Understanding how the tax rate elasticity varies across the distribution is crucial for determining optimal taxes. In Mexico City, a 1% increase in the property tax liability is associated with a 7 times larger reduction in compliance amongst the wealthy than the poor. Unlike similar findings for the elasticity of taxable income, this is not because the rich are able to access certain evasion / avoidance technologies, such as transferring labour to capital income. The results are also not explained by differences in the strength of enforcement across the distribution, or because tax rates are higher at the top. In other words, it seems that the rich are ’fundamentally’ more sensitive to tax rates than the poor. Higher elasticities at the top of the distribution mean that the optimal tax structure will be less progressive.

This paper investigates how wealth taxation affects households' portfolio choices. Leveraging a major wealth tax reform introduced in France in 2017, I estimate the degree of substitution between real estate and financial wealth. To identify causal effects, I use comprehensive administrative-linked income and wealth microdata for France and a difference-in-differences design comparing French residents to non-French residents subject to the wealth tax but not affected by the policy change. Five years after the reform, the stock of real estate held by French taxpayers decreased by an average of 6% This decrease in real estate is driven by investment rather than owner-occupied housing and is mirrored by a surge in dividend income. The reduced-form estimates can be converted into a cross-elasticity of 5: a 1 percentage point increase in the tax rate differential between real estate and financial assets leads to a 5% reallocation of households' housing stock to financial assets.

In Uganda, the revenue authority launched a unit to monitor the tax affairs of the richest, both high net-worth individuals and very important people. We evaluate the impact of such policy on tax filing and payment outcomes of targeted taxpayers, with a standard difference-in-difference framework. We show that the policy has been only partially successful. While the unit increased the probability to file, especially of VIPs, it produced a strategic response by which taxpayers declare less on different margins, with no impacts on tax liability. On tax payments, only a small impact on tax collected is measured, again due to complex offsetting responses across tax heads. Importantly, we also measure the spillover effect on companies associated with the richest – documenting again complex compensating reactions and no meaningful impacts. Lastly, we show that, while deterrence is more effective on HNWIs, taxpayer assistance and public shaming are more relevant for VIPs.