Debates about Wealth Taxes and Strange Numbers

Elizabeth Warren and Bernie Sanders promise that a (likely unconstitutional) wealth tax will allow the United States to pay for all sorts of new and expanded programs. You name it, a wealth tax will pay for it. I’ve written about this several times.

The math of Warren and Sanders does not work, and it is based on misleading (or mistaken) assumptions about how much wealth such a tax might collect.

You absolutely should read Phillip W. Magness’ analysis, Something’s fishy with the Saez-Zucman tax stats. The title truly says it all. The economists cited by Warren played with their data to get the results desired. The evidence?

When Zucman posted the promised data files from SZ-2019 on a new website made specifically for the book, he also initially removed the old online data appendix to the published QJE version of PSZ-2018 from his personal website. He then replaced that file with a “new” version of the PSZ-2018 appendix that conveniently matched the newly released SZ-2019 numbers.

Zucman defended his actions by claiming that the new file constituted an “update” to improve the accuracy of PSZ-2018. Updates to published works, however, normally come from small data refinements. This one involved a major change to the underlying assumptions of the published paper’s methods. Unlike the published version, those changes had not undergone peer review – and likely would not survive it, seeing as the new corporate tax incidence assumptions break sharply from the established literature.

The oddities in Zucman’s behavior only expanded from there, as Kopczuk noticed the newly replaced file for PSZ-2018 lacked any indication that it had been changed from the older published version of the article. Zucman insisted it was only a misunderstanding, and that the old file had simply been moved and relabeled at another place on his website and pointed Kopczuk to that section. Yet something was still off about the data release.

Within minutes, several other economists began reporting that they had noticed the same thing – the old PSZ-2018 files had been replaced without any indication of what happened to the old file. One of them, Jeremy Horpedahl, shared a screen capture of Zucman’s websitefrom about 2 hours earlier showing that the old PSZ-2018 file was clearly missing after being “updated” with the SZ-2019 numbers. Zucman apparently restored the old file to his website sometime after he started to come under criticism for removing it.

What do the wealthiest pay in taxes? That depends on how you calculate wealth and taxes, as Matthew Yglesias notes on Vox.com, though the heading on Vox gives away his conclusion: A striking new fact (or not) that in many ways is beside the point. For Yglesias, the amount paid matters less than the general fact that inequality has increased. By how much? That’s less important than the philosophical position he occupies on the progressive left.

Another great take is offered on the WSJ Opinion website, Just because the New York Times believes a tax claim doesn’t mean readers should, written by James Freeman.

Forbes columnist Elizabeth Bauer describes What Bernie Sanders And Elizabeth Warren Get Wrong About Wealth Taxes. Warren and Sanders propose the highest wealth tax attempted in an OECD nation.

Warren proposes a tax of 2% for net worth over $50 million and 3% for net worth over $1 billion.

Sanders proposes a tax starting at 1% on net worth above $32 million for a married couple, $16 million for singles, increasing to 2% for net worth from $50 million to $250 million, and 4%/5%/6%/7% for incremental brackets up to 8% on wealth over $10 billion.

Plus, most nations with a “wealth” tax end up taxing what American’s consider the upper-middle class. Bauer writes:

And it is true that a small number of other countries do indeed have wealth taxes. But they look nothing like the American version.

Depending on one’s definition, there are as many as six countries in the OECD (that is, the world’s developed/wealthy economies) with wealth taxes. The Tax Foundation summarizes these in a January report:

Belgium taxes stocks at a rate of 0.15%, when they are part of a portfolio worth more than EUR 500,000 ($571,000).

Italy taxes financial assets at 0.2% and real estate at 0.76%, only for Italian taxpayers who own assets/property abroad.

The Netherlands taxes net wealth at rates from 0.61% to 1.61%, excluding the primary residence and “substantial interests in companies.” (This tax is not strictly a wealth tax as such but taxes the “deemed income” from assets, including the “deemed income” from one’s home, based on legislatively-fixed hypothetical annual investment returns.)

Norway taxes net wealth above NOK 1.48 million ($174,400) at a rate of 0.85%.

Spain taxes net worth at rates ranging from 0.2% at EUR 700,000 ($770,000) climbing in level increments to 1.3% at about EUR 2,000,000 (about $2,250,000) and 2.5% at EUR 11.4 million ($12.5 million). However, autonomous communities (Spain’s top-level political division, as “states” in the United States) may establish alternate legislation, and Madrid has eliminated it. In principle, it is also a temporary tax, introduced in 2011 due to Spain’s financial crisis.

Switzerland’s tax varies by canton. In Zurich, the tax begins with a rate of 0.05% for wealth of CHF 77,000 single/CHF 154,000 married (a nearly equal amount in USD), and increasing in steps to 0.3% for wealth above CHF 3,158,000/CHF 3,235,000. In Geneva, the tax begins after a deduction of CHF 82,040 per adult, at the rate of .0175% and climbs in increments up to 0.5625% on wealth above CHF 1,682,068 and 0.585% on wealth above CHF 3,364,137.

In other words, by and large, these taxes fall not just on the super-rich but also on the upper middle class.

Politicians promise everything for nothing. The more likely result will be nothing but at the price of everything. The fact that presidential candidates’ advisers are playing with numbers should cause a lot of concern.