Effect of push-pull technology adoption and dis-adoption on livelihood outcomes of smallholder maize farmers in Homa Bay County, Kenya

Abstract

Dissemination of push-pull technology (PPT), continued use and expansion of land area under PPT remain critical requirements in addressing the major constraints facing maize production. These include infestation by Striga weed and stem borers, coupled with declining soil fertility conditions which jointly result in low yields and poverty in many developing countries. Despite the extensive literature on PPT adoption, its impacts as well as wider dissemination, little is known about drivers of farmers’ decision to abandon it, or to expand the land area under the technology. Similarly, there is limited empirical evidence that demonstrate the effect of continued PPT adoption on smallholder livelihoods. Therefore, this study determined the rate and determinants of PPT dis-adoption and effect of dissemination pathways on the extent of PPT expansion. This study also evaluated the effect of continued PPT adoption on livelihoods of smallholder maize farmers in Homa Bay County. A multistage sampling procedure was used to select a sample of 240 smallholder maize farmers. Data were gathered through face-to-face interviews using a pretested semi-structured questionnaire. Seemly unrelated bivariate probit model, censored tobit model, and propensity score matching model were used to analyze the three objectives, respectively. Descriptive results indicated that adoption, dis-adoption and expansion rates of PPT were 51%, 39.94% and 48.59%, respectively. Bivariate probit results showed that level of education, greater access to extension services, positive perception of stem borer and Striga weed constraints, and smaller land size positively affected PPT adoption. In addition, male-headed households, high education level, large farm sizes, and a large number of livestock units negatively determined PPT dis-adoption decision. Similarly, greater access to extension service, positive perception of stem borer and Striga weed constraint, availability of napier and desmodium seeds, and longevity of PPT negatively and significantly influenced the PPT dis-adoption decision. Tobit results revealed that male-headed households, being in married households, greater access to extension services, longevity of PPT use and availability of napier or brachiaria seeds significantly affected the extent of PPT expansion with positive coefficients 0.146, 044, 0.156, 0.031 and 0.147, respectively. Similarly, the positive perception of the severity of stem borer, dissemination pathways, smaller land size and distance to the nearest market center significantly influenced the extent of PPT expansion. Interestingly, farmer-to-farmer, field days and farmer teachers were found to be the most important and effective dissemination pathways enhancing the extent of PPT expansion. Further, propensity score matching results revealed that continued PPT adoption had a positive and significant effect on household per capita consumption expenditure (KES 47.81 – 59.02 per day) and household dietary diversity (2.76- 2.87); but it had a negative impact on squared poverty gap (-0.07 to -0. 05). These call for policies that will ensure an integrated input development system which involves collaboration of all stakeholders in ensuring affordability, supply, and accessibility to not only desmodium seed but also other agricultural inputs by all gender. Again, there is a need to incorporate a model farmer as a key pathway in technology dissemination. Also, policies that ensure equitable access to quality education, output and input markets, and efficient and effective extension system should be put in place to ensure continuous and extensive use of PPT among maize farmers.