Academic Research

Abstract: Nearly 80 percent of U.S. adults have a credit card, and more than half of them revolve their debt from month to month. Using a large sample of credit bureau data, this paper documents a tight link between available credit (the limit) and credit card debt, and then it offers a model-based interpretation of this linkage. Credit limits change frequently for individuals, increase rapidly on average as people age, and show large changes over the business cycle. Yet credit card debt changes nearly proportionately to credit and at about the same time, so the fraction of credit used is relatively stable over time. The authors build a life-cycle consumption model that includes the joint use of credit cards to pay directly for expenditures, to help smooth consumption against income shocks, and to borrow longer term (revolving indefinitely). The authors estimate the parameters of the model using several data sources, including a large credit bureau database and a new daily diary of consumer payment choices.

Press:

The New Republic: I Worked at Capital One for Five Years. This Is How We Justified Piling Debt on Poor Customers

Bloomberg: Americans Can’t Help Themselves From Borrowing More on Credit Cards

CreditCards.com: Infographic: The life cycle of a credit card

Abstract: Nearly 80 percent of U.S. adults have a credit card, and more than half of them revolve their debt from month to month. Using a large sample of credit bureau data, this paper documents a tight link between available credit (the limit) and credit card debt, and then it offers a model-based interpretation of this linkage. Credit limits change frequently for individuals, increase rapidly on average as people age, and show large changes over the business cycle. Yet credit card debt changes nearly proportionately to credit and at about the same time, so the fraction of credit used is relatively stable over time. The authors build a life-cycle consumption model that includes the joint use of credit cards to pay directly for expenditures, to help smooth consumption against income shocks, and to borrow longer term (revolving indefinitely). The authors estimate the parameters of the model using several data sources, including a large credit bureau database and a new daily diary of consumer payment choices.

Abstract: Using court orders called civil judgments, creditors can garnish wages and seize assets for unpaid consumer debts. We provide the first comprehensive description of civil judgments across states and time. Civil judgments are about twice as common as bankruptcy filings. Civil judgments are 20 times more common in some states than others. Their incidence is the highest around age 40, decreases with credit score, and is concentrated in census tracts with a greater proportion of Black residents. To understand the cross-state variation, we build the most comprehensive panel measuring wage garnishment across states. We show that a decrease in the amount garnishable per week decreases the number of judgments per capita and increases the average civil judgment amount. These findings are consistent with a simple theoretical model where creditors optimally choose which debts to bring to civil court. Yet a decrease in the amount garnishable reduces credit card limits and open credit card accounts. The results suggest that policymakers may need to weigh access to credit when considering protecting debtors.

Non-technical summary

Abstract: This report collects information from the Making Ends Meet survey and Consumer Credit Panel (CCP) to better understand the pandemic’s effects on households. As far as possible, it only includes new research, although some results are slight variants on already published research. The report first describes the pandemic ups and downs of some Americans. It then presents results using the Making Ends Meet survey and the CCP to describe consumer finance before the pandemic and during the pandemic. This work was conducted while I was employed at the Consumer Financial Protection Bureau.

Abstract: Two thirds of all Indian women have migrated for marriage, around 300 million women, but not much is known about this vast  migration. This paper provides a detailed accounting of this large migration and evaluates some of its potential causes.  Marriage migration varies substantially across India, and appears to have changed little over the previous 40 years. Contrary to conventional wisdom, marriage migration does not contribute to risk sharing or consumption smoothing. Nor is it driven by sex ratio imbalances. Instead, this paper introduces a simple model in which parents must search for a spouse for their daughter geographically. The model helps rationalize the correlations between migration, age of marriage, and literacy across districts in India. It suggests that marriage migration is part of the larger puzzle of low workforce participation, education, and bargaining power of women in India, rather than an independent phenomenon.

Abstract: This paper examines the changing distribution of where women and girls live in India at the smallest possible scale: India's nearly 600,000 villages. Village India is becoming more homogeneous in its preferences for boys even as those preferences becomes more pronounced. A consequence is that more than two thirds of girls now grow up in villages where they are the minority. Most Indian women move on marriage, so parents' sex selection decisions are felt well beyond the village. Yet marriage migration does not have an equalizing influence on sexual imbalances across villages. Linking all villages across two censuses, I show that changes in village infrastructure such as roads, power supplies, or health clinics are not related to changes in child sex. Geographically close villages reinforce each other's preferences. The results suggest that there are no easy policy solutions for addressing the increasing masculinization of Indian society.

Abstract: Credit card payments and revolving debt are important for consumer theory but a key data source—credit bureau records—does not distinguish between current charges and revolving debt. We develop a theory-based econometric methodology using a hidden Markov model to estimate the likelihood a consumer is revolving debt each quarter. We validate our approach using a new survey linked to credit bureau data. We estimate that for likely revolvers: (i) 100% of an increase in credit becomes an increase in debt eventually; (ii) credit limit changes are half as salient as debt changes; and (iii) revolving status is persistent.

Ungated working paper version 

Abstract: Buying a house changes a household’s balance sheet by simultaneously reducing liquidity and introducing mortgage payments, which may leave the household more exposed to other shocks. We examine how this change impacts consumer credit beyond the mortgage. Using a large panel, we show that on acquiring a mortgage, credit card debt increases by about $1500 in the short term, severe delinquencies increase by 2.2 percentage points, and credit card utilization—the fraction of a consumer’s credit card limit that is used—increases by 11 percentage points. In the long term, credit card balances increase by $3900 and delinquencies by 9.1 percentage points. In our sample period before the 2008 financial crisis, credit limits increased faster than debt in the long run, pushing down long-term utilization. After the financial crisis, debt increased faster than credit limits in the long run, and credit card utilization rates rose upon the acquisition of a new mortgage, consistent with larger down payments leaving households more constrained.

Ungated working paper version 

Abstract: What impact on local development do immigrants and their descendants have in the short and long term? The answer depends on the attributes they bring with them, what they pass on to their children, and how they interact with other groups. We develop the first measures of the country-of-ancestry composition and of GDP per worker for US counties from 1850 to 2010. We show that changes in ancestry composition are associated with changes in local economic development. We use the long panel and several instrumental variables strategies in an effort to assess different ancestry groups’ effect on county GDP per worker. Groups from countries with higher economic development, with cultural traits that favor cooperation, and with a long history of a centralized state have a greater positive impact on county GDP per worker. Ancestry diversity is positively related to county GDP per worker, while diversity in origin-country economic development or culture is negatively related.

Supplemental Appendix

Replication Files

Abstract: We examine a period during the prevalence of the gold standard in the United States to provide evidence that speculation about a currency peg can have damaging effects on bank balance sheets. In particular, the defeat of the pro-silver candidate in the 1896 presidential election was associated with a large and permanent increase in bank leverage, with the initial impact most pronounced among states where banks held more specie in proportion to their assets and were, therefore, also more committed to paying out deposits in specie. Based on the cross-sectional pattern of changes in leverage observed in 1896, we construct a measure of the credibility of the gold standard spanning the entire sample period. Changes in this measure correlate with changes in aggregate bank leverage, suggesting that uncertainty about the monetary standard played an important role in the 1893 banking panic and its aftermath.

Abstract: This paper examines how the precautionary motive varies with income. I first develop a theoretical benchmark of how we would expect precaution to vary with income starting from a basic version of the buffer-stock model. Emergency savings provide a way for households to smooth over shocks and so give insight into the precautionary motive. Using data from the Survey of Consumer Finances in the United States, I show that as income declines, the desired emergency savings relative to income increases, suggesting that low-income households are more precautionary. Observable differences, such as income uncertainty, do not explain the rise. Instead, I propose and estimate a model with a minimum subsistence level and unexpected expenses. The model implies that low-income households are increasingly exposed to shocks, explaining the increase in precaution. Supporting the approach, I show that expenses on repairs are a larger fraction of the spending of low-income households.

Ungated version and technical appendix. Also circulated as: The Precaution of the Rich and Poor

Abstract: This research note compares sequential and concurrent web-mail mixed-mode approaches and incentives for a survey on US consumers’ use of financial products, especially when their finances are tight. The sample (n = 2,000) was drawn from credit bureau data. We examine the effects on response rates, survey costs, and possible nonresponse bias in an experiment varying two factors in a 2x2 design: (1) using concurrent or sequential web and then mail survey modes, and (2) different incentive amounts given to initial survey nonrespondents ($5 versus $10). The sequential (web-first) design had a significantly lower response rate (3.9 percent) at week five—before the paper questionnaire was mailed—than the concurrent group (11.0 percent). This difference was nearly fully eliminated by the end of the field period. The higher incentive brought in slightly more respondents in the concurrent arm and slightly fewer respondents in the sequential arm, but neither difference is statistically significant. Compared with the sample frame, respondents in both groups were generally older and had many of the characteristics that come from being older: higher credit scores, more open credit cards, lower credit card utilization, greater likelihood of having a mortgage, and lower likelihood of being delinquent on credit card payments. Given the lower initial response rate and the need for more follow-up mailing, the sequential mixed-mode approach resulted in a higher cost per complete survey.

Abstract: What financial services matter for growth? This paper examines the effects national banks had on growth in the United States from 1870-1900. These banks were commercial not investment banks: they made short term loans and could not take land as collateral. I use the discontinuity in entry caused by a large minimum capital requirement to identify the effects of banking. Counties getting a bank increased production per person substantially and tilted production towards agriculture over manufacturing by expanding land under cultivation, not improving yields. The effects are highly persistent and show that the commercial activities of banks matter for growth.

Supplemental Appendix

Replication Files

This article is also the basis of a Cato Research Brief (Cato Institute. Research Briefs in Economic Policy. No 44. February 2016.)

Abstract: This paper examines how the precautionary motive varies with income. I first develop a theoretical benchmark of how we would expect precaution to vary with income starting from a basic version of the buffer-stock model. Emergency savings provide a way for households to smooth over shocks and so give insight into the precautionary motive. Using data from the Survey of Consumer Finances in the United States, I show that as income declines, the desired emergency savings relative to income increases, suggesting that low-income households are more precautionary. Observable differences, such as income uncertainty, do not explain the rise. Instead, I propose and estimate a model with a minimum subsistence level and unexpected expenses. The model implies that low-income households are increasingly exposed to shocks, explaining the increase in precaution. Supporting the approach, I show that expenses on repairs are a larger fraction of the spending of low-income households.

Ungated version and technical appendix. Also circulated as: The Precaution of the Rich and Poor

Abstract: Using a large panel this paper first demonstrates that individuals gain and lose access to credit frequently. The estimated credit limit volatility is larger than most estimates of income volatility and varies over the business cycle. Within a model, variable credit limits create a reason for households to hold both high interest debts and low interest savings at the same time. Using the estimated credit volatility, the model explains why around one third of American households engage in this credit card puzzle. The approach also offers an important new channel through which financial system uncertainty can affect household decisions.

Ungated versions: Federal Reserve Bank of Boston WP 14-8, BC working paper 754

Program Files

Press:

MarketWatch: This is why people carry credit card balances

CreditCards.com: How cutting credit limits hurts cardholders, the economy

Abstract: In India both men and women with more education live in households with greater consumption per capita. Yet aggregating across age cohorts and states, an extra year of education brings male cohorts only 4% more consumption and provides no additional consumption for female cohorts. This result is robust to: (1) accounting for survey measurement error, (2) different measures of household consumption and composition, (3) allowing returns to differ by state and school quality, and (4) age misreporting. The only area with substantial returns is entering into regular wage work which still employs only a small fraction of the population.

Replication files

Also as BC working paper 819

Abstract: Although many view financial access as a means of reducing poverty or increasing growth, empirical studies have produced contradictory results. One problem is that most studies cover only a short time frame and do not consider dynamic effects. I show that introducing credit in a general model of intertemporal consumption creates a boom in consumption and reduces poverty initially, but eventually reduces mean consumption because credit substitutes for precautionary wealth. Using new consistent consumption data that cover a much longer time period than most studies, my empirical findings show that increased access to bank branches in rural India increased consumption initially and reduced poverty, but consumption later fell and poverty rose. The long-term effect is still positive, however, suggesting that credit may have a beneficial role beyond consumption smoothing.

Data Files

Also as BC working paper 741