Abstract

I argue that the measure of credit card debt used by researchers has grown rapidly in part because it captures debt arising from transactions in which a credit card is used because of its advantages over other payment instruments. Increases in debt stemming from such use may not signal greater household financial vulnerability if households are willing and able to repay this short-term debt. However, it may suggest that the cost of using credit cards to pay for purchases has declined relative to other payment instruments. I conclude that had transactions demand remained at its real 1992 levels, rather than growing almost 15 percent per year, measured credit card debt would have grown a bit less than 1 percentage point slower per year between 1992 and 2001. Moreover, I show that removing transactions demand from aggregate consumer credit can alter conclusions about the relationship between credit and consumption.