|
Abstract
|
As of 2023, the United States has accumulated a staggering $1.13 trillion in credit card debt, with delinquency rates on the rise, particularly among younger credit cardholders (Federal Reserve Bank of New York, 2024). Many credit cardholders could reduce their borrowing costs by choosing a more suitable credit card. Yet, selecting an appropriate credit card is challenging. Our study aims to examine whether the complexity of individual credit card choices affects choice quality by characterising this complexity using computational complexity. Method: 26 participants completed a two-alternative forced choice task. Given a fixed 12-month spending and repayment schedule, participants needed to select the credit card that minimised their total borrowing costs. Each credit card had up to two cost features: an annual fee and an annual interest rate, which were sampled from the Australian market. We operationalised the complexity of choosing a credit card in two ways: the number of compounding steps required to compute the total interest payments and credit card feature composition. Complexity was higher with more compounding periods and when a credit card had an interest rate. We also measured subjective complexity based on participants’ perception of task difficulty. Results: We found that as the complexity of a particular credit card choice increased, participants spent more time and made more mistakes, but only when the cost of mistake was less than AU$113 per annum. Subjective complexity was positively correlated with more complex feature composition, but, surprisingly, it was inversely related to the number of compounding steps. Conclusion: We introduce a framework to quantify the complexity of consumer financial decisions based on the computational resources required to make the decision. Our results show that higher complexity choices can impair choice quality and that even simple financial decisions, such as choosing credit cards, are beyond consumers’ cognitive capacity.
|