An Analysis Of The Factors Influencing Participation Of Pastoralists In Commercial Fodder Value Chain For Livelihood Resilience In Isiolo County, Kenya

Abstract

Pastoral communities in the drylands of Kenya face a myriad of challenges including prolonged and recurrent drought coupled with institutional bottlenecks. These factors largely contribute to livestock feed insecurity which impedes livestock production in drylands, thus threatening livelihoods of millions of poor pastoralists. Various interventions have been undertaken by the government and development partners to enhance the coping capacity of pastoralists to the negative effects of climate change. In recent years, the focus of these interventions has been directed to building resilience contrary to disaster risk reduction expenditures such as direct food assistance by various projects. The World Food Programme’s (WFP) project dubbed food for asset creation (FFA) was implemented in thirteen dryland Counties through a collective action approach. The project supports pastoralists in Isiolo County to adopt fodder enterprise in order to address the problem of constant feed shortage. As a result, fodder producer groups and markets have emerged in the County as pastoralists continue to embrace fodder enterprise as a livelihood diversification strategy. However, these developments are happening in the absence of empirical evidence on the socio-economic and institutional factors influencing participation of pastoralists in fodder production and marketing. The present study was conducted in Isiolo County to characterize the commercial fodder value chain, analyze the determinants of individual pastoralist participation in fodder producer groups, and assess the socio-economic and institutional factors influencing pastoralists’ participation in fodder markets. Survey data were collected from 201 randomly selected pastoral households from twenty fodder producer groups. Household interviews, focus group discussions and key informant interviews were used to gather data. Data were analyzed using descriptive statistics, binary Probit model and a Heckman two-step model.