1042-2587-90-152$1.

50
Copyright 1990 by
Baylor University

Structure, Policy,
Services, and
Performance in
the Business
Incubator Industry
David N. Allen
Richard McCluskey

Business Incubators are one of the newest tools on the enterprise development scene;
nearly 400 are now In operation. A business Incubator Is a faclllty that provides affordable
space, shared office services, and business development assistance In an environment
conducive to new venture creation, survlval, and early-stage growth. This article Is a pre-
llmlnary examination of the relatlonshlps among Incubator structure, policy, services, and
performance. A value-added continuum model Is used to describe various kinds of Incuba-
tors and aspects of their operations. Managers of 127 Incubators were surveyed to examine
features of the value-added continuum. Surrogate measures for the concepts that anchor
either end of the continuum-property development and business development-are em-
plrlcally examined. Incubators are found to be poor real estate ventures. Age and size of
faclllty are found to be Important determinants of jobs created and firms graduated. Only
one other structure, policy, or services variable Is Important for explaining business devel-
opment outcomes.

"The business of America" is getting smaller. Large companies are downsizing

to increase competitiveness. In the wake of corporate retrenchment, small businesses
have assumed a prominent role in job creation, investment opportunity, and innovation,
and in so doing they have become the backbone of the future U.S. economy. Although
new business creation has dramatically increased in the past few decades, failure rates
remain high (Bates & Nucci, 1989); depending on macroeconomic conditions, about
half will last five years (Birch, 1987, p. 18). One approach that has gained wide attention
for increasing new firm survival is the business incubator. Business incubators provide
affordable space, shared support services, and business development assistance in an
environment conducive to enterprise creation, survival, and early-stage growth.
This article is a preliminary assessment of the business incubator industry in the
U.S. To conceptually ground the study, two different models are posited-one concern-
ing purpose and structure, the other concerning change. The models suggest how orga-
nizational characteristics and life cycle position of an incubator may affect incubator
program activities and performance. Data were collected from 127 incubators. The
objective of the analysis is to discover relationships between structure, policy, and
services variables with special attention on determining what, if anything, influences
incubator performance. Implications for the research are addressed as recommendations

Winter, 1990 61
for policymakers and other incubator stakeholders. Finally, directions for future research
are offered.

INDUSTRY BACKGROUND

No one is exactly certain when business incubators emerged on the economic land-
scape. Perhaps that is not an important concern because the component parts of business
incubators are not new to enterprise development. Incubators are multi-tenant buildings,
but multi-tenant buildings have existed for decades. Incubators offer tenants shared
office services, but telephone answering services, pooled equipment leasing, temporary
office employees, and so forth have been available for decades too. Incubators help to
fill in the knowledge and skill gaps of start-up team members, but new firms have always
relied on accountants, lawyers, and other professional business advisors to provide
needed expertise. What is new and distinct about incubators is that these features of
entrepreneurship occur at one location-a building or proximate buildings. The combi-
nation of these features is thought to create a synergistic effect that is unique to the
enterprise-development scene. The incubator can be more than a place where a new
company can minimize its start-up costs by accessing affordable space, shared services,
and business assistance. For many entrepreneurs the incubator becomes a place where
they overcome the loneliness of entrepreneurship (Gumpert & Boyd, 1984) by associ-
ating with start-up firms and service providers. These other groups are important to
entrepreneurs because they are people who understand their circumstances, want them to
succeed, provide positive peer role models, may eventually become trade partners, and
help them realize they do not have all the answers. Well-managed incubators not only
make growth more affordable, they help: set realistic company milestones through
graduated rent structures, create a polished professional image, and accelerate the net-
working process (Birley, 1985; Schermerhorn, 1980).
By January, 1990, 385 business incubators were operating in the U.S. (NBIA,
1990). The industry has been growing at a steady rate since the early 1980s (Figure 1).
By 1992 it is likely that about 500 incubators will be operating in the U.S. Various
industry watchers estimated that around 1000 incubators exist throughout the world.

The Existing Body of Knowledge

Although a few business incubators have been operating for slightly more than a
decade, it has not been until recently that one could identify a body of knowledge based
upon experience and research. The first incubator managers did not have detailed models
to assist them in designing and operating their facilities. Managers networked with each
other to gain insights, but they primarily invented as they went along. Essentially, they
figured out on their own what worked and what did not. A series of conferences in 1984
and 1985 sponsored by the U.S. Small Business Administration promoted the incubator
movement, but it left many attendees clamoring for comprehensive, high-quality infor-
mation. Around this time the National Business Incubation Association (NBIA) was
founded, and for its first few years it too primarily promoted the concept, although it did
conduct annual conferences and disseminate research reports. The first systematic body
of information on operating concepts and management procedures was compiled as a
manual in 1988 (COSCAA, 1988). The NBIA is gradually developing research capacity
to conduct analysis that goes beyond their first annual State of the Business Incubation
Industry report (NBIA, 1990).
Research conducted on incubators has slowly become integrated into the existing

62 ENTREPRENEURSHIP THEORY and PRACTICE

 Figure 1
Growth of the Incubator Industry

Pre 1980 National Business Incubator Association

Quarterly Roster of Incubators
Allen (1985)
1984
Smilor and Gill (1986)

1985

1986

1987

1988

1989

1990 385

0 100 200 300 400 500

- Number of Incubators

body of knowledge. Early academic studies (cited in the next section) were primarily
descriptive and without conceptual grounding. Nonacademic research, primarily con-
ducted by consultants for clients interested in incubators, mostly examined the feasibility
of a proposed incubator in a given area. Journalistic accounts of incubators tend to be
ancedotal and mainly summarize the descriptive record (Brown, 1989; Wexler, 1990).
Few truly evaluative studies of incubators exist (Allen & Bazan, 1988, 1989; ILGARD,
1989).

Review of Research on Incubator Performance

A handful of studies have become the body of knowledge on incubator performance.
In different ways these studies have examined incubators from facility, tenant, and
graduate perspectives. Some important findings derived from these studies that convey
a sense of incubator performance are listed below:
• Employment in tenant firms is generally low. One study placed the number of
employees in tenant firms at a median of 2.3 (Allen, 1985), while another study of older
incubators found the median size of graduate firms to be 9 employees (Campbell, 1987).
• "Incubator" periods are generally short. Founders of tenant firms said they
expected to remain in the incubator a median of 29 months (Allen, 1985), with actual
graduates having a median tenancy of two years (Smilor & Gill, 1986; Campbell, 1987).
• "Success" rates are generally favorable. According to one study, for every two
tenants that graduate, one discontinued operations while still a tenant (Allen, 1985);
according to another study, about 14% of the companies that were admitted discontinued
operations (Campbell, 1987).
• Local retention of graduates is generally high. Three major studies found that over
85% of the graduates either expected to relocate (Allen, 1985) or have relocated in the
local area (Smilor & Gill, 1986; Campbell, 1987).

Winter, 1990 63
 • Tenants generally underutilize professional business advisory services and, when
used, often evaluate them unfavorably. Only 40% of the graduate companies reported
using professional services (Campbell, 1987), and over half the tenants said the follow-
ing items available at the incubator had no influence on their business operations:
management knowledge, capital availability, market development, and business strategy
(Allen, 1985). Other research also showed that teaching planning methods or tools to
incubator tenants does not necessarily improve the level of planning (Fry, 1987).
The existing research presents a mixed bag of results concerning incubator perfor-
mance. For example, although success rates are perceived as favorable, firm growth is
not as exceptional as many incubator proponents suggest. This may be explained by
Birch's finding that "among smaller firms, older firms grow faster than young ones"
(1987, p. 26). Also seemingly contradictory to one unique aspect of incubators-on-site
business assistance-are the findings that incubator tenants do not respond to the pro-
fessional services offered in the facility. It could be claimed that because incubators are
relatively recent enterprise-development tools one cannot expect to see stunning perfor-
mance figures and that some inconsistencies will arise as a result of a lack of conceptual
clarity in the research. Overall, however, the performance evidence presented is im-
pressive enough to justify continued incubator development. One purpose of this article
is to increase conceptual clarity with the hope that future studies will have a firmer
foundation for empirical analysis.

THE BUSINESS INCUBATOR CONTINUUM

No two business incubators are alike. Different resources, needs, constraints, and
opportunities are likely to affect incubator missions, policy, services, and performance.
One concept unifies the seemingly disparate dimensions of the incubator industry-how
the facility adds economic value to stakeholders. The value-added continuum (Table 1)
is anchored at either end by two important incubator elements. On one end, an incubator
functions like a real estate development operation. On the other end, an incubator
functions more like an enterprise-development program. One can hypothesize that in-
cubators at the right side are more likely to add economic value to the local economy
(because they emphasize business development) than incubators at the left side of the
continuum. Because the continuum concept is believed to be critical to understanding
differences among incubators it serves as a fundamental motivation for the analysis.
Four organizational ideal types of incubators are posed for the continuum frame-
work. The four types represent a slightly different focus for incubator development and
operation. For-profit property development incubators primarily seek to capture real
estate appreciation (Nyrop, 1986). Non-profit development corporations (Pacholski,
1988) primarily focus on creating jobs and enhancing the entrepreneurial climate. Ac-
ademic incubators seek to commercialize university technology (Smilor, 1987) while at
the same time providing local economic development benefits. For-profit seed capital
incubators are mainly physical embodiments of the seed fund managers wanting to have
the firms in their portfolio located at one location so they can be given maximum
attention. Beyond these primary organizational objectives secondary objectives are iden-
tifiable (see Table 1), and many of these lower order incubator objectives commingle
with the objectives of other organizational types.
Beyond these four organizational types, two other types are worth mentioning. The
first-public-private partnerships-appears fairly frequently in the incubator industry.
This arrangement is a hybrid or combination of two or more of the previously mentioned
four organizational types. It is relatively common for different organizations in a part-

64 ENTREPRENEURSHIP THEORY and PRACTICE

 Table 1

The Business Incubator Continuum: Facility Objectives

Valoe Added Through
Real Estate Business Development
For-Profit Non-Profit For-Profit
Property Development Seed
Development Corporation Academic Capital
Incubators Incubators Incubators Incubators

PRIMARY Real. estate Job creation Faculty-Industry Capitalize

OBJEcnVE appreciation collaboration investment
opportunity
Sell proprietary Positive statement Commercialize
services to tenant of entrepreneurial university
potential research
SECONDARY Create opportunity Generate Strengthen service Product
OBJEcnVE for technology sustainable & instructional development
transfer income for mission
organization
Create investment Diversify economic Capitalize
opportunity base investment
opportunity
Bolster tax base Create good will
between
institution &
community
Complement
existing
programs
Utilize vacant
facilities

nership to divide the real estate operation and the business development mission between
themselves. Because partnership objectives are so diverse, i.e they depend on the mix of
organizational participants, they are not represented in Table 1. The last type of orga-
nizational arrangement is just starting to appear as a distinct group. A few large com-
panies have developed corporate incubators. Generally in these facilities economic value
is added back to the company in the form of new products and services. Because so few
of these exist, and they are more experimental than the other types, they are not repre-
sented in Table 1.

THE FACILITY LIFE CYCLE

A second concept that provides clarity to the operations of business incubators is
drawn from economics-life cycles. Allen's (1988) examination of non-profit incuba-
tors showed that the first stage, start-up, is characterized by gaining control over the
building or site, obtaining construction or renovation financing, and attracting initial
tenants. Because of incubator cash flow requirements, initial tenants are likely to be
companies that can pay rent rather than high-potential start-ups. This stage ends around
the time the facility reaches its financial break-even point.
During the second stage, business development, attention turns from real estate

Winter, 1990 65
concerns to nurturing new businesses. At this time substantial linkages are developed
with local professional business consultants who provide a full range of management
advisory services to tenants. Another characteristic of the business development stage is
the synergy that builds from mutually advantageous trade relations among tenants. When
demand for space is appreciably greater than the space available for tenants, and so-
phisticated, responsive business advisory arrangements are properly functioning, the
incubator moves into the third phase, maturity.
Maturity is a time when the incubator spreads its development activities throughout
the community, becoming a focal point for entrepreneurial assistance in the area. To deal
with space demand pressures, the facility tightens entrance policies, accelerates gradu-
ation, or expands operations into new quarters. The entire life-cycle process has been
seen to take five to six years, but as new incubators are able to learn from past industry
experiences, new knowledge and information-sharing compresses the learning curve.
The life-cycle process suggests one central tension within incubators-balancing
between the enterprise-development mission and the necessity to run the incubator as a
self-sustaining real estate operation. The success of the enterprise-development mission
depends on a sustainable real estate operation, but not necessarily vice-versa. One can
expect that as the facility moves through life-cycle stages, the quality and quantity of
development outcomes produced are higher.

METHODOLOGY AND SAMPLE

During February and March 1987, questionnaires were mailed to business incubators
as identified in a roster compiled by the National Business Incubation Association. Three
weeks passed from the first mailing before a second wave of surveys was sent out.
Telephone follow-up calls were made to second-wave nonrespondents. Only facilities
that were located at a specific site, offered shared office services, and provided tenant
access to business-development advice were considered incubators and included in this
study. Other enterprise assistance programs that do not provide both affordable space
and management services, such as Small Business Development Centers and Enterprise
Zones, were not included in this study. By April, 1987, 127 usable surveys were
returned-a response rate of 70.5% of the facilities that met the inclusion criteria.
Nonrespondents appear to be randomly distributed across the population of facilities.
Questionnaires were completed by the person responsible for day-to-day incubator
operations, usually the incubator manager. Three-quarters of the managers (76.6%) were
male, one-quarter were femaie. Managers' ages ranged from 78 to 26 years, with an
average of 44.9 years. Respondents had managed their facility for a mean of 24.9
months, although 90% had managed their facility for 18 months or less.

FINDINGS
Organizational Structure
The incubator continuum conceptual framework presents ideal organizational types.
In reality, incubators exist in some multidimensional space defined by a myraid of
variables such as organizational type, stage in the life cycle, community entrepreneurial
environment, and a changing tenant mix, to name a few. To better obtain a picture of the
organizational type dimension, information was collected on the building ownership and
facility management. A cross tabulation of those variables is shown in Table 2. Seed
capital and corporate incubators were deleted from this and subsequent analyses because
together they represent less than five percent of the incubators in the known population.

66 ENTREPRENEURSHIP THEORY and PRACTICE

Table 2

Building Ownership by Type of Management

Management
Count (cell % )
Ownership Public Private Academic Partnership Total

Public 53 1 1 2 57
(41.7) (0.8) (0.8) (1.6) 44.9%
Private 2 32 0 0 34
(1.6) (25.2) (0.0) (0.0) 26.8%
Academic 3 0 19 0 22
(2.4) (0.0) (15.0) (0.0) 17.3%
Partnership 2 2 2 8 14
(1.6) (1.6) (1.6) (6.3) 11.0%
Total 60 35 22 10 N = 127
47.2% 27.6% 17.6% 7.9% 100.0%

Non-profit business-development organizations own and operate about 42% of all

incubators (Table 2). Private corporations operating in a for-profit mode constitute
one-quarter of the facilities, while academic organizations constitute 15% of the facil-
ities. Two variations of public-private partnership facilities constitute the remaining
19%. First, incubators explicitly identified as owned and operated as a public-private
facility constitute 8 of the 127 facilities. Implicit public-private partnership facilities are
those that do not share ownership and management within the same sector (i.e., those off
the diagonal in Table 2). These implicit partnership facilities constitute 12% of all
facilities.
The organizational configuration suggests that incubators fit well with different
kinds of institutional enterprise-development objectives. An explanation of why the
incubator concept easily transfers from one sector to another involves compatible hier-
archical objectives and what the facilities do to assist tenants. The combination of
objectives for a particular incubator can be thought of as how the organization addresses
its particular tenant niche. Whether different organizations want to commercialize uni-
versity research, create jobs for minorities, or help promising local inventors is not the
main issue. From an incubator-industry perspective, the important issue is that different
kinds of stakeholders' objectives can be served or achieved by different kinds of incu-
bators.

Incubator Policy Variations

The incubator continuum conceptual framework suggests a relationship between
organizational type and incubator policy and operations. For example, the studies that
examine incubator public policy (Plosila & Allen, 1986; Allen & Weinberg, 1988) state
that public interests for efficiency and equity translate into both termination of tenant
subsidies after a certain point, and admission of particular kinds of tenant (e.g. start-up,
manufacturing, etc.). Organizations' stakeholders and their interests are posited in Table
3. This enumeration of admission criteria was derived from translating primary and
secondary objectives into admission and exit criteria based on the authors' understanding
and relevant literature.

Winter, 1990
67
Table 3

The Business Incubator Continuum: Stakeholder Policy

For-Profit Non-Profit For-Profit

Property Development Seed
Development Corporation Academic Capital
Incubators Incubators Incubators Incubators

STAKEHOLDERS Property developers Economic University officials Managing partners

development
officials
Economic Politicians Economic Limited partners
development development
officials officials
Politicians Economic
development
officials

ADMISSIONS Ability to pay rent Net jobs potential University High growth
POLICY affiliation potential
New firms Technology Harvest potential
intensive
Local ownership Net jobs potential
No retail, Complements
wholesale or university
personal services programs

EXIT POLICY ---------------------- Grow Too Large For Facility ----------------------

-------------------------- Lease Non-Compliance --------------------------
No explicit exit Graduated rent Graduated rent Growth
criteria performance
Time limit for Time limit for
tenancy tenancy
No explicit exit No explicit exit
criteria criteria

Exit (also called graduation) policies, in particular, present a logical hierarchy of real
estate criteria and business-development criteria. For example, private property devel-
opment incubators seldom have tenant graduation criteria. When firms become too large
for the facility or they violate the lease agreement, they relocate. Conversely, non-profit
development corporation and academic incubators logically have the previous two real
estate exit criteria, but beyond that also often have other tenant graduation criteria. Some
incubators in these two arenas incrementally increase rent to cushion tenants' early-stage
cash flow, then raise rental rates above market value to induce relocation during later
years. The philosophy is that the real estate subsidy should be short lived and used when
it is most needed, and if the firm is not meeting the stakeholders' growth objectives,
space should be made available for more promising tenants. Facilities may also place an
explicit, though flexible, time limit (usually three years) on tenancy. Some, similar to
real estate-oriented facilities, have no exit criteria; tenants move out of their own accord
and rents do not rise above market rate.
The data collected provide a quantitative examination of the admission and exit
policies discussed above. Most incubators are willing to accept high value-added firms
as tenants (Table 4). Consequently, high-tech product and R&D firms are accepted by
nearly all incubators. When lower value-added businesses are considered, a marked

68 ENTREPRENEURSHIP THEORY and PRACTICE

Table 4

Acceptable Business Types for Tenancy

% Admitting Particular Type

Public Private Academic Partnership PR>F

High Tech 86.8 96.9 89.5 100.0

R&D 84.9 96.9 95.7 89.5
Light Mfg. 88.7 81.2 73.7 82.6
Mail Order 35.6 71.9 31.6 47.8 .005
Wholesale 35.6 75.0 15.8 43.8 .0001
Non-profit Org. 35.9 71.9 31.6 43.8 .003
Retail 9.4 31.3 10.6 8.7 .04
(N) (53) (32) (19) (23)

Note: PR > F denotes statistical significance below .05.

distinction occurs for private facilities compared to the other three types. Except for
retail firms, over 70% of the private incubators will accept low value-added tenants such
as mail-order, wholesale, and non-profit organizations. Less than 36% of the public and
academic facilities will admit these three types of firms, with a slightly higher accep-
tance percentage for partnerships. Concerning admission criteria, the data show that
public, academic, and partnership incubators are not significantly different, while pri-
vate incubators are generally more willing to accept any tenant that pays rent. Lower
value-added firms are primarily of interest to those facilities that seek to fill space. These
data explain Lumpkin and Ireland's (1988, p. 77) "puzzling" finding that 50% of the
private incubators do not screen tenants for admission. Little or no screening is necessary
when management is primarily interested in collecting rent.
Another aspect of translating incubator objectives into operational policy is exit or
graduation criteria. One misconception about incubators is that at some point a tenant

Table 5

Exit Policy Criteria

% Using Particular Criteria

Criteria Public Private Academic Partnership PR>F

No. of Employees 9.6 3.1 10.5 4.5

Length of Stay 63.5 18.7 63.2 54.5 .000
Business Profits 19.2 6.2 26.3 22.7
(N) (52) (32) (19) (22)

Note: PR > F denotes statistical significance below .05.

Winter, 1990
69
firm will have to leave the facility. As can be seen in Table 5, this is often not the case.
Where graduation is desired, the number of employees and tenants' profitability are not
frequently used exit criteria. A more likely criterion is the length of stay. For all but
private facilities, over half to two-thirds limit the length of tenancy. The most common
approach is to gradually increase the rental rate. For example, the rate could be 20%
below market for the first year, 10% below for the second, market for the third, and 10%
above for the fourth. This implicit graduation policy approach may also be executed with
a short-term lease which after three years is changed to a year-to-year or month-to-month
renewal. If at some time into the renewal period incubator management wants to move
the tenant out, management may decide not to renew the lease. This scenario is less
likely to occur in private facilities; less than 20% use length of stay as an exit criterion.

Incubator Services
Once a firm has become a tenant, company management can avail itself of the
services the incubator provides. One kind of assistance, shared services, is of little
interest for this analysis. Providing shared services such as a photocopier, conference
room, receptionist, personal computer, and so forth, may help reduce tenants' operating
expenses but does little to address management problems that plague new firms (Reyn-
olds, 1987; Kazanjian, 1988).
Across incubators, management problems of tenants are addressed in different ways
with varying results (Fry, 1987). All incubators in this sample provided at least four of
twelve different management services included in the questionnaire. In over 90% of the
facilities accounting, marketing, and business planning assistance is available. In over
75% of the facilities computer training, legal service, government procurement, and
government grant and loan assistance is available. No statistically significant differences
exist between the type of facility and management services provided (Table 6a).
One important difference is seen in the way business assistance is provided by the
local enterprise-support network. Other research (Allen & Hendrickson-Smith, 1986)
has shown that local business-assistance providers work with incubators to gain access
to new clients and that incubator tenants serviced by professional advisors are charged
lower rates than other non-tenant firms served by the same consultants. Managers were
asked how charges were handled for services offered through the external network. Five
different arrangements were given and respondents indicated all those that applied to
their incubator (Table 6b). The two most common arrangements, provider reduces rates
and provider donates services," are also the most advantageous for tenants from a cash
perspective. Donated services occur least frequently in for-profit incubators compared to
the other three types. Tenants pay full rate for services in about 40-55% of the facilities,
depending on the organizational type. The other two arrangements occur infrequently.
This finding suggests that managers in for-profit incubators are less likely to be part of
sophisticated assistance networks where providers make significant contributions in
order to create a broad risk-sharing environment.
Th~ results of this analysis on tenant admission and graduation policy show that
there are some important distinctions between the types of incubator facilities. Greatest
similarity is seen among public, academic, and partnership incubators. For-profit incu-
bators are the most dissimilar. Very few distinctions between kinds of services and how
they are provided are evident for organizational type.

Assessing Incubator Performance

The issue of what characteristics influence incubator ''success'' is an important one.

70 ENTREPRENEURSHIP THEORY and PRACTICE

Table 6

Tenant Access to Business Assistance

A. Services Provided by External Network

% Mentioning Service

Public Private Academic Partnership PR>F

Business planning 41.5 25.0 26.3 30.4

Accounting 77.5 46.4 63.2 73.9
Marketing 64.2 53.I 52.6 52.2
Computer training 47.2 56.3 31.6 52.2
Legal service 75.5 75.0 63.2 73.9
Govt. procurement 41.5 65.6 47.4 52.2
Govt. grants 41.5 68.8 47.4 47.8
(N) (53) (28) (19) (23)

B. Arrangements for Externally Provided Services

% Mentioning Arrangement

Provider reduces
rate 59.6 43.8 57.9 52.2
Provider donates
service 63.5 25.0 63.2 65.2 .003
Tenant pays in full 38.5 56.3 42.1 39.l
Incubator & tenant
divide 15.4 9.4 5.3 13.0
Incubator pays
provider in full 9.6 9.4 5.3 17.4
(N) (53) (28) (19) (23)

Note: PR > F denotes statistical significance below .05.

Incubators are a growth industry. By late 1985, 12 states had programs directed to
supporting incubator facilities, and by 1989 more than 25 states were operating or
beginning incubator programs (Weinberg & Allen, 1989). Sponsors, such as states,
frequently create guidelines to assure incubators' administrative effectiveness (i.e., the
fit with larger enterprise-development objectives), and economic efficiency (i.e., max-
imum development benefit at least cost). These guidelines are based more on accepted
wisdom than empirical evidence.
Data collected from this study can be used to assess performance relative to the ends
of the continuum--real estate and business development. Incubators are speculative real
estate deals. All four types of incubators experienced less than 30% occupancy on
opening day (Table 7). Academic incubators, the smallest in usable square feet, have the
greatest initial leasing compared to the other three. By the first six months the four types
reach virtual equality in occupancy rates. By one year all four types reach the half-full
(or half-empty) occupancy point. Lease rates continue to rise but tend to cap off at the
80-90% occupancy level. Given the turnover of tenants resulting from discontinuance
and graduation, 90% occupancy is about the realistic maximum. Although there are no
data on financial break-even or occupancy comparisons with conventional multi-tenant

Winter, 1990 71
Table 7

Incubator Size and Occupancy Percentage

Public Private Academic P!!!tnership

A. Size XmdN XmdN XmdN XmdN PR>F

79,000 34,000 41 127,750 77,500 12 21,000 13,000 15 130,250 33,000 18 .02

B. Occupancy x N x N x N x N

opening day 13.3 52 11.9 31 27.9 18 12.8 19 .05

6 months 34.7 51 32.9 31 39.1 16 31.5 17
I year 52.5 42 50.2 29 53.6 12 54.0 15
18 months 62.8 36 63.1 26 55.2 12 63.7 12
2 years 70.2 25 79.0 15 60.0 8 71.0 9
3 years 73.9 11 83.5 12 55.0 6 81.3 4
4 years 86.7 6 83.1 10 75.0 5 52.5 2

facilities in the same market area, these figures suggest that business incubators are not
strong real estate ventures.
From the business development side of the continuum, two dimensions, jobs created
and firms graduated, are used, but are by no means a comprehensive or refined set of
evaluative variables. From a state policy perspective, job creation is the primary eval-
uative criterion (Weinberg & Allen, 1990), though Campbell (1987) has cautioned that
measuring jobs created by tenants misses important long-term benefits of incubators.
Because one of the global purposes of an incubator is to nurture fledgling firms into
viable business entities most evaluations examine tenant graduation rates. Graduation
rates alone, however, do not tell the entire story because the measure does not account
for what happens to those firms after they leave the incubator. Acknowledging these
shortcomings does not refute the argument that both variables tell part of the story of
incubators. From a developmental perspective, if long-term benefits are to accrue,
short-term outputs such as these must transpire.
A multiple regression model was devised to assess determinants of incubator per-
formance. The parameters of the model are nonlinear, but through logarithmic trans-
formation the model becomes intrinsically linear (Neter, Wasserman, & Kutner, 1983,
p. 467). The two models employ identical samples, except that in the graduates model,
incubators that have not had any graduates were not included. Two control variables, age
of facility and number of tenants, form the base model for the regressions. The zero-
order correlation between the two is .42. Essentially, these two variables in the equation
remove the effects of older facilities having more time to produce results and larger
facilities having more tenants that can produce jobs or graduates. The age variable can
be seen as a surrogate for the life-cycle process. Together the age and facility size
variables explain 59% of the variance in jobs created by tenant firms (Table 8a) and 52%
of the variance in number of firms graduated (Table 8b).
Various independent variables were tested to determine the effect on incubator
outputs. Dummy variables for organizational type showed no statistically significant
effects. Similarly, dummy variables for business assistance services (both types of
services and service arrangements were tested) showed no statistically significant ef-

72 ENTREPRENEURSHIP THEORY and PRACTICE

Table 8

Determinants of Incubator Performance

A. Results of Regression Analysis for Characteristics of Business Incubators and Job

Generation (logged)
Standardized Regression Coefficients

Dependent Variables Base Model

Age of facility (log) .44** .41**

Number of tenants (log) .47** .47**
Light manufacturing admitted .12*
(dummy variable)
Coefficient of detennination (R-squared) .59 .61
n =115

B. Results of Regression Analysis for Characteristics of Business Incubators and Firms

Graduated (logged)
Standardized Regression Coefficients

Independent Variables Base Model

Age of facility (log) .51**

Number of tenants (log) .35**

Coefficient of detennination (R-squared) .52

n = 110

NOTE: * Significant at .05 level.

** Significant at .01 level.

fects. Policy variables concerning admission and exit criteria were also tested. One
dummy variable, whether or not light manufacturing companies are admitted, was
shown to have a weak (statistically significant at the 0.05 level) effect on jobs created,
but not on firms graduated. This finding suggests that incubators that admit light man-
ufacturing companies, compared to those that do not, have tenants that create more jobs.
Most likely the high labor content of most light manufacturing firms, compared to
non-manufacturing firms, accounts for this finding.

CONCLUSION

Summary of Findings
A clearer picture on the relationships between business incubator structure, services,
and policy to performance has emerged from this research. A framework was devised to
conceptually distinguish between different types of business incubators. A value-added
continuum is ancho_red at either end by o!1e. of t~o major incubator purposes: property
development or busmess development. W1thm this contmuum fall four relatively distinct

Winter, 1990
73
types of incubators. The position of the incubator in the facility life cycle is also
suggested as influencing development outcomes.
Data collected from 127 business incubator facilities operating during the spring of
1987 were examined from the perspective of a four-part typology. Statistically signifi-
cant differences were seen concerning the type of tenants admitted. Privately owned and
operated facilities were seen as much less selective than the three other types, and private
facilities are more likely to take lower value-added business. Tenants in private incu-
bators are less likely to be asked by facility management to leave the incubator. Little
difference by facility type is seen in the kinds of business assistance services provided
to tenants. No attempt was made to assess the quality of business assistance. Tenants in
private incubators, however, are much less likely to receive free services from local
professional providers. The speculative nature of incubators was seen in the relatively
slow lease-up rates for all four types of facilities. Generally speaking, across the indus-
try, incubators are not thought of as good real estate deals.
Through the use of multiple regression techniques relationships between structural,
services, and policy variables and job creation and tenant graduation were examined.
Although two variables, age of facility and number of tenants, explained over half the
variation in jobs created and tenants graduated, virtually none of the structure, service
or policy variables had a statistically significant explanatory effect. One independent
variable was statistically significant in the jobs created model, but the effect was not
particularly strong.

Implications for Incubator Stakeholders

These findings are interesting because they reinforce what some have claimed about
the industry, and call into question some implicit assumptions about industry perfor-
mance. The business incubator industry is similar to the venture capital industry in that
both take a patient perspective on investment. As incubators mature through the life-
cycle process, organizational learning occurs and positive outcomes increase. Public
officials who are looking for a "quick fix" would find incubators inappropriate devel-
opment tools.
The findings call into question the common practice of stakeholders specifying
arrangements, services, or policies to which incubators must conform. Except for one
policy-relevant variable, variables that are indicators of policy prescriptions were not
related to performance outcomes. This general finding reinforces the common conten-
tion that business incubators are idiosyncratic reflections of local conditions. Public
policy prescriptions do, however, have a place in incubator development. They set
parameters that justify the expenditure of public monies on facilities. At the least, they
have not been shown to detract from performance. By creating the bounds for permis-
sible activities, stakeholder policies have an instrumental effect. Other explanations of
incubator performance (i.e., the residuals in the regression models) are likely specific to
a community or facility.
For mcubator managers the findings do not provide a clear sense of optimum policies
or procedures. It may be contended that determining the optimum is presently an unat-
tainable expectation. The industry is still young and about 90% of the managers had been
at the facility only 18 months or less. It may also be contended that determining the
optimum is unrealistic. What works in one incubator may not work for another even if
it shares many of the same enterprise development purposes. The uniqueness of each
incubator seems largely due to the local entrepreneurial environment and the needs and
resources of the stakeholders. Following this argument it becomes necessary for man-

74 ENTREPRENEURSHIP THEORY and PRACTICE

agers to know their entrepreneurial environments as much as possible and carve out a
niche in that environment. Managers must patiently wait out their strategies, and if their
activities seem to be producing results, a larger facility will produce even greater results.
This is the logic of the findings as interpreted by the life-cycle concept.

Future Research
In light of the apparent growth in the industry, much more needs to be known about
incubators as an enterprise-development technique. Until recently the industry has not
been large enough to provide the number of cases necessary for in-depth empirical work.
Now that many incubators have passed into advanced stages of business development
and maturity it is becoming possible to examine sophisticated performance measures.
Little research on incubators has used an approach other than cross-sectional surveys. A
preliminary study comparing incubator tenants with similar firms not located in incu-
bators has been completed (Allen & Bazan, 1990), but more work is necessary. If
incubators are to make a difference they must add value and create outcomes above and
beyond what occurs in the unsheltered local business environment. Important for as-
sessing change and the validity of the life-cycle model are longitudinal data. To date
only a series of unrelated one-time studies have been conducted. The NBIA annual
surveys will provide data for researchers who desire to conduct longitudinal studies.
Unresolved issues of industry performance fall into four domains. First, the area of
how tenant companies benefit by being located in an incubator is largely unexplored. In
particular, the strongly contended synergistic effect (Smilor & Gill, 1986) of tenants has
not been examined. Second, although much is known about which services are available
in incubators, little is known about service quality and how quality services can best be
delivered. Third, the area of financial performance of facilities is unexplored. In par-
ticular, if an incubator is to assume other than a real estate or seed capital role, can it
exist without public subsidy, and if not, how long should the subsidy run? Fourth, what
are the effects of incubators on their local economies? Do incubators play a role in
changing latent entrepreneurs' perceptions, attitudes, and behavior about enterprise cre-
ation? As the incubator industry matures so will research on incubators. The full story
on industry enterprise-creation performance is yet to be told.

REFERENCES
Allen, D. N. (1985). Small business incubators and enterprise development. Report to the U.S. Department
of Commerce, Economic Development Administration, Washington, DC.

Allen, D. N. (1988). Business incubator life cycles. Economic Development Quarterly, 2(1), 19-29.

Allen, D. N., & Bazan, E. J. (1989). Evaluation of Title IX SSED program inducements and enhancements
on multi-tenant buildings as a component in local economic recovery. Report to the U.S. Department of
Commerce, Economic Development Administration, Washington, DC.

Allen, D. N., & Bazan, E. J. (1990). Value added contributions of Pennsylvania's business incubators to
tenant firms and local economies. Report to the Appalachian Regional Commission and the Pennsylvania
Department of Commerce, Washington, DC.

Allen, D. N., & Hendrickson-Smith, J. (1986). Planning and implementing small business incubators and

Winter, 1990 75
enterprise support networks. Report to the U.S. Department of Commerce, Economic Development Ad-
ministration, Washington, DC.

Allen, D. N., & Weinberg, M. (1988). State investment in business incubators. Public Administration
Quarterly 12(2), 196-215.

Bates, T., & Nucci, A. (1989). An analysis of small business size and rate discontinuance. Journal of Small
Business Management, 27(4), 1-7.

Birch, D. L. (1987). Job creation in America: How our smallest companies put most people to work. New
York: The Free Press.

Birley, S. (1985). The role of networks in the entrepreneurial process. Journal of Business Venturing, 1(1),
107-118.

Campbell, C. (1987). Change agents in the new economy: Business incubators and economic development.
Minneapolis, MN: University of Minnesota Press.

Council of State Community Affairs Agencies. (1988). Incubator training manual. Washington, DC.

Fry, F. L. (1987). The role of business incubators in small business planning. American Journal of Small
Business, 12(1), 51-62.

Gumpert, D. E., & Boyd, D. P. (1984). The loneliness of the small business owner. Harvard Business
Review, 62(6), 33-38.

Institute for Local Government and Rural Development. (1989). Assessment of the Edison incubator
program. Report to the Ohio Department of Development, Athens, OH.

Kazanjian, R. B. (1988). Relation of dominant problems to stages of growth in technology-based new

ventures. Academy of Management Journal, 31(2), 257-279.

Lumpkin, J. R., & Ireland, R. D. (1988). Screening practices of new business incubators: The evaluation
of critical success factors. American Journal of Small Business, 12(4), 59-81.

National Business Incubation Association. (1990). The state of the business incubation industry, 1989.
Athens, OH.

Neter, J., Wasserman, W., & Kutner, M. H. (1983). Applied linear regression models. Homewood, IL:
Richard D. Irwin.

Nyrop, K. (1986). Business incubators as real estate ventures. Urban Land, 45(12), 6-10.

Pacholski, R. L. (1988). Hatching an incubator: Obtaining recognition of section 501(c) (3) status for
incubator organizations. The Tax Magazine, 66(4), 273-283.

Plosila, W. H., & Allen, D. N. (1985). Small business incubators and public policy. Policy Studies
Journal, 13(4), 729-734.

Reynolds, P. D. (1987). New firms: Societal contribution versus potential. Journal of Business Venturing,
2, 231-246.

Schermerhorn, J. R. (1980). lnterfirm cooperation as a resource for small business development. Journal
of Small Business Management, 18(2), 48-54.

Smilor, R. W. (1987). Commercializing technology through new business incubators. Research Manage-
ment, 30(5), 36-41.

76 ENTREPRENEURSHIP THEORY and PRACTICE

Smilor, R. W., & Gill, M. D. (1986). The new business incubator. Lexington, MA: D.C. Heath and Co.

Weinberg, M. L., & Allen, D. N. (1989). The state's role in incubator planning and operation. Report to
the National Business Incubation Association and National Council for Urban Economic Development,
Athens, OH.

David N. Allen is Assistant Professor of Business Administration at Pennsylvania State University.

Richard McCluskey is a graduate assistant in the Department of Geography, Pennsylvania State

University.

This paper was presented at the 1988 National Academy of Management Meeting, Entrepreneurship
Division, Anaheim, CA.

Winter, 1990 77