Thread by @D_Burgherr

Finally caught up with the wealth tax literature. A great starting point is this recent @voxeu column by @Marius_Brulhart @jonathangruber1 @MatthiasKrapf @unibas_metrics.

Trying to contribute to the public good that is #econtwitter, a short thread. 1/10

voxeu.org/article/wealth…

Column is based on their recently updated working paper in which they exploit wealth tax reforms in Switzerland (OECD country with highest share of wealth tax revenue) to estimate semi-elasticity of taxable wealth. 2/10

Ungated version: hec.unil.ch/mbrulhar/paper…

Using a variety of subnational wealth tax changes, they estimate that a 1 p.p. drop in wealth tax rate raises reported wealth by 43%! Authors do offer a story why their semi-elasticity is higher than other estimates in literature, which I will come back to in a minute. 3/10

In addition, they decompose the change in reported wealth using micro-level tax records from 2 cantons. Their findings suggest that 24% of reaction is attributable to mobility, 20% to capitalization into housing prices, and 6% to increased savings (incl. mechanical effect). 4/10

This leaves an unexplained residual of about 50% that is attributable to changes in wealth reported by resident taxpayers! Since there is no third-party reporting in Switzerland, these changes are likely linked to avoidance and evasion behavior. 5/10

The authors provide an overview of semi-elasticity estimates in the literature (incl. work by @SeimSeim @juliana_londono @katr_jakobsen @aestellermore) and argue that their results are in line with representative estimate of 8% assumed by Saez and @gabriel_zucman. 6/10

Two main reasons:
1. Cantons are small what enables high taxpayer mobility. Adding up direct and indirect (capitalization) mobility effects (and subtracting international mobility which accounts for one quarter) suggests 34% of response is due to intra-national mobility. 7/10

2. Tax enforcement is weak as wealth is self-reported (remember, 50% of changes are linked to underreporting).

Together, these “Swiss peculiarities” account for 85% of aggregate response. If you mentally adjust their estimate accordingly, it certainly fits in with the lit. 8/10

Note that mentioned 8% are derived as avg. of 3 bunching estimates (Seim; Londoño-Velez/Ávila-Mahecha; Jakobsen et al.) – which are much lower than longer-run diff-in-diff estimates – and Brülhart et al. Comparison might not be too meaningful anyway. 9/10

gabriel-zucman.eu/files/saez-zuc…

Bottom line: Tax enforcement is crucial. Caveats: In Switzerland, avoidance might have acted as substitute for real responses. Additionally, the analyzed wealth tax rate changes are rather small and nowhere near the levels proposed in the US. 10/10

If you prefer listening to reading, there is a short podcast @voxeu in which @Marius_Brulhart explains their paper. I really appreciate the authors’ efforts to make their research more accessible to policy makers and the public.

voxeu.org/vox-talks/weal…

For more institutional context, check out these excellent pieces on the Swiss wealth tax that @MESandbu wrote for the @FT in early 2019.

ft.com/content/c84372…
ft.com/content/87ccaf…
ft.com/content/01e3e3…