California Proposition 3 Explained: Permanent High-Income Tax

Recommendation: YES — moderate confidence. Proposition 3 would keep in place three additional marginal income-tax rates on California’s highest-income taxpayers after the rates otherwise expire at the end of 2030. It does not change anyone’s tax bill today. Relative to the law voters face after 2030, however, it is a permanent tax increase for affected taxpayers. I recommend YES because the existing tax produces material revenue for education and the state budget, and the alternative is a foreseeable multibillion-dollar revenue loss with no replacement specified. The revenue is unusually volatile, and the measure does not guarantee that every dollar funds schools or health care; those facts temper the recommendation but do not outweigh the value of retaining an established revenue source with education protections and reserve rules.

This analysis uses the sherafy.com Civic Outcomes Standard. Los Angeles County readers can return to the Los Angeles County Voter Guide 2026 for the complete ballot.

What a YES or NO vote does

The Secretary of State’s official guide identifies Proposition 3 as an initiative constitutional amendment on the November 3, 2026 ballot. A YES vote makes permanent the higher income-tax rates voters approved in Proposition 30 (2012) and extended through 2030 in Proposition 55 (2016). A NO vote lets those incremental rates expire after tax year 2030. The rest of California’s income-tax schedule remains in place.

The affected rates are marginal: an additional 1 percentage point on one band of income, 2 points on the next, and 3 points on the highest band. The official 2025 inflation-adjusted thresholds are about $371,000 for single filers, $742,000 for joint filers and $505,000 for heads of household. Only about the top 2% of taxpayers pay the added rates, and they currently pay about half of all state personal income tax. The rate applies to income within each band, not to all income once a taxpayer crosses a threshold.

The campaign’s “no new tax” wording is accurate only as a statement about the tax bill now: Proposition 3 does not raise today’s rates. It is misleading if it suggests no change relative to current law’s 2031 sunset. A YES vote continues the higher rates beyond their voter-approved expiration; a NO vote allows them to end. The choice is whether to keep a long-running, temporary tax schedule—not whether to impose it for the first time.

How the money is routed

The higher rates generate revenue through California’s personal income tax. The Legislative Analyst’s Office (LAO) estimates that keeping them in place would maintain roughly $5 billion to $15 billion in annual state revenue after 2030, with the amount varying substantially with financial markets and high-income investment income.

Roughly 40% of this revenue supports K–12 schools and community colleges through the Education Protection Account. Within that education allocation, the constitutional split is 89% for K–12 and 11% for community colleges. School districts and community college districts choose eligible spending through public processes; the funds cannot pay administrative costs and are subject to accounting and audit requirements. The 89/11 split does not mean that 89% of all Proposition 3 revenue goes to K–12 schools. The LAO says the remainder supports other state programs, with part held in budget reserves. Under the existing formula, a share of certain excess revenue can support eligible child and family health services, subject to a cap and emergency provisions; remaining funds support the General Fund and its other obligations.

The current State Controller’s Prop 30/55 tracker reports more than $119 billion generated since 2012, a cumulative projection as of June 30, 2025, and describes education-accounting, district-spending and audit rules. That total combines Proposition 30/55 revenue streams, including earlier sales-tax changes; it is not the amount Proposition 3 alone would generate after 2030, nor does it mean all historical revenue was distributed to classrooms.

The realistic NO baseline

Without Proposition 3, the three added income-tax brackets expire in 2031 under current law. The state would collect less income tax from affected high earners, but the amount lost in a particular year is not fixed. The LAO’s $5–$15 billion range is a forecast of annual revenue maintained by the measure in future years, not a promise that the state will lose exactly that amount every year if voters reject it.

Expiration would create a budget choice, not an automatic list of layoffs or service cuts. The Legislature could reduce spending, use reserves, change other taxes, borrow where lawful, or combine responses. Proposition 3 supporters are right that the state would have fewer resources absent another policy. Their claims that rejection necessarily causes specified teacher layoffs, Medi-Cal losses or school-class-size increases go beyond what the ballot analysis forecasts. Conversely, saying California has “a spending problem, not a revenue problem” is a policy judgment; it does not show that current services can be maintained without replacing the expiring revenue or changing the budget.

The sunset was part of the original voter-approved arrangement: Proposition 30’s increases were temporary, then Proposition 55 extended them through tax year 2030. Proposition 3 asks voters to revise that endpoint. That history strengthens the NO case on voter expectations and legislative accountability. It does not mean the tax has never been reviewed: voters chose the 2016 extension, and they can later amend the state constitution through another voter-approved constitutional change.

Revenue reliability and taxpayer behavior

The tax’s fiscal value comes with real forecasting risk. California’s own 2025–26 revenue analysis explains that high-income taxpayers receive a larger share of income from capital gains and stock-based compensation. The state reported that capital-gains realizations fell 55% from 2021 to 2022, while the top 1%’s income-tax liability fell about 40%, even as the wider economy remained comparatively flat. A 2021 market peak can produce receipts that are not sustainable in a downturn. The LAO’s $5–$15 billion range properly reflects that uncertainty; it is not a guaranteed annual yield.

Reserves help smooth the state’s broader revenue cycles but cannot turn a volatile tax stream into stable recurring income without limits. California’s existing Proposition 2 framework directs some capital-gains-related revenue to reserves and debt payments. That is relevant protection, although it does not eliminate the risk that services funded from the General Fund face pressure when receipts fall.

The evidence about high-earner mobility is mixed. Rauh and Shyu’s peer-reviewed 2024 American Economic Journal: Economic Policy study uses California administrative tax data to examine the 2012 Proposition 30 rate increase. Its main comparison follows upper-income California residents and uses nonresident California filers with California-source income as a control group; the authors also report synthetic difference-in-differences and matched comparisons. The abstract reports that behavioral responses reduced 45.2% of the expected tax “windfall” in the first year and 60.9% within two years, mostly through changes in taxable income among those who remained; migration accounted for a smaller share. These percentages refer to erosion of the modeled windfall, not a loss of 45% or 61% of all California income-tax revenue. The authors note that their tax-return data cannot fully distinguish reduced labor supply from offshore income, avoidance or other channels, and the control comparison relies on assumptions about how federal tax changes affected residents and nonresidents. Their study concerns a sudden tax increase and persistently high earners, so its estimates should not be mechanically applied to Proposition 3’s continuation of rates already in effect.

A separate California administrative-data paper by Varner, Young and Prohofsky examines three tax changes over 25 years and finds little migration response among top filers. It uses difference-in-differences comparisons between affected high-income filers and those just below the new bracket, but is explicitly labeled a work in progress, preliminary and not for citation without permission. The studies differ in design and measure different responses; the preliminary paper does not erase the peer-reviewed finding, and the peer-reviewed study does not establish that this extension would cause ongoing mass departures. The defensible conclusion is that some taxable-income and migration response is plausible, while its size and long-run revenue effect under a continuation are uncertain.

Fiscal and distributional consequences

The burden falls on the high-income taxpayers whose income exceeds the inflation-adjusted thresholds. Lower- and middle-income filers do not pay these added brackets. Since high-income households pay a large share of the state’s income tax, the schedule is progressive, though state income tax is only one component of household tax burdens and living costs.

Public benefits are more broadly distributed through school funding, health and human services, reserves and other state spending. The education allocation is protected through constitutional rules, public spending plans and audits. But the initiative does not earmark every dollar exclusively for schools or guarantee a fixed amount of health spending. When a budget is tight, the education account’s legally determined share remains protected while discretionary services financed from the General Fund can still be cut. When revenue is high, the formula and state budget determine how much supports eligible health services, reserves and other programs.

This distinction matters to both sides. Supporters can reasonably say the tax protects a meaningful source of public revenue and education funding. They cannot prove that the entire $5–$15 billion goes directly to classrooms or that a NO vote automatically causes a particular program cut. Opponents can reasonably object to converting a temporary schedule into a permanent one and to the risks of relying on a concentrated, volatile tax base. They cannot infer from volatility alone that the tax loses money overall or that taxpayers will leave in numbers sufficient to erase the revenue.

Applying the ten Civic Outcomes lenses

1. Human welfare

Revenue helps fund schools, community colleges, health programs and other services with direct effects on opportunity and well-being. The actual service effect of a NO vote depends on future budget choices, and Proposition 3 does not guarantee specific outcomes. The measure preserves fiscal capacity rather than prescribing a service plan.

2. Distribution and inequality

The added rates apply to high-income taxpayers and are marginal, while publicly funded benefits reach a broader population. The tax’s concentration on a small group makes the system progressive but exposes revenue to their market-linked income. Education protections create a distributional benefit across local districts, although they do not equalize all differences in school resources or outcomes.

3. Civil liberties and equal treatment

The measure changes tax liability under a generally applicable income-tax schedule, with thresholds adjusted for inflation and existing tax-administration procedures. No new enforcement power or individualized eligibility rule is central to the proposal. Tax complexity and disputes remain matters for ordinary tax administration.

4. Economic and material effects

Affected taxpayers retain less after-tax income than they would after the scheduled sunset. Public services financed by the revenue may support human capital, health and economic activity, but this measure does not estimate multiplier effects or prove a particular return. Migration and income-timing responses are possible; evidence does not establish an inevitable exodus.

5. Fiscal reality and opportunity cost

The LAO’s $5–$15 billion annual range is material and volatile. A YES vote preserves the revenue and the existing allocation system; a NO vote gives high earners the sunset but leaves the state to adjust spending, reserves or other revenue. Continuing the tax has an opportunity cost for affected taxpayers and a fiscal risk if the state treats volatile receipts as guaranteed recurring funding.

6. Institutional integrity and democratic accountability

The initiative honors the current constitutional allocation rules and uses a statewide vote to decide whether an announced sunset should remain. The NO case is strengthened by the original temporary promise and the repeated extension. Voters, rather than legislators acting alone, would authorize permanence. District-level plans, public reporting and audits provide oversight, though they do not establish that spending is optimally effective.

7. Evidence of effectiveness

The Controller documents the revenue stream, transfers and audit framework. That proves money is generated and routed; it does not establish that the tax caused better test scores, health outcomes or lower inequality. No causal evaluation reviewed here isolates the net service outcomes attributable to Proposition 30/55 revenue. The recommendation rests on preserving fiscal capacity and the education allocation, not a claim of measured program effectiveness.

8. Implementation and administrative capacity

The tax and allocation mechanisms have operated for years through the Franchise Tax Board, Department of Finance, Controller, school districts and community colleges. Continuation avoids a new tax system. Forecasting and budgeting remain challenging because receipts vary with financial markets, but existing estimates, reserve rules and audits provide administrative infrastructure.

9. Unintended consequences and behavioral response

Taxpayers may alter realizations, compensation timing, residence or investment decisions. The California studies disagree on the size of the migration response, and the sharper peer-reviewed estimate addresses a tax increase, not permanent continuation. Volatility also creates risks of overcommitment during boom years. These are reasons to avoid treating the revenue range as a fixed guarantee, not evidence that the revenue has no net value.

10. Reversibility, resilience and future lock-in

Proposition 3 removes the current 2030 endpoint and makes the constitutional tax provisions permanent unless voters later amend them. That is durable, but not literally irreversible: Californians can change the Constitution by another ballot measure. Proposition 3 is not a standing appropriation of $5–$15 billion to one program; allocations continue under existing rules and budgets. The sunset offers a scheduled reconsideration point, which NO preserves. YES should therefore meet a meaningful burden: a foreseeable fiscal need and safeguards must justify removing that checkpoint. The existing revenue scale, education protections, established administration and absence of a specified replacement meet that burden, though only moderately given volatility and limited evidence on long-run behavioral effects.

Strongest case for YES

California relies heavily on personal income tax, and the top 2% pay about half of it. Allowing the added rates to expire would relinquish a revenue stream the LAO estimates at $5–$15 billion in a typical range of future conditions. Roughly 40% supports public schools and community colleges under constitutional allocation and audit rules. Proposition 3 preserves an existing, progressive source while leaving budget decisions and future voter oversight intact. If voters reject it, the state must decide how to absorb or replace the revenue decline; no alternative funding plan is part of the ballot measure.

The best YES argument is not that every dollar is protected for classrooms or that cuts are automatic. It is that continuing a tax already in place avoids a foreseeable fiscal contraction and preserves a tested allocation mechanism while taxpayers retain the option of future constitutional change.

Strongest case for NO

Voters were told the increases were temporary. Proposition 55 extended them once, and Proposition 3 would remove the sunset again, this time indefinitely. A permanent tax on a narrow, financially sensitive income base can make budgets more exposed to market cycles and taxpayer responses. The strongest empirical study of the original increase finds substantial erosion of the expected revenue windfall through taxable-income responses, while the state has not demonstrated that each marginal tax dollar produces outcomes that justify perpetual continuation. The state could instead reduce spending, use reserves, or ask voters to revisit the rates closer to expiration with more current evidence.

The best NO argument is therefore about democratic promise, fiscal resilience and preserving a scheduled review—not a claim that the tax is paid by ordinary earners or that migration necessarily overwhelms its yield. A sunset creates a real choice point; a permanent continuation can normalize a revenue base that is both concentrated and volatile.

Campaign-claims audit

Official ballot arguments are advocacy and are not checked for accuracy by election officials. The claims above are tested against the LAO analysis, legal text, Controller records and empirical research.

Funding and interested parties

The Secretary of State’s campaign-finance totals are assembled from reports by committees primarily formed to support or oppose measures and may omit later contributions or outside spending. CalMatters’ compilation of Secretary of State data, last updated September 28, 2026, showed about $50.1 million for Proposition 3 support committees and $5,500 for opposition committees at that cutoff. The largest listed supporters were education and labor organizations, including the California Teachers Association ($33.28 million), National Education Association ($7.5 million), California School Employees Association ($2.6 million) and California Federation of Teachers ($1.75 million). This snapshot is not a final or comprehensive accounting of every independent expenditure; the official portal warns totals may lag filing deadlines.

Those groups have clear interests in preserving school and public-service funding. That context helps readers understand who is advocating, but it neither validates nor refutes the tax’s merits. The opposition includes taxpayer and business organizations whose stated interests are lower taxes and spending restraint. Contributions are evidence of campaign activity, not evidence that a policy claim is true.

Unknowns and what would change this analysis

The LAO’s range does not predict a precise annual amount after 2030. The future economy, stock-market returns, capital-gains realizations, compensation patterns and tax responses are unknown. The migration literature does not identify a single agreed estimate for the long-run response to maintaining current rates. The record reviewed also does not provide a causal evaluation showing the marginal educational or health outcomes produced by each dollar of Proposition 30/55 funding.

Most importantly, the fiscal consequence of NO depends on future political choices. This article cannot say which taxes would rise or which services would be reduced. Before release, confirm updated finance filings and any new official budget analysis or ballot changes. The campaign-finance snapshot used here ends September 28, 2026.

Evidence that would materially change the recommendation includes a credible, current estimate showing the sunset would have little net revenue effect after behavioral responses; evidence that the existing allocation rules fail to protect the services cited by supporters; or a concrete, durable replacement plan that maintains comparable education and health capacity at lower fiscal risk. Conversely, a strong evaluation linking current funding to significant outcomes or updated evidence showing modest behavioral response under continuation would strengthen YES.

sherafy.com recommendation: YES — moderate confidence

The decisive comparison is between continuing a substantial, existing revenue source and allowing it to expire without a replacement on the ballot. The LAO estimates $5–$15 billion a year after 2030, with the range itself warning against treating this money as stable. The top 2% pay the added marginal rates, while roughly 40% of the resulting funding supports schools and community colleges under constitutional allocation, public spending and audit rules. This is a real fiscal capacity and distributional benefit, even though the measure does not guarantee a specific classroom or health outcome.

The strongest NO evidence is consequential: voters were originally promised a temporary tax, the revenue stream is concentrated and market-sensitive, and peer-reviewed evidence found taxable-income responses after the 2012 increase. Those facts argue for moderate confidence rather than high confidence. They do not establish that continuing rates already in force will cause a comparable ongoing response, nor that the state can replace the projected revenue without equivalent taxes, spending reductions or reserve use.

Proposition 3 is a constitutional change without a sunset, so permanence deserves a higher burden than a short extension. But this is not a new tax on a newly defined population: the rates have operated since 2012, the allocation and audit system exists, and voters can later change the Constitution. Given the recurring public obligations supported by this revenue and the absence of a funded alternative, the foreseeable harm of expiration meets that burden more convincingly than the uncertain incremental harm of continuation. That is an editorial judgment about fiscal resilience and public services, not a finding that every dollar is efficiently spent.

I recommend YES with moderate confidence. A voter who gives greater weight to honoring the sunset promise, reducing exposure to volatile high-earner revenue, or requiring a fresh review before permanence can reasonably vote NO. A NO recommendation would be stronger if a credible replacement plan protected education and other services or if stronger evidence showed the extension’s net revenue would be small after behavioral responses.

Evidence Ledger

References and Further Reading

• California Secretary of State, Proposition 3 official title and summary — Ballot status, measure type, thresholds, education allocation and fiscal label.
• California Secretary of State, Proposition 3 analysis — Impartial voter-guide analysis and detailed fiscal context.
• Complete Proposition 3 proposed constitutional text — Operative constitutional amendments and existing allocation mechanisms.
• Legislative Analyst’s Office, Proposition 3 — Current legal summary, baseline and $5–$15 billion annual estimate.
• State Controller, Track Prop 30/Prop 55 — Revenue and disbursement tracking, education spending and audit framework; cumulative total is a projection through June 30, 2025.
• California Governor’s Budget, 2025–26 revenue estimates — Official evidence on capital-gains and high-income tax volatility. This is an earlier fiscal-year forecast, not a current 2026–27 budget estimate.
• Rauh and Shyu, “Behavioral Responses to State Income Taxation of High Earners: Evidence from California,” American Economic Journal: Economic Policy (2024) (full article PDF copy) — Peer-reviewed study of the 2012 Proposition 30 change; results are not a direct forecast for continuation.
• Varner, Young and Prohofsky, “Millionaire Migration in California: Administrative Data for Three Waves of Tax Reform” — Full working-paper PDF reviewed; authors identify it as work in progress with preliminary estimates, not a peer-reviewed final study.
• California campaign-finance snapshot for Proposition 3 — Secretary of State committee data compiled by CalMatters, last updated September 28, 2026; does not include all later filings or unreported outside activity.
• California Secretary of State, 2026 ballot-measure contribution totals — Official portal and its reporting-cutoff caveat.

Research currency: Updated October 10, 2026. Election records, fiscal estimates, campaign-finance filings and public records can change before Election Day.