California Proposition 40, Reconciliation of the Two Analyses

The two headlines answer different questions: the BSZ analysis reports what the tax collects, the JR analysis a net present value that subtracts income tax lost for ever. Both are put on two outputs. The explorer shows (b); (a) is in the report.

The horizon H is itself disputed: about the 5 payment years for the BSZ analysis, for ever for the JR analysis. So it is a dial in the explorer and a step of the bridge. Present values use the JR analysis's discount-rate range, 1.5% to 4.5%; the BSZ analysis takes no position on the rate.

Conventions. Nominal billions of dollars. The wealth tax is counted at face value when assessed and not discounted over the five installments, as both analyses do. The BSZ analysis values the base on 1 July 2026; the JR analysis grows its 2025 Forbes base by 7% to 31 December 2026. The JR code labels its rate draw r − g and its text r (the JR analysis's mismatch 3).

What this OPA does not estimate. Both analyses, and so this OPA, take the measure as written (docs/initiative/). Neither models the day-to-day movement of the Forbes base. Nothing is added to either analysis here.

BSZ analysis: a revenue estimate

Boll, Saez, Zucman, NBER WP 35218 (Aug 2026) →

Static estimation on the Forbes list with a 10% avoidance allowance.

Main estimate, net total: $106.8B Table 5 row 1, wealth tax: $103.8B (the paper prints $104B)

JR analysis: a net present value estimate

Jaros and Rauh, NBER c15504 (Sep 2026) →

Person-by-person base, observed departures, a litigation weight, lost income tax for ever.

Main estimate, mean NPV: −$38.9B

The BSZ paper prints no (b). It is Table 5 row 1's own columns combined here: $103.8B of revenue, plus $3.68B from asset sales, minus $0.70B, five years of its $0.15B yearly income tax loss. For the JR analysis both outputs are exact means of the authors' Monte Carlo, whose revenue is drawn between zero and $72B.

Jeffrey Hoopes, "Galle v Rauh: Comparing Two Revenue Estimates for California Billionaire Taxation" (SSRN 6428578), is the one published comparison. It compares earlier versions: the February 2026 GGSS expert report (Galle, Gamage, Saez and Shanske), on which the BSZ analysis builds, and the March 2026 version of the JR analysis (Rauh, Jaros, Kearney, Doran and Cosso). His Figure 2 prints no values, so his column is read off the chart (tick fit within $0.1B, about $0.6B per bar; comparison/tools/digitize_hoopes_fig2.py). Beside it is our bridge in his order, ending at the September 2026 version.

Disclosures from the note itself: Hoopes states his priors, "To lay my cards on the table, I believe that the world is complicated, and that when possible, empirical analysis should be done in as detailed a way as possible" (p.2), and thanks "David Gamage for encouraging me to read both of these analyses carefully and compare the two" (p.1); Gamage is a GGSS co-author.

The "Galle estimate" bar is clipped by the chart's frame at $100B; its value is taken from the note's text (p.3).

Every difference

• Vintage. Hoopes's estimate uses the February GGSS text: 213 billionaires, $2,182B of wealth, $109B gross, $99B after 10% avoidance, rounded up to $100B (p.3 fn.1). The July GGSS text covers 250 billionaires and $2,307B: $103.8B after avoidance, rounded down to $100B. The same printed headline hides a rounding of +$1B in one vintage and of −$3.8B in the other.
• Residency. His bar is −$12.6B, close to Rauh's own −$12.84B for dropping Ellison, Houston and Snyder and adding Sacks (Rauh p.5). By July the BSZ analysis has itself dropped Ellison, so our step is a different quantity: the $150B of non-US-citizen residents Rauh omits plus the list and valuation-date difference, −$12.6B in one order.
• Confirmed departures. His bar is −$17.0B; Rauh's own step is −$26.69B before avoidance (94.2 to 67.51, p.11), and ours is −$26.0B at that point of the order. We could not trace his figure to a number in Rauh et al.
• Additional movers. His −$12.5B matches Rauh's expanded ten-departure estimate (67.51 to 55.1, a step of −$12.41B, pp.12-13). The March 2026 version's own −$24.7B mean NPV does not use that estimate, so our bridge has no such step.
• Behaviour. His −$9.2B is close to the March 2026 version's step from the expanded estimate to the literature calibration (55.1 to 45.59, −$9.51B, p.14), a mobility response from a European elasticity. The September 2026 version predicts no such departures; in their place it weights the revenue by the chance the tax survives litigation. Ours, −$29.0B in one order, nets two labelled steps: dropping the 10.0% avoidance allowance (up) and the litigation weight (down).
• Real estate. His bar is drawn near zero (−$1.1B ± 0.8 read off the chart), while his text says the exclusion reduces the base by about $8B (p.4). The two are consistent: $8B of base is about $0.4B of revenue at 5%. Ours is −$0.23B.
• Level before income tax. His revenue level before the income-tax bar is about $47.2B, near the March 2026 version's preferred 'about $40B' (p.14). Ours ends at $36.00B, the expected revenue in the Monte Carlo behind the September 2026 version's mean NPV (its text prefers 'approximately $30B', p.1).
• Income tax. His single bar, −$71.5B, ends at −$24.3B, close to the March 2026 version's −$24.7B. Ours is −$74.98B, the September 2026 version's expected present value of lost income tax against $36.00B of revenue, and we split it into four labelled steps (the share lost, the income tax level, the horizon, asset-sale tax), which interact with the base steps.
• Order and labels. His is one sequential order, with no provenance labels; ours shows Shapley averages over all orders (a Shapley value: a step's effect averaged over every order the inputs could be switched in, so the steps add up to the gap) as well as one order, with each step labelled data, research, guesswork or scenario.
• Steps he does not have. The non-US-citizen residents, the BSZ analysis's avoidance allowance being dropped, the asset-sale income tax, and the litigation weight of the September 2026 version (which postdates his note).

Which dated version of each document the numbers on this site come from. It matters for one reason: the Hoopes comparison above uses earlier versions than this site, so the two do not line up step for step.

One row per input on which the two estimates differ: what each analysis assumes, where it comes from, and what kind of claim it is (data, research, guesswork or scenario). Each row is one step of the bridge and one dial of the explorer. The Row column follows the project's crosswalk, DISPUTES.md, which quotes both analyses on each point.

Row 3: % = 1 − 72 / 100.906, the authors' ceiling after seven departures over the grown base. Their person-level file gives 72.05 (mismatch 4); the script's 72 is used so the comparison returns their mean exactly.

"A vs B" reads the BSZ value's kind against the JR value's; "guesswork on data" is a judgement applied to observed facts.

The LAO (the Legislative Analyst's Office, the state's nonpartisan fiscal adviser) gives no point estimate: responses "probably would reduce state revenues by less than $1 billion per year" (p.2). Walczak (California Tax Foundation, 22 April 2026) puts the loss at $3.53-4.49B a year, spillovers included (p.1); he is credited as an alternative view, not merged into either analysis. All on one scale, with E[f] × E[C] the JR mean share lost times its mean yearly income tax:

Present values use the bridge's rate range. The last column sets each loss, in perpetuity, against the JR analysis's expected revenue, $36.0B, for scale.

• BSZ analysis: Boll, Saez and Zucman (August 2026) reproduced from the authors' workbook, with an explorer over Table 5's guesswork inputs, a report, slides and materials.
• JR analysis: Jaros and Rauh (NBER, September 2026) reproduced from the authors' public code.
• The comparison explorer, the disputes crosswalk and the reconciliation's code, tests and exports in comparison/, with a guide on the materials page.

The numbers on this page reproduce and are tested; what is still being completed is the author's own review of each part, recorded in the last column below, and the credit interview.

This work was a collaboration between the author and Claude Code (Opus and Sonnet models) across many sessions, not an AI working alone. The table records who led the work on each public object of the reconciliation, using the CRediT contributor roles. The answers will come from a questionnaire the author fills in, not from commit history. The table groups the 14 roles into three: ideas and direction (conceptualization, methodology, supervision, project administration), doing the work (data curation, formal analysis, investigation, software, validation, visualization) and writing (original draft, review and editing). Funding and resources do not apply. Each square in a cell is one role, coloured by who led it (key below the table). The last column is not part of CRediT. It records how closely the author has checked the object.

Not yet filled in as of October 2026. The record in comparison/site/credit/CREDIT.md (2026-10-01) carries the answers the author gave once, for the BSZ explorer, and asked to be applied to every object; the per-object interview for the comparison has not been run, so the table is left blank until it is.

Human verification uses three levels: not reviewed, lightly reviewed (checked, but the author cannot yet explain it clearly) and reviewed (the author can explain it to others). The automated test suites are mechanical checks, not an outside review. No other person has verified this work, and the original authors have not reviewed it.