💬 Discussion

California Billionaire Tax

On 3 November 2026 California votes on Proposition 40, a one-time tax on the net worth of the state's billionaires, most of it for Medi-Cal [1]. The initiative comes from the healthcare workers' union SEIU-United Healthcare Workers West. Governor Gavin Newsom, the teachers' union, doctors and the state's taxpayer association oppose it, and two competing measures on the same ballot could stop it even with a majority [1].

  • 213 billionaires lived in California on 1 January 2026, with a combined net worth of about 2,180 billion dollars [4]
  • 3.2 billion dollars of state income tax a year is what billionaires paid between 2023 and 2025 [4][5]
  • 72 percent of that wealth is listed stock, whose value is public [4]

The proposal

Charge everyone who lived in California on 1 January 2026 with a net worth above 1 billion dollars a one-time tax of 5 percent, and spend nine tenths of it on health care.

State (California) A measure of one state, California, not of the federal government. All figures are scored on the same global scale as the federal measures in this area, so they come out smaller.

The debate

Supporters, led by healthcare and teachers' unions, want the money to replace the federal health funding that the budget law of July 2025 took away. They argue that billionaires pay lower tax rates than working families. Opponents warn that the tax drives wealthy residents and their companies out of the state, leaves a lasting hole in the income tax and brings only temporary money. No legislature voted on it; the decision lies with the voters on 3 November 2026.

  • “Prop. 40 makes approximately 200 California billionaires pay a modest, one-time tax on their global wealth to keep local hospitals and ERs open and healthcare affordable for millions of Californians.”

    argues for: Suzanne Jimenez, SEIU-United Healthcare Workers West Argument in favor, Official Voter Information Guide

  • “These residents create new businesses, drive our economic growth and already pay nearly half the income taxes that fund state programs. If they leave, it will create a huge hole in our budget, threatening core services like education, healthcare and public safety or forcing other taxpayers to make up the difference.”

    argues against: René Bravo, California Medical Association Argument against, Official Voter Information Guide

The analysis

California Billionaire Tax

82% of the weight is for, 18% against.

An argument's weight combines three questions: how big the effect is, how certain it is and how much the good at stake matters. The larger the slice, the more the argument weighs.

For

Against

The California billionaire tax is a good proposal on the current analysis.

Very likely better

Very likely better: Under the assumptions used, the proposal brings more than it costs in more than 99 of 100 model runs.

 Overall effect: large

How strongly the proposal works overall, in five steps from very small to very large, comparable across all debates. The ring is always full, even when the overall effect is small; this step says how large it is.

Like the unions behind it, the analysis comes out in favor of the tax, though fairness is not what carries it. The bulk of the weight is money changing hands from a few hundred fortunes to the state budget, so the result turns on where that money does more good. A dollar for Medi-Cal does far more than one in a fortune that regrows within weeks, and the school money adds a gain the campaign barely mentions. The opponents' strongest point, the income tax lost through departures, stays small because liability was fixed before anyone could leave. Only departures on the scale the taxpayer association estimates, lasting for decades, would turn the result.

In detail

All arguments

Arguments for3 arguments

Money for Medi-Cal

17points

About 200 people would pay a one-time tax of 5 percent of their net worth, and nine tenths of it would go to health care for people on low incomes. After departures and avoidance, about 70 billion dollars would arrive between 2027 and 2031.

Value 5 · Public financesImpact 6.1Plausibility 5.5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is money reaching the state of California, nine tenths of it earmarked for Medi-Cal and the hospitals and clinics that serve its patients, the rest for schools and food assistance [3]. Public money is priced at the middle of the scale and at the weight of a household at median income, as everywhere on this site. The measure forbids using the money to replace existing funding, but a state budget is one pot: whether a dollar keeps somebody insured, pays a hospital for care it would otherwise write off, or spares another programme a cut depends on what the state would have done without it. That is why the money is counted as public money rather than at the weight of the poorest households. California loses federal Medicaid funds under the 2025 federal budget law whatever the vote decides [21]: the state's analysts expect the first cuts in 2027 and larger ones from 2028 [33], and across the country three quarters of the reductions fall in 2030 to 2034 [34]; that loss belongs to the comparison without the measure, and the levy would make up part of it for some years. The measure sets no date by which the money must be spent, and the state may keep part of it in reserve [3]. A cross-check against this site's Medicaid evaluations comes out close to the public-money weight: a dollar that keeps a low-income adult covered buys deaths avoided, illness treated and protection against medical bills worth roughly a dollar when added up [19][20]. The people paying appear as the argument opposite. Nothing is counted for the view that the very rich should pay a larger share; that view is what the difference in weights between the two arguments expresses, and counting it again would count it twice. The value is the middle of the scale, because what reaches the state is money, whatever it is spent on.

Impact

The drafters count 213 California billionaires with 2,180 billion dollars of net worth on 1 January 2026, and five percent of that is 109 billion dollars [4]. Two things reduce it. Six billionaires, among them two of the richest people in the world, moved their residence out of the state before the cut-off date [7]; one critical study counts seven departures with 28.6 percent of the base [6], while among the four richest the drafters see only one who could plausibly show he left in time [5]; a quarter is taken off here, in a range from 5 to 30 percent. Then undervaluation, avoidance and tax paid through the deferral account only after the horizon: the drafters assume a tenth, because 72 percent of the wealth is listed stock whose value is public [4], while studies of annual wealth taxes find much larger responses where assets are not reported by banks or brokers [10][11]; 15 percent is used, in a range from 10 to 30. That leaves 69.5 billion dollars collected between 2027 and 2031. The drafters estimate about 100 billion [4]; the critical study arrives at 59 billion before it halves the figure for the risk of losing in court, a risk this evaluation carries in the Plausibility instead [6]; the official analysis says tens of billions [1]. Two smaller receipts come on top. Whoever pays in five annual instalments owes a charge of 7.5 percent a year on the unpaid rest, 15 percent over all five years [3]; a credit line against a large securities account costs only a few points above the reference rate, about 5 to 6 percent in 2026 [32], so instalments pay mainly for those who cannot borrow. How many will choose them has not been measured; a tenth of the tax is used, in a range from none to three tenths, 1.0 billion dollars, and across that whole range the result moves by less than one point. And a third of the tax is expected to be paid by selling shares, most of whose value is untaxed gain [5]; the state income tax on those sales would be 2.5 billion dollars, but owners will sell what costs them least tax, and part of these shares would have been sold within the decade anyway. Studies of state taxes on capital gains find that most of a change in sales persists rather than only shifting in time [35], but no study measures sales forced by a one-time tax; a fifth is counted as additional, in a range from 5 to 40 percent, 0.5 billion dollars, and the whole range moves the result by a quarter of a point at most. Together that is 71.0 billion dollars, spread over the ten years of this evaluation 7.1 billion dollars a year, in a range from 5.7 to 8.5 billion. The Impact is the largest here because it is the tax itself, counted at the weight of public money.

▸ Show calculation ▾ Hide calculation
2,180 billion dollars
Net worth of California's billionaires on 1 January 2026
[4]
× 5 % statutory rate above 1.1 billion dollars
109 billion dollars
One-time tax
× 0.75 Setting, range 0.70 to 0.95: seven departures hold 28.6 percent of the base by one count, far less if their moves fail under the anti-avoidance rule
81.8 billion dollars
Left before the cut-off date
[5][6][7]
× 0.85 Setting, range 0.70 to 0.90: the drafters assume 10 percent loss, annual wealth taxes show more where assets are self-reported
69.5 billion dollars
Undervaluation, avoidance and payment after the horizon
[4][10][11]
+ 69.5 × 0.10 × 15 % Setting, range 0 to 0.30 of the tax paid in instalments: 7.5 percent a year on the unpaid rest over five instalments, while a credit line against shares costs about 5 to 6 percent
1 billion dollars
Deferral charge on instalments
[3][32]
+ 69.5 ÷ 3 × 0.8 × 13.3 % × 0.2 Setting, range 0.05 to 0.40: part of the sales would have happened within the decade; no study measures sales forced by a one-time tax
0.5 billion dollars
Additional income tax on shares sold to pay
[5][35]
= 69.5 + 1.0 + 0.5
71 billion dollars
Received in all
÷ 10 years collected 2027 to 2031, counted against ten years of lost income tax
7.1 billion dollars a year
Spread over the horizon
÷ 1.16 dollars to the euro exchange rate used throughout this evaluation
6.12 billion euro-equivalent a year
In euro-equivalent
× 1.0 the standard weight for a public budget
6.12 billion euro-equivalent a year
Weight of public money
÷ 1 billion euro-equivalent a point scale of this evaluation
6.12
Normalised Impact
Score 6.12 Impact × 5 Value × 5.5 Plausibility ÷ 10 = 17 of 100

Plausibility

The amount is arithmetic: a statutory rate applied to wealth valued on a fixed date, most of it at a market price, so the payment follows from the law wherever the tax stands. The counterfactual is the measure rejected, under which nothing is collected. What holds the Plausibility a little above the middle is whether the tax stands at all. It amends the state constitution, but critics argue that a residency date ten months before the vote, the taxation of people who left during 2026 and the attribution of a resident's entire worldwide wealth to California breach the federal constitution, and that the tax could fall entirely [15]; the drafters reply that no tax has been struck down as retroactive since 1928 and that the apportionment is fair [16]. No neutral assessment puts a number on the outcome, and a challenge is expected in any case, which could delay collection. A remaining risk that the tax is struck down therefore stays with this evaluation. Two competing measures on the same ballot could also stop it from taking effect [1][23]; this evaluation scores the measure as if it takes effect. The confounder for the amount is the valuation of private companies, which the measure sets by formula and which either side can rebut with an appraisal; that is carried in the range. Reverse causation does not arise from a statutory rate. The Plausibility is a little above the middle: the arithmetic is certain where the tax applies, and whether it survives the courts is open.

Evidence: Projection · plausibility at most 6 Evidence of effect: Definitional

A better start at school

4.1points

A tenth of the money goes to schools and food assistance, after the cost of administration. Studies of school spending across the United States find that extra money per pupil raises test scores and college-going, most of all for children from poorer families, whose path upward runs through what they learn and earn.

Value 8 · Educational opportunityImpact 0.9Plausibility 5.5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is what pupils in California's public schools learn because their schools have more money, and what they later earn from it. This site values educational opportunity one step below health and life, because it decides whether a young person can make a living of their own. For a child without family wealth, learning and work are the only route upward: absolute mobility in the United States, the share of children who out-earn their parents, fell from about 90 percent for those born in 1940 to about 50 percent for those born in the 1980s [29]. The earnings are counted at face value, without an income weight; the position of the children concerned is carried by the value. The money that pays for the schooling is counted in the argument on Medi-Cal and the argument opposite it, at the weight of public money. What is counted here is only the later earnings, a stream that the schooling builds and that outlasts the spending, which is why it stands beside the money rather than replacing it. Food assistance is not counted here: what it buys is in the weight of the money. The value is high, because the stream is a young person's chance to earn a living.

Impact

After the cost of administration, a tenth of the 70.5 billion dollars paid into the fund goes to schools and food assistance, and the measure does not fix the split; 60 percent for schools is used, in a range from 30 to 80 percent, because California already pays for free school meals for every pupil: 4.22 billion dollars over about five years, or 0.84 billion a year, under one percent of what the state's public schools spend in a year [1][3]. The meta-analysis of the American studies of school spending finds that 1,000 dollars more per pupil, in 2018 dollars and sustained for four years, raises test scores by 0.0316 standard deviations and college-going by 2.8 percentage points, more for poorer pupils [26]. Prices have risen by about a third since 2018, so the 4.22 billion dollars buy what 3.15 billion bought then. At 0.0079 standard deviations for every 1,000 dollars spent on a pupil, in a range from 0.004 to 0.012, that gives about 25,000 standard deviations, summed over all the pupils who receive a share. The measure forbids replacing existing school funding, but a state budget can still move other money away from schools, so four fifths of the money are counted as additional, in a range from one half to all of it: about 20,000 standard deviations. A lasting standard deviation of achievement is counted like 23.5 percent of a lifetime of earnings, set at about 2.3 million dollars for the United States, which gives a Normalised Impact of 0.93 a year over the horizon, in a range from 0.40 to 2.16. College-going is not added, because it is one of the ways in which the higher achievement turns into earnings. The Impact is about a seventh of the Medi-Cal money, from a tenth of the levy, because money spent on schooling returns in later earnings.

▸ Show calculation ▾ Hide calculation
70.5 billion dollars
Paid into the fund
[4]
− 0.18 billion dollars deducted before the money is split
70.35 billion dollars
Cost of administration
[3]
× 10 % fixed by the measure
7.04 billion dollars
Share for schools and food assistance
[1][3]
× 0.6 Setting, range 0.3 to 0.8: the split is not fixed, and free school meals are already paid for
4.22 billion dollars
Of that, for schools
× 251.1 ÷ 336.9 the studies measure spending in 2018 dollars; consumer prices of September 2026
3.15 billion dollars of 2018
In 2018 dollars
× 0.0079 standard deviations 0.0316 for 1,000 dollars a year over four years; range 0.004 to 0.012
24,850 standard deviations, summed over pupils
Gain in achievement per 1,000 dollars spent on a pupil
[26]
× 0.8 Setting, range 0.5 to 1.0: replacing existing funding is forbidden, other money can still move
19,880 standard deviations, summed over pupils
Additional to what schools would have had
[3]
× 470,000 euro-equivalent value of a lasting standard deviation, 23.5 percent of a United States lifetime income set at 2.0 million euro-equivalent (about 2.3 million dollars)
9.34 billion euro-equivalent
In euro-equivalent
÷ 10 years the same horizon as every argument here
0.93 billion euro-equivalent a year
Spread over the horizon
÷ 1 billion euro-equivalent a point scale of this evaluation
0.93
Normalised Impact
Score 0.93 Impact × 8 Value × 5.5 Plausibility ÷ 10 = 4.1 of 100

Plausibility

The counterfactual is the same schools without the money. The evidence is quasi-experimental: Jackson and Mackevicius (2024) pool the American studies that use sudden changes in school funding, such as court-ordered finance reforms and close referendum votes, and find test-score effects positive in almost every one [26]; Jackson, Johnson and Persico (2016), using court-ordered reforms as an instrument, find that ten percent more spending for twelve years raised adult wages by about seven percent, and by more for children from poor families [27]. The confounder is that districts which raise spending also differ in other ways; the designs use funding changes that districts did not choose. Direction: the funding changes come before the outcomes. The studies measure money per pupil and year, so a five-year inflow counted pupil by pupil is the same dose, and how much of the money adds to what schools would have had is counted in the Impact. What holds the Plausibility in the middle is that the school money exists only if the tax survives the courts, the same doubt as in the argument on Medi-Cal [15][16]. The Plausibility is a little above the middle: school money is well shown to work, and whether this money arrives depends on the courts.

Evidence: Converging studies · plausibility at most 8 Evidence of effect: Quasi-experimental · at most 8

Fortunes that outgrow work

0points

The wealth of California's billionaires has grown more than five times as fast as family incomes for four decades, and the share of business income paid to work is the lowest since records began. Whether a single levy changes anything about that is not shown. The argument is included but not yet scored.

Value 7 · Fair distributionImpact 0Plausibility 2
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream would be a society in which income from wealth and income from work drift apart less, with the trust, cohesion and chances to rise that are said to come with it. This site places fair distribution with broad participation, one step below educational opportunity. It is a different good from the money moved by the levy: the transfer itself, and the view that a dollar matters more to the state than to a billionaire, are already counted in the arguments on Medi-Cal and on the people who pay. Counting a general sense of fairness here as well would count the same transfer twice. The value would be upper middle, because the stream would be the balance between wealth and work itself.

Impact

The trend is measured. Since 1982 the real wealth of California's billionaires has grown by 7.6 percent a year and the income of the average California family by 1.5 percent [5]. Across the United States, the share of income in the nonfarm business sector that goes to work fell from about 63 percent in 1980 to 52.8 percent in mid-2026, the lowest since records began in 1947 [28]. For somebody starting out without wealth, work is the only source of income, and the share of children who out-earn their parents has fallen from about nine in ten to one in two [29]. The levy itself is small against that trend: five percent of the group's wealth equals six to nine weeks of its growth between 2023 and 2025, so the group would reach its old level again within weeks, but would stay about a twentieth below the path it would have taken without the levy [5]. No study prices what concentration of wealth costs everybody else, and this site has no rule for it. The Impact is therefore not set. The Impact is not scored: the trend is measured, but what a single levy changes about it is not.

Plausibility

The chain would be: less concentrated wealth raises trust and fairness as people perceive them, and with them cohesion and the chances to rise. The counterfactual is the same distribution without the levy. The evidence for the links is associational: Americans report being less happy in years of higher inequality, mostly through lower trust and perceived fairness [31], and a panel of countries with an instrument finds inequality lowering trust in advanced economies [30]. The confounder is that the same economic changes may drive both inequality and distrust. Direction: distrust may also raise inequality, through the policies it brings. No study links a one-time levy of this size to any of these outcomes, so nothing carries the claim for this measure. The Plausibility is low: the trend is real, but no measured link runs from this levy to cohesion or mobility.

Recorded, not yet scored: No booking rule and no measured quantity exist for what concentrated wealth costs everybody else; the trend is documented, but a levy equal to six to nine weeks of the group's recent wealth growth would leave the group about a twentieth below its path, small against the trend, and the general fairness of the transfer is already priced in the weights of pro-1 and con-1.

Evidence: Plausibility · plausibility at most 5 Evidence of effect: Associational · at most 5.5 Placement: No carrier, but conceivable · plausibility 1.5–2.5

Open: A study of a recurring state wealth or top-income tax that measures trust, mobility or political participation against a control state would give this argument a carrier and could raise it to about five.

Arguments against6 arguments

About 200 people pay

3.4points

All of the money comes from the roughly 200 people who lived in California on 1 January 2026 and are worth more than a billion dollars. Each pays 5 percent of their net worth once, which for a person worth 10 billion dollars is 500 million.

Value 5 · Household budgetsImpact 1.2Plausibility 5.5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is money leaving the households of the people who pay, priced at the middle of the scale exactly as the money the state receives. This argument exists so that the argument opposite is not counted as though the money came from nowhere. The weight is what differs, and it is the whole substance of this evaluation: a dollar of a billionaire carries 0.2 on this site's scale, the lowest weight it uses, against 1.0 for public money. The true difference in what a dollar means at the two addresses is larger still, but the scale stops at 0.2 so that nobody's money counts for almost nothing. Nothing is counted here for the sense of being singled out by a tax that reaches about 200 named people. The question of a tax whose liability is fixed before the vote is listed separately and left open. The value is the middle of the scale, and this argument exists so that the money for health care is not booked as if nobody paid it.

Impact

The same money seen from the side that pays it: 71.0 billion dollars, the tax with its deferral charges and the additional income tax on sales, spread over the ten years of this evaluation, 7.1 billion dollars a year of real money. It falls on about 200 people, and five percent of net worth is a large sum for each, 500 million dollars for a person worth 10 billion. It is small against what the wealth of this group has done recently: it reached 2,300 billion dollars by mid-2026 after rising 144 percent between 2023 and 2025, and at that pace five percent of it was added every six to nine weeks [5]. Nearly three quarters of it is listed stock, so the tax can be paid by selling or borrowing against a twentieth of a holding [4]. Owners of private companies whose tax exceeds their listed assets and cash can defer it until they sell [1]. A dollar in these hands is counted at 0.2 of a dollar in a public budget, the floor of this site's scale, so the same money carries an Impact of 1.22, in a range from 0.99 to 1.46, against 6.12 for the state. The Impact is a fifth of the argument opposite, built on exactly the same money.

▸ Show calculation ▾ Hide calculation
6.12 billion euro-equivalent a year
The money, seen from the side that pays it
[4]
× 0.2 the lowest weight on this site's scale
1.22 billion euro-equivalent a year
Weight of a billionaire's dollar
÷ 1 billion euro-equivalent a point scale of this evaluation
1.22
Normalised Impact
Score 1.22 Impact × 5 Value × 5.5 Plausibility ÷ 10 = 3.4 of 100

Plausibility

The quantity is the revenue read from the other side, so it carries the same confidence and the same caveats. The counterfactual is the measure rejected. That a tax has two ends is an identity: if the state receives the money, these people pay it. What is judged, on both legs alike, is whether the tax survives the courts and how much of the base it reaches [15][16]. The confounder is the same as opposite and runs the same way: if courts strike down part of the tax or valuations come in low, both arguments shrink together. Reverse causation does not arise. The Plausibility matches the argument it mirrors: the same sum, the same doubt.

Evidence: Projection · plausibility at most 6 Evidence of effect: Definitional

Income tax that leaves

0.63points

Billionaires pay about 3.2 billion dollars a year of California income tax. Those who stay away or leave because the measure passed take their share with them every year, while the levy is collected once.

Value 5 · Public financesImpact 0.2Plausibility 6
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is income tax that California no longer collects because people who would have paid it live elsewhere. Lost public money is priced at the middle of the scale. Only the net loss is counted: the state loses at the weight of public money, 1.0, and the people who moved keep the money at a weight of 0.2, so four fifths of each dollar is a loss. Most of those who left moved to Florida, Texas and Nevada, which tax no personal income, so no other state collects the money instead [7]. What a move costs the people making it is not counted; they chose it. The value is the middle of the scale, and only the four fifths of each lost dollar that pass from public to private hands are counted.

Impact

Billionaires paid about 3.2 billion dollars a year of California income tax between 2023 and 2025, 2.4 percent of the state's income tax [4][5]. The six billionaires who left before 1 January 2026 left because the measure was on its way, not because it passed; they are outside its reach either way, and the vote changes only whether some of them come back. They hold about a quarter of the wealth but pay far less of the income tax, because founders draw little income from their shares: an estimate from stock sales, option exercises and dividends in company filings puts the California income tax of Larry Page at 76 million dollars a year and of Sergey Brin at 52 million on average over 2019 to 2025, from their Alphabet shares alone [5]; billionaires pay about 0.2 percent of their wealth in California income tax [5], which for the others who left adds about 0.1 billion, so all who left together are estimated at about 0.2 billion, in a range from 0.15 to 0.4. Three in ten of them are assumed to return if the measure fails, in a range from one in ten to one in two: 0.06 billion dollars a year. No study has measured how many wealthy people return after a feared tax fails; Boll, Saez and Zucman expect the response to fade [5], the critical study assumes nobody returns [6]. The second group are billionaires who stay liable but may now expect further taxes of this kind and leave over the decade, or who never arrive. Billionaires react strongly to state taxes on wealth: after 2001, states with an estate tax lost about a third of their billionaires relative to other states [8], and when Spain's regions reintroduced their wealth tax, the income tax lost through movers was six times the wealth tax lost [13]. An average shortfall of 8 percent of the remaining 3.0 billion over the horizon is used, in a range from 3 to 20 percent: 0.24 billion dollars a year. That share is a setting too, because no study measures how far a one-time levy makes people expect another; across both ranges the result moves by less than one point. Together that is 0.30 billion dollars a year, in a range from 0.14 to 0.71, against an official estimate of less than 1 billion [1] and the drafters' scenarios of 0.15 to 0.51 billion [5]. A study for the state's taxpayer association arrives at 3.5 to 4.5 billion a year, of which 2.5 to 3.1 billion is income tax and the rest sales tax and spending lost to the local economy, by counting nine departures, some of them only announced, as caused by the measure and by imputing their income from their wealth [14]; a second critical study counts 30 to 60 percent of all billionaire income tax as lost for good [6]. Because the movers keep what the state loses, at a fifth of its weight, the Impact is 0.21 on the scale of this evaluation, in a range from 0.10 to 0.46. The Impact is a thirtieth of the levy per year, but unlike the levy it does not stop after five years.

▸ Show calculation ▾ Hide calculation
California income tax paid by billionaires [4][5] 2023 to 2025 3.2 billion dollars a year
× Paid by those who left before 1 January 2026 estimate from company filings: Page 76 million, Brin 52 million from Alphabet shares, the others less; range 0.15 to 0.4 [5][7] about 6 % 0.2 billion dollars a year
× Part that would have come back had the measure failed Setting, range 0.10 to 0.50: departures that the vote, not the announcement, makes permanent 0.30 0.06 billion dollars a year
+ Billionaires who stay liable but leave or never arrive Setting, range 3 to 20 percent: billionaires are the most mobile taxpayers measured [8][13] 3.0 billion dollars of remaining income tax × 8 % 0.24 billion dollars a year
= Income tax lost each year because the measure passed range 0.15 to 0.67 0.3 billion dollars a year
÷ In euro-equivalent exchange rate used throughout this evaluation 1.16 dollars to the euro 0.26 billion euro-equivalent a year
× Net loss from public to private hands the state loses at 1.0, the movers keep it at 0.2 1.0 − 0.2 = 0.8 0.21 billion euro-equivalent a year
÷ Normalised Impact scale of this evaluation 1 billion euro-equivalent a point 0.21
Score 0.21 Impact × 5 Value × 6 Plausibility ÷ 10 = 0.6 of 100

Plausibility

The counterfactual is the same billionaires with the measure rejected. That the very rich move in response to state taxes is measured with good designs. Moretti and Wilson (2023) use the 2001 change in federal estate tax rules, which suddenly made it matter in which state a billionaire died, in a difference-in-differences with an instrument, and find the number of billionaires in estate-tax states falling by a third [8]. Rauh and Shyu (2024) find, with complete tax records, that California's 2012 top-rate increase drove an extra 0.8 percent of the top-bracket tax base out of the state, while nine tenths of the revenue lost came from those who stayed and reported less income [12]. Those designs are quasi-experimental. The confounder for this measure is that its liability is already fixed for everyone resident on 1 January 2026, so leaving after the vote does not avoid it, and the studies measured recurring taxes; whether a one-time levy creates the expectation of another is not measured anywhere [25]. That transfer is the reason for the deduction from the ceiling, and the size of the effect is carried in the range. Reverse causation would mean the rich leaving for other reasons and the tax taking the blame; the estate-tax study rules that out by comparing older and younger billionaires, for whom the tax mattered differently [8]. The Plausibility is above the middle: that billionaires move is well measured, how far this one vote moves them is not.

Evidence: Converging studies · plausibility at most 6.5 Evidence of effect: Quasi-experimental · at most 8

Open: California's income tax records for 2027 and 2028 will show how many billionaires who were resident on 1 January 2026 have left since, which would replace the setting for later departures and could raise the Plausibility to about seven.

Firms that move away

0.45points

When a rich owner leaves, the firms they control sometimes shrink or follow them. Most of the wealth of California's billionaires sits in large listed companies that stay where they are, so only a small part of that effect applies here.

Value 6 · OutputImpact 0.1Plausibility 5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is output that is not produced: firms that shrink, close or relocate with an owner who leaves, and investment that owners forgo to pay the tax. This site places that with the working order of economic systems and prosperity, because what is lost is production rather than anybody's income in particular. Only the part that is lost to the world is counted; activity that moves from California to Texas is a loss to one state and a gain to another, and not a loss here. The income tax the state loses is the argument before this one and is not counted again. The value sits in the middle-upper part of the scale, and only output that disappears rather than moves is counted.

Impact

A fifth of California billionaire wealth is assumed to sit in privately held operating companies, 436 billion dollars of 2,180 billion, in a range from 14 to 28 percent: 28 percent of it is not listed stock, but that share also holds funds, property and cash [4]; the revenue study by Boll, Saez and Zucman counts half of the wealth outside listed stock, 17.5 percent, as private business [5], and the billionaires whose wealth is mainly a private business hold 16 percent of all billionaire wealth [14]. These are the firms whose fate can depend on where their owner lives. The share of owners whose departure the vote causes follows from the argument before this one: three in ten of the quarter that left, plus 8 percent of the rest, about 13.5 percent, or 59 billion dollars of company value. Such firms produce, as a setting, 15 cents of value added a year per dollar of company value, in a range from 8 to 25 cents: 8.8 billion dollars a year. When a wealthy owner emigrated from Sweden, the value added of their firms fell by a third, although other firms took over nearly half of those that closed [9]. A move within the United States keeps most of that activity in the country and much of it in the firm, so 2 percent of the value added is used as the net loss, in a range from 0.5 to 7 percent. That is 0.18 billion dollars a year, or an Impact of 0.15, in a range from 0.03 to 0.76. The Impact is small because the firms that carry most of this wealth do not move when their owners do.

▸ Show calculation ▾ Hide calculation
436 billion dollars
Billionaire wealth in privately held operating companies
[4][5][14]
× 0.25 × 0.30 + 0.75 × 0.08 = 0.135 the same settings as in the argument on income tax
58.9 billion dollars
Held by owners whose departure the vote causes
× 0.15 Setting, range 0.08 to 0.25
8.83 billion dollars a year
Value added per dollar of company value
× 2 % Setting, range 0.5 to 7 percent: a third of value added when owners emigrate abroad, most of it kept in a move between states
0.18 billion dollars a year
Net loss to the world when the owner moves
[9]
÷ 1.16 dollars to the euro exchange rate used throughout this evaluation
0.15 billion euro-equivalent a year
In euro-equivalent
÷ 1 billion euro-equivalent a point scale of this evaluation
0.15
Normalised Impact
Score 0.15 Impact × 6 Value × 5 Plausibility ÷ 10 = 0.5 of 100

Plausibility

The counterfactual is the same owners with the measure rejected. What is counted is only the output lost when an owner actually leaves, and the finding for that is from Jakobsen and co-authors (2026): an event study of wealthy Scandinavians who emigrated, combined with a difference-in-differences around Sweden's 2006 repeal of its wealth tax, finds that firms whose owner leaves lose a third of their employment and value added [9]. It is a single study. The confounder is that owners who emigrate may already be winding down their firms; the event study follows the firms before the move and finds no decline beforehand [9]. Studies of Norway's wealth tax find owners who stay paying out more and investing less [18], or employing more and saving more [17][24]; they concern owners who stay, which this argument does not count. The transfer from emigration abroad to a move between American states is a deduction of its own. The Plausibility is in the middle: the link is measured, but once, and abroad.

Evidence: Study · plausibility at most 5 Evidence of effect: Quasi-experimental · at most 8

Open: Company filings from 2027 onward will show whether the private firms of the billionaires who left California before the cut-off date changed at all, which could move this towards the ceiling of a measured effect.

Administration and appraisals

0.2points

The tax authority must value private companies, art and trusts for about 200 people, and those who pay will hire appraisers and counsel. The official estimate is tens of millions of dollars a year for the state alone.

Value 5 · Enforcement costImpact 0.04Plausibility 10
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is work spent on administering the tax and valuing what it reaches: tax officials, appraisers and counsel, on both sides. This site counts that as the cost of enforcement, at the middle of the scale. Nothing is counted here for the tax itself; this is only what collecting it uses up. Appraisers and counsel hired by the people who pay are counted alongside the state's own staff, because their work is used up in the same way. The value is the middle of the scale, because the cost is work paid for with money.

Impact

The official analysis expects administration costs for the state of tens of millions of dollars a year for several years [1]; 30 million dollars a year for six years is used, 180 million in all, in a range from 100 to 400 million. This is also the amount taken from the fund before it is split. On the side of those who pay, the measure requires certified appraisals where a private company is not valued by formula and for art above 1 million dollars, and it penalises understatement by 20 to 40 percent [4]. About 1.5 million dollars per taxpayer in valuation and counsel is set, in a range from 0.5 to 5 million, 300 million in all. Together that is 480 million dollars, or 48 million a year over the horizon, in a range from 27 to 120 million. The Impact is among the smallest here, under one percent of what the tax collects.

▸ Show calculation ▾ Hide calculation
Administration by the state tax authority Setting, range 100 to 400: tens of millions a year in the official estimate [1] 30 million dollars a year × 6 years 180 million dollars
+ Appraisals and counsel for those who pay Setting, range 0.5 to 5 million per taxpayer [4] 200 × 1.5 million dollars 300 million dollars
= Cost over the horizon range 265 to 1,214 million dollars 480 million dollars
÷ Spread over the horizon and in euro-equivalent the same horizon as every argument here 10 years, 1.16 dollars to the euro 41.4 million euro-equivalent a year
÷ Normalised Impact scale of this evaluation 1 billion euro-equivalent a point 0.04
Score 0.04 Impact × 5 Value × 10 Plausibility ÷ 10 = 0.2 of 100

Plausibility

Once the tax exists, administering it and valuing what it reaches costs something in every case; the official figure is a projection of how much [1]. The counterfactual is no tax and no administration. The fact of the cost is definitional, and its size is carried in the range. The confounder is that the cost of fighting a valuation rises with the tax at stake, which is why the range reaches far upward. Reverse causation does not arise. The Plausibility is at the top, because the cost arises whenever the tax is collected.

Evidence: Mechanism · plausibility at most 10 Evidence of effect: Definitional

Taxing a past residence

0points

Liability depends on where a person lived on 1 January 2026, ten months before the vote, and people who left during 2026 still owe it. What that does to trust in the stability of tax law belongs to the debate. The argument is included but not yet scored.

Value 9 · Rule of lawImpact 0Plausibility 3
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream would be the trust that a tax applies to circumstances after it has been decided, not before. That expectation belongs to the institutional core of a constitutional order and is valued near the top of the scale on this site. The measure fixes liability on 1 January 2026 so that nobody can escape it by moving once the vote is near, and the critics' legal case rests largely on that date [15]. The drafters point out that the federal income tax of 1913, passed in October, also reached back to income received after 1 March of that year [16]. The value would be high, because the stream would be trust in the rules themselves.

Impact

The concrete consequences of the early date are counted elsewhere: the departures it provoked in the argument on income tax, and the court challenge in the argument on court costs. What would remain is the harm to legal certainty as such, and this site has no rule for pricing it. A price per person reached would be a setting with nothing to anchor it. The Impact is therefore not set. The Impact is not scored: the harm that remains after its concrete consequences has no agreed quantity.

Plausibility

The chain would be: a tax fixed before its vote makes people expect that future taxes may also reach back, and they plan, invest or move accordingly. The counterfactual is the same tax with a cut-off date after the vote. The design is mechanistic; that the date precedes the vote is a fact of the text [1], but that it changes behaviour beyond the departures already counted is untested. The confounder is that people would react to the tax itself whatever its date. Direction: the date precedes any reaction. With the link that carries the quantity untested, the Plausibility stands at three in ten. The Plausibility is low as long as nobody has measured what a retroactive date adds to the tax itself.

Recorded, not yet scored: No booking rule exists for the harm to legal certainty from a tax whose liability date precedes its vote; its concrete consequences, departures and litigation, are counted in con-2 and con-4, and what remains would need a setting for trust in the stability of tax law, which only Julian can decide.

Evidence: Plausibility · plausibility at most 5 Evidence of effect: Mechanistic · at most 6 Placement: Chain open · plausibility 3–3.5

Court costs

0points

Critics argue that the tax breaches the federal constitution, and a challenge is expected as soon as it passes. The measure sends it to the state courts for a quick decision; lawyers and courts on both sides cost money either way.

Value 5 · Enforcement costImpact 0Plausibility 9.5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is the work of lawyers, experts and judges spent on the challenge to the tax, on both sides. This site counts that as a cost of enforcement, at the middle of the scale, like the administration in the argument before. The work is used up whoever wins, so both sides count alike. The tax money at stake is not counted here; it is the subject of the case, not its cost. Nothing is counted here for the possibility that the courts strike the tax down: then nothing is collected and nothing is paid, and that doubt is carried in the Plausibility of the arguments on the money. The value is the middle of the scale, because the cost is work paid for with money.

Impact

Critics argue that a residency date before the vote, the taxation of people who left during 2026 and the attribution of worldwide wealth to California breach the federal constitution [15]; the drafters reply that the tax stands [16]. The measure provides for a quick decision in the state courts [3]. For both sides together, through the state courts and possibly the federal Supreme Court, 50 million dollars is set, in a range from 20 to 150 million. Spread over the horizon that is 5 million dollars a year, in a range from 2 to 15 million. The Impact is the smallest here, a tiny fraction of what is at stake in the case.

▸ Show calculation ▾ Hide calculation
50 million dollars
Court challenge, both sides
[15][16]
÷ 10 years, 1.16 dollars to the euro the same horizon as every argument here
4.3 million euro-equivalent a year
Spread over the horizon and in euro-equivalent
÷ 1 billion euro-equivalent a point scale of this evaluation
0
Normalised Impact
Score 0 Impact × 5 Value × 9.5 Plausibility ÷ 10 = 0 of 100

Plausibility

The counterfactual is no tax and no challenge. That the tax will be challenged is announced by its critics and expected by the official analysis [1][15], so the cost arises whenever the tax passes; its size is carried in the range. The named risk is that the case is settled quickly or not brought in full, which holds the Plausibility just below the top. Reverse causation does not arise. The Plausibility is near the top, because a challenge is all but certain once the tax passes.

Evidence: Mechanism · plausibility at most 9.5 Evidence of effect: Definitional

How robust the result is

Under the assumptions used, the proposal brings more than it costs in more than 99 of 100 model runs.

gains more than it costscosts more than it gains

Each dot stands for 200 of the 20,000 model runs.

A dollar in the hands of a billionaire is counted at a fifth of a dollar in the state budget, the lowest weight this site uses. Counted at the same weight, the measure would still come out ahead in 97 of 100 runs, but only because of what the school money does; without that, in none. The income tax lost through departures is counted for the ten years of this evaluation and only where the vote, not the announcement, caused it. At the size the critics estimate, counting lost sales tax and local spending as well, and lasting thirty years, the measure would come out ahead in a little over three runs in ten. The evaluation assumes that Proposition 40 takes effect: if the courts strike it down or a competing measure on the same ballot blocks it, nothing is collected and nothing is paid. As it stands, the levy lies ahead in practically all runs.

Show the distribution of the model runs

20,000 simulated futures

Very likely better · 100 %

In more than 99 of 100 simulated futures the proposal brings more than it costs.

No effect is known exactly; every argument has a range. The calculation draws a possible size for each argument from its range 20,000 times and counts how often the benefits exceed the costs.

Net effect in points per future. Left of zero the drawbacks outweigh the benefits, right of zero the benefits win. Eight in ten futures fall between +11 and +23 points.
Scale and assessment

Scale of this evaluation: Normalised Impact, unitless and calibrated to this topic. For comparison, one point here corresponds to about 1 billion euro-equivalent a year.

Size class:medium

Effect large (+16) · 93 % of the effect is positive

Areas show individual argument weights. The positive share of the effect nets out money transfers between the two sides; the probability of a positive net effect is determined separately.

How we score →

Review and sources

Last reviewed by Claude Opus 5.5 · October 29, 2026 · 6× AI, 3× human
  1. October 29, 2026AI reviewClaude Opus 5.5record updated

    Page texts for the new page layout (draft G) with quotes from primary sources; numbers unchanged.

  2. October 21, 2026AI reviewClaude Fable 5.1record updated

    Money amounts in argument texts stated in dollars per style-en.md; scores, chains and balance unchanged

  3. October 20, 2026AI reviewClaude Fable 5.1record updated

    Summary and scale note rewritten to the style of 20 Oct 2026; numbers unchanged

  4. October 19, 2026Assessed by a human, implemented by AIClaude Opus 5.5record updated

    Sources added for the estimates (credit costs, private-business share, income tax of leavers, timing of federal Medicaid cuts); where no study exists, the text says so and how little the setting moves the result.

  5. October 19, 2026Assessed by a human, implemented by AIClaude Opus 5.5re-scored

    Revised after an independent review, decisions by Julian: school spending in 2018 prices and counted four fifths additional, deferral charges and additional income tax on sales added, private-company base estimated at a fifth, firm loss at its ceiling, administration at P 10, court costs as their own argument.

  6. October 19, 2026Assessed by a human, implemented by AIClaude Opus 5.5First evaluation

    First evaluation of Proposition 40, reviewed by Julian: the levy as a transfer pair, the school share as an additive earnings stream, departures counted only where the vote rather than the announcement caused them; wealth concentration and the retroactive cut-off date included but not scored.

Evaluations are produced with AI support and reviewed on a schedule for new developments; human passes are marked separately.How we review →

Show all 35 sourcesHide sources
  1. California Legislative Analyst's Office: Proposition 40: Imposes one-time tax on certain taxpayers. Analysis for the official voter guide. lao.ca.gov
  2. California Legislative Analyst's Office: Fiscal estimate of initiative 25-0024, letter to the Attorney General, 11 December 2025. oag.ca.gov
  3. California Secretary of State: Official Voter Information Guide 2026: Proposition 40, title, summary and arguments. vig.cdn.sos.ca.gov
  4. Galle, Gamage, Saez and Shanske: Expert report on the California 2026 Billionaire Tax, updated 16 February 2026. eml.berkeley.edu
  5. Boll, Saez and Zucman, NBER Working Paper 35218: Billionaire wealth and taxes in California. nber.org
  6. Rauh and Jaros, NBER Tax Policy and the Economy 41: The net present value of the California Billionaire Tax Act. nber.org
  7. Fortune: Six billionaires left California before the billionaire tax deadline. fortune.com
  8. Moretti and Wilson, American Economic Journal: Economic Policy 15(2): Taxing billionaires: estate taxes and the geographical location of the ultra-wealthy. eml.berkeley.edu
  9. Jakobsen, Kleven, Kolsrud, Landais and Muñoz, American Economic Review 116(10): Taxing top wealth: migration responses and their aggregate economic implications. nber.org
  10. Jakobsen, Jakobsen, Kleven and Zucman, Quarterly Journal of Economics 135(1): Wealth taxation and wealth accumulation: theory and evidence from Denmark. econ.ku.dk
  11. Londoño-Vélez and Ávila-Mahecha, American Economic Review: Insights 3(2): Enforcing wealth taxes in the developing world: quasi-experimental evidence from Colombia. aeaweb.org
  12. Rauh and Shyu, American Economic Journal: Economic Policy 16(1): Behavioral responses to state income taxation of high earners: evidence from California. nber.org
  13. Agrawal, Foremny and Martínez-Toledano, American Economic Journal: Applied Economics 17: Wealth tax mobility and tax coordination. aeaweb.org
  14. California Taxpayers Association: Wealth tax would cost the state more than 3.5 billion dollars in ongoing annual revenue. caltax.org
  15. Tax Foundation: California Proposition 40: constitutional and legal issues of the wealth tax. taxfoundation.org
  16. Galle, Gamage and Shanske: Correcting the record: the legal case for the California Billionaire Tax Act. eml.berkeley.edu
  17. Bjørneby, Markussen and Røed, Economica 90(358): Does the wealth tax kill jobs?. docs.iza.org
  18. Berzins, Bøhren and Stacescu, ECGI Working Paper: Illiquid owners and firm behavior: financial and real effects of the personal wealth tax. ecgi.global
  19. Miller, Johnson and Wherry, Quarterly Journal of Economics 136(3): Medicaid and mortality: new evidence from linked survey and administrative data. nber.org
  20. Finkelstein, Hendren and Luttmer, Journal of Political Economy 127(6): The value of Medicaid: interpreting results from the Oregon Health Insurance Experiment. pmc.ncbi.nlm.nih.gov
  21. California Legislative Analyst's Office: Considering Medi-Cal in the midst of a changing fiscal and policy landscape, 24 October 2025. lao.ca.gov
  22. Hoover Institution: Proposition 40: the Billionaire Tax Act. hoover.org
  23. CalMatters: Newsom takes sides on the tax propositions on the November ballot. calmatters.org
  24. Ring, Review of Economic Studies 92(5): Wealth taxation and household saving: evidence from assessment discontinuities in Norway. restud.com
  25. Eichengreen, NBER Working Paper 3096: The capital levy in theory and practice. nber.org
  26. Jackson and Mackevicius, American Economic Journal: Applied Economics 16(1): What impacts can we expect from school spending policy? Evidence from evaluations in the United States. ideas.repec.org
  27. Jackson, Johnson and Persico, Quarterly Journal of Economics 131(1): The effects of school spending on educational and economic outcomes: evidence from school finance reforms. doi.org
  28. U.S. Bureau of Labor Statistics: Labor share at its lowest level, 52.8 percent, in the second quarter of 2026. bls.gov
  29. Chetty, Grusky, Hell, Hendren, Manduca and Narang, Science 356: The fading American dream: trends in absolute income mobility since 1940. science.org
  30. Barone and Mocetti, Economic Inquiry 54(2): Inequality and trust: new evidence from panel data. ideas.repec.org
  31. Oishi, Kesebir and Diener, Psychological Science 22(9): Income inequality and happiness. doi.org
  32. Bruce and Hannon, Board of Governors of the Federal Reserve System, FEDS Notes: Estimating securities-based loans outstanding, 2 August 2024. federalreserve.gov
  33. California Legislative Analyst's Office: The 2026-27 budget: Medi-Cal fiscal outlook, November 2025. lao.ca.gov
  34. KFF: Allocating CBO's estimates of federal Medicaid spending reductions across the states: enacted reconciliation package, 23 July 2025. kff.org
  35. Agersnap and Zidar, American Economic Review: Insights 3(4): The tax elasticity of capital gains and revenue-maximizing rates. nber.org

More debates in this area

Very likely better · 100 %
Size: large
A Fully Refundable Credit

Pay the full child tax credit to every eligible child, including those whose parents earn too little to owe income tax.
Very likely worse · 0 %
Size: large
The Forced-Labour Tariffs

An additional duty of 10 or 12.5 percent on imports from sixty economies, set by whether each bans, or has committed to ban, imports made with forced labour.
Unclear · 48 %
Size: medium
School Scholarship Credit

Let every taxpayer send up to 1,700 dollars of federal income tax to scholarship organizations for school children instead of the Treasury, from January 2027.
Very likely better · 100 %
Size: very large
Lifting The Payroll Cap

Charge the Social Security payroll tax on earnings above the taxable maximum of 184,500 dollars, which today stop being taxed altogether.
All debates in Taxes, Tariffs and the Federal Budget →