1. FACTORS AFFECTING IMPROVED WHEAT SEED
TECHNOLOGY ADOPTION: A DOUBLE HURDLE APPROACH
BY
HASSEN BESHIR1, BEZABIH EMANA2, BELAY KASSA3 AND
JEMA HAJI3
1 Department of agricultural Economics, Wollo University, Ethiopia
2 HEDBES Consult, Ethiopia
3 Haramaya University, School of Agricultural Economics and
Agribusiness, Ethiopia
October 9, 2012
2. Outline of the presentation
• Introduction
• Methodology
• Results and discussion
• Conclusion and policy implication
3. 1. INTRODUCTION
• The economic development of Ethiopiahighly
dependent on the performance of its agricultural
sector.
• The average growth rate of agriculture
• 1.68%, Imperial period (1960-1974)
• 3.75% socialist period (1974-1991)
• 5.54% Current period (1991-2011)
4. • The growth rate of agriculture and
GDP low for several decades
mainly due to
severe weather fluctuation,
inappropriate economic policies and
low adoption of improved agricultural
technologies
• Due to this reason Ethiopia received
significant food aid and become
highly dependent on food import
(FAO, 2007).
• The country received 674000 metric
tons of cereal mainly wheat in the
form of food aid in 2006 alone and
more than 1.2 million tone in 2012
5. INTRODUCTION (CONTINUED)
• Farm productivity as indicated by the yield of major crops
and livestock remained low due to mainly to limited
adoption of improved agricultural technologies by
smallholder farmers.
• For example the yield of wheat in south Wollo
• 1.36 ton per hectare in 2008/09
• >2.5 ton per hectare by using improved variety.
• Various empirical studies conducted to identify
determinants of adoption of improved agricultural
technologies whether the farmers adopt or not the
agricultural technologies in Ethiopia, for example (Asfaw et
al., 1997; Tesfaye and Alemu, 2001; Tesfaye et al., 2001;
Mergia, 2002; Kiflu and Berhanu, 2004) but nothing has
been done in the study area.
• Therefore, this study assessed the determinant of the
rate and intensity of improved wheat adoption by
smallholder farmers in the study area in order to fill the
information gap.
6. 2. METHODOLOGY OF THE STUDY
• 2.1. Description of the Study Area
• This study carried out in South
Wollo.
• South Wollo located in the North
East part of Ethiopia.
• South Wollo one of the eleven
administrative zones of the Amhara
Region.
• Among the twenty rural districts,
Dessie Zuria and Kutaber selected
•
7. METHODOLOGY ( CONTINUED)
• 2.2. Sampling Procedure
• Multistage random sampling
• Firstly, Dessie Zuria and Kutaber selected
• At the second stage of sampling a total of 6 FAs
were selected.
• Then 252 sample farmers selected using simple
random sampling technique and distributed
proportionately over the 6 farmers associations.
• 2.3. Data Collection and Sources
• Primary data collected by interviewing the selected
respondents.
• A structured questionnaire designed pre-tested
and refined.
• Data collected from July to November 2009.
8. METHODOLOGY (CONTINUED)
• 3.4. Analytical Models
• 3.4.1. Econometric specification of
agricultural technology adoption model
• The double-hurdle model
• a parametric generalization of the
Tobit model,
• two separate stochastic processes
determine the decision to adopt and the
level of adoption of technology.
9. The double-hurdle model has an adoption (D) decision with
an equation:
if Di*>0 and zero if Di*<0
The level of adoption (Y) decision has an equation:
If Yi*>0 and Di*>0
Otherwise
10. METHODOLOGY (CONTINUED)
• 3.4.2 Definitions of variables for adoption
• 1. The Dependent variables of Probit and
truncated regression models
• The dependent variable of Probit model a
dichotomous value.
• The truncated regression model Dep. Var a
continuous value which should be the intensity,
the use and application of the technology.
• 2. The Independent variables and their
definitions used in double hurdle model
• Adoption literatures provide a long list of
factors that may influence the adoption of
agricultural technologies.
11. TABLE 1 SUMMARY OF DEFINITIONS AND MEASUREMENTS OF PROBIT AND
TRUNCATED MODEL VARIABLES
No definitions and measurements of variables Expected sign
1 Dependent Variable(s)
1.1 Have you used improved wheat? 1. Yes 0. No
How much Kg of improved wheat variety you used?
1.2
2 Independent variables
2.1 Respondent's sex 1. Male 0. female +
2.2 Respondent's age in years -
2.3 Highest Level of schooling of the head in years -
2.4 Number of adult equivalent in the family +
2.5 Number of man equivalent in the family -
2.6 Were you engaged in off-farm activities? 1. Yes 0. No +
2.7 Total cultivated area in hectare +
2.8 Number of plots -
2.9 Total Tropical Livestock Unit +
2.10 Distance from distribution centre for improved wheat variety in minutes -
2.11 Distance from home to nearest market (walking minutes) +
2.12 Distance from home to nearest all weather road (walking minutes) +
2.13 Did extension agent visit you? 1. Yes 0. No +
2.14 Do you participate in credit for production? 1. Yes 0. No +
12. 3. RESULTS AND DISCUSSION
• Improved technologies such as improved seed
and breed, fertilizers and herbicides a
significant role in enabling farmers to increase
the production and hence improve the standard
of living of smallholder farmers.
• The majority of smallholder farmers in Ethiopia
producing both crops and livestock.
• Yield of these activities very low due to low
adoption and application of improved
agricultural technologies
• The study depicted low utilization of improved
wheat which was 3.4% of total cultivated wheat
land in Ethiopia which was 2.6% in south Wollo
(CSA, 2009).
13. TABLE 2 DESCRIPTIVE STATISTICS OF EXPLANATORY VARIABLES AFFECTING
ADOPTION AND INTENSITY OF ADOPTION OF IMPROVED WHEAT VARIETY
(MEANS)
Non- Adopters Total
adopters (57) (252)
Variables
(195)
F Sig.
Distance from home to nearest market 84.12 81.40 83.50 0.117 0.733
Distance from home to nearest all weather road 41.24 14.96 35.30 23.979 0.000
Respondent's age 52.82 54.74 53.25 0.741 0.390
Highest Level of years of schooling of the head 2.09 2.67 2.22 1.311 0.253
Number of man equivalent in the family 3.64 4.68 3.87 14.509 0.000
Number of adult equivalent in the family 4.55 5.66 4.80 13.111 0.000
Total cultivated area in hectare 0.60 0.93 0.68 22.504 0.000
Number of plots 3.58 4.65 3.82 11.361 0.001
Total Tropical Livestock Unit 3.43 4.37 3.64 8.388 0.004
Distance from distribution centre 104.90 38.60 89.90 117.10 0.000
14. TABLE 3 DUMMY VARIABLES CHARACTERISTICS OF FARM HOUSEHOLD
Variable Non-
Characte adopters
r (195) Adopters (57) Total (252) χ2 P-Value
Sex female 47 3 50 9.843 0.002
male 148 54 202
Off-farm no 58 11 69 2.420 0.120
Income yes 137 46 183
DA visit no 105 12 117 19.071 0.000
yes 90 45 135
Credit no 156 48 204 0.507 0.476
Participation yes 39 9 48
15. RESULT AND DISCUSSION
(CONTINUED)
• There are farmers who are adopter and
non-adopter of improved wheat
technologies.
• Adopters are farmers who use improved
wheat seed (for example HAR 1685).
• Non-adopters are farmers who use none of
this technology during the survey year
(2008/2009 production year).
• the likelihood (rate) of adoption of
improved wheat was modest; an average
farmer had 22.62% predicted probability of
adopting the technology
• An average farmer had used improved wheat
seed of 30.77kg with an average cultivated
area of 0.31 hectare for adopters.
16. RESULT AND DISCUSSION (CONTINUED)
• Therefore, the determinant of the rate of adoption
was estimated using Probit model
• whereas the determinant of intensity of use of the
improved wheat was estimated using truncated
regression model.
• Hence double hurdle model was used to estimate
the determinant of the rate and intensity of
adoption of improved wheat variety.
• Accordingly explanatory variables were checked for
problems of multicollinearity, endogeneity and
heteroscedasticity.
• Test statistics of double-hurdle model vs Tobit
model was conducted
• The result suggest the selection of the double
hurdle model
• The result revealed that the calculated
statistical value of likelihood ratio for
improved wheat was 42 which was greater
than the tabulated or critical value of χ2(14) =
32 at 1% level of significance.
17. TABLE 5 IMPROVED WHEAT SEED ADOPTION ECONOMETRIC RESULT
Probit Truncated
Robust Std. Marginal Robust Std. Marginal
Variables Coefficient Err. effect Coefficient Err. effect
Distance to nearest market -0.005** 0.002 -0.0005 -0.71*** 0.181 -0.209
Distance to nearest all weather road -0.020*** 0.005 -0.002 -1.60*** 0.548 -0.471
Sex of the household 1.000*** 0.342 0.067 55.060* 30.817 16.163
Age of the household head 0.005 0.010 0.0005 2.172*** 0.815 0.637
Education of the household head -0.050 0.040 -0.005 2.327 2.020 0.683
House hold size in adult equivalent 0.021 0.175 0.002 -1.391 9.582 -0.408
Labour force in man equivalent -0.056 0.166 -0.006 9.684 9.634 2.843
Farm size 0.290 0.305 0.030 43.96*** 13.274 12.903
Fragmentation 0.025 0.058 0.003 3.653 3.735 1.072
Livestock owned 0.034 0.059 0.004 -6.202 4.419 -1.821
Access to off/non-farm income -0.027 0.316 -0.003 4.142 14.833 1.216
Distance to input supply institution -0.029*** 0.006 -0.003 0.113 0.168 0.033
Access to extension 0.509* 0.300 0.052 62.918** 24.957 18.470
Access to credit 0.104 0.292 0.011 43.26** 21.522 12.699
Constant -0.248 0.919 -247*** 77.631
Wald χ2* (14) = 84 Wald χ2* (14)= 30
Log-L =-291 -
Test statistics No of observation=252 No of observation=57
18. DISCUSSION OF RESULTS
• The gender + and sig. result is consistent
with the finding of Abay and Assefa (2004)
and Teklewold et al. (2006).
• The access to extension + and sig. result is
consistent with the finding of Teklewold et
al. (2006).
• The credit + and sig. result is consistent with
the finding of Abay and Assefa (2004) and
Teklewold et al. (2006).
• The distance of market and road result is
consistent with the finding of Berhanu and
Swinton (2003).
19. 4. SUMMARY AND CONCLUSION
• The objective to identify the major
determinant of the rate and intensity of
adoption of improved wheat in two districts
of south Wollo.
• The study employed cross-section data
• to analyse the effect of farmers
socioeconomic and institutional setting and
physical attributes
• on the determinant of the likelihood and
intensity of improved wheat varieties
adoption.
• Double hurdle model were employed
•
20. SUMMARY AND CONCLUSION (CONT.)
• Selected farmers were interviewed about the
2008/09 cropping production season.
• The study found access and availability of
extension service to be more powerful than other
factors in explaining adoption and intensity of
adoption improved wheat technologies.
• The age of the farmer was significant on intensity
of adoption of improved wheat seed technologies.
• Accumulated knowledge gained through experience
enables older farmers to intensify improved wheat
technologies.
• The resource endowment of the farm household
like farm land size was also significant in affecting
the intensity of improved wheat adoption.
21. SUMMARY AND CONCLUSION (CONT.)
• Physical characteristics like distance from
farmers’ home to markets, roads, and input
supply played a critical role in the adoption of
improved agricultural technologies as
proximity to information, sources of input
supply and credit and markets save time and
reduce transportation costs.
• Given the critical role of proximity to such
centers and better roads for promoting
adoption and productivity gains, the effort of
investment in improved roads infrastructure
should be improved to achieve increased
production.
22. SUMMARY AND CONCLUSION (CONT.)
• Therefore the results of the study
suggest that the adoption of
improved wheat variety should be
increased by raising farm household
asset formation, providing extension
and credit service.
• Such actions may, in turn, reduce
food shortage problem and fasten
economic growth by enhancing
productivity.