Essays on behavioral responses to income taxes

This dissertation consists of an introductory chapter and four empirical essays on behavioral responses to income tax incentives. The introductory chapter presents the conceptual background of the welfare analysis of income taxation, highlighting the crucial roles of behavioral elasticities and the structure of the behavioral response. In addition, I present an overview of recent empirical literature on behavioral responses to income taxes. The first essay studies the extent and significance of income-shifting behavior among the owners of privately held corporations in Finland. The dual income tax system in Finland offers considerable incentives and opportunities for income-shifting between wages and dividends for business owners. Extensive panel data at both the owner and firm level together with Finland's dividend tax reform of 2005 enable us to credibly analyze the empirical significance of income-shifting. The results show that incomeshifting is evident. The income-shifting response is homogenous across different owners and firms. However, the size of the income-shifting incentive affects the extent of income-shifting activity, indicating that the costs of income-shifting are non-negligible. The second essay focuses on distinguishing between real responses and incomeshifting. Previous literature shows that income taxes affect the behavior of high-income earners and business owners. However, it is still unclear how much of the response is caused by real economic activity (labor supply, effort etc.), and how much is due to income-shifting and other tax-planning activities. This distinction is important because it can greatly affect the welfare implications. We build a model that formalizes the analysis of the elasticity of taxable income under income-shifting possibilities. We show that income-shifting accounts for over two thirds of the overall behavioral response among the owners of privately held corporations in Finland. This considerable incomeshifting response halves the marginal excess burden of income taxes among Finnish business owners, compared to the traditional framework where the overall behavioral response determines the welfare loss. The third essay studies the elasticity of taxable income (ETI) in Finland. To identify the average ETI, I use changes in flat municipal income tax rates as an instrument for changes in the overall marginal tax rates. This instrument is not a function of individual income, which is the basis for an exogenous tax rate instrument. In general, instruments based on predicted tax rate changes as used in many previous studies do not have this feature. In addition, I estimate behavioral responses for the subcomponents of taxable income, such as working hours, fringe benefits and tax deductions. This structure of the overall ETI has rarely been studied in previous literature. My preferred estimate of ETI in Finland is 0.27, which indicates moderate but non-negligible welfare effects. The subcomponent analysis suggests that working hours and wage rates respond less than tax deductions and irregular forms of compensation such as fringe benefits. This implies that the overall behavioral response is not driven by changes in deep individual utility arguments, such as the opportunity cost of working. The fourth essay analyzes the role of different frictions in explaining different behavioral responses to similar tax incentives. Recent literature shows that optimization frictions attenuate behavioral responses to tax incentives, compared to the underlying long-run structural elasticity. However, the literature has not addressed the role of different frictions in explaining heterogeneous responses to tax incentives. To study this issue, we analyze to what extent and in what manner taxpayers respond to different tax incentives. We compare the behavioral effects induced by different tax and transfer incentives and institutions within similar or even the same individuals in Finland. We use a local estimation approach and the bunching method in order to produce clear and comparable evidence. We find that taxpayers do not respond at all to small changes in tax incentives, but do respond to large and salient incentives. Moreover, the patterns of response indicate that some taxpayers are unable to respond even to large incentives, and some are unaware of the tax incentives.