Ethnic Divisions, Misallocation and the Organization of Production within Firms

• Development Workshop

Date and Location

Monday, October 31, 2011, 4:10 PM - 5:30 PM

Abstract

Ethnic heterogeneity is believed to lower economic growth. In this paper I provide direct microeconometric evidence on the productivity effects of inter-ethnic rivalries, using daily output data for teams of workers at a large flower-packing plant in Kenya. An upstream worker sorts flowers: two downstream workers then assemble the flowers into bunches. Most workers are members of two rival tribes (the Kikuyu and Luo). The plant uses an essentially random rotation system to assign workers to positions. During the first year of my sample period downstream workers were paid piece rates based on own output and the upstream worker based on the total output of the team. I find that upstream workers "pay to discriminate" and thereby lower output: downstream workers who are not of the upstream worker's ethnicity earn as much as 27 percent less than others. The degree of favoritism increased sharply during a period of conflict between the two tribes after Kenya's December 2007 election. In response, the plant began paying the two downstream workers for their combined output. I show that group-based pay led to a modest decrease in the output of homogeneous teams -- as predicted by standard incentive models -- but an increase in mixed teams' and overall output. Workers' response to conflict and contractual incentives is predicted by a model of joint production with pro-ethnic social preferences. My findings suggest that inter-ethnic antagonism leads to misallocation of intermediate goods and thereby lowers output in the private sector, that the economic costs of ethnic diversity vary with the political environment, and that in high-cost environments firms are forced to adopt "second best" policies to limit the distortions caused by ethnic favoritism.