UC Berkeley

This dissertation studies how heterogeneous agents’ exposure to local markets determines welfare and distributional effects of shocks such as policy interventions. In both chapters recent modeling tools in the international trade literature are applied to the study of spatial inequality in cities and its sources in chapter 1 or scaling up agricultural policy interventions in chapter 2. These tools are used to answer research questions in the context of quantitative general equilibrium models that feature rich heterogeneity but remain tractable. Moreover, the models in this dissertation predict clear relationships between the choices of households, firms and aggregate variables that are disciplined with detailed microdata.

The first chapter Two-Sided Sorting and Spatial Inequality in Cities studies a new economic force underlying the spatial sorting of rich and poor households in cities. On the demand side, households with different incomes choose neighborhoods and differ in their expenditures across various local services. On the supply side, service establishments sort into neighborhoods while taking into account proximity to their consumers. This two-sided sorting leads to endogenous differences in the local price index that amplify the concentration of household groups. A recent literature in urban economics has rationalized spatial sorting of households that is left unexplained by local incomes or housing costs by modeling pure amenity spillovers. In this chapter, I quantify the contribution of endogenous price indices to spatial sorting that is usually projected onto such reduced-form spillovers, and study the implications of two-sided sorting for urban policy. To do so, I develop a quantitative equilibrium model of the city that features two-sided sorting and nests many urban models. I estimate the key parameters of the model using detailed microdata for Los Angeles from 1990-2014. I find that spatial variation in local price indices decreases the estimates of reduced-form spillovers by about 30-50 percent. To shed light on the policy implications, I simulate policy counterfactuals, and compare the effects to the existing framework with only reduced-form amenity spillovers. By studying a number of prominent place-based policies in Los Angeles, I find substantially different effects on neighborhood composition and welfare between both models.

The second chapter Scaling Agricultural Policy Interventions: Theory and Evidence from Uganda studies the welfare and distributional effects of scaling agricultural interventions in general equilibrium. Interventions aimed at raising agricultural productivity in developing countries have been a centerpiece in the global fight against poverty. These policies are increasingly informed by evidence from field experiments and natural experiments, with the well-known limitation that findings based on local variation generally do not speak to the general equilibrium (GE) effects if the intervention were to be scaled up to the national level. In this chapter, we develop a new framework to quantify these forces based on a combination of theory and rich but widely available microdata. We build a quantitative GE model of farm production and trade, and propose a new solution method in this environment for studying high-dimensional counterfactuals at the level of individual households in the macroeconomy. We then bring to bear microdata from Uganda to calibrate the model to all households populating the country. We use these building blocks to explore the average and distributional implications of local shocks compared to policies at scale, and quantify the underlying mechanisms.