Ability to borrow, willingness to borrow, and household debt

The purpose of this study is to investigate consumers\u27 borrowing behavior by incorporating both the ability to borrow and the willingness to borrow. The sample for this study is drawn from the 1989 Survey of Consumer Finances (SCF). Since households\u27 desired debt level is obtained if, and only if, households are able and willing to borrow, a selection situation arises in this study. The selection methods described by Heckman (1979) and Tunali (1986) are applied to avoid inconsistent estimates of population parameters due to the correlation between disturbances of the sample selection equations and the debt equation. The findings show that approximately 19% of American households are not able to borrow and 30% of American households are not willing to borrow. Only 55% of households are both able and willing to borrow. About 26% of U.S. households are able but not willing to borrow, 15% are willing but not able to borrow, and 4% are not able and not willing to borrow. The correlation coefficient between the error terms in the equation of the ability to borrow and the equation of willingness to borrow is very small and not significant. Predicted debt is 121% and 67% higher than observed debt for credit constrained households and households who are not willing to borrow respectively. The impact of the ability to borrow is larger than that of the willingness to borrow on household debt. Desired debt is about two times higher than actual debt for households who are able but not willing to borrow. Households who are willing but not able to borrow would increase their debt level by 122% if credit constraints are removed. Desired debt is 134% more than actual debt for households who are not able and not willing to borrow. These findings suggest that researchers should not combine households who are willing to borrow with households who are not willing to borrow. It is important to incorporate a household\u27s ability to borrow and willingness to borrow simultaneously to understand household borrowing behavior