Both behavioral and traditional approaches to revolving credit card debt suggest that individuals who carry a balance on their credit cards should be more likely, all else being equal, to substitute away from credit cards and into alternative payment methods for purchases. We would expect revolving balances to be associated with a lower fraction 10 of credit card payments and a higher fraction of debit, check, and cash payments. ( credit card)
In Table 4 we present ordinary least squares regressions of the following form, with robust standard errors: F ractionij = β0 + β1RevolvingBalancesi + γxi + i . (2) F ractionij is the fraction of payments made by consumer i, using payment j. RevolvingBalancesi is a binary variable taking the value 1 if an individual regularly carries a credit card balance and 0 otherwise. xi is a vector of socio-demographic and other characteristics of consumer i, which varies with specification but always includes categorical variables for gender, age group, income group, race, education, and a continuous variable for the length of time in years that an individual has held his or her current checking account. ( credit card)
The socio-demographic variables are defined as shown in Table 1. Extended results of these regressions are displayed in Table 5. In Columns 1 and 2 of Table 4 we present regressions with FractionCredit as the dependent variable. In the initial specification of Column 1, we find, controlling for socio-demographics, that having revolving credit card balances is associated with a significant reduction in the fraction of payments for which credit cards are used. ( credit card)
In Column 2, we additionally control for participation in credit and debit card rewards programs.6 Some of the demographic characteristics and the presence or absence of rewards program participation are important determinants of the fraction of payments for which credit cards are used. Across specifications, we find that revolving balances are associated with a reduction of between 2 and 4 percent in the fraction of payments for which credit cards are used. In Columns 3 and 4 of Table 4 we present regressions with FractionDebit as the dependent variable. Socio-demographic characteristics, along with rewards program participation, have strong effects on the fraction of payments for which debit is used. ( credit card)
We find evidence for substitution into debit cards by individuals with revolving balances. Controlling for socio-demographic characteristics and rewards program participation, we find that revolving balances are associated with an increase of 4 to 5 percent in the fraction of payments made by debit card. ( credit card)
It seems that, for individuals with revolving balances, the reduction in credit card 6Although rewards program participation may be endogenous, including it in the regression does not change our results. 11 payments is almost entirely offset by additional debit card payments. ( credit card)
In Columns 5 and 6 of Table 4 we present regressions with FractionCash as the dependent variable. We find no significant increases associated with revolving balances in the fraction of payments for which cash is used. If anything, the implication is that revolvers use cash for a smaller fraction of payments than do convenience users. In Columns 7 and 8 of Table 4, we present regressions with FractionCheck as the dependent variable. We also find no impact of revolving balances on the fraction of payments for which checks are used. ( credit card)
Market size and foundational metrics of consumer use This section reviews a number of market metrics in order to provide a foundation for addressing other topics in the market covered in more detail in subsequent sections—such as the cost of credit cards (which we cover in Section 3) or their availability to consumers in different credit tiers (which is in Section 4). In this section, we review three main aspects of the consumer credit card market. First, we describe the overall size of the market using a number of different measures. By some metrics, such as total credit card debt outstanding, the market has generally grown back to or even surpassed its pre-recession size.1
Second, we look at a number of basic metrics about consumer usage, including cardholding patterns, consumer- and account-level balance and payment behavior, and rewards penetration. Some of these metrics point to potentially significant qualitative differences between the credit card debt held by consumers prior to the Great Recession and the debt held by consumers today. Last, we report on delinquency and charge-off rates. These present a mixed picture. They remain below historic norms, but they are worsening, even as widely relied-upon indicators underlying macroeconomic conditions–like the unemployment rate–do not appear to be deteriorating.
There are two important qualifications to bear in mind when reviewing the findings set forth below. First, when we refer to consumers or accounts in a particular credit tier, we are using scores recorded at a particular point in time. For any credit tier and any point in time, the accounts grouped into that tier will include accounts that migrated into that tier subsequent to origination, as well as accounts that were originated into that tier but that have not (yet) 1 This has caused some parties to express concern that the consumer credit card market is “overheated.” See, e.g., Akin Oyedele, Americans Hold More Credit-card Debt than Ever, and a ‘Major Tipping Point’ Isn’t Far Off, Business Insider (Aug. 8, 2017), available at http://www.businessinsider.com/credit-card-debt-record-2017-8. Others disagree. See, e.g., AFSA Comment Letter, at 1-3. 31
CONSUMER FINANCIAL PROTECTION BUREAU — CONSUMER CREDIT CARD MARKET REPORT
migrated. Second, references to debt encompass both revolving balances and also purchases made in the preceding month or cycle that may be paid in full by the end of the next month or cycle. 2.1 Total market size 2.1.1 Credit card debt The consumer credit card market is one of the nation’s largest consumer credit markets.
The Federal Reserve Board’s G.19 “Consumer Credit” data show that “revolving credit,” which is almost entirely credit card debt, represented over a quarter of all non-mortgage consumer debt outstanding in December of 2016.2 The Federal Reserve Bank of New York’s Quarterly Report on Household Debt and Credit shows that credit cards comprise the fourth-largest source of consumer indebtedness behind mortgages, student loans, and auto loans.3 Figure 1 uses data from the G.19 to show the long-term trend in consumers’ revolving credit balances over time.4
These peaked at just over $1 trillion in 2007 and 2008. During the recession, these balances fell to under $800 billion. They have since rebounded steadily and, according to the latest such data, currently approach $1 trillion once again. 2 The G.19 report uses the term “revolving credit” to refer to debt that can be borrowed up to a prearranged limit and repaid in one or more installments. See Federal Reserve Board, Consumer Credit Data G. 19, (Nov. 17, 2017) (“G.19 Report”), available at http://www.federalreserve.gov/releases/g19/current/default.htm. All credit card outstandings, therefore, are included—not simply the amount of credit card debt revolved by consumers across one or more billing cycles.
The G.19 data exclude loans secured by real estate when considering consumer credit. 3 Federal Reserve Bank of N.Y., Quarterly Report on Household Debt and Credit (Q3 2017), (Nov. 2017), available at https://www.newyorkfed.org/microeconomics/hhdc.html. 4 See generally G.19 Report. 32
CONSUMER FINANCIAL PROTECTION BUREAU — CONSUMER CREDIT CARD MARKET REPORT FIGURE 1: OUTSTANDING REVOLVING CONSUMER CREDIT AND TOTAL CONSUMER CREDIT CARD CYCLE-ENDING BALANCES, ANNUAL AVERAGE
(CCP, FRB G.19)5 These totals include non-credit card revolving debt, such as overdraft lines of credit. Nevertheless, credit card debt comprises the overwhelming share of G.19 revolving consumer debt. To illustrate this, Figure 1 also includes a separate schedule of outstanding credit card balances drawn from the Bureau’s Consumer Credit Panel (“CCP”). Using the relationship between the CCP and G.19 data during the period for which both are available (i.e., from 2005 to the present), we can offer an approximate depiction of total consumer credit card balances from 1990 to the present.
By focusing on CCP data, moreover, we can see that consumer credit card debt now exceeds its pre-recession peak. Figure 2 shows the total of all balances owed by consumers on general purpose and private label credit cards as of the end of each quarter’s last cycle since 2005. In our last report we noted that general purpose credit card balances had more-or-less steadily increased—albeit with seasonal variation—since the end of 2010, but that they had not yet returned to pre-recession levels. That was still true as of mid-2017. In the fourth quarter of 2016, total outstanding general purpose credit card debt reached $716 billion, which remains below 5 This chart displays average cycle-ending balances calculated across each full year, which decreases the effect of seasonality.