Work in Progress


Peer Effects in the Workplace: Experimental Evidence from Agricultural Workers in Malawi (with Lasse Brune and Eric Chyn)

We study a set of employees at a large agricultural firm in Southern Malawi to examine the role of peer effects on their output and attendance, and the mechanisms behind those effects. We will work with a sample of several thousand employees who work picking tea leaves, and who are grouped into gangs that rotate through a fixed set of fields. Each worker is assigned to a specific plot on each field, and we randomize those assignments, allowing us to provide the first field-experimental measure of the causal effect of a worker’s peers on his or her performance. Our study also advances the literature on workplace peer effects by collecting detailed data on the relative positions of each worker, allowing us to separate socialization-based mechanisms for peer effects (which are only possible for immediate neighbors due to the distances involved) from information-based mechanisms (which can operate at a greater distance). We will collect panel data on the workers over the course of several months, and re-randomize workers partway through, allowing us to separate between short-run, transient peer effects and longer-run effects that are indicative of more permanent changes in work culture.

Using the same population of workers, we also plan to explore how workers find jobs in the agricultural offseason, and the role their social networks play in that process. This project, in which we will randomly provide information about a limited number of job opportunities to people and track its spread through social networks, will make valuable contributions in several areas. First, it will help us understand how people seek and find work in the agricultural offseason in rural Africa, which is an important component of how many people across Africa smooth their consumption across the year. Second, it will contribute to the literature on job search and social networks in general. Third, it will help us understand how people make choices about rival information, trading off a desire to help their friends against the cost of potentially not getting a job themselves.

The Literacy Laboratory Project (with Rebecca Thornton, Jeff Smith, and Mango Tree Educational Enterprises Uganda)

The Literacy Laboratory Project is a four-year randomized trial of an education intervention in Northern Uganda. We are currently in the second year of the study, and plan to use the experiment in order to test basic theories about how people’s education and labor market choices respond to their perceptions about the returns to education. We are particularly interested in the household dynamics of this process: how does giving one child in a family the prospect of advancing much farther through the education system affect families’ choices about investing in their other children? Are the benefits of the program complementary to other investments in education, or do households substitute toward the child in the program in order to maximize her chances of a good education and a successful career?