We estimate the effects of income from various sources on charitable giving using administrative German income tax data. We demonstrate that charitable contributions are not uniformly affected by different income types. While business and capital income exhibit a positive effect, the remaining income sources do not influence charity on statistically signifcant levels. This exercise is not new and has been conducted for (at least) three different purposes: 1) Relying on the described results, a public finance researcher would state that business and capital income are more prone to tax evasion than the remaining income sources. 2) An entrepreneurship researcher would conclude that business owners are more generous than employees, and 3) a researcher testing the validity of the life cycle theory (or its behavioral counterpart) would refute the fungibility of income. In contrast, we argue that none of these approaches can answer the intended question if solicitation effects of fundraising or measurement error of the income sources are not taken into account. Applying a fixed effect poisson model, we demonstrate that under certain assumptions the results can have a meaningful interpretation.
Bittschi, Benjamin, Sarah Borgloh and Marc-Daniel Moessinger (2016), On Tax Evasion, Entrepreneurial Generosity and Fungible Assets, ZEW Discussion Paper No. 16-024, Mannheim.