Behavioural answers to top-financial wealth tax: evidence from Belgium

We study the effect of a wealth tax on financial assets of the top 5% wealthiest individuals in Belgium between 2018-2019 – imposing a 0.15% tax rate above 500,000€. Using a unique individual banking data, this paper provides the first scientific evidence on the behavioural answers to specific assets. To do so, we employ three different bunching designs; two static approaches- parametric and non-parametric- and a dynamic bunching model. Our first findings, based on the two static approaches, reveals that taxpayers significantly react to the tax on securities in 2019, but not in 2018 (implementation year). We estimate an elasticity between 8.3 and 13.23, meaning that a 1% increase in the net-of-tax rate leads to a decrease between 8 and 13% in taxable securities in 2019. Taken together, these results seem to indicate a significant adjustment in taxpayers’ taxable securities. However, the dynamic approach - which isolates the response of individuals directly reacting to the threshold - does not reveal significant changes in taxable securities for most treated groups. One exception concerns taxpayers with taxable securities between 520,000€ and 530,000€; in this range, 4.9% exhibit bunching behaviour in 2019, with an average growth rate of 1.7 percentage points lower than the average growth rate in the control group. Nevertheless, additional tests are still required to confirm the conclusions drawn from the dynamic approach. The results of this study aim to contribute to the debate on the legitimacy of wealth taxes, as strong behavioural responses may erode fiscal revenues, thereby undermining the tax’s efficiency.