LAUSR

Abstract Scholarship on African and other developing societies often argues that governments allocate private goods as political patronage, but recent work has called this assumption into question. In this study, I argue that clientelism is a political strategy that emerges only under certain conditions. In a least likely case of land allocation in Kenya, where the literature assumes clientelism to predominate, I ask under what conditions do states distribute resources to fulfil programmatic policy needs versus clientelistic goals? To answer this question, I use a historical process tracing approach that triangulates descriptive data from an original dataset on land allocation programs in Kenya from 1953 until 2017 with archival research and semi-structured interviews. I find that the choice between policy and patronage depends on two domestic conditions: a) the interaction between national political competition and regime type and b) the state’s ability to control resources vis-a-vis local competitors. In an authoritarian regime, competition is conducive to clientelism, but in a democratic regime, competition may well produce programmatic policy, while local actors’ challenges to centralized resource allocation reduce a government’s ability to use clientelism and favor programmatic allocation as ruling elites struggle to control resource allocation for their own political ends. I conclude that the allocation of rights to land in Kenya demonstrates that politics in developing countries occurs in a hybrid neopatrimonial system rather than a purely patronage-based one, in which changes in politics can alter how a government distributes resources. This has ramifications for how scholars and practitioners must rethink questions of property rights, corruption, political linkages and development in many countries where we assume that private goods are allocated on a clientelistic logic.