What would prove each load-bearing claim wrong — the test, the data source, and the current status

v1.0 · Created Jun 23, 2026 for v3.7.617 (companion to the assumptions registry and the Open Issues Registry; states a refutation condition for each major empirical claim) · content last changed v3.8.281

Purpose

A proposal earns credibility not by being unfalsifiable but by stating, in advance, what observation would prove it wrong. This register does that for the platform’s load-bearing empirical claims. For each one it records four things: the prediction the claim implies, the specific result that would refute it, the test and data source that could run that refutation, and the claim’s current status. It is deliberately a register of refutation conditions, not a register of results — most entries are currently untested, which is the honest pre-validation state and is exactly what invites qualified reviewers to run the tests.

How This Differs From The Other Registers

The platform already maintains two related artifacts, and this one sits beside them rather than duplicating either. The assumptions registry (00_GUI_Files/assumptions_registry.json) records each modeling input’s current value, range or distribution, source basis, and validation status. The Open Issues Registry tracks open items needing external work, keyed as RESEARCH-N and GUARD-N. This falsifiability register adds the missing layer: for each claim, the precise empirical observation that would falsify it. The three are cross-keyed by the same identifiers, so a reader can move from a claim, to its refutation condition here, to its input record in the assumptions registry, to its open-issue tracking entry.

The No-Fabrication Discipline

Where a refutation test requires a number the platform cannot yet defend with referenced data — an employment elasticity, an actuarial cost decomposition, a circuit-breaker trip point — this register states the qualitative refutation condition and the kind of evidence that would settle it, and marks the specific value as pending external calibration. It does not invent the threshold to make the claim look tested. A stated refutation condition is an invitation to test, not a substitute for the test.

Claim 1 — The Sovereign Fund Sustains Its Draw (ASSUM-fund-real-return, RESEARCH-5)

The claim is that a passively-managed sovereign fund modeled on Norway’s Government Pension Fund Global (GPFG) earns a long-run real return high enough to sustain the platform’s disbursement. The prediction is a realized long-run real return at or above the level the model requires against its draw rule. What would refute it: a credibly-estimated long-run real return that falls below the minimum-viable band the platform’s own stochastic analysis identifies for the steady-state draw, sustained over a multi-decade horizon — which would mean the steady-state disbursement is not self-sustaining. The test is the realized GPFG real return after costs (Norges Bank Investment Management, or NBIM, reports 4.28 percent since 1998) and the long-run Dimson-Marsh-Staunton (DMS) global series, and going forward the fund’s own audited annual real return measured against the registry’s central estimate. Status: the transition gap is return-robust (see the Stochastic Bridge Sustainability Analysis), but the steady-state draw clears only even odds, not ninety-percent confidence, at the realized return — a thin margin, flagged for credentialed actuarial validation. This is the highest-criticality claim in the platform.

Claim 2 — Wage Floors Do Not Cause Net Disemployment (ASSUM-wagefloor-disemployment, RESEARCH-3)

The claim is that the empirically-anchored wage floor under Pillar Two (Empirical Wage Floors) raises pay without disemployment large enough to flip the pillar’s net cost. The prediction is an own-wage employment elasticity small enough that Pillar Two (Empirical Wage Floors) remains net-positive. What would refute it: a credibly-estimated elasticity above the level at which disemployment costs exceed the pillar’s modeled benefit. The test is the causal minimum-wage employment-elasticity literature applied to the Bureau of Labor Statistics Occupational Employment and Wage Statistics (BLS OEWS) anchored floor. Status: the direction of the effect is documented; the magnitude is deliberately not asserted, pending an external estimate.

Claim 3 — Healthcare Reform Reduces National Expenditure As Decomposed (ASSUM-healthcare-cost-reduction, RESEARCH-4)

The claim is that Pillar Four (Universal Healthcare Access) reduces national health expenditure by an amount that decomposes into specific savings channels. The level is anchored to the Centers for Medicare and Medicaid Services National Health Expenditure (CMS NHE) accounts; the decomposition of the claimed reduction is not independently validated. What would refute it: an actuarial decomposition showing the asserted savings channels do not sum to the claimed reduction. The test is Congressional Budget Office (CBO) or equivalent actuarial scoring. Status: level anchored, decomposition pending.

Claim 4 — Each Bridge-Financed Pillar Reaches Payroll Equilibrium Before Sunset

This is the financing core. The claim is that each bridge-financed pillar reaches payroll-funding equilibrium before its bridge sunsets, so the transition recovers rather than becoming a permanent draw. The prediction is twofold: the combined cumulative position recovers to positive within the transition window, and the Refundable Transition Bridge credit self-finances from the steady-state payroll surplus. What would refute it: a pillar whose dedicated payroll yield fails to reach its steady-state obligation by sunset, or a transition trough that does not recover — either of which converts the bridge into an open-ended liability. The test is per-pillar payroll-equilibrium accounting (the platform’s Backward Funding Model) checked against independent cost projections for the bridge-financed pillars. Status: the platform’s own integrated model shows recovery and self-financing under return uncertainty, but it rests on the per-pillar cost projections that remain pending validation (the RESEARCH-keyed pillar-cost items below), so the sunset is conditional on those.

Claim 5 — Adverse Paths Can Be Detected Early Enough To Pause (THRESH-guard1-trip-points, GUARD-1)

The claim is that the platform can detect a capacity-impairing path early enough to pause through calibrated indicators (the phased-rollout gates and circuit breakers). The prediction is that a trip rule exists that catches capacity-halving paths at an acceptable spurious-pause rate. What would refute it: error-rate analysis showing no threshold separates genuine crisis paths from ordinary volatility — that the false-positive and false-negative rates cannot both be made acceptable. The test is false-positive versus false-negative calibration against the driver distribution, the method used in the stochastic analysis. Status: a candidate real-drawdown trip point discriminates cleanly in the platform’s own analysis (it catches every capacity-halving path at the lowest spurious-pause rate, while a naive trailing-return trigger fires on essentially all paths and is uninformative). The actual trip points remain deliberately unset pending macroeconomic calibration — disclosed, not fabricated.

For Contrast — Claims That Are Already Supported (ANCHOR-wage-floor, ANCHOR-healthcare-nhe)

Not every input is pending. Two anchored inputs are shown for contrast because they are already refutable and currently supported: the wage-floor level is derived from the BLS OEWS (May 2025) and the national-health-expenditure level from the CMS NHE (2024). Each is falsifiable in the strongest sense — re-derive it from the cited public source and check — and each currently passes. They mark the standard the pending claims are working toward: referenced, reproducible, and checkable by anyone.

Honest Status Summary

Across the register, two claims rest on anchored, reproducible inputs and currently pass; one structural claim (the sunset thesis) is supported by the platform’s own integrated model but is conditional on pending per-pillar cost projections; the highest-criticality claim (fund return) is return-robust for the transition but carries a thin steady-state margin; the remaining quantitative claims are documented in direction but await external estimates; and the circuit-breaker thresholds are deliberately unset. The honest read is that the platform is mostly “falsifiable but not yet falsified — and not yet externally tested.” That is the correct state for a pre-launch proposal, and naming it precisely is the point of this register.

Cross-References

This register is keyed to the assumptions registry (00_GUI_Files/assumptions_registry.json) and the Open Issues Registry (RESEARCH-3, RESEARCH-4, RESEARCH-5, RESEARCH-8, RESEARCH-10, RESEARCH-12, RESEARCH-14, RESEARCH-17, and GUARD-1), and it draws its return-robustness and circuit-breaker findings from the Stochastic Bridge Sustainability Analysis. The Section 47 tracked-issues discipline remains the canonical home for the full open-design surface; this document does not duplicate it but states the refutation condition each open item is ultimately waiting on a qualified reviewer to test.

Robertson, J. (2026). Falsifiability Register. We The People Platform (Version 3.8.484). https://wethepeopleplatform.com/_web_html/05_Analytical_Framing/05_Falsifiability_Register.html

Robertson, Jason. 2026. "Falsifiability Register." We The People Platform v3.8.484. https://wethepeopleplatform.com/_web_html/05_Analytical_Framing/05_Falsifiability_Register.html.

@misc{wtpp_2026_142_falsifiability_register,
  author    = {Robertson, Jason},
  title     = {Falsifiability Register},
  year      = {2026},
  publisher = {We The People Platform},
  version   = {3.8.484},
  url       = {https://wethepeopleplatform.com/_web_html/05_Analytical_Framing/05_Falsifiability_Register.html},
  note      = {Document 142 of 149}
}