The fiscal architecture of a country plays a crucial role in how, where and to which extent economic activity develops. When taxing and spending powers are decentralized to sub-central levels of government, one can refer to this task delegation as "fiscal federalism ". The increasing complexity of governmental intervention, the intricate distribution of fiscal powers among levels of government and thé increased exposure to household and business mobility makes this topic a highly relevant field of study, parts of which are still not well understood.
Through decentralizàtion policies, citizens' and firms' preferences might be more closely repre­ sented in the composition of spending or the level.of taxation. From the perspective of households or businesses, this means that choosing a certain jurisdiction rather than another to settle in or build a plant becomes important in terms of fiscal conditions. From the point of view of governments, this fact is known and itself internalized when making fiscal policy decisions. In fact, few tax policies are discussed without considering the possibility that a tax eut might attract investment, or that a tax hike might discourage the creation of new firms. Governmental decision makers, who hence perceive the mobile nature of their funding sources, might as a result engage in fiscal competition. However, sub-central governments do not decide on policy in a legal and institutional vacuum. D ­ centralization in practice also cornes with standards, norms, cost sharing, resources redistribution and other mechanisms implemented to diminish spatial heterogeneity in fiscal conditions. Fiscal instruments at the different government levels within fiscal federations can therefore be influenced by neighboring jurisdictions' ones as well as by upper-level policies and constraints.
This thesis examines some of the economic repercussions of fiscal federalism, bath on sub-central governments' policies and on citizens' decisions. In particlilar, interest lies in understanding how inter-jurisdictional redistribution policies affect local taxes, how delegation of fiscal authority is etfectively implemented, and how jurisdictional fragmentation affects the spatial distribution of incarne in urban areas. The different chapters overall seek to contribute to the body of i"esearch studying how taxes affect the spatial allocation of fiscal resources (e.g. Briilhart et al., 2021; Basten et al., 2017; Schmidheiny, 2006 for households or Giroud and Rauh, 2019; Brülhart et al., 2012 for firms) and to the literature investigating how the mobility of fiscal resources influences govern­ ments' tax-setting behavior (e.g. Parchet, 2019; Eugster and Parchet, 2019; Lyytikainen, 2012). The thesis provides theoretical and empirical insights on questions that help understand how taxes, decentralization and tax-base mobility internet in modern fiscal federations.
In the first chapter, I study the incentive effects created by fiscal equalization on local taxes. I propose two refinements to the empirical estimation of these incentive effects. First, I show that for measuring the extent of equalization it can be important to consider the implications of discrete changes in jurisdictions' tax bases instead of focusing on marginal changes as clone in the literature to date. Given that equalization schedules typically feature non-linearities, focusing on equalization rates with respect to marginal changes can conceal the effect of thresholds in equalization sched­ ules that only enter the analysis when discrete changes are considered. I refer to equalization rates for discrete changes in tax bases as "supramarginal" equalization rates. Second, a policy-relevant measure of equalization rates should not only consider statutory equalization schedules but ought also to condition on jurisdictions' own tax rates. I refer to this as an "effective" equalization rate: frorri the point of view of the local policy maker, any change in transfer amounts through the equalization scheme should be compared to the change in own tax revenue of the jurisdiction.
To identify and quantify the incentive effects of fiscal equalization, I examine a reform of the inter-municipal capacity equalization scheme of the canton Bern, Switzerland. The legislative revision created a change in the statutory equalization rate for some municipalities but not for the others. This naturally delimits a treated and control group which allows me to implement a difference­ in-difference strategy. I first use the reform to estimate how a change in equalization rate impacts local taxes. In a second step, I quantify the effect of marginal/supramarginal, nominal/effective equalization rates on municipal tax rates.
My empirical results show large responses of tax rates to changes in marginal equalization rates (0.08 percentage points for a 1 percentage-point increase in the equalization rate) and even larger responses to changes in supramarginal equalization rates (0.28 percentage points for a 1 percentage­ point increase in the equalization rate). I howèver do not find statistically significant responses to effective equalization rates. Compared to past findings, my estimates on marginal equalization rates are among the largest and the coefficients on supramarginal rates are one order of magnitude higher than those estimated in the most recent literature.
The second chapter, joint with my fellow PhD student Jeremy Zuchuat, examines the sensitivity of individual tax bases with respect to local taxes. More specifically, we study how decentralized progressive income taxation impacts the location choice of individuals relocating in urban areas. We are interested in examining how individuals of different income levels trade-off taxes, housing prices and commuting, and whether taxes affect directly the willingness to commute of individuals. We develop a structural mode!of individual location choices in a fiscally decentralized monocentric city with progressive income taxes. We explicitly include commuting costs, which acknowledge the fact that, within urban areas, a location decision most likely translates into commuting. Our model shows that high-income earners are less sensitive to housing prices compared to taxes and commuting. They are predicted to locate in jurisdictions with higher housing prices. Also, high­ income individuals are more sensitive to taxes relative to housing prices and commuting if the progressivity of the tax schedule is sufficiently high. Whether high incomes value centrality more than lower incomes is however ambiguous. This is explained by the fact that the progressivity of the income tax schedule impacts the opportunity cost of lime of higher-income households to a greater extent. In turn, those might be more willing to commute if they are compensated with lower taxes.
Based on a random utility framework, we estimate how the indirect utility of individuals of different income levels is differently affected by municipal characteristics. For our empirical analysis, we rely on micro-level data on moving decisions within Switzerland, which we combine with de­ tailed records of income, a measure of the comprehensive tax burden and other individual-level socio-demographic information. We additionally have access to a municipality-level measure of housing price expensiveness as well as other jurisdictional characteristics.
Our estimation results show that high-1ncome agents are effectively more sensitive to taxes than low incomes. We also observe that high-income earners tend to sort themselves into municipalities where housing prices are high. Finally, centrality is valued less strongly by higher incomes: our estimates suggest that well-off individuals tend to be more willing to commute. We addition­ ally internet distance with taxes and income and show that lower taxation levels directly increase individuals' willingness to commute. This is consistent with the intuition that local taxes affect the opportunity cost of commuting. Finally, we conduct an illustrative counterfactual exercise where we impose a homogeneous local income tax rate within the urban areas of study. We show that shutting off fiscal decentralization leads to an increase in the demand for central municipalities from high-income movers. Shutting down fiscal federalism would thus entail a gentrification of city centers and a decline in the income levels of peripheral jurisdictions.
In the final chapter of this thesis I propose an empirical method for inferring the effective level
of fiscal autonomy of local jurisdictions. Based on a tax competition model where the jurisdictions' budget is partly determined exogenously by the upper-level government and partly chosen freely by the local government, I derive an empirically testable hypothesis involving endogenous tax rate responses to an exogenous revenue shock. More specifically, looking at how the slope of tax responses to the revenue shock vary with perceived tax-base mobility allows me to distinguish between 'fiscally constrained' and 'fiscally autonomous' behavior. The main prediction from the mode} is that tax decreases resulting from a positive revenue shock are stronger with higher per­ ceived tax-base mobility if the local decision maker is spending-constrained, and vice-versa.
To empirically evaluate the effective fiscal autonomy of local governments based on my theoretical findings, two sources of variation are needed. The first is an exogenous revenue shock and the second is a measure of perceived tax-base mobility. In this research I use a reform of the inter­municipal equalization system of the Swiss canton of Fribourg to identify a budget shock that is.exogenous from the point of view of local decision makers. The reform created a permanent dis­ crepancy between total transfers based on the pre-reform scheme and the total equalization grants actually received after the reform, which I refer to as 'transfer windfall'. I build measures of ex­ posure to mobility of the two main tax bases of the municipalities: firms and households. For this I compute the share of available zoned land affected to residential and industrial use. Reserves in either land-use zones mean that the local jurisdiction can welcome new residents or a new fîrm, which makes tax-base mobility more salient to the decision maker when considering fiscal adjust­ ments.
In my baseline econometric specitication I use the transfer windfall variable as a continuous treatment which I internet with my measures of perceived tax-base mobility to estimate average causal responses of persona! and corporate tax rates, and how these differ between niunicipalities with high or low residential and industrial land reserves.
My main empirical finding is that jurisdictions with high residential land reserves decrease their local taxes on persona! and corporate tax bases more strongly than those with low reserves. I additionally examine the spending side and observe no capital or current expenditure responses to the transfer windfall. Based on the main proposition of my theoretical framework, these empirical re­ sults suggest that local governments are effectively highly constrained in their budgetary decisions. These findings therefore contrast the conclusion one could draw when considering Swiss munici­ palities based ôn the usual fiscal autonomy indicators. In fact, according to traditional 'budget-data' indexes, local jurisdictions of Switzerland and Fribourg would rank among the most autonomous in terms of revenue raising and spending compared to 14 other high-income countries.