Do credit card companies screen for behavioural biases?

BIS Working Papers  |  No 842  | 
12 February 2020
PDF full text
 (542kb)
 |  53 pages

Focus

As they collect and process detailed consumer data, financial firms get better at designing products that serve their customers' needs and preferences. However, these skills also create scope for firms to exploit consumers' fallibilities or "biases", such as limited attention to detail and/or short-termism ("myopia"). This paper investigates whether this is actually the case in the US credit card market.

Contribution

Credit card contracts differ significantly in how far they emphasise back-loaded or front-loaded terms. For instance, some cards offer low introductory rates coupled with high late-payment fees, while others charge higher but uniform annual fees. Back-loaded terms can be attractive for customers with low credit limits, but issuers may also use such terms to extract higher fee income from less perceptive ("myopic") customers. The paper provides evidence for card issuers' strategies by analysing 1.3 million credit card offer letters in the United States.

Findings

The paper shows that issuers do seek to exploit behavioural biases. To this end, they highlight attractive front-loaded features, while "shrouding" unattractive back-loaded features. They do this by placing information more or less prominently, and by adjusting the font size or complexity of language. Moreover, in the case of less-educated consumers, for whom returns on shrouding are likely to be higher, issuers use shrouded back-loaded features more heavily. In principle, such a strategy could result in potentially higher credit risks as customers who are acquired based on their biases may be more likely to default in future. However, the analysis reveals that card issuers consider this trade-off and generally send shrouded offers only after consumers have improved their credit risk profile.


Abstract

Using granular data on the contract terms and design details of more than 1.3 million credit card offers, we document how card issuers shroud unappealing, back-loaded features of an offer (e.g., high default APRs, late or over-limit fees) via the position of the information, font size, or complexity of the language used. More heavily shrouded offers that rely on back-loaded fees are also more likely to be offered to less-educated consumers. In addition, we document a novel interaction between behavioral screening and adverse selection: Using changes in state-level unemployment insurance (UI) as positive shocks to consumer creditworthiness, we show that issuers rely more on shrouded and back-loaded fees when UI increases, especially for less-educated consumers. Card issuers weigh short-term rent maximization against increased credit risk when targeting consumers' behavioral biases.

JEL codes: G02, G1, G21, G23

Keywords: credit card, shrouding, back-loaded