When a credit card company changes interest rates or increases the credit limit of its consumers, how do these consumers respond? The answers are far from predictable, as David B. Gross of the University of Chicago Graduate School of Business and Wharton’s
When a credit card company changes interest rates or increases the credit limit of its consumers, how do these consumers respond? The answers are far from predictable, as David B. Gross of the University of Chicago Graduate School of Business and Wharton’sNicholas S. Souleles have found out.
In a paper titled “Consumer Response to Changes in Credit Supply: Evidence from Credit Card Data,” Gross and Souleles examine the effects of changes in the credit limits and interest rates, and they consider the ability of different models of consumption and saving to rationalize these effects. Their study found that people starting at near their credit limit respond most sharply to changes in credit limits, but even those who are well below their credit limits respond significantly.
The researchers say what is particularly clear is that increased liquidity triggers immediate and large jumps in spending and debt. The average individual’s debt trajectory is fascinating. On average, debt rises by about $40 in the month in which a credit line increases. In two months after a credit line increase, the average debt rises by more than $180. Within a year, the debt rises by more than $350. In fact, each extra $1,000 of liquidity translates into a $130 increase in an individual’s debt.
Unlike most other studies, Gross and Souleles find strong effects from changes in account-specific interest rates. Debt is particularly sensitive to large declines in interest rates, which might explain why lenders use teaser rates so widely. In addition, while individuals often respond to lower interest rates by transferring their balances to low interest accounts, this trait accounts for less than half the response .
This study covers new ground in several areas. For example, why do credit card borrowers simultaneously hold other low yielding assets such as housing equity? Sample this: More than 90% of people with credit card debt have some very liquid assets in checking and saving accounts. And although the authors concede that some transactions cannot easily be made using credit cards, they were surprised to find that a third of borrowers have more than a month’s worth of income in liquid assets, which they say is more than the amount required for cash transactions.
Another surprise:many people seem to set “target” credit-card utilization rates. For instance, if a consumer is originally using 60% of his card’s credit limit and then his limit is increased, which lowers his utilization rate, he might increase his spending on that card to raise his utilization rate back to 60%. Such targeting could reflect rule-of-thumb behavior, but the authors show it is consistent with conventional economic models of fully rational consumers.
The researchers based their conclusions on a unique new data set of several hundred thousand individual credit card accounts, which they tracked each month for 24-36 months. They obtained the data from several different credit card issuers, backed by associated credit bureau data. To protect the identity of the accounts and the issuers, Gross and Souleles report only summary statistics averaged across issuers.
The study by Gross and Souleles covers fresh ground. So far, empirical studies of consumer behavior have found it difficult to track the effects of changes in credit supply and monetary policy. Most such studies have focused almost exclusively on the role of firms, and not consumers. Also, there has been surprisingly little analysis so far of consumer debt, as opposed to consumer assets and corporate debt.
Few will disagree that credit card data is among the best places to probe the effects of credit supply. About a fifth of total personal consumption involves the use of credit cards. With the growth of e-commerce, this figure will increase further. Also, for most households, credit cards represent the main source of unsecured credit. Households borrow about $500 billion on credit cards (Federal Reserve Board, 1999). About three-fourths of households have at least one card, and of these more than half are rolling over debt. The median revolving account borrows more than $2,000, with another $5,000 of balances on other cards, the authors calculate. Clearly, credit card data also allow for powerful, high-frequency event-studies of the effects of changing various aspects of credit supply.
The study does, however, have some limitations. First, there is little information about some potentially important variables such as household assets or employment status. But credit card companies too lack this information, so its absence does not affect the study. Second, the main unit in this analysis is credit card account data, and not the individual borrower. This has been circumvented by using data from the credit bureaus, which cover all sources of credit used by a cardholder, especially other credit card balances.
Gross and Souleles plan to dig deeper into consumer credit trends. When issuers raise credit lines or lower interest rates, they are trading off larger interest payments against larger probabilities of default. In a companion paper using the same data as in this paper, the authors have estimated hazard models of default. They now plan to merge these two papers to identify the profit-maximising credit-supply function. Another area they have identified for research is how credit card issuers and other consumer lenders change credit supply in response to monetary policy and other business cycle shocks. They also want to incorporate the results of this paper into structural models of credit supply and demand.
This study has important implications for both consumer theory and business cycle analysis. As lenders respond to monetary policy or other shocks by changing credit supply, this study shows how individuals change their spending and borrowing in response to changes in credit limits and interest rates. Lenders trying to determine optimal credit policy will also find useful pointers in the Gross-Souleles analysis.