Long-Run Enterprise Responses to Redistribution: Experimental Evidence from Kenya

Cash transfer programs continue to be implemented and expanded by governments and non-government organizations in many low-income countries as a tool for poverty alleviation. When implemented at scale, such programs may have important short- and long-run implications for firms. However, there are relatively few opportunities to study how an economy responds to an exogenous shock of such a magnitude, and the response of the private sector is critical in determining how these types of shocks will propagate through the economy. In particular, there is limited long-run evidence on how enterprises grow after a large shock, and when and how rural village economies reach an equilibrium after a large shock.

This project builds on an existing randomized controlled trial, in collaboration with the NGO GiveDirectly, in which unconditional, unanticipated one-time cash transfers of USD 1,000 each were made to households in 653 villages in Siaya County, in rural Western Kenya. The project was conducted through two-level randomization, in order to experimentally measure spillover effects. This led to significant and economically meaningful results, which persisted two years after treatment. Now, the researchers aim to look at the long-term (i.e. 4-5 years) effects of this large, experimental demand shock. More specifically, the following questions are answered:

  • Are the effects of cash transfers on enterprise formation persistent?
  • Are there long-term effects on enterprise growth?
  • Are effects on enterprise formation and growth heterogeneous based on entrepreneur characteristics, such as personality traits?
  • Are there long-run general equilibrium effects of cash transfers?

To do so, four data collection activities are conducted: firstly, a census of enterprises in the villages involved in the original project; secondly, enterprise surveys; thirdly, a household survey module on enterprise activities; and lastly, quarterly market price surveys. These activities complement an already-planned long-term household follow-up survey.

This project has the chance to influence policy at a variety of levels: firstly, both the local Siaya County government and the national one are interested in the link between entrepreneurship and development in rural areas, as well as in the long-term effectiveness of cash transfers. Outside of governments, NGOs and organizations like GiveDirectly or the Busara Center for Behavioral Economics have already shown a particular interest in this initiative. From a broader perspective, as cash-transfers remain one of the most discussed and widely implemented poverty alleviation tools, any research that furthers our knowledge on this matter is of particular policy relevance.


Johannes Haushofer

Princeton University

Edward Miguel

University of California, Berkeley

Paul Niehaus

University of California, San Diego

Michael Walker

University of California, Berkeley