Taxation of wealth, in its broadest sense, has two main objectives: to generate tax revenues, and to achieve a more equal distribution of wealth.
Until recently, wealth taxes, however, have not been popular among policy makers. Many countries have abandoned wealth taxes in favor of capital income taxes. However, a major drawback is that capital income taxes generally only tax income when it is realized and assets which do not generate taxable capital income remain untaxed. Recently, the interest in wealth taxes has become larger, as many countries face increases in the concentration of capital and face new budgetary pressures from the COVID-19 crisis. A main argument against the wealth tax is that mobility of taxpayers might undermine the objective of wealth taxation in the presence of no-tax countries, making it important to think about policy proposals to harmonize taxation across countries. However, much debate remains over what is the appropriate means of tax coordination.
This debate relates to the reform of the Spanish tax system, under which wealth taxes are currently fiscally decentralized to the regions (Comunidades Autónomas). Prior to 2008, Spain had a mostly uniform wealth tax, which was briefly suppressed. After its reintroduction in 2011, regions have exercised their autonomy to change wealth tax schedules and by now Madrid has become the only region characterized by a zero effective tax rate on wealth. Thus, a millionaire can save thousands of Euros annually by moving from Catalonia to Madrid.
In a recent study focusing on the Spanish setting, we analyze the extent to which mobility may undermine decentralized wealth taxation. Furthermore, we characterize how effective tax coordination can reduce the inefficiencies and inequities created from decentralized taxes on capital.
We exploit two different data sets for our analysis. We make use of the Panel de declarantes de IRPF 1999-2015, which contains information about labor and capital income as well as the fiscal residence of taxpayers. We merge these data with the Panel de declarantes de IP 2002-2007. Both data sets are provided jointly by the Institute for Fiscal Studies (Instituto de Estudios Fiscales) and the National Tax Agency (Agencia Estatal de Administración Tributaria). Based on these data, we first analyze the location choices of wealthy taxpayers by comparing changes of the fiscal residence of individuals likely subject to the wealth tax after decentralization. As we can follow taxpayers over time between 2005-2015, we implement a difference-in-difference design to show that the stock of wealthy individuals living in Madrid has increased following Madrid’s zero-rating the wealth tax after 2011.
Figure 1. Mobility Effect towards Madrid
Figure 1 shows the main results of our analysis. The stock of the wealthy in Madrid who would have paid wealth taxes if living in any other region (those with a level of wealth above the 700.000 euros exemption threshold) increased by 6000 people between 2011 and 2015. At the same time, the other Autonomous Communities lose on average 375 taxpayers each. With the help of various empirical models our study confirms that by five years after the reform, the stock of wealthy individuals in Madrid increases by 9% and falls by 2.5% elsewhere. It is also interesting to have a look at the changes in the stock of the wealthy before the wealth tax differential appeared. The figure shows that the years preceding the tax reform had no noticeably different trends in Madrid and the other regions, validating our research design.
After documenting the mobility effect, our study continues to analyze the impact of tax competition and the resulting mobility on regional tax collection and inequality. It is important to keep in mind that only the richest 0.5% of the population, which holds an estimated 19% of the country’s total net wealth, benefits from avoiding wealth taxes. Lost tax revenue to a region may result in the loss of valuable public services to others.
We perform several simulations to quantify the impact of mobility resulting from tax competition – and policies that would reduce it – on tax revenues. We first analyze the revenue effects of shutting down tax-induced mobility through perfect enforcement and monitoring of fiscal residency. We find that Spain foregoes 5% of total wealth tax revenue due to mobility, with some heterogeneity across regions.
A remaining question is what drives Madrid’s decision to set a zero-tax rate. An often-cited argument of politicians in favor of the zero-tax regime in Madrid is that tax competition creates a more favorable business environment and attracts productive entrepreneurs. However, the average wealth taxpayer is already 65 years old, and most of them are rentiers, that is, people who receive most income in the form of capital income instead of labor income. Furthermore, our results suggest that a share of those moves might rather be fictitious.
Our work also highlights the importance of wealth tax mobility on tax revenues from other sources. Even though rich individuals do not bring wealth tax revenues to a zero-tax jurisdiction, they pay income taxes. This benefits Madrid, given that half of personal income tax revenues can be kept by the regional governments. In our simulations, Madrid benefits from higher income tax revenues due to wealth-tax induced mobility. While the effect on revenue is positive for Madrid, any income tax revenue raised by Madrid from movers is lost by the remaining regions.
Inequality is another dimension affected by decentralized wealth taxation. Figure 2 shows the wealth concentration of the richest 1% of individuals between the scenario with and without mobility. Between 2010 and 2015, the top 1% wealth share was growing almost twice as fast in the presence of mobility (16%) relative to the perfect enforcement scenario (8.7%).
While an aim of our study is to document the causal effect of a zero-tax regime on mobility and, consequently, revenues and inequality, we provide guidance on policy solutions to eliminate the inefficiencies. In particular, our work is helpful to think about how tax coordination might mitigate the inefficiencies we document. We analyze two alternatives: minimum tax rates and harmonization. Using our causally identified mobility responses, we simulate the evolution of revenue (both for wealth taxes and personal income taxes on labor and capital) under a harmonized wealth tax system, which eliminates tax-induced mobility but that also mechanically changes tax revenue by forcibly changing regional tax rates to a common level given by the national default. We show that harmonization to a weighted average of the decentralized tax rates is only a revenue improvement for all regions if the harmonized rate is relatively high, i.e., close to the maximum rate set by Extremadura. As an alternative, we show that a minimum tax rate at the same level could be politically more feasible, as we show that it allows for some diversity in tax schedules across regions and allows everyone to increase revenue.
Figure 2. Regional Wealth Concentration (2005-2015)