Herd growth on shared rangeland: herd management and land-use decisions in Northern Kenya.

This study investigates theoretical and empirical issues characterizing herd growth on a shared rangeland. This analysis explicitly recognizes household and rangeland heterogeneity. Recent developments in the range ecology literature are incorporated into an economic model. Implications of this model are explored using longitudinal data gathered from Gabra pastoral households in Northern Kenya. The study’s findings have important implications for policy measures in pastoral areas and make significant contributions to resource and development economics. Evidence is presented suggesting stochastic herd loss keeps livestock population below ecological carrying capacity in the study area. However, evidence also suggests degradation is occurring in particular rangeland sub-areas. A model is constructed to allow analysis of factors leading to localized degradation. Stochastic herd losses are identified as a factor keeping livestock population below a level impacting rangeland condition in most areas. However, rangeland heterogeneity provides incentives to occupy particular rangeland sub-areas at levels impacting rangeland condition. Analysis of localized degradation requires focus on incentives influencing the spatial distribution of livestock. The analysis utilizes recent developments in simulated maximum likelihood methodology. Empirical analysis identifies factors influencing land use decisions. Food aid availability is found to increase household use of rangeland areas in danger of degradation. Increased household herd size leads to decreased use of these rangeland areas. The latter result suggests the commonly proposed policy measure of decreasing herd sizes to prevent degradation could have the unintended consequence of increasing degradation. Empirical analysis of herd management decisions indicates sales and slaughter decisions are consistent with a model identifying livestock as a form of self-insurance. Herders sell and slaughter more animals as herd size increases. Evidence on livestock transfers indicates transfers are redistributive, but also conditioned on a herder’s transfer record. This suggests repeated interactions support asset redistribution mechanisms. Analysis of herder decision making offers a way to reverse the disappointing record of pastoral development efforts to date. Ecologic and economic success will only be possible if the decision making context facing herders is understood. This study contributes to such understanding