Tax reform and the demand for debt

This article examines the impact on household borrowing from a tax reform that implied a change from progressive to constant marginal tax rates on interest deductions. We use three alternative empirical methods to address the problem of endogenous explanatory variables, which arises when the marginal tax rate itself is a function of the amount of debt: The first approach uses instrumental variables to examine the significance of the marginal tax rate directly and produces mixed results. In the second approach we use tobit regressions based on repeated cross-sectional data to examine the correspondence between debt and income and find that the households' amount of debt were strongly and positively related to income before the tax reform and significantly more income-dependent under progressive tax rates than under the constant tax rate. Finally, using panel data covering the last year prior to the reform and the first year after, we find that the change in a household's marginal tax rate had a significant impact on the change in its amount of debt.