Search

My dissertation consists of two chapters. In the first chapter, I show that the growing trend in financial services digitalization has introduced a new dimension along which commercial banks compete, with consequences for the local economy. Small community banks (SCBs) are slow to implement mobile technologies and lose deposits to larger, better-digitalized banks following mobile infrastructure improvements. This dynamic negatively affects their small business lending, for they have historically relied on information and liquidity synergies with deposits to maintain their competitive advantage in such markets. Larger banks and FinTech firms prove to be imperfect substitutes in this setting, and the local economy benefits less from digitalization in areas where SCBs had an important presence before its advent. The second chapter, co-authored with Prof. Rawley Heimer, focuses on the outcomes of consumers' efforts to achieve restitution for disputed financial services. We find that complaints filed with the Consumer Financial Protection Bureau (CFPB) from low-income and Black zip codes are 30% less likely to be resolved with the consumer receiving financial restitution. The gap in financial restitution was scarcely present under the Obama administration, but grew substantially under the Trump administration. We attribute the change in financial restitution under different political regimes to companies anticipating a more industry-friendly CFPB, as well as to the more industry-friendly leadership of the CFPB achieving less financial restitution for low-income and Black filers. The financial restitution gap cannot be explained by differences in product usage nor the quality of complaints, which we measure using textual analysis.

With American families living on the financial edge and seeking out high-cost loans even before COVID-19, the term financial technology or “fintech” has been used like an incantation aimed at remedying everything that’s wrong with America’s financial system. Scholars and supporters from both the public and private sector proclaim that innovations in financial technology will “bank the unbanked” and open new channels to affordable credit. This exuberance for all things tech in finance has led to a quiet yet aggressive deregulatory agenda, including, as of late, a federal assault via rulemaking on the ability of states to police the cost and privilege of extending credit within their borders. This deregulation and the ethos behind it have made space for growth in high-cost, predatory lending that reaches across state lines via websites and smart phones and aggressively targets cash-strapped families. These loans are made using a business model whereby funds are funneled through a group of lightly regulated banks in a way designed to take advantage of federal preemption. Fintech companies rent out and profit from the special legal status of these bank partners, which in turn keeps the bank’s involvement in the shadows. Stripping down predatory fintech’s practices and showing them for what they really are, this Article situates fintech in the context of this country’s longstanding dual banking wars, both between states and the federal government and between consumer advocates and banking regulators. And it points the way forward for scholars and regulators willing to shake off fintech’s hypnotic effect. This means, in the short term, using existing regulatory tools to curtail the dangerous lending identified here, including by taking a more expansive view of what it means for a bank to operate safely and soundly under the law. In the long term, it means having a more comprehensive and national discussion about how we regulate household credit in the digital age, specifically through the convening of a Twenty-First Century Commission on Consumer Finance. The Article explains how and why the time is ripe to do both. As the current pandemic wipes out wages and decimates savings, leaving desperate families turning to predatory fintech finance ever more, the need for reform has never been greater.

This dissertation contains three essays, each of which uses high-quality data and rigorous econometric methods to further our understanding of key questions in applied labor and public economics. Chapter 1 uses novel data from a sales company whose workers sell pest control services door to door to test for what is called in the behavioral economics literature "reference-dependent preferences." I show that sales workers select daily sales targets based on long-run goals to achieve bonuses paid by the firm at the end of the sales season. I then show that, contrary to standard theory of labor supply, workers substantially reduce their likelihood of continuing to work after achieving their daily sales target holding constant other factors of their work day. This behavior is consistent with loss aversion where workers put forth effort specifically to avoid underperforming relative to their expectations. The results support the theory that narrow goal setting and reference dependence together may act as a commitment device rather than representing a cognitive mistake as standard theory would suggest. These results have broad implications for how firms motivate their workers and show how long-run contract incentives can drive short-run labor supply choices. In Chapter 2, I exploit the 2014 rollout of provisions in the Affordable Care Act to identify the effects of direct subsidies for the purchase of private health insurance on adverse financial outcomes, consumer welfare, and outside parties. I use administrative tax data and credit bureau data to compare outcomes in areas that had high per-capita receipt of these premium tax credits to areas that had low per-capita receipt. To control for pre-treatment differences in trends attributable to the Great Recession, I use a propensity score reweighting and stratification procedure. I find that the premium tax credits substantially reduced the rates of severe mortgage delinquency, consumer bankruptcy, and severe auto delinquency as well as the right tail of the distribution of third-party collections and other debts. I also show that the value of the risk protections against medical debt amount to approximately 10-15% of the cash costs of the subsidies, while the subsidies provided substantial indirect benefits to mortgage lenders, creditors, and hospitals that amount to approximately two-thirds of the subsidy costs. Chapter 3, which is joint work with Michael Lovenheim and Alexander Willén, examines the dynamics of the decline in private-sector unionization rates in the United States over the past 40 years. We take a skill-based approach to studying this decline by accounting for changes in the types of skills covered by unions. We document that, from 1973 to 2017, private-sector unionized jobs shifted toward more non-routine, cognitive skills and fewer routine or manual skills and that women experienced a more pronounced change over this time period than men. After decomposing the changes in skills within the unionized sector to their components, we show that most of the change in unionized worker skills has been driven by the composition of occupations that are unionized rather than within-occupation skill changes. We show how these changes are compatible with a model of skill-biased technological change when we specifically account for the institutional framework surrounding collective bargaining and frictions to union certification or decertification. Finally, we show that accounting for different skills in the unionized sector leads to slightly larger estimates of the union wage premium than shown in the prior literature and that the wage premium remains relatively large for men and women at approximately 20% despite having fallen by over ten percentage points since its peak in the 1980s.

This dissertation addresses many important economic questions surrounding agency problems and vertical incentives. The first chapter documents the institutional background and pricing of U.S. reverse convertible bonds. The primary reason why I look into the reverse convertible bonds market is that it has interesting variation, enabling me to answer research questions in the second chapter and the third chapter. The second chapter shows some suggestive evidence of agency problems and vertical incentives in the reverse convertible bonds market mainly using a reduced-form analysis. Finally, the third chapter estimates the degree of agency problems and the degree of vertical incentives using a structural model. I conclude by analyzing various counterfactual scenarios. The first chapter introduces a financial product called a reverse convertible bond. It documents the product, market structure, and distribution process of the bond. Using a data set on reverse convertible bonds, I calculate the fair value of each bond by which I derive an issuer markup, i.e. how much an issuer earns by issuing each bond. This variable plays a critical role in investigating agency problems and vertical incentive issues in the second and third chapters. I also introduce my second data set on broker firms in this chapter. The second chapter documents some suggestive evidence of agency problems and vertical incentives in the reverse convertible bonds market. Combining two data sets described in the first chapter, I show descriptive evidence of agency problems and vertical incentives which are against the relevant regulations in the market. Specifically, broker behavior indicates agency problems while the relevant regulations such as suitability standard or fiduciary duty prevent them from doing so. In addition, I show that brokers are sensitive to their vertically integrated upstream firm's profit while arms-length transaction laws prohibit them from doing so. The final chapter estimates the degree of agency problems as well as the degree of vertical incentives in the reverse convertible bonds market. Although I show suggestive evidence of agency problems and vertical incentives in the second chapter, it is not straight forward to estimate their severity using a reduced-form analysis. Thus, I introduce a structural model that allows us to estimate the degree of agency problems and the degree of vertical incentives. The estimation results suggest that there are severe levels of both agency problems and vertical incentives in the market. Lastly, the counterfactual analysis shows that there is a substantial consumer surplus loss arising from agency problems. I also document a consumer welfare loss coming from vertical incentives but the magnitude is smaller than the consumer welfare loss from agency problems.

The promise of financial technology (“fintech”) and artificial intelligence (“AI”) in broadening access to financial products and services continues to capture the imagination of policymakers, Wall Street, and the public. This has been particularly true in the realm of fintech credit where platform companies increasingly provide online loans to consumers, students, and small businesses by harnessing AI underwriting and alternative data. In 2019 alone fintech lenders represented nearly 50% of total non-credit card, unsecured consumer loan balances in the United States. One of the most prevalent ways fintech credit firms operate is by securitizing the online loans they help originate. In doing so, fintech lenders are able to access the capital markets and further the spread of borrowed capital and credit risk. Against this backdrop of increasing institutional investment in fintech securitized assets, this Article reveals how consumer finance law is playing a subtle but increasingly important role in commercial financial transactions. I do this by exploring how structured finance has come to operate in the fintech credit marketplace and by comparing it to pre-2008 securitization activity in the home mortgage context. In doing so, I critique algorithm-driven credit securitization and point out certain economic and legal risks that make it similar to pre-2008 mortgage securitization, as well as other risks that are unique to fintech finance. These pertain to, among other things, the opacity of loans underwritten through the use of alternative data and machine learning, the untested efficacy of such underwriting techniques, and the ways consumer finance laws—such as licensing, usury, and standards-based regulation—and the growth of nonbank finance companies intersect with the securitization process. The paper concludes by offering policy recommendations for addressing these future risks.

Consumers in the United States are beset by a number of serious financial issues and concerns that are unlikely to disappear in the near future. When issues such as high levels of consumer debt, financial illiteracy, and suboptimal financial management skills are coupled with the vulnerability to economic fluctuations and persistently low savings levels observed in the U.S., individuals may be rendered less capable of weathering even modest income and expense shocks and may face substantial financial distress over the course of their lives. Consumer credit counseling agencies offer a means of addressing the harms and risks caused by these economic realities.Given the reach of counseling initiatives, which serve millions of people a year, there is a distinct need for rigorous assessments of credit counseling's potential impact on client outcomes. Understanding the impact of these services helps validate credit counseling's potential as a lever for policymakers to employ in addressing the consumer financial issues associated with both economic shocks and longer-term trends in consumer behaviors. This dissertation focuses on two interventions housed within consumer credit counseling agencies: A credit counseling program and a financial coaching program. After outlining a conceptual framework through which the potential impacts of credit counseling may be understood, this dissertation empirically explores a nationwide credit counseling initiative and outlines a profile of counseling clients. To measure the impacts of this credit counseling program relative to the counterfactual, this dissertation estimates a series of differences-in-differences models to track outcomes for over 6,000 counseling clients relative to a matched non-counseled comparison group. The key finding of this analysis is that counseling clients reduce both revolving debt and total debt relative to the comparison group and these reductions hold when controlling for debt write-offs and debt management plan participation. This dissertation then narrows its focus to examine the impact that frontline implementation dynamics have on the outcome of an experimental financial coaching program. The analysis first demonstrates that both program outputs and target group outcomes vary by the frontline worker. It explores the sources of this variation and finds that these differences are associated with differing approaches to client engagement and different levels of worker engagement with the program itself. Specifically, the analysis finds that workers who are capable of successfully engaging clients are significantly more capable of driving program enrollment, while workers who are more engaged with the program's change processes are significantly more capable of driving improved client outcomes.These analyses show not only that credit counseling can drive improvements in client financial outcomes but that understanding how programs are delivered also matters in assessing outcomes. In doing so, this dissertation contributes to the broad general literature on consumer financial policies and programs and to the limited literatures on credit counseling and financial coaching, as well as both the theoretically- and empirically-oriented literatures on social service implementation in the fields of public policy and management. Further, it contributes to the emerging research on the use of behavioral economics to understand and address issues in public policy.

This dissertation examines the financial position of payday borrowers using the most detailed nationally representative survey of American finances - the Survey of Consumer Finances. Payday loans are short term, unsecured loans generally less than five hundred dollars. Borrowers pay dearly for such loans, as fees are typically $15 for every $100 borrowed. Because the term for these loans is never more than two weeks, the annualized percentage rate is quite large ~ 390%. For the first time, in 2007 the Survey of Consumer Finances included a question about use of payday loans, and so it is now possible to expand the analysis of payday borrowing beyond basic demographic information and income. Given these high costs, why would households choose this option for dealing with financial emergencies? I show that payday borrowing is driven more by a lack of resources, in particular available credit alternatives, rather than low incomes. In addition, budgeting is a significant factor. Another alternative that households often look to is financial support from friends and family. Access to financial support is driven by many of the same factors that make these households insecure. Nevertheless, I show "private safety nets" do indeed reduce payday borrowing. But network relations don't just entail receiving. The provision of financial support is correlated with overspending and increases the odds of payday borrowing. This dissertation makes several contributions to our understanding of payday borrowing. First, using the detailed data on savings and budgeting practices, it shows the importance of building resources to managing budgetary shortfalls. Second and relatedly, I show that borrowers have little financial alternatives - they are effectively shut out of credit markets, and at the same time their network relationships can be both a help and a hindrance for avoiding payday loans.

In this dissertation I present three theoretical papers, each as an individual chapter. The first paper is titled "The Economics of Discretion in Multi-Agent Decision Problems." It is premised on the idea that discretion is valuable when knowledge mismatches lead principals to delegate decisions to agents with specialized knowledge. In the paper, I consider a team setting, and I characterize the optimal delegated choice set that is offered to agents when an agency conflict is present. I show that the amount of discretion increases with the value of an agent's private information, with the degree of alignment between the principal and agents, and with the ex ante uncertainty faced by the principal and agents. The latter finding implies that discretion may be used by agents to hedge the risk that they face. Nevertheless, I demonstrate that when all participants have rational expectations, it is never optimal for agents to add strategic uncertainty to receive more discretion ex post. I conclude the paper by applying the theory to delegated portfolio management, which yields novel empirical implications in this setting.The second paper is co-authored with Bruce I. Carlin and Andrew Iannaccone and it is titled "Competition, Comparative Performance, and Market Transparency." In the paper, we study how competition affects market transparency, taking into account that comparative performance is assessed via tournaments and contests. Extending Dye (1985) to a multi-firm setting in which top performers are rewarded, we show that increased competition usually makes disclosure less likely, which lowers market transparency and may decrease per capita welfare. This result appears to be robust to several model variations and as such, has implications for market regulation.The final paper is co-authored with Bruce. I. Carlin and is titled "Political Influence and the Regulation of Consumer Financial Products." In the paper, we explore a theoretical model of product regulation in which the social planner chooses an optimal level of market complexity, given that people have varied sophistication. We investigate how several dimensions affect the quality of regulation: the skill of the social planner, imperfect information, lobbying efforts, voting behavior in elections, and political philosophy. We find that both sophisticated and unsophisticated market participants often vote to elect the least informed and educated planners, which erodes social welfare. Further, when concerns regarding equality are sufficiently large (i.e., a socialistic agenda), the social planner limits the market to one product. In such case, adequacy suffers and all market participants are equally worse off.

The widespread use of computer-based risk models in the financial industry during the last two decades enabled the marketing of more complex financial products to consumers, the growth of securitization and derivatives, and the development of sophisticated risk-management strategies by financial institutions. Over this same period, regulators increasingly delegated or outsourced vast responsibility for regulating risk in both consumer finance and financial markets to these privately owned industry models. Proprietary risk models of financial institutions thus came to serve as a "new financial code" that regulated transfers of risk among consumers, financial institutions, and investors. The spectacular failure of financial-industry risk models in the current worldwide financial crisis underscores the dangers of regulatory outsourcing to the new financial code. This Article explains how financial institutions used the "new financial code" to shift, spread, and price financial risk using the template of the stages of securitization of consumer-credit products, hedging through credit default swaps, and overall portfolio management. This Article then examines several explanations for the failures of risk models, which contributed to the current crisis, including flaws in the design of risk models and agency costs associated with those models. It also outlines several lessons for regulatory outsourcing from the current crisis, including the following: Bank regulators should scrap those provisions of Basel II that allow certain banks to set their own capital requirements according to their internal risk models; Regulators should promote "open source" in code (or the models) used to market financial products to consumers, price securitizations and derivatives, and manage financial-institution risk; and The failure of risk models used to price securitizations and derivatives reveals some of the comparative advantages of equity securities in spreading risk.

My dissertation focuses on two key aspects of the large and growing market for credit cards in the United States: (1) the role of private information in interest rate heterogeneity and (2) the changing search behavior of consumers that may have contributed to falling interest rates in this market in the 1990s. The first question examines whether asymmetric information between banks about consumer credit-worthiness contribute to interest rate heterogeneity in the credit card market. The information set available to a bank with which the consumer has other financial dealings, i.e. a “home bank”, is greater than information available to external banks, which only have access to credit bureau reports. When both types of banks offer cards to consumers, the information asymmetry between sellers arising from the private information of home banks results in a equilibrium where the insiders select the “better” customers by offering them lower rates within each risk cluster that is differentiated on the basis of public information. An endogenous switching regression model incorporating the self-selection by agents is used on consumer-level survey data to test the propositions of the above model. The FIML results show that using private information, home banks are able to select a client pool whose default risk is lower and offer them lower-than-average interest rates. Thus private information arising out of relationship banking in consumer lending contributes in assessing default heterogeneity over and above publicly available credit bureau information and subsequently leads to more accurate pricing of credit card loans. Secondly, my work also investigates whether search costs inhibit consumer search for lower credit card interest rates. The results provide evidence contrary to the existing hypothesis that a high liklihood of rejection of high balance consumers adversely impacts their search behavior and consequently lead to interest rate stickiness. Using the SCF data, I model “consumers’ propensity to search” and their “probability of being denied credit” simultaneously and find outstanding credit card balances induce cardholders to search more even though they face a higher probability of rejection. This result demonstrates a policy impact of the Truth-in-Lending Act, which, together with a high volume of direct solicitation, has lowered the cost of search to consumers in the 1990s.