Kenyan small farm household consumption and production behavior: a policy simulation model

Abstract

This thesis develops an empirical model which can be used for the prediction and evaluation of the effects of price and nonprice food production policy incentives in Kenya's smallholder subsector. The small farm households consumption and production behavior are jointly modeled. An attempt is also made to model both the continuous and discrete households choice behavior. The econometric specifications of the farm households continuous and discrete choice behavior are based on both theoretical and practical considerations. The empirical continuous and discrete choice models are estimated using full maximum likelihood(FIML) estimation technique. The FIML estimation is based on the 1977 long rains cross-section data from

the Monitoring and Evaluation Unit in Kenya's Ministry of Agriculture. Most of the FIML parameter estimates have the appropriate economic signs and are statistically significant. The computed discrete choice v probabilities are meaningful. Using a joint test on the production parameter estimates, I cannot reject the hypothesis of constant returns to scale in the food and nonfood crops production technologies. The parameter estimates are used further for partial ex post policy simulations.. The relevant policy variables are the wage rate and the quantity of the land input. Three main policy conclusions arise from the analysis of the results of the policy simulation exercises. A 25 per cent wage reduction affects both the annual cropping pattern discrete choice probabilities and the magnitudes of the continuous choice variables, namely the food and nonfood crops output, and the labour input. Second, households labour absorption is found to be highly wage rate elastic. generally, Finally, the cropped acreage elasticities are found to be, less than unity implying that a percentage change in cropped acreage will lead to a less than proportionate change in both other inputs and output.