Hasta la Vista, Baby
October 9, 2026

Summary
This November, voters will decide on the California Billionaire Tax Act (Prop 40), which would retroactively impose a 5 percent tax on the wealth of residents worth more than $1 billion. Its supporters say it could raise $100 billion, even though several billionaires have already left the state or plan to.
Beginning in 2020 and intensifying in 2021, a tidal wave of entrepreneurs and businesses left California, including tech billionaires Elon Musk and Larry Ellison. It’s not hard to see why: the state’s high taxes, heavy regulations, and increasingly contentious political climate have made it clear that success is not welcome.
This report examines the real costs of capital flight. Using a synthetic control method, it builds a “statistical twin” modeled after states that mirrored California before 2021 on employment, tax revenue, tech-sector size, economic freedom, educational attainment, and housing prices, but experienced no similar exodus.
Results showed that through 2024, the real California fell behind its twin across the board, losing $100 billion in potential growth, roughly 79,000 jobs, and 813,000 residents. Most ironically, California’s actual tax revenue was about $50 billion less than the twin’s, or about half of what Prop 40 supporters claim it would raise.
The truth is this: when the wealthy leave, they take the tax base with them. As a result, the state collects less money, not more.
Abstract
California will vote this November on whether to retroactively impose a one-time 5 percent wealth tax on its billionaire residents. In response to the ballot initiative, several ultra-wealthy individuals have already left the Golden State, while others are reportedly planning to move. What could a new wave of billionaire departures mean for California’s economy? To answer this question, this paper examines the first wave of high-profile billionaire and corporate departures from California in 2021. Using a synthetic control method, the paper finds that California experienced statistically significant economic contractions following the treatment. Year-over-year estimates show a 2.67 percent drop in economic growth and a 1.70 percent decline in GDP per capita in 2022, alongside a 1.13 percent employment contraction, a 1.66 percent drop in tax revenue in 2023, and sustained population growth decreases through 2024. Relative to the synthetic counterfactual, cumulative losses across the 2021–2024 post-treatment period amount to nearly $100 billion in economic output, $50 billion in tax revenue, 79,000 jobs, and a net population deficit of 813,000 residents.
JEL Codes: H24, H71, H73, R23, R58
Keywords: synthetic control method, wealth tax, billionaire migration, California fiscal policy, tax avoidance
1. Introduction
This November, California voters will consider Proposition 40, a ballot initiative known as the “2026 Billionaire Tax Act.”[1] If approved, the measure would retroactively impose a one-time 5 percent tax on the wealth of individuals with net worth exceeding $1 billion who were California residents at the beginning of 2026. Supporters estimate the tax could generate up to $100 billion in revenue for state programs.[2]
Even before the vote, several billionaires have already left California or announced plans to do so.[3] For example, Google co-founders Larry Page and Sergey Brin departed the Golden State, taking more than $500 billion in wealth and an estimated quarter of the tax’s projected revenue with them.[4] Similarly, while Meta founder Mark Zuckerberg has not officially moved out of California, he has begun purchasing real estate in Florida.[5] These developments have raised questions about the potential economic and fiscal consequences of billionaire migration in response to state-level wealth taxation.
This would not be the state’s first billionaire exodus. Beginning in 2020 and accelerating in 2021, a wave of high-profile entrepreneurs and companies relocated from California,[6] with many citing the state’s high taxes, heavy regulation, and increasingly contentious political climate.[7] Among the most prominent examples were Elon Musk, who relocated Tesla’s headquarters, and Larry Ellison, who relocated Oracle’s.
This earlier episode provides an empirical opportunity to examine a question central to the current debate: What happens to a state’s economy after a substantial departure of high-net-worth individuals and major corporations? This paper examines that question empirically by studying the 2021 wave of high-profile billionaire and corporate departures from California. This event provides the closest observable precedent for assessing the potential consequences of a more pronounced billionaire exodus if the state implements this wealth tax. The objective is not to assume that future departures will replicate the 2021 episode exactly, but rather to use this earlier wave as evidence of the economic effects that may accompany the departure of high-net-worth individuals and the businesses associated with them. This question is particularly relevant given the pending California Billionaire Tax Act and the possibility of another wave of departures.
Specifically, this study employs a synthetic control method to estimate the economic consequences of the 2021 wave of high-profile billionaire and corporate departures from California by comparing California’s post-2021 economic trajectory with that of its synthetic counterfactual, statistically significant post-treatment downturn. Immediate annual impacts include a 2.67 percent decline in economic growth and a 1.70 percent reduction in GDP per capita in 2022, followed in 2023 by a 1.13 percent drop in employment and a 1.66 percent fall in tax revenue, alongside ongoing population growth slowdowns in 2023 (0.64 percent) and 2024 (0.72 percent). Aggregating across the 2021–2024 evaluation window, the treated unit underperformed its counterfactual baseline by approximately $100 billion in total output, $50 billion in tax revenues, nearly 79,000 jobs, and 813,000 in population.
2. Behavioral Responses to Wealth Taxation
One of the fundamental principles of economics is that individuals respond to incentives. Wealth taxes, like other taxes, can alter these incentives by reducing the after-tax return to holding and accumulating wealth. A billionaire wealth tax can therefore generate several behavioral responses, each with potentially different implications for economic activity (Scheuer and Slemrod, 2021). This paper focuses on four channels through which a wealth tax may affect a state economy: distortion, restructuring, evasion, and migration. These mechanisms are not mutually exclusive. Individuals may simultaneously reduce investment, restructure their assets, underreport their wealth, or relocate to another jurisdiction. The relative importance of these responses determines the tax’s ultimate economic and fiscal consequences.
The first channel is distortion. A wealth tax is levied on the stock of wealth rather than solely on the income generated by that wealth. Consequently, it reduces the after-tax return to holding capital and may alter incentives to save, invest, and undertake entrepreneurial activity. The economic effects of a wealth tax therefore need not be limited to the amount of revenue collected by the government. If affected taxpayers reduce investment or entrepreneurial activity, the resulting reduction in capital formation could, over time, affect productivity, employment, and output.
The second channel is restructuring, which encompasses legal adjustments to the composition and structure of assets undertaken to minimize taxable net worth. Common strategies include shifting assets toward preferential or exempt asset classes, increasing debt leverage, and taking advantage of valuation discounts on illiquid holdings. Under the proposed California Billionaire Tax, real estate and certain pension and retirement assets are exempt from the tax. In addition, because the tax is imposed on net worth, liabilities can reduce an individual’s taxable wealth. These provisions create incentives for taxpayers to alter the composition and financing of their assets without necessarily reducing their underlying economic wealth.
A third channel is tax evasion, which involves the illegal concealment or underreporting of taxable wealth. The scope for evasion depends heavily on the observability of assets and the availability of third-party reporting. Publicly traded securities and bank deposits are generally easier for tax authorities to monitor than privately held businesses, art, collectibles, and other difficult-to-value assets. Consequently, the design and enforcement of a wealth tax can influence the extent to which taxpayers respond through evasion rather than through legal restructuring or other behavioral margins.
Finally, geographic migration represents the physical relocation of tax residency to a lower-tax jurisdiction. For a subnational wealth tax, migration is a particularly important behavioral margin because taxpayers can reduce future tax liabilities by changing their state of residence. Moving to a neighboring or otherwise attractive low-tax state can allow taxpayers to retain access to U.S. financial markets, institutions, and business opportunities while reducing or eliminating exposure to California taxation. States that have historically attracted high-net-worth migrants, such as Florida and Texas, do not impose a state individual income tax. Migration may therefore represent a relatively low-cost margin of adjustment for some high-net-worth residents, particularly when the tax liability is large relative to the costs of relocation.
The empirical literature on wealth taxation has focused extensively on estimating the semi-elasticity of taxable wealth, which measures the percentage change in taxable wealth associated with a one-percentage-point change in the tax rate. Studies may express this response relative to either the wealth tax rate
(τw)
:
Semi-Elasticity (εsemi)=%Δ Taxable WealthΔτw
The estimated semi-elasticity captures the aggregate response to the behavioral margins described above, including changes in investment and saving, asset restructuring, tax evasion, and migration. Estimates are therefore conditional on the design of the wealth tax, as well as its enforcement and compliance environment.
Evidence from subnational wealth-tax systems is particularly relevant for evaluating a potential California wealth tax because taxpayers can respond to tax differentials by moving within the same country. In Switzerland, where municipal and cantonal wealth tax rates vary across subnational jurisdictions, Brülhart et al. (2022) estimate an elasticity of reported wealth of 43 percent after six years. The authors further decompose this response and estimate that approximately 24 percentage points are attributable to mobility. This implies a wealth response equivalent to a semi-elasticity of approximately 10.32 associated with migration alone. Similarly, Agrawal, Foremny, and Martínez-Toledano (2025) examine the effects of regional wealth-tax decentralization in Spain, where the Community of Madrid implemented a 100 percent tax credit that effectively reduced its wealth-tax rate to zero. They estimate a wealth-tax elasticity of 7.96.
More directly relevant to California, Rauh et al. (2026) examine the response to the announcement of the initiative to place the California Billionaire Wealth Tax on the state ballot. They report that at least six billionaires left California before January 1, 2026, the retroactive cutoff date under the proposed tax for determining California residency. This initial group included Google co-founders Larry Page and Sergey Brin. Using the Forbes Real-Time Billionaires List and adjusting the data to reflect the proposed California tax design, the authors compute that these departures represented $536.4 billion of the potential tax base, or approximately 28.3 percent of California billionaire wealth. The implied semi-elasticity is 5.67. Under this scenario, they estimate that approximately $40 billion in wealth-tax revenue would be collected.
Rauh et al. (2026) also identify additional potential movers who have indicated an intention to leave California after the January 1, 2026, cutoff date. Including these individuals increases the number of potential movers to ten billionaires, including Meta founder Mark Zuckerberg. Together, these potential movers account for approximately $785.4 billion, or 41.4 percent of California billionaire wealth. If these individuals successfully avoid the wealth tax through migration, the implied semi-elasticity would increase to approximately 8.30. This estimate falls within the range of estimates reported in the Swiss and Spanish subnational wealth-tax literature described above.
In contrast, Boll, Saez, and Zucman (2026) estimate approximately $100 billion in wealth-tax revenue under an assumption that taxpayers avoid 10 percent of their potential tax liability. This assumption implies a semi-elasticity of approximately 2, substantially below the estimates reported in the empirical literature on subnational wealth taxation. Their calculation also does not incorporate the billionaire departures documented by Rauh et al. (2026). The authors subsequently defend this omission on the grounds that residence information is not publicly available (Galle et al. 2026). Ultimately, however, whether an individual qualifies as a California resident for purposes of the proposed tax would be a matter of legal interpretation and, potentially, judicial determination.
Given the willingness of some billionaires to relocate, it is relevant to ask what broader economic consequences could arise if these departures materialize. To address this question, this paper examines a previous episode of substantial outmigration among high-net-worth individuals and businesses, particularly around 2020–2021. Many billionaires denounced high taxes, heavy regulation, and a hostile political climate as the reasons behind these departures.[8] It is therefore the closest observable event to the potential migration response that a wealth tax might generate.
For example, Elon Musk and Larry Ellison relocated from California near the end of 2020.[9] According to Forbes, Musk’s wealth was approximately $151 billion in 2021,[10] while Ellison’s was approximately $93 billion.[11] Relative to total billionaire wealth in California at the end of 2021,[12] the wealth associated with these two individuals alone represented approximately 20 percent of the state’s billionaire wealth. This magnitude is comparable to the approximately 28 percent of billionaire wealth associated with the initial group of departures identified by Rauh et al. (2026). Other high-profile billionaires and companies also relocated during this period, including Charles Schwab and his company.
Recent economic research identifies several channels through which the departure of high-wealth individuals and firms can affect their origin economies. Jakobsen et al. (2026) find that wealth tax–induced migration by wealthy business owners negatively affects their firms, reducing employment and investment. Related evidence suggests that the loss of entrepreneurs and firms can generate broader local effects. Anelli et al. (2023) find that greater emigration of young and highly educated individuals from Italy reduces local firm creation, while Weinstein (2018) shows that firm relocations generate persistent effects on local employment, wages, and labor force participation.
Migration can also affect innovation by weakening the concentration of entrepreneurial and scientific talent that generates knowledge spillovers. Akcigit, Baslandze, and Stantcheva (2016) show that top inventors respond to tax incentives and that their geographic mobility can alter the distribution of innovative activity. Similarly, Moretti and Wilson (2017) find that state tax incentives affect the location of highly productive scientists, with implications for regional innovation.
Tax revenue represents another important spillover. The departure of high-income taxpayers can reduce state revenue directly through the loss of their income-tax payments and indirectly through reductions in the broader tax base associated with their economic activity. Rauh (2022) estimates that taxpayer migration from California between 2000 and 2020 was associated with approximately $50 billion in cumulative net taxable-income losses and $4.8 billion in forgone state tax revenue. Similarly, Rauh and Shyu (2024) find that behavioral responses to California’s Proposition 30, a 2012 measure raising marginal tax rates by up to 3 percentage points on high-income households, eroded 45.2 percent of the expected windfall revenue within its first year and 60.9 percent within two years.
California billionaires generate an estimated $3.3–$5.8 billion annually in state income tax revenue (Rauh et al. 2026). If billionaires were to relocate in response to the wealth tax, the resulting fiscal externality would include a reduction in income-tax revenue collected by the state. Using a Monte Carlo simulation, Rauh et al. (2026) estimate an expected net present value loss of approximately $24 billion from these fiscal effects, with a 71 percent probability that the net present value of the proposed wealth tax would be negative.
Overall, the literature suggests that high-wealth individual and firm mobility can generate spillovers across multiple dimensions of a regional economy. These include employment and firm creation, investment and productivity, innovation, population, and tax revenue.
3. Research Design
The central empirical challenge is straightforward to state but impossible to answer with raw data alone: What would have happened to California’s economy if its billionaires had stayed in 2021? The state’s actual post-2021 economic performance can be observed directly, but the counterfactual path—the trajectory California would have followed absent the wave of high-net-worth departures—cannot be observed and must instead be estimated.
To estimate this counterfactual, the paper employs the synthetic control method (Abadie, Diamond, and Hainmueller, 2010).[13] Using data from other U.S. states, the method constructs a “synthetic California”: a weighted combination of comparison states chosen so that the resulting composite closely tracks California’s economic trajectory in the years before 2021 but has not experienced a comparable wave of billionaire and corporate departures. Because the technique selects donor weights to match pre-treatment outcomes and covariates as closely as possible, it has become the standard approach in applied economics for evaluating the effect of a policy or shock experienced by a single, large geographic unit.[14]
The analysis is built on an annual, state-level panel covering all fifty U.S. states from 1999 through 2024. California is the treated unit; every other state is a potential member of the donor pool used to construct the synthetic control. The treatment year is based on the number of company headquarters relocating from California, which reached a historic high in 2021.[15] Approximately 350 headquarters left California in 2020, followed by nearly 450 in 2021. Meanwhile, the number of companies relocating their headquarters to California remained relatively stable at around 200 per year. As a result, 2021 recorded the state’s largest net loss of corporate headquarters, marking a sharp acceleration in the relocation trend that had begun the previous year.
The synthetic control is constructed by matching California’s pre-2021 trajectory to a weighted combination of donor states on six covariates: Employment Share, Tax Share, Technology Share, the Fraser Institute’s Economic Freedom Index, Educational Attainment (share of the population with a high school diploma or higher), and the FHFA House Price Index. These covariates are intended to capture the structural features—labor-market size, fiscal capacity, exposure to the technology sector, regulatory and economic-freedom climate, human capital, and housing-cost pressure—most likely to predict a state’s economic trajectory and its own risk of a comparable outflow of high-net-worth residents. The donor pool is unrestricted ex ante; the optimization procedure underlying the synthetic control method selects whichever combination of states best reproduces California’s pre-treatment path on each outcome and covariate, subject to the constraint that donor weights are non-negative and sum to one.
All series are drawn from public sources: the U.S. Bureau of Economic Analysis (BEA) for GDP, its technology-sector component, taxes on production and imports, employment, population, and private research and development; the Federal Reserve Economic Data (FRED) system and the U.S. Census Bureau for population and educational-attainment series; the Federal Housing Finance Agency (FHFA) for its House Price Index; and the Fraser Institute for its Economic Freedom of North America Index. Appendix Table A1 defines each raw variable and its source in full.
Six outcomes are analyzed: the annual growth rate of real GDP; the annual growth rate of real GDP per capita; total employment; taxes on production and imports (TOPI); population; and private research and development spending. Because GDP, employment, taxes, and R&D all scale with the size of a state’s economy and population, several of these raw series are converted into shares or per-capita terms before being used in the analysis: for example, Real GDP per Capita (Real GDP ÷ Population) and Tax Share (Taxes ÷ GDP). The full set of derived metrics and the raw variables from which they are constructed is reported in Appendix Table A2. The summary statistics are shown in Table 1:
Table 1. Summary Statistics
| Variable | Obs. | Mean | Std. dev. | Min | Max |
|---|---|---|---|---|---|
| Real GDP Growth | 1,300 | 2.17 | 2.73 | -10.31 | 23.16 |
| Real GDP per Capita Growth | 1,300 | 1.38 | 2.58 | -9.95 | 20.00 |
| Employment Growth | 1,300 | 1.13 | 1.87 | -10.56 | 8.01 |
| Tax Revenue Growth | 1,300 | 4.44 | 6.05 | -53.53 | 106.37 |
| Population Growth | 1,300 | 0.77 | 0.72 | -5.99 | 4.40 |
| Private R&D | 550 | 8.41 | 18.46 | -50.13 | 139.43 |
| Employment Share | 1,350 | 0.60 | 0.05 | 0.47 | 0.82 |
| Tax Share | 1,350 | 0.07 | 0.01 | 0.04 | 0.18 |
| Tech Share | 1,350 | 0.06 | 0.03 | 0.01 | 0.20 |
| Economic Freedom | 1,350 | 6.08 | 0.97 | 3.57 | 8.35 |
| Educational Attainment | 950 | 88.77 | 3.46 | 77.90 | 95.50 |
| Housing Price Index | 1,350 | 545.78 | 248.46 | 219.08 | 1953.02 |
Once the synthetic California is constructed, its post-2021 performance is compared with that of the actual state. Because the synthetic unit is built to resemble California prior to the exodus, any divergence that emerges after 2021 provides an estimate of the economic effect associated with the departure of billionaires and corporate headquarters that year.
To assess whether the estimated post-treatment effects are unusually large relative to what might arise by chance, this study uses a placebo test. The procedure repeatedly applies the synthetic control method to each state in the donor pool, treating each state as if it had received the treatment in 2021, and compares the resulting placebo gaps with the gap estimated for California. If California’s post-treatment divergence is substantially larger than the divergences observed in the placebo tests, this provides evidence that the estimated effect is unlikely to be driven solely by random variation.
4. Results
The results suggest that California paid a measurable economic price following the 2021 exodus. Compared with its synthetic counterpart, actual California experienced slower economic growth, a decline in income per capita, fewer jobs, lower tax revenue, and a faster decline in population. The analysis also finds weaker, less robust evidence that private-sector research and development growth slowed. The sections below walk through each result in turn. In every figure, the solid red line represents actual California and the dashed blue line represents synthetic California; the shaded gray band marks 2021, the year in which billionaire departures intensified; and, where present, the gold-shaded gap represents the estimated post-2021 divergence between actual and synthetic California.
Graph 1 shows the results for the first outcome variable: economic growth. The synthetic California is constructed primarily from Oregon (36 percent), Washington (28 percent), Georgia (26 percent), and Texas (11 percent), with the donor weights reflecting the states that most closely reproduced California’s pre-treatment characteristics. Following 2021, California’s economy underperformed its synthetic counterpart. In 2022, California’s economy did not grow, while the synthetic control grew by 2.67 percent. The placebo test indicates that this negative effect is statistically significant in 2022, with a standardized empirical p-value of 0.02. Over the full post-treatment period, the cumulative difference amounts to nearly $100 billion in lower economic output relative to the synthetic control.

The same pattern emerges when examining real GDP per capita, as Graph 2 demonstrates. The synthetic control is constructed primarily from Washington (38 percent) and Oregon (37 percent), with additional weights assigned to Massachusetts (16 percent) and Rhode Island (10 percent). Following the 2021 treatment year, California’s real GDP per capita diverged from its synthetic control. In 2022, real GDP per capita in California was essentially flat, while it increased by 1.70 percent in synthetic California. Placebo testing indicates that the negative effect is statistically significant in 2022, with a standardized empirical p-value of 0.10.

As observed in Graph 3, the effects on employment emerge somewhat later than those observed for economic output and GDP per capita. The synthetic control is constructed from Washington (52 percent), Rhode Island (30 percent), and Massachusetts (18 percent). In 2023, California had 1.13 percent fewer jobs than its synthetic control. The placebo test indicates that this negative effect is statistically significant, with a standardized empirical p-value of 0.08. The employment gap persisted through the end of the analysis period, amounting to approximately 78,693 fewer jobs relative to the synthetic control.

Perhaps the most important finding concerns state tax revenue represented in Graph 4. The synthetic control is constructed primarily from Oregon (34 percent) and Texas (34 percent), with additional weights assigned to New Mexico (14 percent), Washington (14 percent), and Georgia (4 percent). Following the 2021 treatment year, California’s state tax revenue fell below its synthetic counterpart, with the divergence becoming statistically significant in 2023. In that year, California’s state tax revenue was 1.66 percent below the level of its synthetic control. Permutation placebo testing yields a standardized empirical p-value of 0.08 for the 2023 effect. The gap persisted through 2024. Across the 2021–2024 period, the cumulative difference corresponds to almost $50 billion less in Taxes on Production and Imports (TOPI), which include business-chargeable tax liabilities such as sales, property, and excise taxes.
This result is relevant to the potential fiscal effects of a wealth tax because the fiscal consequences of billionaire outmigration may extend beyond the direct loss of wealth-tax revenue. If high-wealth residents and the economic activity associated with them relocate, the state may also experience reductions in other sources of tax revenue.

As presented in Graph 5, the effects of outmigration extend beyond high-wealth individuals and businesses to California’s broader population. The synthetic control is constructed primarily from New York (55 percent) and Texas (27 percent), with additional weights assigned to New Mexico (12 percent) and Oregon (7 percent). California’s population growth turned negative in 2020, and the gap between California and its synthetic control widened thereafter. Permutation placebo testing indicates that the negative population effect is statistically significant in both 2023 and 2024, with a standardized empirical p-value of 0.08 in each year. By the end of the analysis period, California’s population was approximately 813,000 below its synthetic counterpart.

The pattern for private research and development (R&D) spending is less conclusive than the results for the other outcomes, as shown in Graph 6. However, the data available for this outcome variable starts from 2013. The synthetic control is constructed primarily from Texas (66 percent) and New York (27 percent), with an additional 7 percent weight assigned to Washington. Following the 2021 treatment year, private R&D spending in California fell below its synthetic counterpart. Yet permutation placebo testing does not indicate that the estimated negative effects are statistically significant. Unlike the other outcomes examined above, the results for private R&D should consequently be interpreted cautiously and do not provide statistically significant evidence of a post-2021 decline relative to the synthetic control.

The synthetic control method rests on the assumption that the treatment does not generate spillover or contagion effects in the donor pool. In this study, this assumption implies that California’s billionaire outmigration does not materially affect the economic outcomes of the other U.S. states used to construct the synthetic control. This assumption may be particularly relevant because several billionaires who have left California have reportedly established fiscal residence in states such as Texas and Florida. To address this concern, a robustness check was conducted by excluding Texas and Florida from the donor pool. The results remain largely unchanged, suggesting that the main findings are not driven by the inclusion of these two states in the synthetic control.
5. Discussion
The proposed California Billionaire Tax Act has already been associated with substantial potential outmigration among the state’s wealthiest residents. Rauh et al. (2026) estimate that billionaires representing at least 28.3 percent of California’s billionaire wealth have either left the state or indicated plans to do so. A comparable episode occurred in the early 2020s, when billionaires representing approximately 20 percent of California billionaire wealth left the state amid a period characterized by high taxes, extensive regulation, and a contentious political environment.
This paper examines that earlier episode to assess California’s subsequent economic performance relative to a synthetic counterfactual constructed from comparable states. The 2021 episode provides an empirical benchmark for assessing how California’s economy evolved following a substantial departure of high-net-worth individuals and major corporations. The results indicate that California subsequently underperformed its synthetic counterpart across several dimensions, including economic growth, employment, income per capita, population, and state tax revenue, while the evidence for private R&D is weaker. In particular, the estimated gap in taxes on production and imports (TOPI) approaches $50 billion over the post-treatment period.
These findings highlight that the fiscal consequences of billionaire outmigration may extend beyond the direct loss of wealth-tax revenue. If the proposed wealth tax induces additional migration, the associated effects could operate through multiple channels, including changes in investment, employment, entrepreneurial activity, innovation, consumption, population, and the broader state tax base. The magnitude of these effects would depend on the number and characteristics of individuals who leave, the extent to which their businesses and economic activities relocate with them, and the ability of other taxpayers and firms to replace the associated economic activity. The results should therefore be interpreted as evidence of the potential consequences of substantial outmigration rather than as a forecast of the precise effects of the proposed wealth tax.
An important limitation is the coincidence of the treatment period with the COVID-19 pandemic. The pandemic affected economic activity throughout the United States and could have influenced California particularly strongly. However, because the synthetic control is constructed from states that were also exposed to the national pandemic shock, the methodology is designed to account for common changes affecting California and its donor states. The results should nevertheless be interpreted with caution if California experienced pandemic-related shocks that differed systematically from those affecting the donor pool.
Overall, the findings suggest that evaluating the proposed wealth tax requires considering both its potential revenue from the tax itself and the broader economic and fiscal effects associated with behavioral responses. The 2021 episode does not provide a direct estimate of the effect of the proposed tax, but it offers a useful empirical benchmark for understanding the potential state-level consequences of a substantial increase in billionaire and corporate outmigration.
Appendix: Variables, Sources, and Definitions
| Variable | Definition | Source |
|---|---|---|
| GDP | Gross domestic product; estimates the value of goods and services produced in a state. | U.S. Bureau of Economic Analysis |
| Real GDP | Real GDP, in millions of chained 2017 dollars. | U.S. Bureau of Economic Analysis |
| GDP Tech Sector | Publishing (except internet), data processing/hosting, computer systems design, computer and electronic manufacturing, and telecommunications industries. | U.S. Bureau of Economic Analysis |
| Population | Number of people living in a state. | U.S. Bureau of Economic Analysis |
| Taxes | Taxes on production and imports (TOPI), in millions of dollars. | U.S. Bureau of Economic Analysis |
| Employment | Total employment (number of jobs). | U.S. Bureau of Economic Analysis |
| Economic Freedom of North America | Index measuring the degree to which North American governments permit citizens to make their own economic choices. | Fraser Institute |
| Education | Educational attainment; share of population with a high school diploma or higher. | Federal Reserve Economic Data (FRED) |
| Private Research & Development | R&D activity measured by location of R&D production (performance), not necessarily the location of R&D funding. | U.S. Bureau of Economic Analysis |
| House Price Index (HPI) | A weighted, repeat-sales index measuring average price changes in repeat sales or refinancings of the same properties. | Federal Housing Finance Agency |
Derived Metrics
| Metric | Definition |
|---|---|
| Real GDP per Capita | Real GDP ÷ Population |
| Tech Share | GDP Tech Sector ÷ GDP |
| Tax Share | Taxes ÷ GDP |
| Employment Share | Employment ÷ Population |
| Growth | Year-over-year percentage change |
References
Abadie, Alberto. 2021. “Using Synthetic Controls: Feasibility, Data Requirements, and Methodological Aspects.” Journal of Economic Literature 59, no. 2: 391–425. https://doi.org/10.1257/jel.20191450.
Abadie, Alberto, Alexis Diamond, and Jens Hainmueller. 2010. “Synthetic Control Methods for Comparative Case Studies: Estimating the Effect of California’s Tobacco Control Program.” Journal of the American Statistical Association 105, no. 490: 493–505. https://doi.org/10.1198/jasa.2009.ap08746.
Agrawal, D. R., Foremny, D., and Martínez-Toledano, C. 2025. “Wealth Tax Mobility and Tax Coordination.” American Economic Journal: Applied Economics 17, no. 1: 402–30. https:/doi.org/ 10.1257/app.20220615.
Akcigit, Ufuk, Salomé Baslandze, and Stefanie Stantcheva. 2016. “Taxation and the International Mobility of Inventors.” American Economic Review 106, no. 10: 2930–81. https:/doi.org/ 10.1257/aer.20150237.
Anelli, M., Basso, G., Ippedico, G., & Peri, G. 2023. “Emigration and Entrepreneurial Drain.” American Economic Journal: Applied Economics, 15, no. 2: 218–52. https:/doi.org/ 10.1257/app.20210194.
Bohn, Sarah, David Neumark, Zeya Li, and Vicki Hsieh. 2025. Policy Brief: Are Company Headquarters Leaving California? Public Policy Institute of California. https://www.ppic.org/pdf/policy-brief-are-company-headquarters-leaving-california.pdf.
Boll, Jasper, Emmanuel Saez, and Gabriel Zucman. 2026. “California Billionaires: Wealth, Taxes, and Wealth Tax Revenue Estimates.” NBER Working Paper No. 35218. National Bureau of Economic Research, May 2026, revised July 2026. https:/doi.org/10.3386/w35218.
Brülhart, Marius, Jonathan Gruber, Matthias Krapf, and Kurt Schmidheiny. 2022. “Behavioral Responses to Wealth Taxes: Evidence from Switzerland.” American Economic Journal: Economic Policy 14, no. 4: 111–50. https:/doi.org/10.1257/pol.20200258.
California Office of the Attorney General. 2025. “California Billionaire Tax Act.” Initiative 25-0024A1. Filed 2025. https://oag.ca.gov/system/files/initiatives/pdfs/25-0024A1%20%28Billionaire%20Tax%20%29.pdf.
Galle, Brian, David Gamage, Emmanuel Saez, and Darien Shanske. 2026. “Response to The Net Present Value of the Billionaire Tax Act.” University of Missouri School of Law Legal Studies Research Paper 2026-31. March 17. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6433598.
Jakobsen, Katrine, Henrik Kleven, Jonas Kolsrud, Camille Landais, and Mathilde Muñoz. 2024. “Taxing Top Wealth: Migration Responses and Their Aggregate Economic Implications.” Working Paper No. w32153. National Bureau of Economic Research, 2024. https:// https://www.nber.org/papers/w32153.
Lair, Don. 2026. “8 Billionaires Who Fled California—and One Who Didn’t.” Yahoo! Finance. May 3. https://finance.yahoo.com/economy/policy/articles/8-billionaires-fled-california-wealth-143341289.html.
Moretti, Enrico, and Daniel J. Wilson. 2017. “The Effect of State Taxes on the Geographical Location of Top Earners: Evidence from Star Scientists.” American Economic Review 107, no. 7: 1858–1903. https:/doi.org/ 10.1257/aer.20150508.
O’Keefe, Ross. 2026.“Here Are the Companies That Have Fled California—Including the Company That’s Tracking the Exodus.” New York Post. April 9. https://nypost.com/2026/04/09/us-news/these-companies-have-shifted-their-headquarters-out-of-california.
Rauh, Joshua D., Benjamin Jaros, Gregory Kearney, John Doran, and Matheus Cosso. 2026. “The Fiscal Effects of the California Billionaire Tax Act.” Hoover Institution. March 4. https://www.hoover.org/research/net-present-value-billionaire-tax-act-assessment-fiscal-effects-californias-proposed.
Rauh, Joshua, and Ryan Shyu. 2024. “Behavioral Responses to State Income Taxation of High Earners: Evidence from California.” American Economic Journal: Economic Policy 16, no. 1: 34-86. https:/doi.org/ 10.1257/pol.20200500.
Rauh, Joshua D. 2022. “Taxes, Revenues, and Net Migration in California.” Stanford Graduate School of Business, Hoover Institution, December 22. https://web.stanford.edu/~rauh/TRNMCDec2022.pdf.
Scheuer, Florian, and Joel Slemrod. 2021. “Taxing Our Wealth.” Journal of Economic Perspectives 35, no. 1: 207–30. https:/doi.org/ 10.1257/jep.35.1.207.
Weinstein, Russell. 2018. “Dynamic Responses to Labor Demand Shocks: Evidence from the Financial Industry in Delaware.” Journal of Urban Economics 106: 27–45. https:/doi.org/ 10.17848/wp17-276.
Notes
[1] California Office of the Attorney General, “California Billionaire Tax Act,” Initiative 25-0024A1 (2025) (filed), https://oag.ca.gov/system/files/initiatives/pdfs/25-0024A1%20%28Billionaire%20Tax%20%29.pdf.
[2] Jasper Boll, Emmanuel Saez, and Gabriel Zucman, “California Billionaires: Wealth, Taxes, and Wealth Tax Revenue Estimates,” NBER Working Paper No. 35218 (National Bureau of Economic Research, May 2026, revised July 2026), https:/doi.org/10.3386/w35218.
[3] See, for example, coverage of departing California billionaires in Don Lair, “8 Billionaires Who Fled California Before the Wealth Tax Hit—and One Who Didn’t,” Yahoo! Finance, May 3, 2026, https://finance.yahoo.com/economy/policy/articles/8-billionaires-fled-california-wealth-143341289.html.
[4] Julia Hornstein and Jack Newsham, “Google Cofounders Sergey Brin and Larry Page Moved an LLC out of California Ahead of Proposed Billionaire’s Tax,” Business Insider (via Yahoo! Finance), January 8, 2026, https://ca.finance.yahoo.com/news/california-gov-gavin-newsom-signs-224851354.html.
[5] Stacy Perman, “The Exodus of California’s Tech Billionaires from the Golden State to Florida’s Gold Coast,” Los Angeles Times, March 11, 2026, https://www.latimes.com/entertainment-arts/business/story/2026-03-11/inside-exodus-of-california-tech-billionaires-to-florida.
[6] Sarah Bohn, David Neumark, Zeya Li, and Vicki Hsieh, Policy Brief: Are Company Headquarters Leaving California? (Public Policy Institute of California, 2025), https://www.ppic.org/pdf/policy-brief-are-company-headquarters-leaving-california.pdf.
[7] Ross O’Keefe, “Here Are the Companies That Have Fled California—Including the Company That’s Tracking the Exodus,” New York Post, April 9, 2026, https://nypost.com/2026/04/09/us-news/these-companies-have-shifted-their-headquarters-out-of-california.
[8] Ibid.
[9] Ibid.
[10] “Elon Musk,” Forbes, accessed October 1, 2026, https://www.forbes.com/profile/elon-musk.
[11] “Larry Ellison,” Forbes, accessed October 1, 2026, https://www.forbes.com/profile/larry-ellison.
[12] “Total California Billionaire Wealth (2019 – Present),” California Billionaire Wealth Tracker, accessed October 1, 2026, https://cabillionairetracker.org.
[13] Alberto Abadie, Alexis Diamond, and Jens Hainmueller, “Synthetic Control Methods for Comparative Case Studies: Estimating the Effect of California’s Tobacco Control Program,” Journal of the American Statistical Association 105, no. 490 (2010): 493–505, https://doi.org/10.1198/jasa.2009.ap08746.
[14] Alberto Abadie, “Using Synthetic Controls: Feasibility, Data Requirements, and Methodological Aspects,” Journal of Economic Literature 59, no. 2 (2021): 391–425, https:/doi.org/10.1257/jel.20191450.
[15] Bohn, Neumark, Li, and Hsieh, Policy Brief: Are Company Headquarters Leaving California?



























