Essays on Behavioral Responses to Corporate and Personal Income Taxation

Abstract

This dissertation is a collection of three essays that investigate research questions related to corporate income taxation (Chapters 1 and 2) and personal income taxation (Chapter 3).

Chapter 1 develops a new approach to calculate country-industry-year-specific forward-looking effective tax rates (FLETRs) that also account for typical country-industry-specific financing structures as well as asset compositions. Using these FLETRs, we estimate the tax semi-elasticity of corporate investment in tangible fixed assets. Our results suggest a statistically significant tax semi-elasticity of -0.41, which is at the lower end of previous findings. We further show that different subgroups of firms respond very heterogeneously to tax incentives.

In Chapter 2, we show that effective corporate income taxes are lower in EU NUTS 2 regions where citizens perceive corruption to be comparatively more prevalent due to tax evasion. We develop a new approach for calculating region-industry-year-specific empirical effective income tax rates (EEITRs) using firm-entity-level income statement data. Controlling for proxies for deductions that could legally be claimed, our benchmark model suggests that a one standard deviation increase in corruption leads to a both statistically and economically significant decrease in EEITRs of approximately 0.4 percentage points.

Chapter 3 conducts a social welfare analysis of the negative income tax (NIT), explicitly allowing for nonlinear labor supply responses with respect to the take-back rate. We derive a theoretical model which yields a notion of social welfare as function of the take-back rate that we calibrate using data from two NIT experiments that were conducted in the US in the 1970s. We find (i) that both the theoretical model and the empirical estimates suggest that the labor supply with respect to the take-back rate is nonlinear; and (ii) that the social welfare optimizing take-back rates strongly differ between models calibrated with nonlinear versus linear labor supply functions.