More than two thousand years ago, a simple fable attributed to Aesop, a Greek story teller, illustrated a problem that economists and policymakers still confront today.
A farmer owned a remarkable goose that laid a golden egg every day. The steady and consistent supply of eggs made him wealthy. But eventually, one golden egg a day was no longer enough. Convinced that there must be a store of gold inside the goose, the farmer killed it so he could obtain all the gold at once.
There was none.The farmer had destroyed the productive source of the very wealth he wanted to increase.
The Moral of Aesop’s story was a warning about greed and impatience, not a fable on taxation. But its central lesson applies surprisingly well to public finance: what can be extracted from a productive source today may affect what that source produces tomorrow.
That does not make taxation equivalent to killing the goose, however it is instructive in understanding the second order effects of tax policy. There in lies the balancing act, governments require revenue to finance services, infrastructure, education, public safety and other investments that can themselves support economic activity and make a state more attractive.
Modern taxpayers have one ability Aesop’s goose did not. They can leave.
They can also, just as potentially damaging to a tax base, remain while changing when they realize income, where they invest, where businesses start and expand, what functions and employees should be in a state, how assets are structured and how much taxable activity occurs within a state.
That matters particularly in places such as California and New York because their income-tax systems are already highly progressive, and remarkably dependent upon a comparatively small number of taxpayers.
In this analogy, the “golden goose” is not an individual person. It is the recurring stream of taxable income, investment and economic activity upon which government revenue depends from a very small group of people.
The Golden Goose Challenge
Having wealthy constituents in a state is a luxury any state would love to have. In some states with heavy progressive taxes, leaning on these wealthy residents for revenue makes it much easier on the rest of the electorate, and an easier sell for politicians. However, that knife cuts both ways. Lean enough and you have a steady stream of eggs, lean too much and you may chase the geese and all their gold off. The effects can be dramatic in those states with high concentrations of progressive taxation.
California
Let’s take California for example, their own budget documents describe just how concentrated its revenue system has become.
During the 20 years through 2023, California’s Top 1 percent of income earners paid an average of approximately 45 percent of all resident personal income tax liability. In 2023, roughly 176,000 tax returns in that top 1 percent produced nearly 37 percent of resident personal income taxes. In 2021, their share had reached approximately 50 percent. Not only showing the concentration, but also the uneven revenue stream from this segment. [1]
Personal income taxes themselves account for nearly 60 percent of California General Fund revenue. California’s budget also notes that the tax liability of the top 1 percent fell approximately 40 percent in 2022 even though the broader California economy was roughly flat, largely because high-income tax receipts depend heavily upon capital gains and stock-based compensation.[1]
That is an important distinction. A state does not have to lose the taxpayer to lose some of the eggs.
New York
New York is even more striking.
Preliminary 2024 New York tax data shows that taxpayers reporting at least $1 million of New York source income represented just 0.9 percent of taxpayers but produced 44.6 percent of state personal income tax (PIT) liability. The Top 200,000 taxpayers,1.8 percent of tax filers, produced 51.9 percent of PIT.[9]
And the concentration becomes extraordinary at the very top:
| New York taxpayers, 2024 | Share of Taxpayers | Share of Personal Income Tax |
|---|---|---|
| Top 200 taxpayers | 0.002% | 7.7% |
| NY source millionaires | 0.9% | 44.6% |
| Top 200,000 | 1.8% | 51.9% |
| Top 25% | 25.0% | 89.5% |
| Lower 50% | 50.0% | 0.2% |
The Top 200 individuals and households alone generated about $5.51 billion of New York’s $71.9 billion in personal income tax liability in 2024.[9]
It is easy to understand why policymakers look to high income and high wealth taxpayers when seeking additional revenue: relatively few people can provide substantial sums. New York’s Top 200 tax payers paid more than 30 times more in income taxes that half of the state.
But it also means the fiscal consequences of behavioral changes among those same taxpayers can be disproportionate.
When a small group produces a very large share of the revenue, what happens to that group has a very large impact on state revenue.
California: One Tax, Very Different Estimates
California’s Proposition 40, the so called Billionaire Wealth Tax Act, on the ballot in November provides an unusually good real world demonstration of the challenge.
The November 2026 measure would impose a one time wealth tax of up to 5 percent on not just income, but the entire covered wealth above $1 billion for individuals and certain trusts meeting its California residency requirements with most proceeds dedicated to health care.[6]
The authors who helped draft the initiative originally estimated that the tax would produce approximately $100 billion, or roughly $20 billion per year while payments were being made.[2]
Their updated calculation illustrates how they arrived at that revenue number. They identified 213 California billionaires with approximately $2.182 trillion in wealth. Five percent of that amount is about $109 billion. After assuming a 10 percent reduction for avoidance and evasion, they estimated approximately $99 billion and rounded it to $100 billion.[2]
However, other researchers have produced substantially different estimates.
| Analysis / Scenario | Estimated wealth tax revenue | Estimated effect of departures on revenue |
|---|---|---|
| Initiative drafters’ original estimate | ≈ $100B | 10% avoidance/evasion assumed |
| Boll, Saez & Zucman benchmark | ≈ $98B | ≈ $150M annual temporary loss |
| Boll, Saez & Zucman aggressive-leaver scenario | ≈ $84B | ≈ $530M annual temporary loss |
| Boll, Saez & Zucman broader scenario: missing billionaires + aggressive leavers | ≈ $109B | ≈ $570M annual temporary loss |
| Hoover / Jaros-Rauh | ≈ $40B | Losses produce negative $24.7B NPV* |
| California Legislative Analysts Office (LAO) | Tens of billions | Ongoing PIT reduction potentially < $1B/year |
The updated analysis by Jasper Boll, Emmanuel Saez and Gabriel Zucman, Saez was also among the initiative’s drafters, estimates about $98 billion in its benchmark scenario. Even under what the researchers describe as aggressive assumptions about reported billionaire departures, their estimate falls to only approximately $84 billion, with about $530 million per year of temporarily lost California income taxes. A broader scenario that attempts to account for billionaires omitted from the Forbes list produces an estimate of $109 billion and about $570 million of annual temporary income tax losses.[3]
Stanford University’s Hoover Institution researchers Benjamin Jaros, Joshua Rauh and colleagues arrive at a dramatically different conclusion.
Their model estimates that the wealth tax itself would collect only about $40 billion, less than half the proponents’ $100 billion estimate. They argue that publicly reported migration had already removed nearly 30 percent of the potential wealth-tax base and that additional unobserved departures would further reduce collections.[4]
Hoover identifies six prominent taxpayers as having moved their tax residency out of California before the January 1, 2026 cutoff.
Google cofounders Larry Page and Sergey Brin
Paypal/Palantir founder Peter Thiel
Auto-lending entrepreneur Don Hankey
Uber cofounder Travis Kalanick
Filmmaker Steven Spielberg
Hoover estimates that just these six represented approximately $536 billion, or nearly 30 percent, of the potential billionaire wealth tax base.[4]
The $100 billion versus $40 billion difference may be the most important fact in the entire debate. Both estimates come from credentialed researchers examining essentially the same proposed tax. The difference is largely about assumptions concerning behavior.
Hoover goes one step further. Its researchers estimate that California billionaires currently generate approximately $3.3 billion to $5.8 billion annually in California income taxes. They then estimate how much of that recurring income-tax stream could disappear if some taxpayers permanently establish residency elsewhere.[4]
They report a net present value of negative $24.7 billion. For a non-accountant, the simplest way to understand this is to imagine two piles of golden eggs.
In the first pile is the money California receives from the one-time wealth tax.
In the second pile is the stream of income taxes California would otherwise have received from those taxpayers in future years if they had stayed.
Future dollars are worth somewhat less than dollars today, so economists “discount” those future tax payments back into today’s dollars and call that Net Present Value (NPV). Hoover’s model concludes that the present day value of the future income taxes lost exceeds the wealth tax revenue collected by approximately $24.7 billion. In summary, Hoover is saying that eventually the Wealth Tax will cost more than it generates to California. So, under Hoover’s assumptions, California does lose money versus doing nothing. [4]
Two important qualifications for the Hoover estimate:
- Negative $24.7 billion is not a $24.7 billion annual budget loss, nor is it a bill California would have to pay. It is a comparison between two streams of future revenue expressed in today’s dollars. So they will never have to pay this out, they will just collect less in taxes.
- The result depends heavily upon how long the departing taxpayers, and their tax payments, are assumed to remain gone.
The initiative’s authors specifically challenge Hoover on this point. They argue that Hoover effectively counts the lost income tax payments from departures indefinitely while counting the wealth tax only once. They contend that a truly one time tax should produce a temporary migration response rather than a permanent one. [5] We’ll leave that up to you as to which assumption is correct.
However, that disagreement is consequential. If the income-tax loss lasts only several years, the $100 billion one time revenue side of the calculation dominates. If a meaningful part of the income tax base permanently relocates, the continuing loss becomes progressively more important with every passing year. Hoover tested 100,000 combinations of assumptions about tax revenue, income-tax losses and discount rates. Seventy-one percent produced a negative NPV, with negative $24.7 billion as the mean result.[4]
That is not a prediction that California will definitely lose $24.7 billion, however it gives you a rough order of magnitude to what is at stake.
It also services as a warning about how dramatically the assumption changes once tomorrow’s revenue eggs are included in the calculation.
California’s nonpartisan Legislative Analyst’s Office (LAO) lands between the competing camps. It estimates that the measure would generate tens of billions of dollars temporarily, while potentially reducing ongoing personal-income-tax revenues by less than $1 billion per year.[6]
Most Geese Probably Will Not Fly Away
It is equally important not to exaggerate taxpayer migration.
Research consistently indicates that most wealthy taxpayers do not move following a tax increase.
Rauh and Ryan Shyu’s study of California’s 2012 Proposition 30 found that the baseline departure rate among affected high earners was approximately 1.5 percent. Following the tax increase, an additional 0.8 percent of the affected residential tax base ceased being full-year California residents.[7]
In other words, the overwhelming majority stayed.
Other research by Charles Varner, Cristobal Young and Allen Prohofsky examining 25 years of California administrative tax data found little migration response to changes in top tax rates.[8]
That is an important counterweight to predictions of mass tax flight.
But the Rauh-Shyu study found something else that is particularly relevant to the Golden Goose analogy.
Migration accounted for only a relatively small portion of the behavioral response. Changes in taxable income among people who remained in California mattered much more. Their study estimated that migration and changes in taxable income together reduced the additional revenue that otherwise would have been expected from Proposition 30 by 45.2 percent in the first year and 60.9 percent within two years, with migration accounting for only 9.5 percent of that total erosion.[7]
The lesson: The goose does not necessarily have to leave the farm to lay fewer taxable eggs.
Whether those changes represent less economic activity, changed timing of income, tax planning, changed investment behavior or other responses is much harder to determine. It is also not an apples to apples comparison as a Wealth tax is much harder to dodge as it designed to capture all assets.
New York: Millionaires Leaving. What Did That Cost?
New York provides another useful example.
The state reported that 1,679 taxpayers with at least $1 million in adjusted gross income (AGI) changed their address to another state during 2024. That represented 2.49 percent of the millionaire population covered by the address change data, meaning approximately 97.5 percent did not leave.[10]
So describing the data as a mass millionaire exodus would be difficult to defend.
Indeed, New York reported a record 99,404 New York-source millionaire returns in tax year 2024, up more than 12,000 from 2023.[9]
Yet New York’s own Tax Department also cautions that the tax loss caused by a relatively small number of departing millionaires can be significant because tax liability is so concentrated.[10]
Unfortunately, the state does not publish a figure saying that those specific 1,679 departing taxpayers cost New York exactly $X in annual revenue.
We therefore should not invent one, but we can put the number into a rough order of magnitude estimate.
New York’s 99,404 millionaire returns generated approximately $32.07 billion of state personal-income-tax liability in 2024, or an average of approximately $323,000 per millionaire return.[9]
If, purely as an illustrative benchmark, the 1,679 departing millionaire filers had generated the same average liability, they would represent roughly $542 million of annual tax liability.
That is not an estimate of New York’s actual migration loss. The two taxpayer datasets are not identical, income among millionaires is extremely uneven, and people who move can continue owing New York taxes on New York-source income.
But it illustrates why head counts can be misleading.
The Top 200 New York taxpayers averaged approximately $27.5 million of state PIT liability each in 2024. A migration group containing even a handful of taxpayers from that extreme upper tail could therefore have a fiscal impact many times larger than an average-millionaire calculation suggests.[9]
New York’s broader migration data show the same underlying trend. In 2024, 121,251 part year taxpayers moved into New York while 134,913 moved out, producing net out-migration of 13,662 filers. Among households earning $500,000 or more, net out-migration was approximately 1 percent of resident tax filers.[11]
Again, most stayed. But in a tax system where 0.9 percent of taxpayers generate 44.6 percent of PIT liability, the relevant question is not merely how many leave, but which ones leave, how much taxable income leaves with them, and how much remains taxable after they go.
When a Small Revenue Loss Becomes a Recurring Hole
The scale becomes clearer when these estimates are compared with actual budgets.
California’s enacted 2026–27 budget contains approximately $251.5 billion of General Fund spending.[12]
An ongoing $1 billion annual income tax loss, the upper end of the range implied by the California LAO’s “less than $1 billion” language, would equal less than 0.4 percent of one year’s General Fund spending.
That sounds small. But if a $1 billion revenue stream disappeared and remained absent for ten years, the simple nominal total would approach $10 billion, before considering growth or discounting. This is precisely why the duration of the behavioral effect matters so much.
Hoover’s negative $24.7 billion NPV is equivalent to roughly 10 percent of one year’s current California General Fund spending, but it is not a one-year deficit. It represents Hoover’s estimate of the accumulated present value of a changed future revenue stream.[4][12]
New York faces a similar sensitivity.
Its enacted State Fiscal Year (SFY) 2027 financial plan projects about $127.2 billion in General Fund receipts and $128.8 billion in General Fund disbursements. Its current financial plan projects General Fund gaps of $6.4 billion in SFY 2028, $10.5 billion in 2029 and $14.7 billion in 2030.[13]
Our illustrative $542 million millionaire-migration benchmark would equal only about 0.4 percent of annual General Fund spending.
But against New York’s projected $6.4 billion FY2028 budget gap, the same amount would represent roughly 8 percent of the gap.
Again, $542 million is not a measured migration loss. The comparison simply demonstrates why relatively modest recurring revenue changes can become meaningful when budgets already operate near their revenue limits.
What Else Can Leave With the Goose?
Tax revenue may only be one part of the calculation. Very wealthy taxpayers disproportionately include founders, investors, executives and owners of businesses. If residency changes, some economic relationships may eventually change with it: investment decisions, startup formation, company expansions, payroll, purchases, philanthropy and capital allocation.
But here again, assumptions should not be mistaken for facts.
A billionaire moving his legal residence does not automatically move his company, eliminate its California employees or remove all of its California tax liability. The initiative’s authors make precisely this argument: the location of a shareholder is not necessarily the location of the underlying business.[2]
There is empirical evidence that some broader economic effects exist, however.
A recent NBER study using detailed Scandinavian data found that a one-percentage-point increase in the top wealth-tax rate reduced the long-run stock of wealthy taxpayers by approximately 2 percent. Because many of these taxpayers owned businesses, owner migration also affected their firms. The researchers estimated aggregate reductions of approximately 0.02 percent in employment, 0.07 percent in investment and 0.10 percent in value added.[16]
Those are measurable effects—but notably small aggregate effects.
And they come from Sweden and Denmark under different tax systems, so those percentages cannot simply be transplanted into California or New York.
The study nevertheless illustrates the correct way to frame the question. The consequences do not end with whether a taxpayer changes the address on a return.
Tax Chain:
tax → taxable income/wealth → tax response → investment/business response → employment and economic activity → future tax base.
How large each link becomes is an empirical question.
New York Is Considering the Same Tradeoff
The issue is not confined to California.
New York City Mayor Zohran Mamdani proposed raising the city’s personal-income-tax rate by two percentage points on roughly 33,000 residents earning more than $1 million annually. The city’s Independent Budget Office estimated that such a tax could raise approximately $3 billion annually if Albany authorized it, although that proposal was not enacted as part of the city’s FY2027 budget.[17]
Separately, New York legislators have proposed a Billionaire Mark-to-Market Tax that would tax annual unrealized gains of residents with at least $1 billion of net assets. Its sponsors estimate approximately $23.3 billion in first-year revenue and at least $1.2 billion annually thereafter. The measure remains in committee.[18]
Those may prove to be good policies, bad policies or something between depending on your perspective and the outcomes. But because New York already receives nearly half of its personal-income-tax liability from less than 1 percent of taxpayers, their second-order effects deserve particular attention.
The Lesson of the Goose
No economist can reliably tell California today exactly how many billionaires will ultimately leave because of Proposition 40. No analyst can know precisely how much taxable income those who remain will realize five years from now. No one can say with certainty whether New York’s next tax increase would cause 100 wealthy residents to leave, 1,000 to leave, or virtually none. And no one can perfectly estimate how much investment, hiring, business activity or future tax revenue would accompany those decisions. But uncertainty about the magnitude of second-order effects is not the same as uncertainty about their existence.
People respond to incentives.
- Some will move.
- Most probably will not.
- Some who remain will change their financial behavior.
- Businesses, employees, and economic activity may follow owners in some cases and remain exactly where they are in others.
- Additional tax revenue may finance public investments that create economic value.
- And lost recurring tax revenue may compound for years after a one-time tax has been collected.
The debate should therefore not be reduced to either extreme:
“Tax the wealthy and nothing else changes.”
or
“Raise their taxes and they will all leave.”
The available research supports neither proposition.
The lesson: Aesop’s farmer made his mistake because he considered only the gold he hoped to obtain immediately. He never asked what today’s decision would do to tomorrow’s production.
That may be the useful lesson for modern tax policy.
The first-order question is how much revenue a tax can collect.
The second-order question is what happens to the source of that revenue afterward.
In a highly progressive tax system where a very small population already produces an unusually large share of government revenue, that second question becomes particularly important. The Golden Goose does not tell us how much to tax. It tells us not to assume that tomorrow’s eggs are unaffected by what we do to the goose today.
References
[1] California Department of Finance. (2026). Governor’s Budget Summary 2026–27. State of California.
https://ebudget.ca.gov/2026-27/pdf/BudgetSummary/FullBudgetSummary.pdf
[2] Galle, B. D., Gamage, D., Saez, E., & Shanske, D. (2026). Expert Report on the California 2026 Billionaire Tax: Revenue, Economic, and Constitutional Analysis. Institute on Taxation and Economic Policy.
https://itep.org/expert-report-on-the-california-2026-billionaire-tax-revenue-economic-and-constitutional-analysis/
[3] Boll, J., Saez, E., & Zucman, G. (2026). California Billionaires: Wealth, Taxes, and Wealth Tax Revenue Estimates. National Bureau of Economic Research Working Paper No. 35218.
https://www.nber.org/papers/w35218
[4] Jaros, B., Rauh, J. D., Kearney, G., Doran, J., & Cosso, M. (2026). The Net Present Value of the Billionaire Tax Act: An Assessment of the Fiscal Effects of California’s Proposed Wealth Tax. Hoover Institution / SSRN.
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6340778
[5] Gamage, D., Galle, B. D., Saez, E., & Shanske, D. (2026). Response to “The Net Present Value of the Billionaire Tax Act.” SSRN.
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6433598
[6] California Legislative Analyst’s Office. (2026). Proposition 40: Imposes One-Time Tax on Certain Taxpayers.
https://lao.ca.gov/BallotAnalysis/Proposition?number=40&year=2026
[7] Rauh, J. D., & Shyu, R. (2024). Behavioral Responses to State Income Taxation of High Earners: Evidence from California. American Economic Journal: Economic Policy, 16(1), 34–86.
https://www.aeaweb.org/articles?id=10.1257/pol.20200500
[8] Varner, C., Young, C., & Prohofsky, A. (2018). Millionaire Migration in California: Administrative Data for Three Waves of Tax Reform. Stanford Center on Poverty and Inequality.
https://inequality.stanford.edu/node/9725
[9] New York State Department of Taxation and Finance. (2026). Personal Income Tax: Tax Facts.
https://www.tax.ny.gov/data/stats/taxfacts/personal-income-tax.htm
[10] New York State Department of Taxation and Finance. (2026). Migration.
https://www.tax.ny.gov/data/stats/taxfacts/migration.htm
[11] Office of the New York State Comptroller. (2026). Taxpayer Migration in New York State.
https://www.osc.ny.gov/reports/taxpayer-migration
[12] State of California. (2026). 2026–27 Enacted Budget Detail.
https://ebudget.ca.gov/budget/e/2026-27/BudgetDetail
[13] Office of the New York State Comptroller. (2026). Report on the Enacted Budget and Financial Plan.
https://www.osc.ny.gov/files/reports/pdf/enacted-budget-and-financial-plan-report-1.pdf
[14] U.S. Census Bureau. (2026). Vintage 2025 Population Estimates.
https://www.census.gov/newsroom/press-releases/2026/population-growth-slows.html
[15] Florida Legislature, Office of Economic and Demographic Research. (2026). General Revenue Estimates.
https://flsenate.gov/Committees/DownloadMeetingDocument/7918
[16] Jakobsen, K., Kleven, H., Kolsrud, J., Landais, C., & Muñoz, M. (2026 revision). Taxing Top Wealth: Migration Responses and Their Aggregate Economic Implications. National Bureau of Economic Research Working Paper No. 32153.
https://www.nber.org/papers/w32153
[17] New York City Mayor’s Office. (2026). Fiscal Year 2027 Preliminary Budget materials and remarks concerning the proposed millionaire income-tax increase.
https://www.nyc.gov/mayors-office/news/2026/02/transcript–mayor-mamdani-releases-balanced-fiscal-year-2027-pre
[18] New York State Senate. (2025–2026 Session). S165: Billionaire Mark-to-Market Tax Act.
https://www.nysenate.gov/legislation/bills/2025/S165
