What Happens When Dying Gets Cheaper? Behavioural Responses to Inheritance Taxation [Job Market Paper]

This paper identifies behavioural responses to significant cuts in the inheritance tax, focusing both on the apportionment of estates and on the reporting and assessment of inherited assets. Using the universe of inheritance tax returns of Catalan tax residents from 2008 to 2015, it exploits an age-related tax deduction and the quasi-repeal of the inheritance tax for bequests given to close relatives. The paper provides two main findings. First, spouses are more likely to inherit the entire estate when there is no need to minimize tax payments. Second, reported estates increase up to 40% due to a 99% tax cut. This response is mainly driven by changes in heirs' reporting behaviour. In particular, it is primarily explained by real estate "over-assessment" and, to a lesser extent, by the reporting of assets that otherwise would have been evaded, such as cash, antiques, etc. Although the first practice is not related to inheritance tax evasion, it helps to reduce capital gains in the case of a potential sale, and hence it could imply the evasion of future personal income taxes.

Winner of the Alexandre Pedrós Prize. Media coverage: La Vanguardia and El Punt Avui.

You can find the latest version here.

Avoidance Responses to the Wealth Tax

(with José Maria Durán-Cabré and Alejandro Esteller-Moré)

Former title: "Behavioural Responses to the (Re)Introduction of Wealth Taxes. Evidence from Spain".

In the throes of economic crisis, the Spanish government decided to reintroduce the Wealth Tax, appealing to redistributive motives and its need for greater revenues. This paper studies how individuals reacted to the reintroduction of this tax by drawing on the universe of wealth tax returns submitted to the Catalan Tax Agency between 2011 and 2015. Thus, we exploit the variation in treatment exposure to analyse taxpayers' responses, in terms not only of wealth accumulation, but also of the potential avoidance strategies adopted. Indeed, our results reflect avoidance rather than real responses. They show that while facing higher wealth taxes did not have a negative effect on taxpayers' savings, it did encourage them to change their asset and income composition to take advantage of wealth tax exemptions (mostly business-related) and the existence of a limit on wealth tax liability. This translates into an elasticity of taxable wealth with respect to the net-of-tax rate of return of 0.64, or, put differently, a 0.1 percentage point increase in the average wealth tax rate leads to a reduction in taxable wealth of 3.24% over 4 years. Overall, these avoidance responses are quite marked in terms of tax revenues: they represent a 4-year accumulated revenue loss of 2.6 times the 2011 estimated wealth tax revenues. The existence of such responses has relevant policy implications concerning the design of a wealth tax.

You can find the version published as IEB working paper here.

Detecting Tax Evasion Through Wealth Tax Returns

(with Daniel Mas Montserrat)

In the context of a tax amnesty carried out by the Spanish government in 2012, this paper quantifies the wealth voluntarily disclosed and shows how this type of evasion was distributed across wealth levels. In line with findings from other countries, the data indicates that the probability of voluntarily disclosing hidden assets increases significantly with wealth and that, on average, wealth disclosers were evading around 30% of their net worth. This paper also shows that machine learning methods can be useful tools for governments and tax administrations to detect tax evasion. In particular, it studies whether wealth evaders can be detected with the information they initially report in wealth tax returns - i.e. when evasion is still not disclosed -. We frame tax evasion detection as a binary classification problem and train and evaluate multiple classifiers commonly used in supervised machine learning methods. The main conclusion from this exercise is that the relatively little information available from tax returns, which mostly relates to wealth composition and income levels, it already allows distinguishing evaders from (presumably) non-evaders.

Draft available upon request.