STATISTICAL ANALYSIS OF DETERMINANTS OF CROP PRODUCTIVITY AND HOUSEHOLD POVERTY: THE CASE OF MENZ KEYA GEBRIEL DISTRICT, ETHIOPIA

Abstract

In Ethiopia, agricultural productivity in general and crop productivity in particular is low and has been a critical problem. The study was intended to assess determinants of crop productivity, TE and household poverty in Menz Keya Gebriel district of North Shewa Zone, Ethiopia. The primary data for this study were collected from 207 randomly selected sample households from the district using stratified random sampling based on agro ecology. The analysis basically employed both descriptive and econometric methods. Stochastic frontier and Logistic regression model were used for productivity and poverty analysis, respectively. Cobb- Douglas stochastic frontier model was used to estimate the elasticities of production frontier and inefficiency effect simultaneously by using MLE method for five crops separately. The results of productivity analysis shown that seed, fertilizer, oxen-pair days and labor force affect teff and wheat productivity positively. The inputs (seed, oxen-pair days, labor force) affect bean productivity positively. Fertilizer, oxen-pair days and labor force affect barley productivity positively. Oxen-pair days and labor force has positive effect on sorgum productivity. In most cases, fertilizer, oxen- pair days and labor force affect crop productivity positively. Moreover, the results of efficiency analysis shows that the mean level of technical efficiency for teff, wheat, bean, barley and sorgum crop productivity was 91.11, 88.8, 83.4, 85.6, and 88.2 percent, respectively. The results of the inefficiency analysis shows that sex, education level and age of household head, family size, health status, land quality, number of extension contact, credit, agrochemical and improved seed were the determinants of TE of crop productivity. The study also used logistic regression model for the analysis of determinants of household poverty. The determinants of household poverty were sex of the household head, crop productivity as market value (birr/ha), place of residence, dependency ratio, livestock holding in TLU, non or off-farm activity, non-labor income, and irrigation. Among these significant variables, except place of residence and dependency ratio, all play significance role in poverty reduction. It is suggested that further study conducted on the dynamics of total factor productivity and multi-dimensional poverty of the household in the area.