BUYING FERTILIZER IN KENYA: WHAT ARE THE KEY DETERMINANTS

1
BUYING FERTILIZER IN KENYA WHAT ARE THE KEY
DETERMINANTS?

• Edward Olale
• PhD Student
• 13 September 2007

2
Acknowledgments

• This is a product of a term paper for Advanced
  Agricultural Marketing Course by Dr. John
  Cranfield
• I acknowledge Dr. John Cranfield, Dr. Alfons
  Weersink, Dr. Oliver Masakure and fellow
  classmates Hina Nazli and Henry Anim-Somuah for
  their constructive comments towards the
  development of this paper

3
Summary

• This study develops an analytical framework that
  simultaneously incorporates income
  diversification, transaction costs and production
  risk in farmer fertilizer market participation
  decisions
• The framework is then tested in analyzing the
  probability and intensity of buying fertilizer in
  Kenya, at farm level

4
Outline

• Introduction
• Theoretical Framework
• Empirical Framework
• Data
• Results
• Model Tests
• Conclusion

5
Introduction

• Improved market access is necessary for poverty
  reduction in Kenya and other developing countries
• More attention has been given to output markets
  than input markets
• Specifically, better access to fertilizer can
• increase soil fertility and productivity
• save labour
• Fertilizer consumption levels have been low in
  Kenya and other SSA countries (9 kg of
  nutrients/ha compared to gt70kg/ha in Latin
  America and Asia)

6
Research Problem

• Joint influence of income diversification,
  transaction costs and production risk on input
  market participation has been ignored by past
  studies
• Research on the influence of these factors on
  fertilizer market participation is scanty
• This study aims at
• developing an analytical framework of farmer
  participation in fertilizer markets that
  incorporates all these factors income
  diversification, transaction costs and production
  risk
• empirically testing their influence on farmer
  participation in fertilizer markets

7
Theoretical Framework Diagram
Utility
Feed-back effect
Farmer Income
Fertilizer market participation
Input prices
Output prices
Transaction costs
Prodn risk
Farm hhs xtics
Income diversification
8
Theoretical Framework Assumptions

• Transaction costs in the fertilizer market
• Farmer is risk neutral
• Possibility of production risk, but no price risk
  
• Output is produced for sale
• Farmer can participate in off-farm employment
• Family labour and hired labour are perfect
  substitutes-one wage rate (w)
• With these assumptions, the objective of the
  farmer is to maximize expected income

9
Theoretical Framework Model

• The farmers objective to maximize expected income
  can be written as
• Where,
• M is expected income
• is expected farm income or profit
• TC is transaction costs in the fertilizer
  market
• NFI is off-farm income
• Transaction costs are either proportional (PTC)
  or fixed (FTC)

10
Theoretical Framework Model

• Expanding the expected income equation yields
• Where,
• is output price
• q is expected output
• is the price of fertilizer
• k is fixed expenditure on other inputs
• is per unit proportional transaction
  cost
• x is fertilizer amount
• is family labour available for work
• is the labour requirement in the farm

11
Theoretical Framework Model

• The production function is specified as
• Where,
• is the probability of no crop
  loss and by extension
• is the probability of crop loss
  
• represents farm household
  characteristics
• The above function is substituted into the
  expected income (M) equation

12
Theoretical Framework Model

• If M with fertilizer purchase gt M without
  fertilizer purchase, then the farmer decides to
  buy fertilizer and maximizes the expected income
  equation to yield
• with income
  diversification
• is labour
  supply off-farm

13
Theoretical Framework Hypotheses

• Comparative statics suggest that fertilizer
  demand is
• positively influenced by output price
• negatively influenced by fertilizer price
• negatively influenced by per-unit proportional
  transaction cost
• positively influenced by income diversification
• Test for the influence of production risk
  (probability of no crop loss) is inconclusive
• The above hypotheses are tested empirically

14
Empirical Framework

• Due to data limitations, not all the derived
  input demand equations can be estimated
• Only the derived fertilizer demand function is
  estimated

15
Empirical Framework

• Prob./intensity of fertilizer purchase
• F(input prices, output prices, farm household
  characteristics, transaction costs, production
  risk, income diversification)-linear
• Specifically, 17 variables included in the model
  were
• Prices fertilizer price perception, value of
  crop products
• Farm household characteristics age, gender,
  family size, crop farm size and access to
  production credit
• Transaction costs education, distance to the
  nearest fertilizer market, agricultural extension
  and agricultural group membership
• Production risk use of drought resistant
  varieties, access to permanent water source and
  agro-climatic zone
• Income diversification predicted wage rate
  off-farm, predicted value of livestock and number
  of crops grown

16
Empirical Framework

• Heckmans two step procedure, is used to obtain
  predicted off-farm wage rate and value of
  livestock- since wage rate and value of livestock
  are endogenously determined
• Age, age squared, education and inverse mills
  ratio are used to obtain predicted off-farm wage
  rate
• The same variables in the wage equation, in
  addition to family size and value of farm
  implements are used to obtain predicted value of
  livestock
• Probit and Tobit models for probability and
  intensity respectively

17
Data

• Generated using a semi-structured questionnaire
  administered to 228 farmers in semi-arid areas of
  Eastern Kenya
• GIS random sampling procedure was used in a
  catchment area covering three districts
• Questions were asked on farm household
  characteristics farm enterprise(s) soil
  fertility management technologies and marketing
  and institutional support

18
Results
Significant factors
19
Model Tests

• Likelihood ratio tests showed that, all the 17
  explanatory variables were jointly able to
  explain both probability and intensity of buying
  fertilizer
• Specifically, inclusion of both transaction costs
  and production risk improved the explanatory
  power of both models
• Inclusion of income diversification only improved
  the explanatory power of the probability model

20
Conclusion

• Income diversification positively influence
  fertilizer market participation this may explain
  why use of labour saving inputs increase with
  more off-farm employment opportunities
• Transaction costs and production risk negatively
  influence fertilizer market participation
• The three factors improve the explanatory power
  of the market participation models except for
  inclusion of income diversification in the
  intensity model
• Inclusion of all the three factors is therefore
  recommended in future agricultural input market
  participation studies

21
THANK YOU
I WELCOME QUESTIONS AND COMMENTS