineffective approaches to screen markets.
 123

4.3.3 Intervening Variables

4.3.3.1 Management motivation and commitment to further

internationalise

Arguments can be mounted from the international marketing,

marketing, psychological and management literatures that management’s
motivation and commitment to further internationalise affects management
and organisation effort, then organisation performance.

Evangelist (1994, p. 208), Johanson and Vahlne (1977), Styles and

Ambler (1997, p. 6) and others suggest that the level of management
motivation affect the organisation’s international performance. The further
belief is that motivation and effort are related, with a high motivation being
commonly associated with high effort and low motivation with low effort (e.g.
Craig-Lees, Joy & Browne 1995, p. 363; Evangelista 1994, p. 208; Kolasa
1969, p. 230). In an international marketing situation, effort involves
management allocation of human, financial and other resources to achieve its
goals (e.g. Bodur 1994; Bradley 1995, pp. 114-118; Madsen 1994, p. 33).
Goals have been detailed in the independent variables section. The way the
manager and organisation should behave when expending their effort is the
art and science of international marketing and the subject of ongoing research
activity.

Performance is a function of ability, motivation and environmental

conditions, says the general model of motivation (Bartol et al. 1998, p. 491).
Motivation can be described as the driving force within individuals that may
impel them to action (Schiffman & Kanuk 1991, p. 69). It is invisible, and
cannot be measured directly (Bartol et al. 1998, p. 490). Motivation includes
all those inner striving conditions described as wishes, desires, drives, etc.
(Kolasa 1969, p. 249). Depending on the intensity of the need and the
circumstances, the individual may (or may not) behave in a way that meets
the need. Explanation of the interconnecting links between need, behaviour
and need fulfilment varies with the psychological theorist.
 124

Porter and Lawler (1968) see motivation as a function of both the

expected value of rewards and the perceived effort-performance-reward
probabilities (pp 21-22). This dissertation sees motivation as a function of the
value attached to each of the objectives and the risks involved. Assumed, but
not calculated by the screening framework, is the organisation’s perception of
the achievability of the effort-performance-outputs chain – because the
organisation will not proceed from market screening to in-depth assessment if
this is not so.

If motivation is the strength of desire for goal attainment, commitment

is the endurance aspect of the same desire. Commitment appears less
frequently in psychological and marketing literature, but is a more commonly
encountered word in academic and practical international marketing. Madsen
(1994, p. 33) believes commitment is almost a necessary condition for
successful international ventures because of the long pay-back periods. The
large Aaby and Slater (1989, p. 17) review of 1978-1988 export research
concluded that there is a positive relationship between management
commitment and propensity to export.

This dissertation uses cognitive, affective and behavioural indicators to

assess motivation and commitment to further internationalise, as do Axinn and
Athaide (1991), and Evangelista (1994, p. 208) to assess ‘commitment’.
‘Motivation and commitment’ here are similar to the Johanson and Vahlne
(1977, p. 26) item ‘market commitment’, which they argue is one of the four
ingredients that determine the stage of internationalisation reached by the
firm. Johanson and Vahlne see the item indicating the amount of resources
committed and the degree of commitment (the difficulty of finding an
alternative use for the resources and transferring them to it) (p 27). This
dissertation suggests assessing not only the present level of commitment but
also the force behind the commitment, the motivation. That alteration
changes the variable from a historic indicator to one that helps to roughly
predict the future success of the organisation (at screening, in-depth
assessment, and new market entry).
 125

Commitment to internationalisation needs distinguishing from

commitment to a particular market – the former is more relevant to screening.
In contrast to the Johanson and Vahlne view of market commitment, above,
commitment to internationalisation is a more general concept that predisposes
organisations to begin and continue along the internationalisation path. It
comprises attitudinal and behavioural components (Axinn & Athaide 1991).
Which market is preferred is then determined by a mix of rationality and
attitude – demand, the cost/benefits of particular markets, psychic distance,
existing commitments, resources, etc.. Commitment to internationalisation is
influenced by the four factors (Figure 4.5) of specific market knowledge,
general internationalisation knowledge, cultural distance and the cost/benefit
assessment of further internationalising. Commitment to a specific market
could overwhelm the organisation’s desire to further internationalise through
focusing the organisation’s activity on that market (a market concentration
strategy). Experience with that market will feed into specific market
knowledge, general international knowledge and cultural distance.
Commitment to the idea of further internationalisation is a key screening
matter. Aaby and Slater (1989, p. 17), Leonidou and Katsikeas (1996, p.
535), Reid (1981, p. 105) and others see this as a necessary ingredient that
can retard or halt the process of finding and entering additional markets, and
can influence the speed at which the organisation internationalises. Cavusgil
and Zou (1994, p. 1) found commitment to be a key determinant of export
performance (although their ‘commitment’ can be interpreted as being both
the ‘motivation’ and the ‘commitment’ used in the present research).
 126

4.3.3.2 Marketing mix

Marketing mix is a fundamental concept which appears in domestic

(e.g. Kotler et al. 1998, p. 339; McCarthy et al. 1994, p. 22) and international
(e.g. Douglas & Craig 1995, p. 213; Paliwoda 1993, p. 17) marketing thinking.
Madsen (1984 p. 54) and Cavusgil & Zou (1994 p 7, 14) are examples of
researchers who have empirically found links between marketing mix and
performance in the international setting.

4.3.3.3 Customer Adoptions and Usage Rate

The customer adoptions and usage rate variable is conceptually

necessary to reduce total segment demand to the portion gained by the
organisation. Different psychological models of the adoption process exist
and have been appropriated by marketing and international marketing from
consumer behaviour (Kinnear et al. 1993, p. 179; Schiffman & Kanuk 1991,
pp. 532-44). The hierarchy of effects model has been selected because its six
stages can be aligned to the psychological processes of cognition, affect and
behaviour; these terms have explanatory power when understanding related
matters such as customer needs, promotional techniques, managements’
personal values, etc..

4.3.3.4 Organisation Product-specific Demand

Organisation’s product-specific demand is a calculated figure –

adjusted segment demand in each market can be multiplied by customer
adoption and usage rates. It was not calculated in this research because
adoption rates for each product were not known.

4.3.3.5 Selection Process

This dissertation follows Russow (1989) to become the second

research to empirically select international markets from among all countries
possible. Papadopoulos and Denis (1988) summarise the history of
 127

demonstrating the use of the statistical techniques of principal components

analysis, cluster analysis and multiple discriminant analysis for international
market screening. Among the notable examples that they and other authors
cite for suggesting some or all three of the techniques are Goodnow and
Hansz (1972), Liander, Terpstra, Yoshino and Sherbini (1967), Litvak and
Banting (1973), Papadopoulos (1983), Russow (1989), Sethi (1971), Sherbini
(1967) and Sheth and Lutz (1973). Statistical writers who are quoted as
sources of guidelines for using these techniques include Green (1978), Hair et
al. (1984), Jackson (1973), Sheth (1971) and Tabachnick and Fidell (1983).
Hair et al., (1995) and Tabachnick and Fidell (1996) are modern equivalents.

As cogently put by Russow (1989, pp. 60, 69), there are six objectives
of the principal components analysis stage and four objectives of the cluster
analysis stage.

Principal components analysis objectives are:

1) Treatment of multicollinearity,
2) Establishing the factorability of the original data set,
3) Ascertaining the adequacy of the rotated solution,
4) Determining how many of the dimensions to retain,
5) Identifying and interpreting the dimensions,
6) Comparing the factor solutions across product samples to
determine the degree of similarity.

Cluster analysis objectives are:

1) Determining the existence of country groups that are based on

internal similarities of the objects (the countries),
2) Examining and describing these clusters,
3) Evaluating differences across the product samples,
4) Testing the findings of the analysis.
 128

Cluster analysis requires human involvement to interpret the number of

clusters, so multiple discriminant analysis is used as a partial test of the
robustness of the statistical differences between countries suggested by
cluster analysis.

4.3.4 Moderating Variables

This proposed framework clusters elements of market attractiveness

into aspects related to customers, competition and substitutes, other
environmental matters, and risk levels.

4.3.4.1 Each Market’s Attributes:

Market attractiveness is a recurring theme in international marketing

literature. The Papadopoulos and Denis (1988, p. 40) taxonomy (Figure 1)
lists 12 ‘market grouping methods’ that screen foreign markets to assess
broad environmental aspects (such as political and economic milieus) then
indicate the ‘best’ market. Criteria considered important vary by author, but
none include the markets’ product potentials. The taxonomy also lists 19
approaches termed ‘market estimation methods’ that instead assess product
potential, and which commonly also assess the broader environmental
aspects of market attractiveness. This dissertation concentrates heavily on
product-specific demand and so is a market estimation method. It uses
multiple indices, both micro and macro. Psychology views attractiveness as a
motivator, and degree of attractiveness as a discriminator when selecting
among goals. This dissertation does likewise.
 129

4.3.4.1.1 Potential Customer Aspects

4.3.4.1.1.1 Customer Needs

Management theorists argue that meeting customer needs is a

prerequisite to making a profit (e.g. Koontz & O’Donnell 1968, p. 112).
Marketing and international marketing have adopted psychology’s view that
customers have differing needs, some of which the customers attempt to
satisfy through their purchases (e.g. Dibb & Simkin 1996, p. 58; Douglas &
Craig 1995, p. 196; Kotler et al. 1994, pp. 167-9). Alderfer (1972), Maslow
(1965) and Schwartz (1992) have already been cited as typical of the many
needs theorists. Degree of internationalisation of the market (Madsen 1994,
pp. 34, 38) and the international product life cycle (Vernon 1966; Wells 1968)
affect the kinds of needs to be satisfied. However, practical use of these two
concepts is limited because of the difficulty of predicting what stage the
product has reached (e.g. Kotler 1997, p. 362) and the need for management
to do so for each market without knowledge of each market. An actionable
approach is Means Ends Chains (MECs) (Newell & Simon 1972; Reynolds,
Gengler & Howard 1995; ter Hofstede, Steenkamp & Wedel 1999). This
connects customer values through customer benefits to the product’s
attributes in a chain, identifying segments and the probabilities associated
with each link in the chain.

4.3.4.1.1.2 Segment Population

Segment population size is a necessary multiplier for calculation of the

size of demand in each market (e.g. Bradley 1995, p. 131; Conners 1960;
Terpstra & Sarathy 1994, p. 69).

4.3.4.1.1.3 Segments

Marketing and international marketing both advise that customers with

similar product needs should be aggregated into segments for the
organisation to obtain economies of scale and scope. Segmentation requires
 130

identification of appropriate variables to use in each market, then developing

profiles of the resulting clusters of customers (Dibb & Simkin 1996, p. 15; Jain
1996, p. 380; Kotler et al. 1994, p. 121; Root 1994, p. 56). Consumer
segments can be differentiated on geographic, demographic, psychographic
and behavioural lines, whilst industrial segments can be drawn up using
demographic features, operating variables, purchasing approaches,
situational factors and personal characteristics (e.g. Douglas & Craig 1995, p.
196; Kotler et al. 1994, p. 124). Each segment may adopt the new product or
supplier (move through the steps of the hierarchy of effects) at a different
pace, producing a different adoption rate. Marketing and international
marketing theorists also advise consideration of the macro environmental
influences on customer segments (economic, political, legal, etc. aspects)
(e.g. Douglas & Craig 1995, pp. 11-18; Kotler et al. 1994, p. 44).
Papadopoulos and Denis (1988, p. 40) succeeded at categorising the then
existing screening methodologies into two kinds: those which used macro
segmentation criteria to assess the general market environment, and/or those
which used micro segmentation criteria to assess product demand factors.
The framework proposed in this dissertation takes account of both macro and
micro segmentation influences.

In some cases, segments do flow across national borders and cultural

groupings (e.g. Hassan & Samli 1994, p. 77). However, segmentation is
driven by customer needs, which can vary in each market, yet research on the
needs in each market is rarely available. The organisation also has discretion
about where to draw segment boundaries. In the absence of knowledge
about each market’s customer needs, the boundaries are often assumed to be
the same as in another of the organisation’s markets. Where screening is a
case of old-product-new-to-the-market, the organisation will have more
experience to draw on than when it is a new-product-new-to-the-market. If the
segments are wrong, the subsequent short-list of markets will be wrong and
screening will have failed to recommend the optimum foreign markets. This is
a conceptual weakness of screening. Customer needs and segment
boundaries can both be confirmed during the in-depth market assessment
 131

stage, but this runs the risk of not including all optimum markets in the
screened set per selection error.

4.3.4.1.1.4 Intrinsic Adoption Rate

Intrinsic adoption rate, by market, quantifies the estimated product

segment sales level and is a marketing concept derived from psychology
(Craig-Lees, Joy & Browne 1995, p. 187; Kotler 1994, p. 197; Schiffman &
Kanuk 1991, p. 529). It depends upon the product’s relative advantage,
compatibility, complexity, trialability and communicability to the new customer
(Kotler et al. 1994, p. 197; Stanton 1994, p. 198). Each segment in each
market may have its own adoption rate.

4.3.4.1.1.5 Segment Demand

Segment demand in each market is demand in the segment (see

4.3.4.1.1.3 above).

Where segment demand is unavailable, industry demand is typically

substituted. Marketing and international marketing see this as total market
demand for the product form, some of which the organisation succeeds at
attracting to itself (e.g. Bradley 1991, p. 216; Douglas & Craig 1995, pp. 194-
6; Kotler et al. 1994, p. 241). Economic theory tells us that a market’s
demand is affected by its absolute and comparative advantages (Adam Smith
and David Ricardo), the amount and cost of factors of production (Heckscher-
Ohlin and MacDougall), overlapping demand (Lintner), technological gap
(Kravis, Haufbauer & Douglass), the international product life cycle (Vernon
1966; Wells 1968), follow the leader (Knickerbocker), immiserizing growth
(Bhagwati, Johnson & Alejandro) and the Prebisch import substitution theory
(e.g. see Asheghian & Ebrahimi 1990, p. 27 or Ball & McCulloch 1990, p. 95
for succinct elaboration of most of these theories). Economics does not yet
claim to have a complete explanation for demand causation.
 132

Growth in demand in each market. Madsen (1989, p. 57), Root (1994,

p. 63) and Russow (1989, p. 177; 1996, p. 57) are some of the many authors
who advocate product-specific market size growth as one of the important
market selection criteria. Russow’s research is particularly significant
because it is the first empirical test using this as one of several predictive
variables in international market screening. His work affirmed that growth is
indeed important.

When measuring growth, the question arises about which calculation

method to use. A number of benefits are claimed for the shift-share method
(see section 4.3.2.1.1). Huff and Sherr (1967) provide the original shift-share
method of calculating growth when screening domestic markets, whilst Green
and Allaway (1985) advocated its use internationally. Medina, Saegert and
Merrifield (1996) demonstrate several interesting variations to the shift-share
methodology when screening international markets for a new, novel product.

Both new and existing markets should be included in the screening

assessment. Equity theory (Adams 1965; Daft 1991; Locke 1976; Mowday
1987) says that we perceive the ratio of our inputs and outcomes to be equal
to the ratio of inputs and outcomes for a comparison other. Comparison is
relative, not absolute. The organisation’s comparison base in new market
selection is their present markets, domestic and foreign. Comparison also
sometimes includes competitors who are more successful but whose
performance the organisation assesses it is able to equal. Competitors are
included in this screening framework through the Competition variable,
industry sub-variable.
 133

4.3.4.1.2 Competition and Substitutes

Competition and substitutes in each market is discussed by Porter

(1980, 1990) because of its importance and affect on outcomes, but he
believes there is no theory to fully explain national competitiveness. He
describes (1980, p. 5) competition arising from the five forces of industry
competitors, substitutes, suppliers, buyers and potential new entrants.
Marketing and international marketing theorists consider competition and
substitutes to be major influences on the organisation, its behaviour and the
resulting sales levels (e.g. Bradley 1991; Douglas & Craig 1995; Kotler et al.
1994). Degree of monopolistic advantage and market imperfections (Hymer &
Caves), and whether competition is loose or rigorous (Bradley 1995, p. 270)
affect ease of market entry of a newcomer. The network approach argues
that stage of internationalisation of the market influences the intensity and
level of sophistication of the competition (Johansson & Mattsson 1986;
Madsen 1994, p. 37).

4.3.4.1.3 Other Environmental Matters

Cognitive social psychology sees the environment within which an

organism operates as potentially being a major influence on its behaviour.
The macro environment is also considered important by both marketing and
international marketing to understanding the influences acting on the
customer and the internationalising organisation. Although it is convenient to
use various labels for the components of a county’s environment (economic,
political, cultural, etc.), the components overlap (Jain 1996, p. 202) and
interact with each other, so requiring statistical treatment at some stage in the
screening process.

Which of the ‘other environmental matters’ are considered relevant

varies a little from author to author but all the following influences can
regularly be found in the literature.

4.3.4.1.3.1 Physical
 134

Examples of international marketing writers who advocate considering

the physical influences of the environment on product demand or distribution
include Cooke (1972, p. 28); Douglas, Craig and Keegan (1982, p. 28);
Douglas and Craig (1995, pp. 58, 60), and Terpstra and Sarathy (1997, pp.
93-95).

4.3.4.1.3.2 Economic

Douglas and Craig (1995, pp. 60, 61), Kotler et al. (1994, pp. 65-66)
and Terpstra and Sarathy (1997, pp. 93-95) are just some of the marketing
and international marketing writers who advocate including economic
screening variables. However, few authors suggest which indicators to use.
Root (1987) is something of an exception in that he proposes 178 mainly
economic variables to draw from when selecting markets, but he still leaves
the researcher to determine which, how many, and at what weights are
relevant in their particular situation. Russow (1989, 1996, p. 57) took 21 of
Root’s variables and began the empirical testing that will eventually discover
which variables at what weights are influential during market screening.
Russow found that the following 13 variables were significant in determining
the short-list of markets:
 135

Variable components: Russow’s principal Per cent of

components analysis variance
item name: accounted for *:
GDP, ‘Indirect market size’ 16.4 %
Population,
Manufacturing output as a
proportion of GDP,
Money supply,
Product-specific domestic
production,
International reserves
GDP per capita, ‘Level of economic 13.9 %
Agricultural output as a development’
proportion of GDP,
Average hourly wages in
manufacturing
Growth in product-specific ‘Product-specific 10.1 %
production, imports & market size growth’
exports
Product-specific imports & ‘Product-specific trade’ 9.5 %
exports
* average variance accounted for across six sample products

The discriminating power of the following variables was relatively

low (1996, p. 59):

Capital formation as a Capital spending 8.6 %

proportion of GDP
Population as a proportion of Population density 8.0 %
surface area
Construction as a proportion Infrastructure maintenance 7.2 %
of GDP & development
 136

Russow does not claim to have tested all feasible variables, having
selected from Root’s theorising those variables which he felt most influential
when related to several theories24. This dissertation replicates Russow’s work
to the extent that it includes a large data set of product-specific and economic
variables, but chooses fewer economic variables and adds additional non-
economic variables to capture additional forces which can affect a market
selection.

The economic variables which have been chosen for this research are
the log of GNP (Gross National Product) per capita, growth in GNP per capita,
population, and openness (exports and imports as a per cent of GDP). These
variables are supported by the permanent income hypothesis (Friedman
1957).

The variables selected are useful because:

• GNP sums all economic activity in an economy and determines

each nation’s purchasing power. GNP includes both the domestic
activity (GDP) and the external activity. Dividing it by population
and using the PPP (Purchasing Power Parity) method of calculation
produces a figure which is comparable across countries. The log of
GNP is used to reduce the bias between national extremities and to
reflect diminishing returns in the level of development, as in the
international development literature (e.g. United Nations
Development Program, 1998). To add further sub-components of
GDP or GNP (such as wages or money supply) would double count
those items and be justifiable only if prior research had indicated the
product-specific association with that economic sub-component;

• growth in (as opposed to shrinking) GNP per capita is seen as

desirable because the market is more likely to be expanding,

24
Russow believes these variables assess markets on the basis of cost and availability of
factors of production (per the absolute and comparative advantages of nations and the
Heckscher-Ohlin theorem), level of economic development, and market size and growth.
 137

profitable, less competitively aggressive, and forgiving of strategic

errors. Using the shift-share method to calculate the increase in
each country’s GNP per capita eliminates the biases of either the
absolute or the percentage increase methods;

• population is fundamental because potential domestic demand is

partly driven by the number of potential consumers; and

• the sum of imports and exports as a per cent of GDP indicate

national openness to international trade and foreign investment,
purchasing foreign products, and economic efficiency. Openness
aids national competitiveness (the benefits of which have been
argued in sections 4.2.4.1.2 and 4.3.4.1.2) as well as easing market
entry by whatever method is used.

4.3.4.1.3.3 Technological

Among the marketing and international marketing figures who advocate

considering this variable are Bradley (1995, p. 144), Douglas and Craig (1995,
pp. 60, 61) and Kotler et al. (1998, p. 119).

4.3.4.1.3.4 Legal:

Bradley (1991, pp. 154-161), Cateora (1996, p. 160) and Kotler et al.
(1994, p. 72) are examples of the many marketing and international marketing
writers who believe that legal aspect of the environment should be considered
by organisations working on market selection and entry.

4.3.4.1.3.5 Political

Examples of the numerous marketing and international marketing

writers who believe that political matters have great potential to alter the risks,
 138

and therefore the rewards, in each market include Bradley (1991, p. 138),
Cateora (1996, p. 138), and Kotler et al. (1994, p. 72).

Most writers describe foreign market political forces in negative terms –

as constraints, limitations and risks, not as an area that offers opportunities for
rewards. This results in the variable being badged as ‘political risk’ by many.
Jain (1996, p. 248) categorises all the negative impacts on international
businesses from the political area as stemming ultimately from issues related
to political sovereignty and political conflict. These can lead to market
interventions through such devices as import restrictions, market and price
controls, expropriation, and so on (Bradley 1995, p. 164; Jain 1996, p. 252;
Terpstra & Sarathy 1997, p. 153). The probabilities of occurrence of these
negative aspects can be quantified and are calculated by political risk
estimating organisations – many of which also include economic and/or
financial risks because of the relationship between politics and these other
areas (see Coplin & O’Leary 1994 for discussion about 10 risk assessment
services, including BERI, Moody’s, etc.).

Cateora (1996, p. 8) adds consideration of the home market’s domestic

political forces to the foreign market matters. This is consistent with the
approach taken in this research, which is to include all markets in the
assessment.

Cateora (1996, p. 155) and Terpstra and Sarathy (1997, p. 156) also
discuss how the behaviours of the organisation can influence political
outcomes. An aspect of this is the relationships developed between the
organisation and powerful allies in foreign markets. If relationships have
already been developed by the time that screening is undertaken, they should
be included in the screening framework in ‘Key stakeholder support’. A
further, subsequent aspect is the possibility of negotiating with, or lobbying of,
political decision-makers to change / not change impacting regulations. Whilst
this could be included during screening, it seems likely to have a low level of
accuracy and require a high level of work to estimate it for all markets, so the
matter is desirably normally left to the in-depth assessment stage.
 139

A related matter to networking is the extent to which corruption is likely

to be commonly involved in facilitating positive outcomes in a market. This is
indirectly accounted for through the cultural index (Hofstede 1984), and is not
overtly considered separately. A corruption indicator such as the EIU’s
(Economist Intelligence Unit’s) could be included for concerned organisations.
Whatever ‘legitimate’ or ‘corrupt’ encouragement is given by governments to
foreign firms is excluded from screening as it is entry mode specific
assistance and so deferred for consideration during the in-depth assessment
phase.

Some products are more politically vulnerable than others are (Bradley
1995, p. 168; Cateora 1996, p. 150), and the vulnerability varies from market
to market. However, “ ...there are no absolute guidelines (that) a marketer
can follow to determine whether or not a product will be subject to political
attention” (Cateora 1996, p. 150). This is arguing the need to collect primary
data, along with the ephemeral nature of attitudes and the influence process.
However, a pre-requisite of screening is to not collect market-specific primary
data, the matter is not seen as major for most products, nor is it essential to
be determined during the screening stage. Assessing product-specific
political attractiveness of markets is therefore left to the in-depth research
stage.

This dissertation has thus argued that the impact which the political
dimension should have on screening is delivered solely via the country risk
and cultural variables and not on a product-specific basis.

4.3.4.1.3.6 Cultural

There is a vast literature on the role of culture in international marketing

(see, for example, Albaum, et al. 1994, p. 96; Benito & Gripsrud 1992; Bradley
1991, pp. 109-133; Cannon & Willis 1986; Douglas & Craig 1995, p. 60;
Evangelista 1994; Fletcher & Bohn 1996; Goodnow & Hansz 1972; Gould &
McGillivray 1998; Hansz & Goodnow 1972; Hassan & Blackwell 1994 chapter
 140

5; Hofstede 1980, 1984, 1991, 1994, 2001; Hofstede, Allen & Simonetti 1976;
Hofstede & Bond 1988; Holzmüller & Stöttinger 1994; Jain 1990; Kent 1993,
p. 21; Kotler et al. 1994, pp. 75-77; Kumar, Stam & Joachimsthaler 1994;
Liander et al. 1967; Litvak & Banting 1968; Madsen 1994, p. 39; Munene,
Schwartz & Smith 1998; Paliwoda 1993, pp. 111-2; Papadopoulos & Jansen
1994; Perlmutter 1969; Peterson 1990; Reid 1986; Sagiv & Schwartz c1999;
Schwartz & Bilsky 1990; Schwartz 1992, 1994a, 1994b, 1996, 1997, 1999;
Schwartz & Lilac 1995; Schwartz & Ros 1995; Schwartz & Bardi 1997;
Shoham, Rose & Albaum 1995; Smith & Schwartz 1997; and Styles & Ambler
1997).

These are some of the many who believe that an important market
selection issue is the gap between actual and expected behaviour of the
buyer and seller, and that culture is a fundamental influence on this. In
general, the larger the gap between the home and host cultures, the more
difficult, expensive and threatening it is for both sides to conduct business.
Empirical evidence at this stage is somewhat patchy but a strong case can
provisionally be made (e.g. Arora & Fosfuri 2000, p. 555; Hennart & Larimo
1998; Kogut & Singh 1988; Makino & Neupert 2000, p. 710). The impact of
culture may also be subtle, with it acting as an intervening variable rather than
as either an independent or a moderating variable (e.g. Veiga et al. 2000 used
neural networking to demonstrate culture among the layer of hidden
variables.) It is not likely to be a matter of whether or not cultural distance
should be minimised, ceteris paribus, but of which factors interact with cultural
distance to magnify and retard its impact.

Figure 4.5 will now be used to argue the relevance of culture to

international market screening.
 141

Figure 4.5
Culture-Related Influences
On International Market Screening

The Person:
(with individual
characteristics) Marketing mix:
• abilities & standardise /adapt
aptitudes /standardise
• attitudes & values Their Perception:
(cultural & non-) • cultural distance: Composite frameworks:
• temperament high /low /low • stage of
• cost/benefit internationalisation: Market screening:
Their Learning: assessment: early /intermediate easy /medium
- ve / + ve /+ ve /advanced
• enculturation & /difficult country
acculturation • motivation & • EPRG:
commitment to E /P /RG
• specific market
knowledge: internationalisation:
low /medium /high low /high /high
Other behaviours of
• general the organisation, &
internationalisation Performance influences
knowledge:
low /medium /high
Source: Developed for this
• non-
research based on the
internationalisation
literature; similar to Gould
knowledge
& McGillivray 1998, p. 263
 142

Figure 4.5 (after Gould & McGillivray 1998, p. 263) depicts how cultural
theory interacts with psychological theory and international marketing theory
to affect screening outcomes. Underlying cultural drivers are forces from the
national, organisational and individual levels, and perhaps the industry and
organisation sub-culture (Hallén & Weidersheim-Paul 1989, p. 18; Törnroos &
Möller 1993, p. 113). Feedback occurs between many of the variables.

The presence of so many aspects of culture in an international

marketing situation require delimiting its role to specific situations (Törnroos &
Möller 1993, p. 108). Only factors which interact with culture to influence
screening are therefore considered in detail here, but the Figure includes the
other influential variables in summary form under “Marketing mix” and “Other
behaviours of the organisation, and Performance influences”. The in-depth
market assessment stage requires another iteration through somewhat similar
influences to those on screening.

The diagram was argued in Gould and McGillivray (1998) to depict how
people, each with differing personalities and enculturation, gain additional
learning about international business matters. Each are unique individuals
who have different abilities, attitudes and enculturation; they thus perceive
market opportunities differently when conducting market screening. An
alternative abstract way of expressing this is that behaviour is a function of the
person, their learning, and their perceptions about the prevailing situation
(following Lewin 1936). Just as ability and attitude limit an individual’s
performance at work and play, the composite framework provided by the
stage of internationalisation (ability) and Perlmutter’s EPRG framework
(attitudes) suggest typical limits to the organisation’s international marketing
capability25. The EPRG model can be replaced with Hofstede (1980),
Schwartz (1992), or some other cultural-level quantitative data. The Schwartz
values can be configured at the individual-level to measure the person’s
values (Smith & Schwartz 1997, p. 87), which is advantageous as it allows

25
Perlmutter (1969) proposed the EPG (Ethnocentric, Polycentric and Geocentric) approach,
to which was later added the Regiocentric option.
 143

predicting the individual’s attitudes, then their behaviours, in specific

situations.

Enculturation and acculturation involve the learning process and

develop shared attitudes and values in the minds of the members of the
society. People are also members of sub-groups, for example, nations, ethnic
sub-groups, religions, organisations, industries, etc.. Törnroos and Möller
(1993, p. 113) see five different levels of culture: national, industry,
organisation, organisation sub-units and individuals. Individuals may thus
vary from the entire culture’s average assessment (Nasif et al. 1991, p. 87).
Sub-group memberships develop additional layers of specialist attitudes and
values. These various attitudes and values are just some of those operating
in any individual and exist irrespective of individual differences. Reid (1986,
pp. 25-27) can be interpreted as suggesting that individual enculturation and
acculturation through growing up in, or migrating to, a culture skews
knowledge about that particular culture. This reduces cultural distance,
increases competence in dealing with that culture, reduces uncertainty, and in
turn predisposes decisions in favour of that market. An individual’s
enculturation and acculturation can also be expected to affect that individuals’
attitudes about tolerance of other cultures, interest in other cultures, preferred
learning styles and so on; these influence how individuals learn about other
cultures. It could thus be hypothesised that cultures which facilitate
personality traits such as tolerance (O’Grady & Lane 1996, p. 324), flexibility,
trust, inquisitiveness, and low levels of xenophobia and chauvinism would be
likely to be environments which increase spontaneous learning about other
cultures – both specific and general learning. In turn, Johanson and Vahlne
(1977) argue that learning reduces cultural distance.

The EPRG framework can be related to models of the stages of

internationalisation of an organisation (Shoham, Rose & Albaum 1995, pp. 11-
14). An example is the four step Douglas and Craig (1995, p. 31) approach,
where:
 144

• ‘pre-internationalisation’ is associated with high ethnocentric

management attitudes which inhibit an organisation’s international
activity (the organisation consequently does not trade internationally, or
does so in only a limited, reactive way);

• ‘market entry’ is supported by high polycentric attitudes in the

organisation’s marketers (with the organisation proactively
implementing high local adaptation of its marketing mix to increase
sales) and an adequate number of clients with low ethnocentric
attitudes (who are thus willing to try the non-local product and supplier);

• ‘market expansion’ supported initially by high polycentric attitudes of

the organisation’s managers and, if the clients know the organisation is
foreign, low ethnocentric attitudes in them. At some stage, the
organisation may be influenced by cost or other reasons to alter from
polycentric attitudes to becoming regiocentric or geocentric; and

• ‘market rationalisation’ is accompanied by regiocentric or geocentric

attitude marketers.

Impact of culture on market screening

Figure 4.5 also summarises generally likely outcomes from combining

the various cultural influences on an organisation and its individuals.
Managers with low specific market knowledge, low general internationalisation
knowledge who have had an unsupportive enculturation can be expected to
have high cultural distance, a negative cost/benefit assessment and a low
commitment to internationalisation, and so be likely to be ethnocentric in
attitude. Conversely, people with high specific market knowledge, high
general internationalisation knowledge and having received suitable
enculturation can be expected to have low cultural distance, a positive
assessment of the costs and benefits of internationalisation, a high
commitment to internationalisation, and to be regiocentric or geocentric in
 145

attitude. In between those two extremes will be managers who have

intermediate levels of specific market knowledge and general
internationalisation knowledge. They can be expected to have low cultural
distance, high commitment to internationalisation and a positive assessment
about the costs and benefits of internationalisation, be at an intermediate
stage of internationalisation and have a polycentric attitude.

Consider now the markets selected by market screening for possible

entry by the organisation. They range from ‘easy’ through to ‘difficult’ markets
in terms of cultural distance, commercial and political risk, competition levels,
demand, and so on. The organisation’s sales, profit, etc. performance will be
an outcome from matching the new market’s difficulty with the organisation’s
level of international competence and attitudes. Thus, screening will be
ineffective if it suggests markets that are beyond the capability of the
organisation to achieve entry successfully. The implications of all the
aforementioned for screening are that:

• ethnocentric organisations are less likely to have accumulated the

knowledge, capability and preferences to be able to handle difficult
markets, and should therefore concentrate on low cultural distance,
low competition, and not very different markets. They typically use
a standardised marketing mix, irrespective of the desirability of
doing so;

• polycentric organisations have high motivation but only moderate

accumulated knowledge that equips them to handle intermediate
levels of market difficulty. They typically aid the process by using a
highly adapted marketing mix;

• regiocentric and geocentric organisations have a high level of

accumulated international markets knowledge which (subject to
continued motivation, resources and demand) equips them with the
competence to handle the economic, cultural, political, etc.
 146

dimensions of really difficult markets. Cost and organisational

pressures will abate their earlier desires to adapt their marketing
mix and cause them to introduce standardised, international
marketing mixes wherever possible.

The above demonstration of the complicated inter-relatedness between

many screening variables and culture is believed by the researcher to be
indicative of how most variables are involved with many others. Altering the
value of one of the variables is likely to produce knock-on affects on many
others.

4.3.4.1.3.7 Ecological/environmental

Among the marketing and international marketing figures who advocate

including this variable are Douglas and Craig (1995, p. 378), Kotler et al.
(1994, p. 66) and Terpstra and Sarathy (1997, p. 949). Douglas and Craig
see ecological/environmental issues as recently emerged negative concerns
which pervade all of the organisation’s activities, requiring proactive
conservation of resources and the elimination of toxic products and waste (p
378). On the positive side, some markets represent higher potential for
pollution abatement products than others (eg Hong Kong). A balanced view is
probably that the environment represents both opportunities and threats, and
that the influences vary by country.

It can correctly be argued that much of this variable is already indirectly

embedded in the economic, legal, political and cultural variables. However,
those four variables include many other matters, and do not specifically
include environmental aspects. Also, environmental matters are presently of
much significance in some industries and markets, are increasingly so in
many, and appear likely to continue thus in the foreseeable near future. The
issue is thus judged sufficiently important to merit isolation and specific
attention when discriminating between markets.
 147

4.3.4.1.4 Market Risk

Market risk levels vary by market (see, for example, Bradley 1995;
Douglas & Craig 1995, p. 55; Paliwoda 1993, p. 104). The dominant factor in
explaining the overall impact of internationalisation on firm risk and leverage is
the different risk classes of different countries, believe Kwok and Reeb (2000,
p. 626).

Whilst the different risks can be measured in various ways (Coplin &
O’Leary 1994), recognised assessment of the risks in each market include
those by BERI, EIU and ICRG (respectively Business Environment Risk
Intelligence, Economist Intelligence Unit, and International Country Risk
Guide). The assessment can be made with an emphasis on the nation-wide
perils affecting any organisation in that market (e.g. BERI’s) or can emphasise
the organisation’s individual, commercial risk levels in each market (e.g. an
option with ICRG). However the firm wishes to measure it, this market-
specific risk level can be compared against the risk level set in the
organisation’s objectives then an accept / reject decision made. Alternatively,
as in this research, risk can be made one of numerous matters included in
quantitatively assessing every country’s attractiveness.

The assumption that the commercial, country risk estimating services

do produce a satisfactory proxy of beta in the Capital Asset Pricing Model
(Treynor et al. 196426) has been examined by Harvey over a number of years
(e.g. see Erb, Harvey & Viskanta 1996). He finds that beta is a reliable
indicator of risk in developing countries, but an unreliable indicator in
emerging markets. However, his analysis of the ten leading country risk
estimating services shows that many are good ex-ante indicators of risk in
both developed and developing markets, and that the ICRG Composite Index

26
Capital Asset Pricing Model attributed to Jack Treynor, William Sharpe, and John Lintner.
See Sharpe, W. F. (1964) Capital Asset Prices: A Theory of Market Equilibrium Under
Conditions of Risk, Journal of Finance 19 (September) pp 425-442, and Lintner, J. (1965) The
Valuation of Risk Assets and the Selection of Risky Investments in Stock Portfolios and
Capital Budgets, Review of Economics and Statistics, 47 (February) pp 13-37. Treynor’s
article was not published.
 148

is amongst the better indicators (e.g. Coplin & O’Leary, 1996, p. 44). It was
thus used as the measure in this research.

Conceptual purity would then argue for a slight reduction to the market
risk level where the organisation diversifies its markets. The Capital Asset
Pricing Model separates risk into unique risk (which reduces logarithmically to
near zero by the time the portfolio comprises some 10 or more investments)
and market risk (which comprises risks that affect all businesses) (Brealey &
Myers 1984, p. 126). The level of market risk varies with each market and is
estimated by the abovementioned commercial outlets. Whilst that risk
estimate should then be reduced slightly for each of the first 10 markets
entered, the amount will be small and so has been ignored. The framework is
consequently very slightly conservative in its risk adjusted reward hierarchy of
suggested markets if the organisation is already selling in numerous markets.

Madsen (1989, p. 44) reported how, when experienced export

managers were asked to judge export marketing policy and market
characteristics, they were unable to respond to absolute value levels, instead
comparing against their domestic market. This is reasonable in a situation
where international activities are a small part of total activity, and is consistent
with Simon’s limited set, non-optimising theory of the firm. However,
comparing against only the domestic market is an incomplete assessment that
will sub-optimise.

Considering all existing markets, domestic and foreign, is therefore

necessary when setting screening objectives levels and judging the
attractiveness of the short-listed markets. It may even be that screening will
indicate that the organisation should expand in an existing market rather than
enter a new one.
 149

4.3.4.2 Organisation’s Capabilities and Preferences

The international performance of organisations and the psychological

literatures are the predominant influences on this section of this dissertation,
which describes how the organisation’s unique attributes should shape which
international markets are selected.

Pedersen and Petersen (1998, p. 497) found that the pace at which
firms internationalise is affected by their knowledge, resources and market
volume. The latter two items are found in the screening framework whilst
knowledge is later argued to be a sub-component of competence (Figure 4.6).

Capability and commitment are also believed to play a role in helping to

discriminate good from poor performing firms (Evangelista 1994, p. 207). “...
management characteristics and attitudes ... are widely considered to be
among the most crucial determinants, if not the most crucial determinants of
export performance, although there are also findings which call this into
question (Diamantopoulos & Schlegelmilch 1994, p. 162).” This author
(Gould) believes in a more muted interpretation, where management values
about further internationalisation need to be positive for further
internationalisation to not get vetoed. Management also needs to allocate
necessary resources, encourage staff, monitor performance, and incorporate
international activities in the organisation’s overall strategies and activities. A
threshold level of positive management values is thus necessary, but
insufficient, for success. Its relative importance probably changes over the life
cycle of internationalisation as argued by Bilkey (1978, pp. 34-37, 42) and
Madsen (1994, p. 33).

Changing paradigms, cognitive social psychologists believe that

situation-specific competencies of the person, and the person’s perception of
their capability at using their skills in specific circumstances to achieve goals,
determine which goals are chosen and the height at which standards are set
(Pervin & John 1997).
 150

4.3.4.2.1 Organisation’s Abilities

4.3.4.2.1.1 Sustainable Competitive Advantage (SCA)

Early economic thinking saw land, labour and capital as the

fundamental assets of a business, then later added knowledge. Economist
and strategist Michael Porter (1980, p. 70) views competitive advantage as
the basis for firms choosing between the strategies of cost, differentiation or
focus. Dunning (1980, p. 9) argues that organisation competitive advantage
explains the propensity of an enterprise to engage in international
production27. International marketer Madsen (1989, p. 47; 1994, p. 34) also
sees competitive advantage as the fundamental factor, limiting international
performance by determining feasible target segments. Early international
movers often develop a lead over competitors that is difficult to overtake (Huff
& Robinson 1994; Lieberman & Montgomery 1988; Pan, Li & Tse 1999, p. 82)
for economic, pre-emptive, technological and behavioural reasons (Pan, Li &
Tse 1999, p. 83) and which increases market share and profitability (ibid p.
99). Whatever the sources, economies of scale and scope provide advantage
over competitors to generate higher rents.

Aaker (1995, p. 80) provides a tool that takes account of much of the
theory espoused above, assessing an organisation’s strategic strengths and
weaknesses. It assesses the second component of strategic competitive
advantage – the assets which do not necessarily provide an advantage
because others may have them, but whose absence would create a
substantial weakness (p 100). Aaker’s tool does not endeavour to determine
and quantify the assets and skills that are industry-specific necessities, which
is not problematic in a general screening situation. However, his
questionnaire does require supplementation in the insufficiently covered areas
of labour and knowledge.

27
Dunning believes that the three matters which explain a firm’s propensity to produce
internationally are the extent to which it possesses assets, whether the firm sells or leases
these to other firms, and the involvement of indigenous foreign resources. The latter two
matters relate to choice of mode of entry and are not considered during screening because,
151
 152
Figure 4.6
Locating Screening In The Personality-Learning-Competence-Performance Chain

Duration, intensity, & range of training/experience * Growth & renewal, Efficiency, Stability, Effectiveness (S)
↓ ↓
(Ability & aptitude) (Education, ) (General business competencies * (J) ) Opportunities (J)
(Attitude & values) (Training & ) (International) (Further internationalisation competencies * )
(Experience)(S, J) (Learning)(S, J) (- generic internationalisation competencies) Screening Performance (S, J)
(Motivation & goals) (- market-specific international competencies (J))
Effort (S)
(Environmental variations) (Assumed indicators of the stage of internationalisation which reflect)
( in training/experience )(J) ( the learning necessary for successful further internationalisation )
↑ ↑
↑ ↑
Marketing mix No of markets regularly sold in )
Potential customer aspects Range of cultural distances of markets regularly sold to )
Competition & substitutes No of modes of entry regularly employed )
Risk levels No of recent entries, expansions & rationalisations ) Growth
Other matters Language competencies of international marketing staff ) &
- economic Formal relevant education & training credentials (S) ) renewal
(S)
- technological Amount of practical international marketing, etc. experience )
- legal Whether top decision-makers born / lived / worked abroad )
- political Internationally early / late / lonely / among others )
- cultural No of international training hours (S) )
- ecological No of international marketing research hours )
International marketing staff turnover (S) ) Stability
Satisfaction of key target customer segment (S) ) Effectiveness
Management assessment of performance ) Efficiency & Effectiveness
(S)
= consistent with Sveiby (1997) (J) = consistent with Johanson & Vahlne 1977
* Industry specific quantity and range of necessary skills (e.g. fish & chips vs brain surgery)
Source: Developed for this research based on analysis of the literature
 153

4.3.4.2.1.2 General Business Competencies

Definitive identification of the knowledge and skills that lead to

competitive advantage and to higher performance has yet to be finalised (Zou
& Myers 1999, p. 7), but the literature provides some indicators.

Separating ownership from control of an organisation introduces

professional managers who are more likely to posses higher levels of
competence to improve enterprise performance (Samuelson 1973, pp. 126-8).
Marketing and management theorists have long held that an organisation is
more likely to maximise its performance if it employs people having generic
functional skills in such areas as marketing, finance, management, perhaps
research and development, production, and so on according to the industry’s
needs (e.g. Koontz & O’Donnell 1968, pp. 506-532; Stanton 1994, pp. 7-16).
Zou and Myers (1999) empirically affirmed the importance of R and D,
manufacturing and marketing as being distinctive competencies which lead to
competitive position in global industries (p 8). Competence involves such
components as human abilities, training, experience, task-specific skills and
time (Barney & Griffin 1992, p. 524). Johanson and Vahlne (1977, p. 28)
make the distinction between general (business) knowledge and market
specific knowledge, the former being irrespective of geographical location and
the latter being market specific. Bilkey (1978, p. 36) found that exporting firms
tend to have better management than non exporters.

4.3.4.2.1.3 Further Internationalisation Competencies

Further internationalisation competencies may be either generic or

market-specific. Hundreds of studies into the causes of high international
performance have been undertaken. Their quality and the supposed causes
of performance are assessed by authors such as Aaby & Slater (1989), Bilkey
(1978), Chetty & Hamilton (1993), Madsen (1989), and Zou & Stan (1998).

“Management’s ability to apply appropriate technology, establish

necessary commitment, acquire international knowledge, institute
 154

consistent and realistic export objectives, develop export policy, and

establish the necessary management control systems are important
export competencies”

wrote Aaby and Slater (1989, p. 18) following their review and synthesis of 55
empirical studies covering 9,000 organisations during the 10 years ending
1988. Competence thus involves numerous abilities, each being potentially
unique in: (1) its relative importance to task accomplishment and (2) the level
of proficiency needed. What competencies, at what performance levels, are
relevant to market screening?

Evangelista (1994, p. 212) disentangles general business competence

from international business competence. (Consulting Figure 4.6 at this time
may be helpful.) Internationalisation can be seen as a learning sequence
involving feedback loops (Bilkey 1978, p. 42). Incremental acquisition of
international business experience, again divided into general and market
specific, appears to be a key explanatory variable to understanding
organisational performance (Johanson & Vahlne 1977). Bilkey (1978, p. 42)
can be read to mean that organisation competence increases as its stage of
internationalisation increases.

General internationalisation knowledge comprises such matters as

marketing methods and common customer characteristics. It is knowledge
that is transferable from one country to another i.e. is not market specific
(Johanson & Vahlne 1977, p. 28). It includes knowledge which is acquired
incrementally and experientially (Johanson & Vahlne 1977, p. 28 following
Penrose 1966, p. 53; Yu 1990 quoted in Benito & Gripsrud 1992, p. 463)
through such exposures as time lived overseas, travel, trading with other
cultures and extra-organisational linkages of the organisation (e.g.
Miesenbock 1988; Reid 1981; Reuber & Fischer 1997). However, it would
seem reasonable to expect general international knowledge to be impacted by
the internationally relevant aspects of formal training in functional matters
such as marketing, finance, planning, etc. because these convey general
principles useful to numerous situations. Madsen (1994, p. 34) argues that
 155

top management’s competencies are more important to the organisation

during the initial stages, after which the functional specialists become more
important. Specific market knowledge can be expected to affect general
internationalisation knowledge, and vice versa, because they are an
interaction between the general and the particular.

Screening involves evaluation of alternative markets, so requiring

knowledge about those alternative markets and their environments. It
requires specific market knowledge about some aspects of specific foreign
cultures, their product demands, preferences and other matters. Knowledge
about a foreign country reduces both the cost and uncertainty of operating in
the foreign market (Benito & Gripsrud 1992, p. 462; Buckley & Casson 1981).
This knowledge can originate from either being experienced or being taught.
Much of it is tacit knowledge (implicit, ill-codified, often experientially acquired
knowledge) which Simonin (1999, p. 469) argues is difficult to transfer.
However, learning by trial-and-error and modelling is costly (O’Grady & Lane
1996) so one would hope that better understanding will in future permit the
more efficient teaching of this knowledge. Presumably Johanson and Vahlne
have only isolated part of the answer about learning. Cognitive social
psychologists would argue that experiential learning can be supplemented by
psychological modelling and formal instruction once: (1) we know what
behaviours lead to successful international performance, then (2) reward or
punish behaviour appropriately.

A significant link between experientially obtained market knowledge,

market commitment and high international performance is believed (see, for
example, Styles & Ambler 1997, p. 7). Knowledge acquisition is enhanced by
market specific language skills to aid communication and a deep
understanding (e.g. Arora & Fosfuri 2000, p. 562; Cannon & Willis 1986;
Liesch et al. 2000, pp. 10-11). Network contacts can provide significant
location-influencing benefits (e.g. Chen & Chen 1998, p. 447; Dunning 1997;
Ellis 2000; Johanson & Gunnar-Mattsson 1986; Rangan 2000, p. 205) related
to sales opportunities, strategic resources, market intelligence, technological
know-how, reputation, reduced risk, etc.. The importance of network contacts
 156

in influencing location varies with culture, industry, organisation size, and so

on (Chen & Chen 1998). Specific market knowledge partly explains the
pattern and pace of internationalisation of firms by affecting both market
commitment and the commitment decisions (Johanson & Vahlne 1977, p. 23).
For example, Bilkey (1978, p. 42) argues that stage of internationalisation
affects management’s short term expectations concerning profit and growth
and perceived export inhibitors, and so whether or not the organisation
exports. Reid (1986, p. 24) and Hallén and Weidersheim-Paul (1989, p. 18)
suggest that knowledge skewness in favour of any market may distort
selection by revealing apparently more opportunities there and/or by reducing
uncertainty, thus perceived risk. Knowledge reduces psychic distance by
reducing unknowns about other cultures (Keegan & MacMaster 1983) and by
building trust (Hallén & Weidersheim-Paul 1989, p. 18).

Early writers used ‘cultural distance’ to indicate an individual’s

perceived difference between one particular culture and another. ‘Psychic
distance’ indicated the differences that individual perceived between one
culture and foreign cultures in general. Both distances were considered to
arise from differences in language, nationality, ethnicity, religion, business
etiquette, etc. (see, for example, Hallén & Wiedersheim-Paul 1989; Johanson
& Vahlne 1977, p. 24; Reid 1986). Recent writers interchange ‘psychic
distance’, ‘psychological distance’ and ‘cultural distance’ to blur precision of
their meanings (see, for example, Kogut & Singh 1988, p. 430; O’Grady &
Lane 1996, p. 312). This dissertation does not follow that trend because the
move disadvantages readers by leaving undistinguished the differences in
meaning between the two situations.

Cultural and psychological distances are of interest because the

presumption is that minimising them when selecting markets requires less
learning by managers and employees, and so produces a higher likelihood of
the organisation succeeding at market entry. General internationalisation
knowledge, in conjunction with a supportive enculturation and personality, can
be expected to predispose an individual to a reduced level of psychic distance
from other cultures. When that individual develops specific market
 157

knowledge, they can then be expected to display reduced cultural distance

from a specific culture.

Different stages of internationalisation are associated with different

attitudes toward ethnocentrism, polycentrism, regiocentrism and geocentrism
(Perlmutter 1969; Shoham, Rose & Albaum 1995). Less experienced firms
are typically more ethnocentric whilst very experienced firms are more likely to
be geocentric, although there is debate about the certainty of an organisation
achieving attitudinal geocentrism (Mayrhofer & Brewster 1996). Cultural
distance is related to stage of internationalisation, with increasing cultural
distance being associated with more internationally experienced firms
(Shoham, Rose & Albaum 1995, p. 35). Internationalisation is thus an
evolutionary process of learning to take a series of microsteps along a
continuum of commitment of resources – but an organisation may skip steps
or move at differing velocities (Leonidou & Katsikeas 1996, p. 527).

In the early stages of internationalisation of an organisation, when little

specific market knowledge exists, it can be anticipated that the stage of
internationalisation reached by the organisation is driven by individual
personality – itself affected by enculturation – combining with the level of
general internationalisation knowledge, and the perceptions about psychic
distance, cost/benefit, and commitment to internationalise. If the organisation
does internationalise somewhat, specific market knowledge is added,
affecting the three perceptual matters mentioned, then increasing or
decreasing the stage of internationalisation. The stage of internationalisation
reached is of interest to screening because it is one of the determinants of
how easily the organisation is able to enter markets of differing degrees of
difficulty.

The number of the stages of internationalisation which an organisation

moves through has not been resolved (e.g. Rao & Naidu 1992) but it is agreed
that firms evolve with international experience (Johanson & Vahlne 1977, p.
23; Shoham, Rose & Albaum 1995, p. 17). The arguments have already been
put that stage of internationalisation of employees comprises the platform of
 158

their individual personalities overlayed with specific market knowledge,

general internationalisation knowledge, psychic distance, a cost/benefit
assessment and their level of commitment to internationalisation.

Whilst there is no agreement about the number of stages of

internationalisation, there is considerable congruence among the various
models about the core conceptual characteristics of the international
development process (Leonidou & Katsikeas 1996, p. 534). Empirical testing
reveals a learning process wherein firms gradually become more familiar with
foreign markets and operations (op. cit. p 521). In contrast, Seringhaus
(1986) believes that both experiential and general information are needed in
the early stages, but that more objective and specific information is collected
and used in the later stages. Casual extended observation by the author of
firms and of numerous government export facilitation bodies revealed very
little extensive or ongoing use of objective international data by either
experienced or inexperienced organisations. It seems reasonable to expect
that some knowledge is best learned objectively whilst some is better learned
experientially (Booth di Giovanni 1997, p. 4), and that the optimum learning
method will vary with the specific matter to be learned, the individual’s
attributes, and the stage of internationalisation. However, the question of
what is, and what should be, the balance between experiential and objective
learning is presently unresolved.

Cognitive social psychologists would see internationalisation

competencies as having an interaction with motivation and performance – and
so influencing goals and standards setting, perceptions of achievability of
same, rewards to be derived, effort necessary, and subsequent learning
(Pervin & John 1997, p. 414). These variables are all present in the proposed
screening framework. Cognitive social psychologists disagree with trait theory
psychologists in believing that static trait measures are not helpful predictors
of performance (Pervin 1993, pp. 423-4). They argue that: (1) the
environment requires different traits to be needed in different situations, and
that (2) more learning is possible than the relatively more biologically
determined trait viewpoint accepts. Since international business research has
 159

normally concentrated on individual traits without relating their levels to

different environment variables, and since many environmental variables have
different stages (e.g. stage of internationalisation of the organisation), it is
understandable that research has been unsuccessful at isolating the
antecedent variables for successful performance.

Strong relationships appear to have been found by some researchers

with such personal variables as: (1) age (see, for example, Barrett 1986), (2)
education (see, for example, Barrett 1986; Fletcher 1996) (3) time spent
overseas (see, for example, Barrett 1986; Fletcher 1996). Reid (1981, p. 105)
believes managerial knowledge, attitude and motivation are the critical
individual characteristics for further internationalisation. About individual
characteristics, he continues:

“A number of studies indicate that individual antecedents – such as

type and level of education, foreign nationality, ability to speak foreign
languages and extent of foreign travel – are likely to be associated with
the exporting decision-maker’s existing stock of knowledge, his (sic)
attitudes, and effective preferences concerning foreign markets.” (p
105)

However, attempts at isolating the personal demographic

characteristics that enhance international business performance have so far
produced contradictory results (see, for example, Evangelista 1994, p. 225;
Fletcher & Bohn 1996). This may be because global traits are unable to
capture the variety of environmental influences that determine performance
(trait theory psychologists would disagree with this statement whilst cognitive
social psychologists would agree). It could alternatively be because all the
variables involved have not been assembled together, nor related to the
variations in each of their rhythms. For example, personality traits needed in
management will differ in some ways from those needed by the international
functional specialists; further, these traits should presumably differ between
management and functional specialists according to the stage of
internationalisation of the organisation. This is in line with the views of
 160

cognitive psychology, who see behaviour as a function of an individual’s

personality and their situation (see, for example, Lewin 1936; Pervin 1993,
p. 388). Situational variables, such as industry, organisational resources and
competence, target culture, and so on, can thus be argued to be relevant to
an organisation’s international performance level, so complicating isolation of
the ideal employee characteristics. The proposed screening framework
bypasses these unknowns by targeting attitudes and competencies, rather
than their possible antecedents of age, education, etc..

It is consequently too early to be certain about which performance-

maximising international skills are common necessities for individuals to
posses so as to successfully handle differences in the business cycle, product
life cycle, intrinsic product demand, stage of the organisation’s international
experience and organisational goals. There may also well be sub-specialties
by organisation resources, industry and market. However, from the literature
it seems reasonable to expect that organisations at a higher stage in the
internationalisation process will generally be more internationally competent at
identical international tasks than organisations at an earlier stage.

Moving from the international arena to the general matter of work,

Sveiby (1997, p. 168) provides a conceptual model of competence, along with
suggestions for quantifying the level of an organisation’s competence. He
defines competence as requiring: (1) growth and renewal, (2) efficiency, (3)
stability, and (4) effectiveness (pp 154, 165). Applying Sveiby’s work in the
context of (a) the international business performance competence literature,
and (b) social cognitive theory, leads to the fourteen assumed indicators of the
stage of internationalisation of the organisation which were listed earlier (in
section 4.2.4.2.1.3). Those criteria are a component of Figure 4.6, which
summarises the proposed personality-learning-competence-performance
chain. Those criteria are also among the managerial questions of Appendix
A.

Although we talk about organisation international competence in a way

that seems to imply homogeneity of competence throughout the organisation,
 161

the reality is that every employee of like ability will not automatically have the
same level of competence. Different individuals and functions within the one
organisation may be at different stages of internationalisation, and different
nationalities probably also vary in their version of internationalism (Mayrhofer
& Brewster 1996, p. 768). The level of management having the skill appears
to need to vary according to the organisation’s stage of internationalisation,
with top management’s international competencies being more important in
the initial stages, and the functional specialists more important later on
(Madsen 1994, p. 34). Finally, the comparative level of international
competence of the organisation to its competitors in the market will affect the
enterprise’s performance, so Johanson and Gunnar-Mattson’s (1986, p. 252)
rating of the organisation as an international early starter, a lonely
international, a late starter, or an international among others is thus apposite.

Cultural distance and psychic distance are indications of dissimilarity

between the home market and foreign markets (Reid 1986, p. 23). They
reflect perceived differences in management’s stock of knowledge (op cit.).
They also reflect real differences in the stock of knowledge – what is known
compared with what needs to be known. What is known is, as argued in the
previous variable, composed of general business knowledge and international
knowledge, the latter being further divided into either generic international
markets and market-specific international knowledge (Evangelista 1994, p.
212; Johanson & Vahlne 1977, p. 28). Once a knowledge deficiency is
recognised, Zikmund (1994, p. 19) asserts that there are four decision criteria,
all of which need to be answered affirmatively to justify undertaking market
research: (1) having sufficient time, (2) not already having the necessary data,
(3) having a sufficiently important decision to make, and (4) having a decision
where the benefits of research are greater than the costs of research.

Putting aside ability and interest, training and experience are the
origins of knowledge (Penrose 1966, p. 53). Elements of this knowledge, it
can be argued from Johanson and Vahlne (1977, p. 28), the organisation uses
to develop its competence at collecting and analysing international market
data. Sveiby (1997) suggests quantifiable, conceptual measures of training
 162

and experience, namely efficiency and effectiveness of knowledge, stability in

the firms’ stock of knowledge resident in its people, and the growth and
renewal of the knowledge in those people. Applying Sveiby to the
organisation’s research competence lead to 14 matters (section 4.2.4.2.1.3),
amongst which are four related to international market research
competencies, namely:

• number of hours of international marketing research undertaken,

• amount of formal education and training in international market
research,
• amount of practical experience in international market research, and
• management’s assessment of performance of the international
market research conducted by the organisation.

To the extent that these change year-on-year, they indicate (hopefully

positive) growth and renewal of knowledge. The fourth item also indicates
management’s perception of efficiency and effectiveness. Stability and
effectiveness of research competence are indirectly assessed by the earlier
mentioned indicators of international marketing staff turnover and satisfaction
of the key target customer segment.

Cavusgil and Zou (1994) found that international competence (p. 1),
including export market selection (p. 7) is a key determinant of performance.
International market screening is a specialised part of international market
research. The literature (see Chapter Three, “Literature Review”) advocates
that screening contains certain components:

• quantifiable variables
• inclusion of product-specific variables
• variable weights
• selection of variables based on theory
• consideration of all markets
• use of only secondary data for the market-related variables
 163

• distribution mode neutrality of variables

• a systematic decision system.

These are therefore included in both the screening framework and the
Questionnaire to Management

Just as different cultures have different levels of technical know-how

(Mayrhofer & Brewster 1996, p. 763), we could expect different cultures to
have different commencing levels of proficiency at market screening.

Whilst use of this variable is intuitively logical and arguable from the
literature, Madsen (1989, p. 51) found no significant association between his
‘a priori market research’ variable and export performance among his 83
experienced Danish capital goods manufacturing firms. “The reasons for this
lack of association need yet to be explored” (op cit.). Diamantopoulos and
Schlegelmilch (1994, p. 177) may provide the answer when they suggest that
a threshold level of export experience may be important, beyond which
additional increments may not make any substantial difference. Alternatively
and seemingly more likely, two thresholds may be involved: (1) a minimum
threshold level may be necessary for significant research impact, above which
improved efficiency and effectiveness dividends are possible from additional
training and experience, but then (2) this is capped with a second threshold,
beyond which negligible impact is generated.

4.3.4.2.1.4 Key Stakeholder Support

It may be helpful to here return to Figure 4.2 or 4.3. Management,

marketing and international marketing advocate considering the broad group
of interests that are affected by the business decisions (Jain 1993,p. 487;
Johnson & Scholes 1993 chapter 5; Kotler 1997, p. 65). Some progress on to
distinguish different ability to influence decisions (e.g. Kotler’s 1994, p. 213
users, influencers, buyers, deciders and gatekeepers).
 164

4.3.4.2.1.5 Other Resources

Management, marketing and international marketing theorists all

require that an adequate quantity and quality of resources are available for the
task to be undertaken (e.g. Brealey & Myers 1984; Diamantopoulos &
Schlegelmilch 1994, p. 176; Douglas & Craig 1995, p. 125; Kotler 1997, p. 66;
Madsen 1994, p. 28). However, empirical results are mixed, finding little
agreement about the impact of organisation size on either propensity to export
or export size (Aaby & Slater 1989, p. 17). The probability is that a minimum
level of resources is necessary for the tasks of the particular industry sub-
sector, beyond which the level does not matter greatly (e.g. OECD 1996, p.
54; Rangan 2000, p. 219). The view of this dissertation’s author is that using
organisation size, rather than resources, confounds results from any research
because it combines resources, general competence and international
competence i.e. amalgamating resources with general business competence
and the organisations’ stage of internationalisation. Johanson and Vahlne
(1977, p. 29) conclude that personnel resources are not easily increased in
the short to medium term – personnel or advice from outside the organisation
cannot be greatly used because much of the relevant knowledge is
experientially acquired and specific to the organisation and its market.
 165

4.3.4.2.2 Managerial Values

Values is a familiar term in psychology (e.g. Byrt 1971 p.36; Kolasa

1969 p. 397). Schwartz (1992, p. 4) sees Managerial values as being:
(1) concepts or beliefs, which (2) pertain to desirable end states or
behaviours, (3) transcend specific situations, (4) guide selection or evaluation
of behaviour and events, and (5) are ordered by relative importance. Values
are applied to specific situations to determine attitudes and behaviour. Inter-
relation of the various fragments of values to other fragments makes values
change a generally slow and difficult process that is akin to dominoes
knocking each other down. It is yet to be fully understood by researchers or
clinicians. During the relatively short period of time when an organisation is
screening markets, a change in values by managers is not likely, so values
about strategy, personal matters and risk are uncontrollable, thus moderating,
variables. On the other hand, the actual decisions about preferred strategy,
organisation and personal benefits, etc. are controllable, so among the
independent variables.

4.3.4.2.2.1 Further Internationalisation Values

Previous argument (related to Figure 4.6) was put that an

organisation’s international performance is substantially affected by its stage
of internationalisation. Managerial attitudes towards internationalisation have
been suggested as an explanatory variable of stage of internationalisation in
empirical studies by Barrett and Wilkinson (1986), Bilkey and Tesar (1977),
Cavusgil (1982a), Czinkota (1982), Lim et al. (1991), and Rao and Naidu
(1992) (in Leonidou & Katsikeas 1996, p. 526). The importance to
performance of the further internationalisation values of top management
probably varies with the stage of internationalisation (Madsen 1994, p. 33).
The reasoning here, which is commonly accepted but has yet to be generally
affirmed empirically, is that: (1) less internationalised organisations are initially
more dependent on top management’s knowledge and motivation, whilst (2)
more internationalised organisations have normally developed the
international skills of their lower level functional specialist employees to allow
 166

top management to reduce the amount of its operational involvement. The

characteristics of both top management and the key functional specialists
therefore need to be assessed when determining the organisation’s stage of
internationalisation (Madsen 1994, p. 33). Attitudes are a situation-specific
personality variable which is heavily influenced by values and the particular
situation (Schwartz 1992, p. 4, 1997, p. 80). Values should be more reliable
than attitudes at predicting behaviour (Schwartz 1996, 1997, p. 80).

Management’s perceptions of the obstacles to and disadvantages of

further internationalisation, which research shows are often wrong, relate to
these values (e.g. Aaby & Slater 1989, p. 17; Bilkey 1978, p. 35; Leonidou &
Katsikeas 1996, p. 537). How difficult it seems to the organisation for it to
carry out research, find finance, get government assistance, find foreign
distributors, and so on, is not related to the actual level of difficulty of the
tasks. It is the organisations’ perception of difficulty, and that perception
changes as the organisation internationalises.

4.3.4.2.2.2 Strategic Values

The discipline of logic (Copi 1968, p. 50) argues that decisions are
made on the basis of fact and belief. Psychologists such as Schwartz argue
that a person’s values are beliefs which bias the person’s decisions (1997, p.
80). Management’s selections among strategic options are made after
integrating facts with beliefs, their beliefs reflecting the manager’s attitudes
and values. For example, personality variables (values) such as risk
preference, daring and power can be argued relevant to the kinds of
strategies deemed feasible and preferable, such as whether to confront
competition or not, then that preference flowed on to the foreign markets
preferred. Whilst arguing this four step chain (personality values, strategic
values, strategy selected, market selected) is involved with the screening
decision, this dissertation does not operationalise it due to the complexities
which have yet to be finalised through further research.

4.3.4.2.2.3 Personal Values

 167

A number of paradigms can be argued to be relevant to isolating the

personal values involved with screening, but those in the Schwartz Value
Survey (S.V.S.) (1992) and Hofstede (1980) are appropriate. The ten
Schwartz dimensions were listed in section 4.2.4.2.2.3 (a). They and their 45
sub-components are also compatible with the five work-related ways in which
Hofstede (1984, 1994) suggests that cultures can differ (see for example,
Ralston et al. 1997, p. 191; Schwartz 1994a, p. 107, 1997). Other dimensions
could be substituted to differentiate the organisation’s capability to handle
cultural differences (e.g. composite of stage of internationalisation and EPRG
[Gould & McGillivray 1998]). The S.V.S. items (a) are richer at capturing the
almost infinite variety of human wants and (b) have the benefit of being
applicable to assessment of both cultural distance and some of people’s many
individual wants, and so could then be used to influence other aspects of the
screening framework. A caveat about using the Schwartz Values is that it is
not yet clear that they richly articulate some business aspects, such as
competitive, conflictual and completion values. Ros, Schwartz and Surkiss
(1999) and Schwartz (1999) begin that argument. Further, using Hofstede
yields the functional benefit of accessing cultural difference data on more
countries.

4.3.4.2.2.3 (a) Cultural Distance

An important attribute of both the Hofstede and Schwartz scales, unlike

work by Luostarinen (1980) and Ronen and Shenkar (1985), is that they are
ratio scales and so allow comparison of the relative differences between
countries along cultural dimensions. Reid (1986, p. 24) argues that cultural
distance is composed of affective, cognitive and conative dimensions, and
that it is predominantly influenced by social learning processes (p 31). This
implicitly links learning theory via psychic distance back to individual
capabilities and preferred learning styles, and to individual enculturation and
acculturation.
 168

Psychic distance is affected by acquiring specific markets’ knowledge

(Keegan & MacMaster 1983; Papadopoulos & Jansen 1994, p. 39; Reid 1986,
pp. 23-25), general international experience and mental programming from
the individual’s enculturation and acculturation. Perceived psychic distance
(real or imagined) can impede international business transactions by reducing
general willingness to trade with other cultures (i.e. affect the screening
decision) (Carlson 1975; Hallén & Wiedersheim-Paul 1989, p. 15;
Papadopoulos & Denis 1988, p. 44; Reid 1981, pp. 107-109). Undertaking
commercial transactions with other cultures feeds experience back to affect
the person’s specific markets’ knowledge, general internationalisation
knowledge, perhaps acculturation, and the stage of internationalisation in a
feedback loop which may alter psychic distance, commitment to a specific
market, commitment to internationalisation and the cost/benefit assessment of
internationalisation and/or a specific market. Psychological distance also
influences the strategy of standardisation or adaptation (Reid 1986; Shoham,
Rose & Albaum 1995, p. 14). Further, psychic distance is related to the
EPRG framework (the ethnocentric, polycentric, regiocentric or geocentric
attitudes of management) (Shoham, Rose & Albaum 1995, p. 14). Psychic
distance also affects the way business is transacted, such as the choice of
mode of entry (see, for example, Davidson 1980; Kogut & Singh 1988).

The literature thus recommends that a novice in international trade

should start with a culturally close market, gain experience, proceed to the
next closest market, gain experience, and so on (see, for example, Bilkey
1978, p. 43; Bilkey & Tesar 1977, p. 96; Davidson 1980, p. 18). Although the
cultural difference scales of Hofstede and Schwartz (above) are applied to
describe entire cultures, cultures are composed of sub-cultures comprising
individuals having different religious, ethnic and other aspects (as discussed
under enculturation and acculturation). As it is individuals who make business
decisions, it is the cultural distance between individual buyers and sellers that
is important (Brouthers & Brouthers 2001, p. 186; Hallén & Weidersheim-Paul
1989, p. 16). The cultural difference scales thus need to be applied to the
internationalising organisation’s key decision makers and/or key functional
 169

specialists to determine their general compatibility with each of the various

cultures being screened.

Although there is a solid body of literature supporting psychic or cultural

distance being an influential negative factor, three interesting irregularities
should be mentioned. Firstly, empirical work by Benito and Gripsrud (1992)
found no support for the belief that firms begin investing in culturally close
countries, nor that firms increase cultural distance with each new market
entry. Instead they found that, if the starting point is a distant cultural location,
the next investment will be closer to home, while the opposite is likely when
the starting point is a culturally close location (p 473). Second, the previously
described research by O’Grady and Lane (1996) suggested that a low cultural
distance can cause assumptions of cultural similarity that may retard learning
and so cause a high rate of failure of new international ventures. Perhaps this
is a case of familiarity breeding contempt. Thirdly, Morosini, Shane and Singh
(1998, p. 137) found that increased cultural distance increases performance of
cross-border acquisition performance. If this work is replicated, it could
suggest that there is a turning point on the cultural distance cost curve,
beyond which some internationally experienced organisations acquire benefits
such as innovation, entrepreneurship and decision-making proficiency.
Perhaps American CEO’s such as Bob Joss of Westpac and Paul Anderson
of BHP are analogies of this special case. Lastly, Brouthers and Brouthers
(2001, p. 184) conjecture that there may be a difference based on home
culture. Their reading of the literature suggests to them that high cultural
distance appears to cause U.S. firms to select joint venture, whereas
Japanese and Korean firms select wholly owned, modes of entry.
 170

4.3.4.2.2.3 (b) Risk Preferences

This item is indirectly partly measured by Hofstede’s ‘uncertainty

avoidance’ sub-variable when measuring cultural distance, but is separately
highlighted because of it’s importance in management’s decision-making.
Management attitudes toward risk taking in international business settings
have been investigated by Axinn (1988), Bauerschmidt et al. (1985) and
Cavusgil (1984a) and found to affect the organisation’s behaviours (Aaby &
Slater 1989, p. 17). Bilkey (1978) and Madsen (1994, p. 34) concur. Roux
(1979, p. 95) concluded “international orientation is associated with risk
preference, whereas domestic orientation is related to risk aversion”. Dichtl et
al. (1990) found risk preferences to be one of several factors that determined
managers foreign orientation (cited in Holzmüller & Stöttinger 1996, p. 34).
The work by Davidson (1983, p. 439) can be used to argue that perception of
risk is heightened by lack of information.

Risk is generally seen as a negative matter because it reduces

expected probable profit. However, the psychological makeup of people also
causes some of them to positively value risk to some individually variable
amount for the tension state it causes, the self-esteem and social acceptance
it can produce, etc. (Craig-Lees, Joy & Browne 1995, p. 157). Different
cultures also value risk differently (Hawkins, Best & Coney 1992, p. 40). Risk
is thus both ‘good’ and ‘bad’ depending on the manager, culture and
circumstance. This moderating variable is thus a major cause of the variability
in the independent variable risk level.

4.3.4.2.2.3 (c) Market Research Values

These can be revealed by psychological tests of an individual’s values,

perhaps using a test such as the Schwartz Value System (S.V.S.) (1992) in
future times. The S.V.S. is claimed to be a comprehensive inventory of values
(p. 37) so it should also be able to suggest cultural preferences towards or
against using market research. Pending development of S.V.S. data, this
dissertation questionnaires the key employees of the organisation about their
 171

evaluation of the worth of, and their frequency of use of, international market
research.

What are perceived by an organisation as the costs and benefits of

internationalisation is associated with its stage of internationalisation (see, for
example, Aaby & Slater 1989, p. 17; Leonidou & Katsikeas 1996, p. 537),
although the relationships are complex (Bilkey 1978, p. 35 quotes and
analyses nine studies). The change from advantages and disadvantages of
‘internationalisation’ to advantages and disadvantages of ‘further
internationalisation’ would not seem likely to change the relationships
substantially. Johanson and Vahlne (1977) would predict that knowledge of
the advantages and disadvantages of internationalisation obtained
experientially would be more highly valued, and used more often, than
knowledge obtained through classroom learning. Perlmutter’s EPRG theory
(1969) and later developments (e.g. Mayrhofer & Brewster 1996) would
predict that ethnocentric organisations would project their own values onto
foreign customers, reducing customer and product research, whilst polycentric
organisations would do more research because they emphasise cultural
differences.

4.3.4.2.2.3 (d) Other Personal Values

This is a catch-all category required for conceptual completeness since

cultural distance, risk preferences and market research values comprise only
a small portion of individual values. It is presently not operationalised due to
the absence of country-discriminating data.

4.3.4.3 Data Availability and Quality

Accuracy of a screening assessment depends on the organisation’s

ability to obtain reliable and comparable data (Toyne & Walters 1993, p. 300).
Bartos (1989) gives a series of interesting examples of data deficiencies in
international marketing research situations, including where different credible
agencies in different countries report different figures for the same variable.
 172

Considering the problems of data availability and quality is a frequent topic

among both domestic and international market researchers; the following are
just some of the many who have made interesting comments about problems
with international data: Armstrong (1970, p. 190); Bartos (1989); Cateora and
Hess (1979); Connolly (1987, p. 12); Douglas and Craig (1982, 1992); Evans
and Norris (1976); Liander et al. (1967); Liang and Stump (1996);
Papadopoulos and Denis (1988); Rice and Mahmoud (1984); Samli (1977);
Sheridan (1988, p. 23). Overcoming data problems is treated by far few
authors, but examples include Carr (1978); Douglas and Craig (1983, p. 131)
and Papadopoulos (1983), whilst further references are given by Russow
(1989, p. 212).

The Porter and Lawler (1968, p. 165) motivational model requires

assessment of the probability that effort will lead to reward. This invokes an
assessment of the availability and quality of decision data, because unreliable
data jeopardises the quality of the decision, then places at risk the
expenditure of effort for reward, i.e. management needs to trust the data
used to risk accepting the recommendation, then back it with scarce
resources. Low trust in the data will affect management motivation and
commitment to the next stage of the research process – the in-depth research
of the proposed market or markets.

4.4 Measurement of Dimensions

Each of the dimensions described in this chapter needs to be

measured before computations for market screening can undertaken.
Measurement requires production of specifications about dimension scales,
scores, feasible score ranges and determination of a data source for each
variable. Each organisation may have different objectives, abilities, values,
motivation and commitment, so requiring a questionnaire to collect this
information as primary date from each organisation. Data about countries is
most economically collected as secondary data, principally from reputable
non-government organisations.
 173

Appendix A provides a suggested Questionnaire to Management whilst

Appendix B provides measurement specifications for each dimension.
Appendix F summarises the data used. Although research for this
dissertation included endeavouring to locate existing questionnaire items to
use, no suitable instruments were located. The Questionnaire (Appendix A) is
therefore newly constructed and so has not had the prior use which allows
assessment of its instrument validity and reliability.

‘Low – High used’ (e.g. 0, 100) are two values used in the dissertation’s
computations to assess whether differences among organisations’ abilities, values,
objectives, motivation and commitment have an impact on which markets should
be selected by screening. The ‘High’ represents the maximum score any
organisation could achieve in the dimension. The ‘Low’ does not necessarily
represent the zero point on a dimension’s scale. Instead, it represents the
researcher’s estimate of the minimum feasible levels necessary for probable
success. For example, zero knowledge of some areas is impossible or highly
improbable.

The High and Low thus aim to represent the extremes of the continuums:

- the ‘Low’ is the researcher’s estimate of minimally appropriate

abilities, values, motivation and commitment of a low capability
organisation, an example of which is a small, newly formed,
moderately resourced, new-to-internationalisation organisation with an
un-differentiable product and no market-specific skills (e.g. foreign
language abilities), whilst

- the ‘High’ represents an organisation with maximum ability, maximally

supportive values and high motivation and commitment, typified by a
large, well resourced organisation with considerable international
experience and existing, market-specific abilities (e.g. foreign
language skills).
 174

Between the ‘Low’ and ‘High’ are believed to be all types of organisation,
including:

1) newly formed and old, established organisations,

2) conventional, primarily domestics and born globals,
3) new-to-internationalisation and experienced internationalisers,
4) a range of abilities from high to low,
5) a range of values from unsupportive to supportive, and
6) a range of motivations from high to low.

Using the synthetic data for abilities, values, motivation and commitment,
plus actual 1990 and 1995 market data, markets are selected for a Markets only
situation, the Low organisation and the High organisation for six products (18
solutions) to demonstrate whether there needs to be a solution for a specific
organisation – and whether it is perilous to assert that “the best market for firms to
sell product X in is ..… .” A demonstration is also provided of the impact of adding
objectives by computing markets for a Low and a High capability organisation
marketing one of the products. Identical objectives are used to allow exploration
of whether different markets suit organisations with identical objectives but
different capabilities and preferences.

4.5 Conclusion

This chapter has presented each of the variables used for international
market screening, firstly providing a plain-English description of the
dimension, then with arguments advocating its use from the literature. The
ways to measure each item are provided in Appendix B.
 175

CHAPTER 5

METHODOLOGY AND DATA

(With apologies to statistics)

MORIARTY: How are you at Mathematics?
HARRY SECOMBE: I speak it fluently.
Spike Milligan, The Goon Show

5.1 Introduction

The previous chapter presented a market screening framework, giving

a plain English description of each variable and theoretical justifications for
each variable. That framework requires empirical testing using a scientific
methodology and procedures.

This chapter describes the research design chosen and discusses the
related validity issues. It then traverses instrumentation matters, sampling of
the chosen products, data collection methodology, the variables which were
tested, calculates variable weights, and discusses the three statistical
procedures used.

5.2 Research Design and Validity

A quantitative experimental design was feasible, namely a one shot

case study which examined six cases. Country data were 1995, whilst the
second data point (required to measure the growth in the income per capita
component of the economic variable, and the growth in product-specific
 176

market size) was 1990. The number of countries in the six case studies
ranged from 42 (Internal combustion engines for cars) to 205 (Beef and veal).

A normal weakness of this research design is selection bias from non-

random selection of participants. This was eliminated by analysing all
countries that consumed each product, not a sample of countries. However,
an accepted design frailty is that the data years (1990 and 1995) will not be
likely to be typical of all time envelopes i.e. history and maturation affects may
render the research results a-typical. Some mitigation of this flows from
comparing six products.

The most significant affect on the internal validity of this research is the
matter of whether or not all possible influences on a country screening
decision have been captured. Significant effort was thus expended at the
beginning of the research to isolate all possible moderating and independent
variables – indicated by the large literature review that was undertaken. Over
350 literature items have been reviewed, including all 40 prior researches on
market screening. All influences hypothesised in the literature surveyed, and
additional matters stimulated by the literature and author experience, were
accumulated. This resulted in some 200 independent, moderating and
intervening variables being accumulated, then being either discarded as
illogical or amalgamated into the final 46 aggregate variables. It can thus be
said that there are no known, logically relevant, additional influences which
were not included in this screening framework. Further, that all 46 variable
concepts are supported by literature, so aiding content validity.

Of these 46, 38 could quantified (see Figure 4.3 and Appendix E).
Each of the eight non-quantified variables are partly encapsulated in one or
more of the 38 quantified variables. Two examples are: (1) the levels of
country-specific technological and ecological influences were not quantified
because these forces are already partly included in the levels of product sales
and sales growth in each country, and because these two influences are not
believed to have significant further affect on any of the six particular products
assessed in the research, and (2) country political influences are captured by
 177

the supplier of the data for the market risk variable. Whilst taking full account
of those 10 influences would be desirable, it was technically infeasible but
judged that the partial absences were unlikely to materially affect the research
results obtained. The results obtained bore out that expectancy.

Moving on to external validity, the other significant imponderable is

whether the six products selected are representative of most products. A
diverse sample of commodities was chosen by a two-step heterogeneous-
sample selection process which used strategic criteria and strata. As argued
below in ‘Products selected’, the results are probably indicative but not
statistical generalisable.

5.3 Instrument Selection, Development, Validity and Reliability

A requirement of market screening is that it principally uses secondary

data to minimise cost and maximise data quality and recency. Some 10,000
datum were necessary for the country selection stage, predetermining the
need to not require use of primary data. All secondary data used came from
reliable bodies (principally the United Nations and UN agencies). Cross
checks were made between much of the data acquired from these sources
and alternative sources (e.g. Australian Bureau of Statistics against UN coal
data, UN beef data against WTO beef data). Enquiries were also made with
each source about their assessment of the comprehensiveness, scaling,
reliability and validity of their data. These enquiries led, in one instance, to
selection of an alternative product (in the ETM28 product strata)29.

28
Extensively Transformed Manufacture.
29
The ETM product’s selection involved some work and required moving down the hierarchy
by size of Australian exports. “Computers and office machines, parts, etc.” are larger valued
Australian exports than the eventually chosen product, telephone sets. However, the UN has
been unable to collect computer production data from sufficient countries to be indicative
(personal e-mail communication dated 16 December, 1998 from Mr V. Romanovsky, UN
Statistical Office). “Telecommunications equipment n.e.s. and parts n.e.s.” is the often the
next largest Australian ETM export category group, but UN production data from numerous
countries is only available on the telephone sets sub-group. Whilst telephone sets are only a
small part of the telecommunications category, this data limitation was accepted in return for
obtaining a relatively homogeneous product sub-category of a large Australian export product
 178

The organisation-specific adjustments to the preferred markets involve

test instruments that are a mixture of existing (e.g. Hofstede’s 1980, 2001
cultural differences) and new (e.g. managerial internationalisation values).
Research was conducted in an endeavour to locate existing test instruments,
but without success30. The newly composed items therefore use instrument
design theory (e.g. Balian 1994; Churchill 1991; Zikmund 1994) to extract the
primary data needed but require several applications to assess their reliability,
validity and instrumentation affects.

5.4 Sampling

5.4.1 Product Sample Selection Rationale

There is a universe of some 60,000 distinct products to draw samples

from for this research (Downey et al. 1988). Service products number around
one third of these, but detailed statistics on services imports, exports and local
production by product and country are generally not available31. This absence
of services statistics concentrated the focus of this screening research on
merchandise goods.

The United Nations began collecting and cleaning the current

generation of trade data on tangible products (‘merchandise’ trade) in 196332.
Import and export data are thus available for some 30 years on approximately

type which had been selected on strategic grounds and which was statistically complete and
accurate.
30
Some near misses occurred. For example, the CETSCALE of ethnocentrism was found
(Bruner & Hensel 1996), but inspection of its 17 questions and independent evaluations of its
validity and reliability led to it not being adopted.
31
Personal Enquiries to the Australian Bureau of Statistics and the Australian Coalition of
Service Industries, October, 1996. The Australian Bureau of Statistics began trial publication
of data on a broader, but still limited, range of service products in 1997. A small number of
other countries did likewise at around the same time after several years of co-operative
discussions, but it will be a decade hence before potentially useful data become available for
numerous countries and products.
32
Commonly known as the UN SITC, the original version became effective in 1950 but
contained only 20,000 products. It was subsequently greatly revised (United Nations
Statistical Office (1963) Commodity Indexes for the Standard International Trade
Classification, Revised, Series M No 38 (2 volumes) United Nations, New York). Two
subsequent revisions (in 1981 and 1994) have made useful, but less major, alterations and
improvements.
 179

35,000 products, all tangible, in most countries (United Nations 1994, p. vi)
grouped into 3,121 basic headings (United Nations 1994, p. v). However,
product demand estimation requires not only trade data but also production
data. United Nations production data are based on a different standard, ISIC
(International Standard Industry Classification) Revision 2, the indexes for
which were published back in 1971. Different product definitions between
ISIC and SITC, along with the UN’s inability to move statistical reporting by
many countries to the more detailed ISIC Revision 3 (1990), presently results
in statistical voids on many products33.

Drawing a product sample of sufficient size to be representative of all

manufactured goods is ideal but thus problematic. To produce research
having a low confidence interval and high confidence level, a sample size of
some hundreds of products would need to be randomly drawn then analysed.
This number is beyond the resources of a doctoral dissertation.

The only equivalent work to the present research was conducted by

Russow (1989). He drew six randomly selected, unstratified, merchandise
items which fortuitously happened to represent the agricultural, industrial and
consumer products areas. Without arguing that his sample size was
representative, his methodology and subsequent analysis lead him to believe
that his work was probably indicative of most manufactured products. The
one possible exception was the pork products group, causing him to wonder
whether there could be some differences between agriculture and non-
agriculture related products (p 231). Naturally, a sample size of six is
insufficient to statistically represent all tangible products.

Statistical reliability of probability-based research can usually be

increased by stratifying the sample, then randomly selecting within each
strata. Strata ensure that there are some differences between sub-sample
groups but similarities within sub-samples. The presumption is that there may
be differences in behaviours among buyers and sellers in different strata, as
 180

suggested by marketing theory (e.g. Samli & Hill 1998, p. 156; Zikmund 1994,
p. 459). Stratifying the sample would be a methodologically useful step
toward generalisation of the results obtained from this research but would still
require a sample size which is beyond the resources of this dissertation.

Since random selection is discounted due to sample size needed, an

alternative approach is to choose products which are strategically important to
the country or organisations under consideration. Strategic importance can
be based on such matters as overall international business revenue received,
estimated future revenue, national priorities, marketing strategy, and so on.
The sample should ideally contain a diversity of product kinds, and this can be
brought about by using strata. Strata can be based on features of sub-groups
of either the product or its users. This methodology is a two-step
heterogeneous-sample selection process.

The discipline of marketing obtains stratification through customer-

oriented product classification schemes. Examples are: (1) the consumer -
industrial product groupings, and (2) the tangible - service products
dichotomy, in each of which there are further sub-groups, and (3) the product
class, form and brand categories (e.g. Kotler 1994, pp. 569, 592, 308). In
general, the more homogeneous the needs of the potential customers, the
more accurate the marketing program which can be designed to elicit their
responses. Relating this to selecting products for this research, the finer the
ISIC and SITC product definitions, the more likely the product users’ needs
will be homogeneous.

In contrast to marketing categorisations, merchandise statistics about

imports and exports are commonly grouped by the Australian and other
governments into: Primary Products, Simply Transformed Manufactures
(STM) and Elaborately Transformed Manufactures (ETM). The Primary
Products are then sub-divided into categories such as Unprocessed Food,

33
Statistical harmonisation continues to be only a minor item on trade negotiation agendas,
e.g. the 1999 APEC, New Zealand schedule.
 181

Processed Food, Other Rural, Minerals and Fuels34. These groupings relate
to economic views about transformation processes and to the stages of
development of nations. That is, they fundamentally represent a producer’s
view, rather than the marketing perspective of the buyer’s view, of categories
of goods35. Nevertheless, the government statistical categories reflect a valid
if self-centred perspective of homogeneity which permits product stratification
along their lines.

Finally, ETMs are of particular strategic interest to governments of

countries such as Australia because they employ relative large numbers of the
population, generate exports of a high value and are seen as having high
growth potential. Services are also a growth area, a high employer of labour
and an area which commonly allows use of higher technology.

Operationalising the aforementioned reasoning lead this work to

choosing the following five products:

• The largest overall revenue Australian export item, coal, was

selected;
• The Australian export product categories generating the highest
revenue in each of the three strata of food, STM and ETM in 1995
were also chosen. This caused beef, wool and telephone sets 29 to
be selected;
• The service product, tertiary education, was selected on the
grounds that it is a rare service product for which statistical data are
available, is a large export item, is a government priority area, and
is one about which the author and supervisors have deep
knowledge.

34
e.g. Composition of Trade, Australia, a bi-annual publication by the Australian Department
of Foreign Affairs and Trade, Canberra.
35
The literature makes this point in the general form by distinguishing between marketing
philosophies. Kotler (1994:10) suggests the five views are a production, product, selling,
marketing or societal marketing approach. Stanton (1994:15) sees four choices: production,
sales, marketing and social responsibility. The marketing and societal marketing approaches
are argued to be preferable for the firm to follow because they maximise an organisations’
long term rewards, including profit.
 182

The decision was made to assess the arbitrary number of six products,
so requiring another product to be selected. It was found that, if internal
combustion engines were selected, the product would serve multiple
purposes. It is the next highest ETM category to telecommunications
equipment, it has values which exceed the telephone sets sub-category, and
ETMs are a high priority area to the Australian government. Engines are also
one of those used by Russow (1989), so their inclusion would allow an
element of replication of, and extension to, the only work which is comparable
to the present research. Including this product would also seemingly allow
critique of:

1) whether Russow’s screening results apply to countries other

than the U.S.A. (i.e. determining if his results are source-country
neutral), and
2) whether or not adding variables to Russow’s demand-side only
factors causes a different market to be selected.

However, Russow’s data years were 1975 and 1980, whereas this
research uses 1990 and 1995, so there will not be direct comparability.

Inspection reveals that these six represent a fairly wide range of

product kinds:

• consumer - shopping (telephone sets, tertiary education),

• industrial - materials (coal, meat, wool),
- capital items (telephone sets, internal combustion
engines),
• commodity (coal, meat, wool),
• agricultural (meat, wool),
• manufactured (telephone sets, internal combustion engines) and
• services (tertiary education).
 183

In summary, the products selected lack statistical generalisability. A

diverse sample was therefore chosen by a two-step heterogeneous-sample
selection process which uses strategic criteria and strata.

5.4.2 Products Selected

Six products were selected for statistical treatment and assessment,

these being:

Coking coal (Australia’s largest export, a mineral product)

Beef (the largest Australian food product exported)
Wool (a large Australian STM36 export)
Telephone sets (a large Australian ETM28, 29 export)
Tertiary education (a services export item)
Internal combustion engines (another large ETM and included
by Russow (1989)37)

Product definitions will be found at Appendix D.

These products are imported, exported and/or produced in the

following numbers of countries:

Coking coal 52 countries

Beef 205
Wool 115
Telephone sets 116
Education services 193
I. C. engines 42

36
Simply Transformed Manufacture.
37
The ranking of Australian SITC Rev 3 product categories by export value can change each
year. After telecommunication equipment, it is not uncommon for the next highest
homogeneous category group to be internal combustion piston engines. Additionally, the
engines group is a higher valued group than the telephone sets sub-group.
 184

5.5 Data Collection Method

This dissertation’s framework matches an organisation with a short-list

of markets which have been tailored to the firm or organisation’s abilities,
values and objectives. This required collecting secondary data about all
markets and primary data about the organisation. Sources of data for each
variable are summarised in Appendix F.

5.5.1 Secondary Data

Product-specific data (production, imports and exports by country) were

sourced principally from reputable international agencies (coking coal from the
International Energy Agency [I.E.A.], beef and wool from the UN’s Food and
Agricultural Organisation, tertiary education from UNESCO, telephones from
the UN Statistical Office). Petrol engines statistics came from the commercial
organisation which specialises in collecting this data at source from the
world’s engine manufacturers (Power Systems Research). All these agencies
endeavour to produce quality statistics.

Three (probably slight) qualifications about the product-specific data

from the I.E.A., UN and UNESCO need mentioning. Decomposition of the
U.S.S.R. and Yugoslavia into some 22 independent nations between the data
points of 1990 and 1995 resulted in 1990 product data being apportioned from
1995 proportions. Second, for telephones only, because the UN collect
production data under a different statistical regime (ISIC Revision 2) from
trade data (SITC Revision 3), the product has nearly identical definitions, but it
is possible that some inconsistencies may be hidden in these data. Thirdly,
some of the trade data for telephones needed to be converted to physical
units from values at the average unit prices indicated by the other trade data.
Examination of input and output data suggests that these matters had little
impact on the results.

The data set for each country sometimes required estimates to be

made to complete the observations comprising a variable (e.g. PRS Inc
 185

supplied data for market risk in 130 of the world’s 230 countries so market risk
needed to be estimated for some 100 smaller countries; the World Economic
Forum (W.E.F.) provided data about 53 countries’ competitiveness so a
regression was constructed to estimate the other countries’ values; etc.). The
values for missing countries were estimated after assessment of the criteria
contained in the construct used to assess the countries (e.g. for market risk,
Coplin and O’Leary (1994) describe construction of PRS’ market risk index;
W.E.F. (1998, p. 79) describe construction of their index). Again, examination
of data suggests little impact from use of these estimates.

All the required secondary data were thus generated on 220 markets.
Full information was unable to be collected on the Falkland Islands, Guernsey,
Jersey, Isle of Man, Monaco, Northern Mariana Island, Pitcairn Island, St
Helena, San Marino and Western Sahara. They were thus deleted. These 10
markets account for consumption by around 600,000 persons, a trivial amount
of non-coverage of the six billion world population. The first four of those 10
deleted markets are British dependencies, and Monaco is independent but
bureaucratically linked with France, so most of their statistics are already
embedded in those of Britain and Italy. Western Sahara accounts for some
300,000 persons – almost half the population not included; it is believed to
have a very low GNP per capita (and so probably has low potential to buy
most products and services).

Data were then adjusted for timeliness, accuracy, comparability,

completeness and significance. This required that each cell of each country’s
data be adjusted for the researcher’s estimate of the cell’s datum quality in
each of these five aspects. For example, based on knowledge acquired
during collecting data from the UN Statistics Office and other data sources,
Afghanistan’s coal data were scored 1.00, 0.90, 0.81, 0.71 and 1.00
respectively for timeliness, accuracy, comparability, completeness and
significance. These data quality sub-scores were summed and divided by five
to produce a composite estimate of the country’s data quality (e.g. 0.88
composite for Afghanistan). Each cell of the country’s coal data was then
divided by this estimate so as to increase the cell’s data value – the rationale
 186

being that data imperfections should tend to increase the likelihood of the
country reaching post-screening evaluation. The view was and remains that it
is better to have extra countries in the final set of screened markets than to
unknowingly exclude a market having high potential.

5.5.2 Primary Data

This research aimed to test the two extremes of the range of

organisational abilities and values to demonstrate that a difference in markets
selected should result. It also compared the markets selected for those two
organisations with a selection which had no organisation-specific influences to
demonstrate the nonsensicality of selecting markets without considering
organisation attributes. It also tested, one at a time, the impact of the
organisation’s eight generic competencies and values to demonstrate
mathematically how their affect impacts on the markets selected. The four
different market selections which result are given the nomenclature of Markets
only, Low organisations, High organisations and Organisation with objectives
(as described below). Representative primary data were thus required for the
two organisation types.

Chapters three and four used theory to argue that there are clusters of
competencies and preferences which are relevant to further
internationalisation. These attributes have the capacity to provide efficiency
and effectiveness, so producing such beneficial effects as more rapid market
penetration, higher levels of profit and lower risk. Eight of the clusters of
competencies and values could be quantified, and are shown as black (not
grey scale) variables in Figure 4.3.

Theory, experience and intuition lead to development of a plausible

questionnaire which yielded scores for different levels of competencies and
values in the eight areas. This Management Questionnaire is Appendix A.
That questionnaire has not been tested for validity and reliability as its
purpose is to demonstrate methodology yet retain a quantum of work that is
possible within doctoral research.
 187

When selecting sub-dimensions and setting scores in the

questionnaire, the following types of organisation were considered: (1) newly
formed and old, established organisations, (2) conventional, primarily-
domestic organisations and born-globals, (3) new-to-internationalisation and
experienced internationalisers, (4) a range of abilities from high to low, (5) a
range of values from supportive to unsupportive, and (6) various levels of
motivation from high to low.

Each sub-variable on the questionnaire was scored for two hypothetical

organisations: a minimally competent organisation and a maximally competent
one. ‘Minimal’ is not necessarily zero on many scales: it is often impossible to
have no knowledge or no values (for example, to survive, any organisation
must have some general business competencies). The minimal score thus
represents the estimated minimum necessary for probable success.
Maximum competence or value is always the highest score possible. Variable
and sub-variables’ scores are detailed in Measurement of Dimensions whilst a
summary of variable values will be found at Appendix F.

Assumptions about the Low ability organisation’s generic competencies

were that it had an exportable but “me-too” product (minimal product
differentiation and generic SCA), moderate general business competence, no
pre-existing international experience, and only partial stakeholder support.
Managerial values included a moderately strong risk preference, and modest
levels of motivation and commitment to further internationalisation (on the
assumption that they could not be highly motivated about, and committed to,
something they had not experienced). Market-specific competencies were that
its staff could speak only English, and that it had no country-specific SCA’s, no
foreign contacts and no existing market research.

Assumptions about the generic abilities of the maximally capable

organisation were of high product differentiation, very strong strategic
competitive advantage, considerable general business competence and
further internationalisation competence, and very strong support from key
 188

stakeholders. Managerial values included a high level of risk tolerance, and

high levels of motivation and commitment. Market-specific competencies
were language skills in the 10 main languages (see Appendix G), SCA in a
randomly selected 30 per cent of the world’s countries consuming the product,
contacts in a randomly selected 30 per cent of countries consuming the
product, and current market research in five randomly selected countries
consuming the product.

Data would normally be adjusted for quality, in a similar manner to that

described for secondary data, using input per Appendix C. This was not done
to the primary data in this research. The methodology has been
demonstrated earlier.

5.6 Variables Tested

This section presents and categorises the variables which were used or
not used during the three kinds of market assessment for each of the six
products.

Figure 4.3 summarises the 46 matters indicated by the literature to be

relevant to market screening. Each matter is comprised of either a single
variable or a cluster of sub-variables. Appendix E lists and categorises all
variables. Of the 46 variables, 38 are quantifiable, although eight quantifiable
matters were not required for the six particular products assessed. The
numbers of variables entering SPSS were seven, eight and 11 respectively for
the Markets only, Low organisation and High organisation solutions. The
eight Generic variables and four Objectives variables were used to weight the
eight / 11 Low / High organisation market variables, and so influence the
market selections. Dependent variable is the organisation-specific short-list of
attractive and preferred markets.

As well as the three key selections (best Markets only, Low

organisation and High organisation markets), market selections were also
made for each product based on each individual generic variable. This
 189

involved eight selections for each of the six products, a total of 48, the
purpose being to assess whether the generic variables contributed individually
to unique solutions both individually and collectively.

The moderating / independent / intervening variables not analysed

during this work were:

• Potential customer needs, Segments, Segment population sizes

and Intrinsic adoption rates, which are matters that explain demand
causation. They are necessary for conceptual completeness, and
have practical use when estimating product consumption and
product consumption growth if accurate data on the latter two are
not available. Since product consumption data of quality were
available, these four variables did not need to be activated,

• Physical, Technological and Ecological/ environmental, which were

assessed as not relevant to the particular products investigated by
this research,

• Targets, Strategy decisions, Marketing mix, Strategic values, Other

personal values and the Objectives of Personal benefits and Other
business benefits are complex, multi-dimensional variables. The
researcher believes that these have yet to be automated in a
quantified framework such as the one presented here. They can
presently only be applied during the in-depth assessment stage,

• Customer adoptions and usage rate, and Organisation’s product-

specific demand can be activated if country-specific adoption rates
are known. Where the organisation knows the likely penetration rate
for its product in each market, it can convert the total consumption
rate by country to an expected sales level by the organisation in each
country. The research herein presented on six representative
 190

products could have produced demonstration calculations but left

those for the hypothetically subsequent in-depth stage.

There are several moderating / independent / intervening variables

which were used, but are hidden in differing ways:

• Political was already included by the PRS Group in their data on

Market risk,

• As described above, adjustments were made to the secondary data

on markets to allow for their data quality. Similar adjustments would
normally be made to the primary data on the organisation’s generic
and market-specific variables for their data quality,

• Size of assessment task and Criteria and method of evaluation are

two technical areas which determine the kind of screening
undertaken. The decisions taken within both determine the
fulsomeness and quality of screening.

This sub-section ends with two summaries of the variables used.

 191

Table 5.1
Four Types of Independent, Moderating and Intervening Variables Used
to Generate Three Kinds of Market Selection

Variable type Select Select Select

Markets Markets for Markets for
only Low High
capability capability
organisation organisation
Markets only 7 variables 7 variables 7 variables
Organisation-Markets
variables 1 variable 4 variables
Organisation only (Generic)
variables 8 variables 8 variables
Organisation objectives * 4 variables 4 variables
Totals 7 variables 20 variables 23 variables

(SCA (organisation’s generic))

(General competencies)
(Further internationalisation competencies)
(Key stakeholder support)
(Further internationalisation values)
(Risk values)
(Market research values)
(Motivation & commitment)
[Data availability & quality]
SCA (organisation’s market-specific)
Language
Market research
Contacts
Product consumption
Product consumption growth
Economic
Competition
Legal fairness
Cultural difference
Market risk
Political
(Data availability & quality)
[Physical, Technological, Ecological]

* Obtain sales, Ongoing sales growth, [Obtain other business benefits], Take little
risk, Require little effort

(xxx) = Used to weight values of the un-bracketed variables

[xxx] = Not relevant in this research, but may be to other products and organisations

Source: Developed for this research based on analysis of the literature

 192

Table 5.2
Test of the Impact of Each Generic Variable,
Using Three Types of Independent, Moderating and Intervening Variables

Select Select
Markets for Markets for
Low High
capability capability
organisation organisation
Markets only 7 variables
Organisation-Markets variables 1 variable 4 variables
Test 8 organisation-only (Generic) variables, 1
at a time 1 variable 1 variable
Totals 9 variables 12 variables

It is important to note in Table 5.1 that Markets only, Low organisation

and High organisation are each based upon different quantities of variables.
This comes about because: (1) the Markets only solution involves neither
Organisation-Market variables nor Organisation only variables, and (2) the
Low organisation solution has zero scores for three of the Organisation-
Markets variables (market-specific SCA, Market research and Contacts).

Moving to Table 5.2 each of the eight generic variables is shown being
applied, one at a time, to test whether they contribute to markets selected for
the Low and the High organisations. These five different quantities of
variables produce five different weights for variables, so influencing in a
significant way the markets selected.

5.7 Variable Weights

Each time the number of variables changes, so too does the weight of
each variable. Thus a variable such as Product consumption may have five
different weights38 before it is re-weighted a sixth time by the organisation’s
objectives. With the weight of each variable changing, each market’s unique

38
For Markets only, Low firm with Firm-Markets variables, High firm with Firm-Markets
variables, Low firm with generic variables and High firm with generic variables.
 193

total score changes, so it is likely that six different groups of countries will be
suggested.

Statistically handling the eight clusters of competencies and values is

challenging because many of the variables which differentiate the two kinds of
organisations are not related to specific markets. For example, a high level of
General business competencies causes the organisation to perform better at
marketing, management, finance and production, yet does not directly lead to
a preference for particular markets. A system of market variable re-
weightings was ultimately used to handle the matter, using weightings which
were determined by responses to the questionnaire to management.

An alternative to re-weighting the eight (Low organisation) or 11 (High

organisation) market variables is to impose thresholds for the market to
exceed before including it in the set to be considered (e.g. eliminate markets
having less than $ X GDP per capita). However, variables interact, such that
a strength in one may compensate a weakness in another, and there is
presently no knowledge about where to set the thresholds. Another
alternative to weighting variables is to look for observable breaks in the data
then eliminate markets either side of the break point. However, there are (at
least) three disadvantages of this. The first is that there might not be any such
breaks. Secondly, and conversely, there may be a number of breaks and
selecting the most appropriate (presumably the largest) might be difficult.
Thirdly, and more seriously, this approach is too mechanical, being driven
purely by the data rather than knowledge, conceptual or otherwise, of an
appropriate threshold.

Weights were determined as follows. The Markets only solution

assumes an organisation of median ability. Since a median is always half the
total, the scores for a High organisation need to be doubled. The country
scores on the seven variables which determine the Markets only solution
(which are on a 0 to 1 scale) therefore get doubled (and become a 0-2 scale)
for the High organisations. An organisation having perfect High competence
 194

and values would then be attracted to countries scoring 2 on each of the

seven Markets only variables.

Similarly, a completely imperfect Low organisation would be most

suited to countries scoring 0 on the 0-2 scale. However, no organisation is
capable of having zero knowledge or values. The questionnaire (Appendix A)
results in Low organisations achieving a minimum of .36875 of the High
organisation’s score. Thus the seven Markets only variables need multiplying,
first by two (to get from median to High) then by .36875 (to get to the Low
organisation’s ideal market).

These last two paragraphs can be expressed algebraically as follows:

Low organisation coefficient for each of the 7 Market variables

= 2 (Σ scores from questionnaire re 8 generic matters) / 8
= 2 (.30 + .35 + .13 + .30 + .54 + .50 +.50 +.33) / 8
= 0.7375 (or 2 x .36875)

High organisation coefficient for each of the seven Market variables

= 2 (Σ scores from questionnaire re 8 generic matters) / 8
= 2 (1 + 1 + 1 + 1 + 1 + 1 + 1 + 1) / 8
= 2.0 (or 2 x 1)

The seven Markets only variables each get multiplied by each of these
coefficients for the Low organisation and the High organisation, so changing
each country’s score, and thus the relative attractiveness of each.

An implication of this approach is that an organisation’s competencies

and values can be averaged out – that a low in one aspect can be
compensated by a high in another. The psychological literature generally
supports that view, although there may be lower limits on some components.
The minimum necessary hurdle levels were taken into account when
constructing and scoring scales for the Management questionnaire (Appendix
 195

A) although it was the researcher who set the minimum scores as the
literature has yet to provide final answers to both the criteria and levels
needed for international success.

The four variables relating the organisation’s competencies to specific

markets (‘Organisation-Market’ variables of market-specific SCA, Language,
Market research and Contacts) do not get multiplied by the ability weights of
either .7375 or 2. This is due to their being uninfluenced by the organisation’s
generic competencies levels.

A final task is to bring into play the variables which actualise the
organisation’s Objectives. The Management questionnaire’s Objectives
section assesses the relative importance to the organisation of five areas.
Four of those are implemented, namely those of maximising Sales and Sales
growth, and of minimising Risk and Effort. A fifth goal, Other business-related
benefits, is not synthesised here but could be included using the methodology
provided. The questionnaire responses are used to weight the eight and 11
market variables. This allows production of separate Low organisation and
High organisation screened lists of markets which have been weighted for
unique organisational objectives.

Let ym = the weight that expresses the strength of desire to achieve

each of the organisation’s objectives, and
xn = the related determining variable.

For demonstration purposes, let organisational objectives be the following:

 196

Table 5.3
Demonstration Organisational Objectives Weights

Relative objective level

Low capability organisation = ym = Objectives
High Sales objective = y1 .55
High Sales growth objective = y2 .05
Low Risk objective = y3 .20
Low Effort objective = y4 .20
Total 1.0
High capability organisation
High Sales objective = y1 .35
High Sales growth objective = y2 .35
Low Risk objective = y3 .15
Low Effort objective = y4 .15
Total 1.0

Objectives’ values above have been selected on the basis that they
represent, in this researcher’s experience, the kinds of goals set by Low and
High organisations at the beginning and end of their cycle of
internationalisation learning. At the beginning, organisations generally require
rapid payback to justify continuing the new behaviour to stakeholders. High
sales are thus essential. There is little interest in long-term rewards (growth in
sales), a perception that the organisation will need to take some risk to
succeed at the new endeavour, and little desire to divert energy from existing
activities. In contrast, an organisation that is close to maximum international
learning has already tapped the high sales markets, so expectancies about
large immediate international sales are lower. The organisation also knows
that it must anticipate future sales growth by looking for growth markets (the
sales growth objective). It will be unwilling to take huge risks to gain a
marginal increment in business – much of which will be in the future. It
already has a large international portfolio of markets – providing a momentum
from which it does not need to divert much energy (effort) to enter the new
market.

The values set here, and throughout this research, are an attempt to
represent the ends of the relevant continuums. Another researcher’s
experience may suggest slightly different values for objectives, but it is
 197

believed that a similarly experienced international researcher would pick fairly

similar values to those above for the two situations depicted.

Prior to adjustment for organisational competence or objectives, we have:

Organisation’s Determining Market Weighted

Objectives (weights) Variables Variables
High Sales = y1 Product demand = x1 y1 . x1
High Sales growth = y2 Product demand growth = x2 y2 . x2
Low Risk = y3 Market risk = x3 y3 . x3
Low Effort = y4 (1st Economic PC = x4 + y4 . ( x4 +
(Competition = x5 + x5 +
(Legal fairness = x6 + x6 +
(Cultural difference = x7 + x7 +
(Language abilities = x8 + x8 +
(SCA = x9 + x9 +
(Market research = x10 + x10 +
(Contacts = x11 x11 )

Now, the decision for the High organisation is made using 11 Market
and organisation-market variables, so each variable has a weight of 1/11th.
However, the Low organisation has zero values for three of the organisation-
market variables (SCA, Market research and Contacts), so each Low
organisation variable has a weight of 1/8th.

Therefore, values for each Objective are:

Objectives Low organisation High organisation

Sales .55 [x1 ] / 8 .35 [x1 ] / 11
Sales growth .05 [x2 ] / 8 .35 [x2 ] / 11
Risk .20 [x3 ] / 8 .15 [x3 ] / 11
Effort39 .20 (.2 [x4 + x5 + x6 + x7 + x8]) / 8 .15 (.125 [x4 + x5 + x6 + x7 + x8 + x9 + x10 + x11]) / 11

From previously in the dissertation, adjustments to each of the seven

Market Only variables for competence and values of the organisation were:
.7375 (Low organisation), and
2.0 (High organisation)

39
Effort is composed of 5 sub-variables for the Low organisation or 8 sub-variables for the
High organisation. Each sub-variable is therefore divided by either 5 or 8, then the 5 or 8
sub-variables summed to determine Effort. There is thus an additional number in each of the
Effort equations – these have been made bold for ease of identification.
 198

These adjustments do not apply to country-specific Language (x8), SCA (x9),

Market research (x10) or Contacts (x11) because all four are already scaled for
organisation competence.

Thus the weighted market variables are:

Low organisation:
Sales = .55 {.7375 [x1 ]} / 8= .051 x1
Sales growth = .05 {.7375 [x2 ]} / 8= .005 x2
Risk = .20 {.7375 [x3 ]} / 8= .018 x3
Effort = .20 (.2 {.7375 [x4 + x5 + x6 + x7] + x8}) / 8
= .004 [x4 + x5 + x6 + x7] + .005 x8

High organisation:
Sales = .35 {2 [x1 ]} / 11 = .064 x1
Sales growth = .35 {2 [x2 ]} / 11 = .064 x2
Risk = .15 {2 [x3 ]} / 11 = .027 x3
Effort = .15 (.125 {2 [x4 + x5 + x6 + x7] + x8 + x9 + x10 + x11}) / 11
= .003 [x4 + x5 + x6 + x7] + .002 [x8 + x9 + x10 + x11]

It may be worthwhile noting that, after multiplying out that complex

algebra, the relative (not absolute) coefficients for each variable used on each
country when computing the Education with Objectives solution end up being:

Low High
organisation organisation
Product demand .55 x1 .35 x1
Product demand growth .05 x2 .35 x2
Market risk .20 x3 .15 x3
1st Economic PC .04 x4 .01875 x4
Competition .04 x5 .01875 x5
Legal fairness .04 x6 .01875 x6
Effort

Cultural difference .04 x7 .01875 x7

Language abilities .04 x8 .01875 x8
SCA - .01875 x9
Market research - .01875 x10
Contacts - .01875 x11
Totals 1.00 1.00000
 199

This section ends with a conceptual summary of the various weights

applied to variables.

Table 5.4
Summary of Weights
for Each Independent, Moderating and Intervening Variable
Used to Generate the Seven Kinds of Market Selection

Variables used Select Select Markets Select Markets

Markets for Low for High
only capability capability
organisation organisation
Markets only variables 1/7 ea variable
Markets + Organisation- 1/8 ea variable & 1/11 ea variable &
Markets + Generic (7 Markets only (7 Markets only
(organisation only) variables variables x2x variables x 2 x 1.0)
.36875)
Test 8 Organisation-only 1/8 ea variable & 1/11 ea variable &
variables (7 Markets only (7 Markets only
variables x 2 x .13 variables x 2 x
to .54 ++) 1.0 ++)
Markets + Organisation- 1/8 ea variable & 1/11 ea variable &
Markets + Organisation- (7 Markets only 7 Markets only
only + Objectives variables variables x 2 x variables x 2 x 1.0)
.36875) & each x & each x objectives’
objectives’ weights weights

++
.13 to 2 is the range of scores for the generic competence and values
variables. See Organisation’s capabilities and preferences section for
derivation of these figures.

5.8 Statistical Analysis

Some data preparation and manipulation were undertaken in Excel 97,

then SPSS 8.0.2 and 10.0.5 were used for most statistical calculations.

The following statistical techniques were employed for the tasks and
reasons detailed shortly:

• Principal components analysis (PCA),

 200

• Cluster analysis (CA), and

• Multiple discriminant analysis (MDA).

5.8.1 Principal Components Analysis (PCA)

Principal components analysis (PCA) was used to reduce the four

economic sub-components to a single variable (a single component) that
represented multiple, comparable economic facets of each market.
Tabachnick and Fidell (1996) put it thus: “PCA is the solution of choice for the
researcher who is primarily interested in reducing a large number of variables
down to a smaller number of components” by providing “… an empirical
summary of the data set” (both p 664). PCA extracts maximum variance from
the data, producing a first principal component that is the linear combination of
observed variables which maximally separates markets.

The economic sub-variables so treated were:

• log of GNP PPP per capita (Gross National Product measured in

current international dollars by the Purchasing Power Parity
method) 1995,
• growth in GNP per capita, measured by the shift-share change in
GNP PPP per capita between 1990 and 1995,
• population 1995,
• openness of each economy (imports minus exports divided by
GDP) 1995.

Prior to running PCA, data needs to be assessed to ensure suitability

for treatment (Tabachnick & Fidell 1996, pp. 640-2):

• variables need to be standardised to use the same units of

measurement,
 201

• sample size needs to be adequate – not relevant to this research as

markets analysed comprise the population, not a sample40,
• multivariate normality of data is required if dealing with samples of
population – again not relevant, as above,
• outlier cases are identified and removed or given special treatment
so that a-typical members do not distort the inference;
• factorability of the matrix is indicated by several sizeable
correlations, Bartlett’s test of sphericity and the anti-image
correlation matrix.

There is a difference between the statistical literature and the economic

literature about correlation matrix requirements for factorability of a matrix.
The statistical literature’s usual advice is that there should be several sizeable
correlations over 0.3 minimum (e.g. Hair et al. 1995, p. 374; Malhotra et al.
1996, p. 534; Tabachnick & Fidell 1996, p. 641), and that the Kaiser-Meyer-
Olkin measure of sampling adequacy be over a minimum of 0.6. However,
the aim of this research was to combine four economic variables into a single
composite indicator, and the international development economics literature
regularly employs PCA for this use (e.g. see Ogwan 1994; Ram 1982 for
composite development indicators).

5.8.2 Cluster Analysis (CA)

Cluster analysis (CA) was used to sort markets into groups having
differing levels of potential, isolating especially those of high potential.
Inspection of the relative centroid values of each predictor variable indicates
which cluster is preferred in respect of a single variable. Market potential of
each group is indicated by the value of the group centroid for each variable.
The analyst conducts tests on the CA solutions in MDA then returns to the CA
output to obtain the high potential countries.

40
Every country which used the product was treated. Suppliers of the data – the Australian
Bureau of Statistics, Food and Agricultural Organisation, International Energy Agency, United
Nations, UNESCO and mostly similar ability agencies – indicated that the data sets were
complete.
 202

CA is a statistical procedure which classifies individuals or variables

into a small number of mutually exclusive groups, maximising likeness within
groups and maximising difference between groups (e.g. Malhotra et al. 1996,
p. 570; Zikmund 1994, p. 692). CA has been used by Goodnow and Hansz
(1972), Litvak and Banting (1968), Sheth and Lutz (1973), and others through
a lineage to Russow (1989) for similar purposes to those undertaken in this
research.

There are many ways to calculate the within-cluster and between-

cluster distances (see Malhotra et al. 1996, p. 560 for a diagram of merely the
14 CA procedures they consider to be useful and common). Each method
can produce different numbers of clusters and different cluster memberships.
Hair et al. (1995, p. 442) and Malhotra et al. (1996, p. 562) therefore suggest
employing a hierarchical method and a non-hierarchical method in tandem to
gain the benefits of both. A hierarchical method is used to specify the number
of clusters and the cluster seeds for the non-hierarchical method. Ward’s
method: (1) is a commonly recommended, generally good-performing,
hierarchical method of clustering (e.g. Aldenderfer & Blashfield 1984; Churchill
1995; Everitt 1993, p. 142; Hair et al. 1995; Malhotra et al. 1996), and (2) was
used by the previous quantitative screening researcher, Russow (1989). K-
means is “the most popular” non-hierarchical method according to Churchill
(1995, p. 999), or at least a commonly recommended method (e.g.
Aldenderfer & Blashfield 1984; Hair et al. 1995, p. 442; Malhotra et al. 1996).

This research thus used Ward’s method, at times comparing the results
from it with results from the average linkage method41 to gain insight into
different agglomeration behaviour. The number of clusters and centroid
values generated by Ward’s method were then used to run K-means parallel
threshold CA, so obtaining the benefit of both hierarchical and non-
hierarchical CA. K-means CA data were then used as the input to MDA.
 203

The number of clusters recommended by each method is often

indicated by inspection of the agglomeration schedule, the dendrogram and
the scree plot. However, there are no hard and fast rules (e.g. Churchill 1991,
p. 905; Hair et al. 1995, p. 442). This research required running
approximately one thousand CAs and at times encountered inconclusive or
conflicting assessments of the number of clusters seemingly optimal. In those
cases, the number of clusters was selected using the following rules:

• Following the Ward’s method CA run, centroids for two to 10

clusters were calculated, then K-means CA was run for the two to
10 clusters;
• MDA was then run for the nine solutions (2 to 10 clusters) and the
number of clusters selected which optimised significance of the
functions (≤ .05 for the worst of the several functions). This was a
dominant, go/no-go matter. When several MDA’s exhibited similar
function significance results (typically all .000), preference was
given to the solution which had:

- the higher percentage of countries cross-validated, then

- the higher number of significant variables, and the lower
number of pairs of variables correlated (bivariately and
multivariately), then
- if a two cluster solution appeared qualitatively the same as
a solution based on a higher number of clusters,
preference was given to the non-2 cluster solution. The
rationale is that a two cluster solution forces variance from
all countries, including exceptional ones, into the
regression coefficients and so produces an artificially good
.000 significance for the single discriminant function. A
larger number of clusters allows segregation of higher
variance countries into separate clusters, so producing a
more pure collection of high potential markets.

41
Another widely used, general purpose hierarchical method having minimal statistical side
 204

Sometimes the screened list of markets is too large to be helpful (e.g. it

might be 50-100 markets in a 200 market set having a three cluster solution).
Data and data properties need to determine the number of markets in the high
potential screened set, whereas practitioners require a set of approximately
six markets on which to conduct in-depth assessment42. There is no known
way to hierarchy the high potential sub-set of markets or to otherwise
overcome this limitation of CA.

5.8.3 Multiple Discriminant Analysis (MDA)

Multiple discriminant analysis (MDA) was used in this research to:

1) test the number and composition of clusters indicated by CA,

and
2) assess the relative importance of each screening variable – a
major interest of this study.

Multiple discriminant analysis (MDA) models the relationship between a

dichotomous or multichotomous criterion variable and a set of predictor
(independent) variables (Churchill 1991, p. 1038), identifying the relative
importance of each independent variable (Malhotra et al. 1996, p. 508). The
stepwise method of MDA was used, this being deemed more appropriate than
the alternative, simultaneous (direct) method of MDA, due to the research aim
of testing the necessity for a large number of predictors and, if possible,
reducing the size of the set (e.g. Hair et al. 1995, pp. 197-8; Malhotra et al.
1996, p. 510).

To test the accuracy of MDA results, the population of markets which

consume a product can be divided in two: the first portion is then used to
calculate the coefficients of the discriminant function, then those coefficients

affects.
42
Each additional market on which in-depth market research is conducted can easily cost an
organisation $10,000, so there is pressure to minimise the size of the list of markets.
 205

are used to assign the second portion of the markets population to groups.
Alternatively, cross validation is performed by each country being classified by
the functions derived from all cases other than that country. This latter
method is the standard procedure used by SPSS, and so was used in this
research. The forecasts of membership by group for each country are then
compared with the assignments by the original discriminant functions, and
with the likelihood of those assignments occurring by chance.

Prior to interpreting MDA results, therefore, the discriminant function or

functions need to be determined to be statistically significant and the hit ratio
(or classification accuracy) needs to be judged to be acceptable (e.g. Hair et
al. 1995 chapter 4; Malhotra et al. 1996 chapter 20; Tabachnick & Fidell 1996
chapter 11). Cross-validating CA results with MDA thus involves examining:

1) the Wilks’ λ (lambda) value, which indicates whether the cluster

centroids are statistically different from each other. (Wilks’ λ is
rarely interpreted directly, instead being calculated then
transformed to the more commonly used F statistic), and
2) the hit ratio, the accuracy of the discriminant coefficients at
assigning the markets to the population, indicates the difference
between CA and MDA in assigning markets to the same
clusters. The hit ratio can be compared with the statistic for the
same assignments having been made purely by chance, and
authors typically suggest that it should be at least 25 per cent
greater than chance (e.g. Hair et al. 1995, p. 204; Malhotra et al.
1996, p. 516).

When there is only one discriminant function, the relative importance of

each screening variable is indicated by the magnitude of the standardised
discriminant function coefficients. The predictive accuracy of these
coefficients is suggested by:
 206

1) the size of the hit ratio (being at least 25 per cent greater than
chance, as above), and
2) whether the pooled within-group correlation matrix indicate that
multicollinearity appears to blunt the coefficients’ crispness. By
convention, 0-.33 / .33-.7 / .7-.9 / .9-1 are interpreted as
indicating an unlikely / possible / likely / very likely association
between variable and function or variable and variable (e.g.
Tabachnick & Fidell 1996, p. 677 for a more detailed
classification scheme).

When there two or more discriminant functions, no single method gives

an absolute answer (Tabachnick & Fidell 1996, p. 541; SPSS 1999, p. 268).
Examination of all of the following will yield insight into the relative importance
of each variable, although the methods sometimes provide conflicting
assessments:

• rotated discriminant loadings (structure matrix),

• the potency index,
• the size of the univariate F ratio.

Rotated discriminant loadings are the pooled within-groups correlations

between discriminating variables and standardised discriminant functions.
Variables can be ordered by absolute size of correlation within functions, then
the largest absolute correlation between each variable and any discriminant
function can be located. The ranked order of importance of each variable is
the result of this procedure. A subsequent task is to eliminate any variable
whose correlation falls below a predetermined level – the commonly used
± .33, which has a statistically unlikely level of association (mentioned above),
was used in this research. A final task is to eliminate variables of insignificant
importance by eliminating those whose principal correlation is with a function
which has an eigenvalue ≤ 1.0 and/or those whose principal correlation is with
a function which accounts for less than 5 per cent of variance. (Hair et al.
1995 chapter 4.)
 207

The potency index is a composite measure of the discriminatory power

of an independent variable – and thus its importance relative to other
variables. It includes both the contribution of a variable to each discriminant
function and the relative contribution of the function to the overall solution.
The contribution of each variable to each discriminant function is calculated by
taking the squared value of the discriminant loading, then multiplying by the
relative eigenvalue of that function. The potency index for each variable is
then the sum of the individual potency indices across all significant
discriminant functions. The resulting index allows ranking of variables in order
of importance, but no meaning can be interpreted from the absolute values.
(Hair et al. 1995, pp. 207-8).

The F statistic measures the ratio between one sample variance to

another sample variance, such as the variance between groups to the
variance within groups. It is used to determine whether the variability of two
or more samples differ significantly (Hair et al. 1995, pp. 119, 206; Zikmund
1994, p. 629). To assess relative importance, all independent variables can
be forced to contribute to selections by using stepwise MDA and changing
from default F statistic cutoff levels of 2.84 entry and 2.71 removal to the
probability of F being 1.0 at both entry and removal. The absolute size of
partial F values allows ranking of independent variables, so indicating the
relative discriminating power of each variable.

The above three calculations should be made for each independent

variable, the variables ranked by each method, then comparison made of the
rankings (Hair et al. 1995, p. 209). When uniform rankings result, there is
high certainty about relative variable importance. When variable hierarchies
vary by method, the analyst is left to determine how to interpret the meaning.
 208

CHAPTER 6

RESULTS

Round numbers are always false.

Samuel Johnson (1709-84), Life of Johnson, Vol II (J. Boswell)

6.1 Introduction

Using the methodology, data and procedures described in Chapter five,

what results ensued?

Chapter six presents the results, which are sequenced in the order that
the statistical routines needed to be carried out: principal components
analysis, cluster analysis and multiple discriminant analysis.

6.2 Principal Components Analysis (PCA)

After all input data were standardised, the first economic principal
component was extracted from the four sub-variables, then was used as one
of the variables to select markets. This variable represents the economic
climate in each market and is named the First economic principal component
– abbreviated to 1st economic PC.

A different set of countries was involved for each of the six products, so
necessitating a unique economic component being extracted for each product.
Additionally, data from China or the USA at times exceeded multivariate
outlier levels, necessitating three additional principal economic components
being calculated. There are thus nine 1st economic PCs. Outlier countries
 209

were assessed using Mahalanobis distances from their group mean

calculated as Chi square (χ2) at p < .001 with degrees of freedom equal to the
number of variables (Tabachnick & Fidell 1996, p. 94). For example, in an 11
variable analysis, χ2 (11) p < .001 = 31.264, the score above which countries
are excluded as outliers in those analyses.

The characteristics and quality of principal components can be partially

indicated by such matters as their eigenvalue levels, the amount of variance
they capture, the variance that the sub-variables share with each other, and
the rotated correlations between their sub-variables and their components.
These statistics are now presented.
 210

Table 6.1
Total Variance Explained
The same principal component was extracted and used for all situations for the product,
unless otherwise specified in the notes below

Component Initial Eigenvalues Initial Eigenvalues

Total % of Cumulative Total % of Cumulative
Variance % Variance %

Coking coal (N = 52) Coal No China *1

1 1.847 46.182 46.182 1.833 45.827 45.827
2 1.290 32.249 78.431 1.341 33.523 79.350
3 .681 17.030 95.460 .641 16.034 95.384
4 .182 4.540 100.000 .185 4.616 100.000
Beef (N = 205)
1 1.746 43.661 43.661
2 1.120 28.004 71.665
3 .817 20.423 92.088
4 .316 7.912 100.000
Wool (N = 115) Wool No China *2
1 1.861 46.525 46.525 1.862 46.560 46.560
2 1.098 27.452 73.977 1.127 28.164 74.725
3 .785 19.626 93.603 .756 18.889 93.614
4 .256 6.387 100.000 .255 6.386 100.000
Telephone sets (N = 116)
1 1.873 46.825 46.825
2 1.081 27.016 73.840
3 .780 19.509 93.350
4 .266 6.650 100.000
Tertiary education (N = 193) Education No USA *3
1 1.860 46.493 46.493 1.874 46.846 46.846
2 1.080 26.995 73.488 1.054 26.347 73.193
3 .758 18.948 92.435 .765 19.114 92.307
4 .303 7.565 100.000 .308 7.693 100.000
Internal combustion car engines (N = 42)
1 1.990 49.756 49.756
2 1.193 29.825 79.581
3 .665 16.629 96.210
4 .152 3.790 100.000
Source: SPSS Total Variance Explained tables in PCA
*1 = Markets only and High capability organisations solutions
*2 = Markets only solution
*3 = Low capability organisations with objectives and High capability
organisations with objectives solutions
Only the first principal component is used, so other matters greyscaled

PCA of the economic data describing the countries which consume each
product has produced first economic components which have eigenvalues ranging
between 1.746 and 1.990.
 211

Table 6.2
Communalities

These indicate the amount of variance an economic sub-variable

shares with all other sub-variables in the analysis.

Initial Extraction Initial Extraction

Coking coal Coal No China *1

GNP / capita growth 1.000 .906 1.000 .907
Log GNP / capita 1.000 .905 1.000 .903
Population 1.000 .648 1.000 .671
Openness 1.000 .679 1.000 .693
Beef
GNP / capita growth 1.000 .838
Log GNP / capita 1.000 .826
Population 1.000 .663
Openness 1.000 .540
Wool Wool No China *2
GNP / capita growth 1.000 .871 1.000 .871
Log GNP / capita 1.000 .831 1.000 .835
Population 1.000 .714 1.000 .707
Openness 1.000 .543 1.000 .576
Telephone sets
GNP / capita growth 1.000 .869
Log GNP / capita 1.000 .840
Population 1.000 .704
Openness 1.000 .541
Tertiary education Education No USA *3
GNP / capita growth 1.000 .824 1.000 .818
Log GNP / capita 1.000 .810 1.000 .804
Population 1.000 .792 1.000 .820
Openness 1.000 .514 1.000 .486
Internal combustion car engines
GNP / capita growth 1.000 .889
Log GNP / capita 1.000 .920
Population 1.000 .651
Openness 1.000 .724
Source: SPSS Communalities tables in PCA
*1 = Markets only and High capability organisations solutions
*2 = Markets only solution
*3 = Low capability organisations with objectives and High capability
organisations with objectives solutions

All range between 0.2 minimum and 1.0 maximum.

 212

Table 6.3
Rotated Component Matrix
using Varimax method with Kaiser normalisation

Component Component
1 2 1 2

Coking coal Coal No China *1

GNP / capita growth .949 -.070 .950 -.062
Log GNP / capita .939 .155 .942 .123
Openness -.083 .820 -.073 .829
Population -.156 -.789 -.126 -.809
Beef
GNP / capita growth .915 .032
Log GNP / capita .905 .080
Openness .187 .711
Population .085 -.810
Wool Wool No China *2
GNP / capita growth .930 .082 .929 .088
Log GNP / capita .907 .088 .911 .076
Openness .246 .695 .234 .722
Population .073 -.842 .069 -.838
Telephone sets
GNP / capita growth .930 .067
Log GNP / capita .907 .129
Openness .213 .704
Population .033 -.839
Tertiary education services Education No USA *3
GNP / capita growth .907 -.030 .904 -.028
Log GNP / capita .896 -.088 .893 -.088
Openness .381 -.607 .410 -.563
Population .123 .881 .099 .900
Internal combustion car engines
GNP / capita growth .941 .051
Log GNP / capita .931 .229
Openness -.103 .845
Population -.298 -.750
Source: SPSS Rotated Component Matrix in PCA.
All rotations converged in 3 iterations
*1 = Markets only and High capability organisations solutions
*2 = Markets only solution
*3 = Low capability organisations with objectives and High capability
organisations with objectives solutions
Only the first principal component is used, so other components greyscaled

In the first principal component, GNP / capita growth and Log GNP /
capita are statistically significant and so shown in bold whilst Openness and
 213

Population are statistically non significant at less than ±.33 (Tabachnick &
Fidell 1996, p. 677 quoting Comrey & Lee 1992).

6.3 Cluster Analysis (CA)

The purpose of using CA was to group countries according to their

level of potential – particularly separating the high potential markets from the
others. The task was therefore to determine the number and composition of
clusters for each of the six product-organisation combinations analysed,
including producing a hierarchy of countries by potential.

The optimal number of clusters was determined by examination of the

agglomeration schedule, the dendrogram and the scree plot then cross-
checked by examining significance levels of discriminant functions. Ambiguity
of preference for a specific number of clusters was resolved from MDA
qualitative profiling data. Appendix H provides the MDA cluster profiling data,
together with summaries of various kinds.

Within the preferred cluster set, the best individual cluster is the one
which has the highest centroid values for the most variables. The centroid
value for each variable in each preferred cluster solution follows.
 214

Table 6.4
Final Cluster Centre Values in K-Means CA

Coal
Markets Only Low Capability High Capability
Organisation Organisation
Cluster Cluster Cluster
1 2 3 1 2 3 1 2 3 4
Product consumption .089 .099 .270 .063 .025 .036 .207 .394 .216 .140
Product consumption
Growth .496 .527 .616 .208 .200 .198 .974 1.179 1.004 1.131
1st Economic PC .328 .757 .807 .265 .415 .543 .590 1.516 1.675 1.039
Competition .228 .689 .832 .179 .406 .502 .384 1.489 1.522 .921
Legal fairness .200 .725 .549 .117 .380 .509 .375 1.070 1.676 .750
Cultural distance .347 .803 .163 .357 .574 .532 .598 .449 1.638 1.468
Market risk .394 .844 .922 .287 .501 .632 .703 1.651 1.760 1.415
Language .000 .813 .000 .600 .943 .614 .080
SCA .250 .429 .071 .700
Market research .150 .000 .071 .100
Contacts .300 .571 .214 .300

Beef
Markets Only Low Capability High Capability
Organisation Organisation
Cluster Cluster Cluster
1 2 3 1 2 3 1 2 3
Product consumption .012 .058 .019 .012 .019 .020 .027 .021 .082
Product consumption
growth .384 .387 .396 .286 .291 .277 .763 .805 .776
1st Economic PC .189 .598 .349 .158 .263 .343 .431 .500 .981
Competition .333 .724 .549 .267 .393 .462 .760 .832 1.298
Legal fairness .157 .752 .427 .130 .301 .495 .396 .536 1.237
Cultural distance .515 .723 .539 .361 .426 .496 1.084 .988 1.247
Market risk .489 .862 .719 .390 .499 .597 1.096 1.151 1.592
Language .000 .765 .000 .707 .784 .747
SCA .000 1.000 .181
Market research .012 .020 .042
Contacts .310 .245 .333
 215

Wool
Markets Only Low Capability High Capability Organisation
Organisation
Cluster Cluster Cluster
1 2 3 1 2 3 4 1 2 3 4 5
Product consumption .056 .055 .174 .037 .018 .103 .031 .074 .042 .075 .137 .158
Product consumption
growth .698 .707 .695 .320 .320 .296 .315 .839 .842 .864 .923 .834
1st Economic PC .209 .404 .634 .174 .244 .492 .405 .473 .409 .947 .513 1.227
Competition .301 .516 .701 .241 .339 .551 .463 .609 .653 1.113 .763 1.383
Legal fairness .152 .402 .832 .139 .222 .538 .557 .315 .408 .758 .508 1.567
Cultural distance .519 .617 .787 .368 .445 .605 .546 1.146 1.077 1.086 1.089 1.553
Market risk .493 .745 .903 .389 .489 .657 .648 1.024 1.088 1.553 1.121 1.794
Language .000 .691 .920 .000 .425 .567 .922 .733 .504
SCA 1.000 .000 .167 .292 .333
Market research .000 .133 .056 .000 .000
Contacts .000 .000 .000 1.000 .407

Telephones
Markets Only Low Capability High Capability
Organisation Organisation
Cluster Cluster Cluster
1 2 3 4 1 2 3 4 1 2 3
Product consumption .150 .158 .225 .210 .118 .115 .208 .136 .302 .333 .470
Product consumption
growth .539 .559 .532 .597 .408 .411 .390 .408 1.082 1.131 1.085
1st Economic PC .189 .322 .580 .621 .183 .243 .471 .402 .387 .823 1.132
Competition .281 .475 .658 .691 .253 .352 .542 .461 .592 1.070 1.321
Legal fairness .160 .419 .815 .467 .162 .286 .535 .552 .356 .889 1.583
Cultural distance .538 .569 .859 .412 .385 .397 .594 .552 1.128 1.039 1.651
Market risk .500 .695 .879 .858 .422 .497 .641 .646 1.058 1.450 1.787
Language .000 .730 .927 .000 .625 .947 .357
SCA .225 .472 .043
Market research .025 .057 .043
Contacts .350 .189 .478

Education
Markets Only Low Capability High Capability Organisation
Organisation
Cluster Cluster Cluster
1 2 3 1 2 3 1 2 3 4 5
Product consumption .015 .085 .028 .016 .023 .035 .027 .064 .027 .201 .054
Product consumption
growth .578 .578 .602 .431 .432 .445 1.160 1.189 1.190 1.148 1.186
1st Economic PC .189 .619 .377 .157 .270 .362 .349 .562 .877 1.156 .601
Competition .326 .733 .560 .260 .387 .482 .589 .903 1.177 1.440 .972
Legal fairness .155 .800 .432 .131 .297 .517 .289 .666 .893 1.654 .602
Cultural distance .503 .798 .555 .360 .429 .537 .999 1.175 1.094 1.713 .989
Market risk .479 .871 .716 .379 .491 .604 .872 1.267 1.493 1.736 1.254
Language .000 .760 .000 .625 .744 .905 .464 .842
SCA .054 1.000 .025 .318 .349
Market research .036 .000 .000 .045 .047
Contacts .143 .000 .000 .318 1.000
 216

Education with Low Capability High Capability

Organisational Organisation Organisation
Objectives
Cluster Cluster
1 2 3 1 2 3
Product consumption .007 .114 .345 .012 .197 .595
Product consumption
growth .023 .024 .022 .435 .448 .423
1st Economic PC .009 .013 .013 .011 .016 .017
Competition .013 .018 .017 .017 .023 .021
Legal fairness .010 .013 .009 .013 .017 .012
Cultural distance .016 .021 .013 .021 .027 .017
Market risk .090 .109 .105 .183 .221 .214
Language .014 .017 .008 .014 .013 .019
SCA .005 .008 .014
Market research .000 .001 .000
Contacts .006 .004 .005

Engines
Markets Only Low Capability High Capability
Organisation Organisation
Cluster Cluster Cluster
1 2 3 4 1 2 3 1 2 3
Product consumption .129 .033 .028 .807 .101 .126 .021 .062 .273 1.878
Product consumption
growth .753 .749 .727 .453 .578 .479 .540 1.485 1.467 .818
1st Economic PC .788 .438 .331 .936 .405 .532 .260 .746 1.503 1.942
Competition .729 .581 .220 .914 .478 .508 .218 .803 1.452 1.967
Legal fairness .769 .375 .137 .832 .344 .526 .128 .500 1.467 1.497
Cultural distance .810 .591 .554 .802 .518 .544 .387 1.155 1.546 1.453
Market risk .844 .693 .377 .961 .517 .636 .321 1.041 1.731 1.920
Language .800 .000 .000 .592 .513 1.000
SCA .208 .500 .000
Market research .083 .188 .000
Contacts .375 .250 .000
Source: SPSS Final Cluster Centres tables in K-means CA
Bold figures highlight the highest value cluster centroid for the predictor variable.

To conclude the CA section we now consider the high potential

markets suggested by screening. These are the markets produced by CA on
the basis of the optimal number of clusters and centroid values shown above.
Complete details for all countries, products and organisation are in Appendix I.
The following two tables summarise the numbers and names of the high
potential markets.
Table 6.5
Numbers of Markets in Each Cluster
– All Products
 217

Product Cluster Markets Low High

Number Only Capability Capability
Organisation Organisation
Coal 1 25 19 20
2 21 16 7
3 5* 17 14 *
4 - - 10
Total 51 * 52 51 *
Beef 1 90 77 84
2 34 96 49
3 81 32 72
Total 205 205 205
Wool 1 52 51 16
2 39 33 30
3 23 * 10 18
4 - 21 24
5 - - 27
Total 114 * 115 115
Telephones 1 35 44 40
2 45 40 53
3 22 11 23
4 14 21 -
Total 116 116 116
Education 1 91 77 56
2 26 91 32
3 76 25 40
4 - - 22
5 - - 43
Total 193 193 193
Education
with
objectives 1 171 171
2 17 17
3 4 4
Total 192 192
Engines 1 12 15 24
2 16 14 16
3 11 13 2
4 3 - -
Total 42 42 42
Source: Produced for this research from Appendix I
Bold = the high potential cluster, based on highest number of variables
* = 1 outlier market of unassessed potential not included

After conducting CA, then eliminating the low attractiveness markets,

which markets remain? Table 6.6 provides the answer.
 218

Table 6.6
High Potential Markets Proposed by Screening

Coking coal

Markets Only Low Capability High Capability

Organisation Organisation
Country Cl. Country Cl. Country Cl.
No No No

Chile 3 Austria 3 Australia 3

Japan 3 Belgium 3 Austria 3
Korea, S 3 Chile 3 Belgium 3
Portugal 3 Czech Rep 3 Canada 3
Taiwan 3 Finland 3 Finland 3
France 3 France 3
Germany 3 Germany 3
Hungary 3 Ireland 3
Iceland 3 Netherlands 3
Italy 3 New Zealand 3
Japan 3 Spain 3
Netherlands 3 Sweden 3
Poland 3 United Kingdom 3
Portugal 3 United States 3
Spain 3
Sweden 3
China - Taiwan 3 China -
 219

Beef

Markets Only Low Capability High Capability

Organisation Organisation
Country Cl. Country Cl. Country Cl.
No No No

American Sam 2 Austria 3 American Sam 3

Antigua 2 Belgium 3 Antigua 3
Australia 2 Bhutan 3 Argentina 3
Austria 2 Chile 3 Australia 3
Barbados 2 Colombia 3 Austria 3
Belgium 2 Czech Rep 3 Bahamas 3
Bermuda 2 Denmark 3 Bahrain 3
British Virgin Is 2 Dominica 3 Barbados 3
Brunei 2 Faroe Is 3 Belgium 3
Canada 2 Finland 3 Bermuda 3
Denmark 2 France 3 Brazil 3
Faroe Is 2 French Guiana 3 British Virgin Is 3
Finland 2 Germany 3 Brunei 3
France 2 Greenland 3 Canada 3
Germany 2 Guadeloupe 3 Chile 3
Greenland 2 Hungary 3 Colombia 3
Hong Kong 2 Iceland 3 Cook Is 3
Iceland 2 Italy 3 Costa Rica 3
Ireland 2 Japan 3 Cyprus 3
Israel 2 Lithuania 3 Czech Rep 3
Italy 2 Martinique 3 Denmark 3
Japan 2 Netherlands 3 Dominica 3
Korea, S 2 Norway 3 Faroe Is 3
Kuwait 2 Poland 3 Fiji 3
Malta 2 Portugal 3 Finland 3
Netherlands 2 Saint Pierre & 3 France 3
New Zealand 2 Solomon 3 French Guiana 3
Norway 2 Spain 3 French Poly 3
Singapore 2 Sweden 3 Germany 3
Sweden 2 Switzerland 3 Greenland 3
Switzerland 2 Taiwan 3 Guadeloupe 3
United Kingdom 2 Wallis & Fut 3 Hong Kong 3
United States 2 Hungary 3
Virgin I 2 Iceland 3
India 3
Ireland 3
Israel 3
Italy 3
Jamaica 3
Japan 3
Korea, S 3
Kuwait 3
Macau 3
Malta 3
 220

(Beef cont’d)
Martinique 3
Mauritius 3
Montserrat 3
Netherlands 3
Netherla Antill 3
New Zealand 3
Niue 3
Norway 3
Panama 3
Poland 3
Portugal 3
Puerto Rico 3
Saint Kitts 3
Saint Lucia 3
Saint Pierre & 3
Saint Vincent & 3
Samoa 3
Singapore 3
Solomon 3
Spain 3
Suriname 3
Sweden 3
Switzerland 3
Taiwan 3
Thailand 3
United Kingdom 3
United States 3
Virgin I 3
 221

Wool

Markets Only Low Capability High Capability

Organisation Organisation
Country Cl. Country Cl. Country Cl.
No No No

Australia 3 Australia 3 Australia 5

Austria 3 Canada 3 Austria 5
Belgium 3 Hong Kong 3 Belgium 5
Canada 3 Ireland 3 Canada 5
Denmark 3 Israel 3 Czech Rep 5
Faroe Is 3 Malta 3 Denmark 5
Finland 3 New Zealand 3 Faroe Is 5
France 3 Singapore 3 Finland 5
Germany 3 United Kingdom 3 France 5
Greenland 3 United States 3 Germany 5
Hong Kong 3 Greenland 5
Iceland 3 Hong Kong 5
Ireland 3 Iceland 5
Italy 3 Ireland 5
Japan 3 Israel 5
Netherlands 3 Italy 5
New Zealand 3 Japan 5
Norway 3 Korea, S 5
Singapore 3 Malta 5
Sweden 3 Netherlands 5
Switzerland 3 New Zealand 5
United Kingdom 3 Norway 5
United States 3 Singapore 5
Sweden 5
Switzerland 5
United Kingdom 5
China - United States 5
 222

Telephone sets

Markets Only Low Capability High Capability

Organisation Organisation
Country Cl. Country Cl. Country Cl.
No No No

Australia 3 Australia 3 Australia 3

Austria 3 Barbados 3 Austria 3
Barbados 3 Canada 3 Canada 3
Belgium-Lux 3 Hong Kong 3 Cyprus 3
Canada 3 Ireland 3 Czech Rep 3
Czech Rep 3 Israel 3 Denmark 3
Denmark 3 Malta 3 Faroe Is 3
Finland 3 New Zealand 3 Finland 3
France & Mon 3 Singapore 3 Germany 3
Germany 3 United Kingdom 3 Greenland 3
Greenland 3 United States 3 Hong Kong 3
Iceland 3 Hungary 3
Ireland 3 Iceland 3
Israel 3 Ireland 3
Italy 3 Italy 3
Netherlands 3 Netherlands 3
New Zealand 3 New Zealand 3
Norway 3 Norway 3
Sweden 3 Poland 3
Switzerland 3 Sweden 3
United Kingdom 3 Switzerland 3
United States 3 United Kingdom 3
United States 3
 223

Education services

Markets Only Low Capability High Capability

Organisation Organisation
Country Cl. Country Cl. Country Cl.
No No No

Australia 2 Austria 3 Australia 4

Austria 2 Belgium 3 Austria 4
Belgium 2 Chile 3 Belgium 4
Bermuda 2 Colombia 3 Canada 4
British Virgin Is 2 Czech Rep 3 Czech Rep 4
Brunei 2 Denmark 3 Denmark 4
Canada 2 Dominican Rep 3 Finland 4
Denmark 2 Finland 3 France 4
Finland 2 France 3 Germany 4
France 2 Germany 3 Iceland 4
Germany 2 Hungary 3 Ireland 4
Hong Kong 2 Iceland 3 Italy 4
Iceland 2 Italy 3 Japan 4
Ireland 2 Japan 3 Liechtenstein 4
Italy 2 Liechtenstein 3 Luxembourg 4
Japan 2 Lithuania 3 Netherlands 4
Liechtenstein 2 Netherlands 3 New Zealand 4
Luxembourg 2 Norway 3 Norway 4
Netherlands 2 Poland 3 Sweden 4
New Zealand 2 Portugal 3 Switzerland 4
Norway 2 Solomon 3 United Kingdom 4
Singapore 2 Spain 3 United States 4
Sweden 2 Sweden 3
Switzerland 2 Switzerland 3
United Kingdom 2 Taiwan 3
United States 2
 224

Education services with organisational Objectives

Low Capability High Capability

Organisation Organisation
Country Cl. Country Cl.
No No

Argentina 2 Argentina 2
Australia 2 Australia 2
Brazil 2 Brazil 2
Canada 2 Canada 2
France 2 France 2
Germany 2 Germany 2
Indonesia 2 Indonesia 2
Iran 2 Iran 2
Italy 2 Italy 2
Korea, S 2 Korea, S 2
Mexico 2 Mexico 2
Philippines 2 Philippines 2
Spain 2 Spain 2
Thailand 2 Thailand 2
Turkey 2 Turkey 2
Ukraine 2 Ukraine 2
United Kingdom 2 United Kingdom 2

United States - United States -

 225

Internal combustion engines

Markets Only Low Capability High Capability

Organisation Organisation
Country Cl Country Cl Country Cl
No No No

Germany 4 Austria 2 Japan 3

Japan 4 Belgium 2 United States 3
United States 4 Czech Rep 2
Finland 2
France 2
Germany 2
Italy 2
Japan 2
Netherlands 2
Poland 2
Portugal 2
Spain 2
Sweden 2
Taiwan 2
Source: Produced for this research from Appendix I from SPSS
Cluster Membership tables, K-means CA

- = an exception because of prior exclusion as an outlier market.

Manual assessment desirable to check its true
attractiveness

Discussion about the markets indicated by screening as having high

potential is in Results.

6.4 Multiple Discriminant Analysis (MDA)

6.4.1 Preliminary Matters

Prior to interpreting a MDA (multiple discriminant analysis) result,

homogeneity of the variance-covariance matrices needs to be assessed (the
multivariate homogeneity of variance assumption, sometimes called the
sphericity assumption) (Hair et al. 1995, p. 212; Tabachnick & Fidell 1996,
pp. 512-3). Since there are three situations43 for each of the six products,
there are 18 results to examine.

43
The Markets only, markets for High firms, and markets for Low firms, situations.
 226

Inspection of: (1) scatterplots of the 18 discriminant functions, and

(2) the significance levels of the 18 Box’s M tests, showed satisfactory
equality in overall size of 15, indicating that no remedial data manipulation of
those was required. However, three of the discriminant analysis solutions
appeared optimal in most respects other than that they included a cluster with
no variance, due to comprising only a single country44. Coal Best Markets,
Coal High Capability, and Wool Best Markets comprised a cluster containing
only one country (China). These three solutions were therefore recalculated
after excluding the extreme country. The rationale is that the extremity of the
data describing the outlier country distorts the coefficients of the regression
equation, making coefficients misrepresentative of the bulk of countries. The
outlier country is then analysed manually. Results reported here are those
produced after recalculating the three results to exclude China.

Separate tests for outlier countries were carried out. These involved
considering: (1) whether each country’s squared Mahalanobis distance to its
centroid was less than the critical tabulated values for Chi squared (χ2 ) with
degrees of freedom equal to the number of significant variables at p < .001
(Tabachnick & Fidell 1996, p. 94), and (2) whether plots of the first two
canonical discriminant functions of countries in each cluster show any
countries as outliers. These tests confirmed that all the above three countries
were outliers, further justifying their removal and the recalculation of the three
solutions. This step also revealed some additional countries (typically one to
three countries in each data set of between 42 and 205 countries) which
exceeded the criterion Chi squared value in the tables (e.g. Tabachnick &
Fidell 1996, p. 846). After consideration, these marginal outliers were left in
the analyses as they frequently did not disturb the hyper-sensitive Box’s M
test and/or they did not appear on the plots to be grossly extreme.

Another matter is redundant variables. Tabachnick & Fidell (1996, pp.

85-86) and Malhotra et al. (1996, p. 604) advise that it is not a good idea to

44
These 3 non-optimal solutions were apparent at the CA stage. The normally unnecessary
step of running them through MDA was undertaken to better understand the affects of
population outliers on solutions.
 227

include redundant variables in the same analysis as they inflate the size of
error terms, so weakening the analysis for hypothesis testing – although not
the selection of markets. Pairwise collinearity of the predictive variables was
therefore assessed by inspection of the pooled within-groups correlation
matrix for each of the 18 situations. Bivariate correlation ≥ ± .70 is
Tabachnick and Fidell’s suggested cutoff (pp 84-86) and is the figure used in
this research. Additionally, multicollinearity was assessed by inspection of the
tolerance figure for each variable in each of the 18 situations. Tolerances
should be no less than .10 suggest Hair et al. (1995, p. 127). That .10
equates to a multiple correlation of .95, but the more generous tolerance of at
least .19 (equating to a multiple correlation of .90) was used because of the
exploratory nature of this research. There is neither a universal rule nor
complete agreement about what these bivariate and multivariate figures
should be.

The associations which approached these levels were:

Table 6.7
Collinearity: Competition Variable with Market risk Variable
- Bivariate Correlation

Best Markets Best Markets for Best Markets for

Low Capability High Capability
Organisation Organisation
Coal .692 .838 .742
Beef .730 .861 .832
Wool .719 .812 .760
Telephones .547 .761 .681
Education .746 .866 .786
Engines .599 .777 .733
Source: SPSS Pooled Within-Groups Matrices in MDA
XXX = Pairwise collinearity possibly exists as correlation above ± .70
XXX = Pairwise collinearity unlikely to exist as correlation below ± .70
 228

Table 6.8
Collinearity: Competition Variable with 1st Economic PC Variable
– Bivariate Correlation

Best Markets Best Markets for Best Markets for

Low Capability High Capability
Organisation Organisation
Coal .553 .754 .428
Beef .509 .765 .657
Wool .573 .684 .560
Telephones .523 .721 .684
Education .622 .806 .723
Engines .347 .590 .347
Source: SPSS Pooled Within-Groups Matrices tables in MDA
XXX = Pairwise collinearity possibly exists as correlation above ± .70
XXX = Pairwise collinearity unlikely to exist as correlation below ± .70

Table 6.9
Multicollinearity: Competition Variable
– Multiple Correlation, as Indicated by Tolerance

Best Markets Best Markets for Best Markets for

Low Capability High Capability
Organisation Organisation
Coal .244 .086 .225
Beef .319 .139 .189
Wool .269 .159 .214
Telephones .434 .205 .274
Education .243 .113 .176
Engines .332 .143 .204
Source: SPSS Variables in the Analysis tables in MDA
XXX = Statistical possibility that multicollinearity exists as tolerance ≤ .19
XXX = Statistically unlikely that multicollinearity exists as tolerance > .19

Discussion of the impact of these correlations will be found in the

Interpretations and Conclusions Chapter.

6.4.2 The Null Hypothesis

MDA was used to test the results obtained from CA, per the null and
research hypotheses:

H0 There are no differences among the country groups,

 229

H1 Statistically significant differences exist among the centroids of

the country groups,
H2 The discriminating power of the MDA functions exceeds that of
chance group assignment.

The two matters which determine the response to the null hypothesis
are: (1) the statistical difference between cluster centroids and (2) the hit
ratios (the accuracy of the discriminant coefficients at assigning the markets to
the population). Each will now be dealt with in turn.
 230

Table 6.10
Statistical Significance of the Canonical Discriminant Functions
Which Discriminate Country Groups, Using Wilks' Lambda

Markets only Low Capability High Capability

Organisation Organisation
Test of Sig. Test of Sig. Test of Sig.
Function(s) Functions(s) Function(s)

Coking coal
1 through 2 .000 1 through 2 .000 1 through 3 .000
2 .000 2 .000 2 through 3 .000
3 .000
Beef
1 through 2 .000 1 through 2 .000 1 through 2 .000
2 .000 2 .000 2 .000
Wool
1 through 4 .000 1 through 3 .000 1 through 4 .000
2 .015 2 through 3 .000 2 through 4 .000
3 .010 3 through 4 .000
4 .001
Telephone sets
1 through 3 .000 1 through 3 .000 1 through 2 .000
2 through 3 .000 2 through 3 .000 2 .000
3 .006 3 .008
Tertiary education services
1 through 2 .000 1 through 2 .000 1 through 4 .000
2 .000 2 .000 2 through 4 .000
3 through 4 .000
4 .000
Tertiary education services with organisational objectives
1 through 2 .000 1 through 2 .000
2 .000 2 .000
Internal combustion car engines
1 through 3 .000 1 through 2 .000 1 through 2 .000
2 through 3 .000 2 .000 2 .000
3 .020
Source: SPSS Wilks’ Lambda tables in MDA
Bold figures indicate significance of function is at or below .05

The discriminant functions of consequential size (judged by their

variance) for all six product samples are statistically significant at less than the
.05 level (the conventional criterion per Hair et al. 1995, p. 198). The
probability of chance assignment of markets to groups by all important
discriminant functions is typically .000. The centroids of all six product groups’
clusters are therefore not all equal, meeting the first of the two criteria that
determine whether the null hypothesis is rejected.
 231

The accuracy of forecasting the assignment of countries to clusters

using the discriminant functions examined above will now be examined.

Table 6.11
Classification Results (Hit Ratios)
Compared with the Chance Occurrence of Those Assignments

Product Best Best Best

Markets Markets for Markets for
Low High
Capability Capability
Organisation Organisation

Coal Cross-validation # 98.0 % 96.2 % 96.1 %

Proportional chance ^ 52.4 41.8 35.8
Maximum chance ^ 61.3 43.3 49.0

Beef Cross-validation # 94.6 % 99.0 % 93.2 %

Proportional chance ^ 47.0 47.8 43.7
Maximum chance ^ 53.7 58.5 54.3

Wool Cross-validation # 97.4 % 98.3 % 94.8 %

Proportional chance ^ 45.7 40.4 26.2
Maximum chance ^ 57.0 55.4 32.6

Telephones Cross-validation # 95.7 % 95.7 % 96.6 %

Proportional chance ^ 37.0 39.1 46.1
Maximum chance ^ 49.6 48.5 58.2

Education Cross-validation # 94.3 % 98.4 % 94.8%

Proportional chance ^ 48.9 49.4 27.1
Maximum chance ^ 59.6 58.9 35.0

Education Cross-validation # 99.5 99.5

with Proportional chance ^ 100.2 100.2
Objectives Maximum chance ^ 111.3 111.3

Engines Cross-validation # 95.2 % 95.2 % 83.3 %

Proportional chance ^ 37.6 42.1 60.5
Maximum chance ^ 47.6 44.6 74.4
Source: Constructed for this research from SPSS Classification Results
tables in MDA and calculations based on same
#
Percentage of cross validated grouped cases correctly classified. In
cross validation, each case has been classified by the functions derived
from all cases other than that case. These are more conservative (lower)
figures than hit ratios based on total samples.
^ These figures represent chance plus 25 per cent. Research results
above these hurdle levels are generally accepted as being unlikely to
have occurred through fortuitous assignment of countries to clusters by
the MDA algorithms (see, for example, Hair et al. 1995, p. 204; Malhotra
et al. 1996, p. 516).
 232

With the exception of Education with objectives, all classifications

exceed chance plus 25 per cent. A simple average of the 18 cross-validations
is 95.4 per cent whilst a simple averaging of the severest of the two methods
of calculating chance (maximum chance) plus 25 per cent is 52.3 per cent.
The second condition required for rejection of the null hypothesis is thus
affirmed.

As: (1) the centroids of the clusters are statistically different and (2) the
hit ratio is consistently more than 25 per cent higher than chance, the H0 is
rejected and both H1 and H2 are accepted.

6.4.3 Relative Dimension Importance

The relative importance of each dimension is the next matter to

determine to assess the third research hypotheses:

H3 The product-specific consumption and consumption growth

variables are the most important characteristics for
discriminating among country groups.

This is accepted or rejected by considering: (1) which variables are

statistically significant predictors, then (2) assessing the strength of their
influence when determining the countries selected.

Significance of each variable is examined below.

 233

Table 6.12
Tests of Equality of Group Means for Statistical Significance

Best Markets Best Markets Best Markets

for Low for High
Capability Capability
Organisation Organisation
F Sig F Sig F Sig
Coking coal
Product consumption 1.79 .177 .54 .586 .56 .648
Product consumption growth 1.89 .162 .07 .931 1.67 .188
1st Economic principal component 47.99 .000 13.48 .000 46.80 .000
Competition 75.53 .000 19.34 .000 59.19 .000
Legal fairness 46.36 .000 27.31 .000 44.29 .000
Cultural distance 48.08 .000 12.08 .000 39.73 .000
Market risk 57.27 .000 25.47 .000 36.05 .000
Language (country-specific) 409.50 .000 7.44 .000
SCA (country-specific) 4.59 .007
Market Research (country-specific) .47 .705
Contacts (country-specific) .92 .439
Beef
Product consumption 4.02 .019 .27 .762 2.89 .058
Product consumption growth 0.89 .414 1.12 .327 1.92 .150
1st Economic principal component 214.59 .000 36.69 .000 109.83 .000
Competition 228.40 .000 47.77 .000 98.54 .000
Legal fairness 233.40 .000 81.07 .000 127.80 .000
Cultural distance 22.24 .000 15.59 .000 9.40 .000
Market risk 152.66 .000 40.44 .000 61.84 .000
Language (country-specific) 1507.21 .000 .73 .485
SCA (country-specific) 309.05 .000
Market Research (country-specific) .74 .480
Contacts (country-specific) .55 .577
Wool
Product consumption 6.02 .003 2.55 .060 1.09 .368
Product consumption growth .16 .856 .58 .630 1.49 .209
1st Economic principal component 99.08 .000 44.66 .000 62.90 .000
Competition 129.17 .000 51.35 .000 57.25 .000
Legal fairness 211.54 .000 70.60 .000 59.38 .000
Cultural distance 19.49 .000 15.41 .000 8.12 .000
Market risk 99.11 .000 37.40 .000 36.61 .000
Language (country-specific) 1338.52 .000 4.70 .002
SCA (country-specific) 22.25 .000
Market Research (country-specific) 2.32 .062
Contacts (country-specific) 75.22 .000
 234

Telephone sets
Product consumption 4.20 .007 6.34 .001 6.56 .002
Product consumption growth 2.02 .115 .29 .832 1.08 .345
1st Economic principal component 86.10 .000 35.15 .000 56.00 .000
Competition 106.81 .000 44.45 .000 92.77 .000
Legal fairness 89.43 .000 59.77 .000 115.32 .000
Cultural distance 31.34 .000 15.03 .000 28.17 .000
Market risk 81.86 .000 34.26 .000 78.08 .000
Language (country-specific) 826.26 .000 30.45 .000
SCA (country-specific) 8.82 .000
Market Research (country-specific) .27 .764
Contacts (country-specific) 3.66 .029
Tertiary education services
Product consumption 6.15 .003 .76 .471 4.20 .003
Product consumption growth 2.26 .107 .58 .562 .51 .726
1st Economic principal component 180.31 .000 38.13 .000 50.47 .000
Competition 208.80 .000 50.73 .000 66.57 .000
Legal fairness 246.85 .000 74.52 .000 75.26 .000
Cultural distance 36.54 .000 20.89 .000 27.65 .000
Market risk 143.01 .000 39.73 .000 60.46 .000
Language (country-specific) 1422.29 .000 8.71 .000
SCA (country-specific) 56.89 .000
Market Research (country-specific) .79 .534
Contacts (country-specific) 116.96 .000
Education with Objectives
Product consumption 1201.61 .000 1201.62 .000
Product consumption growth .73 .482 .73 .482
1st Economic principal component 5.53 .005 5.52 .005
Competition 7.45 .001 7.45 .001
Legal fairness 1.58 .208 1.58 .208
Cultural distance 9.30 .000 9.30 .000
Market risk 4.08 .018 4.08 .018
Language (country-specific) .58 .563 1.26 .287
SCA (country-specific) 2.60 .077
Market Research (country-specific) .43 .650
Contacts (country-specific) .19 .824
Internal combustion car engines
Product consumption 70.83 .000 1.64 .207 77.15 .000
Product consumption growth 5.27 .004 3.82 .030 6.67 .003
1st Economic principal component 30.54 .000 10.51 .000 39.67 .000
Competition 30.55 .000 15.63 .000 19.21 .000
Legal fairness 35.22 .000 20.40 .000 33.06 .000
Cultural distance 4.19 .012 3.82 .030 4.17 .023
Market risk 20.40 .000 18.86 .000 20.91 .000
Language (country-specific) 376.07 .000 1.04 .363
SCA (country-specific) 2.49 .096
Market Research (country-specific) .61 .548
Contacts (country-specific) .79 .459
 235

Source: SPSS Tests of equality of Group Means in MDA

Bold figure indicates the variable is significant at or below .05.

SUMMARY:
Number of Times Each Variable’s Coefficient is Significant (≤.05).
Maximum Score Possible = 6 per column (i.e. Education with objectives not included)
Product consumption 5 1 3
Product consumption growth 1 1 1
1st Economic principal component 6 6 6
Competition 6 6 6
Legal fairness 6 6 6
Cultural distance 6 6 6
Market risk 6 6 6
Language (country-specific) 6 4
SCA (country-specific) 5
Market Research (country-specific) 0
Contacts (country-specific) 3
Source: Constructed for this research from data in prior table

Summarising the above tables, when the predictive variables are

considered individually, seven of the 11 are consistently significant at or below
the .05 level. These are: 1st Economic principal component, Competition,
Legal fairness, Cultural distance, Market risk, Language and SCA (country-
specific). The remaining four variables appear to be statistically non-
significant, namely Product consumption, Product consumption growth,
Market research and Contacts. These challenges to conventional wisdom are
interpreted in section 7.2.3.3.

Results regarding the second aspect of research hypothesis three,

determining how strongly each variable influenced the solutions, will now be
presented. However, none of the solutions was based on one discriminant
function (i.e. two clusters), and for two or more discriminant functions, there is
no unambiguous measure of the relative importance of the predictor variables
when discriminating between countries. In this situation there are three
methods available to hierarchy variables:–

1) the Structure Matrix (the pooled within-groups correlations

between discriminating variables and the standardised canonical
discriminant functions),
 236

2) the Potency Index, and

3) the F statistic for the variables entered/ removed.

Appendix K provides those three assessments for each of the six

products in the three situations (Best Markets, Low capability organisation and
High capability organisation), 54 rankings in total.

Table 6.12 (below) summarises the rank order of each variable in each
product and situation.
 237

Table 6.13
Variable Importance: Hierarchy Calculated Using Three Methods
– Numbers Indicate Ranked Order of Importance

Rotated Structure Potency Index F Statistic:

Matrix Variables
Entered/Removed

w/ objectives
w/ objectives

w/ objectives
Education

Education

Education
T’phones

T’phones

T’phones
Engines

Engines

Engines
Wool

Wool

Wool
Coal

Coal

Coal
Beef

Beef

Beef
Best Markets (7 variables)
Competition 1 1 1 1 1 1 1 1 3 3
Legal 3 3 1 3 3 3 5 3 1 3 3 2 2 1 1 1 1 2
1st Economic PC 4 4 2 4 2 3 4 2 4 3 3 2 2 3 2
Market Risk 2 2 2 2 2 2 2 2 4 4
Cultural distance 5 4 4 4 4 4 1 4 3 2 3
Product consumption 1 1 1
Product consumption
growth
Low Organisation’s Best Markets (8 / 16 variables *)
Language (country-specific) 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
Legal fairness 2 2 2 2 2 2 2 4 2 2 2 2 2
Market risk 3 4 3 3 3 4 3 3 4
Competition 4 3 4 4 4 3 4 2 3
1st Economic PC 5 5 5 5 6 5 5 5 3 3
Cultural distance 6 6 5 6 3 4 4 3 2
Product consumption 1 1 5 1
Product consumption
growth
High Organisation’s Best Markets (11 / 19 variables *)
Legal fairness 4 2 6 1 5 2 4 2 5 1 4 2 3 2 3 1 2
1st Economic PC 2 3 1 4 4 3 2 3 1 4 5 3 3 2 6 2
Competition 1 4 2 2 2 5 1 4 2 2 2 5 3
Market risk 3 5 3 3 3 4 3 5 4 3 3 4 4 4
SCA (country-specific) 6 1 6 6 1 6 1 4 4 3
Contacts (country-specific) 4 1 3 1 1 5 1
Cultural distance 5 6 5 7 6 1 4 3 5 2
Language (country-specific) 5 5 6 5 2 5 2 7
Product consumption 1 1 1 1 6 8 1 1
Product consumption 4
growth
Market research (country-specific)
Source: Constructed for this research from the SPSS MDA rotated structure matrices,
potency calculations based on same, and Variables Entered/Removed tables.
Includes only variables which are statistically likely to be significant i.e. variables:
• whose largest rotated pooled within-groups correlations with
their discriminant functions ≥ ± .33, and which have an
eigenvalue of ≥ 1.0, and which account for ≥ 5 % of variance (in
the Structure matrix and Potency index columns);
• whose significance of F to remove is ≤ .05 (in the F Statistic column).
Blank cells indicate the variable has failed to achieve the above minimum hurdles.
 238

* = except Education with organisational objectives, which has 8/20 variables

(Low organisation) and 11/23 variables (High organisation).
Tables 6.11 and 6.12 have not found support for research hypothesis
H3.

Discussion of this section’s interesting findings will be found in the next

chapter.
 239

CHAPTER 7

INTERPRETATIONS AND CONCLUSIONS

“When I use a word,” Humpty Dumpty said in a rather

scornful tone, “it means just what I choose it to mean –
neither more nor less.”
Lewis Carroll, Through the Looking-Glass, Ch 6

7.1 Introduction

The previous chapter’s results need to be interpreted and conclusions

drawn about the results from the three statistical procedures used – principal
components analysis, cluster analysis, and multiple discriminant analysis. Of
particular import are the country assignments for the 20 situations examined,
the tests of the statistical significance of the assignments and the
assessments of the relative importance of each dimension used.

7.2 Interpretations and Conclusions

7.2.1 Principal Components Analysis (PCA) Results

Is factorability appropriate, and do the data meet the other necessary

data quality criteria? Correlations of sub-variables (rotated per Table 6.3;
unrotated: available from the author) all exceed 0.3, so the matrix is factorable
(Hair et al. 1996, p. 374; Tabachnick & Fidell 1996, p. 671),. Since six
populations – not samples – are being dealt with, there are no sampling
issues regarding missing data, sample size, and so on. All first component
eigenvalues are around 1.8 (Table 6.1), well over the 1.0 which would indicate
 240

that it would be better to use some or all of the variables singularly than to use
the component. Communality levels of between 0.2 and 1.0 (Table 6.2)
indicate that appropriate sub-variables have been used. Inspection of the
Correlations matrices, Determinants, Anti-image matrices and the
Reproduced correlation matrices’ residuals all reveal nothing problematic.
Finally, it can be seen (Table 6.1) that PCA of the four economic sub-variables
has produced first components which capture a satisfactory 45.7 to 49.8 per
cent of the variance of the four sub-variables. It is thus reasonable to assert
that the 1st economic PC is a variable which is potentially capable of
influencing market selection.

In terms of the type of influence it can assert, Table 6.3 shows that this
economic variable is primarily driven by the sub-variables GNP / capita growth
and Log GNP / capita, and at that the statistically excellent correlation levels
of .89 to .95 (Tabachnick & Fidell 1996, p. 677). The 1st economic PC thus
selects affluent, or becoming more affluent, markets. The variable will also
correlate with any other variables which have strong traces of wealth or wealth
growth – as happens with a macro version of Competition, and to a lesser
extent with Market risk, as reported in Tables 6.7, 6.8 and 6.9.

Two other sub-variables which were subjected to PCA were Openness

and Population. Their affects are principally captured by the second
component, which is not used. Market size (as partly proxied by Population),
and macro-economic barriers to entry (proxied by Openness) are thus not
significantly represented in the variable. The absence of market size from the
1st economic PC means that future sales potential is not captured as fully as
possible. However, actual growth in product sales and actual growth in per
capita income are both represented among the variables used, so the
framework includes a satisfactory, if not quite as fulsome, level of future sales
potential. There is perhaps an argument here for increasing the number of
economic components from one to two in future research to see if marked
differences are noted in the markets recommended.

7.2.2 Cluster Analysis (CA) Results

 241

7.2.2.1 Optimal Clusters

The material in Appendix H provides much of the data which was used
to make the selection of the number of clusters for each of the solutions.
Optimum solutions ranged between three and five clusters (see Numbers of
Clusters Determining Solutions). The most common optimum solution was
based on three clusters, but there is no a priori, predictable number of clusters
likely to be optimal. The number of clusters also varies across the three
situations for the same product (e.g. Wool Markets Only optimal solution has
three clusters, Wool Low capability organisations has four clusters and Wool
High capability organisations has five clusters).

Within the preferred cluster solution, the cluster having the best
potential is judged from centroid centres – the higher the value, the better the
cluster’s potential (because scales were constructed so that high values
indicated high attractiveness markets). Again, there is no predictability about
which sub-group will be best, although cluster three is frequently the one
having highest potential.

There are differences between solutions. Comments therefore follow

on each product’s optimum cluster centroid results (see Table 6.4).

Coal
Inspection of centroid values indicates that:

• Cluster three in the Markets Only situation has the highest product
sales potential (product consumption and product consumption
growth). It also has the best economic, competitive and risk
environments, but is behind the markets in cluster two for legal and
cultural milieus. Cluster three markets could be typified as having best
sales potential and a relatively easy general environment within which
to conduct business;
 242

• Cluster three contains the best markets for Low Capability

organisations. Those markets score the highest for economic,
competitive, legal and risk environments. However, cluster three
markets are not as high in sales potential (product consumption and
product consumption growth) as cluster one markets, and cluster two
markets are better in terms of culture and language. Cluster three can
be typified as the best compromise between sales potential and
operating environment – the organisation could instead select cluster
one markets but would then trade better sales potential for a tougher,
riskier operating environment;

• Cluster three contains the best markets for High Capability Coal
organisations. Cluster three markets have the most benign
environments in terms of economic, competitive, legal, culture and
market risk, and good but not best potential demand (product
consumption and growth). Cluster three markets have the second best
scores for the languages spoken by the organisation’s employees, the
worst scores for country-specific SCA, third best for country-specific
market research, and worst scores for country-specific contacts.
Cluster three markets can be typified as having reasonably strong
sales potential, good general environment within which to conduct
business, but the markets are ones in which the organisation has weak
country-specific strengths (e.g. has conducted little market research
and built up few contacts)

Beef
• The Markets Only situation for beef markets is best in cluster two.
This cluster has the highest values for six of the seven variables and is
only a little behind cluster two in terms of product consumption growth;

• Cluster three contains the best markets for Low Capability

organisations marketing beef. Those markets score the highest in six
 243

of the eight variables. However, cluster three markets are not as high
in product consumption growth or languages spoken as cluster two
markets;

• Cluster three contains the best markets for High Capability

organisations, having the highest scoring centroids in eight of the 11
variables. However, markets in cluster two are better in terms of
product consumption growth, languages and country-specific SCA.

Wool
• Cluster three is clearly the best for wool on a Markets Only
assessment. It has top centroid scores for product consumption,
economic climate, competitive climate, legal fairness, cultural distance
and market risk. It has strong product consumption growth prospects
although those growth prospects are the weakest of the three clusters;

• Cluster three is again the best for Low Capability organisations, having
the highest valued centroids for six of the eight variables. The less
attractive features of these markets are that they have the least
demand growth, and their legal fairness levels are typically lower than
those in cluster four markets;

• Cluster five contains the best markets for High Capability organisations,
although there are some compromises amongst the markets’
characteristics. Demand is strongest in this cluster of markets but
demand growth is the least of all the clusters. The background
operating environments are good (economic, competitive, legal, cultural
similarity and market risk). However, the organisation has few market-
specific strengths – it has conducted no market research on these
markets, can understand the language in only some of them, and has
patchy county-specific SCA and in-market contacts.

Telephones
 244

• A Markets Only assessment leads towards countries in cluster three.

These have the best product consumption levels and the best legal,
cultural and market risk levels. The downsides to these markets are
that they have slightly poorer economic and competitive environments,
and have the just-worst product consumption growth values;

• A Low Capability organisation should head towards cluster three

countries to sell its telephones. Those markets have the highest
product consumption levels and the best economic, competitive,
cultural and language scores. The softness in those markets is that
they have the weakest product consumption growth levels of all
clusters and have slightly higher risk levels than cluster four markets;

• High Capability organisations will generate highest returns from cluster

three markets. These have the best operating environments
(economic, competition, legal, cultural and market risk) and the
organisation already has contacts in many of the locations. The
denizens of these markets consume the most of the product and have
the second highest levels of product consumption growth. The less
attractive aspects for the organisation are that it has language skills for
only some of the markets, has country-specific SCA in even fewer, and
it has conducted little market research on these markets.

Education
• Ignoring the organisation’s attributes and conducting a Markets Only
assessment of where to market university degrees leads to cluster two
markets. These have the best background environments (economic
and competitive milieus, legal fairness, cultural difference and market
risk), the best product consumption levels, and the equal second-best
product consumption growth levels;

• Adjusting the assessment to include attributes of a Low Capability

organisation sees cluster three become the target group of markets.
 245

These score highest centroid values for seven of the eight variables but
the organisation has a weakness in the language competence area;

• Adjusting Markets Only for the attributes of a High Capability

organisation directs them to cluster four markets. These have the best
operating environments and the best product consumption levels but
weak product consumption growth. However, the organisation has
weak language, country-specific SCA, market research and contacts to
help its market entries.

These Education results suggests that, whilst this research has kept
variable weights uniform by industry, variables should at times be given
different levels of relative importance. The product ‘university degrees’
is more intimately involved with the language of the customer than, say
engines. An engines customer does not need to know the language of
the supplier to consume the product whereas, in general, education
instruction needs to be consumed in the language of the student. It
can thus be argued that language skills of the university should be
given a higher weighting to bias selection towards those markets. The
counter-argument is that the university could hire lecturers capable in
the languages needed. Whilst certainly feasible there are
disadvantages of hiring new employees (for example, that delays
market entry whilst staff are sought, hired and inducted; those new staff
are unknown in terms of capability and motivation; hiring instructors still
raises questions about the appropriateness of applying existing
business strategy and marketing mix to locations where the university
cannot presently communicate, so introducing a range of market-entry
inefficiencies and risks; it opens up the area of motivation of existing
employees and organisational cultural change through introducing
newcomers, and so on). Therefore, the clear preference would seem
to be to give additional weighting to markets in which the university has
language competencies.
 246

Education with objectives

• Re-weighting the Low capability organisation for a set of feasible
objectives changes the high potential cluster from number three to
number two. Seven of the eight variable centroids are highest now
in cluster number two. Language is no longer highest in another
clusters but product consumption is;

• Goal weights set by the High capability organisation reduce the

number of clusters from five to three, and the optimum markets are
now in cluster two (previously in cluster 4).

Adding objectives has produced a fourth kind of solution (additional to

Markets only, Low organisation and High organisation), of which two
examples have been provided here.

Engines
• A Markets Only assessment for engines directs an organisation
towards cluster four markets. Here it will experience countries
generating the highest centroid scores for five of the seven variables
(product consumption, economic, competition, legal and market risk).
The disadvantages of these markets are that they have soft product
consumption growth and marginally weaker cultural distance than
optimum;

• Cluster two holds the best overall set of markets for a Low Capability
organisation. The background environments are the best there but
product consumption growth is weakest of all three clusters, and the
organisation has no language skills;

• High Capability organisations should enter cluster three. That group of

markets has the best product consumption, the weakest product
consumption growth, the best economic, competitive and legal scores,
the second best cultural difference levels, the best market risk scores,
 247

and the organisation can speak all necessary languages. The

downsides of cluster three markets are that the organisation has zero
market-specific SCA, market research and contacts in/on those
locations.

In summary, the centroids of the best clusters tell a story about the
strengths and weaknesses of each of the aggregations of best markets.
Those centroid profiles are unique for each of the 18 situations. An important
consequence is that it is imprudent for governments to tell organisations that
“Markets X, Y and Z are the best to sell product A in”. “Best” markets are
determined by considering characteristics of both the markets the
organisation. Similarly, organisations should include their strengths and
weaknesses, as well as attributes of the markets, when deciding which
markets are best for them to endeavour to enter.

This conceptual approach is not presently part of the rhetoric espoused

by international marketing scholars, practitioners or governments. The result
is that there are probable opportunity costs being borne by society – the costs
of organisations expending additional effort penetrating markets that are more
difficult than optimum, the costs of failure when they do not succeed at
entering the markets they choose, and earning sup-optimum profits from
entering markets with lower potential than the organisation’s level of
competence.

It bears mentioning that China, and to a lesser extent the USA, are
frequently outliers or nearly outliers in terms of their data properties. China is
always the largest market in population size and the USA the most affluent, so
whatever market selection system is used needs to be sensitive to additional
selection variables so as to not invariably direct organisations to the same
market. Ward’s method of CA frequently included China and the USA in a
group of high potential markets whereas Average linkage method of CA
frequently suggested a solutions which placed China and/or the USA in their
own clusters. MDA typically found China to be an outlier.
 248

This leaves the analyst with a conceptual dilemma: when the outlier
classification is marginal, as is frequently the position, is it better to include or
exclude the country? Advantages of inclusion are that regression equation
coefficients represent total coverage of the population (not sample) so all
markets are treated and assessed relatively. That avoids the task of manual
assessment of the excluded markets. On the other hand, the regression
coefficients are distorted by including the extreme values of China and/or the
USA. That stretches the statistical procedures, perhaps questioning the
preciseness of the results they produce. This researcher needs to leave to
more statistically proficient analysts the determination of this question – which
may actually have no general answer.

7.2.2.2 Optimal Markets

We now shift attention to the question: which markets are in each of

those optimum clusters? CA generated the answers to that question.
Appendix H presents alphabetised lists of all the countries in each cluster
whilst Table 6.6 provides a summary of the high potential markets, both by
product and situation.

Inspection of the markets suggested by screening shows that different

markets are typically suggested for the three situations (Markets Only, Low
Capability and High Capability organisations). There is some commonality,
but screening is largely suggesting different markets for different products and
different organisation competencies. The degree of overlap is difficult to
express clearly because of the choice of denominators but the following table
goes part way toward the answer.
 249

Table 7.1
Commonality Among Markets Selected *
including exceptional markets requiring manual analysis +

All 3 Markets Low - High Markets Total No of

situations Only – Low Only - High Markets
i.e. Markets
Only, Low Suggested by
& High Screening +

Coal - 4 8 1 38

Beef 15 15 29 34 138

Wool 8 8 10 23 61

Telephones 6 8 7 18 56

Education 14 14 15 21 94

Engines 1 2 1 2 19

Sub-Totals 44 51 70 99 406

Grand Total 264

Source: Produced for this research from Table 6.6, High Potential
Markets Proposed by Screening.
* i.e. the number of times a market’s name appeared in each of the 4
possible combinations (e.g. if Japan appeared in each of the 4
combinations, a 1 was recorded in each of the 4 columns above; if Japan
appeared in two of the combinations, a 1 was recorded in each of two
columns).

The large markets of Japan and the USA explain a little of the overlap.
The sizes of their statistics perhaps sometimes causes these countries to be
placed in all three situations (Japan for all three situations in three of the six
products while the USA is suggested for all three situations in two of the six
products). Using the current weightings for selector variables, and because of
the relative immensity of data describing these two markets, screening is
saying that these markets are inescapably important opportunities –
irrespective of organisation capability. The USA and Japan only explain 18
(equals [two plus three] times three) of the 264 instances of repeated market
selection.
 250

Apart from Japan, no country appears in all three situations (Markets

only, Low organisation and High organisation) for more than two products.
This says that the framework generally selects different markets for different
products.

The countries which appear in all three situations for two products are:
Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany,
Iceland, Ireland, Italy, New Zealand, Netherlands, Norway, Sweden,
Switzerland, United Kingdom and the United States. These are all
economically developed countries whose overall market attractiveness could
be expected to be high. This suggests that, where the framework repeats its
selection of markets, it is generally tending to select intuitively appropriate
markets.

Not included in the above analysis are the two groups of markets
selected for Education organisations with Low and High competencies and
hypothetical objectives. Examination of Table 6.6 shows that:

• largely different sets of markets were selected after objectives were set
for the firm (i.e. Low organisation compared with Low organisation with
objectives; High organisation versus High organisation with objectives).
This result was expected;

• the same countries have been selected for both the Low organisation
with objectives and the High organisation with objectives. This result
was not expected so will now be commented on.

Examination of the distance of each market from its cluster centroid

provides an incomplete picture of market potential (this is so because no sign
is attached to the distance figure, so there is no indication of whether a market
is highly attractive or highly unattractive). Nevertheless, this distance gives a
broad indication of difference as assessed by the clustering process.
Arranging markets by their distance from the centroid produces very different
 251

hierarchies for Low organisations with objectives from High organisations with
objectives45. It would thus seem that the set of markets in this instance is the
same, but the relative order of importance is quite different. Since the
objectives weights are clearly different between the two organisations, that
suggests either an insensitivity in the objectives scoring system, or an
interaction among the two data sets which happens to produce a similarity in
the outcomes. Since only one product, Education, was calculated, this aspect
requires confirmation.

It is reasonable to conclude from the above that the markets being

selected by screening do vary depending on whether they are for Markets
Only, a Low Capability organisation, a High Capability organisation or an
organisation whose objectives have been included in the calculations. The
objective of using CA (to determine the number and composition of clusters so
as to sort markets into different potentials) has been achieved.

Further, the screening framework appears to be of theoretical and

practical use.

A minor limitation is that the CA and MDA algorithms do not happily

cope with outliers, leaving a small number of countries for manual analysis.
China and the USA regularly feature, and intuition would suggest that the
causes are principally China’s population and resulting total national wealth,
and America’s lesser but still significant population and per capita wealth.
These two countries will regularly need to be analysed manually – not a major
burden to most practitioners, who are familiar with these two exceptional
markets. The number of outlier markets was never large, ranging from one to
four markets for any given product and capability combination. The largest
number of outliers, four markets, occurred for Education, which comprises a
set of 193 markets. Outliers thus do not seem to be a burdensome issue.

45
eg Australia, Indonesia and Iran are the first 3 markets in the set for Low organisations with
objectives, whilst Iran, Canada and Germany are the first 3 for High organisations with
objectives.
 252

A more important problem is that CA has produced lists of markets

which are often too large for in-depth research to be conducted on all. The
number of high attractiveness markets varied between two and 72, and 16 of
the 20 solutions exceeded the arbitrary number of 10 markets (Table 6.5). It
has previously been argued that approximately six markets is a desirable
number of markets to be passed from screening to the in-depth assessment
stage for cost, psychological and comparative reasons. Yet, there is no way
of reducing the number of attractive markets suggested by screening using
CA (e.g. hierarchying markets by potential and, where there are more markets
than required, selecting the top half a dozen)46. This appears to be a critical
flaw in CA which may limit its use in practice.

7.2.3 Multiple Discrimination Analysis (MDA) Results

7.2.3.1 Preliminary Matter – Collinearity and Multicollinearity

Some bivariate and multivariate correlations appeared between the

independent variables Competition and Market risk, and to a lesser extent
between Competition and the 1st Economic principal component (see Tables
6.7 and 6.8 ). Correlations above ± .70 (bivariate) or tolerances above ± .19
(multivariate) indicate possible redundancy of a variable, whilst correlations /
tolerances above ± .90 (bivariate) or .10 (multivariate) indicate almost certain
variable redundancy.

To explore the impact of collinearity on results, Beef was selected as a

sample to have its Best Markets results re-computed, firstly without
Competition, then without Market Risk. Removing either variable was found
to reduce correlations below the ± .70 cutoff and to generate clearly different
clusters, so the null hypothesis continued to be rejected. Based on
discriminant scores, 18 of the same countries appeared in the top 20

46
CA’s cluster centroids are ineffective for this task because, whilst they show distance from
cluster centroid, they do not indicate whether the market is strongly attractive or strongly
unattractive. Likewise, the MDA factor scores produced relate to discriminating between
clusters, not discriminating market attractiveness.
 253

countries of all three Beef analyses. It can thus be concluded that

Competition’s multicollinearity has no major negative impact on the results
reported for Beef.

A correlation matrix was constructed for all variables. This indicate that
there is association between Competition and Market risk, and between
Competition and 1st Economic PC, but not between Market risk and 1st
Economic PC.

The number of correlations reported in Tables 6.7, 6.8 and 6.9 indicate
that Competition is contributing to the solution, but at a statistically
questionable level 46 per cent of the time (26 of 54 cases).

There is thus a possible, but not compelling, argument to eliminate the

Competition variable for parsimony. The above may indicate that this variable
is measuring underlying forces that are replicated in the Market risk and 1st
Economic PC variables. It is also possible that the Competition data are at
fault, because they represents competition at the economy-wide level. If intra-
industry competition data were available to be substituted – a desirable matter
that was discussed in section 4.2.4.1.2 – correlation may disappear. Until
future research clarifies the matter, the author’s view is that the variable
Competition should be replaced with intra-industry competition data where
possible, or eliminated in its present form when that is not possible.

7.2.3.2 The Null Hypothesis

Consider Table 6.9. Statistical significance of each major discriminant

function is .000. As is usual with discriminant analysis, some of the 2nd, 3rd
and 4th functions contain small amounts of residual variation from abnormal
scores, but these account for no more than an inconsequential 2.8 per cent of
variance at significance levels of .001 to .020 (well below the cutoff of .05).
Additionally, the discriminant functions produce classification results which
have a simple average 95.4 per cent across the six products – also an
 254

excellent result. The null hypothesis has therefore been rejected at a very
high level of confidence.

The high classification figure of 95.4 per cent is weighed down by one
relatively poor (but in absolute terms, satisfactory) number of 83.3 per cent for
Engines marketed by High Capability Organisations. This product-
organisation combination is perhaps distorted by three factors related to the
market power and oligopolistic nature of this industry. Firstly, location-
distorting incentives are known to be negotiated with governments by many of
the world’s small number of large engine and car manufacturers. This matter
was not taken into account in this research, but a variable could be added to
allow for government incentives. Secondly, further distortion arises because
car engines are usually manufactured in one country but total production is
consumed across several countries. Demand is thus a function of aggregate
demand in a group of countries, not just the country of manufacture. A perfect
world would have the screening variables’ data set enlarged to include the
additional countries supplied by each engine factory, then countries
aggregated into groups related to each factory – both of which are possible –
then statistical analysis undertaken in the usual way. Finally, collusive
agreements within and between organisations almost certainly affect the
countries they manufacture and sell in. Without determining how difficult it
would be to obtain data about this third matter, remedying the first two matters
can be expected to improve the cross validation, probably substantially. In
light of these points, the relatively poor cross validation for High
Organisations’ Engines seems satisfactorily explicable.

Now consider Table 6.11. Education with objectives just fails to

achieve the cross validation quality test of chance plus 25 per cent. However,
the cause is due to an aberration in the properties of the data and is not
actually problematic. Only one country was reclassified (i.e. 191 of 192
countries were correctly classified), so the difference between the MDA and
 255

CA algorithms’ classification is slight47. The failures also relate to the

mathematics of calculating chance. These in part involve squaring the
number of countries in each cluster, and the outcome of this step is a large
number when there are 171 countries in cluster number one. It will thus be
noted in Table 6.10 that the calculation requires a pass rate which exceeds
100 per cent, an impossibility. This notional quality failure is thus insignificant.

7.2.3.3 Relative Dimension Importance

This involves assessing statistical significance, then the relative

importance, of each variable.

The final Summary of Table 6.12 suggested that seven of the 11

variables were consistently statistically significant. However, Product
consumption, Product consumption growth, Market research and Contacts
were frequently not statistically significant, meaning that they may not be
useful independent variables. That finding is interesting because it questions
espoused theory and intuitive sense. In the case of the first two variables, it
also questions research hypothesis three in this work and Russow’s 1989
empirical finding that these variables are useful.

The apparent lack of support for the third research hypothesis (that
Product consumption and Product consumption growth are the most important
characteristics for discriminating among country groups) will thus be examined
and challenged.

Interpretation of statistical significance of variables has a complexity. It

is possible for a variable to be a statistically significant determinant of market
attractiveness, yet to fail at being a statistically significant determinant of
market (country) groups because of properties of the variable’s data. The
following diagram perhaps explains the point:

47
The country they differ over is Russia, a market whose data is at times considered extreme
by the MDA algorithms when calculating many of the Education solutions.
 256

Figure 7.1
Two Types Of Discrimination Undertaken By Each Relevant
Predictor Variable

Variable

Discriminates Does not

market discriminate
attractiveness market
(e.g. product attractiveness
demand)

Discriminates Does not

market groups discriminate
(e.g. adequate market groups
variance,
little correlation)

Variable gets
used for
screening
markets
(<.05 significance)

To be statistically significant, a variable has to pass two hurdles – to

distinguish attractive markets and to have data properties which allow it to
distinguish market groups.

Why the four important variables seem unimportant can be seen at

Appendix J among the plots of Country Scores By Variable which use
Education as an example. The three sets of plots (for Best Markets, Low
Capability Organisations and High Capability Organisations) of the seven
common variables all look the same – only the weightings of each variable
have changed, not the relationships amongst countries. The plots reveal that
Product consumption, Product consumption growth, SCA, Market research
 257

and Contacts have much less absolute and relative variation among their data
than do the data for the other Education variables. This limited variance
sometimes reduces the statistical significance of the variable, and so the
ability of the CA and MDA algorithms to use data in those five variables to
discriminate among Education markets. There is insufficient information in
each of these variables to allow them to fully discriminate between markets.

However, there are situations where these variables contribute to

market selections. An example of where the same variable is both useful and
non-useful at discriminating market groups can be seen in the plots for
Product consumption growth. An almost straight line is visible and there are
only seven countries clearly separate above and below the common line
(naturally in all three plots). Referring now to Table 6.12 the limited variation
amongst countries’ data in this variable is:

• sufficient to discriminate market groups when there are only seven

variables (the Best Markets situation),

• insufficient when one more variable is added (the Best Markets for
Low Organisations situation). Here the re-weighted but otherwise
identical data for seven variables have data on another powerful
variable, Language, added. The interaction of the properties of the
Language data, in conjunction with the properties of the data
comprising the other variables, is such as to overwhelm the impact
of Product consumption growth. Note how the F test figures change
(Appendix K),

• again sufficient to discriminate market groups when even more

variables are added for the Best Markets for High Organisations
solution.

The cause of the vacillation in the statistical significance of this same

variable (comprising the same data in all three situations) is the interaction
 258

between the properties of its data and the properties of the data representing
the other variables. In only two of the three situations is the variance among
country data sufficient to allow the variable to contribute to the pooled affects
from all variables’ data to distinguish between both markets’ attractiveness
and clusters of markets. The lack of variance almost certainly partly relates to
the raw demand data often not having much variance, and partly to the use of
the shift-share method of assessment of demand growth. Product
consumption and Product consumption growth should therefore be retained
as independent variables.

Therefore, whilst there was not statistical support for the hypothesis
that Product consumption and Product consumption growth are the most
important variables, it would seem likely that they are important – perhaps the
most important – variables from the theoretical and practitioner viewpoints.
After all, if there is inadequate demand or demand growth to attract the
organisation to the market, all other aspects of the market are irrelevant.
Determining the relative importance of demand and demand growth is
important and needs to be resolved by future research, but the answer is
easily guessed to be likely to be resolved in favour of the two variables.

The variable with the weakest case for retention is Market research. It
is significant in none of six cases according to the Structure Matrix method
(the final Summary to Table 6.12). However, lack of variance in data
comprising the variable is again involved (see its plot, Appendix J). Also recall
that the methodology employed by this research was to randomly allocate
scores for this variable to countries, whereas the commercial norm would be
for the organisation to have conducted research on countries which
(hopefully) had been selected for rational if imperfect commercial reasons.
That research should contribute to the screening process because it reflects
an aspect of the organisation’s knowledge, and knowledge is a matter which
influences the optimal markets the organisation is capable of efficiently and
effectively handling. Note also that the F Statistic assesses Market research
as being mildly useful at polishing selections in five of the six situations (Table
 259

6.12). For these various reasons it is thus desirable to retain Market research
until more quantitative experience indicates whether to retain or drop it.

A significant point emerging from this part of the research is that all the
11 variables tested contributed statistically to market selections for some
products – different combinations of all 11 variables were necessary to
achieve the 95 per cent variance accounted for and the 95 per cent cross-
validation rate. Whether or not a particular variable is going to contribute to
the markets selected for a particular product and organisation type is not
known until after having conducted the market selection then analysing the
statistical tests. With time, in the interest of parsimony, experience might
suggest that a reduction in the number of variables is feasible, but that is not
yet appropriate.

There is one final by-product to mention about the finding that a

variable’s influence at market selection is affected by the variance of its data.
Selection of the optimal cluster can not be made using the mean of the
variable centroids, because that mean would incorporate differing levels of
influence of each variable. Selecting the preferred cluster by taking the one
with the highest centroid values for the most variables prevents low variance
variables from being under-represented. For example, Table 6.4 shows
cluster three to be the preferred cluster for Beef in the Low organisation
situation. However, if means of centroids of variables had been used, cluster
two would instead have been selected because the means of the three
clusters are .201 (cluster 1), .370 (cluster 2) and .336 (cluster 3). Yet cluster
two contains clearly low suitability markets such as Rwanda and Somalia.

Having determined that all variables have a role, it would be useful to

know their relative importance in determining market selections. Remember
that there is no unambiguous measure of the relative importance of predictor
variables in discriminating between the groups when there are two or more
functions, as is the case with this research. Nevertheless, Appendix K and
the variable hierarchy table produced from it, Table 6.13, give some
indicators. We will first examine the minutia then draw general conclusions.
 260

• Best Markets:

Competition is ranked as the most important of the seven variables

in eight of the 18 assessments. It plays a negligible role in the
market selections for Wool and Engines. However, this variable
is probably either faulty or redundant as argued in the section on
Collinearity and Multicollinearity;
Legal follows as the top ranked variable in six of the 18
assessments and is 2nd in three assessment;
1st Economic PC, Market risk and Cultural distance play weak to
moderate roles, depending on product and assessment method
considered;
Contrary to theory and practice, the statistical results suggest that
neither Product consumption nor Product consumption growth
appear to play an important role in any consistent way when
selecting markets for most of these particular products. A stark
exception is that Product consumption is the single most
important variable for selecting Engine markets (and by all three
assessment methods). Product consumption and Product
consumption growth are the two variables which most frequently
fail the ± .33 hurdle in the Structure Matrix, superficially
indicating they are unimportant variables when assessed by the
Structure Matrix method, and so also the Potency Index method.
The F Statistic method typically rates these two variables as the
least important of the seven variables. However, low variance of
data is the probable cause, as discussed. Assessing the
importance of these two variables is complex and continues to
be discussed below;
Summarising the Markets Only situation, Competition or Legal,
supported by 1st Economic PC, Market risk and Cultural
difference, usually drive the markets selected. However, the
relative importance of each variable is different for each product,
 261

and there is no consistent pattern of variable hierarchy which

could be applied to all products.

• Best Markets for Low Capability Organisations:

Language and Legal fairness are clearly and uniformly the 1st and
2nd ranked variables by all three assessment methods for all six
products. Language is top in 18 of 18 assessments. Legal
fairness is 2nd in 13 of 18 (but fails both the hurdles of having
eigenvalues of ≥ 1.0, and accounting for ≥ 5 per cent of
variance in two of the six structure matrices);
The size of Language’s correlation coefficients (in the Rotated
Structure Matrices) is consistently huge in relation to the
correlation coefficients of the other variables, and the first
function accounts for 90 to 96 per cent of the variance. In
addition, the size of Language’s F statistic is consistently huge in
relation to those of the other variables. Language thus clearly
dominates all other variables in importance, and very strongly,
and is consistently the principal determinant of the markets
selected for all products for Low Capability organisations;
Whilst Legal is consistently the next most important, the Structure
Matrices and F statistics show that its selection power is only
moderate;
Other variables operate weakly (perhaps trivially) at the margin.
Market risk, Competition and 1st Economic principal component
are commonly, but not always, the 3rd, 4th and 5th most important
variables. However, the F Statistic method disagrees with the
Structure Matrix and Potency Index methods when hierarchying
the importance of these three variables;
Product consumption and Product consumption growth fail the ± .33
correlation hurdle in almost all Structure Matrices. The two
occasions when Product consumption passes ± .33, it accounts
for so little variance that it is of inconsequential import in the
 262

Structure Matrix method of assessing variable importance. The

F statistic method of assessment rates these two variables as
being a little more important than the Structure Matrix and
Potency Index methods, but still of minor influence. However, as
discussed, insufficient variance in the data comprising these two
variables accounts for the two variables’ apparent lack of
importance;
Low organisations with objectives are discussed in a separate
section below;
In summary, when selecting markets for Low Capability
organisations, Language is unequivocally the clearly dominant
matter to select by. Legal comes a clear second, but is of
marginal import. Other variables have negligible impact, and
their relative importance varies by product.

• Best Markets for High Capability Organisations:

Language and Legal are never as important to market selections for

the High Organisation as to selections for the Low Organisation.
They pair up together to be influential at determining markets
selected in only one product (Telephones) as assessed by only
one method (the F Statistic) – and then their relative importance
is reversed, with Language the minor player and Legal the more
important of the two;
Legal is sometimes an important variable (e.g. Telephones) and is
sometimes fairly unimportant (e.g. Coal). However, it often has
a useful level of influence;
In three of the six products, Language fails the ± .33 hurdle, and in
a 4th it is the principal variable in a discriminant function which
accounts for less than 5 per cent of variance. In the two cases
that it survives, Language is ranked by the Potency Index
method as the second least important variable. The F Statistic
method assesses it as being of strong to weak importance;
 263

There is agreement between all three methods of assessment that

the most important variables are: Competition for Coal, SCA
(country-specific) for Beef, Legal fairness for Telephones,
Contacts for Education, and Product consumption for Engines.
So different products are determined by different principal
variables;
Product consumption fails the ± .33 correlation hurdle in four of the
six products, yet is the most important variable when
determining best markets for Engines. Product consumption
growth fails the ± .33 correlation hurdle in six of the six products.
Comment has already been made that inadequate variance is
the cause of the misrepresenting of the importance of these two
variables;
The other variables all play a useful role at different times. Most
vary widely between being an extremely important determinant
of markets for one product to being irrelevant to market
selections for another product. The weakest contributor is
Market research, which has previously been argued to have
special data circumstances which do not justify its removal at
this time;
In summary, all variables contributed at a useful level to selecting
markets for one product or another for High Capability
Organisations. There is also no uniform hierarchy of variable
importance. There is great variation in importance of most
variables, which may be of dominating importance when
selecting markets one product yet irrelevant to another product.
The markets for High Organisations are determined by different
variables for each product.

• Best Markets for Low and High Capability Organisations with

Objectives
 264

Product consumption becomes the dominating variable. No other

variables are significant, for either low or high organisations, as
assessed by the structure matrix of potency index methods.
That is despite the variable weights being different (.35 Low
organisation, .55 High organisation), and despite the previously
discussed lack of variance in the variable which normally
reducing its country discriminating power to less than statistically
probable value.
Using the F to remove method, Cultural difference, Competition and
Market risk are the next most important variables, and for both
Low and High organisations. This is despite the organisations’
objectives re-weighting these variables to only a fraction of their
earlier influence48.
The previously most important variables, Language (Low
organisation) and Contacts (High organisation) have been
overwhelmed by the re-weighting for organisation objectives.
This suggests that the re-weighting requires sensitivity and so
needs to be carefully controlled. Sensitivity analysis is needed
to determine the statistical impacts of objectives to ensure not
unintentionally directing the organisation to markets that it is not
really interested or capable in.

How do we summarise the results from these three situations? Is there

some cohesive theory which we can perceive? The general rules to emerge
are:

• All 11 variables tested were found to contribute to market selections

for some products and organisation. Regrettable from the point of
view of parsimony, no variables could confidently be recommended
for elimination as a result of this research. However, there is need

48
Prior to re-weighting for objectives, the Low organisation weight for each of these 3
variables is 1/8th or .125. After objectives’ weights are applied, the weight of each of these
variables’ becomes 1/8th x .2 x 1/8th = .005. The High organisation re-weights go from 1/11th
or .091 to 1/11th x .15 x 1/8th = .002.
 265

to pursue determining the usefulness of the Competition variable

and the Market research variable;

• Product consumption and Product consumption growth superficially

tested as statistically irrelevant but additional analysis argued that
they should be retained. Future researchers may need to apply
caution to ensure correct interpretation of these two variables;

• For organisations of absolutely minimal abilities and values,

Language should be an extremely important determinant of which
markets they should go to. However, as soon as organisations
increase their abilities and values, this finding probably ceases to
apply to them. This point may therefore be of theoretical interest
but not of practical importance49;

• Variables vary greatly in their levels of importance in each of the

three situations tested (Markets only, Low Organisations and High
Organisations). The cause is that the data, and the properties of
the data, determine which markets are best for each situation. This
produces two major consequences:
♦ All organisations need to use all 11 variables, adjusted for
their levels of ability and values, to calculate the unique
solution for the markets which are best for them. (Phrasing
that point differently to emphasise it, the screened markets
are for a specific organisation-market combination, and so
not useful to other organisations selling the same product);
♦ It is most unwise for governments to produce generalisations
about which markets are “best”. “Best” relates to a
combination of organisation and market characteristics, not

49
Remember that the research was constructed with ‘Low’ and ‘High’ organisations as the
extremes on a continuum representing the limits of practical operating capability and
psychological values. Whilst sensitivity research would need to be undertaken to determine
how much movement along the continuum would be necessary to eliminate Language’s
dominance, the researcher’s sense of the behaviour of the data is that only a small change in
firm characteristics is likely to produce a rapid reduction in Language’s influence.
 266

just to market characteristics. “Best” for one organisation will

not be “best” for many others;

• Nothing can be said about the statistical significance or relative

importance of the eight organisational abilities and preferences
variables and the four objectives variables. Their use to re-weight
the 7-11 market variables definitely changed the markets selected,
and their value can be argued from the literature, but inability to
apply statistical tests to each of them leaves uncertain the role of
each;

• Re-weighting variables for the organisation’s objectives again

clearly alters the relative importance of each variable used to select
markets, and so alters markets selected. Since time allowed
examination of only a single product, Education, there is not a lot of
experience to extrapolate from. It would seem that sensitivity
analysis should be used to determine the way that different
organisational weights can bias the markets selected.
Organisations could then be warned about the selection biases
produced by different organisational weight ranges so they can
make an informed choice about their preferred markets.

7.2.3.4 Individualising the Results for an Organisation

It would have been useful to assess the relative importance of the

organisation-specific capabilities and preferences at discriminating markets.
For example, are these more, or less, important than the market-related
variables at determining the markets chosen? However, it was not possible to
detect the relative influence of the organisation’s capabilities and preference,
given that they appear as weights attached to other variables. The problem
arises because these organisation characteristics do not relate to a specific
market, so cannot directly influence selection of a particular market.
 267

What this research has shown is that adjusting weights of the market
variables for firm capabilities and preferences has produced clearly different
cluster centroids, cluster solutions and sets of markets for low firms,
compared with high firms. The statistical procedures used here have
confirmed the intuitive notion that generic capabilities and values makes a
difference to markets selected.

The second matter requiring coverage under this heading is the

difference made by adding objectives to a given level of abilities and values.
Adding objectives changed the markets suggested for the low and high firm.
However, in the single product examined, the same markets were suggested,
irrespective of firm competence. However, cluster centroids suggest that the
hierarchy of preferred markets will be different for the two organisations.
Definitively determining the answer requires further research.

7.2.3.5 Comparing Engine Results with Russow (1989)

Within the pre-specified constraint that the data periods were different,
results obtained here were compared with those of Russow (1989) to assess
similarity of results, to critique whether his results are source-country neutral
and to consider whether or not adding variables to Russow’s demand-side
only factors causes different markets to be selected.

This research’s Markets only results are the comparable category – as

Russow does not take organisation attributes into account. Russow found 26
countries to have high potential for engines (1989, p. 197), whereas this
research found there to be five countries. Four of this research’s five
countries are members of Russow’s list (viz Germany, Japan, South Korea
and the USA). China is not included by Russow, but the different time period
could be a part explanation because China has since emerged economically.
The conclusion is that there is substantial difference between this and
Russow’s assessments of the high potential markets for engines.
 268

Regarding source country neutrality, appearance of the same four

countries suggests neutrality. However, it could alternatively be argued that
the absence here of so many of Russow’s countries indicates a source
country bias. The dramatic difference in starting number of high potential
countries makes it difficult to assess this point. Inspection of Russow’s
variables indicates no reason for believing source country bias because all his
variables are of an economic nature and assess countries without
consideration of the country of the assessor.

Regarding adding variables to Russow’s demand-side only factors,

again the starting point of Russow’s 26 versus five countries here makes it
difficult to assess the matter. Logic argues that adding variables will alter the
market selection but the matter was not able to be assessed empirically.

Comparison between Russow and this work has been undertaken

because it was set as an original objective, but the exercise is pointless
because the data used in the two researches are radically different. Firstly, it
was known at the start that Russow’s data set was based on a different time
envelope (1975 – 1980 versus 1990 – 1995). Secondly, it appears that
Russow included markets which did not actually produce or consume engines
(because he treated 173 markets whereas there are now only 42). Close
inspection of the UN data set which Russow used shows the cause to be that
it includes: (1) not only engines, but also engine parts (pistons, carburettors,
etc.), and (2) both petrol and diesel engines. This research data includes only
petrol engines so there is a smaller set of markets involved. The UN data
were faulty. There is therefore no opportunity for valid comparison between
the two research works.
 269

CHAPTER 8

MAIN FINDINGS, LIMITATIONS

AND RECOMMENDATIONS

In heaven an angel is nobody in particular.

George Bernard Shaw, Man and Superman, Maxims for
Revolutionists

8.1 Introduction

Chapter eight commences with a summary of the seven previous

chapters in the dissertation, provides a discussion of the limitations of this
research, and then closes with recommendations for future research.

8.2 Summary of Chapters One to Seven

The aim of this research was to develop an international market

selection methodology which selects highly attractive markets, however
defined, and which allows for the diversity in organisations, markets50 and
products.

Conventional business thought is that, every two to five years, dynamic

organisations which conduct business internationally should decide which
additional foreign market or markets to next enter. If they are internationally
inexperienced, this will be their first market; if they are experienced, it might

50
Whilst markets commonly flow across national borders, aggregate statistics rarely do so.
For the purpose of this study and consistent with the literature, markets are defined at the
level of countries.
 270

be, say, their 100th market. How should each select their next international
market?

This is not an easy decision. There are some 230 countries in the
world. Each has its own characteristics in terms of market attractiveness.
Moreover, judging from the relevant literature, a well-informed selection
decision could consider over 150 variables that measure aspects of each
foreign market’s economic, political, legal, cultural, technical and physical
environments (Root 1994; Russow 1989), along with the organisation’s
resources, managerial abilities and preferences. Considering even a fraction
of these variables for a reasonable number of countries would, in the
overwhelming majority of cases, be beyond the resources of the decision-
maker in terms of time, money and expertise. The relevant literature
suggests, therefore, that international market selection ought to be broken up
into two or more stages (see, for example, Ball & McCulloch 1982; Douglas,
Le Maire & Wind 1972; Douglas, Craig & Keegan 1982; Liander et al. 1967;
Root 1987; Russow 1989; Russow & Okoroafo 1996; Samli 1972; Toyne &
Walters 1993; Walvoord 1980; and, for comprehensive surveys,
Papadopoulos & Denis 1988 and Samli 1977).

The first stage involves screening a relatively large number of countries

on the basis of a select, yet limited, number of criteria. The outcome of this
(screening) stage is the selection of a much smaller sub-sample of countries.
The screening stage ought to not only result in the selection of the right group
of markets. As Root (1994, p.55) observes, it should also minimise two
errors: (1) ignoring countries that offer good prospects for the organisation’s
product, and: (2) spending too much time investigating countries that are poor
prospects. Root argues that the first of these errors can be avoided by
including all countries in the initial sample – a significant matter in terms of
resources and computing sophistication. A more intensive, in-depth
assessment of the market potential of the sub-sample of countries selected in
the completion of the screening stage is then recommended during the
second and later stages of market selection. This in-depth assessment
involves consideration of a larger number of variables, possibly measuring
 271

such factors as industry-specific and/or company-specific sales potential,

which may include data gained from primary sources. The outcome of this
stage is the selection of an even smaller sub-sample (usually, but not
necessarily, one country), to which the organisation typically exports.

However, the screening decision can not presently be made properly

because there is insufficient knowledge about which of the abovementioned
150 or more decision variables to use. An unsolicited order (found by Bilkey
(1978) to be the market selection mechanism used by 67 per cent of firms),
the language spoken by a key manager, or even where a manager had
holidayed overseas, are just a few of the common practical geneses. The
consequence is that, in Australia alone, an estimated $A225,000,000 per
annum (see section 1.1) is wasted every year as a result of inefficient and
ineffective foreign market selections.

Empirically determining which variables are important requires

developing a large database. Sufficiently accurate data about meaningfully
defined products to enter into the database only began to become available
during the 1980s. Cross-disciplinary knowledge of economics and psychology
is required. And even a powerful personal computer takes several minutes to
perform the 3,000,000 calculations required for most of the market selection
cluster analyses. The result is that Russow (1989) is the only previous
researcher to have attempted to empirically determine which variables are key
when selecting foreign markets.

This study began by collecting all market selection variables suggested

by 350 items in the literature. It then rationalised them into a comprehensive,
eclectic, theory-driven framework and tested them empirically using principal
components analysis, cluster analysis and multiple discrimination analysis. Of
the 46 potential variables, 30 were used as representative of the 38 that can
be quantified.

Of the framework’s 30 operationalised dimensions:

 272

• seven variables captured attributes of each market (e.g. product

consumption; their economies);
• four variables captured the organisation’s abilities in each market
(e.g. languages; contacts);
• nine variables captured organisation abilities and psychological
values (e.g. general business competencies; values about further
internationalising);
• four variables captured organisation goals (e.g. sales growth; risk
minimisation); whilst
• six variables are required for technical purposes (e.g. to assess
data quality and availability; to set the criteria and method of market
evaluation).

There are eight additional variables which this study was not able to
operationalise as quantifiable criteria (e.g. strategy, marketing mix). A further
eight variables were not required for the particular products and organisations
assessed (e.g. product physical attributes; organisation objective of other
business benefits).

A database of some 10,000 statistics was developed covering six

products and approximately 230 countries. Scores were constructed for two
hypothetical organisations having respectively the minimum necessary, and
the maximum possible, abilities and attitudes. The data were then treated
using the statistical techniques of principal components analysis, cluster
analysis and multiple discriminant analysis. The objective of principal
components analysis was to compress four economic sub-variables into a
single component which was then used to indicate the state of the economy in
each country. Cluster analysis grouped markets into sub-sets having differing
attractiveness. Discriminant analysis tested the cluster analyses, assessed
relative importance of variables, and added shades of understanding about
the solutions developed. Three sets of results were computed for each
product – for a best markets only, for a low capability and values organisation,
 273

and for a high capability and values organisation51. These 18 solutions were
supplemented with two solutions for one product to demonstrate how to add
organisational objectives to the market screening procedures.

The approach herein is prescriptive, being that of the maximising

rational economic person or Allison’s unitary actor (Allison 1971; Schoemaker
1993). It was originally suggested by Liander et al. (1967). Subsequent
decision sub-optimisations due to organisational politics, strategy, human
limits and so on are not considered.

Using the 30 variables produced results which typically accounted for

95 per cent of variance at cross-classification rates of 95 per cent. Significant
findings of the research are as follows:

1) The procedures succeed at allocating markets to groups having

differing levels of attractiveness,
2) Different groups of markets are appropriate (“best”) for
organisations having different abilities and values,
3) Findings (1) and (2) mean that a “One size fits all” approach to
international market selection is not correct. The “best” (most
attractive and appropriate) markets for organisations of high
international capability will usually be different from those for low
capability organisations. The specific attributes of the
organisation (competencies and values) help determine the
market in which it will harvest the largest profit. Governments
are thus incorrect when they indicate that particular markets are
the most appropriate for all their organisations to target,
4) All seven variables describing the markets, and all four variables
describing the organisation’s abilities in each market, were found
to have a role in producing solutions for one or more of the 20
product-organisation combinations tested,

51
As detailed later, these 3 situations consider attributes of the market only, attributes of the
market and a low capability organisation, and attributes of the market and a high capability
organisation.
 274

5) The strength of their importance differed by product and the

organisation’s level of competence,
6) There is no universal rule about which variables are most
important in all situations. However, there are indications about
which variable will dominate market selections for a specific
product and level of organisational competence,
7) The other 19 variables used by the framework describe the
organisation and prescribe the market selection system. They
also all appear to play roles in determining which markets are
most attractive and appropriate but the level of certainty about
this is not as high as for the above 11 variables.

The contributions to knowledge made by this work are that it has:

• collected data on arguably all the influential variables mentioned in

the literature,
• developed a predictive framework within which to house those
variables (see Figure 4.3),
• extended Russow’s (1989) demand-side only market assessment to
include firm-specific and business environment considerations
which affect the market screening decision,
• undertaken the second quantitative test of market screening,
producing results which typically accounted for 95 per cent of
variance at cross-classification rates of 95 per cent, and
• uncovered two significant weaknesses in the cluster analysis
methodology recommended by the lineage descended from Liander
et al. (1967).
 275

8.3 Limitations

Generalisability: One swallow a summer doth not make. Rephrasing

the old saying, Six products a generalisable result doth not make. Whilst the
research design used here was carefully devised to minimise faults,
replication is needed to ensure that we do indeed now have a tool to use to
sensibly pick foreign markets for organisations. Replicating using different
products and time periods is desirable, but most important is volume – to do a
large number of market selections so as to accumulate statistical and
qualitative knowledge about the typical pattern, and any variations from the
norm.

Un-operationalised variables: There were eight variables which were

not quantified, three of which it would be desirable to operationalise, i.e.
Strategic values, Personal benefits and Other personal values. Using these
will only slightly improve the technical accuracy of market selection because
this research left unexplained only 5 per cent of variance. The more valuable
likelihood is that these additional variables will finesse the markets selected in
a qualitative way by picking markets which better meet personal and strategic
desires.

Services: Neither Russow (1989) nor this work included services

products to the extent desirable. The argument for doing so is that, in some
circumstances, services may behave differently from tangible products. The
reason that this research only examined education was (and still is) the
shortage of service product-country data. On present indications, noticeable
improvements to the volume of services product statistics will be available in
about ten years. That would enable determination of whether most services
products are similarly affected by the screening variables examined here. It
can be said, however, that this researcher’s view is that education and the five
tangible products examined here all behaved in a similar way – although each
product responded to different screening variables. On that limited sample,
and whilst we await a more researched answer, the expectancy can
 276

provisionally be that there will be only minor differences in the way that
tangibles and services respond to international market screening.

Field testing the Management Questionnaire: The Questionnaire’s

questions have been developed from theoretical constructs, but most
questions have not been tested on a body of people to test validity and
reliability to eliminate ambiguity. That work remains to be done.

Telephones production data: This product required use of SITC

Revision 3 for trade data and ISIC for production data. ISIC52 data are not
always comprehensive because there is ambiguous support by nations for
provision of production statistics. Prior to use, the UN Telephones production
data were cross-checked against statistics collected by the researcher from
some individual countries’ statistics bureaux, and these all matched. The
product also behaved statistically and qualitatively in the same way as the
other products. However, there can not be quite the same level of confidence
in the telephone production data as in the other data53.

Number of markets selected: CA is unable to reduce the size of the set

of preferred markets, nor hierarchy the set. Whilst the literature shows that
this is unimportant to researchers, it is a major inadequacy to practitioners.
As argued earlier, firms need a maximum of approximately six markets to
consider during in-depth assessment. This appears to be a previously
overlooked and fatal flaw in CA’s use.

8.4 Recommendations for Future Research

Data Variance: This research found lack of support for research

hypothesis three, which predicted that Product consumption and Product
consumption growth would be the most important screening variables. The
cause was traced back to a dramatic lack of variance in data comprising these

52
The International Standard Industrial Classification.
53
In contrast, the other 5 products’ data came from single source, specialist agencies in
whom there is a very high level of confidence.
 277

two variables – lack of variance in the raw demand data, perhaps accentuated
by the shift-share method of calculating demand growth. Inadequate variance
lead to diminished statistical significance of the variable. CA and MDA were
then unable to discriminate between markets using these two variables, so
then made the market selection decision using the remaining significant
variables. It is thus recommended that additional products be screened to
confirm this finding. It will thereafter be necessary to assess whether these
two variables are indeed the most important ones in selection of both
attractive markets and attractive market groups. Intuition plus capitalist
motivation suggest they must be, but proof is necessary to affirm their
statistical role in market screening.

Efficacy of Cluster Analysis: This research found deficiencies in CA,

namely that: (1) inadequate variance in data reduces statistical significance of
variables then reduces the influence of that variable when CA selects
markets, and (2) CA is unable to reduce the list of attractive markets to a
practical number (e.g. the top six markets). This challenges the efficacy of CA
as a screening tool. Future researchers may decide that alternative statistical
methods need to be employed.

Sensitivity Analysis of Objectives: Applying weights for organisational

objectives greatly changed the relative importance of the Markets only
variables, and so the markets selected. Further, applying objectives weights
to the Low and High organisation again changed the markets selected from
the Low and High firms without objectives. However, the same markets were
suggested for the Low and High competency Education firms with objectives.

Sensitivity analysis could confirm whether objectives do select different

markets for Low and High firms with objectives. It could also determine the
way that specific organisational goals can bias the markets selected. That
analysis may reveal limits on the range for each objectives variable, and
implications about particular values (and combinations) of, each objective.
For example, it may be unwise for Effort to be set at less that a certain figure if
that would result in the organisation facing an insurmountable level of
 278

competitive, cultural, legal and economic forces. Organisations could thus be

warned about the implications of different organisational objectives weight
ranges so they could make an informed choice about their preferred markets.

Reducing the Number of Variables: An unfulfilled hope of this research

was to find variables which were redundant. That would simplify the
screening task by reducing the volume of data required and the size of the
statistical task. Instead, it was found that variable importance changes with
the product. It would be pleasing if future research achieved this aim after
examining more products.

Selection Efficacy: Comparisons are needed of screened

recommendations versus the specific organisation’s commercial outcome in
those markets. That is not an easy task, remembering that such events as
the break up of the USSR, stock market gyrations and staff changes are only
some of the discontinuities which distort the implementing of a market entry.
Nevertheless, we will never know whether the screening framework is more
than theoretically correct unless we endeavour to compare recommendations
with outcomes.

Possible Additional or Altered Dimensions

1) Economy: As discussed in section 7.2.1, it is desirable for future

research to explore whether to increase the number of economic
variables from one to two54 to determine whether that makes
marked differences to the markets selected. The purpose is to
increase the influence of future market size, and so the influence
of future market potential, on the framework’s computations.
This would seem particularly important for products whose sales
are driven by increases in both population size and wealth
(presumably many consumer and others products).

54
by including not only the first principal component, but also the second component.
 279

2) Organisational culture: Another variable to consider is

organisational culture. This appeared in the initial research by
Holzmüller and Stöttinger (1996), but almost disappeared during
their second round of analysis. The variable may have an
impact on the psychic stress of the internationalising managers,
then impact the organisation’s foreign orientation, then the
organisation’s international performance.

3) Market risk: Conceptual purity argues for a slight reduction to the

market risk level for each of the first (say) 10 markets entered, to
recognise the reduced risk accruing to the organisation from
diversifying its markets. This is diversification risk, as distinct
from specific risk. The amount is believed to be quite small and
so was ignored during this research. However, future research
should consider slightly improving the screening framework’s
accuracy by making the appropriate adjustment.
 280

APPENDICES
to
INTERNATIONAL MARKET SELECTION –
SCREENING TECHNIQUE:

Replacing intuition with a multidimensional framework

to select a short-list of countries

A doctoral dissertation

Richard R. Gould

R M I T University
Faculty of Constructed Environment
School of Social Science & Planning
GPO Box 2476G
Melbourne Vic 3001
Australia

Email: [email protected]
[email protected]

August, 2002
 281

APPENDIX A

MANAGEMENT QUESTIONNAIRE

This questionnaire needs to be completed by the organisation’s chief operating officer, by each of the individuals in the dominant
coalition of managers, and by the functional specialists who will be deeply involved in expanding the international activities of the
organisation.

A presumption is that your organisation is capable of trading internationally. If in doubt, consider taking the brief “Export
Readiness Test” available from the Australian Trade Commission, and on their world wide web page.

A further presumption is that your organisation’s management is keen to further internationalise, and the question to answer now
is which country you should go to. If you are not indeed highly motivated to further internationalise, it is a waste of time
completing the questionnaire.

Resources

(Show respondent the typical sales and profits graphs e.g. Kotler et al. (1994, p. 308).)
Foreign market entry, to break-even profit stage, typically takes most organisations more than 3 years. Do you expect your
organisation to be different?
Yes / No / Don’t know

Briefly consider what financial, physical, personnel and intangible resources will probably be needed to supply the new foreign
market/s. Also consider how a successful market entry will require management to make unusual mental efforts and be diverted
from their normal duties. Do you feel reasonably confident that you will be able to: (1) marshal whatever resources are
necessary and (2) sustain them for the several years it will take to reach break-even profitability?
Yes / No / Don’t know
 282

If you did not answer “Yes” to the last question, please discontinue completing this questionnaire and seek further advice
because you are indicating a potentially significant inability to further internationalise.

Strategy decisions

The following are implicit in the decisions you make about the market you select. The screening decision framework does not
directly take them into account, but you need to be aware that they are embedded in the decisions you are making. Please
clarify your general preferences by circling the preferred alternative:
For each new foreign market entered, is the preferred strategy to use:
Mass, segmented or niche marketing?
Product standardisation or adaptation?
Competing on cost, differentiation or focus?
Competition or collaboration?
If collaboration, whether it will be direct or indirect?
If competition is chosen, whether the posture will be one of defence, offence, flank or niche?
via pre-emption, confrontation or build up?
Country concentration or diversification?
Incremental or simultaneous market entries?
Becoming the market leader, challenger, follower or nicher?

Every organisation is different in its special capabilities and preferences. Data about you and your organisation needs to be
collected so the market screening process can match you with your ideal additional foreign market. Please answer the following,
bearing in mind that, from here, there are no ‘right’ or ‘wrong’ answers.

Hofstede Survey of Personal Values (Indicative instrument only. Not for actual use – see section 4.2.4.2.2.3 (a) for
reason.)

A1. Are you:

1. Male (married)
 283

2. Male (unmarried)
3. Female (married)
4. Female (unmarried)

A2. How long have you been employed by this organisation?

1. Less than 1 year
2. One year or longer, but less than 3 years
3. Three years or longer, but less than 7 years
4. Seven years or longer, but less than 15 years
5. Fifteen years or longer

About your goals:

People differ in what is important to them in a job. In this section, we have listed a number of factors which people might
want in their work. We are asking you to indicate how important each of these is to you.

In completing the following section, try to think of those factors which would be important to you in an ideal job; disregard
the extent to which they are contained in your present job.

PLEASE NOTE: Although you may consider many of the factors listed as important, you should use the rating “of utmost
importance” only for those items which are of the most importance to you.

With regard to each item, you will be answering the general question:
“HOW IMPORTANT IS IT TO YOU TO …”
(Choose one answer for each line across)
 284

of little importance
importance to me

of very little or no
very important

of moderate
importance

importance
of utmost
How important is it to you to:

A5. Have challenging work to do – work from which you can get 1 2 3 4 5
a personal sense of accomplishment? (IDV, MAS)
A6. Live in an area desirable to you and your family? (IDV, MAS) 1 2 3 4 5
A7. Have an opportunity for high earnings? (IDV, MAS) 1 2 3 4 5
A8. Work with people who cooperate well with one another? 1 2 3 4 5
(IDV, MAS)
A9. Have training opportunities (to improve your skills or to learn 1 2 3 4 5
new skills)? (IDV, MAS)
A10. Have good fringe benefits? (IDV, MAS) 1 2 3 4 5
A11. Get the recognition you deserve when you do a good job? 1 2 3 4 5
(IDV, MAS)
A12. Have good physical working conditions (good ventilation 1 2 3 4 5
and lighting, adequate work space, etc.)? (IDV, MAS)
 285

A13. Have considerable freedom to adopt your own approach to 1 2 3 4 5

the job? (IDV, MAS)
A14. Have the security that you will be able to work for your 1 2 3 4 5
organisation as long as you want to? (IDV, MAS)
A15. Have an opportunity for advancement to higher level jobs? 1 2 3 4 5
(IDV, MAS)
A16. Have a god working relationship with your manager? (IDV, 1 2 3 4 5
MAS)
A17. Fully use your skills and abilities on the job? (IDV, MAS) 1 2 3 4 5
A18. Have a job which leaves you sufficient time for your 1 2 3 4 5
personal or family life? (IDV, MAS)

About the satisfaction of your goals:

In the preceding questions, we asked you what you want in a job. Now, as compared with what you want, how satisfied
are you at present with:

A37. How often do you feel tense or nervous at work? (UAI)

1. I always feel this way
2. Usually
3. Sometimes
4. Seldom
5. I never feel this way.

A43. How long do you think you will continue working for this organisation? (UAI)
1. Two years at the most
2. From two to five years
 286

3. More than five years (but I probably will leave before I retire)
4. Until I retire

The descriptions below apply to four different types of managers. First, please read through these descriptions:

Manager 1 Usually makes his/her decisions promptly and communicates them to his/her subordinates clearly and firmly.
Expects them to carry out the decisions loyally and without raising difficulties
Manager 2 Usually makes his/her decisions promptly, but, before going ahead, tries to explain them fully to his/her
subordinates. Gives them the reasons for the decisions and answers whatever questions they may have
Manager 3 Usually consults with his/her subordinates before he/she reaches his/her decision. He/she then expects all to work
loyally to implement it whether or not it is in accordance with the advice they gave
Manager 4 Usually calls a meeting of his/her subordinates when there is an important decision to be made. Puts the problem
before the group and tries to obtain consensus. If he/she obtains consensus, he/she accepts this as the decision. If
consensus is impossible, he/she usually makes the decision him/herself.

A54. Now for the above types of manager, please mark the one which you would prefer to work under: (PDI)
1. Manager 1
2. Manager 2
3. Manager 3
4. Manager 4

A55. And, to which one of the above four types of managers would you say your own manager most closely corresponds? (PDI)
1. Manager 1
2. Manager 2
3. Manager 3
4. Manager 4
 287

A58. Considering everything, how would you rate your overall satisfaction in this company at the present time:
1. I am completely satisfied
2. Very satisfied
3. Satisfied
4. Neither satisfied nor dissatisfied
5. Dissatisfied
6. Very dissatisfied
7. I am completely dissatisfied

sometimes
frequently

frequently

seldom

seldom
very

very
How frequently, in your experience, do the following occur?

B46. Employees being afraid to express disagreement with their 1 2 3 4 5

managers (PDI)

undecided

disagree
strongly

strongly
Remember, we want your own opinion (even though it may be

agree

agree

agree
different from that of others in your country)

B60. Organisation rules should not be broken – even when the 1 2 3 4 5

employee thinks it is in the organisation’s best interests (UAI)
 288

Objectives of this market selection

As a result of selecting, then entering, an additional foreign market, you and your organisation are aiming to get benefits.
Examples are sales and profits, experience, and travel. Different markets will provide different benefits, so we need to find out
what your goals are, and how important each is to you. You could have multiple reasons for wanting to enter the additional
market.
Please read the table of possible goals, and the explanations of 2 of them, then give your response without tortured delay. You
will note that the first 4 items are positive matters and that the last 2 are inescapable negative matters.
(This section needs to be located after the Hofstede questionnaire, which will have sensitised respondent to psychic rewards)

Goal/s of entering the additional Importance:

foreign market: Very High High Moderate Low None
Obtain sales
Ongoing sales growth
Other business-related benefits *1
Take little risk
Apply little effort *2

*1 Other business-related benefits include: increasing your market share, increasing your corporate image, increased
organisational or political clout, disposing of excess capacity, disposing of products no longer attractive in other markets, deriving
learning (production, marketing, international business, etc.), reducing counter cyclicality, and obtaining resources or technology.
Some of these may be important to you, others not. Overall, how important are any or all of these to your organisation?

*2 Each market requires a different quantity and quality of personnel, finance, physical energy, mental energy and knowledge.
A different length of time may apply in different markets. How important is minimising effort by your organisation?

If you responded that ‘Other business-related benefits’ were of moderate importance or higher, please detail your requirements
so that special market selection techniques can be initiated: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.....................................................................................................
 289

Entry hazards

As well as the above upside potential from entering your additional foreign market, there is downside risk. You may not be
successful and so lose funds, the energy invested in market entry, company reputation, stakeholder support, etc.. Think briefly
about the human and financial capital you will probably invest over, say, the next two years. Compare this with the size of your
present business. Market entry may represent a small or a large proportion of the present total business. From a business point
of view, how damaging would it be to your organisation if you fail at market entry and lose most of that two year investment?

catastrophic very moderate minor negligible

considerable
Failure would
cause ..... damage:

On the other hand, many people enjoy taking an amount of risk. They gamble on horses, participate in adventurous sports, and
make business decisions which involve risk.
In general, Very Moderately Neither pleasurable Moderately Very
do you find risk- pleasurable pleasurable nor unpleasurable unpleasurable unpleasurable
taking to be …

Overall

Please now tell us the relative importance of each area to the other:

increase sales ......%

 290

ongoing sales growth ......

other business-related benefits ......
personal benefits …...
take little risk ......
require little effort ......
Total 100%

Management values, attitudes, motivation and commitment to further internationalisation

Instead of putting your energies into entering another foreign market, you could pursue increasing your domestic sales. You
could also increase your international business volume by expanding in any existing foreign markets you serve, adding products
to your existing range, or could implement tactics aimed at improving profit. Therefore, how keen are you to enter another
foreign market?
Strongly Agree No Disagree Strongly
agree feeling disagree
Our organisation does not presently have the ability to
penetrate an additional international market (F.I.V.)
There are insurmountable external obstacles to my
organisation selling internationally (F.I.V.)
Allocating scarce organisational resources (people,
funds, etc.) to locating and penetrating an additional
foreign market would be cost-effective for our
organisation (F.I.V.)
I think that our organisation will benefit if we locate and
add an additional international market (F.I.V.)
I think that I will benefit if we locate and add an
additional international market (F.I.V.)
Our organisation is not presently enthusiastic about
locating and penetrating an additional international
 291

market (F.I.V.)
I dislike doing international business (F.I.V.)
I dislike dealing with foreign people (F.I.V.)
My organisation tends to react to, rather than initiate,
new foreign opportunities (M&C)
0-20 % 21-40 % 41-60 % 61-80 % 81-100 %
Foreign sales should account for … % of total sales in
an organisation such as mine (F.I.V.)
Very Good Moderate Slight Very
good slight
Our organisation’s general knowledge about foreign
markets is … (F.I.V.)
Our organisation’s general knowledge about other
international business matters is … (F.I.V.)

Not sure Negative Neutral Low Moderate High Very

high
Please now turn back to ‘Objectives of this market
selection’ and re-read the goals you set for the
additional foreign market.
------------------------------
If you confidently expect to achieve those goals, and to
expend the level of effort you specified, how motivated
will you be to enter the new market? (M&C)
If you expect to achieve those goals, how committed
will you be to lasting the distance through adversity until
reaching break-even profitability in the new market?
(M&C)
 292

Very Valuable Moderately Low value Negligible

valuable valuable value
Our organisation believes international
market research to be …
Most organisations doing international
business find market research to be …

Please indicate which statement best reflects what your organisation does about selling on international markets. We (are) (M&C):
Unwilling to conduct international transactions
Fill unsolicited orders, but do not proactively seek foreign business
Experimentally do business in one or more foreign markets
Seriously proactively explore the feasibility of commencing more foreign sales
Have moderate or high experience at selling in numerous international markets
None of the above (please explain) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Please indicate which statement best reflects what your organisation should do about selling on international markets. We
should (be) (F.I.V.):
Unwilling to conduct international transactions
Fill unsolicited orders, but not proactively seek foreign business
Experimentally do business in one or more foreign markets
Seriously proactively explore the feasibility of commencing more foreign sales
Have moderate or high experience at selling in numerous international markets
None of the above (please explain) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Put aside any markets you already sell in. Name any additional foreign markets which you have seriously explored the
possibility of beginning to sell in during the last 2 years. (Include markets which you ceased selling in more than 2 years ago but
then attempted to re-enter during the last 2 years) (M&C)
 293

....................................................................................................

How many person years were devoted by the organisation’s international marketing staff to acquiring new foreign business
during the last 2 years? . . . . . . . . . . . . . . . . . . . . . . . (M&C)

How many person years were devoted by your organisation’s international marketing staff to maintenance of the existing foreign
business during the last 2 years? . . . . . . . . . . (M&C)

How many person years were available for all foreign marketing work (to seek new, and maintain existing, business) during the
last 2 years? . . . . . . . . . . . . . . . . . . . . . . . (M&C)

Customer needs for the product

Please indicate what you believe are the customer needs which cause, and affect the level of, sales of the product (e.g. weather,
population age, need for achievement, etc.): . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
...............................................................................................

Can you indicate the likely importance (weight) of each predictive variable: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.....................................................................................................

What are the descriptions of the preferred target segments?

(consumer products = geographic, demographic, psychographic and/or behavioural descriptions, or
industrial products = demographic features, operating variables, purchasing approaches, situational factors and/ or personal
characteristics)? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Adjustments to demand figures

We will obtain statistics on imports, exports and local production in each market. Adjustments need to be made to those for
permanent or temporary variations in supply and demand.
 294

Consider sources of any: (1) unsolicited foreign orders received (2) unsolicited foreign enquiries received for your product over
the last 2 years. Do these indicate permanent undersupply, or customer discontent with some aspect of the present suppliers, to
those markets? If these orders and enquiries indicate that potential clients are likely to be more easily attracted away from local
suppliers to you, we need to alter attractiveness of those countries:
Country: Country: Country: Country: Country
............................. ............................. ............................. ............................. .............................
+ ...... $US next year + ...... $US next year + ...... $US next year + ...... $US next year + ...... $US next year
+ ...... $US last year + ...... $US last year + ...... $US last year + ...... $US last year + ...... $US last year
+ ...... $US year before + ...... $US year before + ...... $US year before + ...... $US year before + ...... $US year before

Are there any other adjustments to be made for transient surges / slackening in demand (e.g. one-off spending, strikes delaying
production, stock build-up or run-down, price cuts moving purchases forward, etc.)
Country: Country: Country: Country: Country
............................. ............................. ............................. ............................. .............................
± ...... $US next year ± ...... $US next year ± ...... $US next year ± ...... $US next year ± ...... $US next year
± ...... $US last year ± ...... $US last year ± ...... $US last year ± ...... $US last year ± ...... $US last year
± ...... $US year before ± ...... $US year before ± ...... $US year before ± ..... $US year before ± ...... $US year before

Other environmental matters

Different markets have varying levels of demand, competition, etc., which cumulatively help or hinder your marketing mix.

Physical:
To what extent is demand for the product affected by the following physical attributes of the market environment?
Level: Affect on demand:
Climate: temperature high / medium / low increases / none / decreases
rainfall high / medium / low increases / none / decreases
snowfall high / medium / low increases / none / decreases
humidity high / medium / low increases / none / decreases
wind high / medium / low increases / none / decreases
 295

sunshine high / medium / low increases / none / decreases

Topography: hills high / medium / low increases / none / decreases
altitude high / medium / low increases / none / decreases
surface features (specify details) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
high / medium / low increases / none / decreases
Population intensity per squ. km (not to be confused with segment population size)
high / medium / low increases / none / decreases
What importance is availability of transport links to, or within, the market important to product demand?
To market: - air high / medium / low increases / none / decreases
- sea high / medium / low increases / none / decreases
- road high / medium / low increases / none / decreases
- rail high / medium / low increases / none / decreases
Within it: - air high / medium / low increases / none / decreases
- sea high / medium / low increases / none / decreases
- road high / medium / low increases / none / decreases
- rail high / medium / low increases / none / decreases

Technological:
Is it essential (not just desirable) for product use by the key customer segment for the market to have any particular technology
attributes or industries? For example, availability of electricity or other public utilities, telecommunications, computer ownership,
adult literacy, a specific industry such as bioengineering or aerospace, etc. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Cultural: What languages are spoken at home? . . . . . . . . . . . . . . . . . . . (principal) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (other)

What ethnic group do you originate from or identify with? . . . . . . . . . . . . . . . . . . .
Ecological/environmental:
Is the product one which has significant ecological implications for only some markets? (Especially Solid / Liquid / Gas / Noise
pollution affects, and deforestation implications.)
Are the implications positive or negative? Are they strong, moderate, weak or negligible?
Name the markets, indicate the level of affect, and whether positive or negative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 296

Legal:
Do you know of any international treaties between more than 3 countries which affect sales of your product? If so, please name
the treaties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Resources

(Question located at beginning of Questionnaire)

Organisation abilities
~ sustainable competitive advantage
Please assess your company’s absolute levels in the following areas of strategic advantage *:
* When the organisation already operates internationally, some parts of this question may have country-specific answers which
vary from the overall assessment of the company. When this occurs, please specify which countries and their different
assessment.

Very Strong Moderate Weak Very weak

strong
Innovation:
Technical product or service superiority
New product capability
Research and Development
Technologies
Patents
Early Mover advantage (say, 1st or 2nd into market)
Manufacturing of the good or service:
Cost structure
Flexible production operations
 297

Equipment
Access to raw materials
Vertical integration
Work-force attitude and motivation
Capacity
Finance – Access to capital:
From operations
From net short-term assets
Ability to use debt and equity financing
Owner’s or parent’s willingness to finance
If a Multinational Corporation:
Ability to conduct transfer pricing
Ability to shift assets to different countries
Diversity of assets
Management: included in later Competencies
Marketing:
Product quality reputation
Product characteristics / differentiation
Brand name recognition
Trade marks
Breadth of product line – systems capability
Customer orientation
Segmentation / focus
Distribution loyalty and control
Retailer relationship
Advertising / promotion skills
Sales force
 298

Customer service / product support

Customer base:
Size and loyalty
Market share
Growth of segments served
Synergy. Entering an additional market will:
Enhance customer value
Reduce operations costs
Reduce required investment

* Country-specific variations to the above:

Variable name: Country Very Strong Moderate Weak Very weak
strong

~ general business competencies

Very Considerable Moderate A little Negligible
considerable amount
Consider the internal and external staff who
do/will deal with this product. What amount * of
formal education and training do they have in:
management
marketing
marketing research
finance
planning
 299

(* ‘Amount’ of training encapsulates duration, intensity and range of learning i.e. it captures number of years and highest level
reached. It also averages the assessment for the functional area (e.g. all marketing staff).
‘Amount’ does not consider whether there is a sufficient number of people to do the volume of work, that being a Resources
matter.)

Very Considerable Moderate A little Negligible

considerable amount
Consider the internal and external staff who
do/will deal with this product. What amount of
practical experience do they have in:
Management
Marketing
Marketing research
Finance
Planning

How many years have you been marketing this product in the domestic market? . . . . . . . . . years

What is the domestic market share? . . . . . . . % of total industry sales of this product form

~ further internationalisation competencies

Very Considerable Moderate A little Negligible
considerable amount
 300

Consider the internal and external staff who

do/will deal with this product for foreign
markets. What amount of formal education and
training do they have in:
international management
international marketing
international marketing research
international finance
international market planning
export documentation and procedures

Very Considerable Moderate A little Negligible

considerable amount
Consider the internal and external staff who
do/will deal with this product for foreign
markets. What amount of practical experience
do they have in:
international management
international marketing
international marketing research
international finance
international market planning
export documentation and procedures

Excellent Good Moderate Poor

Consider the internal and external staff who do/will deal with
this product. How do you assess their past performance in:
 301

international management
international marketing
international marketing research
international finance
international market planning
export documentation and procedures

How many persons are in the dominant managerial coalition? . . . . . . .

How many of those were born abroad, or have continuously lived or worked abroad for at least 1 year during the last 15 years ?
........

Name the markets you have sold in during the last 3 years, indicating whether each was a new market entry, market
maintenance, market expansion or rationalisation (e.g. China – market entry then exit) . . . . . . . . . . . . . . . . . . . . . . . . . . .
................................................................................................

How many foreign markets do you currently:

- sell to by export? . . . . . . .
- have overseas sales offices in? . . . . . . .
- undertake assembly in? . . . . . . .
- undertake manufacture in (majority owned by your organisation)? . . . . . . .
- minority ownership joint venture with a foreign organisation (irrespective of distribution mode)? . . . . . . .

What foreign languages and abilities do your senior management and international marketing staff have (using the Australian
Foreign Service Language Proficiency Test scores of 0 [no] to 5 [native] *):
Language Speaking Reading Language Speaking Reading
proficiency * proficiency * proficiency * proficiency *
 302

* 0 = No proficiency, 1 = Elementary proficiency, 2 = Limited working proficiency, 3 = Minimum professional proficiency,

4 = Advanced professional proficiency, 5 = Native speaking professional proficiency

Turnover rate last year of international marketing staff . . . . . . . . %

Comparing your organisation with others in your industry, are you: Not yet international / An international early starter / A lonely
international / A late starter / or An international among others (circle one answer only)

Consider your key target customer segment.

(If you are already operating internationally, that should be your key international target customer segment. If you have
not yet begun regularly selling internationally and not determined the key foreign segment, it should be your key local
target customer segment.)
Highly Satisfied Mixed Dissatisfied Very
satisfied views dissatisfied
Overall, how satisfied is that segment with your
organisation’s performance at meeting their needs?

Is that view based on formal market research? Yes / No

What was the average number of days per person per year that your international business staff received in formal training in
international marketing, international finance, international management, international planning, and/or export
documentation and procedures, during the last 2 years? . . . . . . . days / international business person / year
During the last 2 years, what was the average number of weeks of formal, product-specific international market research
undertaken by an experienced researcher, divided by the number of international marketing staff? . . . . . . . weeks /
international marketing person / year
 303

If that research produced generally positive results for any of those markets, name them and indicate whether you have
conducted desk research, visited the market, or both . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
................................................................................................

Do you produce a written, annual, international marketing plan (or a written, annual, business plan with a significant international
marketing component)? Yes / No

At times your home market will boom, stretching production capacity and making a decision necessary about whether to service
it or the international market/s. At such times, do you expect to:
give preference to the home market / treat all markets equally / give preference to the international markets

When entering a new foreign market, do you expect to modify the product’s sales literature, packaging, operating instructions
and warranty message to suit the locally understood language:
all of these always / most of these usually / some of these sometimes / rarely / never

International Market Screening Competencies:

Do you conduct international market screening (as distinct from in-depth) research? Yes / No
If you do conduct market screening:
- during last 3 years, how many times have you conducted international market screening research (not in-depth market
research)? . . . . . . . . . . . .
- do you collect and use only secondary data on markets (as distinct from using primary data)? Yes / No
- are nearly all markets included (>180 countries)? Yes / No
- how many screening variables (e.g. population, GNP, etc.) are used? . . . . . . .
- do the variables include:
- product-specific variables? Yes / No
- organisation related variables? Yes / No
- are any of the variables weighted? Yes / No
- are the variables based on theory, research or experience? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 304

- do you screen assuming that you will enter the foreign market using only one particular mode of distribution (e.g. export,
owning your foreign facilities, etc.)? Yes / No
- please describe the decision system used to process the variables and produce the short-list of markets to be assessed
further: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
(Please specifically mention whether a mathematical method is used, and if so, what method /methods)
- Over the last 3 years, has the screening mostly been initiated for proactive or reactive reasons? Proactive / Reactive

Key stakeholder support and network contacts

As well as customers, who are the key: (1) internal and (2) external people or groups whose support is necessary for you to
successfully penetrate another foreign market? They may be located domestically or overseas.
Examples are: management, functional specialists, shareholders, financial institutions, politicians, bureaucrats, government
export agency, suppliers, customers, distributors, middle men, general public sentiment, press, industry associations, chambers
of commerce, sundry other foreign market network contacts. There may be others (please name them) . . . . . . . . . . . . . . . . . . .
......................................................................................................
......................................................................................................

Please now tick the above stakeholders and contacts who have considerable influence on your decisions – be they decisions
about international business or other matters which impinge on the firm.

Please now consider whether any of these stakeholders and contacts provide an advantage for entering a specific market. If so,
please asterisk them in the earlier question, then relate the asterisk to the market (e.g. *1 = family in N.Z., *2 = distributors in
Singapore, etc..) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Please now consider whether any of these stakeholders and contacts provide an impediment to entering a specific market. If so,
please mark them in the earlier question with a minus sign (-), then relate the asterisk to the market (e.g. - 1 = bank bias against
China, - 2 = director dislike of the U.K., etc..) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

How long will the market entry be likely to take to reach profitability? . . . . . years
 305

Circle the stakeholders who are reasonably likely to discontinue their support if the new market takes longer than this to reach
profitability
 306

APPENDIX B

MEASUREMENT OF VARIABLES

This Appendix details how each of the variables was quantified. For each
dimension, the components of the dimension are given (question, scale, feasible
response range, and either data source [if secondary data] or score used [if an
estimate to substitute for real primary data]). Included here are the scores used to
test the premise that organisations with different competencies, values, etc. should
be directed to different markets from each other. All scores end up with a 0 – 1
range, although many begin as 0 – 100, especially those requiring human
responses. See the Questionnaire in Appendix A for the primary data collection
mechanism. See Appendix F for a summary of all data values (primary and
secondary) in Appendix B.

Variable names (not the related explanatory text) are differentiated in

the following ways:

XXX = Quantified variable,

XXX = Potentially quantifiable variable
{ XXX } = Descriptive (non-quantitative) variable
( XXX ) = Some other irregularity
Q = The Question in the Management Questionnaire, Appendix A

Dependent variable

Organisation-specific short-list of attractive and preferred markets.

Independent variables

Organisation’s Objectives: Questionnaire responses assess both absolute

and relative importance of each organisational objective. These two
perspectives could be, but are not, combined by multiplying each by the
 307

other then rescaling 0-1. They are not combined because

organisations are expected to normally rate all objectives as high to
very high in importance on the absolute scale. However, when
organisations are forced to allocate 100 points across their set of
objectives in the relative scale, they are compelled to make choices
about how important each objective actually is. The absolute question
set is thus used to check the organisation’s response consistency.

- Obtain sales from the target segment/s:

Source: From Appendix A questionnaire item headed
‘Objectives of this Market Selection’
Scale: Very High / high / moderate / low / none (Absolute
importance)
Numeric (Relative importance)
Score: 100 / 80 / 50 / 30 / 0 (Absolute importance)
0 - 100 (Relative importance)
Range: 0 - 100 (Absolute importance)
0 - 100 * (Relative importance)
Score used: 55, 35 (Low-High, Relative importance)

- Obtain ongoing growth in sales to the target segment/s:

Source: From Appendix A questionnaire item headed
‘Objectives of this Market Selection’
Scale: Very High / high / moderate / low / none (Absolute
importance)
Numeric (Relative importance)
Score: 100 / 80 / 50 / 30 / 0 (Absolute importance)
0 - 100 (Relative importance)
Range: 0 - 100 (Absolute importance)
0 - 100 * (Relative importance)
Score used: 5, 35 (Low-High, Relative importance)

- Obtain personal benefits. Manager would rate the importance they

attach to personal benefits
 308

Score used: variable not activated in this research

- Obtain other business-related benefits:

. If activated, the analyst needs to add individually designed,
organisation-specific selection criteria to isolate specific
markets that yield the specified benefits (of decrease risk
by diversifying markets; increase market share; increase
corporate image; increase organisational and political
clout; dispose of excess capacity; dispose of products no
longer attractive in other markets; derive learning
[production, marketing, international business, etc.];
reduce counter cyclicality; and obtain resources or
technology).
Source: From Appendix A questionnaire item headed
‘Objectives of this Market Selection’
Scale: Very High / high / moderate / low / none (Absolute
importance)
Numeric (Relative importance)
Score: 100 / 80 / 50 / 30 / 0 (Absolute importance)
0 - 100 (Relative importance)
Range: 0 - 100 (Absolute importance)
0 - 100 * (Relative importance)
Score used: not activated in this research

- Risk level chosen:

Source: From Appendix A questionnaire item headed
‘Objectives of this Market Selection’
Scale: Very High / high / moderate / low (Absolute importance)
Numeric (Relative importance)
Score: 100 / 80 / 50 / 30 (Absolute importance)
0 - 100 (Relative importance)
Range: 0 - 100 (Absolute importance)
0 - 100 * (Relative importance)
Score used: 20, 15 (Low-High, Relative importance)
 309

(An alternative method of including risk would be to collect the

organisation’s assessment of the significance of loss, then
multiply each market’s risk score by significance of loss then
divide by 100. The questionnaire to management includes this
question but the data was not synthesised or processed.)

- Effort level desired:

Source: From Appendix A questionnaire item headed
‘Objectives of this Market Selection’
Scale: Very High / high / moderate / low (Absolute importance)
Numeric (Relative importance)
Score: 100 / 80 / 50 / 30 (Absolute importance)
0 - 100 (Relative importance)
Range: 0 - 100 (Absolute importance)
0 - 100 * (Relative importance)
Score used: 20, 15 (Low-High, Relative importance)

* all 6 items must add to 100

{Targets: Geographic, demographic, psychographic and/or behavioural

market selection criteria (consumer product), or
demographic features, operating variables, purchasing approaches,
situational factors and/or personal characteristics (industrial product)}

{Strategy decisions: for each market, whether the organisation will use:
.mass, segmented or niche marketing;
.product standardisation or adaptation;
.competing on cost, differentiation or focus;
.competition or collaboration. If collaboration, whether it will be direct
or indirect. If competition is chosen, whether the posture will be one of
defence, offence, flank or niche;
via pre-emption, confrontation or build up;
.country concentration or diversification;
.incremental or simultaneous market entries;
 310

.whether the organisation could, and wants to, become the market
leader, challenger, follower or nicher.
These are embedded in the segment, target and marketing mix
decisions, and subsequent funding of marketing in chosen markets}

(Size of assessment task:

Range: 0 (do not screen) or 1 (screen))
This dissertation elects to screen (naturally!)
If an intermediate value, or 0, is submitted, cease further processing of
the decision

(Criteria and method of evaluation: do not vary. See section 4.3.2.5 for
settings)

Intervening variables

Management motivation and commitment: (to add another market, not to

internationalise in general)
(Motivational cognition and managerial values about further
internationalisation and the Porter-Lawler model (1968) were
especially considered when constructing the following questions.
The details in section 4.4 about High – Low values, and types of
organisations covered, also apply to this section on Intervening
variables)
 311

(Cognitive:)
Q: Please now turn back to page … and re-read the goals you set for
the additional foreign market. If you confidently expect to
achieve those goals, and to expend the level of effort you
specified, how motivated will you be to enter the new market?
Scale: Not sure / negative / neutral / low / moderate / high / very
high
Score: 0 / 0 / 0 / 25 / 50 / 75 / 100 (cease processing and print
warning message if respondent scores 0 because it
indicates a critical level of perceived negative benefit from
further internationalising)
Range: 0 - 100
Low – High used: 50, 100

Q: If you expect to achieve those goals, how committed will you be to

lasting the distance through adversity until reaching break-even
profitability in the new market?
Scale: Not sure / negative / neutral / low / moderate / high / very
high
Score: 0 / 0 / 0 / 25 / 50 / 75 / 100 (cease processing and print
warning message if respondent scores 0 because it
indicates a critical level of perceived negative benefit from
further internationalising)
Range: 0 - 100
Low – High used: 50, 100

(Behaviour:)
(The next six questions assess the number and intensity of foreign
markets actively prospected in)

Q: Please indicate which statement best reflects what your organisation

does about selling on international markets. We (are):
- Unwilling to conduct international transactions
- Fill unsolicited orders, but not proactively seek foreign business
 312

- Experimentally do business in one or more foreign markets

- Seriously proactively explore the feasibility of commencing
more foreign sales
- Have moderate or high experience at selling in numerous
international markets
- None of the above (please explain) . . . . . . . . . . . . . . . . . . . . .
..............................
Scale: nominal (1 item to be agreed with)
Score: 0 / 10 / 50 / 75 / 100 / 0
Low – High used: 50, 100

Q: Put aside any markets you already sell in. Name any additional
foreign markets which you have seriously explored the
possibility of beginning to sell in during the last two years.
(Include markets in which you ceased selling in more than 2
years ago but then attempted to re-enter during the last 2 years)
..............
Scale: A – Z. Convert to a number of markets seriously
explored. Names for qualitative assistance to consultant
Score: For 0, 1 – 5, 6 – 20, ≥ 21 markets,
allocate 0, 100, 50, 30
(even large, internationally experienced, capable firms
have been found by numerous Trade Commissions
to be unable to successfully enter more than a few
new markets each year)
Range: 0- 261
Low – High used: 0, 100

Q: How many person years were devoted by organisation’s

international marketing staff to acquiring new foreign business
during the last 2 years? . . . . . . . . . . . . . .
Scale: numeric
Score: see sub-composite, below
 313

Range: 0 – 10,000
Low – High used: see sub-composite, below

Q: How many person years were devoted by organisation’s

international marketing staff to maintenance of existing foreign
business during the last 2 years? . . . . . . . . . . . . . . . . .
Scale: numeric
Score: see sub-composite, below
Range: 0 – 10,000
Low – High used: see sub-composite, below

Q: How many person years were available for all foreign marketing
work (to seek new, and maintain or rationalise existing,
business) during the last 2 years? . . . . . . . . . . . . . . .
Scale: numeric
Score: see sub-composite, below
Range: 0 – 10,000
Low – High used: see sub-composite, below

Sub-Composite for intensity of prospecting (above three questions):

Scales: various above
Scores: 1. Check that the sum of answers for time spent on:
(i) maintenance and (ii) new business, equals (iii) the
answer about total time available. Suspend
processing and print warning message if there is a
difference
2. Divide person years spent on prospecting for new
business by person years available (percentages
disadvantage organisations with an already large
foreign marketing activity)
3. For percentages of 0, 1 – 5, 6 – 25, 26 – 50, ≥ 50,
allocate 0, 50, 100, 50, 0
Range: 0 - 100
 314

Low – High used: 0, 100

Q: My organisation tends to react to, rather than initiate, new foreign

opportunities.
Scale: Strongly agree / agree / no feeling / disagree / strongly
disagree
Score: 0 / 30 / 50 / 80 / 100
Range: 0 - 100
Low – High used: 50, 100

Composite:
Scales: various above
Scores: sum above, divide by 6
Range: 0 - 100
Low - High used: (200 / 6 =) 33, (600 / 6 =) 100

{Marketing mix: Assume approximately same mix as in home or marker

market. Quantum spend normally larger than in existing markets
because of market size, strategic importance and having to break into a
new market. Funds assumed to be available to the organisation from
various financial sources}

Customer adoption and usage rates:

.The estimated percentage of users, by market, who are expected to be
influenced to buy from the organisation, and their re-buy rate. Would
require organisation to estimate figures for each market after
considering needs, segments, six stage hierarchy of effects, targets,
strategy, marketing mix and probable funding of marketing. Requires a
three (minimum) to five (preferable) year time frame.
Source: Not determined
Range: 0-100% (e.g. 2% year t for each DC market, 1 % for each LDC)
0-100% (e.g. 3% year t + 1 for each DC market, 1½ % for each LDC)
0-100% (e.g. 4% year t + 2 for each DC market, 2 % for each LDC)
Low – High used: Not assessed
 315

Organisation’s product-specific demand:

Conceptually calculated by taking adjusted product-specific demand in
each market then further adjusting for lagged customer adoptions and
usage rates. Not calculated in this research

(Screening selection process and data output: Uses PCA, CA and MDA to
screen markets
Score: produced from secondary and primary data
Range: Country names)

Moderating variables

Each market’s attributes:

- customers aspects:

~ (needs:
This is the basis for later defining the segments.
Assumes fundamental causes of and constraints on
foreign customer needs will be similar to those known
from experience in other markets. Needs may vary from
market to market so requiring different intrinsic adoption
rates by market and segment.
Scale: Insert regression equation if possible – uncommon
– or leave unquantified in descriptive form.
Source: From Appendix A questionnaire item headed
‘Customer needs ...’)

~ segment population sizes:

The number of users in each market
Range: 0 - 10 billion people/ organisations/ cars/ etc..
Units need to be specified (e.g. people, cars, etc.).
 316

~ {segments:
Product-specific. Assumed same as in home or a marker
market unless additional details are known. Usually
based on needs above, these segments are defined by:
(1) geographic, demographic, psychographic and/or
behavioural boundaries (if a consumer product), or
(2) demographic, operating variables, purchasing
approaches, situational factors and personal
characteristics (if an industrial product)}

~ intrinsic adoption rates:

.Product-specific
.Percentage calculated from: (1) number of users, by
market, who are expected to permanently adopt product
form over, say, 3 - 5 years, divided by (2) total population
of possible users. This is an industry-wide figure which is
estimated prior to the organisation applying its marketing
mix. It shows the changing level of product usage – it will
be constant in a saturated market. Estimate permanent
non-adopters here if needed for subsequent adjustment
of ‘Industry demand’. Each segment in each market may
have its own adoption rate. Requires secondary data,
alternatively, an industry expert’s estimate of figures after
considering the six stage hierarchy of effects. Number of
years will vary by product and market, so the 3 to 5 years
allowed for here may need alteration
.If a new product, replacement consumption would be
included here, with appropriate lag
Source: Not collected
Scale: %
Score: In practice, several years would be used because
penetration grows over several years, for example:
0-100% - year t for each market
0-100% - year t + 1 for each market
 317

0-100% - year t + 2 for each market.

However, detailed product-specific knowledge is
required to set these figures, so this dissertation
uses the simplified assumption that all growth
occurs in the first 12 months i.e.
Score = 1 for all 6 products in this research
Range: 0-100%

~ segment demand:
.Product-specific
.Insert segment demand if figure known.
Source: Coal, coking – personal e-mail communication
from the International Energy Agency’s Ms
Jocelyn Troussard 12 December, 1998
Beef – personal e-mail communication from the
Food & Agricultural Organisation’s (FAO’s)
Mr Orio Tampieri 9 November, 1998. Also
on the FAO’s www site
Wool – FAO www site
Telephones – personal e-mail communications
from the United Nation’s (UN’s) Mr Vasyl
Romanovsky 23 January, 1999
(production) and Mr Joseph Habr 4
February, 1999 (imports and exports)
Tertiary education – personal e-mail
communications from the United Nations
Education & Scientific Organisation’s
(UNESCO’s) Ms F. Tandart 17 November,
1998 and 10 December, 1998
I. C. engines – personal e-mail communications
from the UN’s Mr Vasyl Romanovsky 16
December, 1998 (production) and Mr
Joseph Habr 4 February, 1999 (imports
and exports). Also e-mail communication
 318

from Mr Brent Schulz, Power Systems

Research, 29 December, 1999
Scale: Coal, coking – millions of metric tonnes
Beef – millions of metric tonnes, boneless dressed
carcass weight
Wool – metric tonnes
Telephones – number of telephones
Tertiary education – full time student equivalents
I. C. engines – number of engines
Score: 1990 and 1995 data about each reporting
country’s imports, exports and local production
received from the above sources
Range: Coal, coking – 0-999 million
Beef – 0-99 million
Wool – 9 million
Telephones – 0-99 million
Tertiary education – 0-99 million
I. C. engines – 0-99 million

~ Industry demand:
Present industry demand:
.Product-specific
.If causes for industry sales (product form) are known,
insert regression equation to estimate demand by market.
That should eliminate some of the moderating variables in
this framework
.Alternatively, use product-specific data (e.g. SITC Rev 3
and ISIC Rev 2, to 5 digit level if possible) for imports,
exports and local production.
.Alternatively, use various other techniques to estimate
industry (or segment) demand (e.g. extrapolate from
another market’s segmentation data); complimentary or
substitute products’ demand; multifactor index; ordinary
 319

least squares projection; lead-lag analysis; or input-output

analysis
Range: 0 - 1,000 billion $US / units

~ Adjustments:
.Product-specific
.Specify whether to segment or industry
.Add a proportion of: (1) unsolicited foreign orders and
(2) enquiries received for latent and unmet demand.
Q: Consider sources of any: (1) unsolicited foreign orders
received (2) unsolicited foreign enquiries received for
your product over the last 2 years. Do these indicate
permanent undersupply, or customer discontent with
some aspect of the present suppliers, to those markets?
If these orders and enquiries indicate that potential clients
are likely to be more easily attracted away from local
suppliers to you, we need to alter attractiveness of those
countries:
Country: Country:
............................. .............................
...... $US next year ...... $US next year
...... $US last year ...... $US last year
...... $US year before ...... $US year before
Range: $US 0-1,000 billion (check and sum SITC and
ISIC imports, exports and production data to
ensure responses do not exceed maximum
ranges)
Temporary growth/decline in industry demand
(adjust ‘present industry demand’ for management
judgment about unenduring causes):
.Adjust for any transient major projects (capital
equipment) or temporary changes in consumption rates (if
consumable item), with appropriate lag
 320

.Is demand interrelated to, or independent from, demand

for other products (e.g. cars to tyres). Consider adjusting
the forecast demand accordingly, with appropriate lag
.Consider impact from substitute and competitive
products below (not here)
Q: Are there any other adjustments to be made for
transient surges / slackening in demand (e.g. one-off
spending, strikes delaying production, stock build-up or
run-down, price cuts moving purchases forward, etc.)
Country: Country:
............................. .............................
...... $US next year ...... $US next year
...... $US last year ...... $US last year
...... $US year before ...... $US year before
Range: 0 - 1,000 billion $US / units
No adjustments made to demand in this research

~ Composite of Production + Imports – Exports ± Adjustments:

Range: Coal, coking - 0-999 million
Beef - 0-99 million
Wool - 9 million
Telephones - 0-99 million
Tertiary education - 99 million
I. C. engines - 0-99 million

~ growth in demand:
Shift-share of local production + imports – exports
(± inventory if known).
Range: 0 ± 100 %

- Competition and substitutes:

(~ organisation and product-specific:

 321

The organisation’s SCA value substitutes for industry

competition)

~ industry-specific:
Source: Industry concentration ratios, or other, subject to
data availability
Score: not used in this research
Range: 0 - 1

~ economy-wide:
Source: World competitiveness scale of each market from
World Economic Forum (WEF), Geneva
Scale: -3 to +3 converted to 0-10055
Scores: 54 countries ex WEF 1998, with missing
countries scored using researcher’s judgment
Range: 0-100 55

55
Rescaling performed using I i j – I i j min x 100
-----------------
I i j max – I j min
where I = the screening variable, i = the country, & j = the specific indicator of the variable.
 322

- Other environmental matters:

~ physical:
.Product-specific
.From questionnaire to the organisation’s management
(Appendix A) headed ‘Other environmental matters’. If
this variable is activated, the analyst then needs to add
that parameter, implement a scale and a country scoring
system, and collect the data
Scale: high / medium / low = 3 / 2 / 1
increases / none / decreases = + / 0 / -
Score: 0 for all products in this research
Range: 0 - 1

~ economic:
Source: World Bank CD-ROM
Scale & Score:
1. Log GNP PPP per capita in international $, 1995;
2. Shift-share of GNP PPP per capita in international $,
1995 minus 1990;
3. Population, 1995;
4. Openness ([Imports + Exports] / GNP PPP), $US,
1995.
Range: 0 - 10 (log) (GNP PPP per capita)
0 - 100 % (Shift-share of GNP PPP per capita)
0 - 10 billion (Population)
0 - 100 % (Openness)

~ technological:
.Product-specific
.From questionnaire (Appendix A) item headed ‘Other
environmental matters’, insert any essential product-
specific variables here. If this variable is activated, the
 323

analyst then needs to add that parameter, implement a

scale and a country scoring system, and collect the data
Scale: Yes / No
Score: 0 for all products in this research
Range: 0 – 1

~ legal:
EIU’s ‘Arbitrariness/Fairness of Judicial System’ used
- Alternatively, could use whether the country a signatory
to the International Sale of Goods Act (Vienna) 1980 –
but disadvantage is that signing does not equal either
implementing or enforcing the Act.
- Product-specific constraint on sale or consumption of a
product could be indicated by the relevant conventions
signed by each nation. However, like the
abovementioned difference between signing,
implementing and enforcing, this is usually a poor
indicator.
Scale: Very low–Very high (1-5) converted to 0-100
Score: See attachment to 28/1/1999 e-mail from Laza
Kekic, EIU, London for values for 60 countries.
1993-7 average. Missing countries’ scores valued
using researcher’s judgment
Range: 1 - 5 converted to 0 – 100 55

~ political: Included in ‘Market risk levels’ below

Range: 0 – 1

~ cultural:
Scale: Hofstede (1980), standardised as distances from
Australia. See Personal values, below, for details
of matching organisation’s management with the
most compatible cultures
 324

Score: Hofstede’s, plus researcher’s estimates for

missing cultures
Range: -1 to +7 converted to 0 - 100 55

~ ecological:
.Product-specific
.Q: Is the product one which has significant ecological
implications for only some markets? (Especially Solid /
Liquid / Gas / Noise pollution affects, and deforestation
implications.) Are the implications positive or negative?
Are they strong, moderate, weak or negligible? Name the
markets, indicate the level of affect, and whether positive
or negative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
If so, the analyst needs to match to/away from markets
with environmental influences. This might be achieved by
proxies such as:
(1) are significant levels of pollution affecting their
citizens?
(2) are the citizens showing environmental concern
in attitudinal surveys? and
(3) has the country signed ...... international
convention? Check UNEP Register of
International Treaties re the Environment, then
assess differing enforcement or implementation in
each market.
Scale: Strong / Moderate / Weak / None / Weak /
Moderate / Strong
-50 / -30 / -15 / 0 / +15 / +30 / +50
Score: 0 used for all products in this research
Range: -50 to +50 converted to 0 - 100 55
 325

- Market risk levels:

Risks: forecast economic, political and financial risks in each
country, comprising 24 scaled items amalgamated into a
composite score, produced by ICRG Composite Index
from Political Risk Services, East Syracuse, New York.
Scale: see Coplin & O’Leary (1994, p. 248)
Score: actual, by market, as supplied by ICRG’s Mr
Adrian Shute in 28 January 1999 faxed personal
communication
Range: 0 - 100 (highest - lowest risk)
Diversification: offsetting effect from diversification. Could
reduce risk for each of the first 10 markets per Brealey &
Myers (1984, p. 126)
Score: disregarded in framework as amount small and
hard to quantify

- New and existing markets: Ensure existing markets are included, and
that data are comparable with data on the other, less known,
new possible markets.

Organisation’s capabilities and preferences:

- Organisation’s abilities:

~ sustainable competitive advantage (SCA):

(1) Generic SCA:
See Questionnaire to management (Appendix A)
for 38 items from Aaker (1995, p. 80), underpinned
by first of the three items from Dunning (1980, pp.
9 and three items from p 10), minus the
competence items, plus early mover advantage
from Pan, Li & Tse (1999, p. 82).
Scale: very strong / strong / moderate / weak /
very weak
 326

Scores: 100 / 80 / 50 / 30 / 0
Range: 0 - 100
Low – High used: 30, 100
(2) Market-specific variations to the above (included in
‘Organisation’s Market-specific Internationalisation
Competencies’ section). From Questionnaire to
Management.
SCA sub-variable name and Country:
........................................
........................................
Scale: very strong / strong / moderate / weak / very
weak
Scores: 100 / 80 / 50 / 30 / 0
Range: 0 - 100
Low – High used: 0, 100 for 30% of the product’s
markets, selected randomly56

~ general business competencies:

(The next four questions are indicators of accumulated
learning about the domestic market)

Q: Amount of formal education and training in

management, marketing, marketing research,
finance and planning by internal and external staff
who will deal with the product.
(‘Amount’ of training encapsulates duration,
intensity and range of learning i.e. it captures
number of years and highest level reached. It also
averages the assessment for the functional area
(e.g. all management staff). It does not consider

56
Not having real data to use to distinguish between firms of differing capability required
synthesising the situation for data analysis purposes by generating plausible, rational rules.
The low capability firm was assumed to have no market-specific SCA, so scored 0 in all
markets for this variable. The high capability firm was assumed to have SCA advantages
 327

whether there is a sufficient number of people to

do the volume of work, that being a Resources
matter.)
Scale: very considerable / considerable / moderate /
a little / negligible
Scores: 100 / 80 / 50 / 30 / 0
Range: 0 - 100
Low – High used: 50, 100

Q: Amount of practical experience at management,

marketing, marketing research, finance and
planning by internal and external staff who will deal
with the product.
Scale: very considerable / considerable / moderate /
a little / negligible
Scores: 100 / 80 / 50 / 30 / 0
Range: 0 - 100
Low – High used: 30, 100

Q: How many years has the organisation been marketing

this product in the domestic market? . . . . . . . . .
years
Scale: 0, 1, 2, 3 or more
Scores: 0, 40, 75, 100
Range: 0 - 100
Low – High used: 40, 100

Q: What is the domestic market share? . . . . . . . . . % of

total industry sales of this product form
Scale: % market share
Scores: 0 – 5, 5 - 10, 11 – 20, 21 – 30, 31 – 40,
>40, receive: 0, 20, 40, 60, 80, 100

from land, labour, capital or multinationality in an arbitrarily selected number of 30 per cent of
 328

Range: 0 - 100
Low – High used: 20, 100

Composite: (Could use log to reflect diminishing returns)

Scale: 0 - 100
Scores: Blended
Range: 0 - 100
Low – High used: ([140 to 400] ÷ 4 =) 35, 100

~ further internationalisation competencies:

These comprise: (1) generic (non market-specific) and

(2) market-specific further internationalisation
competencies. The organisation’s quantum in
each is determined by accumulating points from
the following indicators, which frequently relate
conceptually to Sveiby’s four drivers of
competence (see Figure 4.6).
“ (S) “
indicates a competence measure that is
specifically mentioned by Sveiby but which has
been modified from a domestic to a foreign
competence situation. Four sub-items are
expected to play a role in both generic and market-
specific competence (SCA, language proficiency,
market research proficiency and network contacts).

Q: Amount of formal education and training in

international management, international marketing,
international marketing research, international
finance, international planning, export
“ (S) “
documentation and procedures

markets, those markets being selected by random numbers then given a score of 100.
 329

Scale: very considerable / considerable / moderate /

a little / negligible
Scores: 100 / 80 / 50 / 30 / 0
Range: 0 - 100
Low – High used: 0, 100
(‘Amount’ of training encapsulates duration,
intensity and range of learning i.e. it captures
number of years and highest level reached. It also
averages the assessment for the functional area
(e.g. all marketing staff). ‘Amount’ does not
consider whether there is a sufficient number of
people to do the volume of work, that being a
Resources matter)

Q: Amount of practical experience at international

management, international marketing, international
marketing research, international finance,
international planning, export documentation and
“ (S) “
procedures
Scale: very considerable / considerable / moderate /
a little / negligible
Scores: 100 / 80 / 50 / 30 / 0
Range: 0 - 100
Low – High used: 0, 100

Q: Assessed performance at international management,

international marketing, international marketing
research, international finance, international
planning, export documentation and procedures
Scale: excellent / good / moderate / poor
Scores: 100, 70, 35, 0
Range: 0 - 100
Low – High used: 0, 100
 330

Q: Number of persons in the dominant managerial

coalition . . . . . . .

Q: How many of those were born abroad, or have

continuously lived or worked abroad for at least 1
year during the last 15 years? . . . . . . . .
Scale: ratio
Score: Number born, lived or worked abroad ÷
number in dominant managerial coalition
Range: 0 - 100
Low – Hi used: 0, 100

Q: Name the markets sold in during the last 3 years,

indicating whether each was a new market entry,
maintenance, expansion or rationalisation (e.g.
China - market entry then exit) . . . . . . . . . . . . . . .
........................................
...............
Scale: A - Z (ensure that all existing home and
foreign markets are included)
converted to:
0 - 261, limit 10057
Range: 0 - 100
Score: 0 - 100 (Sum the number of markets
reported)
Low – High used: 0, 100
(This question quantifies the number of foreign
markets actively serviced. It could be factored up

57
Whilst there are 230 feasible markets, the convenient maximum number of 100 was set.
This is because diminishing returns are expected to apply to the accumulation of the market-
sourced experiential learning, so that any firm which penetrates 100 markets should have
accumulated as much general international learning as is effectively useful. Included in the
100 is the firm’s existing markets to allow comparison against new markets, and for the
possible expansion to be in an existing market instead of entry into a new one. It could be
argued that the scale should be logarithmic as there would be high front-end loading to the
value of learning, although this sophistication was not activated in this exploratory research.
 331

by market share gained and/or whether the sales

were obtained by proactive or reactive means.)

Q: How many foreign markets are currently serviced by

each of the following:
- export? . . . . .
- overseas sales offices? . . . . .
- foreign assembly? . . . . .
- overseas manufacture (majority owned by your
organisation)? . . . . .
- minority ownership joint venture with a foreign
organisation (irrespective of distribution mode)?. . .
...
Scale: above
Scores: 10 for 1st market in each of the 5 options,
16 if organisation has 2 markets in each,
20 if organisation has 3 markets in each,
* maximum of 20 points per mode, and a
combined maximum of 100 points
Range: 0 - 100
Low – High used: 0, 100
(The above question determines the range and
depth of experience with different modes of entry
and with different foreign ownership relationships)

Q: Turnover rate last year of international marketing staff

“ (S) “
........%
Scale: 0 – 5%, 6 – 10, 11 – 15, 16 – 20, 21 – 25,
26 +
Scores: 100, 80, 60, 40, 20, 0
Range: 0 - 100
Low – High used: 60, 100
(This relates to Sveiby’s Stability component of
Competence)
 332

Q: Comparing the organisation with others in the industry,

are they: Not yet international / An international
early starter / A lonely international / A late starter /
or An international among others (Circle one only)
Scale: above
Scores: 0 / 50 / 75 / 25 / 100
Range: 0 - 100
Low – High used: 0, 100
(The above question assesses the organisation’s
experience relative to international competition)

Q: Consider the key target customer segment. (If the

organisation is already operating internationally,
use their key international target customer
segment. If organisation has not yet begun
regularly selling internationally and not determined
the key foreign segment, it should be the key local
target customer segment – unless organisation
indicates otherwise.) Overall, how satisfied is that
segment with the organisation’s performance at
meeting their needs? “ (S) “
Scale: Highly satisfied / Satisfied / Mixed views /
Dissatisfied / Very dissatisfied
Scores: 100 / 75 / 50 / 0 / 0
Range: 0 - 100
Low – High used: 50, 100
 333

Q: Is that view based on formal market research?

Scale: Yes / No
Scores: If Yes, double score from previous
question i.e. add 100 / 75 / 50 / 0 / 0
Range: 0 - 100
Low – High used: 0, 100

(The above two questions relate to Sveiby’s Effectiveness

component of Competence

Q: Number of days per person of formal training of

international business staff in international
marketing, international finance, international
management, international planning, and/or export
documentation and procedures undertaken by
organisation staff during the last 2 years “ (S) “. . . . .
.....
Scale: 0 days, 1-2, 3-6, 7-9, 10+
Scores: 0, 25, 50, 75, 100
Range: 0 - 100
Low – High used: 25, 100

Q: Average number of weeks of formal, product-specific

international market research undertaken by an
experienced researcher, divided by the number of
international marketing staff, during the last 2
years
Scale: 0, 0.1 – 1, 1.1 – 2, 2.1 – 3, >3 weeks
Scores: 0, 25, 50, 75, 100
Range: 0 - 100
Low – High used: 50, 100

(The above two questions relate to Sveiby’s Growth

component of Competence
 334

(The following are additional lead indicators of

competence)

Q: If that research indicated generally positive results for

any of those (formally researched) markets, name
them and indicate whether they had desk
research, a market visit, or both, conducted . . . . .
........................................
(This research indicates an aspect of the
organisation’s accumulated general international
wisdom. It also is market-specific, so these
markets should also receive a positive weighting)
Scale: Desk only, Visit only, or Both
Scores:
(1) Generic international research competencies:
No of new Desk Visit Both59
markets only only
researched58
1st 10 10 30
2nd 10 10 30
3rd 8 8 20
4th 5 5 10
5th 5 5 10
(2) Market-specific research competencies:

58
Many Trade Commissions, including the Australian, British, and the American equivalent,
discourage firms of all sizes and competences from attempting market entry into more than 3
markets at a time, advocating that only one market be entered at a time. See, for example,
BETRO Trust Committee (1976). Experience has repeatedly shown them that firms are
rarely able to successfully digest more. It can therefore be argued that this sub-variable
should give credit to no more than 3 markets because in-depth researching of more is
wasteful. However, it can also be argued that screening needs to generate surplus markets
to allow for in-depth research to subsequently discard several markets as unsuitable. The
number was therefore increased to an arbitrarily determined 5 markets in which the high
ability firm was given credit for previous research undertaken. Whilst in reality it is likely that
these 5 would be amongst the largest markets by product consumption, selecting them in that
way would have caused multicollinearity, so random selection was used.
59
Additional points given when both desk and visiting market research are conducted on the
same market because synergy occurs.
 335

Desk Visit Both 59

only only
35 35 100
(Each market researched gets these scores,
but no more than 5 new markets to receive
preferential weighting). Enter this score below
in ‘Market-specific Internationalisation
Competencies’
Range: 0 – 100 (both)
Low – High used: 40, 100 (Organisation’s generic
international competencies)
0, 100 in 5 markets
(Organisation’s market-
specific competencies) 58

Q: Does the organisation produce a written, annual

international marketing plan (or a written, annual
business plan with a significant international
marketing component)?
Scale: Yes / No
Scores: 100 / 0
Range: 0 - 100
Low – High used: 0, 100

Q: At times your home market will boom, stretching

production capacity and making a decision
necessary about whether to service it or the
international market/s. At such times, do you
expect to:
Scale: give preference to the home market / treat
all markets equally / give preference to the
international markets
Score: 0 / 80 / 100
 336

Range: 0 - 100
Low – High used: 0, 100

Q: When entering a new foreign market, do you expect to

modify the product’s sales literature, packaging,
operating instructions and warranty message to
suit the locally understood language:
Scale: all of these always / most of these usually /
some of these sometimes / rarely / never
Score: 100 / 80 / 50 / 20 / 0
Range: 0 - 100
Low – High used: 0, 100

Q: Foreign language proficiencies (per Australian Foreign

Service Language Proficiency test) of
organisation’s managerial and international
marketing staff:
Speaking Scale: 0 / 1 / 2 / 3 / 4 / 5 for
No proficiency / Elementary proficiency /
Limited working proficiency / Minimum
professional proficiency / Advanced
professional proficiency / Native speaking
professional, respectively, in each
language group.
Reading Scale: same again (0 / 1 / 2 / 3 / 4 / 5)
Combine speaking and reading scores, providing a
score of up to 5 speaking plus 5 reading,
for each foreign language spoken and
written (i.e. 10 points maximum per
language).
Scales: above
Scores: 0 – 10 points x up to 10 languages
maximum (Organisation’s generic
international competencies)
 337

0 – 100 for each language, maximum 100

points per country. (Take speaking
+ reading score, multiply by 10,
then allocate to each country in
which the organisation has a
language proficiency). Enter any
scores for this item in
‘Organisation’s market-specific
internationalisation competencies’
section
Range 0 – 100 (both)
Low – High used: 0, 100 (Organisation’s generic
international
competencies), &
See Appendix F (Organisation’s
market-specific
international
competencies)60
(The above question determines the range and
depth of competence to communicate with different
cultures)
(Markets presently serviced, in which the
organisation has language competence, increase
their general internationalisation competence
assessment. Markets which are not presently
serviced, but in which there is language
competence, should receive a positive weighting

60
Not having a real firm required synthesising the situation for data analysis purposes. The
‘Firm’s market-specific language proficiency variable’ therefore assumed that:
(1) the low ability firm’s employees could speak and read only English, and
(2) the high ability firm had employees with level 5 (native) reading and writing language
proficiency in the 10 major language groups (an arbitrarily selected number which accounts
for 3 billion of the world’s population), but could earn a score of no more than 1.0 in any one
market nor more than 1.0 in more than 10 markets.
See Appendix G for language scoring details by country and assumed level of firm ability.
 338

because of organisational competence to harvest

them)

International market research competence61:

General market research competence:
(The following four indicators are already
included above in Further
international competencies)
Q: Amount of formal education and training
in international market research,
Q: Amount of practical experience in
international market research,
Q: Past performance at international
marketing research, and
Q: Number of person-weeks of marketing
research undertaken pa.
Screening:
Item: Score:
Organisation conducts screening Mandatory *
(as distinct from in-depth) research
Data:
- organisation collects & uses only Mandatory *
secondary data on markets (as
distinct from using primary data)
- all markets are included (>180 Mandatory *
countries)
- number of variables 1 - 5 = 5,
> 5 = 10
Data includes:
- product-specific variables 20

61
This is included as 1 of 18 sub-variables in Generic further international competencies.
However, this ability would seem likely to be a critical further internationalisation skill, so the
score for this sub-item should also be isolated and fed back to the firm to allow it to assess its
need to improve.
 339

- organisation related variables 10

- variables are weighted 10
- variables used have theory, 10
research or experience basis
Method & variables are mode 10
neutral
Decision system:
- is unsystematic 0
- has a quantitative component Mandatory *
(may quantify qualitative items)
- produces a short-list of markets Mandatory *
Proactive initiation of screening (as 10
distinct from reactive)
Amount of experience at market
screening research (not of in-depth
research) during last 3 years:
- 1 previous market screening 5, or
- 2 or more previous screenings 10
Scales: above
Scores: Σ of the above ÷ by 9 multiplied by
100
Range: 0 - 100
Low – High used: 0, 100
* Organisation earns zero for the screening
component of this sub-variable if any
of the following apply: (1) it only
conducts in-depth market research,
(2) it does not consider all markets,
(3) it attempts to collect primary data
on markets when screening, (4) it
does not use a quantitative process to
select the short-list, or (5) it does not
produce a short-list of markets.
 340

Competence at market screening should be

reassessed, perhaps replacing these
indicators, once research confirms a
suitable screening methodology.

Composite non market-specific further internationalisation

competencies:
Scales: Various above
Scores: Sum of 18 scales = (225 ÷ 18, 1800 ÷ 18)
= 13, 100
Range (composite): 0 - 100
Low – High used: 13, 100

- Market-specific further internationalisation competencies

Market-specific Sustainable Competitive Advantage

From question asking about market-specific SCA’s
in the Questionnaire to the organisation’s
management headed ‘Further International
Competencies’

Market-specific Language Competencies

From question asking: ‘What foreign language
abilities do your …’ in the Questionnaire to the
organisation’s management headed ‘Further
International Competencies’

Market Research
From question asking ‘Name those (formally
researched) markets …‘ in the Questionnaire to
the organisation’s management headed ‘Further
International Competencies’

Stakeholder and Network contacts:

 341

Q: Please now consider whether any of these

stakeholders and contacts provide an advantage
for entering a specific market. If so, please
asterisk them, then relate the asterisk to the
market (e.g. *1 = family in N.Z., *2 = distributors in
Singapore, etc.) . . . . . . . . . . . . . . . . . . . . . . . . . . .
........................................
(From Appendix A section headed ‘Key
stakeholder support’. Last of the 3 stakeholder
and network contacts)

Composite market-specific further internationalisation

competencies
Scales: four above
Scores: Σ of the four scores for each market,
each ÷ 4
Range (composite): 0 - 100
Low – High used: 16, 100 (= 65 ÷ 4, 400 ÷ 4)

~ key stakeholder support:

.2 of 3 Questions: From Appendix A section headed ‘Key
stakeholder support’. Do not use second last item
about market-specific contacts here – it appears in
‘‘Market-specific Internationalisation
Competencies’’.
Scale: A - Z
Scores: (1) Key stakeholder support:
(No of influential stakeholders minus
stakeholders likely to discontinue
support) divided by No of influential
stakeholders X 100
 342

(2) Market-specific competencies:

Add or subtract 10 points to each
country for each stakeholder nominated
by the organisation as having a positive
or negative influence in that country.
Maximum ±100 points influence per
country.
Range: 0 – 100 (both)
Low – High used: 30, 100 (Key stakeholder
support)62
0, 100 in 30% of markets63
(Market–specific
competence from
Contacts)

~ other resources: A go/no go assessment.

(Precede this question by showing a plot of the typical

costs, sales and profits cycle e.g. Kotler 1994,
p. 308)

Q: Foreign market entry, to break-even profit stage,

typically takes most organisations more than 3
years. Do you expect your organisation to be
different?
Scale: Yes / No / Don’t know
Scores: Not scored
Low – High used: Not scored

62
For data analysis purposes, no specific markets were given preference in this Generic
internationalisation competencies component. However, a modest minimum level of
stakeholder support (e.g. from bankers, shareholders, management, etc.) for further
internationalisation was assumed necessary for the several years that success normally
requires.
63
For data analysis purposes, 30 per cent of markets were selected at random for the high
ability firm to be given a 100 point preference in. The low ability firm was assumed to have no
contacts in any markets, so scored 0 in all markets for this variable.
 343

Range: Not scored

(This question warns the analyst that the
organisation may have unrealistic expectations
about the speed with which it will reach profitability.
Organisation will waste resources if it does not
have the wherewithal to last until reaching
profitable market entry)

Q: Briefly consider what financial, physical, personnel and

intangible resources will probably be needed to
supply the foreign market/s. Also consider how a
successful market entry will require management
to make unusual mental efforts and be diverted
from their normal duties. Do you feel reasonably
confident that you will be able to: (1) marshal
whatever resources are necessary and (2) sustain
them for the several years it will take to reach
break-even profitability?
Scale: Yes / No / Don’t know
Scores: 1, 0, 0
Range: 0 - 1
Low – High used: 1 (only permitted answer)
(If organisation does not answer “Yes”, market
selection should cease because the organisation
may have insufficient resources to profit optimise.
Worse, an already resource short organisation
risks resource wastage through market entry being
followed by premature market withdrawal. This
question should be located at the beginning of the
Questionnaire)

- Managerial values:
 344

‘Low – High used’ are two values used in the dissertation’s

computations to assess whether differences among organisations’
values have an impact on which markets should be selected by
screening. Appendix F summarises these data.

~ further internationalisation values:

(Cognitive:)
Q: Our organisation does not presently have the ability to
penetrate an additional international market
Scale: Strongly agree / agree / no feeling /
disagree / strongly disagree
Score: 0 / 0 / 25 / 75 / 100
Low – High used: 75, 100

Q: There are insurmountable external obstacles to my

organisation selling internationally
Scale: Strongly agree / agree / no feeling /
disagree / strongly disagree
Score: 0 / 0 / 25 / 75 / 100
Low – High used: 25, 100

Q: Our organisation’s general knowledge about foreign

markets is …
Scale: Very good / good / moderate / slight / very
slight
Score: 100 / 80 / 50 / 25 / 0
Low – High used: 25, 100

Q: Our organisation’s general knowledge about other

international business matters is …
Scale: Very good / good / moderate / slight / very
slight
Score: 100 / 80 / 50 / 25 / 0
 345

Low – High used: 50, 100

Q: Allocating scarce organisational resources (people,

funds, etc.) to locating and penetrating an
additional foreign market would be cost-effective
for our organisation
Scale: Strongly agree / agree / no feeling /
disagree / strongly disagree
Score: 100 / 75 / 25 / 0 / 0
Low – High used: 75, 100

Q: I think that our organisation will benefit if we locate and

add an additional international market
Scale: Strongly agree / agree / no feeling /
disagree / strongly disagree
Score: 100 / 75 / 25 / 0 / 0
Low – High used: 75, 100

Q: I think that I will benefit if we locate and add an

additional international market
Scale: Strongly agree / agree / no feeling /
disagree / strongly disagree
Score: 100 / 100 / 75 / 50 / 0
Low – High used: 50, 100

Q: Foreign sales should account for … % of total sales in

an organisation such as mine
Scales : 0-20 / 21-40 / 41-60 / 61-80 / 81-100
Scores *: 0 / 50 / 100 / 100 / 100
Low – High used *: 50, 100
* based on Australia, which typically comprises 2%
of the world market for a product
 346

Q: Please indicate which statement best reflects what

your organisation should do about selling on
international markets. We should (be):
- Unwilling to conduct international transactions
- Fill unsolicited orders, but not proactively seek
foreign business
- Experimentally do business in one or more
foreign markets
- Seriously proactively explore the feasibility of
commencing more foreign sales
- Have moderate or high experience at selling in
numerous international markets
- None of the above (please explain) . . . . . . . . . . .
........................................
Scale: nominal (1 item to be agreed with)
Score: 0 / 10 / 50 / 75 / 100 / 0
Low – High used: 50, 100

(Affective:)
(The next three questions test whether management would
be adding a market unwillingly - e.g. perceived
unwanted dependence on international markets,
pressure from other stakeholders, relative
preference for domestic sales, etc..)

Q: Our organisation is not presently enthusiastic about

locating and penetrating an additional international
market
Scale: Strongly agree / agree / no feeling /
disagree / strongly disagree
Score: 0 / 0 / 25 / 75 / 100
Low – High used: 75, 100

Q: I dislike doing international business

 347

Scale: Strongly agree / agree / no feeling /

disagree / strongly disagree
Score: 0 / 25 / 50 / 75 / 100
Low – High used: 50, 100

Q: I dislike dealing with foreign people

Scale: Strongly agree / agree / no feeling /
disagree / strongly disagree
Score: 0 / 25 / 50 / 75 / 100
Low – High used: 50, 100
(This question could be replaced by the 17
question CETSCALE of ethnocentrism Bruner &
Hensel [1996]. A highly ethnocentric attitude can
impede education, training and experience re
international learning, so reducing competence
and motivation – see Gould & McGillivray 1998)

Composite:
Scales: various above
Scores: sum above, divide by 12
Low – High used: (650/12 =) 54, (1,200/12 =) 100
 348

~ {strategic values re:

.product standardisation or adaptation (framework
subordinates this to Customer needs and segment
population size → segments → targets);
.competing on cost, differentiation or focus;
.whether competitive posture will be one of defence,
offence, flank or niche
via pre-emption, confrontation or build up;
.country concentration or diversification (framework
assumes concentration by picking single ‘best’ market),
.incremental or simultaneous market entries (framework
assumes incremental market)
.whether the organisation could become the market
leader, challenger, follower or nicher.
.Scale: very strong / strong / medium / weak / very weak
preference}

~ {personal values: not operationalised in this research, except

for the three sub-components cultural difference, risk
preferences and market research values, below.
(When more country data becomes available, use of
Schwartz’s (1992) Value System would allow assessment
of personal needs and allow determination of cultural
distance. S.V.S. seven higher order dimensions are:
Affective Autonomy, Intellectual Autonomy,
Embeddedness, Hierarchy, Egalitarianism, Mastery and
Harmony.
Scale: See 57 item, 9 point scale, questionnaire of
Schwartz (1992, pp. 60-62) and slight amendments
(advised in personal e-mail communication from Schwartz
dated 1 October, 1998).}
 349

- cultural values: the base for calculation of cultural

distance from each market.
Scale: Hofstede (1980). Could use Schwartz
(1992)
Scores: Average Australian Hofstede
Low – High used: Average, not High-Low

- risk preferences: from Questionnaire of organisation’s

management, “Entry hazards” – 2 questions.
Q: Business: Failure (details in preamble) would
cause ..... damage: catastrophic, very
considerable, moderate, minor, negligible
Q: Personal: Do you find risk-taking to be …, very
unpleasurable / moderately unpleasurable /
neither pleasurable nor unpleasurable /
moderately pleasurable / very pleasurable
Alternatively, could use Hofstede (1980) item
‘uncertainty avoidance’ (UAI) comprising averaged
answers to Q’s B60, A43 & A37). UAI = 300 – 30
(B60) – (% staying < 5 years, A43) – 40 (A37).
Scale: Hofstede 1980. Scores: 51 (Australian
average). Range: 0 – 100. Low – High used: 0,
101. Alternatively could use Schwartz’s ‘security’
dimension, questions 8, 13, 15, 22 and 56.
However, these do not seem to directly measure
risk preference, either personally and in
international business.
Match these Hofstede values, PRS’ market risk
figure, and risk preferences expressed in
objectives. Print out at end of analysis run to
consider ambiguities
Scales: see above
Scores: 0, 30, 80 *, 90, 100 ** (Business)
60 *, 70, 100 **, 70, 60 * (Personal)
 350

Range: 0 – 100
Low – High used (combined business and
personal):
(70 * + 30 *)/2 = 50 Low
(100 ** + 100 **)/2 = 100 High

- market research values:

Questions re management valuation of market

research included in Questionnaire located
in ’Management attitudes, motivation and
commitment to further internationalisation’:

Q: Our organisation believes international market

research to be …
Scale: Very valuable / valuable / moderately
valuable/ low value / negligible value
Scores: 100 / 80 / 50 / 30 / 0
Range: 0 – 100
Low – High used: 50, 100

Q: Most organisations doing international business

find market research to be …
Scale: Very valuable / valuable / moderately
valuable/ low value / negligible value
Scores: 100 / 80 / 50 / 30 / 0
Range: 0 – 100
Low – High used: 50, 100

Composite
Scale: as above
Scores: as above
Range: 0 – 100
Low – High used: 50, 100
 351

Data availability and quality:

Markets:

How much of the input data diverge from the target years by
more than ± 2 years? (timeliness)
Scale: 0 / ≤ 5% / ≤ 10% / ≤ 15% / ≤ 20% / > 20%
Score: 100, 95, 90, 85, 80, 75 for each of the 21 input
data cells for each market (excludes the
subsequent cells used to process and aggregate
the input data)
Range: 0 - 100
Value used: varies by product and market

Using a 0 - 100 scale, rate the accuracy of each input data

source used (accuracy)
(100 assigned to Economist Intelligence Unit, PRS Inc,
UNDP, WEF, WTO and World Bank;
90 assigned to CIA, FAO, IEA, UN and UNESCO:
80 assigned to researcher (RG-MMcG) estimates
of missing country Competitiveness, Culture and
Market Risk, and researcher (RG) estimates of
missing country GNP/capita)
Scale: Yes / No
Score: 100, 0
Range: 0 - 100
Value used: varies by product and market

Using a 0 - 100 scale, rate the comparability of each input data

source used (comparability)
(100 assigned to all, except UN ISIC production data
rated at 90)
 352

Scale: Yes / No
Score: 100, 0
Range: 0 - 100
Value used: varies by product and market

Per cent of data cells present (completeness)

Scale: 0 - 100 (low - high data presence)
Score: computer calculates the number of the 21 input
data cells about each market are filled, converts to
a percentage, then reports
Range: 0 - 100
Value used: varies by product and market

If data are missing, do you believe the missing items are likely to
materially affect the outcome of the screening
recommendation? (significance)
Scale: None / a little / a moderate amount / much
Score: 100, 90, 75, 50
Range: 0 - 100
Value used: varies by product and market

Composite for each market:

Σ of the 5 quality scores for each market ÷ 5
Range: 0 - 100
Value used: varies by product and market

Organisation:

Researcher’s assessment of the quality of the organisation’s

data:
Likelihood of Veracity?
Data timeliness?
Capacity to judge itself accurately?
 353

Negligible extremity bias?

Negligible social desirability bias?
Scale: Very high / high / medium / low / very low
Score: 100, 75, 50, 25, 0
Range: 0 - 100
Low - High used: Low High
Likelihood of Veracity? 50 100
Data timeliness? 50 100
Capacity to judge itself accurately? 50 100
Negligible extremity bias? 50 100
Negligible social desirability bias? 50 100

Composite:
Σ of the 5 scores ÷ 5
Range: 0 - 100
Low - High used: 50, 100
 354

APPENDIX C

RESEARCHER CONFIDENTIAL QUESTIONNAIRE

~ Data availability and quality:

Markets: Is much data more than 2 years old? (timeliness) None / a little / a lot
Do you believe that the data sources used in the framework did maximise accuracy? Yes / No
Do you believe that the data sources used in the framework did maximise comparability? Yes / No
Per cent of data cells missing (completeness) ..... % (computer to calculate then report)
If data are missing, do you believe the missing items are likely to materially affect the outcome of the screening
recommendation? (significance) None / a little / a fair amount / much
Organisation: Researcher’s assessment of the quality of the organisation’s data:
Likelihood of: Very high High Medium Low Very low
Veracity
Data timeliness
Capacity to judge
itself accurately
Negligible
extremity bias
Negligible social
desirability bias

~ Without initially revealing the reason, discuss personal benefits from further internationalising (e.g. travel to Mediterranean
countries). These can distort later stage of market selection so may require additional (6th) objective group being inserted (in
questionnaire at section ‘Objectives of this market selection’ & elsewhere). Requires ascertainment of motives, determining
country-selecting mechanism/s, variable weighting, and alteration of statistical processing. Discus with client before
implementing.
 355

APPENDIX D

PRODUCT DEFINITIONS

Coking Coal (Product 1)

Coking Coal is coal with a quality that allows the production of coke suitable to
support a blast furnace charge (International Energy Agency 1998). Metric
tonnes, net weight. No unique SITC or ISIC. Roughly similar to SITC Rev 3:
321 and ISIC Rev 2: 2100-01 but these are product aggregations whereas the
IEA data are refined to be a technically homogeneous product.

Beef (Product 2)

Meat of bovine animals, fresh chilled or frozen, common trade names being
beef and veal. Boneless, dressed carcass weight (weight minus all parts –
edible and inedible – that are removed in dressing the carcass). The concept
varies widely from country to country and according to the various species of
livestock. Metric tonnes, net weight (boneless). FAO Statistical code number
0870. Very similar to SITC Rev. 3: 011 and ISIC Rev. 2: 1511-01 but they
require elimination of weight of bones and buffalo meat.

Greasy Wool (Product 3)

A natural fibre taken from sheep or lambs. Includes fleece-washed, shorn and
pulled wool (from slaughtered animals), but does not include carded or
combed wool. Metric tonnes, net weight. FAO statistical code 0987. SITC
Rev. 3: 268.1

Telephone Sets (Product 4)

Includes telephones which are coin-operated, portable or for mounting on

walls (but note exception 3 below), sealed telephones for use in mines,
 356

military field telephones, ‘Parlophones” for buildings, telephone answering

machines forming an integral part of a telephone set, electrical line telephone
apparatus, telephone apparatus for carrier-current line systems, telephone
sets and subscribers’ telephone sets. Excludes: (1) telephone answering
machines designed to operate with a telephone set but not forming an integral
part of the set, (2) parts of the apparatus (3) radio telephones. Number of
sets. SITC Rev. 3: 764.11; ISIC Rev. 2: 3832-10.

Tertiary Education (Product 5)

Students enrolled in universities and equivalent degree-granting institutions,

including distance learning and other third level institutions (e.g. teacher
training colleges, technical colleges, etc.), both public and private, both part-
time and full-time. Number of students. (UNESCO 1996 Statistical Yearbook
(1997) pp. 3-239, 3-271.) No unique SITC Rev. 3 or ISIC Rev. 2 but appears
in ISIC Rev. 3: 8030.

Internal Combustion Petrol Engines (Product 6)

Internal-combustion piston engines (spark-ignition engines) for motor vehicles

(passenger cars and commercial vehicles) equipped with cylinders, pistons,
connecting-rods, crankshafts, flywheels, inlet and exhaust valves, etc.. Spark-
ignition motor vehicle engines are usually petrol-fuelled but some may employ
other fuel (e.g. propane). Rotary engines of the spark-ignition type are
included. Motorcycle engines are excluded from ISIC but included in SITC.
Number of engines. ISIC Rev 2: 3843-01A plus motorcycle engines; SITC
Rev. 3: 713.21-0A, 713.22-0A.
 357

APPENDIX E

LIST OF VARIABLES

Variables from literature Variable use:

review:
Not Not Used in Used to Manipulated
quant required data adjust in SPSS
ified for these preparation for firm Low High
6 or compet firm firm
products manipulation ence
Markets’ attributes
(Potential customer needs) 9
Segment population sizes 9
{Segments} 9
Intrinsic adoption rates 9
Product consumption (3) 9 9
Product consumption growth (6) 9 9
Competition & substitutes 9 9
Physical 9
Economic (6) 9 9
Technological 9
Legal 9 9
Political 9*
Cultural (21) 9 9
Ecological/environmental 9
Market risk 9 9
(see below re Data availability & quality)
Organisation’s attributes
SCA - generic (38) 9
- market specific 9
General business competencies (4) 9
Further internationalisation
competencies:
- Generic competencies (18) 9
- Market specific competencies:
Language (2) 9 9
Market research 9
Contacts 9
(see above re SCA)
Key stakeholder support 9
Other resources (2) 9
Further internationalisation values (12) 9
{Strategic values} 9
Personal values:
Cultural difference (21) 9
Risk values (2) 9
Market research values (2) 9
{Other personal values} 9
 358

Data availability & quality:

(markets) (5) (organisation) (5) 9
Objectives
Sales 9
Sales growth 9
Other business benefits 9
Risk level 9
{Personal benefits} 9
Effort level (8) 9
{Targets} 9
{Strategy decisions} 9
(Size of assessment task) 9
(Criteria & method of evaluation) 9
Intervening variables
Motivation & commitment to
further internationalise (6) 9
{Marketing mix} 9
Customer adoptions & usage rate 9
Organisation’s product-specific
demand 9
Outcome
(Selection process) 9
(Clusters of markets) output output

Totals:
Quantified = 27 L / 30 H 10 9 8 11
Non-quantified or not required
= 16 L / 16 H 8 8
Source: Constructed for this research from Figure 4.3 and the section on
Measurement of Dimensions in Chapter 4

Τ or Τ = 1 of the 46 variables
Grey scale = a heading
{Grey scale} = a descriptive (unquantifiable) variable
Black italic = a quantified variable
Black = a quantifiable variable, but not required for these 6 products
(Black) = other irregularity
(…) = Number of sub-variables used to compose the variable
No figure beside the variable indicates a single variable
* = Combined with the Market risk variable
L = Low organisation competencies and values
H = High organisation competencies and values

Total variables potentially relevant and quantifiable = 30 + 8 = 38

Additional variables not quantifiable = 8
Total variables relevant to a screening decision = 46
 359

APPENDIX F

VALUES OF VARIABLES – SUMMARY

(See Measurement of Dimensions, Chapter 4, for full details)

Scale Range # Value (Low – High) Source

Countries A–Z File: Countries List CIA Factbook ex WWW modified
# Most variables commence on a 0-100 scale then all get rescaled 0-1

The Organisation – Generic Competencies

SCA- generic (38 items) Aaker’s 1995:80 checklist modified
(see below for market-specific) 0 – 100 30, 100 in Questionnaire to management
General business competencies (12→4) 0 – 100 35, 100 Questionnaire to management
Further internationalisation Questionnaire: In Stage of
competencies – generic (56→18) 0 – 100 13, 100 internationalisation + Competencies
Generic International market research Questionnaire to management. This
competence (4 general items from above item not separately calculated during research
but organisation should be told its score for this
+ [9+5] screening items) 0 – 100 0, 100 ability
62
Key stakeholder support 0 – 100 30, 100 Questionnaire to management
Other Resources (2 items) 0– 1 1, 1 Questionnaire to management
Composite of Organisation’s Generic
Competencies Calculated by framework
- Organisation’s Market-specific International Capabilities (if any) MS
Market-specific SCA’s 56 0 – 100 0 , 100 * Questionnaire to management
60
Language proficiencies (2 items) 0 – 100 See Appendix G Questionnaire to management
58 59
Market research 0 – 100 0, 100 ** Questionnaire to management
63
Stakeholder & network contacts 0 – 100 0, 100 * Questionnaire to management
Composites for each market Calculated by framework
* in 30% of the product’s markets, selected randomly
** in 5 of the product’s markets, selected randomly
 360

- Organisation’s Values
Further internationalisation values (12 items) 0 - 100 54, 100 Questionnaire to management
{Strategic values} – – {Non quantitative descriptive item}
Personal values including 0 - 100 0, 100 Partially quantified, partially not
cultural difference values (21 items) Australia Australia Anchor for cultural difference calculation
risk preferences (2 items) 0 – 100 50, 100 Questionnaire to management
market research values (2 items) & 0 – 100 50, 100 Questionnaire to management
{other values} Not quantified
Composite of Organisation’s values Calculated by framework

Each Market’s attributes

(Potential customer needs) – – Regression equation / Description
Segment population sizes – – Not required for these products
{Segments: geo-, demo-, psycho-, etc.}PS – – {A non quantitative descriptive item}
Intrinsic adoption rates PS MS:
Product 1 - coking coal 0 – 100 1 Questionnaire to management
Product 2 - beef 0 – 100 1 Questionnaire to management
Product 3 - wool 0 – 100 1 Questionnaire to management
Product 4 - telephones 0 – 100 1 Questionnaire to management
Product 5 - tertiary education 0 – 100 1 Questionnaire to management
Product 6 - petrol engines 0 – 100 1 Questionnaire to management
Segment demand volumes PS (3 items):
Local production MS
Product 1 - coking coal 0 - 1,000 million XL file ‘Coking coal data ex IEA’ IEA e-mail 2/12/1998
Product 2 - beef 0 - 1,000 million XL file ‘Beef stat’s FAO587’ FAO e-mail 9/11/1998
Product 3 - wool 0 - 1,000 million XL file ‘Wool production ex FAO’ FAO WWW site
Product 4 - telephones 0 - 1,000 million ASCII file ‘Telephone production UN e-mail 23/1/1999
Product 5 - tertiary education 0 - 1,000 million XL file ‘Education enrolments ..’ UNESCO
Product 6 - petrol engines 0 - 1,000 million XL file ‘Engines production …’ UN e-mail 16/12/1998
 361

Imports MS
Product 1 - coking coal 0 - 1,000 million XL file ‘Coking coal data ex IEA’ IEA e-mail 2/12/1998
Product 2 - beef 0 - 1,000 million XL file ‘Beef stat’s FAO607’ FAO e-mail 9/11/1998
Product 3 - wool 0 - 1,000 million XL file ‘Wool imports ex FAO’ FAO WWW site
Product 4 - telephones 0 - 1,000 million ASCII file ‘Sitcrev3’ UN e-mail 4/2/1999
Product 5 - tertiary education 0 - 1,000 million XL file ‘Education foreign’ UNESCO e-mail 17/11/1998
Product 6 - petrol engines 0 - 1,000 million ASCII file ‘Sitcrev3’ UN e-mail 4/2/1999
Exports MS
Product 1 - coking coal 0 - 1,000 million XL file ‘Coking coal data ex IEA’ IEA e-mail 2/12/1998
Product 2 - beef 0 - 1,000 million XL file ‘Beef stat’s FAO607’ FAO e-mail 9/11/1998
Product 3 - wool 0 - 1,000 million XL file ‘Wool exports ex FAO’ FAO WWW site
Product 4 - telephones 0 - 1,000 million ASCII file ‘Sitcrev3’ UN e-mail 4/2/1999
Product 5 - tertiary education 0 - 1,000 million XL file ‘Education foreign’ & UNESCO e-mail 17/11/1998 &
‘Education UNESCO 1998 …’ UNESCO e-mail 10/12/1998
Product 6 - petrol engines 0 - 1,000 million ASCII file ‘Sitcrev3’ UN e-mail 4/2/1999
PS
(Industry demand volume ) – – Not required for these products
MS
Adjustments to segment/industry demand – – Not required for these products
(Composite P + I – E ± A) 0 - 1,000 million Actual Calculated by framework
Growth in demand (shift-share) (3 items, 2 0 - 1,000 million;
0 - ± 100
MS
years each) Actual Calculated by framework
Competition and substitutes
(organisation & product-specific) (see SCA) –
Industry-specific MS – – Not used in this research
Economy-wide MS -3 - +3 → 0 -100 Global Competitiveness Report WEF 1997
Physical PS MS:
Product 1 - coal 0 – 100 0 Questionnaire to management
Product 2 - beef 0 – 100 0 Questionnaire to management
Product 3 - wool 0 – 100 0 Questionnaire to management
Product 4 - telephones 0 – 100 0 Questionnaire to management
 362

Product 5 - tertiary education 0 – 100 0 Questionnaire to management

Product 6 - petrol engines 0 – 100 0 Questionnaire to management
Economic MS:
log GNP PPP per capita 0 – 10 Actual World Bank 1998 then log
Shift-share of GNP (PPP) per capita 0 – 100 % Actual ) World Bank 1998
Population 0 – 10 billion Actual ) supplemented by
Openness (Exports + Imports / GNP) 0 – 100 % Actual ) CIA 1998, IMF 1998 & 1999,
) UNDP 1998, WTO 1999
Technological PS MS
Product 1 - coal 0 – 100 0 Questionnaire to management
Product 2 - beef 0 – 100 0 Questionnaire to management
Product 3 - wool 0 – 100 0 Questionnaire to management
Product 4 - telephones 0 – 100 0 Questionnaire to management
Product 5 - tertiary education 0 – 100 0 Questionnaire to management
Product 6 - petrol engines 0 – 100 0 Questionnaire to management
Legal environment 1 - 5 → 0 -100 XL file ‘Legal arbitrariness - EIU’ EIU e-mail 28/1/1999
Political – – Included in Market Risk below
Cultural (21 → 5 items) 0 – 100 XL file ‘Cultural distance 3’ Hofstede (1980; 2002 in publication)
Ecological PS MS:
product 1 - coal -50 - +50 → 0 -100 0 Questionnaire to management
product 2 - beef -50 - +50 → 0 -100 0 Questionnaire to management
product 3 - wool -50 - +50 → 0 -100 0 Questionnaire to management
product 4 - telephones -50 - +50 → 0 -100 0 Questionnaire to management
product 5 - tertiary education -50 - +50 → 0 -100 0 Questionnaire to management
product 6 - petrol engines -50 - +50 → 0 -100 0 Questionnaire to management
Market Risk levels MS Country risk levels 0 – 100 Adrian Shute 28/1/1999 fax PRS Composite Risk Index
MS Researcher estimates & XL calculations
Data availability & quality (5 items) 0 – 100 Varies by market
(New and existing markets) – – Check existing markets included
 363

(Composite, each Market’s attractiveness) MS Calculated by framework

Organisation’s
Objectives: Sales 0 - 100 55, 35 (Low, High) Demonstration sample figures
Sales growth 0 - 100 5, 35 (Low, High) Demonstration sample figures
{Personal benefits} – – –
Other business benefits 0 - 100 0, 0 Questionnaire to management
Risk level chosen 0 - 100 20, 15 (Low, High) Demonstration sample figures
Effort (8 items) 0 - 100 20, 15 (Low, High) Demonstration sample figures
{Targets} – – {Non quantitative descriptive item}
{Strategy decisions} – – {Non quantitative descriptive item}
(Size of assessment task) 1 (= screen) 1 (No option allowed)
(Criteria and method of evaluation) – – (Fixed settings)
Data availability and quality (5 items) 0 - 100 50, 100 Demonstration sample figures
Intervening variables
Motivation and commitment (6 items) 0 – 100 33, 100 Questionnaire to management
{Marketing mix} – – {Non quantitative descriptive item}
MS
Customer adoptions and usage rate – – {Non quantitative descriptive item}
Organisation’s product-specific demand MS – – {Non quantitative descriptive item}
(Selection process) – – (Fixed – see Dissertation)

Preferred short-list of markets A-Z Calculated Output – calculated by framework

Source: Constructed for this research from Measurement of Dimensions, Chapter 4

XXX = quantified variable XXX = quantifiable variable XXX = a heading

{XXX} = Non-quantitative variable (XXX) = Some other irregularity M = Million
PS MS
= Product-specific variable = Market-specific variable
Values of ‘Questionnaire to management’ items determined by researcher
 364

APPENDIX G

LANGUAGE SCORES BY COUNTRY AND ORGANISATION COMPETENCE LEVEL

For the purpose of testing the significance of language skill on country selection, the hypothetical low ability and high ability
organisations needed to each be given a score for language competence. To simulate reality, some rules which were logical
and potentially representative needed to be developed to score this variable, “Market-specific language proficiencies”. These
scores need to be recalculated when applying the screening framework in a country whose native language is other than
English.

These two rules were used:

• The low ability Australian organisation is assumed to speak only English. These earn it varying language competence scores
in the specified countries (see below);
• The high ability organisation is assumed to have native speakers of the following 10 languages:
Arabic, Cantonese, English, French, Hindi, Japanese, Mandarin, Portuguese, Russian and Spanish (details again
below);
These have been chosen because they are: (1) the principal languages spoken by over 3 billion persons, and (2) believed to
be the most important commercially (after considering every country’s languages, and after giving additional weight to
countries with populations exceeding 100 million people).
 365

In a normal commercial setting, if the top market for the product used a language different from these 10, it is likely that the high
ability organisation would obtain staff capable of reading and writing that additional language. That further adjustment has
not been made here because it would cause multicollinearity.

These scores have been estimated by the researcher after considering CIA Factbook data about:
• the prevalence of the use of the language in the country,
• the percentage literacy levels of the population,
• which language is the official language,
• ethnic groups,
• colonial history, and
• the geographic contiguity of major language groups.
The scores give an indication of the quantity and quality of understanding of communications between the Australian
business person and foreign business people, and (to a lesser extent) the foreign technical specialists and bureaucrats
with whom the Australian organisation would be expected to at times deal.

Cultural values embedded in communications have not been included, being left to the ‘Culture’ variable.

Country scores for languages spoken fluently:

 366

Albanian: 100 – Albania

Arabic: 100 – Algeria, Bahrain, Comoros, Djibouti, Egypt, Gaza Strip, Jordan, Kuwait, Lebanon, Libya, Mauritania, Morocco,
Oman, Qatar, Saudi Arabia, Syria, Tunisia, Yemen,
80 – Chad, Eritrea, Iraq, Israel, Sudan, United Arab Emirates, West Bank, Western Sahara,
60 – Central African Republic, Ethiopia, Mali, Somalia, Tanzania
Cantonese: 100 – Hong Kong, Macau,
80 – Singapore,
50 – China, Indonesia, Malaysia, Vietnam
Czech: 100 – Czech Republic
Danish: 100 – Denmark, Faroe Islands
Dutch: 100 – Belgium, Netherlands, Netherlands Antilles,
80 – Aruba, Suriname,
60 – Indonesia, Namibia, South Africa
English: 100 – American Samoa, Anguilla, Antigua and Barbuda, Australia, Barbados, Bermuda, British Virgin Islands, Canada,
Cayman Islands, Cook Islands, Dominica, Falkland Islands, Grenada, Guam, Guernsey, Ireland, Jersey, Isle of
Mann, Marshall Islands, New Zealand, Northern Mariana Islands, Pitcairn Island, Puerto Rico, St Helena, St
Kitts, St Vincent and the Grenadines, Turks and Caicos Islands, Tuvalu, U.K., U.S.A.,
80 – Bahamas, Bangladesh, Belize, Brunei, Fiji, Gambia, Ghana, Gibraltar, Guyana, Hong Kong, India, Indonesia,
Israel, Jamaica, Kenya, Kiribati, Lesotho, Malawi, Malaysia, Maldives, Malta, Mauritius, Micronesia (Federated
States of), Montserrat, Namibia, Nauru, Nepal, Netherlands Antilles, Nigeria, Niue, Pakistan, Palau, PNG,
 367

Philippines, St Lucia, Samoa, Seychelles, Singapore, South Africa, Swaziland, Thailand, Tokelau, Tonga,
Trinidad and Tobago, Uganda, Vanuatu, Zambia, Zimbabwe,
60 – Argentina, Aruba, Bahrain, Botswana, Brazil, Cameroon, Costa Rica, Cyprus, Egypt, Ethiopia, French Polynesia,
Gaza Strip, Greece, Jordan, South Korea, Kuwait, Laos, Lebanon, Liberia, Libya, Luxembourg, Macau, Monaco,
New Caledonia, Nicaragua, Oman, Panama, Qatar, Rwanda, Sierra Leone, Somalia, Sri Lanka, Suriname,
Tanzania, United Arab Emirates, Vietnam, Virgin Islands, West Bank
French: 100 – France, French Guiana, French Polynesia, Guadeloupe, Martinique, Monaco, New Caledonia, St Pierre and
Miquelon,
80 – Andorra, Benin, Burkina Faso, Burundi, Canada, Central African Republic, Chad, Comoros, Congo (Democratic
Republic of, The), Congo, (Republic of, The), Cote d’Ivoire, Djibouti, Equatorial Guinea, Gabon, Guernsey,
Guinea, Jersey, Madagascar, Mali, Mayotte, Morocco, Niger, Reunion, St Vincent and the Grenadines, Senegal,
Seychelles, Togo, Wallis and Futuna,
60 – Belgium, Cambodia, Cameroon, Dominica, Egypt, Greece, Grenada, Haiti, Laos, Lebanon, Luxembourg,
Mauritania, Mauritius, Rwanda, St Lucia, Switzerland, Tunisia, Vanuatu
German: 100 – Austria, Germany, Liechtenstein,
80 – Luxembourg, Switzerland,
60 – Italy, Namibia
Greek: 100 – Greece,
80 – Cyprus
Hindi: 80 – Fiji, India, Mauritius, Nepal, Suriname, Trinidad and Tobago, United Arab Emirates
Hungarian: 100 – Hungary
 368

Italian: 100 – Italy, San Marino,

80 – Gibraltar,
60 – Argentina, Libya, Monaco, Somalia
Japanese: 100 – Japan,
60 – Guam
Kazakh: 100 – Kazakhstan
Korean: 100 – North Korea, South Korea
Macedonian: 100 - Macedonia
Mandarin: 100 – China, Taiwan,
80 – Macau, Singapore,
60 – Hong Kong, Malaysia, Mongolia
Polish: 100 – Poland
Portuguese: 100 – Brazil, Portugal, Sao Tome and Principe,
80 – Angola, Guinea Bissau, Macau, Mozambique,
60 – Bermuda, Cape Verde, Gibraltar, Uruguay
Romanian: 100 – Moldova, Romania
Russian: 100 – Russia,
80 – Kazakhstan, Kyrgyzstan,
60 – Belarus, Estonia, Latvia, Lithuania, Moldova, Mongolia, Tajikistan, Ukraine
Serbo-Croatian: 100 – Bosnia and Herzegovina, Croatia, Serbia and Montenegro
Slovak: 100 – Slovakia
 369

Spanish: 100 – Chile, Colombia, Costa Rica, Cuba, Dominican Republic, Ecuador, Panama, Paraguay, Puerto Rico, Spain,
Venezuela,
80 – Andorra, Argentina, Bolivia, Brazil, El Salvador, Equatorial Guinea, Gibraltar, Honduras, Mexico, Nicaragua,
Peru, Uruguay,
60 – Belize, Guatemala, Netherlands Antilles, Portugal, U.S.A., Virgin Islands
Turkish: 100 – Turkey,
80 – Cyprus
Ukrainian: 100 – Ukraine

Bold language names are those assumed to be within the capability set of all high ability organisations;
English is the only language for which the low ability Australian organisation is given country scores.

Source: Constructed for this research after considering language details in the CIA World Factbook 1996-7.
 370

APPENDIX H

SELECTING THE OPTIMAL NUMBER OF CLUSTERS

(Summarising data from various tables in SPSS MDA output)

Coal
7 variables, Markets only, including China
No of No of pairs of No of variables No of significant *3 Significance % of variance % cross-
clusters variables failing tolerance Variables: Functions: of worst included in significant validated
*1 *2
correlated functions functions *4
2
3 *5 1 - 7 of 7 2 of 2 .000 100 % 94.2 %
*5
4 - - 7 3 of 3 .003 98.0 90.4
5 *5 - - 7 4 of 4 .003 98.7 92.3
6 - - 7 4 of 5 .257 98.8 92.3

7 variables, Markets only, excluding China

No of No of pairs of No of variables No of significant *3 Significance % of variance % cross-
clusters variables failing tolerance Variables: Functions: of worst included in significant validated
*1 *2
correlated functions functions *4
2 - - 5 of 7 1 of 1 .000 100 % 92.2 %
3 - - 5 2 of 2 .000 100 98.0
4 - - 5 3 of 3 .007 97.0 92.2
5 - - 5 3 of 4 .896 97.7 92.2
6
I I = Number of clusters indicated by analysis of dendrogram, agglomeration schedule and scree plot
*1 ≥ ± .70 level *2 ≥ .19 level *3 ≤.05 level (*1 & *3 determine support or not for Ho) *4 Eigenvalues ≥ 1.0
*5 Invalid solution as contains a single country cluster
xxx = invalid solution as last discriminant function fails the ≤.05 significance hurdle
Bold = optimal solution.
 371

Beef

7 variables, Markets only

No of No of pairs of No of significant *2 Significance of % of variance included in % cross-
clusters variables correlated *1 Variables: Functions: worst functions significant functions *3 validated
2
3 1 6 of 7 2 of 2 .000 96.8 % 94.6 %
4 1 6 of 7 3 of 3 .000 90 93
5 - 6 of 7 3 of 4 .085 88 96
6 - 6 of 7 4 of 5 .064 92 95
7 - 6 of 7 4 of 6 .322 92 95
8 - 7 of 7 5 of 7 .554 93 96

I I = Number of clusters indicated by analysis of dendrogram, agglomeration schedule and scree plot
*1 ≥ ± .70 level *2 ≤.05 level (*1 & *2 determine support or not for Ho) *3 Eigenvalues ≥ 1.0
xxx = invalid solution as last discriminant function fails the ≤.05 significance hurdle
Bold = optimal solution.
 372

Wool

7 variables, Markets only, including China

No of No of pairs of No of variables No of significant *3 Significance % of variance % cross-
clusters variables failing tolerance Variables: Functions: of worst included in significant validated
correlated *1 *2
functions functions *4
2
3
4 1 5 of 7 3 of 3 .046 87 % 94 %
*5
5 1 7 of 7 4 of 4 .024 93 93
6 - 7 of 7 4 of 5 .248 92 95
7

7 variables, Markets only, excluding China

No of No of pairs of No of variables No of significant *3 Significance % of variance % cross-
clusters variables failing tolerance Variables: Functions: of worst included in significant validated
correlated *1 *2
functions functions *4
2 1 0 6 of 7 1 of 1 .000 100 % 95.6 %
3 1 0 6 2 of 2 .015 98.1 97.4
4 1 0 6 3 of 3 .034 88.6 97.4
5 0 0 6 3 of 4 .494 88.8 94.7
6
I I = Number of clusters indicated by analysis of dendrogram, agglomeration schedule and scree plot
*1 ≥ ± .70 level *2 ≥ .19 level *3 ≤.05 level (*1 & *3 determine support or not for Ho) *4 Eigenvalues ≥ 1.0
*5 Invalid solution as contains a single country cluster
xxx = invalid solution as last discriminant function fails the ≤.05 significance hurdle
Bold = optimal solution.
 373

Telephones

7 variables, Markets only

No of No of pairs of No of significant *2 Significance of % of variance included in % cross-
*1
clusters variables correlated Variables: Functions: worst functions significant functions *3 validated
2
3 - 6 of 7 2 of 2 .000 96 % 97 %
4 - 6 of 7 3 of 3 .006 98 96
5
I I = Number of clusters indicated by analysis of dendrogram, agglomeration schedule and scree plot
*1 ≥ ± .70 level *2 ≤.05 level (*1 & *2 determine support or not for Ho) *3 Eigenvalues ≥ 1.0
Bold = optimal solution.

Education

7 variables, Markets only

No of No of pairs of No of significant *2 Significance of % of variance included in % cross-
clusters variables correlated *1 Variables: Functions: worst functions significant functions *3 validated
2
32 1 6 of 7 2 of 2 .000 97 % 94 %
4 - 6 of 7 2 of 3 .090 89 96
5 - 7 of 7 3 of 4 .177 94 94
6 - 7 of 7 4 of 5 .191 92 95
1
7 - 7 of 7 5 of 6 .347 93 94
1
I I = Number of clusters indicated by analysis of dendrogram, agglomeration schedule and scree plot. = 1st preference, 2 = 2nd possibility
* ≥ ± .70 level
1
* ≤.05 level (* & * determine support or not for Ho)
2 1 2
* Eigenvalues ≥ 1.0
3

xxx = invalid solution as last discriminant function fails the ≤.05 significance hurdle
Bold = optimal solution.
 374

Engines

7 variables, Markets only

No of No of pairs of No of significant *2 Significance of % of variance included in % cross-
clusters variables correlated *1 Variables: Functions: worst functions significant functions *3 validated
2
3 - 6 of 7 2 of 2 .041 95 % 90 %
4 - 7 of 7 3 of 3 .020 98 95
5

I I = Number of clusters indicated by analysis of dendrogram, agglomeration schedule and scree plot
*1 ≥ ± .70 level *2 ≤.05 level (*1 & *2 determine support or not for Ho) *3 Eigenvalues ≥ 1.0
xxx = invalid solution as last discriminant function fails the ≤.05 significance hurdle
Bold = optimal solution.
 375

Coal: generic and market variables [16 (8) and 19 (11) active variables]
No of Variable Weight of No of No of Variab No of significant *2 Significance % of % of Top 2 variables, Struct-
variab being the column clusters pairs of les Variab Functions of worst variance countries based on ‘Variables ure
les explored 1 variable: bold variables failing les functions included in cross- Entered / Removed’ Matrix
2x = best correlated toler significant validated table Agree
*1 ance *4 functions *3 s
=A
16 (8) Lo SCA .30 3 3 2 6 of 8 2 of 2 .000 100 % 96.2 % Language, Legal A
4 2 1 6 2 of 3 .498 99.9 96.2 Language, Legal
5 *5 1 1 8 3 of 4 .383 99.9 94.2 Language, Pdt gth
Lo GBC .35 3 3 2 6 2 of 2 .000 100 96.2 Language, Legal A
4 2 1 6 2 of 3 .497 99.9 96.2 Language, Legal A
5 *5 1 1 8 3 of 4 .383 99.9 94.2 Language, Pdt gth
Lo FIC .13 2 6 3 2 1 of 1 .000 100 98.1 Language, Legal
3 4 2 6 2 of 2 .035 99.5 100 Language, Economic A
4 2 1 6 2 of 3 .497 99.9 96.2 Language, Legal
Lo KSS .30 3 Same as Lo SCA (above)
6 *5 1 0 8 4 of 5 .343 99.2 94.2 Language, Consmn gth
Lo FIV .54 3 2 1 6 2 of 2 .000 100 94.2 Competition, Language
Lo RV .50 3 2 1 6 2 of 2 .000 100 94.2 Competition, Language
4 2 1 6 2 of 3 .483 99.5 92.3 Language, Legal A
Lo MRV .50 3 Same as Lo RV (above)
Lo M & C .33 3 3 2 6 2 of 2 .000 100 96.2 Language, Legal A
4 2 1 6 2 of 3 .497 99.9 96.2 Language, Legal
19 (11) Hi SCA 1.00 2 Same as Hi all generic incl. China (below)
3 Same as Hi all generic incl. China (below)
4 *5 0 0 7 of 11 3 of 3 .002 96.6 84.6 Competition, Consmn
5 *5 0 0 10 4 of 4 .000 96.1 94.2 Competition, Consmn gth
6 *5 0 0 11 5 of 5 .001 97.4 86.5 Competition, Consmn gth
7 *5 0 0 10 6 of 6 .002 96.1 84.6 Competition, Consmn gth
8 *5 0 0 10 7 of 7 .011 97.3 82.7 Competition, Consmn gth A
 376

Coal: generic and market variables [16 (8) & 19 (11) active variables] continued
Hi GBC 1.00 5 Same as Hi SCA (above)
16 (8) Lo all Σ .125 x 8 2 5 3 2 of 8 1 of 1 .000 100 % 98.1 % Language, Legal
generic x Q’aire 3 3 2 6 2 of 2 .000 100 96.2 Language, Legal A
=.36875 4 2 1 6 2 of 3 .415 99.7 92.3 Language, Legal A
Lo all ditto 2 6 3 of 8 1 of 1 .000 100 % 100 % Language, Consmn gth A
generic 3 3 7 2 of 2 .000 100 98.1 Language, Legal
test of 4 2 7 3 of 3 .000 100 96.2 Language, Legal
Pdt con 5 1 8 4 of 4 .000 100 94.2 Language, Consmn gth A
gth 6 1 8 5 of 5 .009 99.7 94.2 Language, Consmn gth A
sig 7 0 8 5 of 6 .104 99.7 90.4 Language, Consmn gth A
19 (11) Hi all Σ .091 x 11 2 2 1 5 of 11 1 of 1 .000 100 92.3 Economic, Mkt risk
generic x Q’aire 3 1 0 6 2 of 2 .004 92.4 86.5 Competition, Legal
(incl. = 1.0 4 *5 0 0 7 3 of 3 .002 96.5 84.6 Competition, Consmn
*5 Competition, Consmn gth
China) 5 0 0 10 4 of 4 .000 96.1 94.2
6 *5 0 0 11 5 of 5 .001 97.4 86.5 Competition, Consmn gth
*5 Competition, Consmn gth
7 0 0 10 6 of 6 .002 96.1 84.6
8 *5 0 0 10 7 of 7 .011 97.3 82.7 Competition, Consmn gth
*5 Competition, Consmn gth
9 0 0 10 8 of 8 .017 96.5 80.8
10 *5 0 0 11 7 of 9 .376 97.6 80.8 Competition, Language
Hi all Σ .091 x 11 2 0 1 5 of 11 1 of 1 .000 100 88.2 Competition, Legal
generic x Q’aire 3 0 0 7 2 of 2 .000 100 94.1 Competition, Language
(excl. = 1.0 4 1 0 7 3 of 3 .000 100 96.1 Competition, Culture
China) 5 1 0 7 4 of 4 .000 100 90.2 Culture, Competition A
*5 Competition, Culture
6 1 0 8 5 of 5 .010 97.0 86.3 A
7 *5 1 0 8 5 of 6 .078 98.5 86.3 Culture, Competition
*1 ≥ ± .70 level *2 ≤.05 level. These 2 matters determine support or not for Ho if ≥ 3 clusters (use variables only if 2 clusters) *3 Eigenvalues ≥ 1.0
*4 Tolerance is < .19, indicating possible multicollinearity (Hair et al 1995, p. 127)
5
* Invalid solution as has a single country cluster
xxx = invalid solution as last discriminant function fails the ≤.05 significance hurdle
Bold = optimal solution.
 377

Beef: generic and market variables [16 (8) & 19 (11) active variables]
No of Variable Weight of No of No of No of significant *2 Significance % of % of countries Top 2 variables, Struct-
variab being the column clusters pairs of Variables Functions of worst variance cross- based on ‘Variables ure
les explored 1 variable: bold variables functions included in validated Entered / Removed’ Matrix
2x = best correlated *1 significant table Agrees
functions *3 =A
16 (8) Lo SCA .30 3 2 6 of 8 2 of 2 .000 95.5 % 99.0 % Language, Legal A
4 1 6 3 of 3 .027 99.7 95.6 Language, Legal A
5 1 7 3 of 4 .195 99.1 93.2 Language, Legal A
Lo GBC .35 3 2 6 2 of 2 .000 95.5 99.0 Language, Legal A
4 1 6 3 of 3 .012 99.6 95.6 Language, Legal A
5 1 7 3 of 4 .075 98.5 96.1 Language, Legal A
Lo FIC .13 3 2 5 2 of 2 .030 99.9 100.0 Language, Legal A
4 2 6 2 of 3 .344 100.0 99.0 Language, Legal A
Lo KSS .30 3 Same as Lo SCA (above)
5 1 7 3 of 4 .195 99.1 93.2 Language, Legal A
Lo FIV .54 3 2 6 2 of 2 .000 100 97.1 Language, Legal A
4 1 6 3 of 3 .006 99.6 96.6 Language, Legal A
5 1 7 3 of 4 .075 98.5 96.1 Language, Legal
9 0 7 4 of 8 .987 97.2 96.6 Language, Legal
Lo RV .50 3 2 6 2 of 2 .000 100 97.1 Language, Legal A
4 1 6 3 of 3 .011 99.6 97.1 Language, Legal A
9 0 7 4 of 8 .988 99.7 96.6 Language, Legal
Lo MRV .50 3 Same as Lo RV (above)
4 1 6 3 of 3 .011 99.6 97.1 Language, Legal A
9 0 7 4 of 8 .988 99.7 96.6 Language, Legal
Lo M & C .33 3 2 6 2 of 2 .000 95.5 99.0 Language, Legal A
4 1 6 3 of 3 .027 99.7 95.6 Language, Legal A
7 1 7 4 of 6 .174 99.0 96.6 Language, Legal
19 (11) Hi SCA 1.0 3 1 6 of 11 2 of 2 .000 100 93.2 SCA (country), Legal A
& 4 1 8 3 of 3 .000 95.1 91.7 SCA (country), Legal A
5 1 9 4 of 4 .000 82.7 91.7 SCA (country), Legal
all 6 1 8 4 of 5 .953 96.3 89.8 Contacts, SCA (country)
 378

other 9 1 8 6 of 8 .979 96.4 93.7 SCA (country), Contacts

Hi’s 14 1 10 8 of 10 .471 98.7 90.7 SCA (country), Contacts
16 (8) Lo all Σ .125 x 8 2 3 5 of 8 1 of 1 .000 100 100 Language, Legal
generic x Q’aire 3 2 6 2 of 2 .000 95.5 99.0 Language, Legal A
=.36875 4 1 6 3 of 3 .012 99.6 95.6 Language, Legal A
5 1 7 3 of 4 .075 98.5 96.1 Language, Legal A
19 (11) Hi all Σ .091 x 11 2 1 5 of 11 1 of 1 .000 100 92.7 Legal, Economic
generic x Q’aire 3 1 6 2 of 2 .000 100 93.2 SCA (country), Legal A
= 1.0 4 1 8 3 of 3 .000 95.1 91.7 SCA (country), Legal A
5 1 9 4 of 4 .000 82.7 91.7 SCA (country), Legal
6 1 8 4 of 5 .953 96.3 89.8 Contacts, SCA (country)
*1 ≥ ± .70 level *2 ≤.05 level. These 2 matters determine support or not for Ho if ≥ 3 clusters (use variables only if 2 clusters) *3 Eigenvalues ≥ 1.0
xxx = invalid solution as last discriminant function fails the ≤.05 significance hurdle
Bold = optimal solution.
 379

Wool: generic and market variables [16 (8) & 19 (11) active variables]
No of Variable Weight of No of No of No of significant *2 Significance % of variance % of Top 2 variables, Struct-
variab being the column clusters pairs of Variables Functions of worst included in countries based on ‘Variables ure
les explored 1 variable: bold variables functions significant cross- Entered / Removed’ Matrix
2x = best correlated *1 functions *3 validated table Agrees
=A
16 (8) Lo SCA .30 2 4 5 of 8 1 of 1 .000 100 % 100 % Language, Legal
3 2 6 2 of 2 .000 100 98.3 Language, Legal A
4 1 6 3 of 3 .010 99.6 98.3 Language, Legal A
5 1 7 4 of 4 .011 98.8 97.4 Language, Legal A
6 1 7 5 of 5 .023 98.1 92.2 Language, Legal A
Lo GBC .35 3 2 6 2 of 2 .000 100 98.3 Language, Legal A
4 1 6 3 of 3 .010 99.6 98.3 Language, Legal A
5 1 7 4 of 4 .009 98.8 96.5 Language, Legal A
Lo FIC .13 2 4 5 1 of 1 .000 100 100 Language, Legal
3 4 4 1 of 2 .155 99.9 100 Language, Culture
4 2 6 2 of 3 .207 99.9 97.4 Language, Legal A
5 2 7 3 of 4 .621 82.0 58.3 Legal, *5 Cultural
Lo KSS .30 2 4 5 1 of 1 .000 100 100 Language, Legal
4 Same as Lo SCA (above)
Lo FIV .54 3 2 6 2 of 2 .000 100 98.3 Language, Legal A
4 1 6 3 of 3 .004 99.4 95.7 Language, Legal A
5 1 7 4 of 4 .003 98.4 98.3 Language, Legal A
6 1 7 4 of 5 .069 97.5 94.8 Language, Legal A
Lo RV .50 2 4 5 1 of 1 .000 100 100 Language, Legal
3 Same as Lo MRV (below)
4 1 6 3 of 3 .004 99.4 95.7 Language, Legal A
5 1 7 4 of 4 .003 98.4 98.3 Language, Legal A
6 1 7 4 of 5 .069 97.5 94.8 Language, Legal A
Lo MRV .50 3 2 6 2 of 2 .000 100 98.3 Language, Legal A
5 Same as Lo RV (above)
Lo M & C .33 2 4 5 1 of 1 .000 100 100 Language, Legal
5 1 7 4 of 4 .009 98.8 96.5 Language, Legal A
 380

6 1 7 5 of 5 .038 98.1 92.2 Language, Legal A

19 (11) Hi SCA 1.0 2 1 5 of 11 1 of 1 .000 100 93.0 Economic, Legal A
& 3 1 6 2 of 2 .000 100 92.2 Contacts, Economic
all 4 1 7 3 of 3 .000 95.3 93.0 Contacts, Economic
other 5 1 8 4 of 4 .001 97.2 94.8 Contacts, Economic
Hi’s 6 1 7 3 of 5 .212 91.1 59.1 Legal, Economic A
16 (8) Lo all Σ .125 x 8 2 4 5 of 8 1 of 1 .000 100 100 Language, Legal
generic x Q’aire 3 2 6 2 of 2 .000 100 98.3 Language, Legal A
=.36875 4 1 6 3 of 3 .010 99.6 98.3 Language, Legal A
5 1 7 4 of 4 .009 98.8 96.5 Language, Legal A
6 1 7 5 of 5 .038 98.1 92.2 Language, Legal A
7 1 8 6 of 6 .012 98.7 93.0 Language, Legal A
8 1 8 7 of 7 .023 97.9 90.4 Language, Legal A
19 (11) Hi all Σ .091 x 11 2 1 5 of 11 1 of 1 .000 100 93.0 Economic, Legal A
generic x Q’aire 3 1 6 2 of 2 .000 100 92.2 Contacts, Economic
= 1.0 4 1 7 3 of 3 .000 95.3 93.0 Contacts, Economic
5 1 8 4 of 4 .001 97.2 94.8 Contacts, Economic
6 1 7 3 of 5 .212 91.1 59.1 Legal, Economic A
*1 ≥ ± .70 level *2 ≤.05 level. These 2 matters determine support or not for Ho if ≥ 3 clusters (use variables only if 2 clusters) *3 Eigenvalues ≥ 1.0
*5 Competition ranked 2nd but disregarded because its tolerance is < .19, indicating multicollinearity (Hair et al 1995, p. 127)
xxx = invalid solution as last discriminant function fails the ≤.05 significance hurdle
Bold = optimal solution.
 381

Telephones: generic and market variables [16 (8) & 19 (11) active variables]
No of Variable Weight of No of No of No of significant *2 Significance % of variance % of Top 2 variables, Struct-
variab being the column clusters pairs of Variables Functions of worst included in countries based on ‘Variables ure
les explored 1 variable: bold variables functions significant cross- Entered / Removed’ Matrix
2x = best correlated *1 (≤ .05?) functions *3 validated table Agrees
=A
16 (8) Lo SCA .30 2 3 2 of 8 1 of 1 .000 100 % 100 % Language, Culture
3 2 6 2 of 2 .000 100 98.3 Language, Legal A
4 2 7 3 of 3 .002 99.3 97.4 Language, Legal A
5 1 7 4 of 4 .036 98.3 97.4 Language, Legal A
6 1 7 4 of 5 .394 98.2 98.3 Language, Legal A
Lo GBC .35 3 2 6 2 of 2 .000 100 98.3 Language, Legal A
4 2 8 3 of 3 .000 99.0 96.6 Language, Legal A
5 1 7 4 of 4 .036 98.3 97.4 Language, Legal A
6 1 7 4 of 5 .394 98.2 98.3 Language, Legal A
Lo FIC .13 2 3 2 1 of 1 .000 100 100 Language, Culture
3 3 2 1 of 2 .139 99.9 100 Language, Culture A
Lo KSS .30 5 Same as Lo SCA (above)
Lo FIV .54 2 3 2 1 of 1 .000 100 100 Language, Culture
3 2 6 2 of 2 .000 100 95.7 Language, Legal A
4 2 7 3 of 3 .003 99.2 95.7 Language, Legal A
5 1 7 3 of 4 .261 98.6 95.7 Language, Legal A
Lo RV .50 3 2 6 2 of 2 .000 100 98.3 Language, Legal A
4 2 7 3 of 3 .008 99.4 95.7 Language, Legal A
5 2 7 3 of 4 .280 99.1 97.4 Language, Legal A
Lo MRV .50 4 Same as Lo RV (above)
Lo M & C .33 3 2 6 2 of 2 .000 100 98.3 Language, Legal A
4 2 7 3 of 3 .012 99.4 95.7 Language, Legal A
5 1 7 4 of 4 .036 98.3 97.4 Language, Legal A
6 1 7 4 of 5 .394 98.2 98.3 Language, Legal A
19 (11) Hi SCA 1.0 3 0 9 of 11 2 of 2 .000 100 96.6 Legal, Language
4 1 9 3 of 3 .000 100 89.7 SCA (country), Legal
& all other 5 1 9 4 of 4 .000 92.9 90.5 SCA (country), Legal
 382

Hi’s 6 1 8 4 of 5 .695 99.6 93.1 Contacts, Legal

16 (8) Lo all Σ .125 x 8 2 3 2 of 8 1 of 1 .000 100 100 Language, Culture
generic x Q’aire 3 2 6 2 of 2 .000 100 98.3 Language, Legal A
=.36875 4 2 7 3 of 3 .008 99.4 95.7 Language, Legal A
5 2 7 3 of 4 .280 99.1 97.4 Language, Legal A
19 (11) Hi all Σ .091 x 11 2 1 7 of 11 1 of 1 .000 100 94.8 Economic, Culture
generic x Q’aire 3 0 9 2 of 2 .000 100 96.6 Legal, Language
= 1.0 4 1 9 3 of 3 .000 100 89.7 SCA (country), Legal
5 1 9 4 of 4 .000 92.9 90.5 SCA (country), Legal
6 1 8 4 of 5 .695 99.6 93.1 Contacts, Legal
*1 ≥ ± .70 level *2 ≤.05 level. These 2 matters determine support or not for Ho if ≥ 3 clusters (use variables only if 2 clusters) *3 Eigenvalues ≥ 1.0
xxx = invalid solution as last discriminant function fails the ≤.05 significance hurdle
Bold = optimal solution.
 383

Tertiary Education: generic and market variables [16 (8) & 19 (11) active variables]
No of Variable Weight of No of No of No of significant *2 Significance % of variance % of Top 2 variables, Struct-
variab being the column clusters pairs of Variables Functions of worst included in countries based on ‘Variables ure
les explored 1 variable: bold variables functions significant cross- Entered / Removed’ Matrix
2x = best correlated *1 functions *3 validated table Agrees
=A
16 (8) Lo SCA .30 2 5 5 of 8 1 of 1 .000 100 % 100 % Language, Legal
3 2 6 2 of 2 .000 95.4 98.4 Language, Legal A
4 2 6 3 of 3 .044 99.7 95.3 Language, Legal A
5 2 7 3 of 4 .292 99.0 95.3 Language, Legal A
6 1 7 4 of 5 .190 99.0 94.3 Language, Legal A
7 1 7 5 of 6 .085 98.9 94.8 Language, Legal A
8 1 7 5 of 7 .355 98.7 95.3 Language, Legal A
Lo GBC .35 2 5 5 1 of 1 .000 100 100 Language, Legal
3 2 6 2 of 2 .000 95.4 98.4 Language, Legal A
4 2 6 3 of 3 .019 99.6 95.9 Language, Legal A
5 1 7 3 of 4 .174 98.7 95.9 Language, Legal A
Lo FIC .13 2 5 5 1 of 1 .000 100 100 Language, Legal
3 4 6 2 of 2 .044 99.9 100 Language, Legal A
4 2 6 2 of 3 .231 99.9 98.4 Language, Legal A
Lo KSS .30 3 Same as Lo SCA (above)
Lo FIV .54 2 3 5 1 of 1 .000 100 97.9 Language, *7 Legal
3 2 6 2 of 2 .000 100 95.9 Language, Legal A
4 2 6 3 of 3 .017 99.6 96.9 Language, Legal A
5 2 7 3 of 4 .166 98.6 95.9 Language, Legal A
Lo RV .50 2 5 5 1 of 1 .000 100 100 Language, Legal
3 2 6 2 of 2 .000 100 95.9 Language, Legal A
4 2 6 3 of 3 .017 99.6 96.9 Language, Legal A
5 2 7 3 of 4 .166 98.6 95.9 Language, Legal A
Lo MRV .50 4 Same as Lo RV (above)
Lo M & C .33 2 5 5 1 of 1 .000 100 100 Language, Legal
3 2 6 2 of 2 .000 95.4 98.4 Language, Legal A
4 2 6 3 of 3 .044 99.7 95.3 Language, Legal A
 384

5 2 7 3 of 4 .247 98.8 95.3 Language, Legal A

19 (11) Hi SCA 1.0 2 2 6 of 11 1 of 1 .000 100 93.3 Legal, Economic
& 3 2 7 2 of 2 .000 100 91.2 Contacts, Legal A
all 4 2 7 3 of 3 .000 100 90.7 Contacts, SCA (country) *5
other 5 2 9 4 of 4 .000 93.8 94.8 Contacts, Legal A
Hi’s 6 2 9 5 of 5 .015 94.6 91.7 Contacts, Legal A
7 2 9 5 of 6 .121 95.3 90.2 Contacts, SCA (country) A
16 (8) Lo all Σ .125 x 8 2 4 5 of 8 1 of 1 .000 100 100 Language, Legal
generic x Q’aire 3 2 6 2 of 2 .000 95.4 98.4 Language, Legal A
=.36875 4 2 6 3 of 3 .019 99.6 95.9 Language, Legal A
5 1 7 3 of 4 .174 98.7 95.9 Language, Legal A
19 (11) Hi all Σ .091 x 11 2 2 6 of 11 1 of 1 .000 100 93.3 Legal, Economic
generic x Q’aire 3 2 7 2 of 2 .000 100 91.2 Contacts, Legal A
= 1.0 4 2 7 3 of 3 .000 100 90.7 Contacts, SCA (country) *5
5 2 9 4 of 4 .000 93.8 94.8 Contacts, Legal A
6 2 9 5 of 5 .015 94.6 91.7 Contacts, Legal A
7 2 9 6 of 6 .121 95.3 90.2 Contacts, SCA (country) A
* ≥ ± .70 level * ≤.05 level. These 2 matters determine support or not for Ho if ≥ 3 clusters (use variables only if 2 clusters)
1 2
*3 Eigenvalues ≥ 1.0
*5 Competition ranked 2nd but disregarded because its tolerance is < .19, indicating multicollinearity (Hair et al 1995, p. 127)
*7 Market risk ranked 2nd but disregarded because its tolerance is < .19, indicating multicollinearity (Hair et al 1995, p. 127)
xxx = invalid solution as last discriminant function fails the ≤.05 significance hurdle
Bold = optimal solution.
 385

Education: generic, market and objectives variables [20 (8) & 23 (11) active variables]

No of Variable No of No of No of No of significant *2 Significance % of variance % of Top 2 variables, Struct-

variab being clusters pairs of variables of worst included in countries based on ‘Variables ure
les explored bold variables failing functions significant cross- Entered / Removed’ Matrix
= best correlated Tolerance *4 functions *3 validated table Agrees
bivariately *1 =A
20 (8) Lo all 2 5 2 1 of 8 1 of 1 .000 100 99.5 Pdt consumptn,
Culture
generic 3 5 2 5 2 of 2 .000 98.6 99.5 Pdt consumptn, A
Culture
with 4 2 1 7 3 of 3 .000 99.0 93.2 Pdt consumptn, Risk A
objectives 5 2 0 7 4 of 4 .000 97.0 95.3 Pdt consumptn, Risk A
6 2 0 7 5 of 5 .001 96.7 92.7 Pdt consumptn, Risk A
7 2 0 7 5 of 6 .209 97.3 92.7 Pdt consumptn, Language A
23 (11) Hi all 2 5 2 6 of 11 1 of 1 .000 100 99.5 Pdt consumptn,
Culture
generic 3 5 2 5 2 of 2 .000 98.5 99.5 Pdt consumptn, A
Culture
with 4 2 1 6 3 of 3 .001 98.9 95.3 Pdt consumptn, Risk A
objectives 5 *5 2 1 7 4 of 4 .000 99.0 94.8 Pdt consumptn, Risk
6 *5 2 0 7 5 of 5 .001 98.1 94.8 Pdt consumptn, Risk A
7 *5 2 0 8 5 of 6 .085 94.8 95.3 Pdt consumptn, Risk A
*1 Correlation ≥ ± .70 level *2 Correlation ≤.05 level. These 2 matters determine support or not for Ho if ≥ 3 clusters
*3 Eigenvalues ≥ 1.0 *4 Correlation ≥.19 level *5 xxx = Invalid solution as has a single-country cluster
 386

Engines: generic and market variables [16 (8) & 19 (11) active variables]
No of Variable Weight of No of No of No of significant *2 Significance % of variance % of Top 2 variables, Struct-
variab being the column clusters pairs of Variables Functions of worst included in countries based on ‘Variables ure
les explored 1 variable: bold variables functions significant cross- Entered / Removed’ Matrix
2x = best Correlated functions *3 validated table Agrees
*1 =A
16 (8) Lo SCA .30 2 3 2 of 8 1 of 1 .000 100 % 100 % Language, Culture
3 1 7 2 of 2 .000 100 95.2 Language, Legal A
4 1 7 2 of 3 .337 97.6 97.6 Language, Legal A
Lo GBC .35 2 3 2 of 8 1 of 1 .000 100 100 Language, Culture
3 1 7 2 of 2 .000 100 95.2 Language, Legal A
4 1 7 2 of 3 .337 99.8 97.6 Language, Legal A
Lo FIC .13 2 3 2 1 of 1 .000 100 100 Language, Culture
3 2 6 2 of 2 .046 99.6 100 Language, Economic
4 1 7 2 of 3 .337 99.8 97.6 Language, Legal A
Lo KSS .30 3 Same as Lo SCA (above)
*4
Lo FIV .54 2 1 6 1 of 1 .000 100 92.9 Culture, Language
3 1 7 2 of 2 .000 100 90.5 Language, Legal
4 1 7 2 of 3 .281 99.0 95.2 Language, Legal A
*4
Lo RV .50 2 1 6 1 of 1 .000 100 92.9 Culture, Language
3 1 7 2 of 2 .000 100 95.2 Language, Legal A
4 1 7 2 of 3 .337 97.7 97.6 Language, Legal A
Lo MRV .50 3 Same as Lo RV (above)
Lo M & C .33 2 3 2 1 of 1 .000 100 100 Language, Culture
3 1 7 2 of 2 .000 100 95.2 Language, Legal A
4 1 7 2 of 3 .337 99.8 97.6 Language, Legal A
19 (11) Hi SCA 1.0 2 1 6 of 11 1 of 1 .000 100 88.1 Economic, Legal A
3 1 7 2 of 2 .000 100 83.3 Consumption, Legal A
& all 4 1 8 3 of 3 .001 100 83.3 Consumption, Competition
other 5 1 10 4 of 4 .000 100 81.0 Consumption, Legal
6 1 10 5 of 5 .045 96.7 78.6 Consumption, Legal
Hi’s 7 1 10 5 of 6 .067 97.8 78.6 Consumption, Contacts
16 (8) Lo all Σ .125 x 8 2 3 2 of 8 1 of 1 .000 100 100 Language, Culture
 387

generic x Q’aire 3 1 7 2 of 2 .000 100 95.2 Language, Legal A

=.36875 4 1 7 2 of 3 .337 99.8 97.6 Language, Legal A
19 (11) Hi all Σ .091 x 11 2 1 6 of 11 1 of 1 .000 100 88.1 Economic, Legal A
generic x Q’aire 3 1 7 2 of 2 .000 100 83.3 Consumption, Legal A
= 1.0 4 1 8 3 of 3 .001 100 83.3 Consumption, Competition
5 1 10 4 of 4 .000 100 81.0 Consumption, Legal
6 1 10 5 of 5 .045 96.7 78.6 Consumption, Legal
7 1 10 5 of 6 .067 97.8 78.6 Consumption, Contacts
* ≥ ± .70 level * ≤.05 level. These 2 matters determine support or not for Ho if ≥ 3 clusters (use variables only if 2 clusters)
1 2
*3 Eigenvalues ≥ 1.0
*4 Competition ranked 1st but disregarded because its tolerance is < .19, indicating multicollinearity (Hair et al 1995, p. 127)
xxx = invalid solution as last discriminant function fails the ≤.05 significance hurdle
Bold = optimal solution.
 388

SUMMARY – All Products

7 variables, Best Markets
Product No of No of pairs of No of significant *2 Significance of % of variance % cross-
clusters variables Variables: Functions: worst functions included in significant validated
correlated *1 functions *3
Coal 4 - 5 of 7 3 of 3 .002 96.4 % 94.1 %
Beef 3 1 6 2 of 2 .000 96.8 94.6
Wool 5 1 7 4 of 4 .024 92.6 93.0
Telephones 4 - 6 3 of 3 .006 97.7 95.7
Education 3 1 6 2 of 2 .000 96.8 94.3
Engines 4 - 7 3 of 3 .020 97.7 95.2

16 (8) variables, Low Organisation’s Markets (excludes Education with objectives)

Coal 3 3 6 of 8 2 of 2 .000 100 96.2
Beef 3 2 6 2 of 2 .000 100 99.0
Wool 4 1 6 3 of 3 .010 99.6 98.3
Telephones 4 2 7 3 of 3 .008 99.4 95.7
Education 3 2 6 2 of 2 .000 95.4 98.4
Engines 3 1 7 2 of 2 .000 100 95.2

19 (11) variables, High Organisation’s Markets (excludes Education with objectives)

Coal 5 0 10 of 11 4 of 4 .000 96.1 94.2
Beef 3 1 6 2 of 2 .000 100 93.2
Wool 5 1 8 4 of 4 .001 97.2 94.8
Telephones 3 0 9 2 of 2 .000 100 96.6
Education 5 2 9 4 of 4 .000 93.8 94.8
Engines 3 1 7 2 of 2 .000 100 83.3
*1 ≥ ± .70 level *2 ≤.05 level. These 2 matters determine support or not for Ho if ≥ 3 clusters (use variables only if 2 clusters) *3 Eigenvalues ≥ 1.0
 389

SUMMARY – By Product and Number of Variables

Product No of No of pairs of No of significant *2 Significance of % of variance % cross-

clusters variables Variables: Functions: worst functions included in significant validated
correlated *1 functions *3
Coal
7 variable 3 - 5 of 7 2 of 2 .000 100 % 98.0 %
16/8 var, Low 3 3 6 of 8 2 of 2 .000 100 96.2
19/11 var, Hi 5 0 10 of 11 4 of 4 .000 96.1 94.2
Beef
7 variable 3 1 6 of 7 2 of 2 .000 96.8 94.6
16/8 var, Low 3 2 6 of 8 2 of 2 .000 100 99.0
19/11 var, Hi 3 1 6 of 11 2 of 2 .000 100 93.2
Wool
7 variable 5 1 7 of 7 4 of 4 .025 92.6 93.0
16/8 var, Low 4 1 6 of 8 3 of 3 .010 99.6 98.3
19/11 var, Hi 5 1 8 of 11 4 of 4 .001 97.2 94.8
Telephones
7 variable 4 - 6 of 7 3 of 3 .006 97.7 95.7
16/8 var, Low 4 2 7 of 8 3 of 3 .008 99.4 95.7
19/11 var, Hi 3 0 9 of 11 2 of 2 .000 100 96.6
Education
7 variable 3 1 6 of 7 2 of 2 .000 96.8 94.3
16/8 var, Low 3 2 6 of 8 2 of 2 .000 95.4 98.4
19/11 var, Hi 5 2 9 of 11 4 of 4 .000 93.8 94.8
Engines
7 variable 4 - 7 of 7 3 of 3 .020 97.7 95.2
16/8 var, Low 3 1 7 of 8 2 of 2 .000 100 95.2
19/11 var, Hi 3 1 7 of 11 2 of 2 .000 100 83.3
*1 ≥ ± .70 level *2 ≤.05 level. These 2 matters determine support or not for Ho if ≥ 3 clusters (use variables only if 2 clusters) *3 Eigenvalues ≥ 1.0
 390

SUMMARY – By Product
Generic and market variables [16 (8) & 19 (11) active variables]
Product No of Variable Weight No of No of No of significant *2 Signific- % of % of Top 2 variables, Struct-
variab being of the clusters pairs of Variables Functions ance of variance countries based on ‘Variables ure
les explored column 1 variables worst included in cross- Entered / Removed’ Matrix
variable: correlated functions significant validated table Agrees
2x *1 functions *3 =A
Coal 16 (8) Lo SCA .30 3 3 6 of 8 2 of 2 .000 100 96.2 Language, Legal A
Lo GBC .35 3 3 6 2 of 2 .000 100 96.2 Language, Legal A
Lo FIC .13 3 4 6 2 of 2 .035 99.5 100 Language, Economic A
Lo KSS .30 3 Same As Lo SCA (above)
*4
Lo FIV .54 3 2 6 2 of 2 .000 100 94.2 Language, Legal
*4
Lo RV .50 3 2 6 2 of 2 .000 100 94.2 Language, Legal
Lo MRV .50 3 Same As Lo RV (above)
Lo M & C .33 3 3 6 2 of 2 .000 100 96.2 Language, Legal A
19 (11) Hi SCA 1.0 5 0 10 4 of 4 .000 96.1 94.2 Competition, Consmn gth
& all Hi
16 (8) Lo all g #1 3 3 6 of 8 2 of 2 .000 100 96.2 Language, Legal A
19 (11) Hi all g #2 5 0 10 4 of 4 .000 96.1 94.2 Competition, Consmn gth
Beef 16 (8) Lo SCA .30 3 2 6 of 8 2 of 2 .000 95.5 % 99.0 % Language, Legal A
Lo GBC .35 3 2 6 2 of 2 .000 95.5 99.0 Language, Legal A
Lo FIC .13 3 2 5 2 of 2 .030 99.9 100.0 Language, Legal A
Lo KSS .30 3 Same As Lo SCA (above)
Lo FIV .54 3 2 6 2 of 2 .000 100 97.1 Language, Legal A
Lo RV .50 3 2 6 2 of 2 .000 100 97.1 Language, Legal A
Lo MRV .50 3 Same As Lo RV (above)
Lo M & C .33 3 2 6 2 of 2 .000 95.5 99.0 Language, Legal A
19 (11) Hi SCA 1.0 3 1 6 of 11 2 of 2 .000 100 93.2 SCA (country), Legal A
& all Hi
16 (8) Lo all g #1 3 2 6 2 of 2 .000 100 99.0 Language, Legal A
19 (11) Hi all g #2 3 1 6 2 of 2 .000 100 93.2 SCA (country), Legal A
 391

Wool 16 (8) Lo SCA .30 4 1 6 3 of 3 .010 99.6 98.3 Language, Legal A

Lo GBC .35 4 1 6 3 of 3 .010 99.6 98.3 Language, Legal A
Lo FIC .13 2 4 5 1 of 1 .000 100 100 Language, Legal
Lo KSS .30 4 Same As Lo SCA (above)
Lo FIV .54 5 1 7 4 of 4 .003 98.4 98.3 Language, Legal A
Lo RV .50 5 1 7 4 of 4 .003 98.4 98.3 Language, Legal A
Lo MRV .50 5 Same As Lo RV (above)
Lo M & C .33 5 1 7 4 of 4 .009 98.8 96.5 Language, Legal A
19 (11) Hi SCA 1.0 5 1 8 of 11 4 of 4 .001 97.2 94.8 Contacts, Economic
& all Hi
16 (8) Lo all g #1 4 1 6 of 8 3 of 3 .010 99.6 98.3 Language, Legal A
19 (11) Hi all g #2 5 1 8 of 11 4 of 4 .001 97.2 94.8 Contacts, Economic
T ’phones 16 (8) Lo SCA .30 5 1 7 of 8 4 of 4 .036 98.3 97.4 Language, Legal A
Lo GBC .35 5 1 7 4 of 4 .036 98.3 97.4 Language, Legal A
Lo FIC .13 2 3 2 1 of 1 .000 100 100 Language, Culture
Lo KSS .30 5 Same As Lo SCA (above)
Lo FIV .54 4 2 7 3 of 3 .003 99.2 95.7 Language, Legal A
Lo RV .50 4 2 7 3 of 3 .008 99.4 95.7 Language, Legal A
Lo MRV .50 4 Same As Lo RV (above)
Lo M & C .33 5 1 7 4 of 4 .036 98.3 97.4 Language, Legal A
19 (11) Hi SCA 1.0 3 0 9 of 11 2 of 2 .000 100 96.6 Legal, Language
& all Hi
16 (8) Lo all g #1 4 2 7 of 8 3 of 3 .008 99.4 95.7 Language, Legal A
19 (11) Hi all g #2 3 0 9 of 11 2 of 2 .000 100 96.6 Legal, Language
Education 16 (8) Lo SCA .30 3 2 6 of 8 2 of 2 .000 95.4 98.4 Language, Legal A
Lo GBC .35 3 2 6 2 of 2 .000 95.4 98.4 Language, Legal A
Lo FIC .13 3 4 6 2 of 2 .044 99.9 100 Language, Legal A
Lo KSS .30 3 Same As Lo SCA (above)
Lo FIV .54 4 2 6 3 of 3 .017 99.6 96.9 Language, Legal A
Lo RV .50 4 2 6 3 of 3 .017 99.6 96.9 Language, Legal A
Lo MRV .50 4 Same As Lo RV (above)
Lo M & C .33 3 2 6 2 of 2 .000 95.4 98.4 Language, Legal A
19 (11) Hi SCA 1.0 5 2 9 of 11 4 of 4 .000 93.8 94.8 Contacts, Legal A
 392

& all Hi
16 (8) Lo all g #1 3 2 6 of 8 2 of 2 .000 95.4 98.4 Language, Legal A
19 (11) Hi all g #2 5 2 9 of 11 4 of 4 .000 93.8 94.8 Contacts, Legal A
20 (8) Lo & obj 3 5 5 of 8 2 of 2 .000 98.6 99.5 Pdt consumptn, Culture A
23 (11) Hi & obj 3 5 6 of 11 2 of 2 .000 98.5 99.5 Pdt consumptn, Culture A
Engines 16 (8) Lo SCA .30 3 1 7 of 8 2 of 2 .000 100 95.2 Language, Legal A
Lo GBC .35 3 1 7 2 of 2 .000 100 95.2 Language, Legal A
Lo FIC .13 3 2 6 2 of 2 .046 99.6 100 Language, Economic
Lo KSS .30 3 Same As Lo SCA (above)
Lo FIV .54 3 1 7 2 of 2 .000 100 90.5 Language, Legal
Lo RV .50 3 1 7 2 of 2 .000 100 95.2 Language, Legal A
Lo MRV .50 3 Same As Lo RV (above)
Lo M & C .33 3 1 7 2 of 2 .000 100 95.2 Language, Legal A
19 (11) Hi SCA 1.0 3 1 7 2 of 2 .000 100 83.3 Consumption, Legal A
& all Hi
16 (8) Lo all g #1 3 1 7 of 8 2 of 2 .000 100 95.2 Language, Legal A
19 (11) Hi all g #2 3 1 7 of 11 2 of 2 .000 100 83.3 Consumption, Legal A
*1 ≥ ± .70 level *2 ≤.05 level. These 2 matters determine support or not for Ho if ≥ 3 clusters (use variables only if 2 clusters) *3 Eigenvalues ≥ 1.0
* Competition ranked 1st but disregarded because its tolerance is < .19, indicating multicollinearity (Hair et al 1995, p. 127)
4

1# Σ .125 x 8 x Questionnaire values = .36875

2# Σ .091 x 11 x Questionnaire values = 1.0
Bold = optimal solution.
 393

Numeric Summary of Which Two Variables are Most Influential in Determining Results
(excludes Education with objectives)
Number of variables in category No. principally determined by Language & Legal, in that order, as assessed by:
Wilks’ Lambda Structure Matrix
Individual variables – Low 6 x 8 = 48 45 of 48 41 of 48
– High 6x1=6 0 *A of 6 0 *A of 6
Low all generic variables 6x1=6 6 of 6 6 of 6
*B
High all generic variables 6x1=6 0 of 6 0 *C of 6
*A *B *C
1 result is Legal and Language (same variables but in reverse order) Legal appears 4 times out of 12 possible Legal appears 6 times out of 12
possible
 394

SUMMARY By VARIABLE

No of Variable Weight of No of No of No of significant *2 Significance % of variance % of Top 2 variables, Struct-ure

Matrix
variab being the column clusters pairs of Variables Functions of worst included in countries based on ‘Variables Agrees
les explored 1 variable: bold variables functions significant cross- Entered / Removed’ =A
2x = best correlated *1 functions *3 validated table
16 (8) Lo SCA .30 3 3 6 of 8 2 of 2 .000 100 % 96.2 % Language, Legal A
3 2 6 of 8 2 of 2 .000 95.5 99.0 Language, Legal A
4 1 6 3 of 3 .010 99.6 98.3 Language, Legal A
5 1 7 4 of 4 .036 98.3 97.4 Language, Legal A
3 2 6 2 of 2 .000 95.4 98.4 Language, Legal A
3 1 7 2 of 2 .000 100 95.2 Language, Legal A
Lo GBC .35 3 3 6 2 of 2 .000 100 96.2 Language, Legal A
3 2 6 2 of 2 .000 95.5 99.0 Language, Legal A
4 1 6 3 of 3 .010 99.6 98.3 Language, Legal A
5 1 7 4 of 4 .036 98.3 97.4 Language, Legal A
3 2 6 2 of 2 .000 95.4 98.4 Language, Legal A
3 1 7 2 of 2 .000 100 95.2 Language, Legal A
Lo FIC .13 3 4 6 2 of 2 .035 99.5 100 Language, Economic A
3 2 5 2 of 2 .030 99.9 100.0 Language, Legal A
2 4 5 1 of 1 .000 100 100 Language, Legal
2 3 2 1 of 1 .000 100 100 Language, Culture
3 4 6 2 of 2 .044 99.9 100 Language, Legal A
3 2 6 2 of 2 .046 99.6 100 Language, Economic
Lo KSS .30 3 Same as Lo SCA (above)
*4
Lo FIV .54 3 2 6 2 of 2 .000 100 94.2 Language, Legal
3 2 6 2 of 2 .000 100 97.1 Language, Legal A
5 1 7 4 of 4 .003 98.4 98.3 Language, Legal A
4 2 7 3 of 3 .003 99.2 95.7 Language, Legal A
4 2 6 3 of 3 .017 99.6 96.9 Language, Legal A
3 1 7 2 of 2 .000 100 90.5 Language, Legal
*4
Lo RV .50 3 2 6 2 of 2 .000 100 94.2 Language, Legal
3 2 6 2 of 2 .000 100 97.1 Language, Legal A
 395

5 1 7 4 of 4 .003 98.4 98.3 Language, Legal A

4 2 7 3 of 3 .008 99.4 95.7 Language, Legal A
4 2 6 3 of 3 .017 99.6 96.9 Language, Legal A
3 1 7 2 of 2 .000 100 95.2 Language, Legal A
Lo MRV .50 3 Same as Lo RV (above)
Lo M & C .33 3 3 6 2 of 2 .000 100 96.2 Language, Legal A
3 2 6 2 of 2 .000 95.5 99.0 Language, Legal A
5 1 7 4 of 4 .009 98.8 96.5 Language, Legal A
5 1 7 4 of 4 .036 98.3 97.4 Language, Legal A
3 2 6 2 of 2 .000 95.4 98.4 Language, Legal A
3 1 7 2 of 2 .000 100 95.2 Language, Legal A
19 (11) Hi SCA 1.00 5 0 10 4 of 4 .000 96.1 94.2 Competition, Consmn gth
3 1 6 of 11 2 of 2 .000 100 93.2 SCA (country), Legal A
other 5 1 8 4 of 4 .001 97.2 94.8 Contacts, Economic
3 0 9 of 11 2 of 2 .000 100 96.6 Legal, Language
5 2 9 4 of 4 .000 93.8 94.8 Contacts, Legal A
3 1 7 2 of 2 .000 100 83.3 Consumption, Legal A
Hi GBC 1.00 5 Same as Hi SCA (above)
16 (8) Lo all Σ .125 x 8 3 3 6 2 of 2 .000 100 96.2 Language, Legal A
generic x Q’aire 3 2 6 2 of 2 .000 100 99.0 Language, Legal A
=.36875 4 1 6 3 of 3 .010 99.6 98.3 Language, Legal A
4 2 7 3 of 3 .008 99.4 95.7 Language, Legal A
3 2 6 2 of 2 .000 95.4 98.4 Language, Legal A
3 1 7 2 of 2 .000 100 95.2 Language, Legal A
19 (11) Hi all Σ .091 x 11 4 1 7 3 of 3 .000 100 96.1 Competition, Culture
generic x Q’aire 3 1 6 2 of 2 .000 100 93.2 SCA (country), Legal A
= 1.0 5 1 8 4 of 4 .001 97.2 94.8 Contacts, Economic
3 0 9 2 of 2 .000 100 96.6 Legal, Language
5 2 9 4 of 4 .000 93.8 94.8 Contacts, Legal A
3 1 7 2 of 2 .000 100 83.3 Consumption, Legal A
*1 ≥ ± .70 level *2 ≤.05 level. These 2 matters determine support or not for Ho if ≥ 3 clusters (use variables only if 2 clusters) *3 Eigenvalues ≥ 1.0
*4 Competition ranked 1st but disregarded because its tolerance is < .19, indicating multicollinearity (Hair et al 1995, p. 127)
Bold = optimal solution.
 396

Numbers of Clusters Determining Solutions

Across the 6 Products *5

No of Markets Individual All Variables

Clusters only Variables
Low High Low High
2 - 2 - - -
3 3 29 3 4 3
4 3 9 - 2 -
5 - 8 3 - 3
*1
Totals 6 48 6 *2 6 *3 6 *4
1
* = 8 variables tested individually for each of 6 products = 6 x 8 = 48
2
* = This is for each of the 8 variables as results were identical for each variable’s Hi solution
3
* = All 8 (16) variables entered simultaneously for each of the 6 products
4
* = All 11 (19) variables entered simultaneously for each of the 6 products
5
* = excludes Education with objectives
 397

APPENDIX I

COUNTRIES IN EACH CLUSTER, BY PRODUCT

AND LEVEL OF ORGANISATIONAL CAPABILITY

Coking coal

Markets Only Low Capability High Capability

Organisation Organisation
Country Cl. Country Cl. Country Cl.
No No No

Albania 1 Albania 1 Albania 1

Algeria 1 Algeria 1 Algeria 1
Brazil 1 Bulgaria 1 Brazil 1
Bulgaria 1 China 1 Egypt 1
Colombia 1 Colombia 1 India 1
Egypt 1 Iran 1 Kazakhstan 1
India 1 Kazakhstan 1 Korea, N 1
Iran 1 Korea, N 1 Macedonia 1
Kazakhstan 1 Macedonia 1 Mexico 1
Korea, N 1 Mexico 1 Pakistan 1
Macedonia 1 Peru 1 Peru 1
Mexico 1 Romania 1 Romania 1
Pakistan 1 Russia 1 Russia 1
Peru 1 Slovakia 1 Slovenia 1
Romania 1 Slovenia 1 Tajikistan 1
Russia 1 Tajikistan 1 Tanzania 1
Slovakia 1 Turkey 1 Ukraine 1
Slovenia 1 Ukraine 1 Yugoslavia 1
Tajikistan 1 Yugoslavia 1 Zambia 1
Tanzania 1 Argentina 2 Zimbabwe 1
Turkey 1 Australia 2 Argentina 2
Ukraine 1 Brazil 2 Chile 2
Yugoslavia 1 Canada 2 Colombia 2
Zambia 1 Egypt 2 Japan 2
Zimbabwe 1 India 2 Korea, S 2
Argentina 2 Ireland 2 Portugal 2
Australia 2 Korea, S 2 Taiwan 2
Austria 2 New Zealand 2 Australia 3
Belgium 2 Pakistan 2 Austria 3
Canada 2 South Africa 2 Belgium 3
Czech Rep 2 Tanzania 2 Canada 3
Finland 2 United Kingdom 2 Finland 3
France 2 United States 2 France 3
Germany 2 Zambia 2 Germany 3
Hungary 2 Zimbabwe 2 Ireland 3
Iceland 2 Austria 3 Netherlands 3
Ireland 2 Belgium 3 Nee Zealand 3
Italy 2 Chile 3 Spain 3
 398

(Coal cont’d)
Netherlands 2 Czech Rep 3 Sweden 3
New Zealand 2 Finland 3 United Kingdom 3
Poland 2 France 3 United States 3
South Africa 2 Germany 3 Bulgaria 4
Spain 2 Hungary 3 Czech Rep 4
Sweden 2 Iceland 3 Hungary 4
United Kingdom 2 Italy 3 Iceland 4
United States 2 Japan 3 Iran 4
Chile 3 Netherlands 3 Italy 4
Japan 3 Poland 3 Poland 4
Korea, S 3 Portugal 3 Slovakia 4
Portugal 3 Spain 3 South Africa 4
Taiwan 3 Sweden 3 Turkey 4
Taiwan 3
China - China -
Source: SPSS Cluster Membership tables, K-means CA

Bold = markets indicated by screening as the optimal ones

Xxxxx = an exception because of prior exclusion as an outlier market.
Manual assessment desirable to check its true attractiveness

Beef

Markets Only Low Capability High Capability

Organisation Organisation
Country Cl. Country Cl. Country Cl.
No No No

Afghanistan 1 Afghanistan 1 Afghanistan 1

Albania 1 Albania 1 Algeria 1
Algeria 1 Algeria 1 Angola 1
Angola 1 Angola 1 Armenia 1
Armenia 1 Armenia 1 Azerbaijan 1
Azerbaijan 1 Azerbaijan 1 Belarus 1
Bangladesh 1 Belarus 1 Belize 1
Belarus 1 Benin 1 Benin 1
Benin 1 Bolivia 1 Bhutan 1
Bosnia & Herz 1 Bosnia & Herz 1 Bolivia 1
Burkina Faso 1 Bulgaria 1 Bulgaria 1
Burma 1 Burkina Faso 1 Burkina Faso 1
Burundi 1 Burma 1 Burundi 1
Cambodia 1 Burundi 1 Cambodia 1
Cameroon 1 Cambodia 1 Cape Verde 1
Cape Verde 1 Cape Verde 1 Central Af Rep 1
Central Af Rep 1 Central Af Rep 1 Chad 1
Chad 1 Chad 1 Comoros 1
Comoros 1 China 1 Congo, Dem R 1
Congo, Dem R 1 Comoros 1 Congo, Rep 1
 399

(Beef cont’d)
Congo, Rep 1 Congo, Dem R 1 Croatia 1
Cote d’Ivoire 1 Congo, Rep 1 Cuba 1
Croatia 1 Cote d’Ivoire 1 Ecuador 1
Cuba 1 Croatia 1 Egypt 1
Djibouti 1 Cuba 1 El Salvador 1
Ecuador 1 Djibouti 1 Eritrea 1
Egypt 1 Ecuador 1 Ethiopia 1
Equatorial Guin 1 El Salvador 1 Georgia 1
Eritrea 1 Equatorial Guin 1 Ghana 1
Ethiopia 1 Eritrea 1 Guatemala 1
Gabon 1 Estonia 1 Guinea-Bissau 1
Gambia, The 1 Gabon 1 Guyana 1
Gaza Strip 1 Georgia 1 Indonesia 1
Georgia 1 Guatemala 1 Iran 1
Ghana 1 Guinea 1 Jordan 1
Guatemala 1 Guinea-Bissau 1 Kazakhstan 1
Guinea 1 Haiti 1 Kenya 1
Guinea-Bissau 1 Honduras 1 Kyrgyzstan 1
Haiti 1 Iran 1 Laos 1
Honduras 1 Iraq 1 Latvia 1
Iran 1 Kazakhstan 1 Lebanon 1
Iraq 1 Korea, N 1 Lesotho 1
Kazakhstan 1 Kyrgyzstan 1 Libya 1
Korea, N 1 Latvia 1 Madagascar 1
Kyrgyzstan 1 Macedonia 1 Malawi 1
Laos 1 Madagascar 1 Mali 1
Latvia 1 Mali 1 Mauritania 1
Lesotho 1 Mauritania 1 Mexico 1
Liberia 1 Mexico 1 Moldova 1
Libya 1 Moldova 1 Morocco 1
Macedonia 1 Mongolia 1 Namibia 1
Madagascar 1 Morocco 1 Nepal 1
Malawi 1 Mozambique 1 Niger 1
Mali 1 Niger 1 Nigeria 1
Mauritania 1 Paraguay 1 Oman 1
Mexico 1 Peru 1 Pakistan 1
Moldova 1 Reunion 1 Papua New G 1
Mongolia 1 Romania 1 Peru 1
Mozambique 1 Russia 1 Philippines 1
Nepal 1 Sao Tome 1 Qatar 1
Nicaragua 1 Saudi Arabia 1 Reunion 1
Niger 1 Senegal 1 Romania 1
Nigeria 1 Slovakia 1 Russia 1
Pakistan 1 Slovenia 1 Rwanda 1
Peru 1 Sudan 1 Saudi Arabia 1
Romania 1 Syria 1 Senegal 1
Russia 1 Tajikistan 1 Sierra Leone 1
Rwanda 1 Togo 1 Slovakia 1
Senegal 1 Tunisia 1 Slovenia 1
 400

(Beef cont’d)
Sierra Leone 1 Turkey 1 Somalia 1
Slovenia 1 Turkmenistan 1 Sri Lanka 1
Somalia 1 Ukraine 1 Swaziland 1
Sudan 1 Uruguay 1 Syria 1
Swaziland 1 Uzbekistan 1 Tanzania 1
Syria 1 Venezuela 1 Togo 1
Tajikistan 1 Yemen 1 Turkey 1
Tanzania 1 Yugoslavia 1 Turkmenistan 1
Togo 1 American 2 Ukraine 1
Turkey 1 Antigua 2 United Arab E 1
Turkmenistan 1 Argentina 2 Uruguay 1
Uganda 1 Australia 2 Uzbekistan 1
Ukraine 1 Bahamas, The 2 West Bank 1
Uzbekistan 1 Bahrain 2 Yugoslavia 1
Venezuela 1 Bangladesh 2 Zimbabwe 1
Vietnam 1 Barbados 2 Albania 2
West Bank 1 Belize 2 Bangladesh 2
Yemen 1 Bermuda 2 Bosnia & Herz 2
Yugoslavia 1 Botswana 2 Botswana 2
Zambia 1 Brazil 2 Burma 2
Zimbabwe 1 British Virgin Is 2 Cameroon 2
American 2 Brunei 2 China 2
Antigua 2 Cameroon 2 Cote d’Ivoire 2
Australia 2 Canada 2 Djibouti 2
Austria 2 Cook Is 2 Dominica 2
Barbados 2 Costa Rica 2 Equatorial Guin 2
Belgium 2 Cyprus 2 Estonia 2
Bermuda 2 Dominica 2 Gabon 2
British Virgin I 2 Egypt 2 Gambia, The 2
Brunei 2 Ethiopia 2 Gaza Strip 2
Canada 2 Fiji 2 Greece 2
Denmark 2 French P 2 Grenada 2
Faroe Is 2 Gambia, The 2 Guam 2
Finland 2 Gaza Strip 2 Guinea 2
France 2 Ghana 2 Haiti 2
Germany 2 Greece 2 Honduras 2
Greenland 2 Grenada 2 Iraq 2
Hong Kong 2 Guam 2 Korea, N 2
Iceland 2 Guyana 2 Liberia 2
Ireland 2 Hong Kong 2 Lithuania 2
Israel 2 India 2 Macedonia 2
Italy 2 Indonesia 2 Malaysia 2
Japan 2 Ireland 2 Maldives 2
Korea, S 2 Israel 2 Micronesia 2
Kuwait 2 Jamaica 2 Mongolia 2
Malta 2 Jordan 2 Mozambique 2
Netherlands 2 Kenya 2 New Caledonia 2
New Zealand 2 Korea, S 2 Nicaragua 2
 401

(Beef cont’d)
Norway 2 Kuwait 2 Paraguay 2
Singapore 2 Laos 2 Sao Tome 2
Sweden 2 Lebanon 2 Seychelles 2
Switzerland 2 Lesotho 2 South Africa 2
United Kingdom 2 Liberia 2 Sudan 2
Untied States 2 Libya 2 Tajikistan 2
Virgin I 2 Macau 2 Tonga 2
Argentina 3 Malawi 2 Trinidad 2
Bahamas, The 3 Malaysia 2 Tunisia 2
Bahrain 3 Maldives 2 Uganda 2
Belize 3 Malta 2 Vanuatu 2
Bhutan 3 Mauritius 2 Venezuela 2
Bolivia 3 Micronesia 2 Vietnam 2
Botswana 3 Montserrat 2 Wallis & Fut 2
Brazil 3 Namibia 2 Yemen 2
Bulgaria 3 Nepal 2 Zambia 2
Chile 3 Netherlands An 2 American 3
China 3 New Caledonia 2 Antigua 3
Colombia 3 New Zealand 2 Argentina 3
Cook Is 3 Nicaragua 2 Australia 3
Costa Rica 3 Nigeria 2 Austria 3
Cyprus 3 Niue 2 Bahamas, The 3
Czech Rep 3 Oman 2 Bahrain 3
Dominica 3 Pakistan 2 Barbados 3
Dominican Rep 3 Panama 2 Belgium 3
El Salvador 3 Papua New G 2 Bermuda 3
Estonia 3 Philippines 2 Brazil 3
Fiji 3 Puerto Rico 2 British Virgin I 3
French Guiana 3 Qatar 2 Brunei 3
French P 3 Rwanda 2 Canada 3
Greece 3 Saint Kitts 2 Chile 3
Grenada 3 Saint Lucia 2 Colombia 3
Guadeloupe 3 Saint Vincent & 2 Cook Is 3
Guam 3 Samoa 2 Costa Rica 3
Guyana 3 Seychelles 2 Cyprus 3
Hungary 3 Sierra Leone 2 Czech Rep 3
India 3 Singapore 2 Denmark 3
Indonesia 3 Somalia 2 Dominican Re 3
Jamaica 3 South Africa 2 Faroe Is 3
Jordan 3 Sri Lanka 2 Fiji 3
Kenya 3 Suriname 2 Finland 3
Lebanon 3 Swaziland 2 France 3
Lithuania 3 Tanzania 2 French Guiana 3
Macau 3 Thailand 2 French P 3
Malaysia 3 Tonga 2 Germany 3
Maldives 3 Trinidad 2 Greenland 3
Martinique 3 Uganda 2 Guadeloupe 3
Mauritius 3 United Arab E 2 Hong Kong 3
 402

(Beef cont’d)
Micronesia 3 United Kingdom 2 Hungary 3
Montserrat 3 Untied States 2 Iceland 3
Morocco 3 Vanuatu 2 India 3
Namibia 3 Vietnam 2 Ireland 3
Netherlands An 3 Virgin I 2 Israel 3
New Caledonia 3 West Bank 2 Italy 3
Niue 3 Zambia 2 Jamaica 3
Oman 3 Zimbabwe 2 Japan 3
Panama 3 Austria 3 Korea, S 3
Papua New G 3 Belgium 3 Kuwait 3
Paraguay 3 Bhutan 3 Macau 3
Philippines 3 Chile 3 Malta 3
Poland 3 Colombia 3 Martinique 3
Portugal 3 Czech Rep 3 Mauritius 3
Puerto Rico 3 Denmark 3 Montserrat 3
Qatar 3 Dominican Re 3 Netherlands 3
Reunion 3 Faroe Is 3 Netherlands A 3
Saint Kitts 3 Finland 3 New Zealand 3
Saint Lucia 3 France 3 Niue 3
Saint Pierre & 3 French Guiana 3 Norway 3
Saint Vincent & 3 Germany 3 Panama 3
Samoa 3 Greenland 3 Poland 3
Sao Tome 3 Guadeloupe 3 Portugal 3
Saudi Arabia 3 Hungary 3 Puerto Rico 3
Seychelles 3 Iceland 3 Saint Kitts 3
Slovakia 3 Italy 3 Saint Lucia 3
Solomon 3 Japan 3 Saint Pierre &I 3
South Africa 3 Lithuania 3 Saint Vincent & 3
Spain 3 Martinique 3 Samoa 3
Sri Lanka 3 Netherlands 3 Singapore 3
Suriname 3 Norway 3 Solomon 3
Taiwan 3 Poland 3 Spain 3
Thailand 3 Portugal 3 Suriname 3
Tonga 3 Saint Pierre &I 3 Sweden 3
Trinidad 3 Solomon 3 Switzerland 3
Tunisia 3 Spain 3 Taiwan 3
United Arab E 3 Sweden 3 Thailand 3
Uruguay 3 Switzerland 3 United Kingdom 3
Vanuatu 3 Taiwan 3 Untied States 3
Wallis & Fut 3 Wallis & Fut 3 Virgin I 3
Source: SPSS Cluster Membership tables, K-means CA

Bold = markets indicated by screening as the optimal ones

Xxxxx = an exception because of prior exclusion as an outlier market.
Manual assessment desirable to check its true attractiveness
 403

Wool

Markets Only Low Capability High Capability

Organisation Organisation
Country Cl. Country Cl. Country Cl.
No No No

Afghanistan 1 Afghanistan 1 Afghanistan 1

Albania 1 Albania 1 Algeria 1
Algeria 1 Algeria 1 Bolivia 1
Armenia 1 Armenia 1 Bosnia & Herz 1
Azerbaijan 1 Azerbaijan 1 Bulgaria 1
Bangladesh 1 Belarus 1 Burma 1
Belarus 1 Bhutan 1 Eritrea 1
Bosnia & Herz 1 Bolivia 1 Estonia 1
Burma 1 Bosnia & Herz 1 Ethiopia 1
Croatia 1 Bulgaria 1 Iran 1
Ecuador 1 Burma 1 Macedonia 1
Egypt 1 China 1 Moldova 1
Eritrea 1 Colombia 1 Poland 1
Ethiopia 1 Croatia 1 Russia 1
Gaza Strip 1 Ecuador 1 Slovenia 1
Georgia 1 Eritrea 1 Syria 1
Guatemala 1 Estonia 1 Albania 2
Iran 1 Georgia 1 Azerbaijan 2
Iraq 1 Guatemala 1 Bangladesh 2
Kazakhstan 1 Iran 1 Belarus 2
Korea, N 1 Iraq 1 Bhutan 2
Kyrgyzstan 1 Kazakhstan 1 Croatia 2
Latvia 1 Korea, N 1 Ecuador 2
Lesotho 1 Kyrgyzstan 1 Egypt 2
Libya 1 Latvia 1 Gaza Strip 2
Macedonia 1 Lithuania 1 Georgia 2
Mali 1 Macedonia 1 Guatemala 2
Mexico 1 Mali 1 Hungary 2
Moldova 1 Mexico 1 Kazakhstan 2
Mongolia 1 Moldova 1 Kyrgyzstan 2
Morocco 1 Mongolia 1 Lebanon 2
Nepal 1 Morocco 1 Lesotho 2
Nigeria 1 Paraguay 1 Libya 2
Pakistan 1 Peru 1 Mali 2
Peru 1 Romania 1 Mexico 2
Romania 1 Russia 1 Mongolia 2
Russia 1 Saudi Arabia 1 Nepal 2
Slovenia 1 Slovakia 1 Nigeria 2
Sudan 1 Slovenia 1 Romania 2
Syria 1 Sudan 1 Tajikistan 2
Tajikistan 1 Syria 1 Turkey 2
Tanzania 1 Tajikistan 1 Turkmenistan 2
Turkey 1 Tunisia 1 Ukraine 2
Turkmenistan 1 Turkey 1 Uzbekistan 2
 404

(Wool cont’d)
Ukraine 1 Turkmenistan 1 Yemen 2
Uzbekistan 1 Ukraine 1 Zimbabwe 2
Venezuela 1 Uruguay 1 Argentina 3
Vietnam 1 Uzbekistan 1 Botswana 3
West Bank 1 Venezuela 1 Chile 3
Yemen 1 Yemen 1 Colombia 3
Yugoslavia 1 Yugoslavia 1 Cyprus 3
Zimbabwe 1 Argentina 2 Greece 3
Argentina 2 Bangladesh 2 Kuwait 3
Bhutan 2 Botswana 2 Macau 3
Bolivia 2 Brazil 2 Malaysia 3
Botswana 2 Cyprus 2 Mauritius 3
Brazil 2 Egypt 2 Namibia 3
Bulgaria 2 Ethiopia 2 Portugal 3
Chile 2 Gaza Strip 2 Saudi Arabia 3
Colombia 2 Greece 2 South Africa 3
Cyprus 2 India 2 Spain 3
Czech Rep 2 Indonesia 2 Thailand 3
Estonia 2 Jamaica 2 Tunisia 3
Greece 2 Jordan 2 Uruguay 3
Hungary 2 Kenya 2 Armenia 4
India 2 Korea, S 2 Brazil 4
Indonesia 2 Kuwait 2 China 4
Israel 2 Lebanon 2 India 4
Jamaica 2 Lesotho 2 Indonesia 4
Jordan 2 Libya 2 Iraq 4
Kenya 2 Macau 2 Jamaica 4
Korea, S 2 Malaysia 2 Jordan 4
Kuwait 2 Mauritius 2 Kenya 4
Lebanon 2 Namibia 2 Korea, N 4
Lithuania 2 Nepal 2 Latvia 4
Macau 2 Nigeria 2 Lithuania 4
Malaysia 2 Pakistan 2 Morocco 4
Malta 2 South Africa 2 Pakistan 4
Mauritius 2 Sri Lanka 2 Paraguay 4
Namibia 2 Tanzania 2 Peru 4
Paraguay 2 Thailand 2 Slovakia 4
Poland 2 Vietnam 2 Sri Lanka 4
Portugal 2 West Bank 2 Sudan 4
Saudi Arabia 2 Zimbabwe 2 Tanzania 4
Slovakia 2 Australia 3 Venezuela 4
South Africa 2 Canada 3 Vietnam 4
Spain 2 Hong Kong 3 West Bank 4
Sri Lanka 2 Ireland 3 Yugoslavia 4
Thailand 2 Israel 3 Australia 5
Tunisia 2 Malta 3 Austria 5
Uruguay 2 New Zealand 3 Belgium 5
Australia 3 Singapore 3 Canada 5
Austria 3 United Kingdom 3 Czech Rep 5
 405

(Wool cont’d)
Belgium 3 United States 3 Denmark 5
Canada 3 Austria 4 Faroe Is 5
Denmark 3 Belgium 4 Finland 5
Faroe Is 3 Chile 4 France 5
Finland 3 Czech Rep 4 Germany 5
France 3 Denmark 4 Greenland 5
Germany 3 Faroe Is 4 Hong Kong 5
Greenland 3 Finland 4 Iceland 5
Hong Kong 3 France 4 Ireland 5
Iceland 3 Germany 4 Israel 5
Ireland 3 Greenland 4 Italy 5
Italy 3 Hungary 4 Japan 5
Japan 3 Iceland 4 Korea, S 5
Netherlands 3 Italy 4 Malta 5
New Zealand 3 Japan 4 Netherlands 5
Norway 3 Netherlands 4 New Zealand 5
Singapore 3 Norway 4 Norway 5
Sweden 3 Poland 4 Singapore 5
Switzerland 3 Portugal 4 Sweden 5
United Kingdom 3 Spain 4 Switzerland 5
United States 3 Sweden 4 United Kingdom 5
Switzerland 4 United States 5
China -
Source: SPSS Cluster Membership tables, K-means CA

Bold = markets indicated by screening as the optimal ones

Xxxxx = an exception because of prior exclusion as an outlier market.
Manual assessment desirable to check its true attractiveness

Telephone sets

Markets Only Low Capability High Capability

Organisation Organisation
Country Cl. Country Cl. Country Cl.
No No No

Algeria 1 Algeria 1 Algeria 1

Azerbaijan 1 Azerbaijan 1 Azerbaijan 1
Bangladesh 1 Belarus 1 Bangladesh 1
Belarus 1 Bolivia 1 Belarus 1
Cameroon 1 Bulgaria 1 Bulgaria 1
Central Af Rep 1 Central Af Rep 1 Cameroon 1
Chad 1 Chad 1 Central Af Rep 1
Congo, Rep 1 China 1 Chad 1
Croatia 1 Colombia 1 Congo, Rep 1
Ecuador 1 Congo, Rep 1 Croatia 1
Egypt 1 Croatia 1 Ecuador 1
Ethiopia 1 Ecuador 1 Egypt 1
 406

(T’phones cont’d)
Guatemala 1 El Salvador 1 Estonia 1
Honduras 1 Estonia 1 Ethiopia 1
Iran 1 French Guiana 1 Guatemala 1
Kyrgyzstan 1 Guadeloupe 1 Honduras 1
Latvia 1 Guatemala 1 Iran 1
Macedonia 1 Honduras 1 Kyrgyzstan 1
Madagascar 1 Iran 1 Latvia 1
Malawi 1 Kyrgyzstan 1 Lithuania 1
Mexico 1 Latvia 1 Macedonia 1
Moldova 1 Lithuania 1 Madagascar 1
Nicaragua 1 Macedonia 1 Malawi 1
Pakistan 1 Madagascar 1 Mexico 1
Peru 1 Martinique 1 Moldova 1
Romania 1 Mexico 1 Morocco 1
Russia 1 Moldova 1 Nicaragua 1
Slovenia 1 Morocco 1 Pakistan 1
Sudan 1 Paraguay 1 Peru 1
Syria 1 Peru 1 Romania 1
Turkey 1 Reunion 1 Russia 1
Ukraine 1 Romania 1 Slovakia 1
Venezuela 1 Russia 1 Slovenia 1
Yugoslavia 1 Saudi Arabia 1 Sudan 1
Zimbabwe 1 Slovakia 1 Syria 1
Argentina 2 Slovenia 1 Turkey 1
Belize 2 Sudan 1 Ukraine 1
Bolivia 2 Syria 1 Venezuela 1
Brazil 2 Tunisia 1 Yugoslavia 1
Bulgaria 2 Turkey 1 Zimbabwe 1
Colombia 2 Ukraine 1 Argentina 2
Costa Rica 2 Uruguay 1 Bahrain 2
Dominica 2 Venezuela 1 Barbados 2
El Salvador 2 Yugoslavia 1 Belgium-Lux 2
Estonia 2 Argentina 2 Belize 2
Fiji 2 Bahrain 2 Bolivia 2
French Guiana 2 Bangladesh 2 Brazil 2
Greece 2 Belize 2 Chile 2
Grenada 2 Brazil 2 China 2
Guadeloupe 2 Cameroon 2 Colombia 2
Hungary 2 Costa Rica 2 Costa Rica 2
India 2 Cyprus 2 Dominica 2
Indonesia 2 Dominica 2 El Salvador 2
Jamaica 2 Egypt 2 Fiji 2
Jordan 2 Ethiopia 2 France & Mon 2
Kenya 2 Fiji 2 French Guiana 2
Lithuania 2 Greece 2 Greece 2
Martinique 2 Grenada 2 Grenada 2
Mauritius 2 India 2 Guadeloupe 2
Morocco 2 Indonesia 2 India 2
Oman 2 Jamaica 2 Indonesia 2
 407

(T’phones cont’d)
Panama 2 Jordan 2 Israel 2
Paraguay 2 Kenya 2 Jamaica 2
Philippines 2 Korea, S 2 Japan 2
Poland 2 Kuwait 2 Jordan 2
Reunion 2 Macau 2 Kenya 2
Saint Kitts 2 Malawi 2 Korea, S 2
Saint Lucia 2 Malaysia 2 Kuwait 2
Saint Vincent &2 Mauritius 2 Macau 2
Saudi Arabia 2 Nicaragua 2 Malaysia 2
Seychelles 2 Oman 2 Malta 2
Slovakia 2 Pakistan 2 Martinique 2
South Africa 2 Panama 2 Mauritius 2
Spain 2 Philippines 2 Oman 2
Sri Lanka 2 Saint Kitts 2 Panama 2
Suriname 2 Saint Lucia 2 Paraguay 2
Thailand 2 Saint Vincent &2 Philippines 2
Trinidad 2 Seychelles 2 Portugal 2
Tunisia 2 South Africa 2 Reunion 2
Uruguay 2 Sri Lanka 2 Saint Kitts 2
Australia 3 Suriname 2 Saint Lucia 2
Austria 3 Thailand 2 Saint Vincent & 2
Barbados 3 Trinidad 2 Saudi Arabia 2
Belgium-Lux 3 Zimbabwe 2 Seychelles 2
Canada ( 3 Australia 3 Singapore 2
Czech Rep 3 Barbados 3 South Africa 2
Denmark 3 Canada ( 3 Spain 2
Finland 3 Hong Kong 3 Sri Lanka 2
France & Mon 3 Ireland 3 Suriname 2
Germany 3 Israel 3 Thailand 2
Greenland 3 Malta 3 Trinidad 2
Iceland 3 New Zealand 3 Tunisia 2
Ireland 3 Singapore 3 Uruguay 2
Israel 3 United Kingdom 3 Australia 3
Italy 3 Untied States 3 Austria 3
Netherlands 3 Austria 4 Canada 3
New Zealand 3 Belgium-Lux 4 Cyprus 3
Norway 3 Chile 4 Czech Rep 3
Sweden 3 Czech Rep 4 Denmark 3
Switzerland 3 Denmark 4 Faroe Is 3
United Kingdom 3 Faroe Is 4 Finland 3
United States 3 Finland 4 Germany 3
Bahrain 4 France & Mon 4 Greenland 3
Chile 4 Germany 4 Hong Kong 3
China 4 Greenland 4 Hungary 3
Cyprus 4 Hungary 4 Iceland 3
Faroe Is 4 Iceland 4 Ireland 3
Hong Kong 4 Italy 4 Italy 3
Japan 4 Japan 4 Netherlands 3
 408

(T’phones cont’d)
Korea, S 4 Netherlands 4 New Zealand 3
Kuwait 4 Norway 4 Norway 3
Macau 4 Poland 4 Poland 3
Malaysia 4 Portugal 4 Sweden 3
Malta 4 Spain 4 Switzerland 3
Portugal 4 Sweden 4 United Kingdom 3
Singapore 4 Switzerland 4 United States 3
Source: SPSS Cluster Membership tables, K-means CA

Bold = markets indicated by screening as the optimal ones

Xxxxx = an exception because of prior exclusion as an outlier market.
Manual assessment desirable to check its true attractiveness

Education services

Markets Only Low Capability High Capability

Organisation Organisation
Country Cl. Country Cl. Country Cl.
No No No

Afghanistan 1 Afghanistan 1 Afghanistan 1

Albania 1 Albania 1 Angola 1
Algeria 1 Algeria 1 Armenia 1
Angola 1 Angola 1 Bangladesh 1
Armenia 1 Armenia 1 Bhutan 1
Azerbaijan 1 Azerbaijan 1 Bosnia & Herz 1
Bangladesh 1 Belarus 1 Burkina Faso 1
Belarus 1 Benin 1 Burma 1
Benin 1 Bhutan 1 Burundi 1
Bosnia & Herz 1 Bolivia 1 Cameroon 1
Burkina Faso 1 Bosnia & Herz 1 Comoros 1
Burma 1 Bulgaria 1 Congo, Dem R 1
Burundi 1 Burkina Faso 1 Cote d’Ivoire 1
Cambodia 1 Burma 1 Croatia 1
Cameroon 1 Burundi 1 Cuba 1
Cape Verde 1 Cambodia 1 Ecuador 1
Central Af Rep 1 Cape Verde 1 Eritrea 1
Chad 1 Central Af Rep 1 Ethiopia 1
Comoros 1 Chad 1 Georgia 1
Congo, Dem R 1 China 1 Ghana 1
Congo, Rep 1 Comoros 1 Guinea 1
Cote d’Ivoire 1 Congo, Dem R 1 Guinea-Bissau 1
Croatia 1 Congo, Rep 1 Haiti 1
Cuba 1 Cote d’Ivoire 1 Honduras 1
Djibouti 1 Croatia 1 Iran 1
Ecuador 1 Cuba 1 Iraq 1
Egypt 1 Djibouti 1 Kazakhstan 1
El Salvador 1 Ecuador 1 Korea, N 1
 409

(Education cont’d)
Equatorial Guin 1 El Salvador 1 Kyrgyzstan 1
Eritrea 1 Equatorial Guin 1 Laos 1
Ethiopia 1 Eritrea 1 Liberia 1
Gabon 1 Estonia 1 Libya 1
Gambia, The 1 Gabon 1 Macedonia 1
Georgia 1 Georgia 1 Mauritania 1
Ghana 1 Guatemala 1 Mongolia 1
Guatemala 1 Guinea 1 Mozambique 1
Guinea 1 Guinea-Bissau 1 Nepal 1
Guinea-Bissau 1 Haiti 1 Niger 1
Haiti 1 Honduras 1 Nigeria 1
Honduras 1 Iran 1 Romania 1
Iran 1 Iraq 1 Russia 1
Iraq 1 Kazakhstan 1 Rwanda 1
Kazakhstan 1 Korea, N 1 Senegal 1
Korea, N 1 Kyrgyzstan 1 Sierra Leone 1
Kyrgyzstan 1 Latvia 1 Slovenia 1
Laos 1 Macedonia 1 Somalia 1
Latvia 1 Madagascar 1 Sudan 1
Lesotho 1 Mali 1 Tajikistan 1
Liberia 1 Mauritania 1 Togo 1
Libya 1 Mexico 1 Turkey 1
Macedonia 1 Moldova 1 Turkmenistan 1
Madagascar 1 Mongolia 1 Uganda 1
Malawi 1 Morocco 1 Vietnam 1
Mali 1 Mozambique 1 Yemen 1
Mauritania 1 Niger 1 Yugoslavia 1
Mexico 1 Paraguay 1 Zambia 1
Moldova 1 Peru 1 Albania 2
Mongolia 1 Romania 1 Algeria 2
Morocco 1 Russia 1 Antigua 2
Mozambique 1 Sao Tome 1 Azerbaijan 2
Nepal 1 Saudi Arabia 1 Belarus 2
Nicaragua 1 Senegal 1 Belize 2
Niger 1 Slovakia 1 Chad 2
Nigeria 1 Slovenia 1 Congo, Rep 2
Pakistan 1 Sudan 1 Cyprus 2
Peru 1 Syria 1 Djibouti 2
Romania 1 Tajikistan 1 Estonia 2
Russia 1 Togo 1 Gambia, The 2
Rwanda 1 Tunisia 1 Hungary 2
Senegal 1 Turkey 1 India 2
Sierra Leone 1 Turkmenistan 1 Israel 2
Slovenia 1 Ukraine 1 Jamaica 2
Somalia 1 Uruguay 1 Kenya 2
Sudan 1 Uzbekistan 1 Korea, S 2
Swaziland 1 Venezuela 1 Latvia 2
Syria 1 Yemen 1 Lesotho 2
Tajikistan 1 Yugoslavia 1 Mali 2
 410

(Education cont’d)
Tanzania 1 Antigua 2 Mexico 2
Togo 1 Argentina 2 Moldova 2
Turkey 1 Australia 2 Niue 2
Turkmenistan 1 Bahamas, The 2 Poland 2
Uganda 1 Bahrain 2 Saint Vincent & 2
Ukraine 1 Bangladesh 2 Saudi Arabia 2
Uzbekistan 1 Barbados 2 Spain 2
Venezuela 1 Belize 2 Swaziland 2
Vietnam 1 Bermuda 2 Syria 2
West Bank 1 Botswana 2 Tanzania 2
Yemen 1 Brazil 2 Zimbabwe 2
Yugoslavia 1 British Virgin Is 2 Argentina 3
Zambia 1 Brunei 2 Bahamas, The 3
Zimbabwe 1 Cameroon 2 Bahrain 3
Australia 2 Canada 2 Barbados 3
Austria 2 Cook Is 2 Bermuda 3
Belgium 2 Costa Rica 2 Botswana 3
Bermuda 2 Cyprus 2 British Virgin Is 3
British Virgin I
2 Dominica 2 Brunei 3
Brunei 2 Egypt 2 Chile 3
Canada 2 Ethiopia 2 Colombia 3
Denmark 2 Fiji 2 Cook Is 3
Finland 2 French P 2 Costa Rica 3
France 2 Gambia, The 2 Dominican Rep 3
Germany 2 Ghana 2 Greece 3
Hong Kong 2 Greece 2 Hong Kong 3
Iceland 2 Grenada 2 Indonesia 3
Ireland 2 Guyana 2 Jordan 3
Italy 2 Hong Kong 2 Kuwait 3
Japan 2 India 2 Lebanon 3
Liechtenstein 2 Indonesia 2 Lithuania 3
Luxembourg 2 Ireland 2 Macau 3
Netherlands 2 Israel 2 Maldives 3
New Zealand 2 Jamaica 2 Malta 3
Norway 2 Jordan 2 Mauritius 3
Singapore 2 Kenya 2 Morocco 3
Sweden 2 Korea, S 2 Namibia 3
Switzerland 2 Kuwait 2 Netherlands An 3
United Kingdom 2 Laos 2 Oman 3
United States 2 Lebanon 2 Puerto Rico 3
Antigua 3 Lesotho 2 Samoa 3
Argentina 3 Liberia 2 Seychelles 3
Bahamas, The 3 Libya 2 Singapore 3
Bahrain 3 Luxembourg 2 Solomon 3
Barbados 3 Macau 2 Sri Lanka 3
Belize 3 Malawi 2 Taiwan 3
Bhutan 3 Malaysia 2 Thailand 3
Bolivia 3 Maldives 2 Tonga 3
 411

(Education cont’d)
Botswana 3 Malta 2 Tunisia 3
Brazil 3 Mauritius 2 Uruguay 3
Bulgaria 3 Namibia 2 Vanuatu 3
Chile 3 Nepal 2 Australia 4
China 3 Netherlands An 2 Austria 4
Colombia 3 New Zealand 2 Belgium 4
Cook Is 3 Nicaragua 2 Canada 4
Costa Rica 3 Nigeria 2 Czech Rep 4
Cyprus 3 Niue 2 Denmark 4
Czech Rep 3 Oman 2 Finland 4
Dominica 3 Pakistan 2 France 4
Dominican Rep 3 Panama 2 Germany 4
Estonia 3 Papua New G 2 Iceland 4
Fiji 3 Philippines 2 Ireland 4
French P 3 Puerto Rico 2 Italy 4
Greece 3 Qatar 2 Japan 4
Grenada 3 Rwanda 2 Liechtenstein 4
Guyana 3 Saint Kitts 2 Luxembourg 4
Hungary 3 Saint Lucia 2 Netherlands 4
India 3 Saint Vincent &2 New Zealand 4
Indonesia 3 Samoa 2 Norway 4
Israel 3 Seychelles 2 Sweden 4
Jamaica 3 Sierra Leone 2 Switzerland 4
Jordan 3 Singapore 2 United Kingdom 4
Kenya 3 Somalia 2 United States 4
Korea, S 3 South Africa 2 Benin 5
Kuwait 3 Sri Lanka 2 Bolivia 5
Lebanon 3 Suriname 2 Brazil 5
Lithuania 3 Swaziland 2 Bulgaria 5
Macau 3 Tanzania 2 Cambodia 5
Malaysia 3 Thailand 2 Cape Verde 5
Maldives 3 Tonga 2 Central Af Rep 5
Malta 3 Trinidad 2 China 5
Mauritius 3 Uganda 2 Dominica 5
Namibia 3 United Arab E 2 Egypt 5
Netherlands An 3 United Kingdom 2 El Salvador 5
Niue 3 Untied States 2 Equatorial Guin 5
Oman 3 Vanuatu 2 Fiji 5
Panama 3 Vietnam 2 French P 5
Papua New G 3 Virgin I 2 Gabon 5
Paraguay 3 West Bank 2 Grenada 5
Philippines 3 Zambia 2 Guatemala 5
Poland 3 Zimbabwe 2 Guyana 5
Portugal 3 Austria 3 Madagascar 5
Puerto Rico 3 Belgium 3 Malawi 5
Qatar 3 Chile 3 Malaysia 5
Saint Kitts 3 Colombia 3 Nicaragua 5
Saint Lucia 3 Czech Rep 3 Pakistan 5
 412

(Education cont’d)
Saint Vincent & 3 Denmark 3 Panama 5
Samoa 3 Dominican Re 3 Papua New G 5
Sao Tome 3 Finland 3 Paraguay 5
Saudi Arabia 3 France 3 Peru 5
Seychelles 3 Germany 3 Philippines 5
Slovakia 3 Hungary 3 Portugal 5
Solomon 3 Iceland 3 Qatar 5
South Africa 3 Italy 3 Saint Kitts 5
Spain 3 Japan 3 Saint Lucia 5
Sri Lanka 3 Liechtenstein 3 Sao Tome 5
Suriname 3 Lithuania 3 Slovakia 5
Taiwan 3 Netherlands 3 South Africa 5
Thailand 3 Norway 3 Suriname 5
Tonga 3 Poland 3 Trinidad 5
Trinidad 3 Portugal 3 Ukraine 5
Tunisia 3 Solomon 3 United Arab E 5
United Arab E 3 Spain 3 Uzbekistan 5
Uruguay 3 Sweden 3 Venezuela 5
Vanuatu 3 Switzerland 3 Virgin I 5
Virgin I 3 Taiwan 3 West Bank 5
Source: SPSS Cluster Membership tables, K-means CA

Bold = markets indicated by screening as the optimal ones

Xxxxx = an exception because of prior exclusion as an outlier market.
Manual assessment desirable to check its true attractiveness

Education services with organisational Objectives

Low Capability High Capability

Organisation Organisation
Country Cl. Country Cl.
No No

Afghanistan 1 Afghanistan 1
Albania 1 Albania 1
Algeria 1 Algeria 1
Angola 1 Angola 1
Antigua 1 Antigua 1
Armenia 1 Armenia 1
Austria 1 Austria 1
Azerbaijan 1 Azerbaijan 1
Bahamas 1 Bahamas 1
Bahrain 1 Bahrain 1
Bangladesh 1 Bangladesh 1
Barbados 1 Barbados 1
Belarus 1 Belarus 1
Belgium 1 Belgium 1
Belize 1 Belize 1
 413

(Education cont’d)
Benin 1 Benin 1
Bermuda 1 Bermuda 1
Bhutan 1 Bhutan 1
Bolivia 1 Bolivia 1
Bosnia & Herz 1 Bosnia & Herz 1
Botswana 1 Botswana 1
British Virgin Is 1 British Virgin Is 1
Brunei 1 Brunei 1
Bulgaria 1 Bulgaria 1
Burkina Faso 1 Burkina Faso 1
Burma 1 Burma 1
Burundi 1 Burundi 1
Cambodia 1 Cambodia 1
Cameroon 1 Cameroon 1
Cape Verde 1 Cape Verde 1
Central Af Rep 1 Central Af Rep 1
Chad 1 Chad 1
Chile 1 Chile 1
Colombia 1 Colombia 1
Comoros 1 Comoros 1
Congo, Dem R 1 Congo, Dem R 1
Congo, Rep 1 Congo, Rep 1
Cook Is 1 Cook Is 1
Costa Rica 1 Costa Rica 1
Cote d’Ivoire 1 Cote d’Ivoire 1
Croatia 1 Croatia 1
Cuba 1 Cuba 1
Cyprus 1 Cyprus 1
Czech Rep 1 Czech Rep 1
Denmark 1 Denmark 1
Djibouti 1 Djibouti 1
Dominica 1 Dominica 1
Dominican Rep 1 Dominican Rep 1
Ecuador 1 Ecuador 1
Egypt 1 Egypt 1
El Salvador 1 El Salvador 1
Equatorial Guin 1 Equatorial Guin 1
Eritrea 1 Eritrea 1
Estonia 1 Estonia 1
Ethiopia 1 Ethiopia 1
Fiji 1 Fiji 1
Finland 1 Finland 1
French P 1 French P 1
Gabon 1 Gabon 1
Gambia, The 1 Gambia, The 1
Georgia 1 Georgia 1
Ghana 1 Ghana 1
Greece 1 Greece 1
Grenada 1 Grenada 1
 414

(Education cont’d)
Guatemala 1 Guatemala 1
Guinea 1 Guinea 1
Guinea-Bissau 1 Guinea-Bissau 1
Guyana 1 Guyana 1
Haiti 1 Haiti 1
Honduras 1 Honduras 1
Hong Kong 1 Hong Kong 1
Hungary 1 Hungary 1
Iceland 1 Iceland 1
Iraq 1 Iraq 1
Ireland 1 Ireland 1
Israel 1 Israel 1
Jamaica 1 Jamaica 1
Jordan 1 Jordan 1
Kazakhstan 1 Kazakhstan 1
Kenya 1 Kenya 1
Korea, N 1 Korea, N 1
Kuwait 1 Kuwait 1
Kyrgyzstan 1 Kyrgyzstan 1
Laos 1 Laos 1
Latvia 1 Latvia 1
Lebanon 1 Lebanon 1
Lesotho 1 Lesotho 1
Liberia 1 Liberia 1
Libya 1 Libya 1
Liechtenstein 1 Liechtenstein 1
Lithuania 1 Lithuania 1
Luxembourg 1 Luxembourg 1
Macau 1 Macau 1
Macedonia 1 Macedonia 1
Madagascar 1 Madagascar 1
Malawi 1 Malawi 1
Malaysia 1 Malaysia 1
Maldives 1 Maldives 1
Mali 1 Mali 1
Malta 1 Malta 1
Mauritania 1 Mauritania 1
Mauritius 1 Mauritius 1
Moldova 1 Moldova 1
Mongolia 1 Mongolia 1
Morocco 1 Morocco 1
Mozambique 1 Mozambique 1
Namibia 1 Namibia 1
Nepal 1 Nepal 1
Netherlands 1 Netherlands 1
Netherlands An 1 Netherlands An 1
New Zealand 1 New Zealand 1
Nicaragua 1 Nicaragua 1
Niger 1 Niger 1
 415

(Education cont’d)
Nigeria 1 Nigeria 1
Niue 1 Niue 1
Norway 1 Norway 1
Oman 1 Oman 1
Pakistan 1 Pakistan 1
Panama 1 Panama 1
Papua New G 1 Papua New G 1
Paraguay 1 Paraguay 1
Peru 1 Peru 1
Poland 1 Poland 1
Portugal 1 Portugal 1
Puerto Rico 1 Puerto Rico 1
Qatar 1 Qatar 1
Romania 1 Romania 1
Rwanda 1 Rwanda 1
Saint Kitts 1 Saint Kitts 1
Saint Lucia 1 Saint Lucia 1
Saint Vincent & 1 Saint Vincent & 1
Samoa 1 Samoa 1
Sao Tome 1 Sao Tome 1
Saudi Arabia 1 Saudi Arabia 1
Senegal 1 Senegal 1
Seychelles 1 Seychelles 1
Sierra Leone 1 Sierra Leone 1
Singapore 1 Singapore 1
Slovakia 1 Slovakia 1
Slovenia 1 Slovenia 1
Solomon 1 Solomon 1
Somalia 1 Somalia 1
South Africa 1 South Africa 1
Sri Lanka 1 Sri Lanka 1
Sudan 1 Sudan 1
Suriname 1 Suriname 1
Swaziland 1 Swaziland 1
Sweden 1 Sweden 1
Switzerland 1 Switzerland 1
Syria 1 Syria 1
Taiwan 1 Taiwan 1
Tajikistan 1 Tajikistan 1
Tanzania 1 Tanzania 1
Togo 1 Togo 1
Tonga 1 Tonga 1
Trinidad 1 Trinidad 1
Tunisia 1 Tunisia 1
Turkmenistan 1 Turkmenistan 1
Uganda 1 Uganda 1
United Arab E 1 United Arab E 1
Uruguay 1 Uruguay 1
Uzbekistan 1 Uzbekistan 1
 416

(Education cont’d)
Vanuatu 1 Vanuatu 1
Venezuela 1 Venezuela 1
Vietnam 1 Vietnam 1
Virgin I 1 Virgin I 1
West Bank 1 West Bank 1
Yemen 1 Yemen 1
Yugoslavia 1 Yugoslavia 1
Zambia 1 Zambia 1
Zimbabwe 1 Zimbabwe 1
Argentina 2 Argentina 2
Australia 2 Australia 2
Brazil 2 Brazil 2
Canada 2 Canada 2
France 2 France 2
Germany 2 Germany 2
Indonesia 2 Indonesia 2
Iran 2 Iran 2
Italy 2 Italy 2
Korea, S 2 Korea, S 2
Mexico 2 Mexico 2
Philippines2 Philippines 2
Spain 2 Spain 2
Thailand 2 Thailand 2
Turkey 2 Turkey 2
Ukraine 2 Ukraine 2
United Kingdom 2 United Kingdom 2
China 3 China 3
India 3 India 3
Japan 3 Japan 3
Russia 3 Russia 3
United States - United States -
Source: SPSS Cluster Membership tables, K-means CA

Bold = markets indicated by screening as the optimal ones

Xxxxx = an exception because of prior exclusion as an outlier market.
Manual assessment desirable to check its true attractiveness

Internal combustion engines

Markets Only Low Capability High Capability

Organisation Organisation
Country Cl Country Cl Country Cl
No No No

Australia 1 Argentina 1 Argentina 1

Austria 1 Australia 1 Botswana 1
Belgium 1 Botswana 1 Brazil 1
Canada 1 Brazil 1 China 1
 417

(Engines cont’d)
Finland 1 Canada 1 Colombia 1
France 1 India 1 Malaysia 1
Italy 1 Indonesia 1 Pakistan 1
Korea, S 1 Korea, S 1 Poland 1
Netherlands 1 Malaysia 1 Romania 1
Spain 1 Pakistan 1 Russia 1
Sweden 1 Philippines 1 Slovenia 1
United Kingdom 1 South Africa 1 South Africa 1
Argentina 2 Thailand 1 Turkey 1
Botswana 2 United Kingdom 1 Ukraine 1
Brazil 2 United States 1 Yugoslavia 1
China 2 Austria 2 India 1
Colombia 2 Belgium 2 Thailand 1
Czech Rep 2 Czech Rep 2 Mexico 1
Hungary 2 Finland 2 Hungary 1
India 2 France 2 Iran 1
Indonesia 2 Germany 2 Philippines 1
Malaysia 2 Italy 2 Venezuela 1
Philippines 2 Japan 2 Indonesia 1
Poland 2 Netherlands 2 Bulgaria 1
Portugal 2 Poland 2 Australia 2
South Africa 2 Portugal 2 Austria 2
Taiwan 2 Spain 2 Belgium 2
Thailand 2 Sweden 2 Canada 2
Yugoslavia 3 Taiwan 2 Czech Rep 2
Slovenia 3 Bulgaria 3 Finland 2
Ukraine 3 China 3 France 2
Venezuela 3 Colombia 3 Germany 2
Iran 3 Hungary 3 Italy 2
Bulgaria 3 Iran 3 Korea, S 2
Mexico 3 Mexico 3 Netherlands 2
Russia 3 Romania 3 Portugal 2
Romania 3 Russia 3 Spain 2
Turkey 3 Slovenia 3 Sweden 2
Pakistan 3 Turkey 3 Taiwan 2
Germany 4 Ukraine 3 United Kingdom 2
Japan 4 Venezuela 3 Japan 3
United States 4 Yugoslavia 3 United States 3
Source: SPSS Cluster Membership tables, K-means CA

Bold = markets indicated by screening as the optimal ones

Xxxxx = an exception because of prior exclusion as an outlier market.
Manual assessment desirable to check its true attractiveness
 Legal 1st Economic PC Product consumption

0.25000
1.00000
0.00000
0.25000
0.75000
1.00000
0.00000
0.25000
0.50000
0.75000
1.00000

$
$

0.00000 $$$
0.50000 $
ghanis ghanis

$
ghanis

$$
Af Angola Af Angola Af Angola

$$
$$$
Armenia Armenia Armenia

$
Azerbaij Azerbaij Azerbaij

0.75000 $$ $
$
Banglade Banglade Banglade

$ $$
$
Belgium Belgium Belgium

$ $ $$
Bermuda Bermuda Bermuda

$
$
a a a

$
Bosnia Bosnia Bosnia

$
British British British

$
$ $
Burkina Burkina Burkina

$
Cambodia Cambodia Cambodia

$$
$
$
Ver Ver Ver

$ $ $ $$$$$$
$ $
CapeChile CapeChile CapeChile
Comoros Comoros Comoros

$
$

0.50000 $ $ $ $
Cook Isl Cook Isl Cook Isl

$
Croatia Croatia Croatia

$ $
$
Czech Re Czech Re Czech Re

$ $$
$
$
Dominica Dominica Dominica

$
Salv a Salv a Salv a

$
$
ElEstonia ElEstonia ElEstonia

$ $$ $ $ $ $

$
Finland Finland Finland

$ $$

$
Gabon Gabon Gabon

$$ $ $$$
Germany Germany Germany

$
$

$
Grenada Grenada Grenada

$
$

$
Guinea-B Guinea-B Guinea-B

$
$ $ $
$

$
Honduras Honduras Honduras

$
$
Iceland Iceland Iceland

$
Iran Iran Iran

$$$

$$ $$
$
Israel Israel Israel

$ $ $ $$$
$$

$$ $
$$ $ $
Japan Japan Japan

$
Keny a Keny a Keny a

$
Kuwait Kuwait Kuwait

$$
$
$
Latv ia Latv ia Latv ia

$$
$
Liberia
Lithuani Liberia
Lithuani Liberia
Lithuani

$
Macedoni Macedoni Macedoni

$
$

$ $$$$ $$$ $$ $$ $ $$ $ $ $$ $
Malay sia Malay sia Malay sia

$
Malta Malta Malta

$ $
Mexico Mexico Mexico

$ $

$
$
Morocco Morocco Morocco

Country
Country
Country

$ $$ $
$$ $ $ $$$$

$ $$$$$$$ $ $ $ $$ $$ $$$$$ $$$ $

Nepal Nepal Nepal

$
$

$$
Nicaragu
Niue Nicaragu
Niue Nicaragu
Niue

$
$

$
$
Best Markets for Education

Pakistan Pakistan Pakistan

$
Paraguay Paraguay Paraguay

$ $ $ $ $ $$ $$
$
Poland Poland Poland

$
Qatar Qatar Qatar

$
$$ $ $ $
$ $ $ $ $ $$
Rwanda Rwanda Rwanda

$
$$
Vi Vi Vi

$
$

$
Saint Ar
Saudi Saint Ar
Saudi Saint Ar
Saudi

$
Sierra L Sierra L Sierra L
$
enia enia enia

$
Slov Slov Slov

$
South Af South Af South Af

$$
$

$$ $ $$ $
$ $$$
Sudan Sudan Sudan

$ $ $
$
Sweden Sweden Sweden

$$ $$ $
Taiwan Taiwan Taiwan $$

$
Thailand Thailand Thailand

$ $$
$ $$$$
Trinidad Trinidad Trinidad

$
$$

$
$

$
Turkmeni Turkmeni Turkmeni

$ $ $

$$$$$ $$$$$$$$$$$ $ $$ $
$
United A United A United A

$ $$
$

$
Uruguay Uruguay Uruguay $ $$$$$ $ $$ $$$$$$$$$$$$$$$$$$$ $$$$$$$$$$ $$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$$$$$$$$$$$$$$$$

$$
Venezuel Venezuel Venezuel

$$$$$

$ $$$$$$ $$$$$ $$$$ $$$$$$ $$$$$$$$ $ $ $$$$$ $$$$$$ $$$$$$$$$$$$$$ $$$$$ $ $$$$$$$$$$$ $$$$$
West Ban West Ban West Ban $$$$$$$$$$$$$$$$$$$$$$$$$ $ $ $$$ $$$ $ $ $$$ $$$$$ $$ $$$$$$$$$$$ $ $$$$$$
Zambia Zambia Zambia
418

APPENDIX J

Cultural distance Competition Product consumption growth

0.00000
0.25000
0.75000
1.00000
0.00000
0.25000
0.50000
0.75000
1.00000
0.00000
0.25000
0.50000
0.75000
1.00000

$
$$

ghanis ghanis ghanis

$
$

$
$$
Af Angola Af Angola Af Angola

$$ $
$$
$

Armenia Armenia Armenia

$
$
$
Azerbaij Azerbaij Azerbaij

$
$$

$
$

$$ $ $
Banglade Banglade Banglade

$$ $
$$

Belgium Belgium Belgium

$
$

$ $$
$$ $
$
$ $$ $

Bermuda Bermuda Bermuda

$
$

$
$

a a a

$
$$
Bosnia Bosnia Bosnia

$ $$ $ $

$ $ $
$ $

$
$
British British British

$
$

$ $ $
$
$
$

Burkina Burkina Burkina

$
$$

Cambodia Cambodia Cambodia

$
$
$

$
$
$

Ver Ver

$$ $
Ver
$

$$
$
CapeChile CapeChile CapeChile

$
$

Comoros Comoros Comoros

$ $
$

Isl Isl
$$
Cook Cook Cook Isl
$

$
$

$
Croatia Croatia Croatia

$
Re Re
$

Czech Czech Czech Re

$

$
$

Dominica Dominica Dominica

$
$$$$ $ $

$ $
Salv a Salv a Salv a
$

$
$

$
$

$ $ $
ElEstonia ElEstonia ElEstonia
$

$
$

$ $$ $
Finland Finland Finland
$

$
$

$ $$ $
$

Gabon Gabon Gabon

$ $$$
$$
Germany Germany Germany

$
$$

$ $$ $$
Grenada Grenada Grenada
$ $
$

Guinea-B Guinea-B Guinea-B

$

$$
$

Honduras Honduras Honduras

$

$
$

$
Iceland Iceland Iceland
$

$$$
$

$
$$ $

$
Iran Iran Iran
$
$

$ $ $ $ $

$
$

Israel Israel Israel

$ $
$

Japan Japan Japan

$$ $

$
Keny a Keny a Keny a

$
Kuwait Kuwait Kuwait

$
$

Latv ia Latv ia Latv ia

$ $$
$ $

$
$

Liberia Liberia Liberia

$$ $
Lithuani Lithuani Lithuani
$

$
$ $ $ $ $$$$ $$$ $ $$$ $$

$
Macedoni Macedoni Macedoni
$

Malay sia Malay sia Malay sia

$
$

$ $$

$
Malta Malta Malta

$
Mexico Mexico Mexico
$

$
$ $$

$
Morocco Morocco Morocco

Country
Country
Country

$ $$
$$ $$ $

$$ $ $$ $ $
$

Nepal Nepal Nepal

$

$
$

$ $
Nicaragu Nicaragu Nicaragu
$

$
$

$$
$

Niue Niue Niue

$
$

Pakistan Pakistan Pakistan

$
$

$$
$

$$
$

Paraguay Paraguay Paraguay

$$$$ $ $$ $

$$
$
$

Poland Poland Poland

$
Qatar
Rwanda Qatar
Rwanda Qatar
$ $$ $ $ $$

Rwanda
$

$
$

Vi Vi Vi
$

$
$

$
$

$ $$

Saint Ar
Saudi Saint Ar
Saudi Saint Ar

$$ $
Saudi
$ $ $

$ $$ $ $ $$
$
L L L
$$ $

Sierra Sierra Sierra

$
$

enia enia enia

$
Slov Slov Slov
$$

$ $
Af Af Af
$

$
South South
$

South

$$

$ $
$

$
Sudan Sudan Sudan
PLOTS – TERTIARY EDUCATION: COUNTRY SCORES BY VARIABLE

$
Sweden Sweden Sweden
$$

Taiwan Taiwan Taiwan

$$
Thailand Thailand Thailand

$
$

$
Trinidad Trinidad Trinidad
$

$
$$

$
$

Turkmeni Turkmeni Turkmeni

$$$ $ $ $ $ $ $ $$ $ $ $ $$ $

$ $
$ $$
$

United A United A United A

$ $
$

$$ $ $
$$

$
$$$$$$ $$$ $$$$ $$

Uruguay Uruguay Uruguay

$

$$
$

0.50000 $$$ $$$ $$$$$$$$ $ $ $$ $$$$ $ $$ $$ $$ $$ $$$$$$$$$$$$ $ $ $$$

$ $ $$ $$

Venezuel Venezuel Venezuel

$ $$ $

$$ $$
$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$$$ $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$ $$$ $$$$

West Ban
Zambia West Ban
Zambia West Ban
Zambia
 Marke t risk

0.00000
0.25000
1.00000

$
$
0.50000 $
0.75000 $

ghanis
$
$
$

Af Angola
$
$

Armenia
$
Azerbaij
$ $$ $

Banglade
$$

Belgium

$
$ $$

Bermuda
$
$

Bosnia a
$

$
British

$
Burkina
$
$
$$

Cambodia Ver
$

$
CapeChile
$
$

Comoros
$
$ $$

Cook Isl
$

Croatia
Czech Re
Dominica
$ $ $$
$

Salv a
$ $ $

ElEstonia
$

Finland
Gabon
$ $$ $ $ $$$ $ $

$$ $ $

$$

Germany
$
$

Grenada
$

Guinea-B
$

Honduras
$$ $ $ $

$
$ $ $ $ $$ $

Iceland
$

Iran
Israel
$ $

Japan
$

Keny a
$

Kuwait
$
Latv ia
$

Liberia
$ $
$ $ $

Lithuani
Macedoni
$$
$

Malay sia
$ $

Malta
$
$ $ $$$ $ $

Mexico
$ $ $

Morocco

Country
$
$ $ $

Nepal
$
$ $ $ $

Nicaragu
$
$

Niue
$

Pakistan
$ $

Paraguay
$

Poland
Qatar
$ $ $$ $ $
$ $

Rwanda Vi
$

Saint Ar
$

Saudi
$ $ $ $ $ $$$$$$ $

Sierra L
$

enia
$

Slov
$

South Af
$
$ $

Sudan
$ $

Sweden
$$ $ $

Taiwan
Thailand
$
$

Trinidad
$
$

Turkmeni
$$$ $
$ $ $

United A
$

Uruguay
$

$
$ $ $ $ $$ $$ $

Venezuel
$

West Ban
Zambia
419
 Low all ge nerics Legal fairne ss Low all gener ics 1st Economic PC Low all generics Product consumption

0.00000
0.40000
0.60000
0.20000
0.40000
0.60000

0.00000 $$$
ghanis ghanis

$$
ghanis

$$
Af Angola Af Angola

$$
Af Angola
Armenia Armenia Armenia

$
Azerbaij Azerbaij

$$$$

$$ $
Azerbaij

$
Banglade Banglade

0.60000 $ $
Belgium Belgium Banglade

$ $ $$
Belgium

$ $
Bermuda Bermuda

$
Bermuda

$
a a

$
Bosnia
British Bosnia
British Bosnia a

$
British

$
Burkina Burkina

$ $
Burkina

$ $
Cambodia Cambodia

$$
Cambodia

$
Ver Ver

$ $ $ $$$ $
Ver

$ $
$
CapeChile CapeChile

$
CapeChile

$
Comoros Comoros

$
Comoros

$
Cook Isl Cook Isl Cook Isl

$
Croatia Croatia

$
Croatia

$
Czech Re Czech Re Re

$
Czech

$
Dominica Dominica

0.20000 $ $ $ $ $$$
Dominica

$
$
Salv a Salv a Salv a

$
$
ElEstonia ElEstonia

$ $ $$$ $
ElEstonia

$
Finland Finland

$ $$
Finland

$
$$
Gabon
Germany Gabon
Germany Gabon
Germany

$$
Grenada Grenada

$ $ $$
Grenada

$
Guinea-B Guinea-B

$
Guinea-B

$ $ $
$
Honduras Honduras

$
Honduras

$
Iceland Iceland Iceland

$
Iran Iran

$$$

$$ $$
Iran

$$$$ $
$
Israel Israel Israel

$ $$
$$

$$
Japan Japan

$
$ $
Japan

$$ $
Keny a Keny a Keny a

$
Kuwait Kuwait Kuwait

$
$
Latv ia Latv ia

$
Latv ia

$$
$
Liberia
Lithuani Liberia
Lithuani Liberia
Lithuani
Macedoni Macedoni Macedoni

0.40000 $ $ $$$$$$ $$$$$$$ $$ $$ $$ $$$ $$ $$

Malay sia Malay sia

$$
Malay sia

$
Malta Malta Malta

$ $
Mexico Mexico Mexico

$
$ $
$
Morocco Morocco Morocco

Country
Country
$$$ $$ $

$ $$$$$$$ $ $ $ $$ $$ $$$$$ $$$ $

Country

$
Nepal

$
Nepal Nepal

$$
$
Nicaragu
Niue Nicaragu
Niue Nicaragu
Niue
Pakistan

$
Pakistan Pakistan

$$

$
$$
Paraguay Paraguay Paraguay

$ $ $ $ $ $$ $$
Poland
Qatar Poland
Qatar Poland
Qatar
$
Rwanda

$ $ $ $ $ $$
Rwanda Rwanda

$$$ $ $ $ $$
Vi

$$
Vi Vi

$
Saint Ar

$ $ $
Saudi Saint Ar Saint Ar

$
Sierra L Saudi L Saudi

$
Slov enia Sierra Sierra L
$
Slov enia enia

$
Af Slov

$
South Af

$$
South Af

$$ $ $$ $
South

$ $ $ $
Sudan

$ $$$
Sudan

$ $$ $
Sudan

$
Sweden

$
Taiwan Sweden
Taiwan Sweden
Taiwan
Thailand Thailand Thailand $$

$ $ $$
$$$$
Trinidad Trinidad

$
Trinidad

$
$$

$
$

Turkmeni Turkmeni Turkmeni

$$$$$ $$$$$$$$$$$ $ $$ $
$

$
United A United A

$ $ $$

$ $$
United A

$
Uruguay Uruguay Uruguay

$ $$
Venezuel Venezuel $ $$$$$ $ $$ $$$$$$$$$$ $ $$$ $$ $$$$ $$ $ $$$$$ $$$ $$ $ $$$ $$$$$ $$ $$$$

$ $$$

0.20000 $ $$$$$$$$$$$ $$$$$$$$$$ $$$$$$$$$ $ $$$$$ $$$$$$ $$$$$$$$$$$$$$ $$$$$ $$$$$$$$$$$$ $$$$$
Venezuel
0.00000 $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$
West Ban
Zambia West Ban
Zambia West Ban
Zambia
420

Low all gener ics Cultural distance Low all generics Competition Low all generics Product consumption growth

0.00000
0.20000
0.40000
0.60000

0.00000
0.20000
0.60000
0.00000
0.20000
0.60000

$
ghanis ghanis

$
ghanis
0.40000 $

$
$$
Af Angola Af Angola

$$ $
$

$$
Armenia Armenia Af Angola

$
Armenia

0.40000 $$
$
Azerbaij Azerbaij

$
$

$
Azerbaij

$$ $ $
Banglade Banglade

$$ $

$ $
Belgium Belgium Banglade

$
Belgium

$$ $
$
Bermuda Bermuda

$
$
$$$ $

a a Bermuda

$
$$
$$

Bosnia Bosnia

$ $$ $ $
a

$ $ $
Bosnia

$$
$
British British

$
British

$
$
Burkina Burkina

$ $$
$

Cambodia Cambodia Burkina

$
$

$
$
Ver Ver Cambodia

$
$

$
Ver

$$
$
CapeChile CapeChile

$
CapeChile

$
$

Comoros Comoros

$ $

$
Comoros

$$
Cook Isl Cook Isl

$
$

$
Croatia Croatia Cook Isl

$
Croatia

$
Czech Re Czech Re Re
$$

Dominica Dominica Czech

$ $
$$$$ $ $

$
Dominica

$ $
Salv a Salv a $

$
$

$
Salv a

$ $ $

$
ElEstonia ElEstonia
Best Markets for Low Capability Education Organisations

$
ElEstonia

$
Finland Finland
$

$
$ $$ $
Finland
$

Gabon Gabon

$
Gabon

$$
Germany Germany

$
Germany

$
Grenada Grenada
$

Guinea-B Guinea-B Grenada

$
Honduras Honduras Guinea-B

$
$

Honduras
$

$
$

Iceland Iceland
$

$$$
$

$
Iceland
$ $

$$ $ $$
Iran Iran
$
$ $

Iran

$
$

Israel Israel

$$ $
$

$ $$ $$$ $
Israel
$

Japan Japan
$$ $

$ $
$
Japan

$
$

Keny a Keny a

$
$$$
Kuwait Kuwait Keny a

$
$ $ $ $
$

Latv ia Latv ia Kuwait

$ $$
$ $

$
Liberia
Lithuani Liberia
Lithuani Latv ia
Liberia
$

$
$

$
Macedoni Macedoni Lithuani
$

Malay sia Malay sia Macedoni

$$
Malta Malay sia

$
Malta Mexico Malta
$
$

Mexico
$ $$

$$
$
Morocco Mexico
$ $$

Country
Morocco

Country
$$ $$ $ Morocco
$

$ $ $$ $ $
Nepal
$

$
Nepal
Country

$
$

Nicaragu Nepal
$

$
Nicaragu

$$
$

Niue Nicaragu
$

$ $
Niue
$
$

Pakistan Niue

$
Pakistan
$$

$$
Paraguay Paraguay Pakistan
$

$
$
Poland Paraguay
$$

Poland
$

Qatar Qatar Poland

$
$$ $ $ $ $$ $ $ $ $$$ $ $ $$

Rwanda Qatar
$ $$ $ $

$
Rwanda
$

Vi
$
Vi Rwanda
$

$
Saint
$

Saint Ar Vi
$

Ar Saudi Saint
$ $ $

Saudi

$ $ $ $ $ $$
L Ar
$$ $

Sierra L Sierra Saudi

$
enia Slov enia $ Sierra L
$$
$

Slov Af enia
$
$

South Af South Slov

$$
Af
$

$ $
Sudan South

$
Sudan
$

Sweden Sudan
$$

Sweden
$
$ $

Taiwan Taiwan Sweden

$ $

Thailand Thailand Taiwan

$

$
Thailand
$

$ $
Trinidad Trinidad
$
$

$$
Turkmeni Turkmeni Trinidad
$

$$$ $ $ $ $ $ $$ $ $ $ $ $

$ $
Turkmeni
$ $$

United A United A
$ $

A
$

$$ $ $
$

United

$
Uruguay Uruguay
$$ $

$$
$ $$

$$ $ $$$$ $$$$$$$$ $ $$ $$ $$$ $$$$ $$ $ $$ $$$ $$$$$$$ $$$$$$ $$$$ $$$

Uruguay
$ $$

Venezuel Venezuel
$ $$ $

$ $ $ $$ $$
Ban Ban Venezuel
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

West
Zambia West
Zambia West Ban
Zambia
 Low Language (country-specific) Low all generics Marke t risk

0.25000
0.50000
0.75000
0.00000
0.20000
0.40000

$
$$
ghanis
Af Angola ghanis
Af Angola
$
Armenia
Azerbaij Armenia
Azerbaij

$
$ $$$$$
Banglade Banglade

$$
Belgium Belgium

$
0.60000 $ $$

$
Bermuda Bermuda
$$ $
$

$$$ $
Bosnia a Bosnia a

$
British British

$
Burkina Burkina

$
$$
Cambodia Ver Cambodia Ver $

$
CapeChile CapeChile

1.00000 $$ $ $ $ $ $
Comoros Comoros

$
$$ $ $
Cook Isl Cook Isl
$ $
$

Croatia Croatia

$
Czech Re Czech Re
$

Dominica
Salv a Dominica
Salv a
$ $ $

ElEstonia ElEstonia
$
$ $ $$ $
Finland Finland
$$

$ $ $$ $ $$ $ $$ $
Gabon Gabon
$

$
$$ $ $

Germany Germany $

$
$
Grenada Grenada
$
$$ $ $ $ $ $$ $ $

Guinea-B
Honduras Guinea-B
Honduras
$$ $ $ $

$
$

$
Iceland Iceland
$

Iran Iran
$

Israel Israel
$$ $

Japan Japan $
$

Keny a Keny a
$

Kuwait Kuwait

$
Latv ia Latv ia
$

$$$$$$$$$
Liberia Liberia
$
$ $
$$ $

Lithuani
Macedoni Lithuani
Macedoni
$

Malay sia Malay sia

$ $

Malta Malta
$

Mexico Mexico
$
$

$
$
$

Morocco Morocco

Country
Country
$$ $$
$ $ $ $$$ $ $ $

Nepal Nepal
$ $ $ $

Nicaragu
Niue Nicaragu
Niue
$

Pakistan Pakistan
$$ $ $ $ $

Paraguay Paraguay
$

Poland Poland $
Qatar Qatar

$ $$
$ $$$ $ $
$ $

Rwanda Vi Rwanda Vi
$

Saint Ar Saint Ar
$

Saudi Saudi
$ $ $ $ $ $$$$$$ $

L Sierra L $
$

Sierra enia
$

enia Slov
$

Slov Af
$

Af South
$

South
$ $

Sudan Sudan
$ $

Sweden
$$ $

Sweden
$

Taiwan
Thailand Taiwan
Thailand
$
$

Trinidad Trinidad
$
$

$$
$

Turkmeni Turkmeni
$$$ $
$ $ $

United A United A
$

$ $$ $$ $ $$$ $ $ $$$
Uruguay Uruguay
$

0.00000 $$$$$$$ $$$$$$ $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$ $$ $$$$$$ $$$ $$$$$$$$$$$$$$$$$$$$

$ $ $ $ $$ $$ $

Venezuel Venezuel
$

$ $$ $$$$$$$ $ $$$$$ $$ $$$$$ $$$$ $ $ $$$$ $ $$

West Ban
Zambia West Ban
Zambia
421
 Hi all ge nerics Legal fairne ss Hi all gener ics 1st Economic PC Hi all generics Product consumption

0.50000
2.00000
0.00000
0.50000
1.50000
2.00000
0.00000
0.50000
1.00000
1.50000
2.00000

$
$

0.00000 $$$
1.00000 $
ghanis ghanis

$
ghanis

$$
Af Angola Af Angola Af Angola

$$
$$$
Armenia Armenia Armenia

$
Azerbaij Azerbaij Azerbaij

1.50000 $$ $
$
Banglade Banglade Banglade

$ $$
$
Belgium Belgium Belgium

$ $ $$
Bermuda Bermuda Bermuda

$
$
a a a

$
Bosnia Bosnia Bosnia

$
British British British

$
Burkina Burkina

$
$ $
Burkina

$ $ $ $$$$
Cambodia Cambodia a

$$
Cambodi

$
Ver Ver

$
Ver

$
$ $
$
CapeChile CapeChile CapeChile
Comoros Comoros Comoros

1.00000 $ $ $ $
Cook Isl Cook Isl Cook Isl

$
Croatia Croatia

$ $
Czech Re Czech Re Croatia
Czech Re

$
$
Dominica Dominica

$ $$$
Dominica

$
Salv a Salv a Salv a

$
$
ElEstonia ElEstonia

$ $$ $ $ $ $
ElEstonia

$
Finland Finland

$ $$
Finland

$$
Gabon Gabon

$
Gabon

$ $ $$$
Germany Germany

$
Germany

$
Grenada Grenada

$
Grenada

$
Guinea-B Guinea-B

$
Guinea-B

$ $ $

$
$
Honduras Honduras

$
Honduras

$
Iceland Iceland Iceland

$
Iran Iran

$$$

$$ $$
Iran

$
Israel Israel Israel

$ $ $ $ $ $$$
$$

$$
Japan Japan

$
$ $
Japan

$$
Keny a Keny a Keny a

$
Kuwait Kuwait

$ $$$$
Kuwait

$
$
Latv ia Latv ia

$
Latv ia

$$
$
Liberia Liberia Liberia

$
Lithuani Lithuani

$
Lithuani

$
Macedoni Macedoni Macedoni

$ $$$$ $$$ $$ $$ $$ $$ $$$ $$ $$

Malay sia Malay sia Malay sia

$
Malta Malta

$$ $
Mexico Mexico Malta
Mexico

$ $

$
$

$
Morocco Morocco Morocco

Country
Country
$$

$ $$$$$$$ $ $ $ $$ $$ $$$$$ $$$ $

Country
Nepal Nepal

$
$

$$
Nicaragu Nicaragu Nepal
Nicaragu
Niue Niue

$
Niue

$
Pakistan Pakistan

$
Pakistan

$
Paraguay
Poland Paraguay
Poland Paraguay
$
Qatar Qatar Poland
Qatar

$ $ $ $ $$
Rwanda Rwanda

$$ $ $ $
Rwanda

$$
Vi Vi

$
Vi

$
Saint Ar
Saudi Saint Ar
Saudi Saint Ar

$ $ $
L L Saudi

$
Sierra Sierra Sierra L
$
Slov
South enia
Af Slov
South enia
Af Slov enia
Af

$$
South

$$ $ $$ $
$ $$$
Sudan Sudan

$ $ $
Sudan

$
Sweden Sweden

$$ $$ $
Taiwan Taiwan Sweden

$
Taiwan

$ $ $ $ $
Thailand Thailand Thailand $$

$ $ $$
$
Trinidad Trinidad Trinidad

$
$$

$
$

$
Turkmeni Turkmeni Turkmeni

$ $ $ $ $ $$ $$ $ $ $ $ $

$$$$$ $$$$$$$$$$$ $ $$ $
$
United A United A United A

$ $$
$

$
Uruguay Uruguay Uruguay

$ $$
Venezuel Venezuel Venezuel $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ $$$$$$ $$$$$$$$$$$$$$$$ $$$$$$ $$$

$$$$$

$ $$$$$$ $$$$$ $$$$ $$$$$$ $$$$$$$$ $ $ $$$$$ $$$$$$ $$$$$$$$$$$$$$ $$$$$ $ $$$$$$$$$$$ $$$$$
$ $$$$ $ $$ $$ $$$$ $ $$ $ $$$$ $$$$$ $ $$$$$$$$$ $$ $$$$$$$$$$$ $$ $$$$$$$$
West Ban
Zambia West Ban
Zambia West Ban
Zambia
422

Hi all gener ics Cultural distance Hi all generics Competition Hi all generics Product consumption growth

0.00000
0.50000
1.50000
2.00000
0.00000
0.50000
1.00000
1.50000
2.00000
0.00000
0.50000
1.00000
2.00000

$
ghanis ghanis
$

$
ghanis

$
$$
Af Angola Af Angola

$$ $
1.50000 $

Armenia Armenia Af Angola

$
Armenia

$$
$
Azerbaij Azerbaij

$
$ $$

$
Azerbaij

$$ $ $
Banglade Banglade

$$ $
Banglade
$

Belgium Belgium

$ $$
Belgium

$$ $
$
Bermuda Bermuda

$
$
$$$ $

a a Bermuda

$
$$
$

Bosnia Bosnia

$ $$ $ $
a

$ $ $
$ $
Bosnia

$
$
British British

$
$ $ $
British

$
$
Burkina Burkina

$
$

Cambodia Cambodia Burkina

$
$

$
$
Ver Ver Cambodia

$$ $

$
$

$
Ver

$$
$
CapeChile CapeChile CapeChile

$
$

Comoros Comoros

$ $
Comoros
$$
Cook Isl Cook Isl
$

$
$

$
Croatia Croatia Cook Isl

$
$ $
Czech Re Czech Re Croatia

$
$

Dominica Dominica Czech Re

$
$$$$ $ $

$
Dominica

$ $
Salv a Salv a
$

$
$

$
Salv a
$

$
ElEstonia ElEstonia
$

$
ElEstonia

$ $ $
Finland Finland
$

$
Best Markets for High Capability Education Organisations

$ $$ $
Finland
$

Gabon Gabon

$
Gabon

$$
Germany Germany Germany

$$

$ $$ $$
Grenada Grenada
$

Guinea-B Guinea-B Grenada

$

$$
Honduras Honduras Guinea-B

$
$

Honduras
$

$
$

Iceland Iceland
$

$$$
$

$
Iceland
$$ $

$$ $ $$
Iran Iran
$

$ $ $ $ $

$
Iran
$
$

Israel Israel

$ $
$

Japan Japan Israel

$$ $

$
$

Keny a Keny a Japan

$
Kuwait Kuwait $$$ Keny a

$
$ $ $ $
Kuwait
$

Latv ia Latv ia

$ $$
$ $

$
Latv ia

$
Liberia Liberia

$$$ $
Lithuani Lithuani Liberia
$

$
$
Lithuani
$
$ $ $$

Macedoni
Malay sia Macedoni
Malay sia Macedoni
$

$$
Malay sia

$
Malta Malta

$
Malta

$ $ $ $
Mexico Mexico
$

Mexico
$ $$

$
Morocco Morocco

Country
Country
$ $$
$$ $$ $
Morocco

$ $$$
Nepal Nepal
$

$
Country

$
$

Nepal
$

Nicaragu Nicaragu
$ $$

$
$

$$
Nicaragu
$

Niue Niue
$
$

Niue

$
Pakistan Pakistan

$$
$$

Paraguay Paraguay Pakistan

$

$$
$

Paraguay

$
$$$

Poland Poland
$

Qatar Qatar Poland

$
$

Qatar
$ $$ $ $ $$

Rwanda Rwanda
$

Vi Vi Rwanda
$

$
$

Saint Ar Saint Ar Saint Vi

$$ $
Saudi Saudi
$$$ $$$

$ $$ $ $ $$
Ar
$
L L Saudi
$$ $

Sierra Sierra

$
enia enia Sierra L
$
Slov Slov
$$
$

enia

$
Af Af Slov
$
$

South South

$$

$ $
Af
$

$
Sudan
Sweden Sudan
Sweden South
Sudan
$$

$
$ $

Taiwan Taiwan Sweden

$$
$ $

Thailand Thailand Taiwan

$

$
Thailand
$

$
Trinidad Trinidad
$

$
$

$
Trinidad
$

Turkmeni Turkmeni
$

$$$ $ $ $ $ $ $ $ $ $ $ $

$ $
Turkmeni
$ $$

United A United A
$ $
$

$$ $ $
$

United A

$
Uruguay Uruguay
$

$$
$$

1.00000 $$$ $$$ $$$$$$$$ $ $ $$ $$$$ $ $$$$ $$ $$ $$$$$$$$$$$$ $ $ $$$

Uruguay
$ $ $$ $$

Venezuel Venezuel
$ $$ $

$ $ $ $$ $$
Ban Ban Venezuel
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

West
Zambia West
Zambia West Ban
Zambia
 High Conta cts (country-spe cific) High SCA (country-spe cific) Hi all generics Marke t risk

0.25000
0.50000
0.75000
0.25000
0.50000
0.75000
0.00000
0.50000
2.00000

$
$
1.00000 $
ghanis
Af Angola ghanis
Af Angola ghanis
Af Angola 1.50000 $
$
Armenia
Azerbaij Armenia
Azerbaij Armenia
Azerbaij
Banglade
Belgium Banglade
Belgium Banglade
Belgium $$

$
$ $$

$
Bermuda Bermuda Bermuda
$$ $
$

Bosnia a Bosnia a Bosnia a

$
British British British

$
Burkina Burkina Burkina
$$
$$
Cambodia Ver Cambodia Ver Cambodia Ver
$

$
CapeChile CapeChile CapeChile
$
$

Comoros Comoros Comoros

$
Cook Isl Cook Isl Cook Isl $ $$
$ $
$

Croatia Croatia Croatia

$
$

Czech Re Czech Re Czech Re

$

Dominica Dominica Dominica

$

Salv a Salv a Salv a

$ $ $

ElEstonia ElEstonia ElEstonia

$

$
$ $$$$$$ $ $ $ $$ $

Finland Finland Finland

$$

Gabon Gabon Gabon

$

Germany Germany Germany

$

1.00000 $ $ $$$$$$$$$$$$ $$$$$$$$ $$$

$
Grenada Grenada Grenada
$
$

Guinea-B Guinea-B Guinea-B

$ $$ $ $ $$$ $ $ $

Honduras Honduras Honduras

$$ $ $ $

$
$

Iceland Iceland Iceland

$

Iran Iran Iran

$

Israel Israel Israel

$ $

Japan Japan Japan

$

Keny
Kuwaita Keny
Kuwait a Keny
Kuwait a

$
Latv ia Latv ia Latv ia
$

Liberia Liberia Liberia

$ $
$ $$$ $

Lithuani Lithuani Lithuani

$ $$$$$
Macedoni Macedoni Macedoni
$

Malay sia Malay sia Malay sia

$ $

Malta Malta Malta

$
$ $$$ $ $

Mexico Mexico Mexico

$
$

Morocco Morocco Morocco

$$ $

Country
Country
Country
$ $ $

Nepal Nepal
$

Nepal
$ $ $

Nicaragu Nicaragu Nicaragu

$

Niue Niue Niue

$$

Pakistan Pakistan Pakistan

$$ $ $ $ $

Paraguay Paraguay Paraguay

$

Poland
Qatar Poland
Qatar Poland
Qatar
$ $$ $ $
$ $

Rwanda Vi Rwanda Vi Rwanda Vi

$

Saint Ar Saint Ar Saint Ar

$

Saudi Saudi Saudi

$ $ $ $ $ $$$$$$ $

Sierra L Sierra L Sierra L

$

enia enia enia

$

Slov Slov Slov

$

Af Af Af
$

South South South

$
$ $

Sudan Sudan Sudan

$ $

Sweden Sweden Sweden

$$ $

Taiwan
Thailand Taiwan
Thailand Taiwan
Thailand
$

Trinidad Trinidad Trinidad

$
$

Turkmeni Turkmeni Turkmeni

$ $$
$$$ $
$ $ $

United A United A United A

$

Uruguay Uruguay Uruguay

$

$ $$$$$$$$$$$ $$$$$ $ $$$$$$$$

$$ $ $ $$ $ $

Venezuel Venezuel

0.00000 $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
Venezuel
$

1.00000 $$$ $ $$$ $ $$$$$$ $$$$$ $$$ $$$$$$$$$$$ $$$ $$$$ $$$$ $$$$$ $$$$ $$$$ $

0.00000 $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

West Ban
Zambia West Ban
Zambia West Ban
Zambia
423

High Marke t research (country-spe cific) High Language (country-spe cific)

graphs
0.25000
0.50000
0.75000

0.25000
0.50000
0.75000
1.00000

ghanis
Af Angola ghanis
Af Angola
Armenia
Azerbaij Armenia
Azerbaij
Banglade
Belgium Banglade
Belgium
Bermuda
Bosnia a Bermuda
Bosnia a
0.00000 $$ $$$ $$ $$

British British
Burkina
$$ $ $$

Burkina
Cambodia Cambodia
CapeChileVer Ver
CapeChile
Comoros
Cook Isl Comoros
Cook Isl
Croatia
Czech Re Croatia
Czech Re
Dominica
$ $

Dominica
Salv a
ElEstonia Salv a
ElEstonia
Finland
Gabon Finland
Gabon
Germany Germany
$$

Grenada
Guinea-B Grenada
Guinea-B
Honduras
Iceland Honduras
Iceland
Iran
Israel Iran
Israel
Japan
Keny a Japan
Keny a
Kuwait Kuwait
Latv ia
$$$$

Latv ia Liberia
$$$ $ $$ $$$$ $ $$

Liberia
Lithuani Lithuani
$ $ $

Macedoni Macedoni
Malay sia
$

Malay sia
Malta Malta
Mexico
$

Mexico Morocco
Country

Morocco

Country
$

Nepal Nepal
Nicaragu
Nicaragu
Niue Niue
Pakistan
Pakistan
Paraguay Paraguay
$ $$

Poland Poland

Source: Constructed for this

Qatar
Rwanda Qatar
Rwanda
Saint
Saudi Vi
Ar Saint Ar
Saudi Vi
Sierra
Slov eniaL Sierra
Slov eniaL
South Af South
SudanAf
Sudan
Sweden Sweden
Taiwan Taiwan

research using SPSS interactive

Thailand Thailand
$ $$ $$ $$$ $$$ $ $$$$$$$ $$$$$$ $$$$$$$ $$$$$$$ $$ $$$ $$$ $

$ $ $$ $

Trinidad Trinidad
$

Turkmeni
United A Turkmeni
United A
Uruguay
1.00000 $$$$$$$$$$ $$$$$$$$$$$$$$$$$$$$$ $$ $$$$$$$ $$$$$$$$$$$$$$$ $$$$$$$$$$$$$$$$$$$$$$$ $$$$$$$$$$

Uruguay
$$$ $ $$ $ $

Venezuel
$$

Venezuel
0.00000 $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

West Ban
Zambia West Ban
Zambia
 425

% of variance 96.8 3.2

Market only variables continued
Rotated Structure Matrix Potency index Variables Entered/Removed
WOOL Excludes China Function F Statistic
1 2
Legal fairness 0.747 -0.016 Legal fairness 54.7 Legal fairness 211.5
1st Economic PC 0.373 0.348 1st Economic PC 13.9 1st Economic PC 97.6
Cultural difference 0.228 -0.010 Competition 9.6 Market risk 72.8
Market risk 0.144 0.910 Product consumption 6.0 Cultural difference 57.0
Competition 0.296 0.724 Cultural difference 5.1 Competition 46.4
Product consumption 0.243 -0.351 Market risk 3.6 Product consumption 39.0
Product consumption growth -0.049 0.124 Product consumption growth 0.3 Product consumption growth 33.3

Eigenvalue 8.0 0.2

% of variance 98.1 1.9

TELEPHONES Function
1 2 3
Competition 0.717 -0.219 0.239 Competition 43.5 Competition 106.8
Market risk 0.691 -0.084 0.035 Market risk 39.7 Legal fairness 63.0
Legal fairness 0.615 0.488 -0.094 Legal fairness 35.0 Cultural distance 49.2
Cultural distance 0.034 0.776 0.027 Cultural distance 8.9 1st Economic PC 40.1
Product consumption growth 0.099 -0.224 -0.047 1st Economic PC 8.6 Market risk 33.8
1st Economic PC 0.282 -0.075 0.904 Product consumption growth 1.6 Product consumption growth 28.0
Product consumption -0.029 0.078 0.331 Product consumption 0.4 Product consumption 23.8

Eigenvalue 5.8 1.0 0.2

% of variance 83.0 14.7 2.3
 426

Market only variables continued

Rotated Structure Matrix Potency index Variables Entered/Removed
EDUCATION Function F Statistic
1 2
Competition 0.765 0.415 Competition 57.2 Legal fairness 246.9
Market risk 0.660 0.436 Market risk 42.7 1st Economic PC 137.7
Legal fairness 0.631 -0.190 Legal fairness 38.7 Cultural distance 103.8
1st economic PC 0.552 -0.124 1st economic PC 29.6 Competition 85.8
Cultural difference 0.067 -0.548 Product consumption growth 1.7 Product consumption growth 69.1
Product consumption growth 0.122 0.309 Cultural difference 1.4 Product consumption 58.0
Product consumption 0.034 -0.210 Product consumption 0.3 Market risk 49.7

Eigenvalue 6.0 0.2

% of variance 96.8 3.2

ENGINES Function
1 2 3
Product consumption 0.818 -0.017 -0.072 Product consumption 57.2 Product consumption 70.8
Product consumption growth -0.287 0.117 0.047 Legal fairness 7.8 Legal fairness 50.0
1st economic PC -0.044 0.750 -0.031 Product consumption growth 7.2 Competition 39.8
Legal fairness -0.161 0.670 0.227 1st economic PC 7.1 Product consumption growth 30.3
Cultural difference -0.065 0.373 -0.071 Cultural difference 2.1 1st economic PC 25.1
Competition -0.059 -0.037 0.774 Competition 1.7 Cultural difference 20.2
Market risk -0.086 0.035 0.606 Market risk 1.5 Market risk 16.6

Eigenvalue 16.8 2.4 0.5

% of variance 85.4 12.3 2.3
Bold = largest absolute pooled within-groups correlations between a Variable and any Discriminant function and correlation > +/- .33
xxxxx = variable falls below +/- .33 correlation xxxx = variable below .19 tolerance, indicating possible multicollinearity
xxxxx = variable's largest correlation (bold) in function which accounts for less than 5% of variance or whose eigenvalue is less than 1.0
xxxxx = variable's largest correlation (bold) in function which accounts for less than 5% of variance or eigenvalue is less than 1.0, and falls below +/- .33
xxxxx = variable fails .05 significance in F to remove (a harsh test indicating variable is a possible candidate for elimination)
 427

Low Capability Organisation

Rotated Potency index Variables Entered/Removed
Structure Matrix
COAL Function
1 2 F Statistic
Language (country-specific) 0.904 0.276 Language (country-specific) 77.3 Language (country-specific) 409.5
Legal fairness 0.235 0.795 Legal fairness 8.9 Legal fairness 144.6
Market risk 0.217 0.772 Market risk 8.0 Market risk 106.8
Competition 0.203 0.666 Competition 6.5 Cultural distance 80.9
1st Economic PC 0.145 0.563 Cultural distance 4.7 Product consumption 64.7
Cultural distance 0.196 0.434 1st Economic PC 3.9 1st Economic PC 53.4
Product consumption -0.042 -0.084 Product consumption 0.2 Product consumption growth 45.1
Product consumption growth -0.014 -0.039 Product consumption growth 0.0 Competition 38.6

Eigenvalue 25.7 1.6

% of variance 94.1 5.9

BEEF Function
1 2
Language (country-specific) 0.962 0.198 Language (country-specific) 88.5 Language (country-specific) 1507.2
Legal fairness 0.265 0.900 Legal fairness 10.3 Legal fairness 494.5
Competition 0.226 0.640 Competition 6.7 Cultural distance 333.9
Market risk 0.195 0.624 Market risk 5.4 1st Economic PC 252.9
1st Economic PC 0.191 0.583 1st Economic PC 5.0 Product consumption 203.1
Cultural distance 0.117 0.393 Cultural distance 2.0 Competition 168.9
Product consumption growth 0.005 -0.062 Product consumption 0.0 Product consumption growth 144.1
Product consumption 0.018 0.042 Product consumption growth 0.0 Market risk 125.5

Eigenvalue 19.0 0.9

% of variance 95.5 4.5
 428

Low Capability continued

Rotated Structure Matrix Potency index Variables Entered/Removed
WOOL Function
1 2 3 F Statistic
Language (country-specific) 0.926 0.035 0.064 Language (country-specific) 79.7 Language (country-specific) 1338.5
Legal fairness -0.029 0.810 -0.082 Legal fairness 4.4 Legal fairness 342.9
Competition 0.039 0.599 0.087 Competition 2.5 1st Economic PC 197.7
Market risk 0.002 0.593 -0.259 Market risk 2.3 Cultural distance 139.4
1st Economic PC 0.023 0.552 0.272 1st Economic PC 2.1 Product consumption 109.2
Cultural distance 0.018 0.338 0.009 Cultural distance 0.8 Product consumption growth 89.1
Product consumption 0.023 -0.034 0.562 Product consumption 0.2 Market risk 75.3
Product consumption growth -0.011 -0.015 -0.189 Product consumption growth 0.0 Competition 66.0

Eigenvalue 42.4 3.0 0.2

% of variance 93.0 6.6 0.4

TELEPHONES Function
1 2 3
Language (country-specific) 0.937 0.074 -0.032 Language (country-specific) 78.8 Language (country-specific) 826.3
Legal fairness 0.033 0.831 -0.171 Legal fairness 6.8 Legal fairness 249.9
Market risk 0.029 0.628 -0.136 Market risk 3.9 1st Economic PC 150.7
Competition 0.063 0.577 0.152 Competition 3.6 Cultural distance 113.1
1st Economic PC 0.011 0.487 0.338 1st Economic PC 2.4 Product consumption 88.1
Product consumption -0.010 -0.017 0.642 Cultural distance 0.9 Market risk 72.3
Cultural distance -0.021 0.287 0.345 Product consumption 0.3 Competition 61.1
Product consumption growth 0.004 0.023 -0.169 Product consumption growth 0.0 Product consumption growth 52.6

Eigenvalue 25.3 2.7 0.2

% of variance 89.7 9.7 0.6
 429

Low Capability continued Rotated Structure Matrix Potency index Variables Entered/Removed
Function
EDUCATION 1 2 F Statistic
Language (country-specific) 0.947 0.211 Language (country-specific) 85.7 Language (country-specific) 1422.3
Legal fairness 0.216 0.884 Legal fairness 8.1 Legal fairness 451.2
Competition 0.201 0.682 Competition 6.0 Cultural distance 307.4
Market risk 0.168 0.628 Market risk 4.5 Competition 234.5
1st Economic PC 0.172 0.598 1st Economic PC 4.5 Product consumption 188.9
Cultural distance 0.109 0.474 Cultural distance 2.2 1st Economic PC 157.8
Product consumption 0.020 0.090 Product consumption 0.1 Product consumption growth 135.2
Product consumption growth 0.008 0.075 Product consumption growth 0.0 Market risk 118.2

Eigenvalue 18.9 0.9

% of variance 95.4 4.6

Education: Low capability organisation with objectives

Function
1 2
Product consumption 0.983 0.043 Product consumption 95.1 Product consumption 1201.6
Cultural distance 0.000 0.692 Cultural distance 0.7 Cultural distance 269.9
Competition 0.047 0.407 Competition 0.5 Competition 181.3
Market risk 0.036 0.288 1st Economic PC 0.3 Market risk 138.4
1st Economic PC 0.049 0.277 Market risk 0.2 1st Economic PC 111.6
Legal fairness 0.005 0.271 Legal fairness 0.1 Language 93.6
Product consumption growth -0.003 0.199 Product consumption growth 0.1 Legal fairness 80.6
Language -0.011 0.172 Language 0.1 Product consumption growth 70.7

Eigenvalue 13.1 0.2

% of variance 98.6 1.4
 430

Low Capability continued Rotated Structure Matrix Potency index Variables Entered/Removed

ENGINES Function
1 2
Language (country-specific) 0.919 0.183 Language (country-specific) 79.8 Language (country-specific) 376.1
Legal fairness 0.210 0.750 Competition 7.6 Legal fairness 113.3
Market risk 0.221 0.719 Market risk 7.6 Market risk 86.2
Competition 0.240 0.613 Legal fairness 7.4 Cultural distance 65.2
1st Economic PC 0.149 0.538 1st Economic PC 3.8 Product consumption growth 52.3
Cultural distance 0.116 0.310 Cultural distance 1.8 Competition 43.3
Product consumption 0.073 0.207 Product consumption 0.7 Product consumption 36.7
Product consumption growth 0.022 -0.178 Product consumption 0.2 1st Economic PC 31.4

Eigenvalue 27.9 1.7

% of variance 94.2 5.8

Bold = largest absolute pooled within-groups correlations between a Variable and any Discriminant function and correlation > +/- .33
xxxxx = variable falls below +/- .33 correlation xxxx = variable below .19 tolerance, indicating possible multicollinearity
xxxxx = variable's largest correlation (bold) in function which accounts for less than 5% of variance or whose eigenvalue is less than 1.0
xxxxx = variable's largest correlation (bold) in function which accounts for less than 5% of variance or eigenvalue is less than 1.0, and
falls below +/- .33
xxxxx = variable fails .05 significance in F to remove (a harsh test indicating variable is a possible candidate for elimination)
 431

High Capability Organisation

Rotated Structure Matrix Potency index Variables Entered/Removed
COAL Excludes China Function
1 2 3 F Statistic
Competition 0.753 -0.114 0.219 Competition 38.3 Competition 59.2
1st Economic PC 0.646 -0.066 0.070 1st Economic PC 27.8 Cultural distance 42.0
Market risk 0.560 0.017 0.316 Market risk 21.5 Language (country-specific) 33.8
Legal fairness 0.531 0.054 -0.261 Legal fairness 19.3 Legal fairness 30.1
Market research (country-specific) -0.072 0.036 -0.063 Cultural distance 9.1 Contacts (country-specific) 25.1
Cultural distance 0.171 0.524 -0.073 Language (country-specific) 3.1 Product consumption growth 22.0
Language (country-specific) 0.114 -0.265 -0.233 SCA (country-specific) 1.6 Market research (country-specific) 20.1
Product consumption 0.053 -0.080 0.022 Product consumption growth 1.1 SCA (country-specific) 18.6
SCA (country-specific) -0.001 0.033 0.449 Contacts (country-specific) 0.5 1st Economic PC 17.0
Product consumption growth 0.089 -0.039 0.266 Market research (country-specific) 0.4 Market risk 15.5
Contacts (country-specific) 0.048 -0.091 0.132 Product consumption 0.4 Product consumption 13.9

Eigenvalue 10.3 4.0 1.2

% of variance 66.2 25.9 7.9
 432

High Capability continued Rotated Structure Matrix Potency index Variables Entered/Removed
BEEF Function
1 2 F Statistic
SCA (country-specific) 0.930 0.038 SCA (country-specific) 60.9 SCA (country-specific) 309.0
Product consumption growth 0.076 0.018 Legal fairness 19.3 Legal fairness 210.9
Language (country-specific) 0.048 0.025 1st Economic PC 16.1 1st Economic PC 172.2
Contacts (country-specific) -0.028 0.023 Competition 14.5 Cultural distance 133.5
Legal fairness 0.162 0.769 Market risk 8.9 Product consumption 107.6
1st Economic PC 0.130 0.709 Cultural distance 1.0 Language (country-specific) 90.1
Competition 0.127 0.673 Product consumption growth 0.4 Contacts (country-specific) 77.6
Market risk 0.091 0.531 Product consumption 0.4 Market research (country-specific) 68.2
Cultural distance -0.044 0.169 Language (country-specific) 0.2 Market risk 60.4
Product consumption 0.004 0.110 Market research (country-specific) 0.1 Competition 54.6
Market research (country-specific) 0.017 0.059 Contacts (country-specific) 0.1 Product consumption growth 49.5

Eigenvalue 4.3 1.8

% of variance 70.4 29.6
 433

High Capability continued Rotated Structure Matrix Potency index Variables Entered/Removed
WOOL Function
1 2 3 4 F Statistic
1st Economic PC 0.789 0.021 -0.049 .200 1st Economic PC 33.5 Contacts (country-specific) 75.2
Competition 0.729 0.074 -0.036 .039 Competition 28.6 1st Economic PC 68.4
Market risk 0.590 0.002 -0.021 .009 Contacts (country-specific) 26.9 Legal fairness 54.6
Contacts (country-specific) -0.073 0.897 -0.168 .010 Market risk 18.7 SCA (country-specific) 49.9
Product consumption 0.034 0.080 0.034 .045 Legal fairness 13.4 Language (country-specific) 39.1
Language (country-specific) 0.288 0.027 -0.626 -.020 Language (country-specific) 8.7 Product consumption growth 32.4
Legal fairness 0.403 0.133 0.614 -.083 SCA (country-specific) 2.3 Market research (country-specific) 27.6
Cultural distance 0.018 0.044 0.452 .024 Cultural distance 2.3 Cultural distance 24.1
Product consumption growth 0.033 0.098 -0.200 .015 Product consumption growth 0.8 Market risk 21.4
SCA (country-specific) -0.019 -0.022 0.088 .882 Market research (country-specific) 0.7 Competition 19.2
Market research (country-specific) -0.079 -0.085 0.057 -.224 Product consumption 0.3 Product consumption 17.5

Eigenvalue 5.2 3.2 1.0 0.3

% of variance 53.5 32.8 10.9 2.8
 434

High Capability continued Rotated Structure Matrix Potency index Variables Entered/Removed
TELEPHONES Function
1 2 F Statistic
Legal fairness 0.739 -0.127 Legal fairness 39.3 Legal fairness 115.3
Competition 0.697 0.185 Competition 35.5 Language (country-specific) 63.2
Market risk 0.630 0.030 Market risk 28.2 Competition 56.7
1st Economic PC 0.538 0.075 1st Economic PC 20.7 Cultural distance 47.2
Product consumption 0.147 -0.120 Language (country-specific) 9.7 SCA (country-specific) 41.8
Language (country-specific) -0.007 0.579 Cultural distance 7.9 Contacts (country-specific) 37.8
Cultural distance 0.190 -0.429 SCA (country-specific) 2.9 Product consumption 34.2
SCA (country-specific) 0.004 0.314 Product consumption 2.0 Market risk 30.6
Contacts (country-specific) 0.001 -0.201 Contacts (country-specific) 1.2 Market research (country-specific) 27.5
Product consumption growth 0.033 0.108 Product consumption growth 0.4 Product consumption growth 24.7
Market research (country-specific) 0.029 0.043 Market research (country-specific) 0.1 1st Economic PC 22.4

Eigenvalue 3.6 1.5

% of variance 71.0 29.0
 435

High Capability continued Rotated Structure Matrix Potency index Variables Entered/Removed
EDUCATION Function
1 2 3 4 F Statistic
Contacts (country-specific) 0.932 -0.124 -0.108 0.108 Contacts (country-specific) 41.5 Contacts (country-specific) 117.0
Competition -0.045 0.808 -0.002 -.148 Competition 20.6 Legal fairness 95.5
Market risk -0.099 0.778 0.077 -.191 Market risk 19.6 SCA (country-specific) 88.6
1st Economic PC -0.078 0.689 -0.090 -.023 Legal fairness 15.4 Market risk 73.2
Legal fairness -0.060 0.683 0.009 .337 1st Economic PC 15.2 Cultural distance 61.4
SCA (country-specific) -0.123 -0.064 0.921 .001 SCA (country-specific) 14.2 1st Economic PC 53.2
Cultural distance -0.049 0.201 0.079 .544 Cultural distance 3.3 Language (country-specific) 46.9
Language (country-specific) -0.035 0.181 0.016 -.535 Language (country-specific) 2.9 Product consumption 41.3
Product consumption 0.039 0.046 0.038 .247 Product consumption 0.5 Market research (country-specific) 36.8
Product consumption growth -0.017 0.042 0.035 -.123 Market research (country-specific) 0.5 Competition 33.0
Market research (country-specific) 0.076 -0.051 -0.041 .108 Product consumption growth 0.2 Product consumption growth 29.7

Eigenvalue 3.8 2.6 1.3 0.5

% of variance 46.9 31.1 15.8 6.2
 436

High Capability continued Rotated Structure Matrix Potency index Variables Entered/Removed
Education with objectives Function F Statistic
1 2
Product consumption 0.962 0.157 Product consumption 91.1 Product consumption 1201.6
SCA 0.048 -0.014 Cultural distance 1.2 Cultural distance 269.9
Cultural distance -0.070 0.669 Language 0.3 Competition 181.3
Competition 0.005 0.400 Competition 0.2 Market risk 138.4
Market risk 0.007 0.283 SCA 0.2 1st Economic PC 111.6
1st Economic PC 0.020 0.273 1st Economic PC 0.2 Language 93.9
Legal fairness -0.022 0.262 Legal fairness 0.2 Legal fairness 81.0
Language 0.047 -0.196 Market risk 0.1 Product consumption growth 71.0
Product consumption growth -0.023 0.192 Product consumption growth 0.1 Contacts 63.1
Market research -0.016 0.146 Market research 0.1 Market research 56.8
Contacts -0.001 -0.062 Contacts 0.0 SCA 51.5

Eigenvalue 13.4 0.2

% of variance 98.5 1.5
 437

High Capability continued Rotated Structure Matrix Potency index Variables Entered/Removed
ENGINES Function
1 2 F Statistic
Product consumption 0.761 0.212 Product consumption 44.9 Product consumption 77.2
Product consumption growth -0.260 0.003 Legal fairness 12.1 Legal fairness 45.4
Language (country-specific) 0.118 -0.055 1st Economic PC 10.7 1st Economic PC 34.8
Legal fairness -0.126 0.668 Market risk 6.5 Market risk 28.4
1st Economic PC 0.014 0.661 Product consumption growth 5.1 Product consumption growth 24.2
Market risk -0.049 0.510 Competition 4.9 Contacts (country-specific) 21.5
Competition 0.047 0.439 SCA (country-specific) 2.4 Competition 18.8
Cultural distance -0.071 0.245 Cultural distance 1.9 SCA (country-specific) 16.6
SCA (country-specific) -0.147 0.180 Language (country-specific) 1.1 Cultural distance 14.6
Market research (country-specific) -0.074 0.088 Market research (country-specific) 0.6 Market research (country-specific) 12.8
Contacts (country-specific) -0.040 -0.066 Contacts (country-specific) 0.2 Language (country-specific) 11.3

Eigenvalue 7.3 2.4

% of variance 75.6 24.4

Bold = largest absolute pooled within-groups correlations between a Variable and any Discriminant function and correlation > +/- .33
xxxxx = variable falls below +/- .33 correlation xxxx = variable below .19 tolerance, indicating possible multicollinearity
xxxxx = variable's largest correlation (bold) in function which accounts for less than 5% of variance or whose eigenvalue is less than 1.0
xxxxx = variable's largest correlation (bold) in function which accounts for less than 5% of variance or eigenvalue is less than 1.0, and
falls below +/- .33
xxxxx = variable fails .05 significance in F to remove (a harsh test indicating variable is a possible candidate for elimination)

Source: Constructed for this research from the SPSS MDA rotated structure matrices, potency calculations based on same, and
Variables Entered/Removed tables.
 438

BIBLIOGRAPHY

Akhter, Syed H. (1994) Global Marketing: Concepts, Strategies and

Practice, South-Western College Publishing, Cincinnati
Albaum, Gerald and Peterson, Robert A. (1984) Empirical Research in
International Marketing: 1976-1982, Journal of International Business
Studies 1 (Spring/summer) pp 161-173
American Metals Market Online (1999) Cahner’s Business Information
Service World steel statistics, by country, 1997.
www.amm.com/ref/90120isi.htm 3 March, 1999
Amine, Lyn S. and Cavusgil, S. Tamer (1986) Export Marketing Strategies in
the British Clothing Industry, European Journal of Marketing 20 (7)
pp 21-33
Anderberg, Michael R. (1973) Cluster Analysis for Applications, Academic
Press, New York
Andersen, Poul H. and Strandskov, Jesper (1998) International Market
Selection: A Cognitive Mapping Perspective, Journal of Global
Marketing 11 (3) pp 65-84
Anderson, E. and Gatignon, H. A. (1986) Modes of Entry: A Transaction
Cost Analysis and Propositions, Journal of International Business
Studies 3 (3) (Fall) pp 1-26
Athukorala, P., Jayasuriya, S., and Oczkowski, E. (1995) Multinational Firms
and Export Performance in Developing Countries: Some Analytical
Issues and New Empirical Evidence, Journal of Development
Economics 46 (1) (Feb) pp 109-122
Aulakh, Preet S. and Kotabe, Masaaki (1993) An Assessment of Theoretical
and Methodological Development in International Marketing: 1980-
1990, Journal of International Marketing 1 (2) pp 5-28
Aulakh, Preet S. and Kotabe, Masaaki (1997) Antecedents and
Performance Implications of Channel Integration in Foreign Markets,
Journal of International Business Studies 28 (1) pp 145-175
Aurifeille, J. M. and Quester, P. (1999) Clusterwise Regression: Concept
and Application, Australasian Journal of Market Research 7 (1)
AusIndustry (1996) Winning Exports: Export Action Workbook, Pitman,
Melbourne
Australian Bureau of Statistics (1993) Australian and New Zealand Standard
Industrial Classification, Catalogue No 1292.0, Canberra
Australian Bureau of Statistics (1996) Balance of Payments and
International Investment Position 1994-95, Canberra
Australian Bureau of Statistics (1998) Australian exports of specific products
by value and unit volume. Personal e-mail communications from Mr
Rod McCrea on 7 September, 1998 and Ms Julia Kirk on 30 December
1998 and 18 January, 1999
Australian Government Publishing Service (1995) Australia’s Export Coal
Industry: A Project of the Coal Promotion Program, Coal Branch,
Department of Primary Industries and Energy, 2nd Ed., Canberra
 439

Australian Trade Commission (1994) Intelligent Exports … and the Silent

Revolution in Services, Australian Trade Commission, Sydney
Australian Trade Commission (1996) Australian Trade Commission Annual
Report 1995-96, Australian Trade Commission, Sydney
Axelsson, B. and Johanson, J. (1992) Foreign Market Entry – the Textbook
vs the Network View, in Industrial Networks: A New View of Reality,
Axelsson and Johanson, Eds., pp 218-234, Routledge, New York
Balian, Edward S. (1994) The Graduate Research Guide Book: A Practical
Approach to Doctoral/Masters Research, 3rd Ed., University Press of
America, Lanham MD
Béliveau, Donald (1987) Export Market Selection Decisions for Small Firms:
A Role for Government, In Rosson, P. J. and Reid, S. D. (Eds.)
Managing Export Entry and Expansion: Concepts and Practice,
pp 159-185 Praeger, New York
Bennett, Roger (1995) International Marketing: Strategy, Planning, Market
Entry and Implementation, Kogan Page, London
Bilkey, Warren J. (1985) Development of Export Marketing Guidelines,
International Marketing Review 2 (1) (Spring) pp 31-40
Blight, Denis (1995) International Education: Australia’s Potential Demand
and Supply, Research paper presented to the 1995 International
Education Conference, Brisbane (October) IDP Education Australia,
Canberra
Bryant, Fred B. and Yarnold, Paul R. (1995) Principal Components Analysis
and Exploratory and Confirmatory factor Analysis, pp 99-136. In
Grimm, Laurence G. and Yarnold, Paul R. Reading and
Understanding Multivariate Statistics, American Psychological
Association, Washington, DC
Buzzell, Robert D., Quelch, John A. and Bartlett, Christopher A. (1995)
Global Marketing Management: Cases and Readings, 3rd Ed.,
Addison-Wesley, Reading
Cateora, Philip R. (1993) International Marketing, 8th Ed., Irwin, Burr Ridge,
Il
Cavusgil, S. Tamar and Li, Tiger (1991) International Marketing: An
Annotated Bibliography, American Marketing Association, Chicago
Cavusgil, S. Tamer; Kiyak, Tunga, and Aulakh, Preet S. (c1997)
Complimentary Approaches to Preliminary Foreign Market Opportunity
Assessment: Country Clustering and Country Ranking, draft working
paper, Michigan State University, East Lansing
Cheng, B. and Titterington, D. (1994) Neural Networks: A Review from a
Statistical Perspective, Statistical Science 9 (1) pp 2-54
Chetty, S. K. and Hamilton, R. T. (1993) Firm-level Determinants of Export
Performance: A Meta-analysis, International Marketing Review 10 (3)
pp 26-34
Chiesl, Newell E. and Lamb, Steven W. (1983) Segmenting International
Industrial Markets, Journal of the Academy of Marketing Science
11 (3) (Summer) pp 250-258
Coddington, Ian (1995) Secondary Data Sources for Australian Exporters,
Working Paper, UTS, Sydney
Craig, C. Samuel and Douglas, Susan P. (2000) International Marketing
Research, 2nd Ed., John Wiley, Chichester, West Sussex
 440

Crossborder Monitor (1995) Indicators of Market Size for 115 Countries,

3 (34) (August 30) pp 1-12
Czinkota, Michael R. and Ronkainen Ilkka A. (1995) International Marketing,
4th Ed., Harcourt Brace, Orlando
Czinkota, Michael R. and Ronkainen Ilkka A. (1998) International Marketing,
5th Ed., Harcourt Brace, Orlando
Czinkota, Michael R. and Ronkainen Ilkka A. (1996) Global Marketing,
Dryden, Fort Worth
Denis, Jean-Emile and Depelteau, Daniel (1985) Market Knowledge,
Diversification and Export Expansion, Journal of International
Business Studies 16 (3) (Fall) pp 77-90
Department of Employment, Education, Training and Youth Affairs (1997)
Overseas Student Statistics 1996, Canberra
Department of Foreign Affairs and Trade (1989) Australian Treaty List,
(1989 No 38), Canberra
Department of Foreign Affairs and Trade (1994) Australian Treaty List:
Supplement 1, (1994 No 41), Canberra
Department of Foreign Affairs and Trade (1997) Treaty Action 1997,
(1997 No 1), Canberra
Department of Foreign Affairs and Trade (1995) Composition of Trade
Australia 1994, Canberra
Department of Foreign Affairs and Trade (1996) Composition of Trade
Australia 1995, Canberra
Department of Foreign Affairs and Trade (1999) Composition of Trade
Australia 1998, Canberra
Department of Foreign Affairs and Trade (1997) Trade in Services Australia
1996-7, Canberra
Dickensheets, R. J. (1963) Basic and Economical Approaches to
International Marketing Research, Proceedings of the American
Marketing Association, Chicago pp 359-377
Dissertations Abstracts International database (via FirstSearch) 1861-5/1999
Doyle, Peter and Gidengil, Zeki B. (1977) A Strategic Approach to
International Market Selection, Proceedings of the American
Marketing Association pp 230-234
Dudley, L. and Montmarquette, C. (1976) A Model of the Supply of Bilateral
Foreign Aid, American Economic Review 66 (1) pp 132-142
Dunteman, George H. (1989) Principal Components Analysis, Sage,
Newbury Park, CA
Dutta, P. K. and Radner, R. (1994) Profit Maximisation and the Market
Selection Hypothesis, University of Wisconsin – Madison, Social
Systems Research Institute. Working Paper No 9425
Encyclopaedia Britannica, The New (1995) Encyclopaedia Britannica,
Chicago
Evangelista, Felicitas U. (1999) Export Information and the Role of
Marketing Research, Australasian Journal of Market Research 7 (2)
(July) pp 3-16
Fifield, Paul and Lewis, Keith (1995) International Marketing Strategy,
Butterworth Heinemann, Jordan Hill
Fletcher, R. (1996) The Role of Countertrade in the Internationalisation of
the Australian Firm, PhD Thesis, University of Technology, Sydney
 441

Fletcher, Richard (1992) Information Sources for International Business,

Working Paper, UTS, Sydney
Fletcher, Richard and Bohn, Jennifer (1998) The Impact of Psychic Distance
on the Internationalisation of the Australian Firm, Journal of Global
Marketing, (12) 2 pp 47-68
Gale Research Country & World Rankings Reporter 2nd Edition, (1997)
Brigitte T. Darnay Ed., New York
Gittins, Ross (2000) “Stand Back to Distinguish the Wood from the Trees”,
“Age, Business” p 5, 3 June, 2000
Goodnow, James D. (1985) Developments in International Mode of Entry
Analysis, International Marketing Review 2 (3) (Autumn) pp 17-30
Goodnow, James D. and Kudek, Donald R. (1997) Export Business
Planning for Small and Medium-Sized Companies, 4th Ed., Bradley
University, Peoria, Il
Gould, R. R. (1996) International Market Screening: A Conceptual
Framework, Proceedings of the Australian and New Zealand
Marketing Educators’ Conference, Auckland (November) 1 pp 34-53
Hardin, Alexis and Holmes, Leanne (1997) Services Trade and Foreign
Direct Investment, Industry Commission, Canberra
Hibbert, E. P. (1989) Marketing Strategy in International Business, McGraw
Hill, Maidenhead
Hodgson, Raphael W. and Uyterhoeven, Hugo E. (1962) Analyzing Foreign
Opportunities, Harvard Business Review 40 (March-April) pp 60-79
Ietto-Gillies, Grazia (1998) Measuring the Degree of Internationalisation:
Conceptual Frameworks and Empirical Analysis of Various Indices,
Proceedings of the 25th AIB Conference, London (April) pp 324-340
Ingram, Thomas N.; LaForge, Raymond W. and Schwepker, Charles H.
(1997) Sales Management: Analysis and Decision Making, 3rd Ed.,
Dryden, Fort Worth
Jeannet, Jean-Pierre and Hennessey, Hubert D. (1992) Global Marketing
Strategies, 2nd Ed., Houghton Mifflin, Boston
Jeannet, Jean-Pierre and Hennessey, Hubert D. (1995) Global Marketing
Strategies, 3rd Ed., Houghton Mifflin, Boston
Jeannet, Jean-Pierre; Gale, Christopher; Kashani, Kamran and Turpin,
Dominique (1995) Cases in International Marketing, 2nd Ed.,
Prentice-Hall, Englewood Cliffs
Johansson, J. K. and Moinpour, R. (1977) Objectives and Perceived
Similarity of Pacific Rim Countries, Columbia Journal of World
Business XII (Winter) pp 65-76
Johansson, Johny K. (1997) Global Marketing: Foreign Entry, Local
Marketing, and Global Management, Irwin, Chicago
Johnson, Bryan T.; Holmes, Kim R., and Kirkpatrick, Melanie (1998) 1998
Index of Economic Freedom, The Heritage Foundation and Dow
Jones & Company, Washington, DC and New York, NY respectively
Joint Coal Board (1991) Australian Black Coal Statistics, Sydney
Joint Coal Board (1995) Australian Black Coal Statistics, Sydney
Joint Coal Board (1995) 1995 Annual Report, Sydney
Journal of International Business Studies 25 Year Index 1969-94, 25 (4)
Journal of International Business Studies
Journal of International Marketing via WWW entry. 1 (1) 1993 to 3 (3) 1995
 442

Kalyanaram, G. and Krishnan, V. (1997) Deliberate Product Definition:

Customizing the Product Definition Process, Journal of Marketing
Research 34, May pp 276-285
Keegan, Warren J. (1995) Global Marketing Management, 5th Ed.,
Prentice-Hall, Englewood Cliffs
Keegan, Warren J. and Green, Mark C. (1997) Principles of Global
Marketing, Prentice-Hall, Upper Saddle River
Kendall, Ian (1996) Kendall’s Australian Directory: A Comprehensive
Directory of Occupational & Vocational Tests, Inventories &
Organisational Survey Questionnaires, PsychMetrics, Melbourne
Kirplani, V. H. and Kuan, Xu (1987) Effective International Market Potential
Assessment: China. In Wallendorf, Melanie and Anderson, Paul F.
(Eds.) Advances in Consumer Research Association for Consumer
Research 14 pp 398-402
Kramer, Roland Laird (1970) International Marketing, 3rd Ed., South-
Western, Cincinnati
Krause, Joyce H.; Wilson, William W. and Dooley, Frank J. (1995) Global
Market Segmentation for Value-Added Agricultural Products,
Agribusiness 11 (3) (May/June) pp 195-206
La Porta, Rafael; Lopez-de-Silane, Florencio; and Shleifer, Andrei (1996)
Law and Finance, NBER Working Paper No 5661
Lawley, Meredith and Blight, Denis (1997) International Students: Reasons
for Choice of an Overseas Destination, Research paper presented to
the 11th Australian International Education Conference, Melbourne
(September) IDP Education Australia, Canberra
Lensink, R. and Van Bergeijk, P. A. G. (1991) The Determinants of
Developing Countries’ Access to the International Capital Market,
Journal of Development Studies 28 (1) pp 86-103
Lock, Jody (1999) International Market Selection Methods Used by Export
Management Consultants, Honours thesis, University of Southern
Queensland, November, 1999
Luthans, Fred (1995) Organizational Behavior, 7th Ed., McGraw-Hill,
New York
Mahoney, D; Trigg, M; Griffin, R and Pustay, M (1998) International
Business: A Managerial Perspective, Longman, Melbourne
McGillivray, Mark and Oczkowski, Edward (1992) A Two-part Sample
Selection Model of British Bilateral Foreign Aid Allocation, Applied
Economics 24 pp 1311-1319
McGillivray, Mark (1992) The Human Development Index: Yet Another
Redundant Composite Development Indicator, Asian Journal of
Economics and Social Studies 11 (1) (January)
McGillivray, Mark and White, Howard (1993) Aid Principles and Policy: An
Operational Basis for the Assessment of Donor Performance, Institute
of Social Studies, The Hague Working Paper No 52
Michelsen, C. (2000) The Number of Clusters: Stopping Rules for Cluster
Analysis, e-mail communication on ELMAR, 7 April, 2000
Morisset, Jacques (1997 draft) Why Are Some Countries Investing Abroad
and Others Exporting? Evidence from OECD Countries
 443

Moyer, Reed (1968) International Market Analysis, Journal of Marketing

Research 5 (November) pp 353-360. Also (1973) International
Market Analysis, in Sethi, S. P. and Sheth, J. N. (Eds.) Multinational
Business Operations: Marketing Management, Goodyear Publishing,
Pacific Palisades, pp 163-177
Mueller, Dennis C. (1991) Entry, Exit, and the Competitive Process. In
Geroski, P. A. and Schwalbach, J., (Eds.) Entry and Market
Contestability: An International Comparison, Blackwell, Oxford
Nachum, L. (1994) The Choice of Variables for Segmentation of the
International Market, International Marketing Review 11 (3) pp 54-67
Narver, John C. and Slater, Stanley F. (1990) The Effect of a Market
Orientation on Business Profitability, Journal of Marketing (October)
pp 20-35
Nicolas, Stephen and Maitland, Elizabeth (1996) Comparative Market Entry:
Mode Choice and Clustering, University of Melbourne Working Paper
No 5 (November)
Norman, Neville (1997) The Lyncroft Report, Melbourne (audio tape)
OECD (1996) Geographical Distribution of Financial Flows to Aid
Recipients, Paris
O’Neill, Helen (1996) Globalisation, Competitiveness and Human Security:
Challenges for Development Policy and Institutional Change,
Conference of the European Association of Development Research &
Training Institutes, University of Economics & Business Administration,
Vienna
Onkvisit, Sak and Shaw, John J. (1993) International Marketing Analysis
and Strategy, 2nd Ed., Macmillan, New York
Onkvisit, Sak and Shaw, John J. (1997) International Marketing Analysis
and Strategy, 3rd Ed., Prentice Hall, Upper Saddle River
Organisation for Economic Co-operation and Development (1996) Services:
Statistics on International Transactions 1970-1993, Statistics
Directorate, OECD, Paris
Organisation for Economic Co-operation and Development (1997)
Globalisation and Small and Medium Enterprises, Volumes 1 & 2,
OECD, Paris
Paliwoda, Stanley J. and Thomas, Michael J. (1998) International Marketing,
3rd Ed., Butterworth Heinemann, Oxford
Papadopoulos, Nicolas and Jansen, Derek (1993) Target Countries for
Export, Department of Industry, Science & Technology, Ottawa
Papadopoulos, Nicolas and Rosson, Philip (1996) Inventory of Canadian
Contributions to Exporting, Carleton University, ISBG Occasional
Papers Series (August) Ottawa
Parmerlee, David (1993) Identifying the Right Market, NTC Business
Books, Lincolnwood
Quelch, John A.; Kashani, Kamran and Vandermerwe, Sandra (1994)
Cases in European Marketing Management, Irwin, Burr Ridge
Ramaswamy, Kannan; Kroeck, K. Galen and Renforth, William (1996)
Measuring the Degree of Internationalization of a Firm: A Comment,
Journal of International Business Studies 27 (1) (First Quarter)
pp 167-177
Ramond, C. (1974) The Art of Using Science in Marketing, Harper & Row,
 444

New York
Robinson, Richard D. (1967) International Management, Holt, Rinehart &
Winston, New York
Rochester, J. D. (1990) New Business Uses for Computing, I/S Analyzer
23 pp 1-16
Rokeach, Milton (1973) The Nature of Human Values, Free Press,
New York
Root, Franklin R. (1984) International Trade and Investment, 5th Ed.,
South Western, Cincinnati
Salvacruz, Joseph C., Reed, Michael R. and Mather, David (1992)
Research Reports: Food Demand in Other Countries – Market
Assessment Models for U. S. Agricultural Exports, Journal of Food
Distribution Research February pp 119-125
Salvacruz, Joseph and Reed, Michael (1993) Identifying the Best Market
Prospects for U. S. Agricultural Exports, Agribusiness 9 (1) pp 29-41
Samiee, S. and Walters, P. G. P. (1991) Segmenting Corporate Exporting
Activities: Sporadic versus Regular Exporters, Journal of the Academy
of Marketing Science 19 (Spring) pp 93-104
Science Applications International Corporation (1987) 10260 Campus Point
Drive, M/S 71, San Diego, CA
Seringhaus, F. H. Rolf (1987) Market Entry Decisions and Problems of
Industrial Exporters. In Leeflang, P. and Rice, M. (Eds.)
Contemporary Research in Marketing, Proceedings of the European
Marketing Academy, Toronto (June)
Sethi, S. P. and Curry, D. (1973) Variable and Object Clustering of Cross-
Cultural Data: Some Implications for Comparative Research and Policy
Formulation. In Sethi S. P. and Sheth, J. N. (Eds.) Multinational
Business Operations, Vol 2 Goodyear, Pacific Palisades, pp 19-49
Sethi, S. P. and Holton, R. H. (1969) Review of Bertil Liander et al.
Comparative Analysis for International Marketing, Journal of
Marketing Research 6 (November) pp 502-3
Sharda, R. (1994) Neural Networks for the MS/OR Analyst: An Application
Bibliography, Interfaces 24 (2) pp 116-130
Sheth, J. N. (1971) The Multivariate Revolution in Marketing Research,
Journal of Marketing, 35 (January), pp 13-19
Sheth J. and Eshghi A. (1989) Global Marketing Perspectives, South-
Western, Cincinnati
Sullivan, Daniel (1994) Measuring the Degree of Internationalization of a
Firm, Journal of International Business Studies 25 (2)
(Second Quarter) pp 325-342
Sullivan, Daniel (1996) Measuring the Degree of Internationalization of a
Firm: A Reply, Journal of International Business Studies 27 (1)
(First Quarter) pp 179-192
Sweetland, Richard C. and Keyser, Daniel J. (Editors) (1991) Tests, 3rd Ed.,
Pro-ed, Austin, TX
Towards an Export Initiative, (1997) Department of Trade and Industry,
United Kingdom
Transparency International and Gottingen University (1997) Internet
Corruption Perception Index, www uni-goettingen.de/~uwvw
Tryon, Robert C. and Bailey, Daniel E. (1970) Cluster Analysis, McGraw-
 445

Hill, New York

UNEP (United Nations Environment Programme) (1993) Register of
International Treaties and Other Agreements in the Field of the
Environment, UNEP, Nairobi
United Nations Statistical Office (1971) Recommendations for the 1973
World Programme of Industrial Statistics, Part 2, List of Selected
Products & Materials, United Nations, New York
United Nations Statistical Division (1993) Commodity Trade Statistics 1989,
Statistical Papers Series D Vol XXXIX, No 1-21 New York
United States Department of State (1998) Background Notes
Usunier, Jean-Claude (1993) International Marketing: A Cultural Approach,
Prentice-Hall, Hemel Hempstead, Hertfordshire
Usunier, Jean-Claude (1996) Marketing Across Cultures, 2nd Ed., Prentice-
Hall, Hemel Hempstead, Hertfordshire
Walsh, L. S. (1993) International Marketing, 3rd Ed., Pitman, London
Wedel, Michel and Kamakura, Wagner A. (1998) Market Segmentation:
Conceptual and Methodological Foundations, Kluwer, Boston
Weymes, Ed (1987) Six Steps to Success: A Strategy for Examining
Potential Markets in the U.S. Business Quarterly 52 (2) (Fall) pp 101-
103
White, Howard and McGillivray, Mark (1995) How Well is Aid Allocated?
Descriptive Measures of Aid Allocation: A Survey of Methodology and
Results, Development and Change 26 (1) (January) pp 163-183
World Almanac and Book of Facts 1996, The (1995) Robert Famighetti,
(Ed.), Funk & Wagnall, New Jersey
World Book Encyclopaedia (International), The (1994) World Book
International, London, V 10
World Economic Forum (1997) The Global Competitiveness Report 1997,
World Economic Forum, Geneva
World Outlook 1994, Economist Intelligence Unit, London
World Outlook 1995, Economist Intelligence Unit, London
World Outlook 1997, Economist Intelligence Unit, London
Yip, George S. (1995) Total Global Strategy: Managing for Worldwide
Competitive Advantage, Prentice-Hall, Englewood Cliffs
Young, S., Hamill, J., Wheeler, C., and Davies, J. R. (1989) International
Market Entry & Development: Strategies & Management, Harvester
Wheatsheaf, Hemel Hempstead
 446

REFERENCES

Aaby, Nils-Erik and Slater, Stanley F. (1989) Management Influences on

Export Performance: A Review of the Empirical Literature 1978-88,
International Marketing Review 6 (4) pp 7-26
Aaker, David A. (1995) Strategic Market Management, 4th Ed., Wiley,
New York
Adams, J. Stacey (1965) Inequity in Social Exchange. In L. Berkowitz (Ed.),
Advances in Experimental Social Psychology, 2 Academic, New York
Aguilar-Manjarrez, Rafael; Thwaites, Des and Maule, John (1997) Modelling
Sport Sponsorship Selection Decisions, Asia-Australia Marketing
Journal 5 (1) (December)
Albaum, Gerald; Strandskov, Jesper; Duerr, Edwin and Dowd, Laurence
(1994) International Marketing and Export Management, 2nd Ed.,
Addison-Wesley, Wokingham
Aldenderfer, Mark S. and Blashfield, Roger K. (1984) Cluster Analysis,
Sage, Beverly Hills, CA
Alderfer, Clayton (1972) An Empirical Test of a New Theory of Human
Needs, Organizational Behavior and Human Performance (May)
pp 142-175
Alexandrides, C. G. (1973) A Methodology for Computerization of
International Market Research. In International Business Systems
Perspectives, Georgia State University, Atlanta, GA pp 185-193
Alexandrides, C. G., and Moschis, George P. (1977) Export Marketing
Management, Praeger, New York
Allison, G. T. (1971) Essence of Decision: Explaining the Cuban Missile
Crisis, Little Brown, Boston
Armstrong, J. Scott (1970) An Application of Econometric Models to
International Marketing, Journal of Marketing Research 7 (May)
pp 190-198
Arora, A. and Fosfuri, A. (2000) Wholly Owned Subsidiary Versus
Technology Licensing in the Worldwide Chemical Industry, Journal of
International Business Studies 31 (4) pp 555-572
Asheghian, Parviz and Ebrahimi, Bahman (1990) International Business:
Economics, Environment and Strategies, Harper Collins, New York
Australian Bureau of Statistics and Australian Trade Commission (2000) A
Portrait of Australian Exporters: A Report Based on the Business
Longitudinal Survey, Catalogue No 8154.0, Canberra
Axinn, C. N. (1988) Export Performance: Do Managerial Perceptions Make a
Difference? International Marketing Review 5 (Summer) pp 61-71
Axinn, C. N. and Athaide, G. (1991) Export Commitment: Building
Conceptual Consensus, School of Management Working Paper,
Syracuse University, New York
Ayal, I. and Zif, J. (1978) Competitive Market Choice Strategies in
Multinational Marketing, Columbia Journal of World Business 13
(Fall), pp 72-81
Ball, Donald A., and McCulloch, Wendell H. Jr. (1982) International
Business: Introduction and Essentials, Business Publications, Plano
 447

Ball, Donald A., and McCulloch, Wendell H. Jr (1990) International

Business: Introduction and Essentials, 4th Ed., Irwin, Homewood, Il
Barnes, W. N. (1980) International Marketing Indicators, European Journal
of Marketing 14 pp 83-136
Barney, Jay B. and Griffin, Ricky W. (1992) The Management of
Organisations: Strategy, Structure, Behavior, Houghton Mifflin, Boston
Barrett, N. J. (1986) A Study of the Internationalisation of Australian
Manufacturing Firms, PhD Thesis, University of NSW, Sydney
Barrett, N. J. and Wilkinson, I. F. (1986) Internationalisation Behaviour:
Management Characteristics of Australian Manufacturing Firms by
Level of International Development. In P. W. Turnbull and S. J.
Paliwoda, Eds, Research in International Marketing, pp 213-233
Bartels, Robert (1963) Outline for Comparative Marketing Analysis. In
Comparative Marketing: Wholesaling in Fifteen Countries, Irwin,
Homewood, Il pp 299-308
Bartol, Kathryn; Martin, David; Tein, Margaret and Matthews, Graham (1998)
Management: A Pacific Rim Focus, 2nd Ed., McGraw-Hill, Sydney
Bartos, R. (1989) International Demographic Data? Incomparable!
Marketing and Research Today, 4 (4) pp 205-212
Bauerschmidt, A.; Sullivan, D. and Gillespie, K. (1985) Common Factors
Underlying Barriers to Export: Studies in the US Paper Industry,
Journal of International Business Studies, Fall pp 111-123
Benito, G. R. G. and Gripsrud, G. (1992) The Expansion of Foreign Direct
Investments: Discrete Rational Location Choices or a Cultural Learning
Process? Journal of International Business Studies 23 (Third
Quarter) pp 461-476
BETRO Trust Committee (1976) Concentration on Key Markets, British
Overseas Trade Board (BOTB), London
Bilkey, W. J. and Tesar, G. (1977) The Export Behaviour of Smaller-Sized
Wisconsin Manufacturing Firms, Journal of International Business
Studies 8 (1) pp 93-98)
Bilkey, Warren J. (1978) An Attempted Integration of the Literature on the
Export Behaviour of Firms, Journal of International Business Studies
9 (1) pp 33-46
Bodur, Muzaffer (1994) Foreign Market Indicators, Structural Resources and
Marketing Strategies as Determinants of Export Performance,
Advances in International Marketing 6 JAI Press, Greenwich
pp 183-205
Booth di Giovanni, Heather (1997) The Role of Government Trade
Promotion Organisations in Trade Promotion and Internationalisation
of Companies: The Case of the Department of Trade and Industry in
the United Kingdom, Conference paper at UNCTAD: Commission on
Enterprise, Business Facilitation and Development Meeting of Experts
on Government and Private Sector Roles and Interactions in SME
Developments, Geneva July, 1997
Bradley, Frank (1991) International Marketing Strategy, Prentice Hall,
Hemel Hempstead
Bradley, Frank (1995) International Marketing Strategy, 2nd Ed., Prentice-
Hall, Hemel Hempstead
Brealey, Richard and Myers, Stewart (1984) Principles of Corporate Finance
 448

2nd Ed., McGraw-Hill, Tokyo

Brewer, Paul (2000) Strategy, Statistics and Circumstance in International
Market Selection, Proceedings of the Australian & New Zealand
International Business Academy Conference (ANZIBA), Auckland
(October) pp 53-64
Brewer, Paul (2001) International Market Selection: Developing A Model
from Australian Case Studies, International Business Review 10
pp 155-174
Brouthers, K. D. and Brouthers, L. E. (2001) Examining the National Cultural
Distance Paradox, Journal of International Business Studies 32 (1)
(First quarter) pp 177-189
Bruner, Gordon and Hensel, Paul (1992) Marketing Scales Handbook: A
Compilation of Multi-Item Measures, American Marketing Association,
Chicago, Il
Bruner, Gordon and Hensel, Paul (1996) Marketing Scales Handbook: A
Compilation of Multi-Item Measures, Volume ΙΙ, American Marketing
Association, Chicago, Il
Buckley, Peter J. and Casson, James G. (1981) The Optimal Timing of a
Foreign Direct Investment, Economic Journal 91 (1) pp 75-87
Byrt, W. J. (1971) People and Organisations, McGraw-Hill, New York
Cannon, Tom and Willis, Mike (1986) How to Buy & Sell Overseas,
Hutchinson, London
Carlson, S., (1975) How Foreign is Foreign Trade? Acta Universitatis
Upsaliensis, Studia Oeconomiae Negotiorum 11, Uppsala
Carr, Richard P. Jr. (1978) Identifying Trade Areas for Consumer Goods in
Foreign Markets, Journal of Marketing (October) pp 76-80
Cateora, Philip R. (1996) International Marketing, 9th Ed., Irwin, Burr Ridge,
Il
Cavusgil, S. Tamer (1982a) Some Observations on the Relevance of Critical
Variables for Internationalization Stages. In M. R. Czinkota and G.
Tesar, Eds, Export Management: An International Context, pp 276-
285
Cavusgil, S. Tamar (1985) Guidelines for Export Market Research,
Business Horizons 28 (6) (Nov-Dec) pp 27-33
Cavusgil, S. Tamer and Zou, Shaoming (1994) Marketing Strategy-
Performance Relationship: An Investigation of the Empirical Link in
Export Market Ventures, Journal of Marketing 58 (January) pp 1-21
Chen, Homin and Chen, Tain-Jy (1998) Network Linkages and Location
Choice in Foreign Direct Investment, Journal of International Business
Studies 19 (3) (Third Quarter) pp 445-467
Churchill, Gilbert A. Jr. (1991) Marketing Research: Methodological
Foundations, 5th Ed., Dryden, Fort Worth
Churchill, Gilbert A. Jr. (1995) Marketing Research: Methodological
Foundations, 6th Ed., Dryden, Fort Worth
CIA World Factbook 1998, The (1996) http://www.odci.gov/cia/
publications/factbook/index.html
CIA World Factbook 1996-7, The (1996) Brassey’s, Washington
Collins Dictionary of the English Language (1979) William Collins, Sydney
 449

Comrey. A. L. and Lee, H. B (1992) A First Course in Factor Analysis,

2nd Ed., Erlbaum, Hillsdale, NJ
Conners, R. J. (1960) World Market Potentials as Developed for 3M’s
Overseas Operation. In Dynamic Marketing for a Changing World,
American Marketing Association, Chicago pp 461-466
Connolly, S. G. (1987) Finding, Entering and Succeeding in a Foreign
Market, Prentice-Hall, Englewood Cliffs
Cooke, P. (1972) Market Analysis Utilizing Cultural Anthropological
Indicators, European Journal of Marketing 6 (1) pp 26-34
Copi, Irving M. (1968) Introduction to Logic, 3rd Ed., Macmillan, London
Coplin, William D. and O’Leary, Michael K. (1994) The Handbook of Country
and Political Risk Analysis, International Business Communications,
East Syracuse
Craig-Lees, Margaret; Joy, Sally and Browne, Beverly (1995) Consumer
Behaviour, John Wiley & Sons, Brisbane
Cromley, Robert G., Hempel, Donald J. and Hillyer, Clayton L. (1993)
Dimensions of Market Attractiveness: Competitively Interactive Spatial
Models, Decision Sciences Journal 24 (4) pp 713-738
Cundiff, Edward W., and Hilger, Marye T. (1984) Marketing in the
International Environment, Prentice-Hall, Englewood Cliffs
Cuyvers, L.; De Pelsmacker P.; Rayp G. and Roozen, I. T. M. (1995) A
Decision Support Model for the Planning and Assessment of Export
Promotion Activities by Government Export Promotion Institutions –
the Belgian Case, International Journal of Research in Marketing 12
pp 173-186
Czinkota, Michael R. (1982) Export Development Strategies: U.S. Promotion
Policy, Praeger, New York
Daft, R. L. (1991) Management, Dryden, Florida cited in #88 pp 507, 888
Daniels, John D. and Radebaugh, Lee H. (1995) International Business
Environments and Operations, 7th Ed., Addison-Wesley, Reading
Davidson, William H. (1980) The Location of Foreign Direct Investment
Activity: Country Characteristics and Experience Effects, Journal of
International Business Studies (Fall) pp 9-22
Davidson, William H. (1983) Market Similarity and Market Selection:
Implications for International Marketing Strategy, Journal of Business
Research 11 (Dec) pp 439-456
Deschampsneufs, Henry (1967) Marketing Overseas, Pergamon Press,
Oxford
Diamantopoulos, A. and Schlegelmilch B. B. (1994) Linking Export
Manpower to Export Performance: A Canonical Regression Analysis of
European and U.S. Data. In S. T. Cavusgil (Series Ed.) C. Axinn
(Volume Ed.) Advances in International Marketing 6 pp 161-181
JAI Press, Greenwich
Dibb, Sally and Simkin, Lyndon (1996) The Market Segmentation Workbook:
Target Marketing for Marketing Managers, Routledge, London
Dichtl, Erwin; Köeglmayr, Hans-George and Müeller, Stefan (1990)
International Orientation as a Precondition for Export Success,
Journal of International Business Studies, 1 pp 23-40
Douglas, S.; Le Maire, P.; and Wind, Y. (1972) Selection of Global Target
Markets: A Decision-Theoretic Approach, Proceedings of the XXII
 450

ESOMAR Congress, September, France, pp 237-251

Douglas, S. P. and Craig, C. S. (1982) Information for International
Marketing Decisions, Handbook of International Business. In Walters,
I. And Murray, T. (Eds.), J. Wiley, New York pp 29-1 – 29-33
Douglas, S. P.; Craig, C. S. and Keegan, W. J. (1982) Approaches to
Assessing International Marketing Opportunities for Small and Medium
Sized Companies, Columbia Journal of World Business Fall
pp 26-32
Douglas, S. P. and Craig, C. S. (1983) International Marketing Research,
Prentice-Hall, Englewood Cliffs
Douglas, Susan P. and Craig, C. Samuel (1992) Advances in International
Marketing, International Journal of Research in Marketing 9 (1)
pp 291-318 (December)
Douglas, Susan P. and Craig, C. Samuel (1995) Global Marketing Strategy,
McGraw-Hill, Singapore
Downey, Lloyd; Watson, David and van Beurden, Irene (1987) Market
Intelligence Task Force Paper No 1 (First of the ‘MARIS Reports’),
Australian Trade Commission, Canberra. Unpublished internal report
Downey, Lloyd; Watson, David; Wasson, Merrilyn and van Beurden, Irene
(1988a) Market Intelligence Task Force Papers No 2, 3 & 4 (The
‘MARIS Reports’), Australian Trade Commission, Canberra.
Unpublished internal report
Downey, Lloyd; Watson, David; Wasson, Merrilyn and Neumann, Susan
(1988b) Market Intelligence Task Force Papers No 6, 7, 8, & 9 (The
‘MARIS Reports’), Australian Trade Commission, Canberra.
Unpublished internal report
Downey, Lloyd; Watson, David and Neumann, Susan (1988c) Market
Intelligence Task Force Paper No 14 (The ‘MARIS Reports’),
Australian Trade Commission, Canberra. Unpublished internal report
Dretske, F. I. (1981) Knowledge and the Flow of Information, MIT Press,
Cambridge, MA
Dunning, John H. (1980) Toward an Eclectic Theory of International
Production: Some Empirical Tests, Journal of International Business
Studies Spring/Summer pp 9-31
Dunning, John H. (1997) Re-evaluating the Benefits of Foreign Direct
Investment. In Dunning, John H. (Ed.) Alliance Capitalism and Global
Business, Routledge, London
Economist Intelligence Unit (1999) Arbitrariness of Judicial System in 60
Countries 1993-97: Average. 29 January, 1999 personal e-mail
communication from Mr Laza Kekic, economist, EIU, London
Ellis, Paul (2000) Social Ties and Foreign Market Entry, Journal of
International Business Studies 31 (3) (Third Quarter) pp 443-469
Erb, Claude B.; Harvey, Campbell R. and Viskanta, Tadas E. (1996)
Expected Returns and Volatility in 135 Countries, Financial Analysts
Journal Nov/Dec
 451

Evangelista, Felicitas U. (1994) Export Performance and its Determinants:

Some Empirical Evidence from Australian Manufacturing Firms,
Advances in International Marketing 6 JAI Press, Greenwich, CT
pp 207-229
Everitt, Brian S. (1993) Cluster Analysis, 3rd Ed., Arnold, London
Fletcher, Richard and Bohn, Jennifer (1996) The Role of Psychic Distance in
the Internationalisation of the Firm, Proceedings of the Academy of
International Business, South East Asian Regional Conference,
University of Otago, Dunedin (June)
Food and Agriculture Organization (1996) FAO Yearbook: Trade Vol 49,
1995, FAO Statistical Series No 132, United Nations, Rome
Food and Agriculture Organization (1997) FAO Yearbook: Production Vol
50, 1996, FAO Statistical Series No 135, United Nations, Rome
Food and Agriculture Organization, Statistics for beef imports, exports and
total production, boneless, 1990-96. Personal e-mail communication
from FAO’s Mr Orio Tampieri 9/11/1998
Food and Agriculture Organization, Statistics for wheat imports, exports and
total production, 1990 and 1995, from search of FAO database at their
www site
Food and Agriculture Organization, Statistics for wool imports, exports and
total production, 1990 and 1995, from search of FAO database at their
www site
Friedman, Milton (1957) A Theory of the Consumption Function, Princeton
University Press, Princeton, NJ
Global Outlook 1997, Economist Intelligence Unit, London 2nd quarter
Goodnow, James D. and Hansz, James E. (1972) Environmental
Determinants of Overseas Market Entry Strategies, Journal of
International Business Studies 3 (45) (Spring) pp 33-50
Gould, R. R. and McGillivray, M. (1998) Culture-Related Interactions During
International Market Screening, Proceedings of the Association of
International Business Conference, London (April) 1 pp 250-263
Gray, Jerry L. and Starke, Frederick A. (1988) Organizational Behavior:
Concepts and Applications, 4th Ed., Macmillan, New York
Green, Paul E. (1978) Analyzing Multivariate Data, Dryden, Hinsdale, Il
Green, Robert T. and Allaway, Arthur W. (1985) Identification of Export
Opportunities: A Shift-share Approach, Journal of Marketing 49 (1)
(Winter) pp 83-88
Hallén, L. and Weidersheim-Paul, F. (1989) The Evolution of Psychic
Distance in International Business Relationships. In Between Markets
and Hierarchy, Haag, I. and Weidersheim-Paul, F. Eds. University of
Uppsala pp 15-27
Hansz, James E. and Goodnow James D. (1972) A Multivariate
Classification of Overseas Country Market Environments, American
Marketing Association Conference Proceedings (Aug) pp 191-198
Hair, J. F. Jr; Anderson, R. E; Tatham, R. L. and Grablowsky, B. J. (1984)
Multivariate Data Analysis, Macmillan, New York
Hair, J. F. Jr; Anderson, R. E; Tatham, R. L. and Black, W. C. (1995)
Multivariate Data Analysis, 4th Ed., Prentice Hall, Englewood Cliffs
Harrigan K. R. (1985) Strategies for Joint Ventures, Heath & Co, Lexington
Hassan, S. S. and Blackwell, R. D. (1994) Global Marketing Perspectives
 452

and Cases (International Edition), Harcourt Brace, Orlando

Hassan, Salah S. and Samli, A. Coskun (1994) The New Frontiers of
Intermarket Segmentation. In Hassan, Salah S. and Blackwell, Roger
D. Global Marketing Perspectives and Cases, Harcourt Brace,
Orlando
Hawkins, Del I.; Best, Roger J. and Coney, Kenneth A. (1992) Consumer
Behavior: Implications for Marketing Strategy, Fifth Ed. Irwin, Burr
Ridge
Hennart, Jean-Francois and Larimo, Jorma (1998) The Impact of Culture on
the Strategy of Multinational Enterprises: Does National Origin Affect
Ownership Decisions? Journal of International Business Studies
29 (3) pp 515-538
Hofstede, ter, Frenkel – see ter Hofstede
Hofstede, Geert; Kraut, Allen I. and Simonetti, S. H. (1976) The
Development of a Core Attitude Survey Questionnaire for International
Use, Working Paper No 76-17, EIASM, Brussels
Hofstede, Geert (1980) Culture’s Consequences: International Differences in
Work-Related Values, Sage, Beverly Hills
Hofstede, Geert (1984) Cultural Dimensions in Management and Planning,
Asia Pacific Journal of Management 1 (2) (January) pp 81-99
Hofstede, Geert (1991) Cultures and Organisations: Software of the Mind,
McGraw-Hill, Maidenhead
Hofstede, Geert (1994) The Business of International Business is Culture,
International Business Review 3 (1) pp 1-14
Hofstede, Geert (2001) Culture’s Consequences: Comparing Values,
Behaviors, Institutions and Organizations Across Nations 2nd Ed.,
Sage, Thousand Oaks
Hofstede, Geert and Bond, Michael H. (1988) The Confucius Connection:
From Cultural Roots to Economic Growth, Organizational Dynamics
16 (4) pp 5-21
Holzmüller, Hartmut H. and Stöllnberger, B. A. (1994) A Conceptual
Framework for Country Selection in Cross-National Export Studies,
Advances in International Marketing 6 JAI Press, Greenwich pp 3-24
Holzmüller, Hartmut H. and Stöttinger, Barbara (1996) Structural Modelling
of Success Factors in Exporting: Cross-Validation and Further
Development of an Export Performance Model, Journal of
International Marketing 4 (2) pp 29-55
Huff, Lenard C. and Robinson, William T. (1994) The Impact of Leadtime
and Years of Competitive Rivalry on Pioneer Market Share
Advantages, Management Science 40 (10) pp 1370-1377
Huff, David L. and Sherr, Lawrence A. (1967) Measure for Determining
Differential Growth Rates of Markets, Journal of Marketing Research
4 (November) pp 391-395
Institute for Training in Intercultural Management (ITIM) Additional Hofstede
values received by mail 20/10/1999 from Niels Noorderhaven
International Energy Agency (1994) World Energy Outlook, Paris
International Energy Agency (1995) Oil, Gas & Coal Supply Outlook, Paris
International Energy Agency (1998) Coal Information 1997, Paris
 453

International Energy Agency (1998) Statistics on coal imports, exports and

local production by country, 1990 and 1995. Personal e-mail
communication from Ms Jocelyn Troussard 1/12/1998
International Energy Agency (1998) Statistics on coking coal imports,
exports and local production by country, 1990 and 1995. Personal
e-mail communication from Ms Jocelyn Troussard 12/12/1998
International Institute for Development Management (1998) The World
Competitiveness Yearbook 1998, IMD, Lausanne
Jackson, B B. (1973) Multivariate Data Analysis: An Introduction, Irwin,
Homewood, Il
Jain, Subhash C. (1990) International Marketing Management, 3rd Ed.,
PWS-Kent, Boston
Jain, Subhash C. (1993) International Marketing Management, 4th Ed.,
Wadsworth, Belmont
Jain, Subhash C. (1996) International Marketing Management, 5th Ed.,
South-Western, Cincinnati
Johanson, Jan and Vahlne, Jan-Erik (1977) The Internationalization Process
of the Firm – A Model of Knowledge Development and Increasing
Foreign Market Commitments, Journal of International Business
Studies Spring/Summer pp 23-32
Johanson, Jan and Gunnar-Mattsson, Lars (1986) International Marketing
and Internationalization Processes – A Network Approach. In Turnbull,
Peter W. and Paliwoda, Stanley J. Research in International
Marketing, Croom Helm, Beckenham, Kent pp 234-265
Johnson, Gerry and Scholes, Kevan (1993) Exploring Corporate Strategy:
Text and Cases, 3rd Ed., Prentice Hall, Hemel Hempstead
Keegan, Warren J. and MacMaster, Norman A. (1983) Global Strategic
Marketing. In Kirpalani, V. H. (Ed.) International Marketing:
Managerial Issues, Research and Opportunities, American Marketing
Association, Chicago pp 94-105
Kent, David H. (1993) Foreign Direct Investment Criteria: A Comparison of
Chinese and Japanese General Managers, Management Working
Paper Archive, University of Alaska
Kinnear, Thomas C.; Taylor, James R.; Johnson, Lester, and Armstrong,
Robert (1993) Australian Marketing Research, McGraw-Hill, Sydney
Kogut, Bruce and Singh, Harbir (1988) The Effect of National Culture on the
Choice of Entry Mode, Journal of International Business Studies
19 (3) pp 411-432
Kolasa, B. J. (1969) Introduction to Behavioural Science for Business, John
Wiley, New York
Koontz, Harold and O’Donnell, Cyril (1968) Principles of Management,
4th Ed., McGraw-Hill, New York
Kotler, Philip; Chandler, Peter C.; Brown, Linden and Adam, Stewart (1994)
Marketing: Australia and New Zealand, 3rd Australian Ed., Prentice
Hall, Sydney
Kotler, Philip (1997) Marketing Management: Analysis, Planning,
Implementation and Control, 9th International Ed., Prentice-Hall,
Upper Saddle River
 454

Kotler, Philip; Armstrong, Gary; Brown, Linden and Adam, Stewart (1998)
Marketing, 4th Ed., Prentice-Hall, Sydney
Kramer, Roland Laird (1964) International Marketing, 2nd Ed., South-
Western, Cincinnati
Kumar, V.; Stam, Antonie and Joachimsthaler, Erich A. (1994) An
Interactive Multicriteria Approach to Identifying Potential Foreign
Markets, Journal of International Marketing 2 (1) pp 29-52
Kwok, Chuck C. Y. and Reeb, David M. (2000) Internationalization and Firm
Risk: An Upstream-Downstream Hypothesis, Journal of International
Business Studies 31 (4) (Fourth quarter)
Lawler, Edward E. III (1973) Motivation in Work Organizations,
Brooks/Cole, Monterey, CA
Leonidou, Leonidas C. and Katsikeas, Constantine S. (1996) The Export
Development Process: An Integrative Review of Empirical Models,
Journal of International Business Studies 27 (3) (Third quarter)
pp 517-551
Lewin, Kurt (1936) A Dynamic Theory of Personality, McGraw-Hill, New
York
Liander, B.; Terpstra, V.; Yoshino, M. Y. and Sherbini, A. A. (1967)
Comparative Analysis for International Marketing, Allyn and Bacon,
Boston
Liang, Neng and Stump, Rodney L. (1996) Judgmental Heuristics in
Overseas Vendor Search & Evaluation: A Proposed Model of Importer
Buying Behavior, The International Executive 38 (6) pp 779- 806
New York
Lieberman, Marvin B. and Montgomery, David B. (1988) First Mover
Advantages, Strategic Management Journal (9) pp 41-58
Liesch, Peter W.; Welch, Lawrence S.; McGaughey, Sara L.; Welch, Denice;
Petersen, Bent and Lamb, Peter (2000) Reflections on Firm
Internationalisation Research: Old Theories and New Ideas,
Proceedings of the Australian & New Zealand International Business
Academy Conference (ANZIBA), Auckland (October)
Lim, Jenn-Sui; Sharkey, Thomas W. and Kim, Ken I. (1991) An Empirical
Test of an Export Adoption Model, Management International Review
31 (1) pp 51-62
Lintner, J. (1965) The Valuation of Risk Assets and the Selection of Risky
Investments in Stock Portfolios and Capital Budgets, Review of
Economics and Statistics, 47 (February) pp 13-37
Lipsey, Richard G.; Langley, Paul C. and Mahoney, Dennis M. (1986)
Positive Economics for Australian Students, 2nd Ed., Weidenfeld &
Nicolson, London
Litvak, Isaiah A. and Banting, Peter M. (1968) A Conceptual Framework for
International Business Arrangements. In Marketing and the New
Science of Planning, American Marketing Association Conference
Proceedings King, Robert L, Ed. Chicago, 1968 Fall Conference
pp 460-467
Locke, Edwin A. (1968) Toward a Theory of Task Motivation and Incentives,
Organizational Behavior and Human Performance, May 1978
pp 157-189
 455

Locke, E. A. (1976) The Nature and Causes of Job Satisfaction. In M.

Dunnette (Ed.), Handbook of Industrial and Organizational
Psychology, Rand McNally, Chicago
Loudon, David L. and Bitta, Albert J. (1993) Consumer Behavior, 4th Ed.,
McGraw Hill, New York
Luostarinen, Reijo (1980) Internationalization of the Firm, Helsinki School of
Economics, Helsinki
Luthans, Fred (1989) Organizational Behavior, 5th Ed., McGraw-Hill,
New York
Macquarie Dictionary, The (1981) Macquarie University, Sydney
Madsen, T. K. (1989) Successful Export Marketing Management: Some
Empirical Evidence, International Marketing Review (6) 4 pp 41-57
Madsen, Tage K. (1994) A Contingency Approach to Export Performance
Research. In S T Cavusgil (Series Ed.), C Axinn (Volume Ed.),
Advances in International Marketing, 6 JAI Press, Greenwich,
pp 25-42
Makino, S. and Neupert, K. E. (2000) National Culture, Transaction Costs,
and the Choice Between Joint Venture and Wholly Owned Subsidiary,
Journal of International Business Studies 31 (4) pp 705-713
Malhotra, Naresh K.; Hall, John; Shaw, Mike and Crisp, Mike (1996)
Marketing Research: An Applied Orientation, Prentice Hall, Sydney
Maslow, Abraham H. (1965) Eupsyhchian Management, Dorsey-Irwin,
Homewood, Il
Mayrhofer, Wofgang and Brewster, Chris (1996) In Praise of Ethnocentricity:
Expatriate Policies in European Multinationals, The International
Executive 38 (6) pp 749-778
McCarthy, E. Jerome; Perreault, William D. Jr.; Quester, Pascale G.;
Wilkinson, John W. and Lee, K. Y. (1994) Basic Marketing: A
Managerial Approach, 1st Australasian Ed., Irwin, Sydney
McCloskey, Donald N. (1982) The Applied Theory of Price, Macmillan New
York
McGillivray, M. and Gould, R. (1997) Selection Bias in International Market
Screening, Proceedings of the Australian & New Zealand Marketing
Educators’ Conference, Melbourne (December) 3 pp 1252-1261
Medina, J. F.; Saegert, J. and Merrifield, J. (1996) A Simple Export Country-
Market Screening Technique for Small US Businesses: A Packaged
Good Illustration, Journal of Business Strategies 13 (2) (Fall)
pp 126-150
Miesenbock, Kurt J. (1988) Small Business and Exporting: A Literature
Review, International Small Business Journal 6 (2) pp 42-61
Morosini, Piero; Shane, Scott and Singh, Harbir (1998) National Cultural
Distance and Cross-Border Acquisition Performance, Journal of
International Business Studies 19 (1) First Quarter pp 137-158
Mowday, R. T. (1987) Equity Theory Predictions of Behavior in
Organizations. In R. M. Steers and L. W. Porter (Eds.), Motivation and
Work Behavior, McGraw Hill, New York
 456

Munene, John C.; Schwartz, Shalom H. and Smith, Peter B. c1998 (press
details unknown) Cultural Dimensions Relevant to Development in
Sub-Saharan Africa. Personal communication from S. Schwartz by
e-mail, 1 October, 1998
Nasif, Ercan G.; Al-Daeaj, H.; Ebrahimi, B. and Thibodeaux, M. S. (1991/1)
Methodological Problems in Cross-Cultural Research: An Updated
Review, Management International Review 31 pp 79-91
National Trade Data Bank (1997) How Do I Determine Where to Export?
Stat-USA and U.S. Department of Commerce, Washington, DC
Newell, Allen and Simon, Herbert A. (1972) Human Problem Solving,
Prentice Hall, Englewood Cliffs, NJ
O’Grady, Shawna and Lane, Henry W. (1996) The Psychic Distance
Paradox, Journal of International Business Studies 27 (2)
pp 309-333
Ogwan, Tomson (1994) The Choice of Principle Variables for Computing the
Human Development Index, World Development 22 (12) pp 2011-
2014
Oxford Illustrated Dictionary (1962) Oxford University Press, London
Paliwoda, Stanley J. (1993) International Marketing, 2nd Ed., Butterworth-
Heinemann, Oxford
Pan, Yigang; Li, Shaomin and Tse, David K. (1999) The Impact of Order
and Mode of Market Entry on Profitability and Market Share, Journal
of International Business Studies 30 (1) (First Quarter) pp 81-103
Papadopoulos, Nicolas G. (1983) Assessing New Product Opportunities in
International Markets, New Product Developments, Amsterdam:
ESOMAR, Athens pp 69-89
Papadopoulos, Nicolas (1987) Approaches to International Market Selection
for Small and Medium Sized Enterprises. In Managing Export Entry
and Expansion: Concepts and Practice, Rosson, P. J. and Reid, S. D.
(Eds.), Praeger, New York pp 128-58
Papadopoulos, Nicolas, and Denis, Jean-Emile (1988) Inventory, Taxonomy
and Assessment of Methods for International Market Selection,
International Marketing Review (Autumn) pp 31-44
Papadopoulos, Nicolas and Jansen, Derek (1994) Country and Method-of-
Entry Selection for International Expansion: International Distributive
Arrangements Revisited, Dimensions of International Business No 11
Carlton University, Ottawa
Pavord, W. C. and Bogart, R. G. (1975) The Dynamics of the Decision to
Export, Akron Business and Economic Review Spring pp 6-11
Pedersen, Torben and Petersen, Bent (1998) Explaining Gradually
Increasing Resource Commitment to a Foreign Market, International
Business Review 7 (5) pp 483-501
Penrose, E. (1966) The Theory of the Growth of the Firm, B. Blackwell,
Oxford
Perkett, W. O (1963) An Analysis of the Obstacles to Increased Foreign
Trade which Confront British Columbia Industrial Machinery
Manufacturers, PhD dissertation, University of Washington, Seattle,
Washington
 457

Perlmutter, Howard V. (1969) The Tortuous Evolution of the Multinational

Corporation, Columbia Journal of World Business (Jan/Feb) pp 9-18
Pervin, Lawrence A. (1993) Personality Theory and Research, 6th Ed.,
Wiley, New York
Pervin, Lawrence A. and John, Oliver P. (1997) Personality Theory and
Research, 7th Ed., Wiley, New York
Peterson, Robin T. (1990) Screening is the First Step in Evaluating Foreign
Markets, Marketing News 24 (14) (9 July) p 13
Porter, Lyman W. and Lawler, Edward E. III (1968) Managerial Attitudes and
Performance, Irwin, Homewood, Il
Porter, Michael E. (1980) Competitive Strategy: Techniques for Analyzing
Industries and Competitors, Free Press, New York
Porter, Michael E. (1985) Competitive Advantage: Creating and Sustaining
Superior Performance, Free Press, New York
Porter, Michael (1990) The Competitive Advantage of Nations, Macmillan,
London
Power Systems Research (1999) petrol engine data (production, imports
and exports, by country, 1990 & 1995. Personal e-mail communication
from Mr Brent Schulz, Vice President, 29 December, 1999
PRS Group Inc (1995) Composite country risk ratings (political, financial and
economic) for 130 countries as at November, 1995. Personal fax
communication from Mr Adrian Shute dated 28 January, 1999
Rahman, Syed H. (2000) Towards Developing an International Market
Selection Decision Framework, Proceedings of the Australian & New
Zealand Marketing Educators Conference, Gold Coast (December) pp
1029-1033
Rahman, Syed H. (2000) Business Operating Environmental Considerations
in International Market Selection Process: A Study of Successful
Australian International Businesses, Proceedings of the Australian &
New Zealand International Business Academics Conference
(ANZIBA), Auckland (October) pp 471-9
Ralston, David A.; Holt, David H.; Terpstra, Robert H, and Yu, Kai-Cheng
(1997) The Impact of National Culture and economic Ideology on
Managerial Work Values: A Study of the United States, Russia, Japan
and China, Journal of International Business Studies 28 (1)
(First Quarter) pp 177-207
Ram, R. (1982) Composite Indices of Physical Quality of Life, Basic Needs
Fulfilment, and Income: A Principal Component Representation,
Journal of Development Economics 11 (2) pp 227-247
Ramanujam, V. and Venkatraman, N. (1987) Planning System
Characteristics and Planning Effectiveness, Strategic Management
Journal 8 pp 453-468
Rangan, Subramanian (2000) Search and Deliberation in International
Exchange: Microfoundations to Some Macro Patterns, Journal of
International Business Studies 31 (2) (second quarter) pp 205-222
Rao, T. R. and Naidu, G. M. (1992) Are the Stages of Internationalisation
Empirically Supportable? Journal of Global Marketing 6 (1/2)
pp 147-170
 458

Reid, Stan D. (1981) The Decision-Maker and Export Entry and Expansion,
Journal of International Business Studies 12 (2) (Fall) pp 101-112
Reid, Stanley (1986) Migration, Cultural Distance and International Market
Expansion. In Turnbull, P. W. and Paliwoda, S. J. (Eds.) Research in
International Marketing, Croom Helm, Kent pp 22-34
Reuber, A. Rebecca and Fischer, Eileen (1997) The Influence of the
Management Team’s International Experience on the
Internationalization Behaviors of SMEs, Journal of International
Business Studies 28 (4) Fourth Quarter pp 807-825
Reynolds, Thomas J; Gengler, Charles E. and Howard, Daniel J. (1995) A
Means-End Analysis of Brand Persuasion Through Advertising,
International Journal of Research in Marketing 12 (October)
pp 257-266
Robertson, K. R. and Wood, V. R. (2001) The Relative Importance of
Information in the Foreign Market Selection Process, International
Business Review 10 (3)
Robinson, Richard D. (1978) International Business Management: A Guide
to Decision Making, 2nd Ed., Dryden, Hinsdale
Roethlisberger, F. J. and Dickson, W. J. (1939) Management and the
Worker: An Account of the Research Program Conducted by the
Western Electric Company, Hawthorne Works, Chicago, Harvard
University Press, Cambridge
Ronen, Simcha and Shenkar, Oded (1985) Clustering Countries on
Attitudinal Dimensions: A Review and Synthesis, Academy
Management Review 10 (3) pp 435-454
Root, Franklin R. (1982) Foreign Market Entry Strategies, Amacom,
New York
Root, Franklin R. (1987) Entry Strategies for International Markets,
Lexington Books, Lexington
Root, Franklin R. (1994) Entry Strategies for International Markets,
Lexington Books, New York
Ros, Maria; Schwartz, S. H. and Surkiss, S. (1999) Basic Individual Values,
Work Values, and the Meaning of Work, Applied Psychology: An
International Review 48 (1) pp 49
Roux, E. (1979) The Export Behaviour of Small and Medium Size French
Firms: The Role of Manager’s Profile. In L. G. Mattsson and F.
Weidersheim-Paul (Eds.), Recent Research in the Internationalisation
of Business, Uppsala
Russow, Lloyd Curtis (1989) Global Screening: the Preliminary Identification
of Existing Product Specific Market Potential Using Macroeconomic
and Demographic Factors, PhD dissertation Georgia State University
Russow, Lloyd C., and Solocha, Andrew (1993) A Review of the Screening
Process within the Context of the Global Assessment Process,
Journal of Global Marketing 7 (1) pp 65-85
Russow, Lloyd C. and Okoroafo, Sam C. (1996) On the Way Towards
Developing a Global Screening Model, International Marketing Review
13 (1) pp 46-64
 459

Sagiv, L. and Schwartz, S. H. (c1999) A New Look at National Culture:

Illustrative Applications to Role Stress and Managerial Behavior. In
Ashkanasy, N. N.; Wilderom, C. and Peterson, M. F. (Eds.) The
Handbook of Organizational Culture and Climate, Sage, Newbury
Park, CA
Samli, A. Coskun (1972) Market Potentials Can Be Determined at the
International Level, Australian Journal of Market Research Aug/Nov
pp 85-92
Samli, A. Coskun (1977) An Approach for Estimating Market Potential in
East Europe, Journal of International Business Studies 8 (2) pp 49-
53
Samli, A. Coskun and Hill, John S. (1998) Marketing Globally: Planning and
Practice, NTC Business Books, Lincolnwood, Il
Samuelson, Paul A.; Hancock, Keith and Wallace, Robert (1975)
Economics: Second Australian Edition, McGraw-Hill, Sydney
Sarkar, Mitrabarun and Cavusgil, Tamer S. (1996) Trends in International
Business Thought and Literature: A Review of International Market
Entry Mode Research: Integration and Synthesis, The International
Executive 38 (6) pp 825-847
Schoemaker, P. J. H. (1993) Strategic Decisions in Organisations: Rational
and Behavioural Views, Journal of Management Studies 30 pp 106-
124
Schiffman, Leon G. and Kanuk, Leslie, L. (1991) Consumer Behavior,
4th Ed., Prentice Hall, Englewood Cliffs
Schwartz, Shalom H. and Bilsky, Wolfgang (1990) Toward a Theory of the
Universal Content and Structure of Values: Extensions and Cross-
Cultural Replications, Journal of Personality and Social Psychology
58 pp 878-891
Schwartz, Shalom H. (1992) Universals in the Content and Structure of
Values: Theoretical Advances and Empirical Tests in 20 Countries,
Advances in Experimental Social Psychology 25 pp 1-65
Schwartz, Shalom H. (1994a) Beyond Individualism/Collectivism: New
Cultural Dimensions of Values. In U. Kim, H. C. Triandis, C.
Kagitcibasi, S-C. Choi, and G. Yoon (Eds.), Individualism and
Collectivism: Theory, Method and Applications pp. 85-119 Sage,
Newbury Park, CA
Schwartz, Shalom H. (1994b) Are There Universal Aspects in the Structure
and Contents of Human Values? Journal of Social Issues 50 (4)
pp 19-45
Schwartz, Shalom H. and Sagiv, Lilach (1995) Identifying Culture-Specifics
in the Content and Structure of Values, Journal of Cross-Cultural
Psychology 26 (1) pp 92-116
Schwartz, Shalom H. (1996) Value Priorities and Behavior: Applying a
Theory of Integrated Value Systems. In C. Seligman, J. M. Olson and
M. P. Zanna (Eds.), The Psychology of Values: The Ontario
Symposium 8 pp 1-24
Schwartz, Shalom H. and Ros, M. (1995) Values in the West: A Theoretical
and Empirical Challenge to the Individualism-Collectivism Cultural
Dimension, World Psychology (1) pp 91-122
 460

Schwartz, Shalom H. and Bardi, Anat (1997) Influences of Adaptation to

Communist Rule on Value Priorities in Eastern Europe, Political
Psychology 18 (2) pp 385-410
Schwartz, Shalom H. (1997) Values and Culture. In D. Munro, J. F.
Schumaker, and S. C. Carr (Eds.), Motivation and Culture, Routledge,
New York pp 69-84
Schwartz, Shalom H. 1 October, 1998 personal communication (e-mail)
which attached latest version of the Schwartz Value System cultural
questionnaire (altered from the Schwartz 1992 version)
Schwartz, Shalom H. 19 October, 1998 personal communication (e-mail)
which provided data on 6 African national cultural values
Schwartz, Shalom H. (1999) ‘A Theory of Cultural Values and Some
Implications for Work’, Applied Psychology: An International Review
48 (1) (January). In Press at 1 October, 1998 when e-mail version
received
Sekaran, Uma (1992) Research Methods for Business: A Skill-Building
Approach, 2nd Ed., John Wiley, New York
Seringhaus, F. H. Rolf (1986) The Role of Information Assistance in Small
Firms’ Export Involvement, International Small Business Journal 5 (2)
pp 26-36
Sethi, S. Prakash (1971) Comparative Cluster Analysis for World Markets,
Journal of Marketing Research 8 (August) pp 348-354
Sharpe, W. F. (1964) Capital Asset Prices: A Theory of Market Equilibrium
Under Conditions of Risk, Journal of Finance 19 (September)
pp 425-442
Sherbini, A. A. (1967) Classifying and Comparing Countries. In Liander, B.;
Terpstra, V.; Yoshino, M. Y. and Sherbini, A. A. Comparative Analysis
for International Marketing, Allyn and Bacon, Boston pp 57-173
Sheridan, John Vincent (1988) An Examination of International Market
Selection Methods in Small and Medium Sized Technology Firms, M.
Mgt dissertation, Carleton Uni., Canada
Sheth, J. N. and Lutz, R. J. (1973) A Multivariate Model of Multinational
Business Expansion. In Sethi, S. P. and Sheth, J. (Eds.) Multinational
Business Operations: Marketing Management, Goodyear Publishing,
Pacific Palisades pp 96-103
Shoham, Aviv; Rose, Gregory M. and Albaum, Gerald S. (1995) Export
Motives, Psychological Distance, and the EPRG Framework, Journal
of Global Marketing 8 (3/4) pp 9-37
Shoham, Aviv and Kropp, Fredric (1998) Explaining International
Performance: Marketing Mix, Planning and their Interaction, Marketing
Intelligence and Planning 16 (2) pp 114-123
Simmonds, K. and Smith, H. (1968) The First Export Order: A Marketing
Innovation, British Journal of Marketing Summer, 1968 pp 93-100
Simon, Herbert A. (1947) Administrative Behaviour, Macmillan, New York
Simon, Herbert A. (1997) Administrative Behaviour, 4th Ed., Free Press,
New York, pp 118-122
Simonin, Bernard L. (1999) Transfer of Marketing Know-How in International
Strategic Alliances: An Empirical Investigation of the Role and
Antecedents of Knowledge Ambiguity, Journal of International
Business Studies 30 (3) (Third quarter) pp 463-390
 461

Simpson, C. L. Jr and Kujawa, D. (1974) The Export Decision Process: An

Empirical Enquiry, Journal of International Business Studies Spring
pp 107-117
Sinai, C. C. (1970) An Investigation of Selected Characteristics of Export-
Participating Manufacturing Firms, DBA dissertation, University of
Washington, Seattle, Washington
Smith, Peter B. and Schwartz, Shalom H. (1997) Values. In J. W. Berry, M.
H. Segall and C. Kagitcibasi (Eds.), Handbook of Cross-Cultural
Psychology, Vol. 3, 2nd Ed., Allyn & Bacon, Boston pp 77-118
Snavely, W. P.; Weiner, P.; Ulbrich, H. H. and Enright, E. J. (1964) Export
Survey of the Greater Hartford Area, Vols 1 & 2, The University of
Connecticut, Storrs, Connecticut
Solloway, Graeme and Ashill, Nicholas (1995) An International Market
Selection Framework: A Case Study for a New Zealand Agribusiness
Industry, Australian Agribusiness Review 3 (1) (June)
Stanton, William J; Miller, Kenneth E. and Layton, Roger A (1994)
Fundamentals of Marketing, 3rd Australian Ed., McGraw-Hill, Roseville
SPSS Base 8.0 Applications Guide (1998) SPSS Chicago
SPSS Base 10.0 Applications Guide (1999) SPSS Chicago
Sveiby, Karl Erik (1997) The New Organizational Wealth: Managing and
Measuring Knowledge-Based Assets, Berrett-Koehler, San Francisco
Tabachnick, Barbara G. and Fidell, Linda S. (1983) Using Multivariate
Statistics, Harper & Row, New York
Tabachnick, Barbara G. and Fidell, Linda S. (1996) Using Multivariate
Statistics, 3rd Ed., Harper Collins, New York
ter Hofstede, Frenkel; Steenkamp, J-B E. M. and Wedel, Michel (1999)
International Market Segmentation Based on Consumer-Product
Relations, Journal of Marketing Research 36 (1) (February) pp 1-17
Terpstra V. and Sarathy R. (1994) International Marketing, 6th Ed., Dryden,
Fort Worth
Terpstra V. and Sarathy R. (1997) International Marketing, 7th Ed., Dryden,
Fort Worth
Tesar, G. (1975) Empirical Study of Export Operations Among Small and
Medium-Sized Manufacturing Firms, PhD dissertation, The University
of Wisconsin, Madison, Wisconsin
Törnroos, Jan-Åke, and Möller, Kristian E. (1993) The Cultural Factor and
the Formation of International Business Relationships in Industrial
Markets, Advances in International Marketing 5 pp 107-121
Toyne, Brian, and Walters, Peter G. P. (1993) Global Marketing
Management: A Strategic Perspective, 2nd Ed., Allyn & Bacon,
Needham Heights
UNCTAD/GATT (1968) The Compilation of Basic Information on Export
Markets, Geneva, Switzerland
United Nations Development Program (1998) Human Development Report,
OVP, New York
United Nations Development Program, national account statistics from WWW
February, 1999 Source: http://www.undp.org/hydro/indicators.html &
http://www.undp.org/hydro/98hdi3.htm
UNESCO Publishing & Bernan Press (1997) Statistical Yearbook 1996,
Paris
 462

UNESCO Division of Statistics (1998) Enrolment in Higher Education and

Number of Foreign Students, spreadsheet attached to e-mail from Ms
F. Tandart, 17 November, 1998
UNESCO Division of Statistics (1998) Foreign Students by Host Country
and Country of Origin, spreadsheet attached to personal e-mail
communication from Ms F. Tandart, 17 November, 1998
UNESCO Division of Statistics (1998) Foreign Students by Country of
Origin, Table 3.14 from the ‘Statistical Yearbook 1998’, a table of
imports and exports for the top 50 countries, attached to personal e-
mail communication from Ms F. Tandart, 11 December, 1998
UNESCO Web Site www.unesco.org, for statistics and documents
United Nations Department of Economic and Social Affairs (1996)
Demographic Yearbook, 48th Issue United Nations, New York
United Nations Statistical Office (1963) Commodity Indexes for the Standard
International Trade Classification, Revised, Series M No 38 (2
volumes) United Nations, New York
United Nations Statistical Office (1981) Commodity Indexes for the Standard
International Trade Classification, Revision 2, Series M No 38/Rev, (2
volumes) United Nations, New York, NY
United Nations Statistical Division (1994) Commodity Indexes for the
Standard International Trade Classification, Revision 3, Series M No
38/Rev 2, (2 volumes) United Nations, New York
United Nations Statistical Office (1971) Indexes to the International
Standard Industrial Classification of All Economic Activities,
2nd Revision, Series M No 4, Rev 2, Add. 1, United Nations, New York
United Nations Statistical Office (1990) International Standard Industrial
Classification of All Economic Activities, 3rd. Revision, Series M No 4,
Rev 3, United Nations, New York
United Nations Statistical Office (1998) List of Selected Products & Materials
in Terms of PRODCOM. Attached to personal e-mail from Mr V.
Romanovsky 22 December, 1998 United Nations, New York
United Nations Statistical Office (1998) Revised List of Industrial Products &
Materials (Draft). Personal e-mail communication from Mr V.
Romanovsky 22 December, 1998 United Nations, New York
United Nations Statistical Office (1999) World production statistics by
country for coal & beef, petrol engines, & telephones. Personal e-mail
communication from Mr V. Romanovsky dated 7 January, 1999,
16 December, 1998, and 23 January, 1999 respectively United
Nations, New York
United Nations Statistical Office (1999) World production statistics by
country for coal, beef, wool, petrol engines & telephones. Personal
e-mail communication from Mr J. Habr dated 4 February, 1999 United
Nations, New York
United Nations Statistical Office (1991) Industrial Commodity Statistics
1982-1991, New York
United Nations Statistical Office (1997) Industrial Commodity Statistics
Yearbook, Production Statistics 1986-1995, New York
United Nations Statistics Division (1986) Statistical Yearbook 1983/84
34th Issue, New York
United Nations Statistics Division (1997) Statistical Yearbook 1995,
 463

42nd Issue New York

Veiga, John F.; Lubatkin, Michael; Calori, Roland; Very, Philippe and Tung,
Alex (2000) Using Neural Network Analysis to Uncover the Trace
Effects of National Culture, Journal of International Business Studies,
31 (2) (second quarter) pp 223-238
Vernon, Raymond (1966) International Investment and International Trade
in the Product Cycle, Quarterly Journal of Economics (May) pp 190-
207
Vroom, Victor (1964) Work and Motivation, Wiley, New York
Walvoord, Wayne R (1980) Export Market Research, American Import
Export Bulletin May pp 82-91
Wells, Louis T. Jr (1968) A Product Life Cycle for International Trade?
Journal of Marketing July pp 1-6
Wood, Van R. and Robertson, Kim R. (2000) The Importance of Information
by Industry, by Country of Destination, and by Type of Export
Transaction, International Marketing Review 17 (1) pp 34-55
World Bank (1997) World Development Indicators on CD-ROM, Washington
DC
World Bank (1999) World Development Indicators on CD-ROM, Washington
DC
World Economic Forum (1998) The Global Competitiveness Report 1998,
World Economic Forum, Geneva
World Trade Organization (1996) The International Markets for Meat 1995/6
Geneva
World Trade Organisation (1999) World merchandise and services imports
and exports statistics, by country, in 1995. Source:
http://www.wto.org/wto/statis/stat.htm
Yandle, Bruce (1978) Identifying Brand Performance by Shift-Share
Analysis, Journal of the Academy of Marketing Science 6 (1) (Winter)
pp 126-137
Yu, Chwo-Ming Joseph (1990) The Experience Effect and Foreign Direct
Investment, Weltwirtschaftliches Archiv, 126 (3) pp 561-580
Zikmund, William G., (1994) Exploring Marketing Research, 5th Ed.,
Dryden, Fort Worth
Zou, Shaoming and Stan, Simona (1998) The Determinants of Export
Performance: A Review of the Empirical Literature Between 1987 and
1997, International Marketing Review 15 (5) pp 333-356
Zou, Shaoming and Myers, Matthew B. (1999) The R & D, Manufacturing
and Marketing Competencies and the Firm’s Global Competitive
Position: An Empirical Study, Journal of Global Marketing 2 (3) (July)