Five-second answer

Claim we investigated

The state's revenue estimate for the millionaires tax assumes high earners won't change their behavior, though the leading study of California's 2012 increase found responses erased 45-61% of the expected new revenue, mostly through less reported income rather than moving.

True. The fiscal note assumes taxpayers won't reduce income; a California study found responses erased 45-61% of new revenue. The ESSB 6346 fiscal note states the no-response assumption in so many words, OFM's I-645 estimate adds only an IRS-enforcement cut and a compliance rate, and Rauh and Shyu's tax-return study of California's Proposition 30 measured 45.2% and 60.9% erosion, about 90% of it from reduced reported income.

Quote or cite this fact-check

FactHub rates the claim “The state's revenue estimate for the millionaires tax assumes high earners won't change their behavior, though the leading study of California's 2012 increase found responses erased 45-61% of the expected new revenue, mostly through less reported income rather than moving” True (high confidence): The fiscal note assumes taxpayers won't reduce income; a California study found responses erased 45-61% of new revenue. Reviewed October 22, 2026. https://facthub.org/n/article/wa-millionaires-tax-static-estimate/3

One-minute summary

What the evidence establishes

  • Established. Washington's fiscal note for the millionaires tax (ESSB 6346) states: "In response to the new tax, affected taxpayers do not reduce taxable income."
  • Established. The only other adjustments are a 4.71% cut to reported income for weaker IRS enforcement and a 90-95% compliance rate; neither is a response to Washington's tax.
  • Established. Rauh and Shyu found high earners' responses erased 45.2% of California's expected Proposition 30 windfall in the first year and 60.9% within two years; migration was 9.5% of that.
  • Established. Federal estimators (the Joint Committee on Taxation) include behavioral responses; state fiscal notes, including Washington's, generally do not.
  • Context. Most similar taxes still beat their official estimates: 12 of 18 comparisons we compiled for 10 taxes, with every shortfall but two in the 2008-09 recession.
  • Context. Smaller studies found migration alone cost about 8.5% (Multnomah County) and 1.5% (New Jersey) of the new revenue.
  • Not shown. The evidence does not show Washington will collect less than its $12.9 billion four-year estimate; market swings have mattered more than behavior.
Important context: Both halves of the claim hold, but they do not add up to "Washington will collect far less than projected." Behavioral erosion is measured against what the same taxpayers would have paid in the same economy, not against the official forecast. In 18 comparisons we compiled for 10 similar taxes, revenue beat the official estimate 12 times, including California's Proposition 30, whose income-tax piece came in 24-35% above the Legislative Analyst's estimate in the years checked despite the erosion Rauh and Shyu measured. Of the six comparisons that fell short, four were in 2008-09 recession years; the other two were Washington's capital gains tax in 2023-24 and Metro's housing tax against its original budget. Market and economic swings have moved these revenues far more than behavioral responses have.

The evidence

The conclusion rests on the combined record. Expand each item to see its role and limitations.

Independence note: This check uses 5 methods across 17 institutions and 1 country or international grouping.
Evidence map
Method Institution / country Year(s) What it shows Key limitation Primary links
Reading of the ESSB 6346 fiscal note (OFM package 77653) and OFM's Fiscal Impact Statement for Initiative 645 Washington Department of Revenue, Washington Office of Financial Management United States 2026 The fiscal note lists its assumptions in bullets, including that affected taxpayers do not reduce taxable income. Its only other haircuts are a 4.71% cut to taxpayers' federal adjusted gross income, which it ties to ITEP's estimate of weaker IRS enforcement, and compliance of 90% in tax year 2028 and 95% after. OFM's fiscal impact statement for I-645 repeats both adjustments; its only use of the word behavior refers to the compliance rate. On this basis the fiscal note projects $12.894 billion of gross collections over its first four years of collections (FY2029-32). Both documents are produced by the state agencies that administer and forecast the tax; we report what they assume, not whether the assumption is reasonable. The fiscal note is a point estimate and gives no range. Washington Office of Financial Management Washington Office of Financial Management FactHub dataset
Tax-return study comparing California top earners with matched out-of-state taxpayers after Proposition 30 raised top rates by up to 3 points Stanford University, Hoover Institution United States 2012, 2013, 2014 Rauh and Shyu found that the response of high earners who stayed (an elasticity of 2.5-3.2) plus extra departures eroded 45.2% of the expected windfall in the first year and 60.9% within two years. Migration accounted for 9.5% of the loss: an additional 0.8% of the top-bracket tax base left California in 2013. They note that the LAO never said whether its projections allowed for a behavioral response. Erosion is measured against the authors' no-response counterfactual, not against the official forecast, and the authors do not compare their result with LAO or Department of Finance projections or with collections. Proposition 30 raised rates on an existing tax, applied retroactively to 2012, and coincided with federal rate increases that moved income between years, so part of the response is timing. Joshua Rauh and Ryan Shyu
Migration studies using IRS county-to-county flows (Multnomah) and New Jersey tax returns (New Jersey 2004 increase) University of New Hampshire, Yale Budget Lab, Reed College, Stanford University, Princeton University United States 2004, 2006, 2021, 2022 Conway, Iselin and Rork estimate that added out-migration after Multnomah County's 2021 Preschool for All tax cost about $16 million, or 8.5% of the static revenue estimate, in 2022. Young and Varner put New Jersey's 2004 millionaire-tax yield at $1.08 billion in 2006 and the migration loss at $16.4 million a year, about 1.5% (our calculation). Both studies measure migration only, not reduced reported income, so they understate total behavioral response. The Multnomah paper is a working paper, not yet peer reviewed, and calls its estimate an upper bound; the New Jersey share is our calculation from the paper's figures. Conway, Iselin and Rork Cristobal Young and Charles Varner FactHub dataset
Review of federal and state revenue-estimating practice and the elasticity-of-taxable-income literature Joint Committee on Taxation, Montana Legislative Fiscal Division, University of California, Berkeley, Pew Charitable Trusts United States 2002, 2011, 2012, 2019, 2022 The Joint Committee on Taxation's conventional estimates include behavioral responses such as accelerating bonuses ahead of a tax increase. The literature puts the long-run elasticity of taxable income at 0.12 to 0.40 overall, and higher (0.57 in Gruber and Saez) for high earners. A 2022 Montana survey found no state produces dynamic fiscal notes as part of its routine work. Pew traced the growing volatility of state revenue largely to capital gains. Federal elasticities come from federal tax changes; state-level responses can be larger because moving or shifting income across a state line is easier. The Montana survey is a snapshot from 2022. Joint Committee on Taxation Emmanuel Saez, Joel Slemrod and Seth Giertz Jon Gruber and Emmanuel Saez Montana Legislative Fiscal Division Pew Center on the States and Rockefeller Institute of Government
Our compilation of 18 comparisons of an official (or, once, independent) revenue estimate with later collections or official re-estimates, for 10 state and local taxes on high incomes, 2004-2025 California Legislative Analyst's Office, Massachusetts Department of Revenue, Maryland Department of Legislative Services, Washington Department of Revenue, Metro (Portland), Tufts University United States 2006, 2007, 2009, 2012, 2014, 2017, 2022, 2024, 2025, 2026 Revenue beat the estimate in 12 of 18 comparisons, from 1.21 times (California Proposition 63) to 2.20 times (Washington's 2.9% capital gains surcharge). All four comparisons in 2008-09 recession years fell short (Maryland at 0.56 and 0.77, Oregon at 0.67, New York at 0.90); the other two shortfalls were Washington's capital gains tax in 2023-24 (0.84) and Metro's housing tax against its original budget (0.87), which Metro partly attributes to "a potential behavioral response." California's LAO warned in 2012 that taxpayers' uncertain responses made the revenue hard to estimate, yet Proposition 30 raised $5.6 billion in income tax in 2013-14 against $4.4 billion estimated. The set is small and not a random sample, and several outcomes are later official estimates or preliminary certifications rather than final collections; two rely on news reports. A ratio shows the forecast error, not its cause, so it cannot by itself show how large behavioral responses were. FactHub dataset FactHub dataset California Legislative Analyst's Office California Legislative Analyst's Office Massachusetts Department of Revenue Massachusetts Department of Revenue Tufts University Center for State Policy Analysis Maryland Department of Legislative Services Metro (Portland regional government) Washington Department of Revenue Washington Office of Financial Management
1. The state's estimates assume no behavioral response Reading of the ESSB 6346 fiscal note (OFM package 77653) and OFM's Fiscal Impact Statement for Initiative 645

The fiscal note lists its assumptions in bullets, including that affected taxpayers do not reduce taxable income. Its only other haircuts are a 4.71% cut to taxpayers' federal adjusted gross income, which it ties to ITEP's estimate of weaker IRS enforcement, and compliance of 90% in tax year 2028 and 95% after. OFM's fiscal impact statement for I-645 repeats both adjustments; its only use of the word behavior refers to the compliance rate. On this basis the fiscal note projects $12.894 billion of gross collections over its first four years of collections (FY2029-32).

Limit: Both documents are produced by the state agencies that administer and forecast the tax; we report what they assume, not whether the assumption is reasonable. The fiscal note is a point estimate and gives no range.

Verify: Multiple Agency Fiscal Note Summary, ESSB 6346 (2026), OFM Fiscal Note package 77653 Fiscal Impact Statement for Initiative 645 (July 2026) Washington millionaires tax (ESSB 6346): official revenue estimates by fiscal year

2. California's 2012 increase: 45-61% of the windfall eroded Tax-return study comparing California top earners with matched out-of-state taxpayers after Proposition 30 raised top rates by up to 3 points

Rauh and Shyu found that the response of high earners who stayed (an elasticity of 2.5-3.2) plus extra departures eroded 45.2% of the expected windfall in the first year and 60.9% within two years. Migration accounted for 9.5% of the loss: an additional 0.8% of the top-bracket tax base left California in 2013. They note that the LAO never said whether its projections allowed for a behavioral response.

Limit: Erosion is measured against the authors' no-response counterfactual, not against the official forecast, and the authors do not compare their result with LAO or Department of Finance projections or with collections. Proposition 30 raised rates on an existing tax, applied retroactively to 2012, and coincided with federal rate increases that moved income between years, so part of the response is timing.

Verify: Behavioral Responses to State Income Taxation of High Earners: Evidence from California (NBER Working Paper 26349; published in American Economic Journal: Economic Policy 16(1), 2024)

3. Other studies find smaller, migration-only losses Migration studies using IRS county-to-county flows (Multnomah) and New Jersey tax returns (New Jersey 2004 increase)

Conway, Iselin and Rork estimate that added out-migration after Multnomah County's 2021 Preschool for All tax cost about $16 million, or 8.5% of the static revenue estimate, in 2022. Young and Varner put New Jersey's 2004 millionaire-tax yield at $1.08 billion in 2006 and the migration loss at $16.4 million a year, about 1.5% (our calculation).

Limit: Both studies measure migration only, not reduced reported income, so they understate total behavioral response. The Multnomah paper is a working paper, not yet peer reviewed, and calls its estimate an upper bound; the New Jersey share is our calculation from the paper's figures.

Verify: Mult-No-More? Migration responses to Multnomah County's Preschool for All tax (working paper, April 2026) Millionaire Migration and State Taxation of Top Incomes: Evidence from a Natural Experiment (National Tax Journal 64(2), 2011) Top-rate and millionaire taxes: published behavioral erosion estimates

4. Federal estimators count behavior; state fiscal notes usually don't Review of federal and state revenue-estimating practice and the elasticity-of-taxable-income literature

The Joint Committee on Taxation's conventional estimates include behavioral responses such as accelerating bonuses ahead of a tax increase. The literature puts the long-run elasticity of taxable income at 0.12 to 0.40 overall, and higher (0.57 in Gruber and Saez) for high earners. A 2022 Montana survey found no state produces dynamic fiscal notes as part of its routine work. Pew traced the growing volatility of state revenue largely to capital gains.

Limit: Federal elasticities come from federal tax changes; state-level responses can be larger because moving or shifting income across a state line is easier. The Montana survey is a snapshot from 2022.

Verify: Overview of the Revenue Estimating Process (February 2019) The Elasticity of Taxable Income with Respect to Marginal Tax Rates: A Critical Review (Journal of Economic Literature 50(1), 2012) The Elasticity of Taxable Income: Evidence and Implications (NBER Working Paper 7512; Journal of Public Economics, 2002) Dynamic Fiscal Notes Study (March 2022) States' Revenue Estimating: Cracks in the Crystal Ball (2011)

5. Most similar taxes still beat their official estimates Our compilation of 18 comparisons of an official (or, once, independent) revenue estimate with later collections or official re-estimates, for 10 state and local taxes on high incomes, 2004-2025

Revenue beat the estimate in 12 of 18 comparisons, from 1.21 times (California Proposition 63) to 2.20 times (Washington's 2.9% capital gains surcharge). All four comparisons in 2008-09 recession years fell short (Maryland at 0.56 and 0.77, Oregon at 0.67, New York at 0.90); the other two shortfalls were Washington's capital gains tax in 2023-24 (0.84) and Metro's housing tax against its original budget (0.87), which Metro partly attributes to "a potential behavioral response." California's LAO warned in 2012 that taxpayers' uncertain responses made the revenue hard to estimate, yet Proposition 30 raised $5.6 billion in income tax in 2013-14 against $4.4 billion estimated.

Limit: The set is small and not a random sample, and several outcomes are later official estimates or preliminary certifications rather than final collections; two rely on news reports. A ratio shows the forecast error, not its cause, so it cannot by itself show how large behavioral responses were.

Verify: Top-rate and millionaire taxes: official revenue estimates vs. actual collections (ratio by comparison) Top-rate and millionaire taxes: comparisons above and below the estimate Proposition 30: Temporary Taxes to Fund Education (ballot analysis, November 2012) Proposition 30 revenues (November 2014) Fiscal Year 2024 Revenue Collections Totaled $40.800 Billion (August 9, 2024) Fiscal Year 2025 Revenue Collections Totaled $43.708 Billion Evaluating the Massachusetts Millionaires Tax (January 2022) Fiscal and Policy Note, House Bill 784 (2012) Supportive Housing Services FY 2024-25 year-end revenue report Capital Gains Tax - Tax Year 2025 Initial Collections Multiple Agency Fiscal Note Summary, SSB 5096 (2021), capital gains tax, OFM package 63363

Forecast vs. actual: Capital gains tax revenue (gross), FY2023 (tax year 2022)
Forecast
527 US dollars, millions, by Washington Department of Revenue,
Actual
782 US dollars, millions (Tax year 2022 net payments as of June 9, 2026) · Actual revenue as a multiple of the estimate
Error (actual minus forecast)
+255 US dollars, millions (+48.39%)
Note
Gross capital gains tax lines of the fiscal note; the note's $442 million state total also subtracts a B&O tax credit.

Source: Washington Office of Financial Management (DOR fiscal note) · View in the dataset

Forecast vs. actual: 4% surtax revenue (dynamic estimate), 2023
Forecast
1300 US dollars, millions, by Tufts Center for State Policy Analysis,
Actual
2199 US dollars, millions (FY2024 (DOR preliminary certification)) · Actual revenue as a multiple of the estimate
Error (actual minus forecast)
+899 US dollars, millions (+69.15%)
Note
Approximate date (January 2022). Estimate is for calendar 2023; compared with FY2024, the first full fiscal year of collections. The static estimate was $2.1 billion.

Source: Tufts University Center for State Policy Analysis · View in the dataset

Most top-rate tax estimates were beaten; recession years fell shortActual revenue divided by the official estimate for 18 comparisons across 10 state and local taxes on high incomes. Above 1 means revenue beat the estimate.{"width":960,"height":540,"plot":{"left":90,"top":70,"right":926,"bottom":482},"yDomain":[0,2.5],"series":[{"name":"Expansion","color":"#176b87"},{"name":"Recession (2008-09)","color":"#b6542c"},{"name":"Market decline (2022)","color":"#684b9b"}],"points":[{"series":0,"x":90,"y":277.15,"xLabel":"CA Prop 30 FY12","yLabel":"1.2"},{"series":0,"x":139.18,"y":277.15,"xLabel":"CA Prop 30 FY13","yLabel":"1.2"},{"series":0,"x":188.35,"y":270.23,"xLabel":"CA Prop 30 FY14","yLabel":"1.3"},{"series":0,"x":237.53,"y":259.03,"xLabel":"CA Prop 30 FY17","yLabel":"1.4"},{"series":0,"x":286.71,"y":282.92,"xLabel":"CA Prop 63 FY06","yLabel":"1.2"},{"series":0,"x":335.88,"y":279.3,"xLabel":"CA Prop 63 FY07","yLabel":"1.2"},{"series":0,"x":385.06,"y":256.22,"xLabel":"MA surtax FY24 (DOR)","yLabel":"1.4"},{"series":0,"x":434.24,"y":203.16,"xLabel":"MA surtax FY24 (Tufts)","yLabel":"1.7"},{"series":1,"x":483.41,"y":333.68,"xLabel":"NY surcharge FY10","yLabel":"0.9"},{"series":1,"x":532.59,"y":355.76,"xLabel":"MD $1M bracket TY08","yLabel":"0.8"},{"series":1,"x":581.76,"y":390.37,"xLabel":"MD $1M bracket TY09","yLabel":"0.6"},{"series":1,"x":630.94,"y":372.08,"xLabel":"OR Measure 66 2009-11","yLabel":"0.7"},{"series":0,"x":680.12,"y":206.78,"xLabel":"Multnomah PFA FY22","yLabel":"1.7"},{"series":0,"x":729.29,"y":255.56,"xLabel":"Multnomah PFA FY25","yLabel":"1.4"},{"series":0,"x":778.47,"y":338.95,"xLabel":"Metro SHS FY25","yLabel":"0.9"},{"series":2,"x":827.65,"y":237.44,"xLabel":"WA cap gains TY22","yLabel":"1.5"},{"series":0,"x":876.82,"y":343.24,"xLabel":"WA cap gains TY23-24","yLabel":"0.8"},{"series":0,"x":926,"y":119.77,"xLabel":"WA surcharge TY25","yLabel":"2.2"}]}Most top-rate tax estimates were beaten; recession years fell short0.00.51.01.52.02.5CA Prop 30 FY12CA Prop 30 FY14CA Prop 63 FY07MA surtax FY24 (Tufts)MD $1M bracket TY09Multnomah PFA FY22WA cap gains TY22WA surcharge TY25Tax and revenue periodExpansionRecession (2008-09)ExpansionRecession (2008-09)Market decline (2022)Revenue equal to estimateFactHub · Actual revenue ÷ estimate
Actual revenue divided by the official estimate for 18 comparisons. Points above the line beat the estimate; the four recession-year comparisons fell short. A ratio shows the forecast error, not how much revenue behavioral responses removed. View source chart

Common questions

“Rauh and Shyu prove Washington will collect only about half the projected revenue.”

Short answer: No

Their 45-61% is the share of California's expected windfall lost against a no-response counterfactual in the same economy, not a comparison with the official forecast. Proposition 30's income-tax revenue still came in about 24-35% above the LAO's estimate in the years we checked, because rising markets added more than responses took away. Their result also comes from a rate increase on an existing tax with a retroactive year and federal fiscal-cliff timing. It is evidence that responses can be large, not a forecast for Washington.

“If most of these taxes beat their forecasts, behavioral responses must be small.”

Short answer: Not necessarily

Beating a forecast and losing revenue to behavior can happen at once. Multnomah County's Preschool for All tax collected $187 million in its first year against a $112 million forecast, while researchers estimate migration cost about 8.5% of its static revenue. Massachusetts collected $2.2 billion in FY2024, well above both the Revenue Department's range and Tufts' behavior-adjusted $1.3 billion, but that does not tell us how much more it would have collected with no response.

“The 4.71% cut in the state's estimate is a behavioral adjustment.”

Short answer: No

Both state documents tie it to reduced IRS enforcement, which ITEP expects to lower the income people report to the IRS. It would apply whether or not Washington had a new tax. The compliance rates (90%, then 95%) are the only behavior-related assumption, and they concern payment, not how much income people earn, realize or report.

“Millionaires are leaving Washington, so the revenue won't come.”

Short answer: Too early to measure

No collections or post-enactment migration data exist yet; the first payments are due in April 2029. The research suggests moving is the smaller channel: in California migration was 9.5% of the measured erosion, and in Multnomah and New Jersey migration alone cost about 8.5% and 1.5% of the new revenue. Reduced or shifted reported income matters more, and Washington's 2026-to-2028 gap gives high earners time to shift income.

Related claim: Columbia Bank is moving its top executives to Arizona over Washington taxes. — Columbia says new Washington taxes cost it about $8 million a year; its CEO and most executive roles are moving.

“So the millionaires tax will raise less than the state projects.”

Short answer: Unknown

A static estimate is likely high compared with what the same economy would produce, but actual revenue will depend mostly on markets. Washington's capital gains tax collected $782 million in its first year, 1.48 times the fiscal note's gross capital gains estimate of $527 million (1.77 times if measured against the note's $442 million whole-bill total, which also subtracts a B&O tax credit), and then 0.84 times the gross estimate over the next two years. The 2025 surcharge raised about 2.2 times its estimate. Pew found capital gains are the main source of volatility in state revenue forecasts.

“Other states account for behavior in their estimates, so Washington is an outlier.”

Short answer: No, it is typical

None of the official estimates we checked in California, Maryland, New Jersey, New York, Oregon or Washington stated a behavioral adjustment for the income tax, and a 2022 survey found no state produces dynamic fiscal notes routinely. Federal estimators do include behavior. Washington's estimate is standard state practice, which is also why its assumption is worth knowing.

Conclusion

The claim is accurate on both counts. Washington's fiscal note for ESSB 6346 says plainly that "affected taxpayers do not reduce taxable income," and the OFM fiscal impact statement for I-645 adds only two adjustments: a 4.71% cut to reported income because ITEP expects weaker IRS enforcement, and a 90-95% compliance rate. Neither is a response to Washington's tax. That is ordinary state practice (a 2022 Montana survey found no state routinely produces dynamic fiscal notes), but it differs from federal practice: the Joint Committee on Taxation's conventional estimates include behavioral responses. The research says such responses are real and can be large at the state level. Rauh and Shyu found Proposition 30's top-bracket windfall eroded 45.2% in the first year and 60.9% within two years, with migration only 9.5% of the total; smaller studies found migration alone cost about 8.5% of Multnomah County's Preschool for All revenue and about 1.5% of New Jersey's 2004 increase. What the evidence does not support is the step from "erosion" to "shortfall." Erosion is measured against a counterfactual in the same economy; the official forecast is a different benchmark. In our compilation of 18 estimate-vs-actual comparisons for 10 taxes, revenue beat the estimate in 12. California's Proposition 30, Multnomah's Preschool for All and the Massachusetts surtax all beat their forecasts even though researchers measured or projected behavioral losses for each; in Multnomah, the researchers' static simulation ($189 million) almost exactly matched actual 2022 revenue ($187 million). The only pre-enactment estimate in the set that cut revenue for behavior, Tufts' $1.3 billion for Massachusetts, was the furthest below actual collections ($2.2 billion in FY2024). All four shortfalls in the 2008-09 recession (New York, Maryland twice, Oregon) came when incomes fell; Oregon's first year was taxed retroactively, so no one could have changed behavior to avoid it. Washington's own capital gains tax shows the same pattern: 1.48 times the fiscal note's gross estimate in tax year 2022, then 0.84 times it for 2023-24, driven by a handful of very large payments. We rate the claim True with high confidence because both halves rest on primary documents and a peer-reviewed study. Three cautions limit what it implies. California's 45-61% measured a rate increase on an existing tax with a retroactive year and fiscal-cliff timing, so it is an upper-end reference, not a forecast for Washington. Washington's tax is new rather than an increase, its rate is higher (9.9%), and it was enacted in March 2026 but applies from January 2028, which gives high earners time to shift income into 2027 or move, something California's Proposition 30 and Washington's 2025 surcharge did not allow. Whether Washington ends up above or below its $12.9 billion four-year estimate will depend more on markets than on behavior, and that remains unknown until collections begin in April 2029.

How we checked

  1. Claim definition: we rated two linked statements: that Washington's official revenue estimate for ESSB 6346 assumes no behavioral response, and that the leading study of California's 2012 increase found responses erased 45-61% of the expected new revenue, mostly through reported income rather than migration.
  2. We read the ESSB 6346 fiscal note (OFM package 77653) and OFM's I-645 fiscal impact statement in full for every assumption about taxpayer behavior, and identified what the 4.71% adjustment represents.
  3. We read the Rauh and Shyu working paper (NBER w26349) for its erosion measure, its counterfactual and what it says about official projections.
  4. To test whether behavioral erosion shows up as revenue below forecast, we compiled every top-rate or millionaire tax since 2004 for which both an official estimate and a later collection figure were published (18 comparisons, 10 taxes), preferring legislative analysts' and revenue departments' own documents, and marked values from news reports as unverified. Taxes with no published bracket-level collections (New Jersey 2018 and 2020, New York 2021) were excluded.
  5. We compared Washington's capital gains fiscal note with Department of Revenue collections on a gross basis, because the note's state total also subtracts a B&O tax credit.
  6. We reviewed federal practice (Joint Committee on Taxation), a survey of state practice (Montana Legislative Fiscal Division) and the elasticity literature for context.
  7. Limits recorded: erosion estimates and forecast errors measure different things; several outcomes are re-estimates; the comparison set is small and dominated by the 2008-09 recession and the later expansion.

Editorial note: Scope: this article checks what Washington's official revenue estimate assumes about behavior and what the leading research found, and it compares official estimates with collections for similar taxes. It does not forecast Washington's revenue. The 18-comparison table is our compilation; some outcomes are later official estimates or preliminary certifications, and two rely on news reports (marked in the dataset). FactHub does not take a position on Initiative 645.

Expert-review status

No external specialist review has been recorded.

Funding and conflicts

Funding: No external funding for this piece.

Ownership: Published by the FactHub Research Team.

Conflicts: None known. FactHub takes no position on Initiative 645.

Sources

Sources are labeled by kind, with what this fact check takes from each, so readers can check the evidence independently.

  1. Government Washington Office of Financial Management. Multiple Agency Fiscal Note Summary, ESSB 6346 (2026), OFM Fiscal Note package 77653

    The Legislature's fiscal note for the millionaires tax as passed, with the Department of Revenue's revenue estimate and assumptions.

    Reliability: Prepared by the Department of Revenue for the Legislature in March 2026; the official basis for the tax's revenue estimate.

    • The estimate assumes taxpayers subject to the tax do not reduce their taxable income in response to it.

      What the source says
      In response to the new tax, affected taxpayers do not reduce taxable income.
      Department of Revenue note, assumptions for the tax on income

    • The 4.71% cut to reported income reflects ITEP's estimate of weaker IRS enforcement, not a response to Washington's tax.

      What the source says
      Due to the reduced IRS enforcement budget, the Institute on Taxation and Economic Policy (ITEP) predicts a decline in federal collections. Based on the ITEP analysis, we applied a 4.71% downward adjustment to taxpayers' AGI before estimating collections.
      Department of Revenue note, assumptions for the tax on income

    • The estimate assumes 90% compliance in the first tax year and 95% after.

      What the source says
      90% revenue collections in tax year 2028, and…95% revenue collections in tax year 2029 and thereafter.
      Department of Revenue note, assumptions for the tax on income
  2. Government Washington Office of Financial Management. Fiscal Impact Statement for Initiative 645 (July 2026)

    OFM's official estimate of the revenue lost if I-645 passes, printed in the voters' pamphlet.

    Reliability: Required by state law for every initiative; based on the June 2026 ERFC forecast.

    • OFM's I-645 estimate applies the same 4.71% income cut, tied to ITEP's IRS-enforcement analysis.

      What the source says
      Based on the ITEP analysis, a 4.71 percent downward adjustment is applied to taxpayers' Adjusted Gross Income (AGI) prior to estimating tax collections.
      Assumptions, Revenue

    • The statement's only reference to taxpayer behavior concerns the compliance rate.

      What the source says
      Due to assumed taxpayer behavior, an assumption about taxpayer compliance is applied to revenue estimates.
      Assumptions, Revenue
  3. FactHub record FactHub dataset. Washington millionaires tax (ESSB 6346): official revenue estimates by fiscal year

    FactHub compilation of the fiscal note, ERFC and OFM estimates by fiscal year and window of years.

    • The fiscal note projects $12.894 billion of gross collections over FY2029-32, its first four years of collections.

  4. Peer-reviewed study Joshua Rauh and Ryan Shyu. Behavioral Responses to State Income Taxation of High Earners: Evidence from California (NBER Working Paper 26349; published in American Economic Journal: Economic Policy 16(1), 2024) Free copy

    Tax-return study of how California's highest earners responded to Proposition 30's top-rate increase.

    Reliability: Uses confidential California tax returns; Rauh is a Stanford professor and Hoover Institution fellow. The journal version is peer reviewed.

    • Responses erased 45.2% of California's expected windfall in the first year and 60.9% within two years; migration was 9.5% of the total.

      What the source says
      These responses together eroded 45.2% of state windfall tax revenues within the first year and 60.9% within two years, with the extensive margin accounting for 9.5% of this total.
      Abstract

    • An additional 0.8% of the top-bracket tax base left California in 2013.

      What the source says
      an additional 0.8% of the residential tax base that landed in the top bracket left California in 2013
      Abstract

    • The authors say the LAO never stated whether its Proposition 30 projections allowed for a behavioral response.

      What the source says
      the LAO never explicitly mentioned the degree to which (if at all) the possibility of a behavioral response factored into their projections
      Section 3
  5. Academic Conway, Iselin and Rork. Mult-No-More? Migration responses to Multnomah County's Preschool for All tax (working paper, April 2026)

    Working paper estimating out-migration and revenue effects of the Portland-area Preschool for All income tax.

    Reliability: Economists at the University of New Hampshire, the Yale Budget Lab and Reed College; not yet peer reviewed. Uses public IRS migration data.

    • Added out-migration cost about $16 million, or 8.5% of the static Preschool for All revenue estimate.

      What the source says
      The $1.4 billion in AGI would have brought in approximately $16 million in PFA revenue, or about 8.5 percent of the conventional revenue estimate.
      Section 6.3

    • Qualifies: The authors regard their migration effects as an upper bound.

      What the source says
      This is another reason why we see our estimated effects as an upper bound on the effects of PFA.
      Section 2

    • Qualifies: The researchers' static simulation almost exactly matched actual 2022 revenue, so the tax met a no-behavior benchmark despite the migration loss.

      What the source says
      Notably, this closely matches actual 2022 PFA revenue of $187 million
      Section 6.2
  6. Peer-reviewed study Cristobal Young and Charles Varner. Millionaire Migration and State Taxation of Top Incomes: Evidence from a Natural Experiment (National Tax Journal 64(2), 2011)

    Peer-reviewed study of migration after New Jersey's 2004 tax increase on incomes over $500,000 (copy posted by the Vermont Legislature).

    Reliability: Uses New Jersey tax-return data; measures migration only.

    • Qualifies: The 2004 increase yielded about $1.08 billion in 2006 on a static basis.

      What the source says
      The difference attributable to the new tax, shown in column three, is $1.08 billion.
      Section on revenue yield, Table 4

    • Qualifies: Migration cost New Jersey about $16.4 million a year in income tax revenue.

      What the source says
      a loss of 69.7 millionaire tax filers results in losses of $194.7 million in taxable income and $16.4 million in state income tax revenues
      Section on revenue effects
  7. Government Joint Committee on Taxation. Overview of the Revenue Estimating Process (February 2019)

    Congress's nonpartisan tax scorekeeper explains how its revenue estimates are built.

    Reliability: Official description of federal revenue-estimating practice.

    • Federal conventional revenue estimates include behavioral responses.

      What the source says
      A conventional JCT estimate incorporates behavioral responses in projecting tax revenues, but generally assumes that these tax and behavioral changes do not change the size of the US economy
      Limits of the Conventional Estimate

    • One response federal estimators count is pulling bonuses forward ahead of a tax increase.

      What the source says
      Acceleration of bonuses in anticipation of an individual income tax increase.
      List of behavioral responses
  8. Government Montana Legislative Fiscal Division. Dynamic Fiscal Notes Study (March 2022)

    Survey of how state legislatures produce fiscal notes and whether they include economic or behavioral effects.

    Reliability: Staff study for the Montana Legislature surveying other states' practice.

    • No state produces dynamic fiscal notes as routine practice.

      What the source says
      While no states provide dynamic fiscal notes as part of their daily work for legislatures, a few states publish occasional, limited dynamic analysis.
      State by State Survey Summary
  9. Peer-reviewed study Emmanuel Saez, Joel Slemrod and Seth Giertz. The Elasticity of Taxable Income with Respect to Marginal Tax Rates: A Critical Review (Journal of Economic Literature 50(1), 2012)

    Leading review of research on how reported income responds to tax rates.

    Reliability: Survey by economists at UC Berkeley, the University of Michigan and the University of Minnesota.

    • The best long-run estimates of the elasticity of taxable income range from 0.12 to 0.40.

      What the source says
      While there are no truly convincing estimates of the long-run elasticity, the best available estimates range from 0.12 to 0.40.
      Conclusion
  10. Peer-reviewed study Jon Gruber and Emmanuel Saez. The Elasticity of Taxable Income: Evidence and Implications (NBER Working Paper 7512; Journal of Public Economics, 2002)

    Study estimating how taxable income responds to tax rates, by income group.

    Reliability: Uses federal tax-return panel data from the 1980s.

    • Taxpayers with incomes above $100,000 had an elasticity of 0.57, much higher than others.

      What the source says
      taxpayers who have incomes above $100,000 per year, who have an elasticity of 0.57
      Abstract
  11. Other Pew Center on the States and Rockefeller Institute of Government. States' Revenue Estimating: Cracks in the Crystal Ball (2011)

    Study of state revenue forecast errors from 1987 to 2009.

    Reliability: Nonpartisan research by Pew and the Rockefeller Institute, based on NASBO data for all 50 states.

    • Qualifies: Volatile capital gains and growing reliance on income taxes are the main causes of rising forecast volatility.

      What the source says
      The main causes of the increase in volatility appear to be states' growing reliance on income taxes and the ways in which highly volatile capital gains affect income tax revenue.
      Summary
  12. Government California Legislative Analyst's Office. Proposition 30: Temporary Taxes to Fund Education (ballot analysis, November 2012)

    Official ballot-pamphlet fiscal analysis of Proposition 30.

    Reliability: Nonpartisan fiscal adviser to the California Legislature.

    • The LAO warned that taxpayers' uncertain responses made Proposition 30's revenue hard to estimate, without quantifying them.

      What the source says
      Due to these swings in the income of these taxpayers and the uncertainty of their responses to the rate increases, the revenues raised by this measure are difficult to estimate.
      Fiscal Effects
  13. Government California Legislative Analyst's Office. Proposition 30 revenues (November 2014)

    LAO estimate of what Proposition 30 raised by fiscal year.

    Reliability: LAO estimate; it notes amounts remain subject to change.

    • Qualifies: Proposition 30 raised $5.6 billion of income tax in 2013-14, against $4.4 billion in the LAO's 2012 estimate.

      What the source says
      $7.0 billion of revenues ($5.6 billion PIT and $1.4 billion sales tax)
      2013-14 entry
  14. Academic Tufts University Center for State Policy Analysis. Evaluating the Massachusetts Millionaires Tax (January 2022)

    Pre-enactment estimate of the Massachusetts surtax, with and without behavioral responses.

    Reliability: Nonpartisan university policy center; the only pre-enactment estimate in our set that cut revenue for behavior.

    • Qualifies: Tufts projected $1.3 billion for 2023 after cutting about 35% for avoidance and migration.

      What the source says
      Factoring in expected behavioral changes by high earners, the Massachusetts millionaires tax would raise about $1.3 billion in 2023
      Key findings
  15. Government Massachusetts Department of Revenue. Fiscal Year 2024 Revenue Collections Totaled $40.800 Billion (August 9, 2024)

    DOR press release reporting FY2024 collections, including the surtax certification.

    Reliability: Official preliminary certification; later certifications may revise it.

    • Qualifies: The surtax raised $2.199 billion in FY2024, its first full fiscal year.

      What the source says
      On July 24, 2024, DOR certified that the preliminary estimate of surtax revenue collected in fiscal year 2024 was $2.199 billion
      Paragraph 4
  16. Government Massachusetts Department of Revenue. Fiscal Year 2025 Revenue Collections Totaled $43.708 Billion

    DOR press release reporting FY2025 collections, including the surtax certification.

    Reliability: Official preliminary certification for FY2025.

    • Qualifies: The surtax raised $2.987 billion in FY2025.

      What the source says
      On July 21, 2025, DOR certified that the preliminary estimate of surtax revenue collected in fiscal year 2025 was $2.987 billion
      Body text
  17. Government Maryland Department of Legislative Services. Fiscal and Policy Note, House Bill 784 (2012)

    Fiscal note quoting the Comptroller on revenue from Maryland's expired 6.25% millionaire bracket.

    Reliability: Nonpartisan legislative staff quoting the Comptroller's Office.

    • Qualifies: Maryland's millionaire bracket raised $78.2 million and $60.8 million in 2008 and 2009, against estimates of $102.1 million and $109.3 million.

      What the source says
      The Comptroller's Office advises that the revenue increase from the 6.25% rate totaled $78.2 million in tax year 2008, $60.8 million in tax year 2009, and $75.0 million in tax year 2010.
      Current Law
  18. Government Metro (Portland regional government). Supportive Housing Services FY 2024-25 year-end revenue report

    Metro's report on collections of its income tax for supportive housing.

    Reliability: Metro administers and spends the tax.

    • Metro attributes its 2023-24 shortfall partly to a possible behavioral response by taxpayers or migrants.

      What the source says
      FY 2023-24 collections were below expectations, reflecting both poorer local economic conditions and a potential behavioral response from taxpayers and/or migrants to the area.
      FY 2024-25 Tax Collections
  19. Government Washington Department of Revenue. Capital Gains Tax - Tax Year 2025 Initial Collections

    DOR's report of capital gains tax net payments by tax year.

    Reliability: Figures are net payments as of June 9, 2026, and may change.

    • Qualifies: Washington's capital gains tax collected $782.0 million for tax year 2022, $394.7 million for 2023 and $583.8 million for 2024.

      What the source says
      2024 - $583.8 million, 6,727 net payments…2022 - $782.0 million, 3,625 net payments
      Net payments by tax year
  20. Government Washington Office of Financial Management. Multiple Agency Fiscal Note Summary, SSB 5096 (2021), capital gains tax, OFM package 63363

    The Legislature's fiscal note for Washington's 2021 capital gains tax.

    Reliability: Department of Revenue estimate, revised June 22, 2021.

    • The capital gains estimate projected gains forward from 2018 returns, with no behavioral response to the tax.

      What the source says
      Future amounts of capital gains for taxpayers equal the amounts reported in 2018 tax returns, adjusted by using forecasted rates of growth and decay.
      Department of Revenue note, assumptions
  21. FactHub record FactHub dataset. Top-rate and millionaire taxes: official revenue estimates vs. actual collections (ratio by comparison)

    Our compilation of 18 estimate-vs-actual comparisons for 10 taxes.

    • Qualifies: Actual revenue ranged from 0.56 to 2.20 times the estimate; California's Proposition 30 came in at 1.24 to 1.35 times the LAO's estimate in the years checked.

    • Qualifies: Against the capital gains fiscal note's gross tax estimate ($527M for FY2023), Washington's tax came in at 1.48 times for tax year 2022 and 0.84 for 2023-24; the note's $442M whole-bill total, net of a B&O credit, gives 1.77. The 2025 surcharge: 2.20.

  22. FactHub record FactHub dataset. Top-rate and millionaire taxes: comparisons above and below the estimate

    Counts of comparisons above and below the estimate, by economic setting.

    • Qualifies: Revenue beat the estimate in 12 of 18 comparisons (11 of 16 using verified values only).

    • Qualifies: All 4 comparisons in 2008-09 recession years fell short; 12 of the other 14 came in above.

  23. FactHub record FactHub dataset. Top-rate and millionaire taxes: published behavioral erosion estimates

    Published estimates of how much revenue behavioral responses removed.

    • Measured erosion ranges from about 1.5% and 8.5% (migration only) to 45.2% and 60.9% (California, all responses); Washington's estimate assumes none.

  24. FactHub dataset Top-rate and millionaire taxes: official revenue estimates vs. actual collections

    • Tables: forecast-vs-actual

      Saved analysis ratio-by-comparison

    • Tables: forecast-vs-actual

      Forecast wa-capgains-fy2023

    • Tables: forecast-vs-actual

      Forecast ma-surtax-fy2024-tufts

  25. FactHub dataset Washington millionaires tax (ESSB 6346): official revenue estimates by fiscal year

    Tables: estimates-by-fy

    Saved analysis window-totals, fiscal note FY2029-32

  26. FactHub chart top-rate-taxes-actual-vs-estimate

Update history

  1. First version.
  2. Version 2: new main photo; stated that capital gains ratios use the 2021 fiscal note's gross capital gains estimate ($527 million for FY2023), and noted that the note's $442 million whole-bill total gives 1.77; disclosure wording updated. First Look (version 1) reported no findings.
  3. Version 3: the context note now says six comparisons fell short, four of them in 2008-09 recession years, matching the evidence and the dataset.

Cited by

Version details

The exact citing versions, and the version of this article each cites.

  • Briefing wa-i-645 — version 6 (published) cites version 3
  • Briefing wa-i-645 — version 4 (published) cites version 3

For search, research, and AI systems

Machine-readable fact-check

This page includes standardized ClaimReview metadata describing the exact claim, FactHub's finding, publisher, and review dates. You can inspect the record below and confirm that it matches the article you just read.

View ClaimReview JSON 
{"@context":"https://schema.org","@type":"ClaimReview","url":"https://facthub.org/n/article/wa-millionaires-tax-static-estimate/3","datePublished":"2026-10-22","dateModified":"2026-10-22","author":{"@type":"Organization","name":"FactHub","url":"https://facthub.org/"},"claimReviewed":"The state\u0027s revenue estimate for the millionaires tax assumes high earners won\u0027t change their behavior, though the leading study of California\u0027s 2012 increase found responses erased 45-61% of the expected new revenue, mostly through less reported income rather than moving.","reviewRating":{"@type":"Rating","alternateName":"True"},"image":"https://facthub.org/node-assets-v2/article/wa-millionaires-tax-static-estimate/3/bellevue-skyline/social-1200x630.jpg"}

This metadata helps machines discover the fact-check. It does not prove the conclusion or guarantee special treatment by a search engine.

Challenge this fact-check

Every source above is linked so you can check it yourself. Identify the specific conclusion you dispute and provide verifiable evidence that could change it. Credible challenges deserve review and a visible response.

Submit evidence Send general feedback instead

Technical details: node handle article:wa-millionaires-tax-static-estimate