Prop 40 comparison: Overview · Open Output (here) · Open Analysis: Report and Slides · Open Materials Also in this project: BSZ analysis explorer · JR analysis explorer

Why do the two analyses of Proposition 40 disagree?

The BSZ analysis (Boll, Saez and Zucman, NBER, August 2026) finds that the billionaire tax raises money for California. The JR analysis (Jaros and Rauh, NBER, September 2026) finds that it loses money. Open Policy Analysis overview This page puts both on one measure, the net fiscal effect: the wealth tax collected, plus extra income tax from asset sales, minus the income tax lost to people who leave. This is part of an Open Policy Analysis. See more here:

BSZ main estimate1 $106.8B Net total: the wealth tax, plus income tax from assets sold to pay it, minus income tax lost to leavers over 5 years, in today's dollars.
JR main estimate1 −$39.0B Net present value: the wealth tax minus the income tax lost for ever, in today's dollars.

Both numbers rest on decisions this comparison makes, not the authors. The BSZ number adds income tax effects to the paper's headline, the wealth tax alone. The JR number is the average of the authors' 100,000 draws, which spread from −$208.8B to $47.7B. Each keeps its own horizon, 5 years against for ever. The six decisions, in the report.

What separates the two numbers

Each amount is what its dials change when they are moved to the JR values, term by term from left to right, as the Walk the path button does.2

The inputs

Each dial (a slider) moves one disputed input between the two analyses' values. Every dial has its smaller value at the left end and its larger value at the right end, and each end names whose value it is. The BSZ panel starts at the BSZ values, the JR panel at the JR values, so a dial can start at either end. Tap or hover over a symbol or an end for what it means and the kind of each value.

BSZ: $106.8B

The BSZ analysis's main estimate1 (Boll, Saez and Zucman, NBER, August 2026)

With these sliders $106.8B

  • Wealth tax collectedWealth tax collected = 5% × (W − D) × (1 − a) × q

    The tax takes 5% of the wealth. W is the wealth on the list, D the part counted as moved out in time, a the share lost to avoidance, q the chance the tax survives in court. moreThe four dials multiply, so one dial's effect depends on the others. W and D are valued on the same Forbes list.

    • W, wealth on the list. The billionaires' wealth the tax applies to, from one of two Forbes lists.
      Forbes 2025JRJR: the 2025 list, 212 billionaires, grown 7%, the Act's growth figure, to the valuation date (31 December 2026): $2,027B. Kind: guesswork on data.Forbes 2026Forbes 2026BSZBSZ: the Forbes list on 1 July 2026, 250 California billionaires worth $2,307B, or $2,157B without the non-US citizens (a separate dial). Kind: data.
    • D, wealth counted as moved out. It left before the 1 January 2026 residency date. Middle stop: a BSZ scenario counting Page, Thiel, Hankey and Kalanick as moved.
      None namedBSZBSZ: half of the 10% avoidance, 5% of the wealth, is people who moved in time. No one is named without evidence of a legal move in time. Kind: guesswork on data.None named7 movesJRJR: seven counted as moved in time, 28.6% of the wealth, on public reports (Page, Thiel, Hankey, Kalanick, Brin, Spielberg, Sacks). Kind: guesswork on data.
    • a, avoidance share. The share of the wealth that escapes the tax through avoidance and evasion, including moves no one names (D holds the named ones).
      0.0%JRJR: none beyond the departures it counts, since US securities are third-party reported. Its draws cover unreported moves. Kind: guesswork.10.0%10.0%BSZBSZ: a small allowance argued from enforcement. Half is people who moved out in time, who also cost income tax (f); half is avoidance. Kind: guesswork.
    • q, chance it survives court. The chance that the tax survives a court challenge. Below 100%, it discounts the revenue for the chance courts strike the tax down.
      50%JRJR: a discount; the wealth tax is multiplied by the chance it survives in court. The authors set it at the 'more likely than not' threshold of ASC 740-10. Kind: guesswork.100%100%BSZBSZ: no discount; the tax is assessed at face value, as written. The authors' expert report judges it drafted to stay clear of constitutional issues. Kind: guesswork.

  • Income tax lostIncome tax lost = f × C × A(H, r)A(H, r): what $1 a year for H years is worth today at rate r.

    f is the share of billionaires' income tax lost, C that tax per year, H the years it is counted, and r the rate that discounts it. moref, C and A(H, r) multiply, so one dial's effect depends on the others. f also moves with D and a, so its end labels differ between the panels.

    • f, share lost. The share of all billionaires' income tax lost to leavers. Both ends include the movers in a (at the JR end, this OPA's addition), and the BSZ end also the leavers in D.
      5%BSZThe BSZ rule: the share of wealth assumed to leave, except named leavers, who pay little income tax for their wealth. Kind: guesswork.5%45%JRThe JR rule: more than the 28.6% of wealth counted as gone, since income tax need not leave in step with wealth. Kind: guesswork.
    • C, yearly income tax. Billionaires' yearly California income tax revenue, the amount at risk each year.
      $3.06B/yrBSZBSZ: from California tax data, taking billionaires to have about half the income of the state's highest earners. A study of top wealth holders finds that ratio. Kind: research.$3.06B/yr$4.55B/yrJRJR: from California tax data, taking billionaires to be among the state's 212 to 500 highest earners. The value is the middle of $3.3 billion to $5.8 billion. Kind: guesswork.
    • H, years counted. How many years the lost income tax is counted. Middle stops: 10, 20 and 50 years.
      5 yearsBSZBSZ: about the payment period, treating the departures as temporary. $1 a year for 5 years is worth $4.6 at 3%. Kind: scenario.5 yearsfor everJRJR: leavers are assumed not to return once the Act permanently lifts the intangible property tax cap. Kind: scenario.
    • r, discount rate. The discount rate that turns a yearly loss into today's dollars. To see how much it matters inside the JR draw, use the JR explorer's Rate 1.5% and 4.5% settings.
      2.73%JR2.73%: worth the same as the JR draw (1.5% to 4.5%) of r − g, the interest rate minus the lost tax's growth ($1 a year for ever is worth $36.6). Kind: derived.3%3%BSZ3%: the rate this OPA applies to the BSZ analysis's 5-year loss, the middle of the JR range. BSZ give no rate. Kind: scenario.
  • One switch for three inputs that each move the estimate by less than $7 billion. Open the fold below to set them one by one.
    BSZ valuesBSZBSZ: non-citizens taxed, no real-estate deduction, the extra tax from asset sales counted.BSZ valuesJR valuesJRJR: non-citizens left out, real estate deducted, no extra tax from asset sales.
  • Show the three inputs one by one
    • Wealth of California billionaires who are not US citizens.
      $0BJRJR: not taxed, as too uncertain to collect (treaties, apportionment): the 28 non-citizens are left out of the base. Kind: guesswork.$150B$150BBSZBSZ: taxed. The non-citizen billionaires are in the base, like any other California resident. Kind: data.
    • The share of the base deducted for directly held real estate.
      0.0%BSZBSZ: no deduction. Directly held real estate stays in the base. Kind: guesswork.0.0%0.5%JRJR: a deduction. Directly held real estate, about 0.5% of the whole base, is taken out of it. Kind: data.
    • The share of the tax paid by selling assets, whose gains raise extra income tax.
      0%JRJR: not counted. No extra income tax from selling assets is included. Kind: guesswork.1/31/3BSZBSZ: counted. One third of the tax is paid by selling assets, and the gains on them are taxed. Kind: guesswork.
The BSZ panel's estimate is $106.8B and the JR panel's is −$39.0B.
Net fiscal effect, $ billions: wealth tax collected minus income tax lost, plus asset-sale tax (one of the three minor inputs). Move every slider in a panel to its far end and the two bars meet.

JR: −$39.0B

The JR analysis's main estimate1, the exact average of the authors' model (Jaros and Rauh, NBER, September 2026)

It averages the authors' 100,000 draws. 90% of them fall between −$107.9B and $15.9B, and all between −$208.8B and $47.7B.

With these sliders −$39.0B

  • Wealth tax collectedWealth tax collected = 5% × (W − D) × (1 − a) × q

    The tax takes 5% of the wealth. W is the wealth on the list, D the part counted as moved out in time, a the share lost to avoidance, q the chance the tax survives in court. moreThe four dials multiply, so one dial's effect depends on the others. W and D are valued on the same Forbes list.

    • W, wealth on the list. The billionaires' wealth the tax applies to, from one of two Forbes lists.
      Forbes 2025JRJR: the 2025 list, 212 billionaires, grown 7%, the Act's growth figure, to the valuation date (31 December 2026): $2,027B. Kind: guesswork on data.Forbes 2025Forbes 2026BSZBSZ: the Forbes list on 1 July 2026, 250 California billionaires worth $2,307B, or $2,157B without the non-US citizens (a separate dial). Kind: data.
    • D, wealth counted as moved out. It left before the 1 January 2026 residency date. Middle stop: a BSZ scenario counting Page, Thiel, Hankey and Kalanick as moved.
      None namedBSZBSZ: half of the 10% avoidance, 5% of the wealth, is people who moved in time. No one is named without evidence of a legal move in time. Kind: guesswork on data.7 moves7 movesJRJR: seven counted as moved in time, 28.6% of the wealth, on public reports (Page, Thiel, Hankey, Kalanick, Brin, Spielberg, Sacks). Kind: guesswork on data.
    • a, avoidance share. The share of the wealth that escapes the tax through avoidance and evasion, including moves no one names (D holds the named ones).
      0.0%JRJR: none beyond the departures it counts, since US securities are third-party reported. Its draws cover unreported moves. Kind: guesswork.0.0%10.0%BSZBSZ: a small allowance argued from enforcement. Half is people who moved out in time, who also cost income tax (f); half is avoidance. Kind: guesswork.
    • q, chance it survives court. The chance that the tax survives a court challenge. Below 100%, it discounts the revenue for the chance courts strike the tax down.
      50%JRJR: a discount; the wealth tax is multiplied by the chance it survives in court. The authors set it at the 'more likely than not' threshold of ASC 740-10. Kind: guesswork.50%100%BSZBSZ: no discount; the tax is assessed at face value, as written. The authors' expert report judges it drafted to stay clear of constitutional issues. Kind: guesswork.

  • Income tax lostIncome tax lost = f × C × A(H, r)A(H, r): what $1 a year for H years is worth today at rate r.

    f is the share of billionaires' income tax lost, C that tax per year, H the years it is counted, and r the rate that discounts it. moref, C and A(H, r) multiply, so one dial's effect depends on the others. f also moves with D and a, so its end labels differ between the panels.

    • f, share lost. The share of all billionaires' income tax lost to leavers. Both ends include the movers in a (at the JR end, this OPA's addition), and the BSZ end also the leavers in D.
      5%BSZThe BSZ rule: the share of wealth assumed to leave, except named leavers, who pay little income tax for their wealth. Kind: guesswork.45%45%JRThe JR rule: more than the 28.6% of wealth counted as gone, since income tax need not leave in step with wealth. Kind: guesswork.
    • C, yearly income tax. Billionaires' yearly California income tax revenue, the amount at risk each year.
      $3.06B/yrBSZBSZ: from California tax data, taking billionaires to have about half the income of the state's highest earners. A study of top wealth holders finds that ratio. Kind: research.$4.55B/yr$4.55B/yrJRJR: from California tax data, taking billionaires to be among the state's 212 to 500 highest earners. The value is the middle of $3.3 billion to $5.8 billion. Kind: guesswork.
    • H, years counted. How many years the lost income tax is counted. Middle stops: 10, 20 and 50 years.
      5 yearsBSZBSZ: about the payment period, treating the departures as temporary. $1 a year for 5 years is worth $4.6 at 3%. Kind: scenario.for everfor everJRJR: leavers are assumed not to return once the Act permanently lifts the intangible property tax cap. Kind: scenario.
    • r, discount rate. The discount rate that turns a yearly loss into today's dollars. To see how much it matters inside the JR draw, use the JR explorer's Rate 1.5% and 4.5% settings.
      2.73%JR2.73%: worth the same as the JR draw (1.5% to 4.5%) of r − g, the interest rate minus the lost tax's growth ($1 a year for ever is worth $36.6). Kind: derived.2.73%3%BSZ3%: the rate this OPA applies to the BSZ analysis's 5-year loss, the middle of the JR range. BSZ give no rate. Kind: scenario.
  • One switch for three inputs that each move the estimate by less than $7 billion. Open the fold below to set them one by one.
    BSZ valuesBSZBSZ: non-citizens taxed, no real-estate deduction, the extra tax from asset sales counted.JR valuesJR valuesJRJR: non-citizens left out, real estate deducted, no extra tax from asset sales.
  • Show the three inputs one by one
    • Wealth of California billionaires who are not US citizens.
      $0BJRJR: not taxed, as too uncertain to collect (treaties, apportionment): the 28 non-citizens are left out of the base. Kind: guesswork.$0B$150BBSZBSZ: taxed. The non-citizen billionaires are in the base, like any other California resident. Kind: data.
    • The share of the base deducted for directly held real estate.
      0.0%BSZBSZ: no deduction. Directly held real estate stays in the base. Kind: guesswork.0.5%0.5%JRJR: a deduction. Directly held real estate, about 0.5% of the whole base, is taken out of it. Kind: data.
    • The share of the tax paid by selling assets, whose gains raise extra income tax.
      0%JRJR: not counted. No extra income tax from selling assets is included. Kind: guesswork.0%1/3BSZBSZ: counted. One third of the tax is paid by selling assets, and the gains on them are taxed. Kind: guesswork.

What separates the two numbers, step by step

Apply to the BSZ panel, from left to right.2

BSZ estimate −Billionaires who moved out in time
(D) −Chance the tax does not survive litigation
(q) −Lost income tax
(H, f, C, r) +Other inputs together
(W, a, the minor switch) =JR estimate

The BSZ panel with these dials: $106.8B

But what are the actual quantities?

Each dial's two ends above give both analyses' values. The overview's table of differences sets them side by side with their sources.

Most of these are empirical quantities, and none should depend on whether one supports or opposes a wealth tax. H and r are scenario choices. An expert should be able to judge which of the two values is more credible.

Take the survey: which values are more credible?

The survey is primarily for economists and policy analysts. Share this survey with people you think have expertise in this area, to learn which values they find more credible.

Notes on the two estimates and the conventions

1 The BSZ number is this OPA's net total over 5 years at 3% (BSZ's own headline is the wealth tax alone), and the JR number is the authors' mean net present value in perpetuity. The paper prints −$38.9B, the average of one seeded run of 100,000 draws. The exact average of the same formula is −$38.98B, used throughout this page (more in the report: The two main estimates).

2 The equation adds up only when the terms are switched one after another, in this order, from left to right. In another order the same dials split the gap differently among the terms: moved first, q would take −$53.75B instead of −$39.44B, because it would also halve the tax on the billionaires who already left. The Apply buttons work from left to right, each unlocking the next, so each press changes the estimate by its amount in the equation (more in the report: The explorer's equation).

The net fiscal effect puts both on one definition, and the panels start at its two ends. The BSZ end is $106.8B and the JR end −$39.0B, a gap of $145.8B. The discount rate is a dial: 3% at the BSZ end, the rate the BSZ overview uses, and 2.73% at the JR end, the one rate worth the same as the authors' random draw of it (more in the report: The discount rate).

The litigation dial, q, is at 50% at the JR end. The JR explorer's litigation dial reads 1 at the main estimate, because there the revenue draw from 0 to the ceiling already halves the mean revenue. This explorer applies the same half as q = 50% of the ceiling (more in the report: The litigation dial).

PolicyEngine's calculator does the same walk in its own model, with each paper's assumptions as a setting. It was built independently of this project and lands at +$106.2B and −$37.2B. Its code is public (more in the report: Two other comparisons).

Conventions. All amounts are billions of dollars. Present values are as of the date the wealth tax is assessed, and the first year of lost income tax is counted one year later (more in the report: Conventions).

Where the numbers come from: code, data and kinds

Where the numbers come from. Every number on this page is computed by comparison/R/ from the two analyses' export contracts (bsz-analysis/export/r/, rjkdc-analysis/export/r/: the files each pipeline writes for others to read) and the transcriptions in comparison/data/. The estimating function is score_common() in comparison/R/model.R, with a Python twin (R and Python agree to 1e-9 on every exported number; see comparison/tests). The browser runs model.js, a copy of it, and a test runs that copy with node over 600 slider mixes against the R values, to 1e-9. The report derives each number step by step and shows the bridge from the BSZ end to the JR end; the materials page lists the code, tests and files.

Data. ends.csv (the two ends, at each horizon setting), slider_sample.csv (300 slider mixes per panel with their values, the sample the tests run model.js on), inputs.csv (every disputed input and its sources), DISPUTES.md (the crosswalk of the disagreements, one row each, with what each analysis says and where), dial_path.csv (the terms of the equation at the top, dial by dial), comparison-data.js (what this page reads). A test fails if comparison-data.js or this page differs from the generator's output.

Kinds. Each end's tip ends with the kind of that analysis's value: data (settled by observable facts), research (a published estimate), guesswork (an assumption with no data or study behind it), guesswork on data (a judgement applied to observed facts), scenario (a choice of which world is estimated) and derived (computed from the analysis's own inputs). A kind says what evidence is behind a value, not whether the quantity itself can be observed.

The function. The report states the estimating function line by line. The BSZ panel's sliders move only the BSZ bar and the JR panel's sliders only the JR bar.

Survey links. This page records, without any identifying data, that a survey link was opened and which form it drew.