Consumer-finance experts usually advise people with balances on multiple credit cards to pay off the card with the highest interest rate first.
However, relatively few people do that, and research suggests why. According to University of Nottingham’s John Gathergood and Jörg Weber, Chicago Booth’s Neale Mahoney, and University of Warwick’s Neil Stewart, consumers try to manage multiple debts at once rather than pay off the most expensive one. That flawed approach can be costly.
The most cost-effective thing people with two or more credit-card balances can do is to first make the minimum payment on each, then repay as much as possible on the card with the highest interest rate. Only after paying off the highest-rate card in full should they pay down the lower-rate cards. This minimizes overall interest payments and maximizes debt reduction.
Where credit is due
Consumers carrying balances on two credit cards with varying interest rates tend to pay off the cards more evenly than is optimal.
The researchers looked at 100,000 individuals in the United Kingdom who held a revolving balance on two credit cards over a two-year span. The average difference in interest rates between the two cards was 6 percentage points.
On average, the amount people paid monthly toward their higher-interest card was £259, only marginally more than the £230 paid toward the other card. The researchers calculate that, after paying the minimums, individuals should have allocated 97 percent of the remaining payments to the higher-interest card, although in practice they actually allocated 51.5 percent to that card.
Gathergood, Mahoney, Stewart, and Weber consider a number of explanations, including that people could be repaying the credit card with the highest capacity, the highest balance, or the lowest balance. They pit these explanations against their own theory, “balance matching,” in which payments are proportionate to overall amounts owed.
Using statistical approaches, the researchers find that it’s far more likely that people were engaging in balance matching. “Consistent with the balance matching results, the machine learning models confirm that balances are hugely important for predicting behavior,” they write. The other explanations predict people should concentrate on repaying a single credit card, but the data suggest people instead juggle multiple debts—focusing on high balances rather than high interest rates.