Date of Award

Spring 1-1-2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Economics

First Advisor

Peters, Michael

Abstract

Economic development encompasses more than just growth of average productivity and income per capita. Growth cannot occur without a broader process of structural transformation that reallocates economic activity across production units, sectors, and locations. This reallocation depends on workers’ choices of where and how to supply their labor and consumers’ choices of what to consume. The three chapters of my dissertation examine different forces that influence worker and consumer choices – such as labor market frictions, market access costs, uninsurable risk, and technological spillovers – and these choices’ implications for the aggregate path of the economy. Services are the largest employment sector in the world, and recent studies show that it is a significant source of productivity growth in developing countries. But little is understood about how the service sector can generate this growth, especially in low-income settings, where it primarily consists of basic consumer services like retail and food service. The first chapter of my dissertation argues that the key driver of service-led growth in developing countries is the service transformation: the ongoing process of modern service firms like supermarkets and restaurants replacing traditional micro-entrepreneurs like street vendors and food hawkers. I use micro-data to show that this process reflects two underlying forces: workers moving out of subsistence employment in traditional services to more productive modern jobs, and consumers increasing their relative demand for modern services as they get richer. I build these two forces into a general equilibrium model of the service sector. The model micro-founds income effects on consumer demand with fixed shopping costs to access modern services. It also features search frictions for wage work in the labor market, which micro-found the marginal productivity gap between modern and traditional services. Together, these features imply that small improvements in the technology used by modern service firms set off a cycle of amplified productivity growth. As workers transition to the modern sector, they increase aggregate income and redirect demand to modern services, which pulls in more workers and generates further growth. A significant portion of this growth comes from reallocation of workers to more productive jobs, rather than mechanical growth from making workers more productive in the jobs they already hold. I estimate the model with Brazilian micro-data on consumer service expenditures, worker earning and employment dynamics, and the local allocation of service employment and income. I validate the estimated model by testing it against a number of untargeted moments, including a natural experiment from Brazil’s trade liberalization. Counterfactual simulations of the estimated model show that the main catalyst of Brazilian service transformation from 2000 to 2010 was improving technology in the modern service sector. Demand effects amplified this transformation by a factor of 1.22. The resulting transformation generated service-led productivity growth. Over half of this growth was reallocation gains from shifting service workers into the more productive modern sector. The second chapter of my dissertation assesses environmental arguments against growth in light of ongoing structural transformation, recognizing that growth is not simply a scale-up of production, but a change in what is produced. With coauthors, I build a model with non-homothetic preferences and innovation that can be directed towards reducing the cost of material production or improving the quality of final goods. As real income grows, consumers demand higher-quality goods, and innovation shifts towards increasing the quality rather than the quantity of consumer goods. Higher quality is associated with a higher service input, so that over time the economy becomes more reliant on service labor and uses environmentally harmful material inputs less intensively. If quality improvements are incompletely measured, the shift to quality-led growth will appear in data as a growth slowdown. The chapter uses aggregate and micro data to discipline the model and finds that it fits cross-sectional and time-series patterns of the US economy. The calibrated model shows that innovation subsidies can accelerate the transition to quality-led growth and prevent excessive environmental damages. In the third chapter of my dissertation, I study whether the need to manage uninsured risk influences the destination choice of migrants in low-income settings. I develop a parsimonious model of migration as a risk management strategy. Risk-averse households have an incentive to migrate between origin-destination pairs whose economic shocks have low covariance, because these location pairs are better hedges for one another. Under common preference specifications, the model generates a risk-augmented gravity equation that I estimate using migration data from the Philippines. Empirical results confirm that migration is substantially higher between origin-destination pairs with low covariance of shocks.

Download

Share

COinS