Eight responses about millionaire migration
A reply to report by Pioneer Institute’s Greg Sullivan
A MASSACHUSETTS-BASED FREE MARKET THINK TANK, the Pioneer Institute, produced a paper titled Eight Reasons to Question Professor Cristobal Young’s Conclusions about Millionaire Migration. I welcome the critical reflection on my work. Indeed, scholarly research is always subjected to interrogation and criticism during the publication review process, and there is no reason this should cease simply because the research has been published. (1)
That said, these criticisms are misguided. Most are simple misunderstandings that can be readily clarified. I provide a point-by-point response to Sullivan’s eight questions about my work, as concisely as possible. I finish with a new look at census data on top income earners in New Jersey, Massachusetts, and Florida.
Question 1: Millionaires are people with net worth of $1M+
The lead criticism is that I did not consider people who are millionaires by net worth. That is largely true. The reason for this is that I am studying taxes on top incomes. There is not a tax on net worth in the U.S., nor is one currently proposed in any U.S. state. If people have high net worth, but their income is below $1M, they will not be affected at all by the “Fair Share Amendment” or any other millionaire tax in the U.S. My research focuses purely on the people who would be affected by a millionaire tax, and does not consider those not affected.
Sullivan rhetorically asks, “Is a millionaire someone who makes a million dollars once?” (p 6). As a philosophical question, I am not sure what the right answer is. For the purposes of the Massachusetts millionaire tax, the answer is definitely “yes.” Indeed, earning a million dollars in a single year is the only requirement to be eligible for the proposed new tax rate.
Incidentally, I have analyzed billionaires by net worth in my book, using the Forbes list of billionaires (Young 2017:32-34). This analysis assumed that most billionaires have very high annual incomes, and so would be affected by the top tax rate in their state. The results show that billionaires are no more likely to live in low tax states than high tax states (slight preference for higher tax states). Billionaires also have lower migration rates than the middle class or the poor. Over a five year period (2010-2015), billionaires were more likely to pass away than to move across state lines. In other words, the data on net-worth billionaires reiterates what we see from (income-defined) millionaires.
Question 2: One-Time Millionaires
Sullivan writes that “Young’s definitions overemphasize one-time millionaires – people who sell a home, a business, or another once-in-a-lifetime asset they have been counting on” (8).
People who make $1M+ are typically at the very peak of their careers, and will only make this much money for a few years of their lives. In other words, the millionaire tax mostly applies to people who are having a remarkably good financial year. In my view, this is part of the reason why the tax does not cause much migration – people affected will only pay the tax for a few years.
Nevertheless, this criticism is incorrect. We published results for people who were one-time millionaires, as well as for people who earned million-dollar incomes for many years – a distinction we called “millionaire persistence” (Young et al 2016:433). We showed that highly persistent millionaires have lower migration rates – they are more tied to place than one-time millionaires. However, when they do move, they are more likely to make their destination a lower-tax state. These two effects largely cancel each other out. In any event, it is incorrect to say we over-emphasize one-time millionaires; we provide estimates for both one-time and long- term millionaires.
Question 3: What if people move just before claiming a large capital gain?
My work has defined millionaires as those making a million dollars in the year that they moved. It is possible, however, that people can foresee when they are about to earn $1M+, and move to a low-tax state just before this happens. This is most credible in the case of capital gains income, where people can choose when they exercise stock options. However, most millionaires are the “working rich,” and most of their income is from wages and salaries. Nevertheless, the idea of moving to exercise large capital gains is a question I am currently studying with my colleague Charles Varner using California income tax data.
Question 4: The Florida Effect.
Sullivan argues that “Professor Young pays too little heed to the Florida effect.” This is odd, because in Young et al (2016), the word Florida appears on 33 occasions, and there is a sub- heading in the article titled, “The Florida Effect.”
In the press release, Sullivan writes:
“Saying there’s virtually no tax migration when you exclude Florida is like saying that if you exclude Muhammad Ali, Louisville hasn’t produced many great boxers,” said Sullivan.
This sounds like a clever one-liner, but it is very misleading. We did not exclude Florida. We simply point out that it is Florida that drives the entire result, and that all other no-income-tax states see no evidence of attracting millionaires. We are emphasizing that Florida is important, nothing more or less.
Our overall estimates of tax-flight migration assume that all migration to Florida is strictly for tax purposes. We see this as a conservative estimate. A lingering question is, if Florida adopted a millionaire tax, would Texas or New Hampshire become the new destination for millionaire migration? Or would east coast elites continue to see Florida as an attractive location? I believe that at least some of the migration to Florida is strictly for tax purposes. But I also suspect that sun, sand, and palm trees would be attractive even if Florida had the same tax policies as New Jersey or New York.
Question 5: What about capital gains taxes?
Capital gains taxes are of course interesting. However, few millionaires make their money from capital gains. Only 11 percent of millionaires primarily make their income (75% or more) from capital (Young et al 2016:433). Most millionaires are the working rich, employed and earning salaries as managers, financial executives, accountants, lawyers, and doctors.
I have analyzed migration separately for millionaires with high and low capital gains. Neither migration nor tax flight is notably different between “capitalists” (those with high capital gains) and the working rich (those with high salaries) – as reported in Young et al (2016:433). Both have low rates of migration, and both have a level of tax flight that is essentially negligible for the millionaire population in a state.
As noted above, I am conducting ongoing research on capital gains responsiveness.
Question 6: State-Level Estate (Inheritance) Taxes
Sullivan states that I did not consider taxes on millionaire estates. First, this is an aside for the “Fair Share Amendment.” The proposed millionaire tax does not tax estates – nor do any of the other state millionaire taxes in the U.S. So, for this issue, there is no reason to be talking about an estate tax.
Second, it is wrong to say I did not consider estate / inheritance taxes. I explored inheritance taxes at the state level, and found they had no effect on millionaire migration. Because readers have limited patience for null results, I left this to a footnote, saying “in alternative specifications, we included a coarse dummy variable for a state-level inheritance tax” (Young et al 2016: fn6).
Currently, it is difficult to estimate the actual level of inheritance taxation in different states, due to the complexity of state inheritance tax laws. Scholars are currently working on a formal state- level estate tax calculator, which will allow more detailed research on this question in the future.
Question 7: “Important Caveats”
Item 7 is where the comments by Sullivan become comedic. I limit my response to two of the especially silly criticisms.
7.1 “Young notes that ‘Little is known about the migration patterns of the rich and their broader demography.’” Elsewhere, I am quoted as saying “the existing evidence… is limited. ”
Of course, I am not describing my own evidence as limited. Indeed, my evidence comes from the IRS tax returns of every top income earner in the country over more than a decade. The data set is remarkable, including roughly 45 million observations. What I am saying in this quote is that past research on millionaire migration is limited. This quote is an opening statement where I aim to fill in this knowledge gap, and provide a lot of new evidence. The quote only means that prior to my research we didn’t know much about this topic.
Specifically, some of the key facts I show are that top income earners are much more likely to be married, have children at home, and to own a business – all factors that root people in place and lower their migration. I also show that millionaires have much lower migration rates than the poor. With these data, we now know a lot about the migration patterns of the rich and their broader demography.
In the past, politicians have made statements like “there is nothing more portable than a millionaire and his money.” (2) This is something that sounded true, but turns out to be completely wrong. At best, one could say that millionaires are slightly more mobile than the upper-middle class, but less mobile than most people in the country.
7.2 “Professor Young advocates for a global millionaire tax”
I do not advocate for a global tax on wealth. I mentioned that Thomas Piketty, in his famous book Capital in the 21st Century, advocates for a global tax. This is the complete opposite of a state-level millionaire tax. A central conclusion from my research is that this kind of coordinated tax policy is not necessary: “States can make policy choices that contribute to the reduction of inequality without waiting for national or international agreements” (Young et al 2016: 440). I do not understand why the author would make up criticisms that are refuted by the original paragraph he quoted from.
Question 8: Cumulative Effect of Millionaire Migration
Finally, Sullivan argues that my work disregards the cumulative effect of millionaire migration over the long term. He also forecasts out revenue losses that accumulate every year through to the year 2045, based on different assumptions of millionaire out-migration. This forecast is not based on any actual evidence – it is entirely a chalk board exercise.
This argument about cumulative effects comes, as Sullivan notes, from a critique and replication of my work on millionaire migration in New Jersey – in an a study by the then Chief Economist of New Jersey, Charles Steindel and co-authors (Cohen, Lai, and Steindel 2015). After criticizing my methods, they proceeded to find the same results I had reported; specifically, their estimates were within the margin of error published with my estimates: substantively a small, or very small, effect of the New Jersey millionaire tax on migration (Young and Varner 2015).
However, to make this effect seem larger, they forecasted migration losses beyond the horizon of the data. In my response, I showed that most of the estimated migration happened in the first year after the tax came into effect, and diminished sharply thereafter. This approach of forecasting tax migration effects long after the tax had been passed was contradicted by the evidence.
In section 8, Sullivan also shows some confusing slippage between a measure of net migration between all states, and net out-migration from a single state (such as Massachusetts). Annual net migration of millionaires across any state line in the U.S. is 2.4 percent. This statistic does not provide an estimate of tax-flight migration. Indeed, the central question is how much of this migration is for tax purposes? As I report in my book (Young 2017: 23-27) only about 15 percent of millionaire moves come with a net tax advantage (meaning that after moving, only 15 percent of migrating millionaires overall live in a lower tax state – 85 percent of millionaire migration is just churn between states with no particular tax advantage).
This is an important issue. In Sullivan’s press release, he refers to “Young’s 2.4 percent
estimate” as if I had estimated that 2.4 percent of the Massachusetts millionaire population may leave the state after the millionaire tax is passed. This is completely incorrect. Based on my experience evaluating millionaire taxes in New Jersey, California, and across the U.S. with IRS data, I predict two migration results of the Massachusetts millionaire tax: (1) there will be many anecdotes of millionaires leaving the state; and (2) in aggregate, these flows will be too small to matter for the size of the millionaire population.
What will greatly affect the size of the millionaire population is the next recession. It has been nine years since the end of the last recession – which is historically longer than a typical economic expansion. An economic downturn is likely within the next few years. When that happens, the number of people earning million-dollar incomes in Massachusetts will fall sharply – not because they move away, but because they are earning less income. Forecasting revenues from a millionaire tax based on growth years is risky business. It would be wise to ensure that a millionaire tax contributes to the state’s rainy day fund, not just to current spending and investment. This is a basic principle of sound budgeting across the business cycle, and would equally apply to a new sales or excise tax. It is important to set aside some of these revenues for when the inevitable downturn occurs.
A Long-Term Analysis of Top-Income Earners
The long-term effects of top tax rates are very important – a view that Sullivan and I share. New research I am developing focuses on long-term, cumulative differences between high and low tax states. I provide a simple, vivid look at new evidence central to debates about taxing high- income earners.
The long-term, cumulative analysis is to simply look at the population of top income earners in a state over time. This captures both long-term migration patterns, as well as other dynamics such as economic opportunities that affect the top-earner population. Do high income earners tend to concentrate in low-tax states? Are high-tax states “less competitive” and less able to cultivate, attract, and retain top earners?
I reviewed Census data on the share of the adult population earning $200,000+ per year in constant dollars, from 1990 to 2016. I focus on $200k as that is the highest income level we can see over time in the census – above this point many people’s incomes are top-coded (concealed) for confidentiality reasons. (3)
New Jersey has long had one of the most progressive state income taxes in the U.S. (meaning relatively high taxes on top incomes), while Florida has never had a state income tax.
Massachusetts lies between the two, with a flat 5.1 percent income tax. The basic features of these state income taxes have been in place for generations. So, if higher taxes seriously reduce a state’s population of top earners, there will be a strongly diminishing population of top earners in New Jersey, a modest decline in Massachusetts, and a growing top earner population in Florida.
The data show some striking basic facts that are completely contrary to the tax flight hypothesis. In 1990, about 1.6 percent of Floridians earned $200k (in constant 2016 dollars), compared to 2.5 percent in Massachusetts and 3.5 percent in New Jersey.
Thirty years ago, there was much higher concentration of top income earners in New Jersey and Massachusetts than in Florida. Since then, the gap has only widened. In New Jersey, the top earner population rose sharply from 3.5 percent in 1990 to 8.0 percent in 2016. In Massachusetts, the rise of $200k earners was from 2.5 percent to 7.1 percent. In low-tax Florida, the rise was from 1.6 percent to 2.8 percent.
New Jersey and Massachusetts are leading centers in the country for elite talent and high income earners. Florida is not. Indeed, Florida is falling behind in comparison. These are basic facts that anyone debating state tax policy needs to take into account.
The basic reason for these patterns, I believe, is that the economic opportunities to earn very high income are found in New Jersey and Massachusetts, but much less so in Florida. The Sunshine State is a good place to retire, but not a good place for those aspiring to earn top incomes. Higher-tax places are in fact highly competitive – and centers for elite talent – especially when they offer a high-services, high-infrastructure equilibrium that is attractive to young people launching their careers.
Many of the arguments against state millionaire taxes predict a problem of trying to raise more and more revenue from a dwindling population of top earners. That is far from the experience of New Jersey, Massachusetts, and Florida.
For the last three decades, income has been accumulating at the top in ways we have not seen in generations. A large portion of all economic growth today accrues to the top one percent of income earners. (4)
Modest millionaire taxes at the state level are prudent ways to address growing income inequality in America, as well as providing revenues for public services and investments in people. These taxes come with little risk of millionaire tax flight. These taxes are largely paid by the late-career working rich, who are embedded in their states for a host of social, family, and economic reasons. Moving is a young person’s game, but earning top incomes is most common among professionals at the very peaks of their careers, tied to place by their employers, clients, families, and rich social networks that are not easy to move.
State budgets bear responsibility for kindergarten to grade 12 education, public colleges and universities, as well as much of the transportation infrastructure, police and judicial systems, and social services. These systems make a big difference for quality of life and opportunity in a state. New revenue streams come with a deep obligation to ensure those monies are spent in ways that make the state a better place to live, work, and thrive.
(1) I accept complete responsibility for the results of my research and any possible errors or omissions. However, it must be stated that Charles Varner has been a long-term and essential collaborator and deserves recognition. Moreover, much of what has been learned about millionaire migration would not have been possible without Ithai Lurie and Rich Prizinzano. Erin Cumberworth provided excellent research assistance.
(2) Yamamura, Kevin. 2011. “Plans to ‘tax the rich’ hold risks and rewards for California.” Sacramento Bee. Dec 27, 2011.
(3) This threshold considers a wider population of top earners than millionaires, consistent with Sullivan’s concerns about incomes below a million (though admittedly, this does not include net worth). Going forward, I plan to replicate this and additional analyses with the IRS tax data on all top income earners.
(4) See, for example, “Distributional National Accounts.”
Young, Cristobal. 2017. The Myth of Millionaire Tax Flight: How Place Still Matters for the Rich. Stanford University Press.
Young, Cristobal, Charles Varner, Ithai Lurie, and Richard Prisinzano. 2016. “Millionaire Migration and the Taxation of the Elite: Evidence from Administrative Data.” American Sociological Review. Vol. 81(3):421-446.
Cohen, Roger, Andrew Lai, and Charles Steindel. 2015. “ A Replication of “Millionaire Migration and State Taxation of Top Incomes: Evidence from a Natural Experiment.” Public Finance Review. Vol. 43(2).
Young, Cristobal, and Charles Varner. 2015. “A Reply to “A Replication of “MillionaireMigration and State Taxation of Top Incomes: Evidence from a Natural Experiment”
Public Finance Review. Vol. 43(2): 226-234.