The World’s Largest Open Access Agricultural & Applied Economics Digital Library

This document is discoverable and free to researchers across the

globe due to the work of AgEcon Search.

Help ensure our sustainability.

Give to AgEcon Search

AgEcon Search
http://ageconsearch.umn.edu
[email protected]

Papers downloaded from AgEcon Search may be used for non-commercial purposes and personal study only.
No other use, including posting to another Internet site, is permitted without permission from the copyright
owner (not AgEcon Search), or as allowed under the provisions of Fair Use, U.S. Copyright Act, Title 17 U.S.C.
 African Journal of Agricultural and Resource Economics Volume 8 Number 1 pages 58 - 68

Is food security enhanced by agricultural technologies in rural Ethiopia?

Degye Goshu*
School of Agricultural Economics and Agribusiness, Haramaya University, Ethiopia. E-mail: [email protected]

Belay Kassa
School of Agricultural Economics and Agribusiness, Haramaya University, Ethiopia. E mail: [email protected]

Mengistu Ketema
School of Agricultural Economics and Agribusiness, Haramaya University, Ethiopia. Email: [email protected]

*Corresponding author

Abstract

This paper investigates the interdependence of decisions on the adoption of agricultural technology
and the simultaneous interaction between adoption and food security situations of smallholders,
using a sample of 260 households from rural Ethiopia. Three agricultural technologies and two
food security measures were estimated with simulated maximum likelihood (SML) multivariate
probit models to measure the link between the adoption of agricultural technology and food security
indicators and to identify their underlying determinants. The simulation results suggest that
households’ decisions about the adoption of agricultural technology and their food security
situations were strongly and positivity interdependent, with very low likelihood of adoption and
food security. The common underlying factors of technology adoption and food security situations
were also identified. The results generally imply that a concerted effort is required to enhance
household food security through the accelerated introduction and dissemination of appropriate
agricultural technologies in rural Ethiopia.

Keywords: food security; technology adoption; multivariate probit; simulation

1. Introduction

The eradication of poverty in Ethiopia, where smallholder farming is the dominant livelihood
activity and the source of vulnerability to poverty and food insecurity, is an overriding objective of
the incumbent government (FDRE 1996, 2004, 2012; MoFED 2006; Brown & Teshome 2007).
Achieving agricultural growth and development and thereby improving rural household welfare
requires increased efforts to provide yield-enhancing resources. Agricultural technology can
contribute to increased food production (food availability) and increased agricultural and rural
incomes (better access to food), and entails positive spillovers to other sectors and contributes to
economy-wide growth. Agricultural productivity growth is also vital for stimulating growth in other
sectors of the economy (Moreno & Sunding 2003; Kidane et al. 2006).

Studies on the adoption of technology date back to the exploration of the economics of
technological change by Griliches (1957), and the formal adoption and diffusion models applied by
Mansfield (1963), Feder et al. (1985) and then by Green and Ng’ong’ola (1993). Since then,
adoption and diffusion have been conceived as the processes governing the utilisation of
innovations, and studies of adoption behaviour emphasise factors that affect the adoption of
agricultural technologies. In developing countries like Ethiopia, agriculture is a strong option for
spurring growth, overcoming poverty, and enhancing food security, and this has necessitated the
need to increase agricultural productivity through the introduction and use of improved agricultural
AfJARE Vol 8 No 1 Goshu et al.

technologies (Moreno & Sunding 2003; World Bank 2008). A number of area- and commodity-
specific studies of technology adoption conducted in Ethiopia have focussed minimally on the link
between the adopted technologies and food security, their interactions, and the effect of the former
on the latter. Some of the latest adoption studies in Ethiopia include the evidence provided by
Chilot et al. (1996), Asfaw et al. (1997), Negatu and Parikh (1999), Mulugeta et al. (2001), Nega
and Sanders (2006), Aklilu and Graaff (2007), Hassen et al. (2011) and Workneh and Puskur
(2011).

Food security is assumed to exist “when all people, at all times, have physical and economic access
to sufficient, safe and nutritious food to meet their dietary needs and food preferences for an active
and healthy life” (FAO 1996). The four dimensions of food security are food availability, stability
of supply, accessibility of food, and quality and safety of food. In the literature there are three
categories of indicators of food security, each with limited capacity to capture the extent of food
security and hunger: process, outcome, and trend indicators (Hoddinott 1999; Bickel et al. 2000;
Swindale & Ohri-Vachaspati 2005; Smith & Subandoro 2007). The two basic outcome measures of
diet quantity available to a household are daily food energy consumption per capita or per adult
equivalent, and percentage of households or people that are food energy-deficient
(FAO/WHO/UNU, 1985). The second diet quantity indicator is the percentage of households in a
population group that do not consume sufficient dietary energy. If the estimated total energy in the
food that the household acquires daily is lower than the sum of its members’ daily requirements, the
household is classified as food energy-deficient or food insecure.

On the other hand, the three basic indicators of diet quality of a household are diet diversity,
percentage of food energy from staples, and quantities of foods consumed daily per adult equivalent
(Hoddinott 1999; Smith & Subandoro 2007). Improved diet quality is associated with improved
birth weight and child nutritional status, and reduced mortality (Ruel 2002, 2003). It is also
recognised that inadequate diet quality rather than insufficient energy consumption is becoming the
main dietary constraint facing poor populations.

Empirical evidence of food security in Ethiopia indicates the prevalence of a high level of food
insecurity, with significant idiosyncratic and spatial characteristics. The specific food security
studies by Samuel (2004), Berhanu (2004), Freihiwot (2007), Bogale and Shimelis (2009), Zegeye
and Hussien (2011), Hadleya et al. (2011), Abebaw et al. (2011) and Hailu (2012) generally suggest
that the depth and intensity of food insecurity are high, influenced by poor functioning of marketing
systems and other household and socioeconomic factors. However, all the studies have focussed
little on the role and measurement of adoption of agricultural technologies, and their
interdependence with the food security situation of households. To account for these shortcomings,
the objective of this study was to measure the multivariate interdependence of households’ adoption
decisions relating to agricultural technologies and their food security situation, and to identify their
underlying determinants. To this end, univariate and multivariate probit estimation and simulation
techniques were employed to verify the consistency and interaction of all outcome variables.

2. Research Methodology

2.1 Dataset and variables

Agricultural systems in Ethiopia are classified into four types, namely the highland mixed farming
system, low plateau and valley mixed agriculture, pastoral livestock production of the arid and
semi-arid zones, and commercial agriculture (Ayele 1980). This study was conducted in four
districts, selected from two major sedentary sub-farming systems (Central and Eastern highlands)
covering about 40% of the total sedentary farming system in Ethiopia. To account for the expected

58
AfJARE Vol 8 No 1 Goshu et al.

heterogeneity in the samples operating in different farming systems, a stratified two-stage random
sampling procedure was employed and a total of 260 households were randomly and
proportionately sampled.

The major endogamous variables considered in the analysis include the household’s food security
status, determined from daily calorie availability per adult equivalent compared to the
predetermined daily minimum calorie requirement for Ethiopia (2 200 kilocalorie), dietary diversity
status, determined from the number of food groups consumed out of seven food groups, and
adoption status of three agricultural technologies (chemical fertiliser, high-yielding crop varieties,
and improved livestock breeds).

The food security and adoption status of agricultural technologies in Ethiopia was generally
hypothesised to be determined by family size (head count), gender of the household head (binary),
literacy status (binary), farming experience (years), total cultivated land and its allocation to
production of staples and cash crops (hectares), irrigation water use (binary), quantity of fertiliser
(quintals) used for crop production, livestock holding in tropical livestock units (TLU), gross
income earned, access to credit (binary), participation in off-farm activities (binary), distance
(kilometres) to major town, nearest road and development station as a proxy for market information
on food and agricultural technologies, transaction cost, and access to government extension services
respectively, and a dummy variable for the two farming systems as a proxy to capture omitted
location-specific characteristics (Hoddinott 1999; Bouis & Hunt 1999; Bickel et al. 2000; Ruel
2002, 2003; Moreno & Sunding 2003; Aklilu & Graaff 2007; Smith & Subandoro 2007; Kennedy
et al. 2011; Hassen et al. 2011; Workneh & Puskur 2011).

2.2. Estimation and simulation techniques

Two food security measures and the adoption of three agricultural technologies were analysed.
Chemical fertiliser, high-yielding crop varieties and improved livestock breeds were adopted by
43%, 29% and 32% of the sample households respectively, suggesting that the majority of the
households were non-adopters. The interest in this regard was to measure the likelihood of
households to adopt an agricultural technology, not their intensity of adoption. Accordingly, food
security status and the adoption of each type of agricultural technology were estimated by a
univariate probit model (Maddala 1983; Long 1997; Long & Freese 2005; Cameron & Trivedi
2009; Greene 2012):
y i*  xβ  u i (1)
1 if y i*  xβ  u i  0
yi   (2)
0 if y i*  0.
where yi* is the binary latent variable for food security status, dietary diversity status, fertiliser use,
adoption of high-yielding crop varieties or improved livestock breeds (observed if yi*  0 , 0
otherwise); and x is a vector of household-specific and other socioeconomic factors determining
food security status and the adoption of an agricultural technology.

However, the above univariate probit estimation of food security measures and adoption status of
each agricultural technology would be misleading for the expected problem of simultaneity. The
adoption of one type of agricultural technology would be dependent on the adoption of the other,
since household adoption decisions are interdependent, suggesting the need to estimate them
simultaneously. To account for this problem, a seemingly unrelated multivariate probit simulation
model was employed (Long 1997; Chib & Greenberg 1998; Cappellari & Jenkins 2003):

59
AfJARE Vol 8 No 1 Goshu et al.

fert i*  x'1 β 1   1i ,
hyvci*  x'2 β 2   2i (3)
hyvl  x'3 β 3   3i ,
*
i
where
  1i   0 1 12 13 
    
  2i  ~ N  0  12 1  23  , (4)
   0  13  23 1 
 3i   
E  x   0
Var  x   1 (5)
Cov  x    .

where fert i , hyvci , and hyvyi are the households’ adoption status of chemical fertiliser, high-
yielding crop varieties and improved livestock breeds respectively; x1 to x3 are vectors of
independent variables determining the respective adoption variables; β' s are vectors of simulated
maximum likelihood (SML) parameters to be estimated;  1i to  3i are correlated disturbances in a
seemingly unrelated multivariate probit model; and  ’s are tetrachoric correlations between
endogenous variables.

In the trivariate case there are eight joint probabilities corresponding to the eight possible
combinations of successes (a value of 1) and failures (a value of 0). If we focus on the probability
that every outcome is a success, for instance, the probabilities that enter the likelihood function of
the technology adoption simulation are explained as

Pr  fert i  1, hyvci  1, hyvli  1

  2 (β 1' x 1 , β '2 x 2 , β' 3 x 3 ,  ) (6)
 Pr  1i  β'1 x1 ,  2i  β' 2 x 2 ,  3  β'3 x 3 .

where  2 is the bivariate standard normal distribution.

To estimate the interdependence of household decisions to adopt agricultural technology and the
household’s food security objectives, the above trivariate probit simulation model was extended to a
multivariate probit simulation of five endogenous variables, including both groups of indicators for
adoption and the food security situation:
fert i*  x'1 β 1  v1i ,
hyvci*  x'2 β 2  v 2i
hyvli*  x'3 β 3  v3i (7)
sec uri*  x'4 β 4  v 4i
hddsi*  x'5 β 5  v5i ,

60
AfJARE Vol 8 No 1 Goshu et al.

 v1i   0 1 12 13 14 15 

    
 v 2i   0   21 1  23  23  25 
 v  ~ N  0    32 1  34  35  (8)
 3i   
31

 v 4i  0 
  41   42  43 1  45 
   0   
 v 5i    51  52  53  54 1 

where sec uri and hddsi are the households’ food security and dietary diversity status respectively;
x1 to x 5 are vectors of independent variables determining the respective latent variables; v1i to v5i
are correlated disturbances in a seemingly unrelated multivariate probit model, and the other
notations and assumptions are as explained earlier.

3. Results and Discussion

3.1 Univariate estimation results

Two important dimensions of the food security situation were measured and estimated in this study.
Household dietary quantity was measured by daily calorie availability per adult equivalent
(kilocalorie, kcal) and percentage of households that are food energy-deficient. Accordingly,
households’ food security status was measured by comparing the level of the daily calorie
availability per adult equivalent with the predetermined caloric requirement per adult equivalent for
Ethiopia, which is 2 200 kcal per day. The distribution of the estimated daily calorie availability per
adult equivalent was a bit right-skewed, suggesting that the higher number of food-insecure
households fell below the mean value. Counts of food groups and food items consumed by a
household were assumed to capture the level of diet quality and diversity of the household. Seven
major food groups and the total counts of food items consumed by a household were analysed to
measure the level of diet diversity as an indicator of diet quality. The frequency distribution of
counts of food items suggested that most households consumed five kinds of food items grouped in
three food categories.

The determination of the food security status of households based on their daily calorie availability
showed that 42.7% of the sample households were food secure, while the majority were food
insecure or calorie-deficient. The number of food items consumed in each food group was counted
and a total of 26 food items were identified. The analogous binomial classification of households by
their level of dietary diversity into two, as ‘medium diversity’ and ‘low diversity’ status, showed
that only 40% of the households consumed more than three food groups.

Before the estimation of their interdependence, food security indicators and the adoption of the
three agricultural technologies were estimated by univariate probit models (Table 1). Food security
status was significantly and positively determined by female heads, cultivated land, quantity of
chemical fertiliser used, annual gross income, and access to credit, but negatively influenced by
family size, land allocated to staples, and irrigation water use. The results indicated that the
likelihood of households to be food secure was about 42.3%. The factors determining the
households’ diet diversity were literacy status, livestock holding, annual income, farming system
and other exogenous shocks. The probability of households having a semi-diversified diet was only
37.2%.

The adoption of chemical fertiliser was enhanced and determined by irrigation water use,
agricultural income, distance to research institution and the farming system. Adoption of high-
yielding crop variety, on the other hand, was improved by land allocated to cash crops and the
61
AfJARE Vol 8 No 1 Goshu et al.

farming system. Factors determining the adoption of improved livestock breeds were farming
experience, land cultivated, irrigation water use, distance to nearest road, and the farming system.
The magnitude of marginal effects of farming systems (48% to 54%) was very pronounced in
enhancing the adoption of all three agricultural technologies. The second largest marginal effect
was that of agricultural income (29%) on adoption of chemical fertiliser, followed by land allocated
to cash crops (23%) on adoption of high-yielding crop variety. The predicted probabilities
suggested that the likelihood of households to be food secure, to have a semi-diversified diet, and to
adopt chemical fertiliser, high-yielding crop varieties and improved livestock breeds, were 42%,
37%, 34%, 20% and 21% respectively.

Table 1: Marginal effects of univariate probit estimates of adoption and food security status
Determinants Food Dietary Fertiliser Crop Livestock
security diversity variety breeds
Female heads 0.18* 0.05 -0.07 -0.05 -0.08
Family size -0.06*** 0.01 -0.001 0.02 0.01
Literacy status - 0.18*** - - -
Farming experience - - -0.002 -0.003 -0.01***
Land cultivated 0.28*** 0.03 - - 0.15***
Land allocated to staples -0.26** - - - -
Land allocated to cash crops - - 0.11 0.23*** -
Livestock holding (TLU) 0.01 0.04*** - - -
Irrigation water use -0.21*** -0.08 0.19** 0.08 0.14*
Quantity of fertiliser 0.12* -0.04 - - -
Annual gross income (log) 0.09* 0.15*** - - --
Gross agricultural income (log) - - 0.29*** 0.05 0.04
Access to credit 0.16* -0.05 - - -
Off-farm activity -0.10 -0.04 - - -
Distance to nearest road - -0.01 0.01 0.004 -0.04***
Distance to development station - - 0.02 0.001 0.01
Farming system -0.06 0.54*** 0.53*** 0.48*** 0.54***
Log likelihood -145.06 -135.04 -123.15 -109.23 -112.94
Pseudo R2 0.18 0.23 0.29 0.30 0.30
Predicted probability 0.42 0.37 0.34 0.20 0.21
Note: ***, ** and * signify significance levels of 1%, 5% and 10% respectively.

3.2 Interactions of adoption decisions

The expected multivariate interdependence of adoption of chemical fertiliser, high-yielding crop

varieties and improved livestock breeds was accounted for by employing the multivariate probit
simulation of the adoption of the three agricultural technologies (Table 2). The null that the
tetrachoric correlations are jointly zero and the three adoption decisions are independent was
rejected at the 1% level. The SML estimation results suggested that there was positive and
significant interdependence between household decisions to adopt chemical fertiliser and high-
yielding crop varieties, and high-yielding crop varieties and improved livestock breeds, but not
between adoption of chemical fertiliser and improved livestock breeds.

The adoption of chemical fertiliser was enhancing the adoption of high-yielding crop varieties,
since the households’ decision to adopt one type of technology reinforced adoption of the other,
specifically when they were adopted in the same enterprise. The adoption of chemical fertiliser was
improved by irrigation water use, agricultural income and the farming system, but adversely
affected by other exogenous shocks. All the covariates were similar to the results from the
univariate probit estimation results. However, the probability of households adopting fertiliser was
39%, which was significantly higher than the probability predicted from the univariate probit model
(34%).
62
AfJARE Vol 8 No 1 Goshu et al.

Adoption of a high-yielding crop variety was improved by land allocated to cash crops, agricultural
income and the farming system, but significantly and negatively affected by other exogenous
shocks. The predicted probability of adopting of high-yielding crop variety was 29%, nine
percentage points higher than the probability estimated in the univariate estimation (20%). Farming
experience, land cultivated, irrigation water use, distance to nearest road, farming system and other
exogenous shocks were significant determinants of the adoption of improved livestock breeds. The
likelihood of households adopting improved livestock breeds was 31%, which was 10 percentage
points higher than the probability estimated in the univariate analysis (21%).

If households were able to adopt all three agricultural technologies, their joint likelihood of
adopting these technologies would be only 12%. It was unlikely for households to adopt all three
agricultural technologies simultaneously. This was justified either by the fact that simultaneous
adoption of all the technologies was unaffordable for the smallholders, or that all three technologies
were not simultaneously accessible in the study areas. However, their joint probability of not
adopting all the agricultural technologies was 44%, implying that the households were more likely
to fail. This evidence suggests the need to launch a progressively developing package and scheme
of agricultural technology adoption, and points to the importance of mobilising additional resources
to augment households’ efforts at accelerated technology adoption. It was also shown that the
adoption of chemical fertiliser, high-yielding crop varieties and improved livestock breeds was
determined by similar underlying factors of adoption, suggesting the feasibility of launching a
synergetic programme for the adoption and dissemination of agricultural technologies in Ethiopia.

Table 2: Multivariate probit simulation results of agricultural technology adoption

Variables Coefficients (adoption equations)
Fertiliser Crop variety Livestock breeds
Female heads -0.19 -0.13 -0.29
Family size -0.02 - 0.02
Farming experience 0.004 -0.01 -0.02***
Land cultivated - - 0.46***
Land allocated to cash crops 0.27 0.88*** -
Irrigation water use 0.49** 0.28 0.45**
Gross agricultural income (log) 0.81*** 0.20* 0.14
Distance to major town - -0.003 -
Distance to nearest road 0.02 - -0.13***
Distance to development station 0.04 - 0.04
Farming system 1.53*** 1.63*** 1.76***
Constant -9.37*** -3.98*** -2.86***
Predicted probability 0.39 0.29 0.31
 0.41***
 21
 0.12
 31
 0.24*
 32
Number of simulations (draws) 100
Log likelihood -339.43
Wald  2 25 175.88

Wald test of rho,  21   31   31 =0, P   2 3 0.01

Joint probability (success) 0.12
Joint probability (failure) 0.44
Note: ***, ** and * signify significance levels of 1%, 5% and 10% respectively.

63
AfJARE Vol 8 No 1 Goshu et al.

3.3 Interdependence of adoption and food security

The adoption status of the three agricultural technologies was simulated with two food security
indicators to identify the presence of nonlinear interdependence among them. The SML probit
estimates suggest the presence of some interdependence among agricultural technology adoption
decisions and the food security conditions of the households (Table 3). It was also indicated that the
factors determining each equation were consistent with the results of the previous simulation of
agricultural technology adoption.

Table 3: Multivariate probit simulation results of technology adoption and food security
Variables Coefficients (equations)
Food Dietary Fertilizer Crop Livestock
security diversity variety variety
Female heads 0.52** 0.12 -0.17 -0.06 -0.26
Family size -0.15*** 0.02 -0.02 - 0.01
Literacy status 0.30 0.53*** - - -
Farming experience - - -.003 -0.01 -0.02***
Land cultivated 0.75*** 0.13 - - 0.47***
Land allocated to staples -0.65** - - - -
Land allocated to cash crops - - 0.31 0.84*** -
Irrigation water use -0.65*** -0.22 0.51*** 0.30 0.49**
Quantity of fertiliser 0.23 -0.18 - - -
Livestock holding (TLU) 0.02 0.10*** - - -
Gross annual income (log) - 0.45*** - - -
Gross agricultural income (log) 0.25 - 0.80*** 0.20* .1426348
Access to credit 0.41* -0.13 - - -
Off-farm activity -0.27 -0.12 - - -
Distance to major town - - - -0.003 -
Distance to nearest road - - 0.02 - -0.13***
Distance to development stationa - - 0.04 - 0.04**
Farming system -0.14 1.66*** 1.57*** 1.59*** 1.79***
Constant -2.25 -6.02*** -9.30*** -3.94 -2.87***
Predicted probability 0.43 0.40 0.39 0.29 0.31
Joint probability (success) 0.04
Joint probability (failure) 0.17
Number of simulations (draws) 5
Log pseudo likelihood -615.52
Wald  2 48 376.28

Likelihood ratio test of rhob = 0, Pr   10  0.03

2

Notes: a: Distance to development station was not a good proxy for access to government extension services.
b: For lack of space, tetrachoric correlations are not reported.
***, ** and * signify significance levels of 1%, 5% and 10% respectively.

The simulation results indicate that the probability that households are food secure and have a semi-
diversified diet was 43% and 40% respectively. These probabilities were a bit higher than the
probability estimated from the univariate probit models of food security status (42%) and dietary
diversity status (37%). However, it was verified that households’ food security conditions were not
largely impacted on by the adoption of agricultural technologies, since the households’ likelihood to
adopt these technologies was very much limited. To enhance food security and dietary diversity
status at the household level it would be necessary to facilitate the adoption of agricultural
technologies. The joint probabilities of success and failure of the five variables also suggest that it
would be unlikely for households to adopt all the three technologies and achieve their food security
objectives simultaneously, for their likelihood to do so was only 4%.

64
AfJARE Vol 8 No 1 Goshu et al.

4. Conclusion and Recommendations

The empirical evidence in this paper suggests that the food security situation of rural households in
Ethiopia is very poor and that a great majority of people suffer from deficiencies in their daily
calorie intake and from problems relating to dietary diversity. The adoption of agricultural
technologies was very low (not more than 39%) and the majority of households were non-adopters
of these simple agricultural technologies. From the likelihood of households to adopt the
technologies and the unsatisfactory effects of these technologies on the food security situation, it is
evident that the introduction and dissemination of yield-enhancing agricultural technologies is very
limited and that the food security situation is bad. Even if the adoption of agricultural technologies
was positively and significantly associated with households’ food security situation, the limited
success in the introduction and dissemination of the technologies at household level dilutes the
opportunity to enhance food security through the use of agricultural technologies.

It is evident that the effects of government extension services and marketing infrastructural facilities
are generally insignificant and sometimes irrelevant, suggesting the need to improve access to
packages of agricultural technologies. In order to boost agricultural production and productivity, a
concerted effort is required to generate, introduce, integrate and disseminate appropriate agricultural
technology packages, which, in turn, will improve food security. Moreover, to accelerate food
production and enhance the food security conditions of rural households, the functioning of input
and output agricultural markets, including technological inputs and their expected net returns, need
to be improved.

References

Abebaw S, Janekarnkij P & Wangwacharakul V, 2011. Dimensions of food insecurity and adoption
of soil conservation technology in rural areas of Gursum District, Eastern Ethiopia. Kasetsart 32:
308–18.
Aklilu A & Graaff de J, 2007. Determinants of adoption and continued use of stone terraces for soil
and water conservation in an Ethiopian highland watershed. Ecological Economics 61(2-3): 294–
302.
Asfaw N, Gungal K, Mwangi W & Beyene S, 1997. Factors affecting adoption of maize production
technologies in Ethiopia. Ethiopian Journal of Agricultural Economics 2: 52–69.
Ayele A, 1980. Agro-climates and agricultural systems in Ethiopia. Agricultural Systems 5(1): 39–
50.
Berhanu A, 2004. The food security role of agriculture in Ethiopia. eJADE 1(1): 138–53.
Bickel G, Nord M, Price C, Hamilton W & Cook J, 2000. Guide to measuring household food
security. United States Department of Agriculture (USDA), Alexandria VA, USA.
Bogale A & Shimelis A, 2009. Household level determinants of food insecurity in rural areas of
Dire Dawa, Eastern Ethiopia. African Journal of Food, Agriculture, Nutrition and Development
9(9): 1914–26.
Bouis H & Hunt J, 1999. Linking food and nutrition security: Past lessons and future opportunities.
Asian Development Review 17(1, 2): 168–213.
Brown T & Teshome A, 2007. Implementing policies for chronic poverty in Ethiopia. Chronic
Poverty Research Centre, Addis Ababa, Ethiopia.
Cameron C & Trivedi TK, 2009. Microeconometrics using Stata. StataCorp Ltd, Texas, USA.
Cappellari L & Jenkins SP, 2003. Multivariate probit regression using simulated maximum
likelihood. The Stata Journal 3(3): 278–94.
Chib S & Greenberg E, 1998. Analysis of multivariate probit models. Biometrika 85(2): 347–61.

65
AfJARE Vol 8 No 1 Goshu et al.

Chilot Y, Shapiro BI & Mulat D, 1996. Factors influencing the adoption of new wheat technologies
in Wolmera and Addis Alem Areas of Ethiopia. Ethiopian Journal of Agricultural Economics 1(1):
63–84.
FAO (Food and Agriculture Organization), 1996. World Food Summit Plan of Action. Accessible at
http://www.fao.org (Accessed 20 May 2012).
FAO/WHO/UNU, 1985. Energy and protein requirements. Report of a Joint FAO/WHO/UNU
Expert Consultation. Geneva, World Health Organization Technical Report Series 724
FDRE, 1996. Food Security Strategy. Addis Ababa, Ethiopia.
FDRE, 2004. Food Security Program 2004-2009. Addis Ababa, Ethiopia.
FDRE, 2012. Ethiopia’s progress towards eradicating poverty: An interim report on poverty
analysis study (2010/11). Addis Ababa, Ethiopia.
Feder G, Just RE & Zilberman D, 1985. Adoption of agricultural innovations in developing
countries: A survey. Economic Development and Cultural Change 33(2): 255–98.
Freihiwot F, 2007. Food security and its determinants in rural households in Amhara Region. Addis
Ababa, Ethiopia: Ethiopian Development Research Institute (EDRI).
Green DA & Ng’ong’ola DH, 1993. Factors affecting fertilizer adoption in less developed
countries: An application of multivariate logistic analysis in Malawi. Journal of Agricultural
Economics 44(1): 99–109.
Greene WH, 2012. Econometric analysis. Seventh edition. New Jersey: Pearson Hall.
Griliches Z, 1957. Hybrid corn: An exploration in the economics of technological change.
Econometrica 25(4): 501–22.
Hadleya C, Linzerb DA, Tefera B, Abebe G, Fasil T & Lindstrome D, 2011. Household capacities,
vulnerabilities and food insecurity: Shifts in food insecurity in urban and rural Ethiopia during the
2008 food crisis. Social Science and Medicine 73: 1534–42.
Hailu M, 2012. Causes of household food insecurity in rural Boset Woreda: Causes, extent and
coping mechanisms to food insecurity. Germany: Lap Lambert Academic Publishing.
Hassen B, Bezabih E, Belay K & Jema H, 2011. Factors affecting adoption and intensity of use of
improved wheat varieties in North Eastern Ethiopia: A double hurdle approach. Ethiopian Journal
of Agricultural Economics 8(1): 133–58.
Hoddinott J, 1999. Choosing outcome indicators of household food security. Washington, DC:
International Food Policy Research Institute (IFPRI).
Kennedy G, Ballard T & Dop M, 2011. Guidelines for measuring household and individual dietary
diversity. Rome: FAO.
Kidane W, Maetz M & Dardel P, 2006. Food security and agricultural development in Sub-Saharan
Africa (SSA): Building a case for more public support. Rome: FAO.
Long JS & Freese J, 2005. Regression models for categorical dependent variables using Stata.
Second edition. USA: Stata Press, Texas.
Long JS, 1997. Regression models for categorical and limited dependent variables (RMCLDV).
Thousand Oaks CA: Sage Press.
Maddala GS, 1983. Limited-dependent and qualitative variables in econometrics. Cambridge:
Cambridge University Press.
Mansfield E, 1963. The speed of response of firms to new techniques. Quarterly Journal of
Economics 77: 290–311.
MoFED (Ministry of Finance and Economic Development), 2006. Ethiopia: Building on progress:
A Plan for Accelerated and Sustained Development to End Poverty (PASDEP). Addis Ababa,
Ethiopia.
Moreno G & Sunding DL, 2003. Simultaneous estimation of technology adoption and land
allocation. Paper read at the American Agricultural Economics Association Annual Meeting, 27–30
July, Montreal, Canada.
Mulugeta E, Belay K & Legesse D, 2001. Determinants of adoption of soil conservation measures
in Central Highlands of Ethiopia. The case of three districts of North Shoa. Agrekon 40(3): 313–35.

66
AfJARE Vol 8 No 1 Goshu et al.

Nega GW & Sanders JH, 2006. Farm-level adoption of sorghum technologies in Tigray, Ethiopia.
Agricultural Systems 91: 122–34.
Negatu W & Parikh A, 1999. The impact of perception and other factors on the adoption of
agricultural technology in the Moret and Jiru Woreda districts of Ethiopia. Agricultural Economics
21: 205–16.
Ruel MT, 2002. Is dietary diversity an indicator of food security or dietary quality? A review of
measurement issues and research needs. Food Consumption and Nutrition Division Discussion
Paper 140, IFPRI, Washington, D.C.
Ruel MT, 2003. Operationalizing diet diversity: A review of measurement issues and research
priorities. Journal of Nutrition 133(11): 3911S–26S.
Samuel G, 2004. Food insecurity and poverty in Ethiopia: Evidence and lessons from Wollo.
Working Paper No. 3/2004, Ethiopian Economics Association (EEA)/Ethiopian Economic Policy
Research Institute (EEPRI), Addis Ababa, Ethiopia.
Smith LC & Subandoro A, 2007. Measuring food security in practice. Washington, DC: IFPRI.
Swindale A & Ohri-Vachaspati P, 2005. Measuring household food consumption: A technical
guide. Washington, DC: Food and Nutrition Technical Assistance (FANTA) Project, Academy for
Educational Development (AED).
WHO (World Health Organization), 1985. Energy and protein requirements. Technical Report
Series 724, WHO, Geneva.
Workneh A & Puskur R, 2011. Financial benefits of box hive and the determinants of its adoption
in Atsbi Wembera district of Tigray, Ethiopia. Ethiopian Journal of Agricultural Economics 8(1):
159–84.
World Bank, 2008. Agriculture for development. World Development Report 2008, World Bank:
Washington, DC.
Zegeye T & Hussien H, 2011. Farm households’ food insecurity, determinants and coping
strategies: The case of Fadis district, Eastern Oromia, Ethiopia. Ethiopian Journal of Agricultural
Economics 8(1): 1–35.

67