Washington State has proposed a tax on billionaires equal to 1% of their non-tangible financial assets. Most of the revenue generated from this tax would come from four individuals that live in Washington, Jeff Bezos, Bill Gates, MacKenzie Scott and Steve Ballmer. So, what if these folks just leave Washington?
One CNBC article I was reading cited the Washington State representative, Noel Frame, as stating that “The idea that they will just pick up and leave is a cynical view and it’s not supported by evidence.”* The evidence provided in the news article is from a research paper from the American Sociological Review about income taxes on high earners. Is that the only paper that looks at this issue, and, what does that paper actually find? What is the actual “evidence”?
I have increasingly seen people’s use of the term “evidence”, or “science”, to back up their arguments, when often the evidence or science on any complex issue is far from clear. Complex issues are, well, complex, and in many cases, those bandying “evidence” or “science” have little of either.
So, what does the evidence actually say? First, despite what Ms. Frame states, that particular paper does find evidence of millionaire tax flight. However, it’s not a lot of flight. Like, instead of seeing a whole flock of millionaires fleeing a state, you might see one fly away. As the paper notes, “We find that millionaire tax flight is occurring, but only at the margins of statistical and socioeconomic significance.” So, if the flight is small, why should we care? Which people do we expect to be most affected by such taxes? My guess is those with the weakest ties to the state (say, those that are retired, or soon to be retiring), and who will take the largest financial hit? Maybe. It seems like if we are imposing a tax that will generate nearly all of its revenue from four people, losing even one of them would make the revenue take very different. But what does the “evidence” actually say about who may fly?
There are more papers that just this one. Two of the authors of that paper, Professors Young and Varner, have other papers that also look at this issue. They likewise find little (but not no) responsiveness to a millionaire tax in New Jersey. What little effect there is is concentrated among “people living on investment income rather than wages, and people who work (and pay tax) entirely in-state.” Interestingly, in an extremely rare situation, this particular study was replicated and its findings are questioned in this published paper, which suggests much higher responsiveness, to which the original authors have responded. Studying California, these same two authors also find limited evidence of outward migration of millionaires, but, interestingly, do find an effect among “high-income earners who become divorced”. Interestingly, 2/4 of the four billionaires who would provide most of the revenue are recently divorced (from each other!).
There are many other papers about income taxes and mobility, as summarized by this excellent review article by Scheuer and Slemrod, and, this excellent review article. The general summary, by Scheuer and Slemrod, is that “Studies of the European wealth taxes often, but not always, find a substantial behavioral response, although the nature of the response varies. We emphasize that any lessons drawn from the European experience must be applied to the recent US proposals with substantial caution, because the design features of recent proposals—rate schedule, broadness of the base, and enforcement provisions—are very different from any previous wealth tax.”
Some examples of these studies: Agrawal and Foremny find that a Spanish tax reform did cause “significant” movement. A paper by Martinez finds that a Swiss Canton that cut its tax rate experienced significant inward mobility of high-income taxpayers. Likewise, another paper by Schmidheiny and Slotwinski find a responsiveness to taxes in Switzerland, especially for households without children. Other papers look at specific types of high-income folks. Akcigit, Baslandze, and Stantcheva find evidence that scientists are sensitive to taxes, as do Moretti and Wilson. Likewise, soccer players are sensitive to taxes.
All of that evidence is interesting, but, to some extent, it is irrelevant. It is about an income tax, and, the proposed tax is a wealth tax. People are more sensitive to wealth taxes than income taxes. Especially for the four folks in Washington, income taxes and wealth taxes would have dramatically different effects. While Bezos, for example, on any given day may qualify as the wealthiest individual on earth, he is probably not even in the top 10,000, or way more, individuals for the highest income, as far as taxable income goes. Almost all of the Bezos wealth is from unrealized capital gains, which are, so far, not taxable. While Jeff Bezos owns in excess of $170 billion in wealth, his total compensation from Amazon amounts to less than $2 million. While he may occasionally sell shares to finance his consumption or have other sources of income, my guess is that the wealth tax of, say, $1.7 billion would be orders of magnitude more than any income tax he currently pays on a regular basis.
So, what does the literature tells us about migration due to wealth taxes?
There are no examples of true wealth taxes in the U.S. The estate tax is similar, but, is imposed only once, after you are dead. Bakija and Slemrod find some, but very modest, propensity to move as a result of state-level estate taxes. Using a change in Denmark, Jakobsen, Jakobsen, Kleven and Zucman find significant effects on wealth accumulation as a result of wealth taxes. The authors study only the wealth that stayed in Denmark, so, suggest that their finding less wealth as a result of the wealth tax may stem from wealth fleeing (likely the capital itself, and not the actual families, although they do not rigorously test this). Agrawal, Foremny, and Martinez-Toledano find significant evidence of people moving as a result of a Spanish wealth tax.
So, what does the “evidence” actually say? Every paper finds some responsiveness to both wealth and income taxes. It’s often not large, especially for income taxes. The responsiveness to wealth taxes is larger than income taxes. The kinds of people who respond to these taxes are households without children, recently divorced, who rely on investment income.
What does this mean for Washington? It’s anyone’s guess. Would one of these people move, and, what would that mean? I don’t know. The literature suggests that Washington would likely see some capital flight if it imposed these types of taxes, but given how concentrated the tax base here is (four people), it’s tricky to know what would happen. It’s complicated, just like any issue of this type, and, anyone, politician or otherwise, who simply dismisses concerns suggesting that the “evidence” makes the issue clear usual is dealing with a limited view of what the evidence actually is.
* I emailed Representative Noel Frame for any comment on her summary of the evidence, and, will update this if I hear back. **Updated 11/2/2021 – No reply from Representative Noel Frame.
