FAS Theses and Dissertations

This dissertation contains three essays in econometrics and international trade. A common theme is the development of econometric tools to support informed policy decisions, with a focus on quantitative trade and spatial models. The first chapter considers sampling uncertainty in quantitative trade and spatial models. Economists use these models to make counterfactual predictions. Because such predictions aim to inform policy decisions, it is important to communicate the uncertainty surrounding them. Three key challenges arise in this setting: the data are dyadic and exhibit complex dependence; the number of interacting units is typically small; and counterfactual predictions depend on the data in two distinct ways—through the estimation of structural parameters and through the description as the status quo. I propose a new Bayesian bootstrap procedure that is tailored to this setting and that addresses all these challenges. The procedure is simple to implement and provides both finite-sample Bayesian and asymptotic frequentist guarantees. I illustrate the practical advantages of this approach by revisiting the applications in Waugh (2010), Caliendo and Parro (2015), and Artuç et al. (2010).

The second chapter considers measurement error in quantitative trade and spatial models. Counterfactuals in these models are functions of the current state of the world and the model parameters. Common practice treats the current state of the world as perfectly observed, but there is good reason to believe that it is measured with error. This chapter provides tools for quantifying uncertainty about counterfactuals when the current state of the world is measured with error. I recommend an empirical Bayes approach to uncertainty quantification, and show that it is both practical and theoretically justified. I apply the proposed method to the settings in Adao et al. (2017) and Allen and Arkolakis (2022) and find non-trivial uncertainty about counterfactuals.

The third chapter, which is co-authored with Isaiah Andrews and Raj Chetty, considers policy choice when both experimental and observational evidence are available. We characterize when and how these sources of evidence should be combined to guide treatment adoption at a new site. We show that the optimal linear predictor for the site-specific treatment effect is a weighted average of the cross-site experimental ATE and the local observational estimate, with weights determined by the covariance matrix of site effects and observational estimands. We provide unbiased estimators for this covariance in settings with both large and small sites, quantify the effect of mismatch between experimental and target sites, and derive easy-to-interpret breakdown points.

This dissertation investigates how households acquire information, form expectations, and make financial decisions in the presence of limited attention and reasoning. The first chapter documents selective inattention to interest rates. A combination of existing and newly-designed surveys shows that households close to durables purchases actively acquire more information about interest rates and have more accurate expectations. These empirical findings motivate a heterogeneous-agent model with endogenous attention to interest rates. Selective inattention shifts the composition of spending responses to interest rate cuts, accelerates the impact of larger cuts, and generates dampened responses to changes in volatility that are closer to empirical evidence. The second chapter introduces a theory of shallow thinking, where agents reason through only a limited number of steps in the causal structure of the economy. New survey data show that households mentally process just 2.6 steps of economic propagation on average. When incorporated into New Keynesian and real business cycle models, shallow thinking accounts for several empirical puzzles, including the asymmetric response of long-term interest rates to different shocks, the predictive power of inflation expectations for bond returns, and persistent output responses to productivity shocks. The third chapter studies how and why households adjust their consumption, saving, and borrowing in response to transitory income shocks. Drawing on new large-scale survey data, the chapter quantifies intertemporal marginal propensities to consume and deleverage and finds significant heterogeneity across households. Standard socioeconomic characteristics explain only a fraction of this variation; psychological factors, past experiences, and expectations improve explanatory power substantially. A machine learning approach identifies four behavioral types, highlighting that similar financial decisions often arise from fundamentally different motivations.

This thesis studies the problem of what keeps regions from growing.

The first two chapters focus on the effects that the geographic mobility of financial capital has on regional growth. That is, they show what happens to the development of regions when financial capital can more easily move from regions where it is abundant to those where it is scarce.

The first chapter studies the geographic integration of American banking markets between the early fifties and early eighties. We show that this financial integration was due to rising nominal rates during the Great Inflation---introducing what we term the ``nominal rate channel'' of financial integration---and to technological improvements in banks' access to national financial markets. Financial integration explains part of the higher growth of the South and West, relative to the average US state, as well as part of the relative decline of the Northern financial centers. This introduces a new framework to jointly study the dynamics of regional growth in an environment where workers and financial capital are both mobile across regions.

The second chapter studies the geographic integration of mortgage markets in the US between 1933 and 1940. This integration was due to government policies that created a national mortgage market, facilitating mortgage capital to move from the financial centers to the rest of the country. Cities that had higher mortgage rates before the policy---and where mortgages became cheaper as a result of financial integration---saw higher growth in rates of homeownership, population, housing construction, and house prices. We also find effects on fertility, as young households witnessed higher birth rates in cities where mortgages became more affordable.

The third chapter concerns why regional transfers to poor regions can fail to generate growth. I offer a theoretical explanation that hinges on local political economy constraints that arise when local governments are in charge of spending these transfers. Local governments' objective to be re-elected can be at odds with maximizing regional growth, transforming a policy aimed at sustaining productivity into one that depresses economic activity. This wedge comes about because local incumbent voters might rationally prefer subsidizing declining incumbent industries instead of attracting new ones, and I find evidence of these political constraints using data from the EU Cohesion policy.

This dissertation studies economic interactions with behavioral agents using both theory and experiments across three chapters. The first two chapters study human misperceptions of GenAI capabilities and their consequences for AI usage. The first introduces the notion of Human Projection: people project human features onto AI when forming beliefs about its performance, which affects their usage decisions. A lab experiment shows that people project human task difficulty when evaluating AI, leading them to overestimate AI performance on human-easy tasks and underestimate it on human-difficult ones. A field experiment shows that among mistakes made by AI, those deemed less reasonable---from a human perspective---induced significantly larger breaches of trust in AI and further reduced user engagement. The second chapter studies the consequences of projection for equilibrium adoption of AI. Projection-prone beliefs can produce an all-or-nothing strategy: users either fully delegate tasks to AI or fully reject it, even when optimal use is task-contingent. Manipulating the framing of AI to remove anthropomorphic cues raises welfare-maximizing adoption, thereby highlighting a potential pitfall of anthropomorphism. The last chapter studies the role of audience effects on displayed moral universalism. When anticipating future interactions with their audience, experimental subjects slant their decisions toward perceived audience preferences. This strategic display is effective in raising audience cooperation and increasing subjects' payoff. Results caution against inferring intrinsic social preferences from public behavior when strategic image concerns are active.

This thesis studies three topics related to household heterogeneity in macroeconomics. In the first chapter, co-authored with Pierfrancesco Mei and Prof. Stefanie Stantcheva, I study how and why households adjust their spending, saving, and borrowing in response to transitory income shocks. I leverage new large-scale survey data to quantitatively assess households’ intertemporal marginal propensities to consume (MPCs) and deleverage (MPDs) (the “how”) and to examine households’ motivations and decision-making processes (the “why”). My findings are as follows. First, I provide evidence that surveys can reliably predict actual economic behavior by comparing responses to hypothetical financial scenarios with observed actions in past studies. Participants’ predicted reactions closely align with real-life behavior. Second, I show that MPCs are higher immediately after an income shock and decline over time, with substantial variation in cumulative MPCs over a one-year horizon. I also show that MPDs play a critical role in household financial adjustments. The heterogeneity in both MPCs and MPDs is not easily explained by socioeconomic or financial characteristics alone, but is better accounted for once psychological factors, past experiences, and expectations are incorporated. Third, using specifically designed survey questions, I document a broad range of motivations behind households’ financial decisions. Based on these motivations, I identify four household types using machine learning: Strongly Constrained, Precautionary, Quasi-Smoothers, and Spenders. Similar financial actions stem from diverse motivations, challenging the predictability of financial behavior based solely on socioeconomic and financial characteristics. Finally, I use these findings to address several puzzles in household finance.

In the second chapter, co-authored with Prof. Tommaso Monacelli, I provide micro-data evidence on heterogeneous perceptions of household income risk. I develop an incomplete information framework with Bayesian learning to provide a micro-foundation for heterogeneity in perceived income risk. I then study the determinants of aggregate fluctuations and the monetary and fiscal policy multipliers in a New Keynesian model with heterogeneous perceptions of income risk. I obtain three main results. First, MPCs are heterogeneous because perceived income risk responds differently to earnings realizations across the income distribution. Second, the response of aggregate output to demand shocks hinges crucially on how perceived income risk co-varies with the cyclicality of individual income across the income distribution. Third, the general equilibrium effects of monetary and fiscal policy can be summarized by a set of observable cross-sectional sufficient statistics. Conditional on the estimated signs of these statistics, heterogeneity in perceived income risk amplifies the response of output to demand shocks.

In the third chapter, I study the optimal level of government debt in a stylized heterogeneous agents open economy model. First, I show that a Ramsey planner faces a key trade-off between providing liquidity to domestic agents and manipulating interest rates to extract monopoly rents from foreign agents holding domestic debt. Second, I derive necessary conditions to characterize the optimal level of debt and interest rates. The planner optimally sets lower interest rates than in the closed-economy first best in order to reduce interest payments to foreign agents. This trade-off intensifies when foreign demand for domestic debt is more inelastic. Finally, I show that the planner can achieve the first-best allocation by issuing two separate debt instruments to domestic and foreign agents.

In this dissertation, I study classic labor economics questions with an eye towards the macroeconomic impact of the answers. In Chapter 1, I study the general equilibrium properties of the childcare industry and whether those properties could motivate designing an economic stimulus that is more effective than traditional stimulus programs. I build a model of an economy with a childcare sector. The model will demonstrate how wages and employment decisions can change in the presence of a childcare sector. Then, in order to calibrate my model, I estimate how much childcare employment changes in response to demand and supply factors. Specifically, when industries where lots of workers use paid childcare expand by 1000 employees, there are on average 18 more childcare workers. Conversely, when industries that compete with childcare for labor expand by 1000 employees, the childcare sector is on average smaller by about 3 workers. Unfortunately, robustness checks cast doubt on the soundness of these main empirical results. In Chapter 2, we decompose employment growth into contributions from supply and demand factors. Policymakers require an accurate understanding of recent, high-frequency fluctuations in order to form accurate near-term forecasts and to craft appropriate fiscal and monetary policy actions. We adapt the methodology and sign-restriction approach in Shapiro (2024) to infer the presence of supply and demand shocks at the sector level and then aggregate to the labor market. We use data from the Current Employment Situation (CES) to measure employment levels across industries and data from the Employment Cost Index (ECI) to measure real compensation to workers. We demonstrate the importance of accounting for changes in employment composition by comparing our main results to a specification using CES wages instead. In Chapter 3, we generate upper bounds for the GDP increase that could occur if there was a substantial increase in housing supply in America's most productive areas. We write a relatively simple model of production and spatial equilibrium. We vary two parameters determine the maximal GDP increase: the extent to which local GDP increases with employment and how many workers are reallocated across space. If the elasticity of GDP with respect to employment is close to 1 (0.96), then the increase in GDP associated with GDP-maximizing reallocation ranges from 7% (if 16.3 million workers are moved) to 24% (if 64 million workers are moved). If the elasticity is much lower (.8125), then even moving 43 million workers will only increase GDP by 8 percent. Reducing barriers to building are only likely to generate larger increases to national income if there are dynamic benefits from agglomeration, either at the individual or place level.

This dissertation studies the provision, operation, and benefits of transportation-related public goods. It focuses on how policy, regulation, and market design can improve these urban systems. The three chapters examine highway procurement, road maintenance, and public transit. Together, they use new data and quasi-experimental variation to study why infrastructure is costly to build, who bears the costs of experiencing poor infrastructure, and how public agencies should set prices and service quality.

The first chapter studies whether limited competition in procurement auctions can explain the high and rising cost of U.S. road infrastructure. I assemble a new dataset covering the near-universe of state highway auctions between 2002 and 2024. I first document thin competition: one- or two-bidder auctions account for a third of awards, and this share has risen over the past decade. Using spatial variation in interstate bidder locations, I estimate that an additional bidder reduces prices by ten percent. I then develop a semi-parametric structural auction model to decompose bids into markups and production costs. The estimates imply that recent price growth is driven primarily by rising markups rather than rising production costs. Embedding the markup estimates in an entry model, I estimate large auction and market entry costs, consistent with an important role for procurement complexity and regulatory barriers.

The second chapter, coauthored with Edward Glaeser and Gabriel Kreindler, studies the distributional costs of road roughness. The chapter measures road roughness throughout the United States using vertical acceleration data from Uber rides across millions of road segments. It estimates drivers’ willingness to pay to avoid roughness from speed changes around salient changes in road quality, including town borders and repaving events. One standard deviation of road roughness generates losses of $0.33 per driver-mile. Rough roads are concentrated in poorer places and neighborhoods with larger Black populations, while resurfacing is only weakly targeted to the roughest roads.

The third chapter, coauthored with Lia Petrose, studies pricing and service quality in public transit. The chapter considers the problem of a welfare-maximizing transit agency subject to a budget requirement. Using rich MBTA data and quasi-experimental variation from a fare increase and salient service slowdowns, it estimates price and quality elasticities in a discrete-choice model with heterogeneity by neighborhood income. Riders are highly price-sensitive and moderately responsive to quality. Lower-income riders place greater value on lower fares and reliable service, but face fewer choice options. The estimates are used to study optimal subsidies and efficient pricing.

This dissertation comprises three essays examining how economic incentives shape political attention and how political attention shapes equilibrium outcomes. The first essay, joint with Shakked Noy, links the contemporary American culture war'' to changes in media technologies in the 1980s and to the cable news networks which responded to the new incentives of the era. We trace cable news' emphasis on cultural over economic issues to a distinctive business strategy: culture attracts viewers who would otherwise not watch news (mobilization''); economics attracts viewers who would otherwise watch competing news outlets (poaching''); and the number mobilized by culture is greater than the number poached by economics, so culture increases viewership. We show these incentives can account for a significant fraction of the time-series increase in cultural conflict since 1996. The second essay, joint with Leonardo Bursztyn, Georgy Egorov, Ingar Haaland, and Christopher Roth, examines the role of rationales in facilitating dissent. We experimentally show that liberals are more willing to express opposition to the movement to defund the police, are seen as less prejudiced, and face lower social sanctions when able to cite credible scientific evidence supporting their position. Analogous experiments with conservatives demonstrate that the same mechanisms facilitate anti-immigrant expression. The third essay, joint with Leonardo Bursztyn, Jonathan Kolstad, Pietro Tebaldi, and Noam Yuchtman, documents a pattern of political adverse selection'' in the health insurance exchanges established under the Affordable Care Act: Republicans enrolled at lower rates than Democrats and independents, a gap driven by healthier Republicans. This selection raised public subsidy spending by approximately $155 per enrollee annually (3.2% of average cost).

This thesis contains three chapters in macroeconomics. In the first chapter, co-authored with Robert Minton, we study how firms forecast their own future costs, and the implications of this for inflation. Firms’ forecasts of their own future costs are central to the propagation of shocks into inflation. Using survey data on US firms, we establish five novel facts about these forecasts: they (1) are driven by idiosyncratic cost movements, (2) are overly sensitive to a firm’s own costs (a failure of rational expectations), (3) over-react consistently across time, sectors, and firm size, (4) incorporate cost movements slowly over time, and (5) under-react to aggregate shocks until costs move. We estimate a model of firms’ beliefs that is able to match these facts, and embed this in a New Keynesian model. The New Keynesian Phillips curve becomes less forward-looking and steeper, and we confirm this in external pricing data. Supply shocks are made more inflationary, because they hit costs quickly, leading firms’ beliefs to over-react, while demand shocks are made less inflationary, as firms fail to anticipate future wage pressure. We show that optimal monetary policy is less reliant on forward guidance and commitment, since it is difficult to move firms’ beliefs without first moving their costs. In contrast, current interest rates remain powerful, and may have lasting effects, suggesting a sharp optimal reaction to inflationary shocks.

In the second chapter, I study the role of inventories in propagating demand fluctuations. Inventory investment accounts for a significant share of the business-cycle variation in GDP. I develop a tractable New Keynesian model with inventories to analytically explore their role in propagating demand fluctuations. Persistent increases in demand are amplified upstream, while transitory ones are smoothed out. The slope of the supply curve is increasing in the persistence of demand fluctuations, as markups respond more to more persistent shocks. And inventories amplify convexity in the supply curve, making larger positive demand shocks even more inflationary. In a quantitative model, I show that this demand channel of inventory investment amplifies the peak effect of monetary policy on GDP and inflation, but only when the shock is sufficiently persistent. Likewise inventories boost the peak GDP response to very persistent demand shocks, but dampen the response to transitory shocks. Finally, I show that the supply curve is more convex when shocks are more persistent, suggesting that the key to a nonlinear Phillips curve is that the demand shocks be both large and persistent.

In the third chapter, co-authored with Gert Bijnens, Helene Hall, and Laura Nicolae, we use quasi-randomly assigned wage indexation events in Belgium to study the effects of wages on employment. In Belgium, nearly all employees’ wages are indexed to inflation. Firms are grouped into long-standing labor agreements that determine the exact timing and frequency at which wages are indexed, e.g. every year versus every month. Using firm-level administrative data, we leverage the resulting variation in real wages across firms to estimate the employment response. We find that firms that are forced to increase wages by 1% react by cutting employment by 0.4% over four quarters. This result is robust to including NACE sector-date fixed effects and to using only variation in firms’ indexation timing, controlling for their chosen indexation frequency. About one-third of the response comes via anticipation of future wage increases. The elasticity is more than twice as large in magnitude in the post-pandemic period than before it, suggesting strong nonlinearities. Overall, these results show that, by preventing inflation from reducing real wages, inflation indexation reduces employment.

In this dissertation, I use tools from labor and public economics to study the effectiveness of policies aimed at improving access to public goods and services and how the benefits of these policies are distributed across key decision-makers and recipients.

In the first essay, joint with Shreya Tandon, we estimate the causal effects of new primary care physician arrivals to underserved areas on net physician supply, healthcare utilization, and health outcomes of local residents using quasi-random variation in the timing of physician placements through the National Health Service Corps (NHSC) program. Our primary data include the roster of NHSC physician recipients and a sample of Medicare claims from 1999--2019. We find that physician arrivals increase overall primary care supply. Half of the non-inpatient healthcare provided by new entrants would have been satisfied by the incumbent primary care physicians, while the other half constitutes an increase in overall utilization. The additional utilization is concentrated in more advanced preventive care services, such as advanced testing and imaging. We also find an overall increase in elective inpatient procedures, particularly related to cardiovascular diagnoses. The greater healthcare utilization is associated with significant reductions in emergency hospitalizations and mortality. Health improvements persist for at least four years following physician arrival and are particularly pronounced for beneficiaries with chronic conditions and those living in more disadvantaged and rural locations.

In the second essay, joint with Leonardo D'Amico, we examine how a persistent change in mortgage rates affects city growth, home prices and construction, and household formation. We study the revolutions in mortgage financing that took place in the U.S. between 1933 and 1940, which created a national mortgage market and allowed mortgage capital to move from the financial centers to the rest of the country. By digitizing city-level census data and a new sample of loan-level data, we show that differences in mortgage rates across cities went from nearly 300 basis points to just over 100 in only six years. This national mortgage market allowed initially capital-scarce places to grow more than initially capital-abundant ones. In the decades following the housing policies, cities where mortgage rates declined more as a result of the integration of mortgage markets saw higher growth in rates of homeownership, population, and housing construction. House prices moved only modestly, suggesting that, in 1940--50, housing supply responded robustly to higher demand for homes. These macro-level policies also influenced intra-household decision-making: we find that women who experienced lower mortgage interest rates during their childbearing years had more children.

In the third essay, joint with Will Dobbie and Crystal Yang, we study a common feature of policies that deploy predictive algorithms in high-stakes settings: despite the availability of algorithmic recommendations, human decision-makers are typically kept in the loop and can exercise discretion to override them. We ask whether these discretionary overrides add valuable private information or merely reintroduce human biases and mistakes, in the context of bail decisions. We develop new quasi-experimental tools to measure the impact of human discretion over an algorithm on the accuracy of decisions, even when the outcome of interest is only selectively observed. We find that 90 percent of the judges in our setting underperform the algorithm when they make a discretionary override, with most making override decisions that are no better than random. Yet the remaining 10 percent of judges outperform the algorithm in terms of both accuracy and fairness when they make a discretionary override. We provide suggestive evidence on the behavior underlying these differences in judge performance, showing that the high-performing judges are more likely to use relevant private information and are less likely to overreact to highly salient events compared to the low-performing judges.