Between May 1989 and November 1991, the prime rate charged by commercial banks dropped from 11.5 percent to 7.5 percent, and the interest rate on large-denomination certificates of deposit fell from around 9 percent to about 5 percent. During this entire period, bank credit card rates barely moved, with the largest issuers holding their rates fixed at 18 to 20 percent.
This recent stickiness of credit card rates repeated a familiar story. During several episodes in the 1980s, when other interest rates rose or fell, credit card rates changed little. At the same time, credit cards consistently earned higher returns than most other bank products. A carefully done study by Ausubel (1991) concludes that during the 1980s, bank credit card operations earned three to five times the rate of return earned in the banking industry at large.
This observed performance is intriguing when one considers the fragmented structure of the industry. There are numerous providers of credit cards and no major barriers to entry into the industry; one would expect such a market structure to lead to competitive performance, whereby prices would align with costs and issuers would earn a normal rate of profit.1 Ausubel argues that the industry deviates from the perfectly competitive model because consumers (cardholders) do not conform to the behavioral assumptions of perfect competition. He argues that discrepancies from the outcome of the perfectly competitive model could result from any or all of the following causes: consumers face search costs; consumers face switch costs; and firms would face an adverse selection problem if they were to unilaterally reduce their interest rates.
The present paper presents empirical evidence in support of this argument, drawing on data from the Federal Reserve's 1989 Survey of Consumer Finances. Unlike previous studies, we provide specific evidence about consumer behavior, which is at the core of the theories on credit card rate stickiness. Overall, our analysis suggests that each of the three factors cited by Ausubel has contributed to the observed performance of the credit card market.2 In addition, the paper advances two further arguments supporting Ausubel's contention that credit card issuers may face an adverse selection problem.
Ausubel posits that many cardholders, particularly those representing minimal default risk, end up borrowing more than they expect to a priori. He shows that such consumer irrationality can naturally induce an adverse selection problem. We show that search costs or switch costs also can induce an adverse selection problem. We argue first that consumers who face high search costs tend to maintain higher credit card balances than consumers who face low search costs. Second, consumers with high credit card balances may face relatively high switch costs. Either way, a firm that competes on interest rates will disproportionately draw customers who maintain lower balances and hence yield lower profits.
Our principal empirical findings can be summarized as follows. Holding constant demand for and access to credit, we find that credit card indebtedness is inversely related to an individual's propensity to comparison shop »for the best terms» on loans or deposits. This result suggests that consumers with substantial search costs tend to have high balances. We interpret this as evidence that card issuers face an adverse selection problem induced by search costs. Although this result may also be consistent with Ausubel's reasoning that consumers holding large balances are often irrational or naive in their search behavior, we find no direct evidence (no inconsistencies in consumer responses) to indicate that consumers may be underestimating their propensity to borrow.
In addition, we find that households with larger outstanding card balances are more likely to have applications for credit denied and are more likely to have experienced payment problems. These findings suggest that holding large card balances signals credit risk and makes it harder for a consumer to obtain alternative credit. We interpret these findings as evidence that card issuers face an adverse selection problem induced by switch costs.
The paper is organized as follows. Section 2 reviews the interest rate and profit performance of the bank card market and the market's apparent divergence from the textbook model of perfect competition. In Section 3, we review the theoretical explanations for this observed performance, and introduce our two new arguments pertaining to adverse selection. In section 4, we describe our data and present our empirical analysis of consumer search behavior. Section 5 provides our empirical evidence on adverse selection due to consumer switch costs. Section 6 presents conclusions and some policy implications of our findings.
Download
PDF Ebook Consumer Behavior And The Stickiness Of Credit Card Interest Rates
