Errors-In-Variables and Estimated Income and Price Elasticities of Charitable Giving‪.‬

Many important tax policy debates hinge on how responsive individuals are to taxes. For example, the efficacy of the U.S. federal income tax deduction for charitable contributions depends, at least in part, on the magnitude of the price elasticity of giving. It is commonly assumed, at least implicitly, that the data used to estimate income and price elasticities provide true measures of their theoretical counterparts. This is an important assumption because even a single badly measured regressor will result in the estimated coefficient of the badly measured regressor being biased toward zero (attenuated) as well as the others though in unknown directions (Levi, 1973). Since some individuals misreport their income for tax purposes, after-tax income may be badly measured when self-reported tax data are used to estimate tax price models. The purpose of this paper is twofold: (1) to investigate the effect of errors-in-variables due to intentional reporting errors for tax purposes on estimated coefficients in tax price models with an application to charitable giving, and (2) to evaluate a potential solution for bias due to errors-in-variables, described in the literature. The effect of taxes on charitable giving is a natural issue to investigate in this context. It is an important economic activity, representing over $100 billion in annual transfers. In addition, charitable giving has been the subject of numerous studies using reported tax data (Randolph, 1995; Bakjia, 1999; Auten et al., 2002).