The Credit Card Debt Puzzle: The Role of Preferences, Credit Risk, and Financial Literacy 
As of 1983 in the United States, and later in Denmark and the United Kingdom, researchers have documented a type of consumer behavior that has come to be known as the "credit card debt puzzle": individuals who choose to revolve unsecured high-interest credit card debt while also holding low interest-bearing monetary assets that could be used to pay down these balances. The authors use a dataset constructed from the 1979 National Longitudinal Survey of Youth (NLSY79) in order to test different theoretical explanations for the credit card debt puzzle. The NLSY dataset permits examination of the credit card debt puzzle over different time periods, it contains measures of intelligence, financial literacy, and risk preferences, and provides details on household income, assets and liabilities. Based on reported amounts of credit card debt and liquid monetary assets, the authors divide the NLSY79 respondents into four categories: borrower-saver (the puzzle group holding both debt and assets), borrowers (debt, no assets), neutral (no debt, no assets), and savers (assets, no debt).
Key Findings
- Respondents in the puzzle (borrower-saver) group are very similar to savers in terms of intelligence, education, financial literacy, and financial knowledge. Borrower-savers and savers score higher on these attributes than do members in the borrower and neutral groups, and have more income and wealth. Members of the puzzle group have slightly lower annual family income (averaging $89,295 in 2004) and wealth (averaging $322,574 in 2004) compared to the saver group (in 2004, average annual income for this group is $97,195 and average wealth is $492,574).
- Borrower-savers are distinguished from savers by having a higher appetite for credit, higher discounts rates (being more present-biased), middle levels of risk aversion, and higher actual and perceived credit risk. Changing perceptions about credit risk are essential for predicting transitions among the puzzle group and the saver group. Respondents whose credit risk increases over time are likely to transition from being savers to being borrower-savers, while those individuals whose credit risk decreases over time are more likely to transition from the puzzle group to the saver group, conditional on other observable variables.
- Overall, borrower-savers are a very heterogeneous group; many of them can simultaneously hold revolving credit card debt and liquid assets without experiencing financial trouble, while others cannot. It appears that for at least some subset of the puzzle group, the behavior associated with the credit card debt puzzle may be informed by some financially sophisticated reasoning informing strategic choices. In particular, the authors find support for the hypothesis that the puzzle group engages in precautionary borrowing in anticipation of future adverse income or credit shocks.
Implications
Changing borrower-saver behavior will be difficult, as it is partly driven by individual preferences toward risk and time discounting. However, better financial knowledge may mitigate the effect of these preferences.