Household Need for Liquidity and the Credit Card Debt Puzzle

Abstract

In the 2001 U.S. Survey of Consumer Finances, 84% of households that revolve credit card debt simultaneously hold significant liquid balances. The so-called "credit card debt puzzle” is: given the average 14% APR on credit cards and 1-2% interest on deposit accounts, why not pay down the debt? In this work, I evaluate the following hypothesis: households that accumulate credit card debt may not pay it down using their liquidity because they expect to use it for goods for which credit cards cannot be used. First, I document in aggregate and survey data that liquid assets are a substantial part of household portfolios, and that consumption in goods requiring liquid payment appears to have a sizeable unpredictable component. This may warrant holding precautionary balances in liquid accounts. Second, I analyze the issue by extending standard monetary theory to incorporate consumer debt, to show qualitatively how the inability to pay for certain goods by credit yields co-existence of money and debt in portfolios of rational households. Third, to evaluate the hypothesis quantitatively, I adapt the model for computation to a partial-equilibrium setting with exogenously incomplete markets, calibrate the model to match key properties of the data, and solve it. I find that the occasional inability to pay with credit and the degree of uncertainty in expenditures that households face induce any household to keep a portion of its assets liquid - even if it revolves debt simultaneously. The calibrated model accounts for 81% of households in the data that keep both money and debt, and for up to 55 cents of every dollar that they keep liquid. Finally, I discuss realistic institutional constraints, such as cash advances, that make portfolio rebalancing costly, but are omitted in the benchmark model. I find that such restrictions may increase liquidity demand in the benchmark case by up to 48%, potentially adding significantly to the ability of the hypothesis to account for the credit card debt puzzle. I conclude that the puzzle may largely not be puzzling; resulting instead from rational behavior of optimizing households in an incomplete-market economy with frictions.