Hidden fees can distort consumer decision-making. In response, regulators historically have (a) improved disclosure to make fees more salient or (b) standardized products to restrict what fees can be charged. We use Chilean administrative data and a multi-stage natural experiment to separately identify the effects of disclosure and standardization on repayment. We find that disclosure reduces delinquencies by 13.7 percentage points (40%) and default by 1.68 percentage points (98%), whereas standardization has no effect. We find no effect on initial loan terms, suggesting that disclosure’s effects are specific to repayment behavior: specifically, disclosure improves borrowers’ understanding of their credit obligations.

Bankruptcy can provide a “fresh start” to consumers whose crushing debt leaves them unable to flourish economically or psychologically. Research shows the benefits of giving borrowers a clean slate: bankruptcy filing increases the probability of homeownership by 13.2 percentage points, increases annual earnings by $5,562, decreases five-year mortality by 1.2 percentage points, decreases five-year foreclosure rates by 19.1 percentage points, and increases the probability of a debtor creating a new business. Debt reduction from interest payments (i.e., debt restructuring) decreased the probability of filing for bankruptcy by 33% and increased the probability of employment by 5.1% among other positive outcomes. Despite these benefits, evidence suggests that fewer debtors file for bankruptcy than would benefit from doing. Our paper asks why so few debtors file for bankruptcy. Past research proposes that the bankruptcy system is underused because it is expensive: insolvent Canadian debtors must pay between $1,800 to $6,500 in fees on top of a portion of their debts. Although we do not doubt that cost plays a role, our paper explores a second channel: stigma. Our paper explores this second channel in three parts. First, we develop a theoretical model that explains why stigma presents a barrier to bankruptcy filing. Specifically, we distinguish two forms of bankruptcy stigma––other-directed and self-directed––that plausibly limit filing in different ways. Second, we provide the first empirical evidence about the levels of self-stigma amongst debtors who are considering filing for bankruptcy. To do so, we develop a novel scale to measure bankruptcy self-stigma, adapted from research on self-stigma about depression. We then administered this scale to a population of 267 debtors who were actively considering bankruptcy, insofar as they responded to an advertisement for insolvency services. We find that bankruptcy carries stigma that is significantly higher than depression, which is itself highly stigmatized. Furthermore, debtors who are considering filing for bankruptcy display higher ratings of subclinical anxiety, depression, and stress than prisoners. Financial self-stigma is a large positive predictor of depression, anxiety, and stress within our participants: those with the lowest levels of stigma displayed normal ratings on each scale, whereas those with the highest stigma ranged from moderate to extremely severe.  Finally, we consider the implications of stigma for policies that aim to increase participation in the bankruptcy system. We consider how interventions might directly reduce stigma by correcting misinformation about the moral character of insolvent debtors, the causes of bankruptcy, and the bankruptcy process. We also consider the implications of stigma for policies that seek to either increase or remove the human interactions within the bankruptcy system, including counselling and pro se bankruptcy.

Sociologists have theorized U.S. universities as a heterogenous organizational ecology. We use this lens to compare student debt and college prices for low-income students across public universities according to their research intensiveness and varied state grant aid policies. We show that students at research-intensive public universities have had an easier time repaying student loans than at other schools. By linking multiple data sets, we also provide the first comprehensive assessment for all 50 states of state-level need-based grant aid programs, which might alleviate loan repayment challenges. We find large disparities. California, Washington, Wyoming, and New Jersey spent more than $4,000 on aid per low-income student in 2015, more than the federal expenditure on Pell Grants for their state. Most states spend little in comparison. Contra the Bennett hypothesis, we also find that state need-based aid is strongly associated with both lower net prices and lower student loan nonrepayment rates.

How do biased beliefs about the interest rate distribution affect search, negotiation, and loan terms in consumer credit markets? In collaboration with Chile's financial regulator, we conducted a randomized controlled trial with 112,063 loan seekers. Randomly eliciting beliefs about interest rates led participants to search more and obtain lower rates. Most participants underestimated both interest rate levels and dispersion. Showing them a price comparison tool we built using administrative data on the universe of consumer loans caused them to update their beliefs, negotiate more, obtain lower rates without searching more, and be more likely to take out a loan.

We estimate the causal effects of a financial windfall on consumption, savings, and wage earnings using linked administrative data on a large sample of Canadian lottery winners. By combining income tax records with credit bureau data, we estimate how lottery winnings affect consumption measured through credit card spending, savings through financial asset accumulation and debt repayment, and wage earnings from tax records. We then examine how these responses vary across the income distribution. We find that high-income individuals allocate a larger share of lottery winnings to savings and leisure, while low-income individuals spend a larger share on consumption. We illustrate the policy relevance of these heterogeneous responses through three applications: designing fiscal stimulus programs, evaluating a Universal Basic Income (UBI), and determining optimal top tax rates on wage earnings and savings.

We examine the impact of a hyper-local payday loan supply shock on debtor uses of formal credit, by matching debtor-level credit bureau data with location of individual payday lender entry and exit in a difference-in-differences setting. We find that payday lender entry into a neighborhood worsens the financial stress of borrowers who do not have the ability to borrow against housing, while only increasing credit card balances for those who do. However, we find that payday borrowing helps borrowers who are credit constrained, as their credit card balances under stress increase and credit scores drop significantly when a payday lender exits their neighbourhood. We also exploit exogenous variation in provincial regulation of payday lenders and find that 7 day cool-down periods between payday loans increase credit card stress, which persists for two years after payday lender entry. Regulations that restrict borrowers to one loan per lender increase their credit card balances under stress and damage their credit scores, though these effects are relatively short-lived. These results suggest that there are important heterogeneities in how payday borrowing interacts with formal credit products.

We exploit a unique natural experiment to structurally estimate the information frictions associated with switching costs. Specically, we study a Chilean policy that simplied and standardized the presentation of loan characteristics in contracts and quotes. Using administrative data from the banking regulator, we exploit how this policy change affected the price-sensitivity in consumer decisions to identify the reduction in information frictions. We then incorporate this estimate into a dynamic structural model to explore the link between reduced informational frictions and welfare in long-term market equilibrium. We find that after the policy information frictions fell around 10 percent, which translated into an interest reduction of 180 basis points. We estimate a welfare improvement for consumers of 15 percent in the long run.

By relieving individuals from crippling debt, bankruptcy has been shown to increase home ownership, annual earnings, employment, and entrepreneurship. In addition to improving filers’ future financial outcomes, bankruptcy may also improve psychological well-being by relieving financial stress. However, despite the benefits that insolvency offers, many debtors lack access to bankruptcy due to costs, technological barriers, and stigma. We conduct, to our knowledge, the first randomized controlled trial (RCT) to examine whether lowering the financial barriers to insolvency improves access to this important financial institution. In partnership with a licensed insolvency trustee (LIT) in Canada, we randomly subsidize potential filers with $1,000 toward bankruptcy or consumer proposal filing fees. A separate treatment arm will offer debtors a surprise subsidy after they have decided to file so that we can separately identify income effects from the economic effect of lowering the financial barrier to access bankruptcy. We also use moment-to-moment experience sampling and sub-clinical questionnaires to track participants’ stream of consciousness and mental health throughout the filing process to determine whether debt relief improves well-being by relieving financial stress.