Response to “The Net Present Value of the Billionaire Tax Act”

Response to “The Net Present Value of the Billionaire Tax Act” March 4, 2026

Brian Galle (UC Berkeley), David Gamage (University of Missouri), Emmanuel Saez (UC Berkeley), Darien Shanske (UC Davis)

This report responds to the study by Rauh et al. (2026) published on March 4, 2026 that purports to examine the fiscal consequences of the 2026 California Billionaire Tax Act (CBTA) ballot initiative (full text here) and to criticize our expert report.

Rauh et al.’s analysis and conclusions simply do not rest on an estimate of the CBTA as written. The CBTA is a one-time 5% billionaire wealth tax, payable over five years and so roughly equivalent to a 1% annual tax assessed for only five years plus an interest-deferral charge. Yet Rauh et al. estimate taxpayer mobility responses as though taxpayers face a recurring 5% annual wealth tax as a permanent ongoing tax. Their revenue estimate of that policy also depends on a crucial and highly unlikely assumption: they count the income tax loss due to billionaires leaving in perpetuity because of this large recurrent 5% annual tax as a cost but on the revenue side, they assume that the wealth tax applies only once. This unreasonable asymmetry is the largest flaw in their analysis.

On top of this, the Rauh et al. analysis is also based on incorrect claims about California law and about who counts as a California resident for tax law purposes. There is no real evidence that any significant number of individuals on the Forbes billionaire list for California at the end of 2025 will avoid the CBTA under the legal standard that will actually apply. Below we elaborate as to the specific issues.

Analyzing an imaginary permanent 5% wealth tax

The CBTA is a one-time tax on the net wealth of California billionaires. The rate is 5%, but individuals can elect to spread payment over five years, yielding an annual burden of 1% per year, plus an interest charge. We estimate that this tax raises $100 billion with comparatively very modest reductions in California income tax revenue.

In contrast, Rauh et al. (2026) estimate the mobility effects of a permanent and annual 5% tax on billionaire wealth (using behavioral elasticities from the empirical literature on mobility responses to wealth taxes). On the cost side, they count the income tax loss due to billionaires leaving in perpetuity but on the revenue side, they assume that the wealth tax applies only once. Yet if the tax were permanent then surely the revenue it would raise should be included, which would of course add hundreds of billions of dollars in revenue. To emphasize, the Rauh et al. estimate depends on the assumption that the costs of the tax will last forever, but the revenues will be one time only.

1

By way of apparent justification for analyzing a different policy than the one that is actually in the CBTA, Rauh et al. appear to argue that some taxpayers may assume that the wealth tax would be permanent (and ongoing with a 5% annual rate assessed every year). They do not explain the basis for that assumption. Under a standard model of rational expectations, taxpayers’ behavioral responses are tied to expectations with quick corrections if these expectations are not met. If the wealth tax turned out not to be permanent, then surely the model should reflect some updating of taxpayer assumptions. If the CBTA is not followed by another wealth tax, taxpayers will have no reason to act as though the CBTA is permanent and the Rauh et al. projections would grossly overstate taxpayers’ behavioral responses.

Put another way, we think it is highly unlikely that taxpayers would act as though a tax were permanent unless that tax were, in fact, permanent. If so, then a serious estimate of the revenue effect of the tax should include those permanent future revenues. Yet Rauh et al. appear to assume that the supposedly permanent tax they analyze will not collect a single dollar of revenue in future years. (This could only be true if every single billionaire left California in the first year and with no new taxpayers becoming billionaires in California thereafter, but there is absolutely no justification for assuming or concluding this from Rauh et al.’s analysis, in addition to this being completely implausible).

Although Rauh et al.’s assumptions are implausible, the underlying notion that either the introduction or passage of the CBTA might affect expectations as to future policy is plausible and would be worth analyzing with a serious methodology. Such an analysis would be inherently speculative, and should thus be considered along with other speculative possible effects of the CBTA, both positive and negative, such as potential effects as to combating harms resulting from rising inequality. Most importantly, such an analysis should also include estimates of the economic effects of the CBTA’s revenues being used to address California’s expected health care crisis. Regardless, even if it might be plausible to speculate that either the introduction or the passage of the CBTA might increase expectations as to the likelihood of a future wealth tax, it is not plausible to assess this as though taxpayers would expect a guaranteed future tax at five times the effective annual rate of the CBTA enacted immediately and sustained permanently. Further, even under their assumptions on mobility responses, a 5% recurrent annual tax would raise much more revenue than what is lost through income tax losses (see below).

Further, the enactment of the CBTA has no legal effect on the ability of the California legislature or the voters to enact a subsequent wealth tax. If voters believe that the legislature or the voters are likely to adopt additional wealth taxes in the future, then they should already hold those beliefs and the CBTA should have only at most modest incremental effects on those expectations.

Most critically, there is no plausible justification for assuming that taxpayers would expect a 5% annual tax, rather than the roughly 1% annual tax imposed by the CBTA. Even if

2

somehow the CBTA made a future tax more likely, and if that future tax were for some reason the exact same as the CBTA, then voters should expect a 1% annual tax rate.

This is a point worth emphasizing again: almost the entirety of the Rauh et al. analysis rests on the assumption that voters treat the CBTA as informative of the likelihood of a future wealth tax, but their supposed analysis offers no explanation for why that future wealth tax would be a 5% annual tax. Why wouldn’t the future tax rate voters expect be .4% (the cap currently in the California Constitution)? Or 1%, the effective rate in the CBTA? If it were, their estimates would be vastly different.

As a result of this mix and match methodology and series of hidden and undefended assumptions, including counting losses as permanent and using their 1.5% discount rate, the annual income tax loss is magnified by a factor 67 - almost two orders of magnitude. This explains the negative net present value they find.1

Incorrect assumption that substantial movement has already occurred

Rauh et al. also significantly undercount the revenue potential of the CBTA because they assume that “[n]early 30% of the Act’s wealth tax base has already departed.” This high number is based on press reports about billionaires claiming that they have taken steps to try to change their residency either before January 1, 2026 (6 billionaires listed in their Table 6) or after January 1, 2026 and up to early March (4 additional billionaires listed in their Table 7). These assumptions misunderstand both the CBTA and California residency law. Buying real estate, opening office space out ofnstate, or announcing a move is not the same as successfully severing California tax residence which requires, as a matter of law, uprooting most of the individual’s personal and business ties to California.

If voters approve the CBTA, it will fall on any billionaire who was a California resident as of January 1, 2026. Residence for purposes of the CBTA is based on longstanding residence rules under California’s personal income tax. Among other aspects of this approach, which is almost a century old, California residents are presumed to remain in California until they are able to demonstrate that their absence from California was not temporary. When California courts make that determination, they look to whether a person’s closest

1 Rauh et al. (2026) state that “The net present value of the Act is −$24.7 billion.” That is, they say the present value of permanently lost income tax revenue more than offsets the one-time wealth tax collections. For a sense of how improbable that number is, consider that using the elasticity they employ to reach that conclusion, the revenue-maximizing permanent annual wealth tax rate would be nearly 10%. (Elasticity roughly refers to how much taxpayers are likely to respond to a tax). Formally, the revenue maximizing rate for a permanent wealth tax is given by the inverse of the semi-elasticity which Rauh et al. assume to be around 10 (see Boll, Saez, and Zucman 2026 for details).

3

personal, business, family, and political ties are to California, or instead to somewhere else, Thus, simply buying real estate, opening office space out of state, or announcing a planned move is not the same as successfully severing California tax residence. As courts recognize, it is very easy to claim to move, but very hard to actually do it. Many Californians have claimed to have “moved” for tax purposes, only to find that their move was not effective (at least not as early as they hoped).

Given that the CBTA is a one-time tax in 2026, there is an obvious and powerful incentive for a potential taxpayer to take up, or claim that they have taken up, a temporary abode somewhere else. Billionaires already own, and can readily purchase, new vacation homes elsewhere, and that will be obvious to courts. In light of the obvious incentives and opportunities for billionaires to misrepresent their residency plans, the close attention California tax authorities can give to each of the two hundred or so CBTA taxpayers, and the inherent difficulty of resettling deep personal, business, and other ties in such a short period, it is unlikely that many billionaires have actually taken the steps necessary to establish a change of residency as a matter of law. An analysis by experts in California residency tax planning legal services explains this in persuasive detail.2

Put another way, little in the current public record offers real reason to think that any significant number of individuals on the Forbes billionaire list for California at the end of 2025 will avoid the CBTA. Certainly we see no evidence to suggest that individuals who now say they are spending time elsewhere will remain outside the state, rather than, say, returning to Silicon Valley in January of 2027. At best, Rauh et al. offer a kind of worst-case scenario for the CBTA’s revenue; we do not believe that presenting only the worst case is a reasonable way to project the CBTA’s effects.

Finally, from press reports and direct discussions, some billionaires apparently initially misunderstood the text of the initiative and thought they would be taxed based on their voting share in their business rather than the economic value of their share. As a result, many, including the Google founders Larry Page and Sergei Brin, incorrectly thought their tax liability would be much more than the 5% tax. We hope that this issue has now been cleared up (see our explanation) and that billionaires now understand that the tax will never exceed 5% of the economic value of their wealth. If this was indeed a pivotal reason for wanting to leave, it is possible that some billionaires who announced their intent to depart will revisit their decision, especially as it also becomes clear, as we argue above, that these paper moves were not likely effective anyway.

2 Rauh et al. argue that billionaires who cannot establish residency by January 1 elsewhere may be hoping that the January 1, 2026 date will be eventually struck down as unconstitutional. This is an unlikely scenario, as we have explained in depth elsewhere. In short, the current residency rule is longstanding in California and other states, and the Supreme Court has since the Nineteenth Century repeatedly approved tax laws that extend back to the beginning of the calendar year or further.

4

Smaller Points

California income tax paid by billionaires. Rauh et al. (2026) estimate that California billionaires pay $4.55 billion (in their central scenario) or about .24% of their wealth in annual California income tax in 2025 while we estimate that they pay around $3 billion per year. This moderate size discrepancy arises primarily because they assume that the 12 billionaires have reported individual income equal to 78% of the highest 212 income earners. We assume, based on direct evidence from Balkir et al. (2025) matching US wide billionaires to income tax data, that billionaires’ income is only 50% of the reported income of the highest income earners3. This discrepancy however is relatively modest (compared to the scoring issues discussed above).

Rauh et al. (2026) also assume that the California income tax loss due to billionaires leaving is proportional to their wealth with Google founders Larry Page, Sergei Brin, and Facebook founder Mark Zuckerberg accounting for slightly over 90% of the wealth that has left (in both their shortlist in Table 6 and their longer list in Table 7). Using public filings on stock sales, executive compensation, and dividends paid out, Boll, Saez, and Zucman (2026) find that the California centi-billionaires pay a much lower California income tax (relative to their wealth) than mere California billionaires overall. In particular, they find that Page, Brin, and Zuckerberg paid only $269 million4 in

California income tax between the three of them in 2025, which is only .04% of their wealth in 2025, much less than the .24% of wealth paid by “ordinary” California billionaires overall estimated by Rauh et al. (2026). As a result, even if these 3 centibillionaires have or will depart California, something we very much doubt, the consequences for California income tax collections are almost negligible ($269 million is about .2% of the $120 billion expected total California income taxes collected in 2025).

Is Larry Ellison a California resident? Forbes listed Larry Ellison, whose wealth at the end of 2025 was $245 billion, slightly over 10% of total CA billionaire wealth, as a California resident up to Jan 10, 2026, at which point Forbes started listing him as a Florida resident. Forbes had also listed him as a Hawaii resident for a couple of years during COVID. Rauh et al. (2026) argue that he left California years ago. The reality is that California residency for tax purposes is not public information and hence we cannot know for sure. We followed the Forbes list which is based on careful investigative reporting but we recognize that it is possible that Larry Ellison, who had already taken steps to move out of California well before the billionaire tax initiative, may not be a resident by January 1, 2026, and hence not liable for the tax.

4 Boll, Saez, and Zucman (2026) provide those estimates annually. On average from 2019 to 2025, Brin, Page, and Zuckerberg paid only $222 million per year in California income taxes.

3 This 50% number is also what comes out when analyzing Survey of Consumer Finance data that provides information on both reported income and wealth for a representative sample of very high wealth/high income families (see Saez and Zucman 2019, p. 466).

5

Conversely, it is likely that Forbes misses some California billionaires as well, either heirs of old fortunes who are not as visible as they tend to have diversified wealth or new billionaires who haven’t yet attracted public attention. For example, the Wall Street Journal listed 255 California billionaires in 2024 when Forbes listed only 196.

Non US-citizens California residents. Both Rauh et al. (2026) and our report failed to include the California billionaires who are not US-citizens but liable for California income (or wealth) tax as residents.5 Most of them are very long-term residents of California. There are 24 of them at the end of 2025 with a total wealth of $110 billion, about 5% of the total, that should be included in the revenue estimates, an omission we will correct in our next iteration.

Concluding Points

Rauh et al. can reach their extreme conclusions only through a series of improbable assumptions and misunderstandings large and small. Taken on its own terms, the policy they analyze (which is not, on any set of reasonable assumptions, the effects of the actual CBTA) would raise hundreds of billions of dollars. Our prior analysis as well as nonpartisan analysis by California’s Legislative Analyst's Office each conclude those policies would on net raise tens of billions to offset California’s health-care funding crisis.

5 The Forbes data provides citizenship information but only state level residence for US and Chinese residents. This explains why many users of the Forbes data (including ourselves) fail to include non-US citizens in their analysis.

6