Soil erosion and land degradation have become major environmental concerns and present a formidable threat to food security and sustainability
of agricultural production in Kenya. The biggest challenge currently facing the Kenyan government is how to achieve the triple developmental goals
of food sufficiency, better nutrition and poverty reduction without increasing the land devoted to food crops. This paper responds to a paucity
of empirical information on the impact of land degradation on farm productivity and poverty in Kenya. The paper builds on the few existing
studies in this area and explores the impact of conventional inputs and adoption of soil conservation practices on farm yields per acre controlling
for the effects of institutional factors. Panel data collected from farming households is utilized to achieve the study objectives. The paper tests and
rejects the hypothesis that adoption of soil conservation practices has no effect on farm productivity or on poverty reduction. We test this
hypothesis controlling for impacts of conventional farm inputs and institutional factors notably the property rights regimes. Due to the joint
determination of soil conservation practices and productivity, we use the fixed effects – instrumental variables (FE-IV) method to isolate the
productivity effects of soil conservation. We further simulate the impact of key policy changes on productivity on one hand and on poverty (head
count ratio) on the other. The key finding of the paper (from the FE-IV) is that controlling for unobserved heterogeneity, soil conservation has a
large measurable impact on farm productivity and on poverty reduction in the long-run. We also find that well specified property rights are
associated with higher farm productivity, a debate that is widely contested in the literature. Policy simulations indicate that privatization of
common land and adoption of particular soil conservation practices can play a decisive role in increasing agricultural yields and in reducing poverty
in semi-arid areas.