This dissertation consists of three essays on various aspects of development and the political economy of developing countries. The first two chapters share a focus on issues of political economy in South Asia, the first examining the influence of politics over public goods allocations, and the second the effects of ethno-religious politics on voter behavior, violence, and policy outcomes. The third chapter shares with the first two its geographic setting, being located in South Asia, but focuses on education, employing an RCT design to evaluate the efficacy of public-private partnerships in delivering high-quality primary education to remote communities. The first chapter examines the role of political parties in India's national government in shaping public goods allocations. Party preference is often regarded as important for shaping policy outcomes, but the empirical literature has yielded mixed results, with some research finding substantial party effects, and other research little to none. The discrepancies in estimated party effects are likely due to a combination of heterogeneous party characteristics and institutional context, as well as the the nature of political competition itself, with parties facing a trade-off in the promotion of their most preferred policies against the electoral incentive to cater to the median voter. To generate random random variation in party identity, I make use of the assassination of the Congress party leader, Rajiv Gandhi, in the midst of India's 1991 national elections, which had the effect of dramatically increasing the probability of Congress victory for a subset of constituencies. Using this variation, I find that representation by the ruling Congress party leads to a substantial increase in the provision of public goods favored by the poor, consistent with the party's expressed populist agenda. Among the salient changes are increases in the availability of drinking water and declines in infrastructure such as productive electrification and paved roads. I also estimate party effects using a regression discontinuity identification strategy, which generates variation in party identity for closely contested elections. Here I find little effect of Congress representation on public goods allocations. I argue that the reason for the differences between the results estimated with the two identification strategies is the importance of both the identity of the winning party, as well as the margin of victory. The second chapter examines the role of ethno-religious propaganda in generating support for political parties espousing ideologies of ethno-religious nationalism. A significant literature has shown the effects of political campaigns and media bias in influencing voter behavior. Ethnic identity often figures prominently in campaigns of voter mobilization, particularly in developing countries, where ethnic identities tend to be more salient, and state resources more subject to capture through power over the state. A large body of research has shown the ways in which, not only does ethnic diversity create an environment conducive to the ethnicization of political competition, but political competition itself contributes to the increased salience of ethnic identity.Prior to India's 1991 national elections, the leader of the Hindu-nationalist BJP political party toured northern India on a "pilgrimage" to the city of Ayodhya, holding numerous rallies along the way to promote the construction of a Hindu temple there. Causal identification of the campaign's effects comes through the incidental exposure of localities due to their lying along the road joining the cities which were the ultimate destinations of the campaign. The main result is that the campaign increased the BJP's vote share by 5-9 percentage points in visited constituencies, which translated to a 10-20 percentage points increase in the probability of victory. I also find that the campaign significantly increased the probability of riots, which were 9 percentage points more likely to occur in constituencies through which the campaign passed; and that the riots associated with the campaign increased the party's vote share by 3.5 percentage points. There is also evidence that the campaign increased the availability of local public goods, with the sub-district through which the campaign directly passed showing a 3-6 percentage points increase in a variety of public goods, such as electrification, drinking water, and primary schools. The third chapter, which is jointly authored with Leigh Linden, Felipe Barrera-Osorio, Dhushyanth Raju, and Matthew Hoover, examines the efficacy of public-private partnerships for delivering high-quality primary education to remote, and underserved, communities. Private entrepreneurs were enlisted to establish and operate primary schools throughout rural Sindh province in Pakistan, for which they were paid a per-child subsidy, with all local children between the ages of 5 and 9 allowed tuition-free enrollment. To address potential sources of endogeneity, the intervention was designed as a randomized control trial (RCT): 263 villages were identified as qualifying for the program, of which 200 were randomly assigned a school. In addition, half of the treatment villages were assigned a subsidy scheme whereby entrepreneurs were paid slightly more for girls than boys. The program proved remarkably effective, with enrollment increasing by 30-50 percentage points. Child test scores also improved considerably, with children in treatment villages scoring 0.67 standard deviations higher on administered exams. Interestingly, there was no differential effect on female enrollment for either subsidy scheme, which we attribute to the lack of a pre-existing gender gap in enrollment.

This dissertation presents three essays in Political Economy with different approaches, but a single line of inquiry: how can political institutions shape individual behaviors by modifying the incentives of political actors? Krugman and Wells (2005) defines economics as "the study of economies, at both the level of individuals and of society as a whole" and an economy as "a system for coordinating society's productive activities." Political Economy, in parallel, can be seen at the study of politics, at both the level of individuals and of institutions as a whole, where institutions are defined as systems to coordinate individuals' interactions. The two dimensions are important: although politics consists in decisions taken at the individual level, the outcomes are shaped by the institutional rules which thus partly determines those choices. The three chapters presented here consider particular cases of this interdependence between individual political actors and political institutions. Chapter 1 analyzes how the effective super-majority in the US Senate along with the role of parties as imperfect coordinators of politicians' actions affect the incentives of the centrist senators; and suggests in a stylized model that, counter-intuitively, a smaller minority might be more successful in its effort to fight the majority's priorities. Chapter 2 studies empirically how changes in a country's constitutional executive term limits affect the incentives of politicians and the consequences on a country's default probability by considering the effect those shocks have on the perception that international investors have of a country's financial soundness. Chapter 3 completes the parallel between the standard definition of Economics and Political Economy by investigating the understudied extension of markets for goods to markets for votes, and shows that the idiosyncratic characteristics of votes imply that a typical market performs badly in allocating the decision power to the parties valuing it the most. This dissertation not only tackles a series of problems in Political Economy, but also discusses and develops a wide range of methods which are available to understand those issues. Chapter 1 proposes a participation game model where a certain number of contributors are required to pay in order for a public good to be provided. The main theoretical contribution of this paper is to show that when the contribution cost falls in the number of ex-post contributors, not only individual participation is more likely when the required number of participants increases with the size of the group, but the provision probability increases too. On the contrary, this does not occur in a fixed cost model. One practical implication of the model suggests that if a party in the US Senate keeps its majority while losing seats at the center of the political spectrum, it might be more successful in overcoming a cloture vote without any change in policy ideology. This chapter then uses a laboratory experiment to test the model's predictions and underlines how, generally, simple experiments can guide theorists to first find identifiable, testable comparative statics predictions, and second, design experiments which would not be easily replicated in the field and provide clean identification. The experimental results also show the importance of using models with testable implications: although the theory's predictions on individual behavior are qualitatively borne out by the data, the quantitative deviations from standard "rational" behavior as expressed in game theoretical solution concepts differ across the set of parameters and generate aggregate outcomes which do not match the theory exactly. Optimization-based models with additional, behavioral elements, or models of bounded rationality which are discussed in part in that chapter should thus also be an integral part of political economy models: a general equilibrium model which answers its motivating question under the assumption of perfect rationality will only be of limited use if it is not robust to the individual deviations from this assumption that we observe in reality. Chapter 2, co-authored with Laurence Wilse-Samson, is an empirical study which uses an event-study methodology to uncover the impact of changes in a country's constitutional executive term limits on international investors' perception of that country's risk, by analyzing the evolution of bond market spreads around the time of those changes. It provides two main contributions, one methodological, and the other empirical.. The flourishing literature on institutions mainly considers the impact of institutions on low-frequency variables such as fiscal outcomes, while this study uses high-frequency financial data. The trade-off in these two approaches is informative. With high frequency data and using event-studies, the identification is clear: any movement in financial markets can be linked to the institutional change under investigation. However, failures of rational expectations means that this impact on expectations might differ from the effect on realized economic variables. This chapter thus emphasizes that while these two types of analyses are complementary, high-frequency analyses are underused. On the empirical side, the chapter considers the unresolved debate over the impact of term limits on fiscal outcomes, as underlined by contradictory results in the empirical literature. Moreover, theories developed on term limits also suggest ambiguous effects: for instance, do term limits prevent insiders from controlling the political process, or do they prevent elections from creating incentives for the executive to behave well? The chapter considers the movement of bond spreads around term-limits "shocks" and shows that although bond spreads fall after restrictions on term limits, there is no significant impact of extensions. Furthermore, it provides suggestive evidence that the impact of such shocks is larger in relatively weakly institutionalized countries, and that the separation of branches also matter to investors since restrictions implemented by the judiciary also generate strong movements. Finally, Chapter 3, co-authored with Alessandra Casella, is motivated by the simple question of whether in a committee of members belonging to two opposing parties and voting on a binary decision, markets, which have been thoroughly studied in economic theory and are considered to function quite well in allocating goods to the agents valuing them the most, can work in allocating votes and decision power in the same way. Generally, one question in thinking about voting mechanisms has been that formulated by Dahl (1956): "What if a minority prefers an alternative much more passionately than the majority prefers a contrary alternative? Does the majority principle still make sense?". A market for votes appears like an intuitive way to allow members of a committee to sell and buy votes using a numeraire, but this chapter shows that it is unable to do so in an efficient way and usually performs worse than majority voting, in particular in a large electorate. A market for votes indeed yields a competition between the higher-intensity member of each party irrespectively of the size of those parties, which generates a systematic bias in favor of the minority which will win too often. In particular, it is shown that for any party sizes, the probability of a minority victory converges to a half as the electorate becomes infinitely large. The model also emphasizes other inefficiencies: this institution implies intra-party trade and supermajorities. Importantly, the implications of the model have been tested in a laboratory experiment in a previous paper and are generally verified by the experimental results.

This dissertation has been motivated by my two broad areas of interest - the nonprofit sector and health care, - and in particular by the growing number of community health centers in the United States in the last decade. The dissertation consists of three chapters: In the first chapter, I estimate the causal impact of community health centers on the utilization of emergency rooms in California, in 2006-2010. In the second chapter, I study the San Francisco healthcare reform of 2007-2008. And in the third chapter, I analyze California's Nonprofit Integrity Act of 2004 and its impact on governance and revenue growth of nonprofit organizations. The First Chapter is titled "The Effect of Community Health Centers on Utilization of Emergency Care: Evidence from California." The United States has a long list of programs that aim to reduce disparities in access to health care, including public health insurance, subsidies, and health insurance mandates on employers, insurers, and private citizens. This paper focuses on an alternative but complementary way of bringing medical care to medically underserved populations: community health centers (CHCs). CHCs provide primary health care services to two main groups - the uninsured and Medicaid enrollees - who face barriers to accessing primary health services and often turn to their local emergency departments (EDs) for non-urgent medical care. To examine the effect of CHCs on ED visits that do not result in an immediate hospital admission, I first use an event study that relies on an ED patient's proximity to a CHC and exploits the variation in the number of clinics in each zip code over time. To deal with endogenous clinic entry, I use a second differences-in-differences strategy that uses proximity to a highway to model the patient's level of access to a CHC. I find that recent expansions in the CHC sector have been effective in reducing the number of non-hospitalization ED visits by uninsured non-elderly adults and have had no impact on the number of ED visits by patients with private or public insurance. The results are robust to different measures of access, sample composition, and choices of specification. The Second Chapter is titled "Impact of the San Francisco Health Care Reform of 2007-2008 on the Utilization of Emergency Departments." In July 2007, the city and county of San Francisco launched the Healthy San Francisco Program (HSF) that aimed to substantially increase access to a broad range of medical services for uninsured residents of San Francisco. In March 2011, about 54,500 uninsured San Franciscans between the ages of 18 and 64 (or roughly two thirds of the city's non-elderly adults without any medical coverage) were enrolled in HSF to take advantage of affordable and well-coordinated primary and preventive health care. In the first half of 2008, the city also enacted the Health Care Security Ordinance (HCSO), by which San Francisco-based employers must spend a certain minimum amount on employee health care either through the provision of employer-sponsored health insurance, or by contributing to the Healthy San Francisco or Health Savings Accounts. In this paper, I analyze the impact of Healthy San Francisco and the Health Care Security Ordinance on the utilization of emergency rooms by non-elderly adults in San Francisco. My main specification is based on a differences-in-differences strategy, in which I take zip codes in the nearby Bay Area as a comparison group. I also perform a within-San Francisco analysis. Both strategies produce a consistent set of results for uninsured patients. Overall, I find that HSF and HCSO reduced the number of non-admission ED visits by uninsured by an average of 30.1% between 2007 and 2009-2010. However, because this reduction coincides with a higher use of emergency rooms by patients with private insurance and by those covered by county funds, the lower number of ED visits by uninsured may be partially due to their "switch" to a different payor category, rather than their actual substitution away from EDs as a source of regular care. I discuss the validity of these results in the context of various control groups and data limitations. The Third Chapter is titled "Impact of the Nonprofit Integrity Act (2004) on Governance and Growth of Charities in California." The Nonprofit Integrity Act (NIA) took effect in California on January 1, 2005. Its goals are to improve governance and financial transparency in the charitable sector, to increase its oversight, and to prevent fraudulent solicitations. NIA imposes certain restrictions on charity managers and boards, and explicitly requires organizations whose annual revenue is at or above $2 million (net of government grants) to have an independent audit of financial statements. In this paper, I evaluate NIA's impact on large charitable organizations in California. Specifically, I use a simple differences-in-differences framework to examine the effect of NIA on various financial and governance outcome measures. I find that in the two years following the enactment of NIA, California charities with an average 2002-2004 (pre-treatment) income "just below" the $2 million threshold grew by about 7.7-7.9% less than similar charities in other U.S. states, and by 15-18% less than California charities with the pre-treatment revenue "just above" $2 million. These differences in growth rates are not detectable by year 2007. The Nonprofit Integrity Act appears to have had little impact on other major financial measures of charities. With respect to the available set of governance measures, it seems that the majority of California's charitable organizations affected by NIA had already used select NIA-prescribed governance and management practices before the law came into effect. It is also possible that the adoption of some new governance measures (e.g., audit committees) and their incorporation into by-laws did not happen immediately following NIA's enactment because of natural organizational rigidities.

This dissertation consists of three chapters on different aspects of economic inequality. In the first chapter, I study the aggregate implications of health risk and access to health care. At the individual level, health influences earnings potential, while income affects access to medical care. I investigate how this interaction shapes the joint dynamics of inequality in health and earnings over the life cycle, and I measure the redistributive impact of policies that improve access to health care. For that, I introduce health shocks and health care spending in an incomplete markets model with heterogeneous agents. Earnings risk is partially determined within the model due to the health-income feedback, and negative shocks may drive agents into a low income-low health trap, thus magnifying inequality along the life cycle. I estimate the process for health shocks and I calibrate the key parameters of the model using survey data. The calibrated model successfully reproduces the joint dynamics in health and earnings inequality in the life cycle. Like in the data, it predicts that life cycle inequality in health is driven by a sharp decline in health status for the lowest percentiles of the health distribution. I find that the health-income feedback accounts for 9 percent of total earnings inequality at retirement age as measured by the coefficient of variation of earnings, and that it increases by almost seven times the persistence of shocks to productivity. I also find that health care policies that facilitate access to health care have redistributive effects, mostly through earnings improvements for those at the bottom of the earnings distribution. The second chapter, joint with Stefania Albanesi, studies the connection between recent trends in earnings inequality and the behavior of labor supply of married women in the U.S. The entry of married women into the labor force and the rise in women's relative wages are amongst the most notable economic developments of the twentieth century. These phenomena were particularly pronounced in the 1970s and 1980s, when participation of married women grew from 38\% in 1975 to a peak of 60\% in 1996 and the male to female ratio in hourly wages dropped from 1.60 to 1.34. Since the early 1990s, the growth in these indicators has stalled, especially for college graduates. This development is puzzling in light of the continued rise in women's educational investments relative to men and their entry into professional occupations. In this paper, we link the decline in married women's participation and wages relative to trend since the early 1990s to the growth of the skill premium, which substantially accelerated in those years. Our hypothesis is that the growth in wages for highly educated men generated a negative wealth effect on the labor supply of their female spouses, reducing their labor supply and their wages relative to men. Disaggregated evidence on skill premia by gender, gender wage gaps by education and labor force participation of wives provides descriptive support for this mechanism. Specifically, starting in the early 1990s, the growth in the skill premium was lower for women, while convergence in wages across gender slowed more for college graduates. Finally, participation of married women declined starting in the early 1990s especially for college women, women married to men with a college degree or to men in the top percentiles of the earnings distribution. We develop a model of household labor supply which can qualitatively reproduce a negative effect on wives' participation of a rise in husbands' earnings. We show that a calibrated version of the model can account for more than half the decline relative to trend in married women's participation in 1995-2005, and more than two thirds for college women. The model can also account for one third of the rise in the gender wage gap for college graduates relative to trend in the same period. In the third chapter I study the dynamics of earnings risk and inequality over the life cycle for women, and document the gender differences in earnings stochastic processes faced by workers. Female workers have a weaker average attachment to the labor force than their male counterparts, and career interruptions have an impact on earnings. Therefore, it is to be expected that the average earnings process differ by gender, and in this paper I study if this is the case. The main empirical gender asymmetries I find in the profiles of earnings are: i) inequality is lower amongst women than amongst men, ii) inequality peaks twice over the life cycle for women: once during the fertile years, and the again later at retirement age, iii) the differences in inequality evolution between educational groups are larger for men than for women. I estimate the statistical properties of the earnings process, with and without heterogeneity in age profiles, and find that the specification without profile heterogeneity seems to fit the female workers dynamics better.

This dissertation investigates the sources of real business cycle fluctuations in emerging countries, using a combination of real business cycle theory and econometric techniques. The first chapter consists of two main sections. In the first section, I empirically evaluate the canonical dynamic stochastic general equilibrium model of a small open emerging economy using bayesian methods. I show that estimated dynamic models of business cycles in emerging countries deliver counterfactual predictions for the country risk premium. In particular, the country interest rate predicted by these models is acyclical or procyclical, whereas it is countercyclical in the data. The second section proposes and estimates a small open economy model of the emerging-market business cycle in which a time-varying country risk premium emerges endogenously through a variant of the financial accelerator mechanism as in Bernanke, Gertler, and Gilchrist (1999). In the proposed model, a firm's borrowing rate adjusts countercyclically as the productivity default threshold depends on the state of the macroeconomy. I econometrically estimate the proposed model and find that it can account for the volatility and the countercyclicality of the country risk premium as well as for other key emerging market business cycle moments. Time varying uncertainty in firm specific productivity contributes to delivering a countercyclical default rate and explains more than 65 percent of the variances in the trade balance and in the country risk premium. Finally, I find that the predicted contribution of nonstationary productivity shocks in explaining output variations falls between the high estimate reported by Aguiar and Gopinath (2007) and the low estimates reported by Garcia-Cicco, Pancrazi, and Uribe (2010). In the second chapter, I investigate the extent to which global financial conditions contribute to the macroeconomic fluctuations in emerging economies. Using a panel structural VAR model, I find that global risk shocks are important contributors to the dynamics of the country risk premium and real macroeconomic variables. In particular, I find that global risk shocks explain about 20 percent of movements both in the country risk premium and in the economic activity in emerging economies. The contribution of U.S. real interest rate shocks to macroeconomic fluctuations in emerging economies is negligible. I argue that the role of U.S. interest rate shocks in driving the business cycles in emerging economies, as emphasized in the previous literature, is taken up by global risk shocks. The country risk premium shock also has significant explanatory power of emerging economy real business cycle fluctuations. Global financial shocks altogether account for about 45 percent of the aggregate fluctuations in emerging economies. I find that domestic macroeconomic variables including domestic banking sector risk have sizable impact on the country risk premium fluctuations. I argue that the linkage between the economic activity and the country risk premium is the key mechanism through which global risk shocks are transmitted to emerging economies.

This dissertation consists of three essays on macroeconomics and business cycles. In the first chapter, written with Nicolas Crouzet, we ask whether news shocks, which change agents' expectations about future fundamentals, are an important source of business-cycle fluctuations. The existing literature has provided a wide range of answers, finding that news shocks can account for 10 percent to 60 percent of the volatility of output. We show that looking at the dynamics of inventories, so far neglected in this literature, cleanly isolates the role of news shocks in driving business cycles. In particular, inventory dynamics provide an upper bound on the explanatory power of news shocks. We show, for a broad class of theoretical models, that finished-good inventories must fall when there is an increase in consumption and investment induced by news shocks. When good news about future fundamentals lowers expected future marginal costs, firms delay current production and satisfy the increase in demand by selling from existing inventories. This result is robust across the nature of the news and the presence of different types of adjustment costs. We therefore propose a novel empirical identification strategy for news shocks: negative comovement between inventories and components of private spending. Estimating a structural VAR with sign restrictions on inventories, consumption and investment, our identified shock explains at most 20 percent of output variations. Intuitively, since inventories are procyclical in the data, shocks that generate negative comovement between inventories and sales cannot account for the bulk of business-cycle fluctuations. The second chapter looks into the dynamics of durables over the business cycle. Although transactions of used durables are large and cyclical, their interaction with purchases of new durables has been neglected in the study of business cycles. I fill in this gap by introducing a new model of durables replacement and second-hand markets. The model generates a discretionary replacement demand function, it nests a standard business-cycle model of durables, and it verifies the Coase conjecture. The model delivers three conclusions: markups are smaller for goods that are more durable and more frequently replaced; markups are countercyclical for durables, resolving the comovement puzzle of Barsky, House, and Kimball (2007); and procyclical replacement demand amplifies durable consumption. In the third chapter, written with Ricardo Reis, we study the macroeconomic implications of government transfers. Between 2007 and 2009, government expenditures increased rapidly across the OECD countries. While economic research on the impact of government purchases has flourished, in the data, about three quarters of the increase in expenditures in the United States (and more in other countries) was in government transfers. We document this fact, and show that the increase in U.S. spending on retirement, disability, and medical care has been as high as the increase in government purchases. We argue that future research should focus on the positive impact of transfers. Towards this, we present a model in which there is no representative agent and Ricardian equivalence does not hold because of uncertainty, imperfect credit markets, and nominal rigidities. Targeted lump-sum transfers are expansionary both because of a neoclassical wealth effect and because of a Keynesian aggregate demand effect.

My research has two main themes --- the link between political economy and economic development, and the causes and effects of economic crises and long recessions. This dissertation samples from some of this ongoing research. The relationship between economic development and democracy is key in political economy. Many commentators have suggested that economic growth increases support for democracy. One proposed mechanism is that modernization, by reducing the demand for low-skilled labor, increases the willingness of elites, particularly in agriculture, to extend the franchise. In Chapter 1 I use subnational variation in South Africa to test this mechanism. I employ national shocks to the mining sector's demand for native black workers and cross-sectional variation in labor market competition induced by apartheid to estimate the effect of black labor scarcity on wages, capital intensity, and changes in partisan voting preferences. I find that reductions in the supply of foreign mine labor following the sudden withdrawal of workers from Malawi and Mozambique (and the increased demand for native black workers) increased mechanization on the mines and on farms competing with mines for labor. I then show that these induced structural changes resulted in differential increases in pro-political reform vote shares in the open districts relative to closed districts, even as mining districts became more conservative and voted more to maintain the non-democratic regime. Chapter 2 also explores issues related to the close relationships between economic and political institutions. In this chapter, together with my coauthor Sebastien Turban, we show how sovereign debt spreads are impacted by news about executive term limits. Political institutions matter for countries' cost of borrowing. We use an event-study to analyze the markets' response to new information about executive term limits over 101 events in seven emerging markets. Investors respond significantly to news about restrictions on those limits, lowering risk spreads. The one day abnormal returns following news about a restriction is 2 percentage points. Over ten days, the cumulative abnormal return is 5 percentage points. News about term limits extensions are not significant in the medium run. The results are robust to a non-parametric test and are confirmed when looking at the behavior of sovereign CDS prices. Chapter 3 starts the second part of this dissertation which is an investigation into the housing-related aspects of the recent crisis which began as a "subprime crisis" before it became the "Great Recession". In particular, this chapter focuses on the institutional details underpinning these markets. It also serves to set up the analysis in the following chapter which looks at one of the potentially important mechanisms which amplified the severity of the housing crisis. One important feature emerging from this analysis is that it appears that protections for home mortgage creditors were strengthened in the period preceding the subprime crisis. This may have both increased lending, but also the difficulty of modifying home loans ex post. This is more problematic to the extent that there are negative externalities from foreclosures. Chapter 4, co-authored work with David Munroe, shows that completed foreclosures cause neighboring foreclosure lings. We estimate this relationship using administrative data on home foreclosures and sales in Cook County, IL, instrumenting completed foreclosures with randomly assigned chancery-court judges. A completed foreclosure causes 0.5 to 0.7 additional foreclosure lings within 0.1 miles, an effect that persists for several years. Contagion is driven by borrowers on the margins of default, not those severely at risk. We find evidence that borrowers learn about lender behavior from neighboring foreclosures. Finally, a foreclosure causes an increase in housing sales among relatively low-quality properties.

A central concern in international economics and urban economics is explaining the distributions of economic assets and activity across space. This dissertation contains three essays examining the pattern of specialization across US cities. Chapter 1 investigates the determinants of quality specialization within products. A growing literature suggests that high-income countries export high-quality goods. Two hypotheses may explain such specialization, with different implications for welfare, inequality, and trade policy. Fajgelbaum, Grossman, and Helpman (2011) formalize the Linder (1961) conjecture that home demand determines the pattern of specialization and therefore predict that high-income locations export high-quality products. The factor-proportions model also predicts that skill-abundant, high-income locations export skill-intensive, high-quality products (Schott, 2004). Prior empirical evidence does not separate these explanations. I develop a model that nests both hypotheses and employ microdata on US manufacturing plants' shipments and factor inputs to quantify the two mechanisms' roles in quality specialization across US cities. Home-market demand explains at least as much of the relationship between income and quality as differences in factor usage. In Chapter 2, Donald R. Davis and I develop a theory to jointly explain the distributions of skills, occupations, and industries across cities. Our model incorporates a system of cities, their internal urban structures, and a high-dimensional theory of factor-driven comparative advantage. It predicts that larger cities will be skill-abundant and specialize in skill-intensive activities according to the monotone likelihood ratio property. We test the model using data on 270 US metropolitan areas, 3 to 9 educational categories, 22 occupations, and 21 manufacturing industries. The results provide support for our theory's predictions. Chapter 3 examines whether larger cities are attractive to consumers. Popular and academic discussions celebrate the virtues of large cities for consumption and leisure. But the standard spatial-equilibrium account says that the consumer attractions of larger cities cannot account for their higher nominal wages and more skilled populations. This chapter revisits that conclusion and shows that the consumption motive can play a first-order role in spatial variation in wage distributions when individuals are heterogeneous. I present a general-equilibrium model in which larger cities offer a greater variety of local goods and services, attracting higher-income individuals who value such variety relatively more. Despite the absence of production-related agglomeration economies, the equilibrium outcomes match a series of facts about spatial variation in wage distributions. I present evidence on the spatial choices of retirees, who consume but do not produce, that is consistent with consumption-driven agglomeration.

This dissertation uses administrative data to study regulatory issues in the American real estate market. The first chapter studies spillovers from home foreclosures in Cook County, Illinois. Random assignment of foreclosure cases to judges allows for estimation of the causal effect of foreclosure (relative to a foreclosure case being dismissed) on neighboring foreclosure filings and housing transactions. When a property forecloses, the local housing market is disrupted--prices fall and more lower quality homes sell--and neighbors are more likely to end up in default and going through the foreclosure process. The second chapter examines how discontinuously applied transfer taxes distort the mar- ket for real estate sales in New York and New Jersey. These transfer taxes distort not only the price of real estate transactions that occur near the discontinuity, corresponding to sellers bearing the entire incidence of the tax, but also the volume of sales that occur--productive transactions that would occur if the tax were not discontinuous disappear from the market. The third and final chapter estimates the market-level response of home equity loans to two discontinuous mortgage policies--the home mortgage interest deduction, and real estate appraisal regulations in the Financial Institutions Reform Recovery and Enforcement Act. The estimates therein imply that home equity debt is very responsive to both the after-tax interest rate as well as lender underwriting requirements.

The increasing availability of detailed, individual-level data from retail settings presents new opportunities to study fundamental issues in product design, price discrimination, and consumer behavior. In this set of essays I use a particularly rich data set provided by a major fashion goods manufacturer and retailer to illustrate how observed firm strategies correspond to predictions from producer theory. I present evidence on the importance of multidimensionality in consumer preferences, both within the theory of price discrimination and as a factor in actual firm decisions. Finally, I explore the applicability of concepts from signaling theory and behavioral economics in explaining consumer purchase decisions. The first chapter describes the empirical setting used throughout the entire dissertation. Data is provided by a luxury goods firm that dominates its category of fashion goods in the United States. The firm operates hundreds of stores in the US, with different types of stores differing markedly in their geographic accessibility to consumers. I present and estimate a model of demand that admits consumer heterogeneity in two dimensions: travel sensitivity and product age sensitivity. I show that consumer heterogeneity in these two dimensions outweigh that in observable characteristics, such as household income. Furthermore, I estimate a high correlation in the two dimensions, such that consumers who are most averse to travel are also those for whom product newness is most valuable. The second chapter focuses on the firm's store location and product introduction strategies. I introduce a model of product introduction wherein the firm selects only the parameters of the distribution of product characteristics, rather than the characteristics of each new product. This dramatically simplifies the firm's optimization program. I use this model to simulate counterfactual product assortments given alternative store location decisions. I show that the optimality of observed store locations depends substantially on the correlation in consumer values for travel distance and product quality. I also show that increased differentiation in geographic accessibility enables the firm to profitably increase differentiation in product quality. The third chapter studies how consumers respond to different price signals conditional on store visitation. Many firms employ price comparisons as a selling strategy, in which actual prices are framed as discounted from a high list price, occasionally even when no units are sold at list prices. I show that high list prices enhance demand both on product and store levels. I present evidence that suggests that consumers infer quality from list prices. I also demonstrate that these demand-enhancing effects are dependent on characteristics of the retail context, such as the general level and dispersion of discounting. These essays study in isolation components of consolidated selling strategies that have been widely adopted by US manufacturers and retailers across a wide variety of categories. My hope is to achieve a deeper understanding of the aspects of consumer behavior and firm incentives that have led to the prevalence of these selling strategies. This understanding is central in forming prescriptions for managers as well as measuring welfare implications, both of which I leave for future work.