Two thirds of United States households had credit cards by the end of 1998. More than half of credit card holders actually revolve debt on their cards,while nearly half of card holders declare that they do not usually pay off their credit card balance. Half of those who revolve debt actually declare themselves as systematic debt revolvers and have at the same time liquid financial assets of at least $500 and no less than one-half of total monthly income. The actual size of liquid assets among households in this group makes it unlikely that this co-existence can be attributed simply to a motive for holding transactions balances. Median liquid financial assets, excluding cash holdings, are $4850 for all such households, while their median credit card debt is $1900 and the median credit card interest rate on their card with the highest balance is equal to 15 percent. Failing to arbitrage between the high credit card interest rates and the low rates on their liquid assets is costly. Despite these pronounced tendencies to revolve credit card debt, by the time they are included in the 50 to 59 age range United States households manage to attain median ratios of total wealth (net of credit card debt, mortgages,and education loans) to annual total after-tax income of the order of 3.5. The surprising portfolio co-existence of high-interest credit card debt, low-interest liquid asset holdings, and substantial accumulation of assets for retirement is a challenge for the modern theory of saving and the focus of this paper.
Gross and Souleles (2002) used a unique, proprietary panel data set of thousands of individual credit card accounts from several different card issuers and identified two co-existence puzzles: one refers to the co-existence of credit card debt with illiquid assets, and the other to the co-existence of credit card debt with liquid assets. Two subsequent papers, namely Laibson,Repetto and Tobacman (2000) and Bertaut and Haliassos (2002), proposed explanations for these puzzles that were based on self-control considerations.
Support for considering self-control explanations of saving and borrowing behavior can be found in a number of different sources. Information on attitudes of credit card holders in the 2001 Survey of Consumers shows that about forty percent of card holders perceive self-control problems emanating from credit cards and the possibility of overspending, although fewer are willing to admit that they personally face such problems (Durkin, 2002). Self-control considerations are stressed in the Marketing literature and in research on consumer psychology. Hoch and Loewenstein (1991) argue that self-control problems occur when the benefits of consumption occur earlier and are dissociated from the costs, as is the case with credit card spending. There is evidence suggesting that liquidity enhances both the probability of making a purchase and the amount one is willing to pay for a given item being purchased, over and above any effects due to relaxation of liquidity constraints (see Shefrin and Thaler, 1988; Prelec and Simester, 2001; and Wertenbroch, 2002). Indeed, Soman and Cheema (2002) present experimental and survey evidence to argue that consumers interpret their available credit lines as indications of future earnings potential when deciding their level of consumption expenditures. More generally, costly self-rationing as a means of self-control has been stressed in Marketing research and in economic research on mental accounting. Interestingly, the idea that self-control considerations and “low-willpower techniques” are important in understanding saving behavior is now openly considered by policy makers.
In an intriguing study, Laibson, Repetto and Tobacman (2000) focused on the co-existence of credit card debt and accumulation of assets for retirement. They showed that a standard consumption-saving model with credit cards and consumers discounting future utility exponentially could not account simultaneously for the proportion of households revolving debt on their credit cards and for the considerable pre-retirement ratios of total wealth to total after-tax, non-asset income (including income from transfers, bequests, and lump-sum windfalls). Intuitively, they showed that if a household, as commonly analyzed, valued the future so much as to accumulate such high levels of wealth for retirement, it could not be acting so impatiently as to tend to revolve debt on the credit card at anywhere close to the observed frequency.
Their proposed solution was a model of hyperbolic discounting. In such a model, the current incarnation of the self discounts the immediate future considerably, thus justifying high-interest credit card borrowing, but discounts the distant future by less than he anticipates his future incarnations to discount it. In other words, his current willingness to accumulate for retirement is greater than the willingness he expects to have at a later stage in his life. So, he decides to accumulate illiquid assets for retirement now, imposing heavy (in the limit, infinite) liquidation costs to the future selves who will be prone to act impatiently, and thus helping to ensure that enough assets will have been accumulated by the time of retirement. In this model, accumulation of illiquid assets can co-exist with credit card debt revolving, since hyperbolic consumers can act impatiently with respect to short-run objectives and patiently with respect to longer-run objectives. Credit card borrowing is intended for short-term consumption smoothing, while accumulation of illiquid assets is used as an instrument of self-control, in particular as an instrument of controlling future selves. As the authors mention, however, this model with its emphasis on temporally separated selves cannot simultaneously address the other unhappy co-existence, namely that of credit card debt with low-interest liquid assets.
Downlaod
PDF Ebook Credit Card Debt Puzzles
