Bipartisan legislation to cap credit card interest rates at 10 percent was recently introduced in the U.S. Senate.[1] In this research note, we explore the potential effects of a 10 percent interest rate cap on consumer access to general purpose credit card (bank card) borrowing using data from the Federal Reserve Board’s triennial Survey of Consumer Finances (SCF).
Our analysis finds that the proposed cap will adversely affect access to card credit for those cardholders who regularly use their cards for borrowing by rolling over a portion of their balance from month to month. As many as two-thirds of such cardholders are likely to see their credit lines curtailed or eliminated. Most vulnerable to loss of their credit line are consumers with imperfect credit histories, who would generally be those most in need of the backup liquidity provided by bank cards.
Thus, while the proposed cap is a well-intentioned effort to reduce the high debt burden some households are facing, it would harm consumers’ access to card credit. Yet bank cards, because of their relatively low minimum required monthly payments, are a vital and affordable source of backup liquidity for many households, particularly for those with volatile incomes and limited savings.
For example, according to a recent Federal Reserve study, 37 percent of respondents to a nationwide survey said they would have difficulty covering an unexpected $400 expense out of their income or savings. More than 40 percent of those households indicated they would cover such an expense by borrowing on their credit card.[2]
The rationale for imposing rate caps on credit cards is questionable to begin with, as there is little evidence of market inefficiencies that would require such extreme intervention, particularly in light of the changes implemented by the Credit Card Act of 2009 (the CARD Act). These modifications to how terms and conditions on credit cards are structured and disclosed addressed historical concerns about non-transparent terms and associated inefficiencies. They also addressed practices seen as unduly harmful to a consumer’s financial health, such as interest rate increases affecting a customer’s pre-existing balance.
Research has demonstrated that the CARD Act reforms made the credit card market more efficient and led to more equitable pricing that benefited consumers.[3] Consequently, there should be a high bar for demonstrating the necessity of further regulatory interventions in the credit card market, especially for such a draconian measure as a 10 percent interest rate cap, which will have a negative effect on consumers’ ability to access credit.
Our analysis relies primarily on the SCF sample from 2019, a more stable macroeconomic environment compared to 2022, which was distinguished by sharply rising interest rates, although both samples tell a similar story. The 2019 data indicate that 22 percent of bank card users in the United States, corresponding to 21.1 million individuals and families, rarely paid their bank card balances in full — they regularly carried a revolving bank card balance on which they pay interest.
By our assessment, a 10 percent cap would have adversely affected at least two-thirds of these borrowers, or 14.3 million individuals and families, resulting in their credit lines being eliminated or significantly reduced.[4] The cap would have had a particularly severe effect on consumers with riskier credit profiles, close to three-fourths of whom would have been adversely affected. Access to card credit can be especially important for these consumers, as a path to reestablishing a favorable credit record and improving their credit scores.
Our analysis is based on data from 2019 and 2022, the most recent survey years for the SCF. Market interest rates then were substantially lower than today. In today’s higher interest rate context, a significantly larger proportion of consumers would see their access to card debt curtailed as a result of a rate cap.
Bank Card Interest Rates in the 2019 SCF
Our analysis relies on detailed information about family financial circumstances, including on their use of bank cards, from the 2019 and 2022 SCF, an interview survey of U.S. families sponsored by the Board of Governors of the Federal Reserve System with the cooperation of the U.S. Department of the Treasury.[5] The survey collects information on family income, debt, financial assets, wealth, pensions, labor force participation, demographic characteristics, homeownership status and use of financial services. As previously noted, we rely primarily on the 2019 sample.
The SCF is designed to provide reliable information both on attributes that are broadly distributed in the population (such as on homeownership) and on those that are more narrowly applicable (such as on closely held businesses).[6] Families were interviewed between May and December of the survey year. Most of the information collected is current as of the time of the interview; the primary exception is information on families’ total income before taxes, which pertains to the calendar year preceding the survey.[7]
For respondents who have bank card accounts, the SCF records whether the consumer “rarely, sometimes or always” makes full payment on the balance owed each month. Those who rarely pay in full have revolving balances on which they generally would be making interest payments; those who sometimes pay in full would make occasional interest payments; while those who always pay in full are non-revolvers. In addition, the SCF records the interest rate on the account with the largest balance or (for consumers that pay their balances in full) the card acquired most recently.
The 2019 SCF sample includes 4,560 families who have at least one bank card, corresponding to about 96 million families in the total U.S. population. Table 1 shows the distribution of the population across the three specified payment behavior segments and the median and 75th percentile interest rates for the overall population and by segment.
Among these consumers, 58 percent always pay their balances in full and 22 percent rarely pay in full. Consumers who rarely pay their balances in full exhibit distinctly higher interest rates: their median rate is 17.1 percent, which compares to 16 and 15 percent for always and sometimes pay-in-full, respectively.
Table 1: Interest Rate by Payment Behavior Segment
Source: 2019 Survey of Consumer Finances. SCF weights are used except for count of respondents.
These cross-segment interest rate differences likely reflect consumers’ risk profiles. Consumers capable of paying their balances in general are less likely to have had incidents of late payment and would tend to have lower credit line utilization rates, attributes associated with higher credit scores, enabling them to qualify for cards with lower interest rates. In contrast, consumers who rarely pay in full are relatively likely to exhibit risk factors associated with lower credit scores and thus less likely to qualify for cards with lower rates.
Consumers who rarely pay in full are the most relevant population for evaluating the effects of an interest rate cap, since they consistently pay interest on their revolving balances. On this basis, and because they are less likely to qualify for cards with lower rates, we focus our discussion on this cohort of consumers.
Possible Effects of an Interest Rate Cap: Key Considerations
Because we only have data from a market not subject to a rate cap, the effects of imposing a cap are difficult to analyze. Banks and consumers may adapt to a cap in ways not yet observed.
The loss of interest income will make it less profitable for banks to continue to provide credit to borrowers with interest rates that a priori exceed the cap, without a corresponding reduction in the cost of providing the credit lines or an increase in annual or other fees. There are several ways that banks might respond, short of terminating these credit lines entirely.
One possible response would be to reduce credit limits and raise minimum required payments, which would generally have the effect of reducing the risk of borrowers becoming delinquent and thereby reduce expected credit losses.[8] However, curtailing credit limits may also reduce revenue, so it may be necessary to take additional measures to reduce costs, such as offering fewer benefits to cardholders, including reduced rewards for purchases. Moreover, for higher-rate accounts associated with higher expected credit losses, it may not be possible to sufficiently compensate for the loss of interest income through cost reductions, compelling banks to withdraw the credit lines completely.
Second, banks might seek to reduce costs by engaging in less marketing and promotional activity, including by eliminating or curtailing zero-interest or low-rate promotional offers. They might also seek to increase fee income by adding an annual fee, less frequently waiving late payment fees and increasing their fees for balance transfers and cash advances. Again, for higher-rate accounts associated with higher expected credit losses, it may not be possible to sufficiently compensate for the loss of interest income through such measures.
Third, for consumers whose pre-existing rates marginally exceed the cap, banks might reduce interest rates to comply with the cap and accommodate the marginally lower profitability of their credit card operations. The degree to which they would be willing to do so depends on current profitability of their credit card activities relative to alternative banking and financial activities, which is difficult to assess ex-ante. As credit cards encompass both payment services and provision of credit lines to consumers and small businesses, and typically are provided by diversified financial institutions, measuring their profitability is challenging and subject to error.
Given the highly competitive nature of credit card lending and related activities, the scope for excess profitability is no doubt limited. Conceivably, though, marginal excess profits might presently be sustained by frictions affecting consumers’ ability to shop for a lower rate or to switch between issuers. For instance, consumers may make mistakes in shopping for or accepting offers of card credit that leave them paying higher rates than they might otherwise be eligible for, enabling banks to earn some excess profit.[9] If so, there may be some leeway for banks to grant marginal interest rate reductions in response to the cap.[10]
To address such knowledge gaps, we shall invoke two assumptions. First, we allow that banks would reduce card interest rates within a margin above the 10 percent cap, such that consumers with rates as high as 12 percent would retain their existing credit lines unaffected by the cap. This is a conservative assumption in the sense of allowing that banks might substantially offset the reduced revenue from the cap. Second, we assume that among consumers observed in the data to have zero or very low introductory rates, none would be affected by the cap.[11] This assumption is also conservative—in a rate cap environment, introductory rates would likely be curtailed.
Effect of a Rate Cap as Inferred from the Distribution of Credit Card Interest Rates
Figure 1 presents a histogram depicting the full distribution of interest rates for 2019 SCF respondents who rarely pay their card balances in full. About two-thirds of this population have interest rates greater than 12 percent, our assumed threshold for the 10 percent cap to be binding. These consumers are likely to see their credit lines reduced or eliminated as a result of the cap.
Figure 1 also indicates that 7 percent of consumers (those with rates above 10 and less than or equal to 12 percent) would benefit from a 10 percent cap under the assumption that banks would grant interest rate reductions of up to 2 percentage points. However, these consumers receiving lowered rates might also see some curtailment of benefits or an increase in account-related fees.
Additionally, Figure 1 indicates that (as of their interview dates) 11 percent of consumers had interest rates in the 0 to 4 percent range (most of whom had a 0 interest rate), which presumably are low introductory rates assigned for a specified period of time. An unknown number of consumers with interest rates in the 5 to 10 percent range may also have introductory rates.
Figure 1: Distribution of Interest Rates — Rarely Pay in Full
Interest Rates in Relation to Consumer Risk Profile
Typically, a major factor influencing bank card interest rates is the consumer’s credit quality. Consumers assessed to have a greater risk of default would pay a higher interest rate to compensate the lender for bearing this risk. These consumers can be expected to be more severely affected by an interest rate cap.
Ideally, we would distinguish consumers’ credit quality based on a credit score, which is the primary indicator of creditworthiness that banks rely on for approving applications for credit cards. However, the SCF does not include credit scores for respondents. Therefore, we rely on the following three indicators, which are based on information available in the SCF and are expected to strongly correlate with having a below-prime credit score (indicating “high-risk”).
The first indicator, which we term “flawed credit,” is for whether the consumer fell behind schedule on a debt payment obligation during the 12 months prior to the interview date or had filed for bankruptcy or had a home in foreclosure during the past five years. Any such credit event typically would have a significant, negative impact on a consumer’s credit score.
The second is an indicator for whether the consumer had been turned down on an application for non-mortgage credit during the 12 months preceding the interview date (“turned down”). A common reason for having an application for credit denied is that the applicant’s credit score does not satisfy the lender’s requirements for the credit product in question. Thus, we expect that this indicator correlates with having a below-prime credit score.
In addition, we consider a less direct indicator, which is whether the consumer’s bank card utilization ratio, as calculated from balance and credit limit information in the SCF, exceeds 50 percent (“high-utilization”). The SCF records the total balance still owed on a consumer’s bank card accounts after the last payments were made on all of these accounts, and the aggregate credit limit of all of these accounts. The utilization ratio is calculated as the ratio of these two items. Often, high credit card utilization rates have a significant, negative impact on a consumer’s credit score.[12]
Note that while the above factors should correlate with a below-prime credit score, the identification is necessarily imprecise. Some of the SCF respondents would have credit scores in the below-prime range but not be identified as high-risk based on these factors, and the approach may misclassify some consumers with prime scores as being high-risk.[13]
Figure 2 shows the proportion of consumers in each and any of these risk categories. About three-fifths of the sample exhibit at least one of the three risk factors. That this would be the case is not surprising, since the sample consists of consumers who rarely pay their card balances in full, who tend to have higher utilization rates and lower credit scores than non-revolvers.
Figure 2: Distribution of Consumers by Risk Category
Figure 3 presents histograms of bank card interest rates for the sample split by risk category: consumers who are identified as high-risk based on any of the three indicators, and those who are not. These interest rate distributions confirm that elevated risk of default is reflected in higher interest rates on card credit. The median and 75th percentile interest rates are 18 and 23 percent, respectively, for consumers who are identified as high-risk, which compares to 15 and 19 percent for those who are not.
Nearly three-fourths of the high-risk population segment and three-fifths of the low-risk segment have interest rates greater than 12 percent, our assumed threshold for the cap to be binding. They are likely to see their credit lines reduced or eliminated as a result of the cap.
Traditionally, many consumers paying a high interest rate at a given point in time due to below-prime credit scores or other risk factors obtain lower-cost credit within a year or two by improving their credit scores over that period.[14] Higher-risk consumers who lose access to bank card credit will find it more difficult to improve their credit scores by demonstrating reliable performance with credit, prolonging their loss of access to credit.
Figure 3: Distribution of Interest Rates by Risk Category
High-risk
Not high-risk
Table 2 summarizes our calculations by risk segment for the population that rarely pays in full. The more than two-thirds of consumers we assess would be harmed by the proposed rate cap corresponds to 14.3 million U.S. families.
It is important to bear in mind that we have been restricting attention to consumers who rarely pay their card balances in full. An interest rate cap would also adversely affect many of the 18.7 million U.S. families (as of 2019) who sometimes pay in full, many of whom may revolve a balance over much of the year and have relatively high interest rates.
Moreover, card interest rates in 2019 averaged about 17 percent, whereas currently they average about 23 percent.[15] Therefore, it is almost surely the case that if a 10 percent cap were applied today, then the share of consumers that would see their access to card debt curtailed would be substantially larger than indicated by this analysis.
Table 2: Effect of a 10 Percent Cap by Risk Cohort
Source: 2019 Survey of Consumer Finances. SCF weights are used for all calculations.
Consumers with revolving balances
Finally, we repeat the analysis using the 2022 sample. As noted, credit card interest rates increased steeply during 2022, reflecting the rising interest rate environment in financial markets generally. Card interest rates began the year about a percentage point below 2019 levels and were about 3 percentage points higher by the end of the year.
Table 3 shows the distribution of the population in 2022 across the three payment behavior segments and the information about card interest rates analogous to Table 1. The payment behavior distribution is almost the same as in 2019.
Table 4 summarizes our calculations of the share of consumers for whom the 10 percent rate cap would be binding by risk segment, for the population that rarely pays in full. Again, we assess that about two-thirds of consumers overall, corresponding to 15.1 million families, would be harmed by the proposed rate cap.
The difference in proportion affected by the cap for high-risk consumers versus others is narrower compared to 2019. This conceivably might reflect the rising interest rate environment. For instance, the higher risk cohort may have been overrepresented among those surveyed earlier in the year.
Table 3: Interest Rate by Payment Behavior Segment – 2022 SCF
Source: 2022 Survey of Consumer Finances. SCF weights are used for all calculations.
Consumers with revolving balances
Table 4: Effect of a 10 Percent Cap by Risk Cohort – 2022 SCF
Source: 2022 Survey of Consumer Finances. SCF weights are used for all calculations.
Consumers with revolving balances
Concluding Remarks
This research note has explored the potential effects of a 10 percent interest rate cap on consumer access to card credit using data from the 2019 and 2022 SCF. As of 2019, among the 21 million U.S. families that rarely pay their bank card balances in full, at least 14.3 million, or 68 percent, would likely have had their access to credit curtailed if subject to the cap. The 2022 sample yields a similar assessment. These assessments represent a “best case scenario” that reflects lower interest rate environments than prevail today and applies conservative assumptions.
For instance, the analysis assumes that among consumers who exhibit zero or very low introductory rates, none would be affected by the cap. In actuality, banks would likely reduce the offering of such programs in response to the cap. In the 2019 sample, about 2.3 million families among those who rarely pay in full had introductory rates (11 percent of this population). Many of them would have had their access to credit curtailed in the presence of a cap.
An interest rate cap would disproportionately curtail access to card credit among consumers with riskier credit profiles. They may then lose an opportunity to boost their credit scores by demonstrating reliable repayment performance, causing their loss of access to credit to persist.
Our analysis focuses on the population of consumers who rarely pay their card balances in full. However, among the additional, 18.7 million U.S. families (as of 2019) who sometimes pay in full, many may revolve a balance over much of the year and also could experience curtailed access to credit as a result of the cap.
[1] See https://www.sanders.senate.gov/press-releases/news-sanders-hawley-introduce-bill-capping-credit-card-interest-rates-at-10/
[2] The other 40 percent indicated they would have to borrow from a friend or family member, sell an asset, use a payday loan or simply not pay the expense. See “Report on the Economic Well-Being of U.S. Households in 2023,” Board of Governors of the Federal Reserve System. https://www.federalreserve.gov/publications/files/2023-report-economic-well-being-us-households-202405.pdf
[3] A 2014 paper from the National Bureau of Economic Research found that the CARD Act enhanced the efficiency of the credit card market as reflected in reduced borrowing costs for consumers. See Sumit Agarwal, Souphala Chomsisengphet, Neale Mahoney and Johannes Stroebel, “Regulating Consumer Financial Products: Evidence from Credit Cards”, Working Paper 19484, National Bureau of Economic Research, June 2014. Also, a 2015 report from the Consumer Financial Protection Bureau found that the Card Act substantially reduced the overall cost of card credit. The study also noted a few practices associated with particular card products that may merit further attention.
[4] As explained in detail later, we evaluate the impact of a 10 percent rate cap by determining the share of consumers for whom a 12 percent cap would be binding given their reported interest rate, thus allowing that marginal rate reductions might be granted (possibly with other changes in terms) without reducing the credit line.
[5] For convenience, we use the term families to connote single-person households as well as families.
[6] This is accomplished by means of a two-part sample design: a standard, geographically based random sample and a special oversample of relatively wealthy families. Missing data in the survey are imputed using a multiple imputation technique involving five separate imputation replicates (implicates). Thus, the public version of the data set consists of 28,885 records for 5,777 families. Each record is assigned a weight to adjust for the non-random nature of the sample design, such that the weighted set of observations will be representative of the U.S. population. We utilize all five implicates and the assigned weights for calculating sample statistics and frequency distributions.
[7] For additional details, see https://www.federalreserve.gov/econres/scfindex.htm
[8] In general, raising the minimum required payment reduces long-run credit risk due to the debt is being repaid faster. This may not be the case for those consumers who are already just scraping by; that is, those for whom meeting the a priori required minimum is already difficult. However, even for those consumers an increase in the share of required principal payment that just offset the reduced interest payment under the cap would, at worst, be neutral.
[9] See, for example, Igor Livshitz, “Regulating Consumer Credit and Protecting (Behavioral) Consumers,” Federal Reserve Bank of Philadelphia Economic Insights, First Quarter 2020, pp. 14-19. Available at: https://www.philadelphiafed.org/-/media/frbp/assets/economy/articles/economic-insights/2020/q1/eiq120-consumer-protections.pdf
[10] Informational constraints on lenders’ ability to evaluate applicants for credit are another potential friction, impeding a consumer’s ability to swap an existing card credit line for a new, lower cost line with a different institution. See, for example, Calem, Paul S., Michael B. Gordy, and Loretta J. Mester (2006), “Switching Costs and Adverse Selection in the Market for Credit Cards: New Evidence.” Journal of Banking and Finance 30, pp. 1653-1685. Arguably, these barriers have lessened over time due to advances in credit scoring data and methods.
[11] The period for an introductory rate often extends a full year and sometimes to 18 months; per the CARD Act it cannot be shorter than six months, after which the introductory rate is replaced by a much higher rate.
[12] Most consumers indicated to be high-risk by either having flawed credit or a turned-down non-mortgage credit application are also high-utilization.
[13] For example, being turned down for credit may reflect factors other than credit score, such as the consumer requesting a larger loan or credit line than the lender is willing to provide based on the borrower’s assessed capacity to repay.
[14] The dynamic nature of consumer credit scores is explored by Sarena Goodman and Steve Ramos in “Understanding Changes in Household Debt by Credit Risk Category: The Role of Credit Score Transitions,” Federal Reserve Bank of New York (2019) https://www.federalreserve.gov/econres/notes/feds-notes/the-role-of-credit-score-transitions-20190625.htm
[15] See Board of Governors of the Federal Reserve System, Consumer Credit – G.19 report, available at https://www.federalreserve.gov/releases/g19/current/
