Mobility responses to the establishment of a residential tax haven: Evidence from Switzerland

Abstract

I analyze mobility responses to a tax reform that established the Swiss canton of Obwalden as a tax haven in 2006. The reform, which included a regressive income tax schedule, was explicitly aimed at attracting the top 1%. Difference-in-Differences (DiD) estimations comparing Obwalden to all other cantons confirm that the reform successfully attracted high-income taxpayers: by 2016, the share of top earners in the canton had doubled, and average income per taxpayer was 16% higher relative to 2005. Based on individual tax return data, I estimate the mobility elasticity with a two-stage least square (2SLS) approach, which isolates the identifying variation in the tax rate stemming from the 2006 reform only. I find a large elasticity of the stock of high-income taxpayers of 1.5–2 with respect to the net-of-average-tax rate. The corresponding flow elasticity is 7.2. Despite these large behavioral responses, the reform did not increase revenue per capita in the canton. Finally, I find small positive effects on local employment. However, in-movers with high incomes were not more likely to also work in the canton, and I cannot rule out that employment effects were driven by the simultaneous reduction in corporate income taxes.