Consumer debt has traditionally been analyzed within the Fisher framework of shifting consumption between periods and smoothing consumption over the lifecycle. While the shifting of consumption to earlier periods via various debt instruments will necessarily have a negative impact on future consumption as debt is repaid with interest, the assumption is usually made that consumers will indeed over time successfully repay the debt acquired earlier in life. However, with the introduction of non-secured lines of credit with flexible payments i.e., credit cards the traditional patterns of consumption smoothing and debt repayment may no longer be the same for U.S. households. Unlike mortgages, student loans, auto loans, and other installment loans with which the U.S. population has a relatively long history, substantial credit card debt is a fairly recent phenomenon, and predictable debt accumulation and repayment trends have yet to be firmly established. Besides the flexible repayment feature of credit card debt, the possibility of default and bankruptcy poses further uncertainties about how consumers will repay their debt over the lifecycle.
To understand debt accumulation and repayment patterns over time and make reliable predictions, a lifecycle analysis on credit card debt and payoff rates is necessary. Due to data limitations, most previous studies analyzed consumer debt in a static or comparative static context. Using a synthetic cohort approach, this research incorporates a time dimension to provide evidence on credit card debt and payoff rates from a lifecycle perspective. The objectives of this study are: to empirically estimate the lifecycle profiles of credit card borrowing and repayment behavior for 14 different birth cohorts and compare the patterns with the implications of the simple lifecycle model, and to examine the financial, socioeconomic, demographic, and psychological determinants of credit card debt and payoff rates.
A two-way fixed effect, pseudo-panel data model is proposed to characterize both the cohort effects and the time effects, and a two-step estimation procedure which greatly simplifies the estimation process and allows for more flexibility in model specification is utilized in the regression to separately address the above two objectives. The results suggest that younger American consumers are borrowing more heavily and repaying at lower rates on credit cards than earlier generations. If the current borrowing and repayment habits persist, a substantial buildup of credit card debt at a later period in life may jeopardize the financial well-being of the elderly and cause instability in the credit card market. Therefore, new policies are needed to regulate the credit card market and guide consumer behavior.
The paper proceeds as follows. The next section briefly reviews relevant literature on credit card debt and lifecycle modeling. Section III discusses the data and descriptive statistics. Section IV proposes the econometric model and estimation procedure. Section V presents the empirical results. Section VI concludes with a discussion of future research directions.
Download
PDF Ebook A Synthetic Cohort Analysis of Credit Card Debt and Payoff Rates
