Abstract
Due to increasing inequality and government debt, taxation is at the forefront of both academic and policy debates. With this in mind, the criteria economists use to evaluate tax policies are crucial. Typically, the criterion has been utilitarianism. However, utilitarianism has problematic tax policy implications. This dissertation considers other fairness views, such as horizontal equity, equal sacrifice and intergenerational mobility. The main contribution is therefore to expand the fairness views considered in designing tax policy.
The first chapter shows how to reveal inequality aversion from observed tax policy when governments restrict the information they exploit. Governments have increasing access to information about individuals, but they exploit little of it in setting taxes. I build on the inverse optimal tax problem, which considers the observed tax system, combined with information about the income distribution and how individuals react to taxes, to reveal the weight society assigns to each individual’s consumption. The first contribution is to map these weights into the concerns for vertical and horizontal equity. While vertical equity provides the standard inequality aversion rationale for redistributive taxation, horizontal equity introduces a restriction against tax discrimination. Here, I develop a theory and optimal tax algorithm to reveal the priority on each concern. The second contribution is to apply the model to gender-specific taxation in Norway. My result is that inequality aversion is overestimated if horizontal equity is ignored.
The second chapter (joint with Paolo G. Piacquadio) proposes and axiomatically characterize a family of welfare criteria that prioritize individuals making larger sacrifices. By combining efficiency with a concern for equality of sacrifice, our criteria avoid serious shortcomings of utilitarianism. We illustrate our results within the Mirrleesian optimal taxation framework. Our simulated equal-sacrifice optimal tax schedule rationalizes the Californian tax system at higher incomes when sacrifice is proportional, and suggests marginal tax rates that are about 20 percentage points lower than the utilitarian recommendation.
In the third chapter (joint with Shafik Hebous), we ask the questions: Does parental wealth inequality impact next generation labour income inequality, and if so, does a tax on parental wealth affect the labour income distribution of the next generation? We tackle both questions empirically using detailed intergenerational data from Norway, focusing on effects on wages rather than capital income. Our results suggest that a net wealth of NOK 1 Million increases the yearly wages of the children by NOK 14,000. Children of wealthy parents also have a higher labour income mobility. The estimated hypothetical wage distribution without the wealth tax is more unequal. In addition, suggestive evidence indicates parental wealth is associated with higher labour risk taking.
In the fourth chapter (joint with Thor O. Thoresen), we study the elasticity of taxable income and the self-employed. The elasticity of taxable income summarizes the welfare effects of behavioural responses to taxes. We show how behavioural responses may create bias in the measurement of the elasticity of taxable income. Using register data on the self-employed in Norway, we find that their taxable income response stems from responses in working hours, tax evasion and organizational shifts. In particular, organizational shifts create a selection problem and, in this chapter, we show how to correct the measurement bias due to this response.