UC Berkeley

This dissertation applies tools developed in labor economics to empirically study questions in labor, development, and urban economics. Each chapter attempts to decompose a problem into competing explanations. The first decomposes racial segregation in US cities. The second decomposes differences in wages between agricultural workers and and non-agricultural workers. And finally, the last decomposes the heterogeneous responses of workers to a new monitoring technology.

In the first chapter, I revisit the question of whether residential segregationin US cities emerged in the mid-twentieth century as a consequence of decentralized location choices in combination with white antipathy toward black residents or whether it reflected institutionalized constraints on the availability of neighborhoods that black families could access. The chapter analyzes rich population data from the 1930 and 1940 censuses to disentangle these channels. I first lay out a simple discrete choice model of residential choices by white and black families that depends on the local price of housing and on the fraction of black residents in each neighborhood. I show how the preferences of both race groups can be identified using information on the impacts of exogenous inflows of white and black residents to different neighborhoods. White and black rural inflows constituted a major source of inmigration to major cities during this time period; I construct a pair of novel instrumental variables for these inflows by connecting the distributions of white and black surnames in rural areas to earlier migrants living in different census tracts in 1930. The resulting structural estimates confirm that white families had a relatively high willingness to pay to avoid black neighbors, consistent with an important role for preferences in the evolution of neighborhood segregation. Combining white and black preferences, however, I also find strong evidence that black residents faced supply side constraints on their neighborhood choices. I conclude that about one half of the overall degree of neighborhood segregation observed in 1940 was due to the different preferences of white and black families, while a comparable share was due to implicit or explicit constraints on which neighborhoods black families could move into.

While the first chapter interpreted incumbent residents' responses to migrants as reflective of their racial preferences, the second chapter, based on joint work with Marieke Kleemans, Joan Hamory Hicks, and Ted Miguel, directly studies the migrant experience itself. Recent research has pointed to large gaps in labor productivity between the agricultural and non-agricultural sectors in low-income countries, as well as between workers in rural and urban areas. Most estimates are based on national accounts or repeated cross-sections of micro-survey data, and as a result typically struggle to account for individual selection between sectors. We use long-run individual-level panel data from two low-income countries (Indonesia and Kenya). Accounting for individual fixed effects leads to much smaller estimated productivity gains from moving into the non-agricultural sector (or urban areas), reducing estimated gaps by over 80%. Estimated productivity gaps do not emerge up to five years after a move between sectors. We evaluate whether these findings imply a re-assessment of the conventional wisdom regarding sectoral gaps, discuss how to reconcile them with existing cross-sectional estimates, and consider implications for the desirability of sectoral reallocation of labor.

Finally, the third chapter is based on joint work with Ernesto Dal Bó, Frederico Finan, and Laura Schechter and empirically studies models of task assignment within organizations in a developing country context. Standard models of hierarchy assume that agents and middle managers are better informed than principals about how to implement a particular task. We estimate the value of the informational advantage held by supervisors (middle managers) when ministerial leadership (the principal) introduced a new monitoring technology aimed at improving the performance of agricultural extension agents (AEAs) in rural Paraguay. Our approach employs a novel experimental design that, before randomization of treatment, elicited from supervisors which AEAs they believed should be prioritized for treatment. We find that supervisors did have valuable information---they prioritized AEAs who would be more responsive to the monitoring treatment. We develop a model of monitoring under different allocation rules and roll-out scales (i.e., the share of AEAs to receive treatment). We semi-parametrically estimate marginal treatment effects (MTEs) to demonstrate that the value of information and the benefits to decentralizing treatment decisions depend crucially on the sophistication of the principal and on the scale of roll-out.