The S5 instrument, page by page, with response data.
Pre-registered national online study. S5 (locked, N=392). Completed N = 479; analysis cohort N = 392 after the pre-registered exclusion code and a vote-eligibility proxy for citizenship (S5 did not ask citizenship directly; see Method notes). OSF pre-registration osf.io/br3u2. Every number on this page recomputes from the frozen snapshot at public/data/rthg-s5.json.
Headline result
Ballot support, overall and across political subgroups.
state ballot (rated 5+).
A further 4.6% were neutral and 5.1% opposed, putting support-or-neutral at 94.9%. Wilson 95% CI on the yes-rate: [87.0%, 92.9%].
Pre-registered on OSF (osf.io/br3u2). Locked dataset, N = 392 strict-clean and American. Every figure recomputes from the frozen snapshot.
Yes-rate by party and 2024 vote
95% CI [86.8%, 95.3%]
95% CI [82.8%, 94.9%]
95% CI [81.6%, 93.3%]
95% CI [79.4%, 91.9%]
95% CI [88.5%, 95.8%]
95% CI [72.5%, 92.0%]
Yes = rated 5 or higher on a 1 to 7 support scale. Bars show the yes-rate; neutral (rated 4) and opposed (rated 3 or lower) are not shown per subgroup. Brackets are Wilson 95% CI. Subgroups with fewer than 30 respondents are not shown. These are descriptive subgroup rates, not a pre-registered confirmatory test; see the post-stratification check in Appendix A for the party-weighted whole-sample estimate.
Step 1 of 6
Section B principles, before any mechanism is described.
Respondents see Section B first, on a page by themselves, before the ballot text or any mechanism. They rate each item on a 1 to 7 agreement scale (Strongly disagree to Strongly agree), with a no-opinion option. Order is stratified-randomised per respondent. The instrument page tells respondents to focus only on whether they agree with the principle itself, setting aside who supports it, how it could be used, and whether a specific law would work.
Headline result
Wilson 95% CI [98.2%, 99.9%]. Computed on the 392-respondent analysis cohort. Mean is over the 13 substantive items in Section B (11 governance principles plus the cross-partisan-vocabulary item B-bal and the reverse-coded honesty check B-rev). One fixed attention check (B-att) is interleaved but not counted in the principle-mean.
Per-item breakdown (analysis cohort, N = 392)
Each row shows the share of respondents who rated the item at 5 or higher on the 1 to 7 scale, with Wilson 95% CI. B-rev is reverse-coded: its agree-rate is the share who disagreed with the negatively-worded statement (the rating is 8 minus the raw response, on the same scale as the other items). Items the respondent answered “no opinion” on are excluded from the denominator for that item only.
- 1.
B-balCross-partisan vocabulary
Government officials of both parties should explain their decisions honestly to the people they serve.
99.7% Agreed (rated 5+) (390/391)Wilson 95% CI [98.6%, 100.0%]Mean 6.79 / 7No opinion: 1 - 2.
P1Loyalty
Government officials serve the public, not private interests, and have a duty to act in the public’s best interest.
98% Agreed (rated 5+) (383/391)Wilson 95% CI [96.0%, 99.0%]Mean 6.83 / 7No opinion: 1 - 3.
P2Reasoning
When the government makes a major decision, officials should have to honestly explain their reasoning, including the evidence they used, the alternatives they considered, and the strongest objections to their choice.
98.7% Agreed (rated 5+) (387/392)Wilson 95% CI [97.0%, 99.5%]Mean 6.61 / 7 - 4.
P3Reasoning
Major government decisions that affect the public should be based on evidence of how the decision achieves the public’s goals, not ideology, loyalty to a party, or the preferences of donors.
98% Agreed (rated 5+) (384/392)Wilson 95% CI [96.0%, 99.0%]Mean 6.63 / 7 - 5.
P4Reasoning
When the government makes a significant decision, the public has a right to see the reasoning behind it.
99% Agreed (rated 5+) (388/392)Wilson 95% CI [97.4%, 99.6%]Mean 6.69 / 7 - 6.
P5Disclosure
Officials should be required to disclose conflicts of interest before voting on, or otherwise deciding, matters that could personally benefit them or their donors.
99.5% Agreed (rated 5+) (389/391)Wilson 95% CI [98.2%, 99.9%]Mean 6.73 / 7No opinion: 1 - 7.
P6Accountability
When government officials commit fraud to benefit themselves or their donors at the public’s expense, there should be real consequences, including removal from office for egregious cases.
100% Agreed (rated 5+) (392/392)Wilson 95% CI [99.0%, 100.0%]Mean 6.91 / 7 - 8.
P7Accountability
If a government decision is proven to harm the public for private gain, the public should have the ability to require the government use clear reasoning to make a better choice for the people they serve.
98% Agreed (rated 5+) (383/391)Wilson 95% CI [96.0%, 99.0%]Mean 6.64 / 7No opinion: 1 - 9.
P8Accountability
When officials purposefully harm the public for private gain, if proven, accountability should be possible without waiting for the next election.
97.7% Agreed (rated 5+) (383/392)Wilson 95% CI [95.7%, 98.8%]Mean 6.72 / 7 - 10.
P9Equal application
Ethics rules for government officials should be enforceable, not just aspirational, and should apply equally to officials of every party.
99.2% Agreed (rated 5+) (388/391)Wilson 95% CI [97.8%, 99.7%]Mean 6.79 / 7No opinion: 1 - 11.
P10Citizen challenge
Americans should have a real way to inspect and challenge the government’s stated reasoning for major decisions, and the government should have to respond directly, with reasons.
97.2% Agreed (rated 5+) (378/389)Wilson 95% CI [95.0%, 98.4%]Mean 6.49 / 7No opinion: 3 - 12.
P6-narrowNarrow trigger
When a government official puts false statements into the official public record of a major decision, and the evidence shows honest intent is unlikely, consequences should be proportionate to the harm caused and the evidence of intent.
96.4% Agreed (rated 5+) (370/384)Wilson 95% CI [94.0%, 97.8%]Mean 6.50 / 7No opinion: 8 - 13.
B-revReverse-coded honesty check (reverse-coded)
Government officials should be allowed to knowingly put false reasons into the official public record of a major government decision.
100% Disagreed (rated 5+ after reverse-scoring) (392/392)Wilson 95% CI [99.0%, 100.0%]Mean 6.88 / 7
Sponsor mask
During recruitment the survey was administered under a neutral sponsor label, “American Institutions Study 2026.” The Right to Honest Governance Institute was not named to respondents until the debrief page. The mask keeps the support measurement clean of acceptance-bias on a sponsor-favourable proposal.
Step 2 of 6
The ballot text, exactly as respondents read it.
Ballot measure
The right
The people of this state have the right to honest government. Public officials serve as fiduciaries of the people, owing loyalty, care, and good faith.
The requirement
Every major government decision must be made with honest reasoning on the public record: the key facts relied on, the alternatives considered, and a response to the strongest objections.
When there is doubt
When there is credible evidence of dishonest reasoning on the public record, state citizens may petition for review. Courts may promptly dismiss frivolous or duplicative petitions and combine petitions about the same decision.
Who reviews
Who reviews each petition — and in what combination of courts, randomly selected citizens similar to a grand jury, and independent professional advisors — is the subject of the next page of this survey. There are several reasonable ways to design it. For now, we will call the reviewing body the Public Reasoning Panel. Different review designs appear on the next page. On this page, please rate the rule itself before seeing those design options.
How the review works
The Public Reasoning Panel asks whether the reasoning was honest, whether it accounted for the key facts on the public record, and whether it was consistent with the official's duty to the public rather than to personal or private benefit. If the panel finds the reasoning failed any of these tests, the decision is sent back to be remade with honest reasoning on a corrected public record.
How findings are made
Required procedures ensure that the panel's findings are grounded in evidence and reasoned in writing, so its decisions cannot be arbitrary.
Courts and consequences
Courts keep final legal authority. The panel cannot write laws, cancel laws, or remove officials on its own. In the most serious cases, such as knowingly placing false information on the public record, destroying or concealing evidence, or willfully obstructing review, personal consequences may be imposed only after court findings under heightened proof standards. For those acts, this amendment sets constitutional minimum consequences of removal from current office and disqualification from future office. Courts may impose additional consequences, including criminal referral. Officials acting in documented good faith are protected from personal penalty.
The goal
To sufficiently discourage dishonest service of the people and refocus government on: what the people's goals are, how best to achieve them, how to represent them well, and how to bring them together on issues where they genuinely disagree.
Lasting protection
No statute or official action may materially narrow, delay, or burden the rights and remedies created by this measure. Reasonable rules that speed review, combine duplicate cases, and protect fair process are allowed, but rules that substantially weaken access or remedies are not. Any conflicting law or action is void to the extent a court finds a conflict.
Implementation
The legislature has two years to set up the system, consistent with the principles of honest governance in this measure, and subject to review. If it does not pass a qualifying law within two years, the default implementation contained within the amendment takes effect automatically until the legislature enacts compliant laws.
The ballot vote
Prompt: “If this were on your state ballot today, how likely would you be to vote for it?” Rated on a 1 to 7 Likert scale.
90.3%
Yes (rated 5+)
4.6%
Neutral (rated 4)
5.1%
Opposed (rated 3 or lower)
Wilson 95% CI on the yes-rate: [87.0%, 92.9%]. The ballot vote was followed by a four-option describe-your-answer item: 40.8% yes unconditional, 51.3% yes conditional on mechanism, 3.3% maybe, 4.6% no.
Step 3 of 6
Four mechanism designs. Respondents see all four.
After the ballot vote, respondents see four mechanism designs with a shared callout that names what is identical across all four: the three honest-reasoning tests, the Source Reliability Framework, advanced analytical tools under human fact-check, and courts keeping final legal authority. What changes across the four designs is who reviews first, how the case gets there, and how directly citizens participate.
Respondents do two distinct things. First, they rate each design independently on a 1 to 7 likely-to-work scale. Then they forced-choose the one they would trust most on a difficult case, with an explicit “none of these” option. The two measure different things: the rate-all-four block shows whether each design is broadly acceptable; the forced-choice shows which one wins when respondents must pick.
Rate-all-four (1 to 7 likely-to-work scale)
Citizen routing panel chooses the review path
4.78 / 7Citizen panel with source-reliability audit
4.77 / 7Public audit with court referral
4.78 / 7Courts only
4.02 / 7Forced-choice: “If these designs appeared on a real ballot today, which would you vote for?”
Citizen routing panel chooses the review path
A small grand-jury-style panel reviews each petition first; its only job is to decide which review path fits this case (court, full citizen panel, or expert-prepared review with citizen oversight). The downstream path then applies the three honest-reasoning tests.
Forced-choice share. Wilson 95% CI [29.9%, 39.3%].
Citizen panel with source-reliability audit
Random citizens drawn from jury lists audit the official’s published reasons under the Source Reliability Framework, supported by an independent legal officer (replaceable for cause). Courts review afterward and keep final legal authority.
Forced-choice share. Wilson 95% CI [22.6%, 31.4%].
Public audit with court referral
State citizens may file a formal audit request alleging the public record is incomplete or dishonest. An independent auditor verifies the request under the same Source Reliability Framework. A verified audit triggers court review.
Forced-choice share. Wilson 95% CI [20.7%, 29.2%].
Courts only
A petition that a major decision was made with dishonest reasoning goes directly to a court. The judge applies the three honest-reasoning tests, with advanced analytical tools assisting analysis under human fact-check. No citizen panel.
Forced-choice share. Wilson 95% CI [8.0%, 14.2%].
None of these four; I would prefer something else
3.3%13 respondents picked this on the forced-choice. Wilson 95% CI [1.9%, 5.6%].
After the mechanism page: did the protections answer your concerns?
Respondents are asked directly whether the named protections (against capture, gaming, weaponization, expertise gaps, and delay) answered their concerns.
Yes, the protections answer my concerns
Partially
I had no concerns
No, the protections do not answer my concerns
3.6% answered “no” (Wilson 95% CI [2.1%, 5.9%]). The pre-registered bound for this item was ≤10%.
The Source Reliability Framework
A standard for vetting the evidence a review relies on. What it checks: track record, independence, methodology, transparency, expert consensus. What cannot support findings: sources with proven dishonesty or hidden conflicts. What it covers: the sources findings rely on, and the sources officials cite in the decision under review. Weak or unreliable material can be examined; findings cannot rest on it.
Step 4 of 6
Panel composition follow-up.
Once respondents picked a mechanism, the survey asked a separate page of questions about how the citizen panel itself should be set up. This page is parameter-tuning for the design that won, not a mechanism evaluation. Per the pre-registration, panel composition is exploratory rather than confirmatory.
How should panel members be selected?
- 46.2%Mostly random, with a minimum share of seats reserved for screened civic or professional members181 respondents. Wilson 95% CI [41.3%, 51.1%].
- 31.1%All seats filled randomly from ordinary jury rolls, with conflict-of-interest screening122 respondents. Wilson 95% CI [26.7%, 35.9%].
- 17.1%All seats filled by nomination from professional or civic associations, with conflict-of-interest screening67 respondents. Wilson 95% CI [13.7%, 21.1%].
- 5.6%Not sure22 respondents. Wilson 95% CI [3.7%, 8.4%].
Step 5 of 6
The re-vote. Within-respondent shift after seeing the mechanism.
After the mechanism and panel-composition pages, respondents are asked the same vote question again: “Now that you have considered how the review might work, how likely would you be to vote for this measure?”
Within-respondent mean shift
Movers split modestly upward. Mean is the small number above; modal outcome is no change.
Step 6 of 6
Open-text reflection.
After the re-vote, the instrument asks 2 required open-text prompts and 1 optional follow-up.
The prompts
- Required. “In one or two sentences, how would you explain this proposal to a friend in your own words?”
- Required. “What pulled you toward your vote, what gave you hesitation, and what change would move you the most? Even a rough idea is valuable.”
- Optional. “What stood out most as you were reading? Any word, sentence, or idea that drew your attention.”
The s5 instrument did not include data-publication language at consent time. For s5 we publish only de-identified, aggregate results and the analysis code, and we do not quote individual s5 written answers.
Appendix A
Pre-registered hypothesis tests.
The five pre-registered hypothesis tests (OSF prereg questions a through e), each with the pre-registered bound, the observed value on the headline cohort (N = 392), and the Wilson 95% CI. Rules and bounds are the OSF prereg verbatim; observed values are computed from the locked dataset.
Principle agreement
- Pre-registered rule
- ≥ 95% of clean respondents have a Section B principle-mean ≥ 5 of 7.
- Observed
- 99.5% had a principle-mean ≥ 5. Wilson 95% CI [98.2%, 99.9%]
Ballot support
- Pre-registered rule
- ≥ 80% support the ballot initiative (rated ≥ 5 of 7).
- Observed
- 90.3% yes (rated ≥ 5). Wilson 95% CI [87.0%, 92.9%]
Within-respondent aversion delta
- Pre-registered rule
- The abstract-to-specific shift is equivalent to zero within ± 0.5 Likert points (TOST).
- Observed
- +0.02 mean shift. Within-respondent split: 55.1% same, 25.3% up, 19.6% down.
Post-stratification check
Re-weighting the analysis cohort to the published US-adult party benchmark (Democrat 27% / Republican 27% / Independent 46%) yields a weighted yes-rate of 89.8%, 0.5 point off the unweighted 90.3%. Party blocs: Democrat 92% [86.8%, 95.3%]; Independent 88.7% [81.6%, 93.3%]; Republican 90.4% [82.8%, 94.9%].
Appendix B
Lock manifest and dataset hashes.
Every number on this page recomputes from the frozen snapshot. The hashes below tie the published files to the figures shown.
- Wave
- S5
- Snapshot generated
- 2026-06-13T20:52:03.854Z
- Lock cohort (completed)
- 479
- Analysis cohort (American)
- 392
- Sensitivity (no ‘Prefer not to say’)
- 383
- Dataset SHA256
ea73c3bf562f7e3affa42d2c31eddce241002d6d7b6233c05e83e73cdfd77a03- OSF preregistration
- osf.io/br3u2
Method notes
How the analysis cohort was cleaned.
Two layers of cleaning sit on top of the 479 completed responses in the lock cohort.
1. Pre-registered exclusion code. Attention-check failure, speed-click thresholds, straight-line patterns, and the rest of the rules pre-registered on OSF. The exclusion logic lives at lib/rthg-study-flags-s5.ts and lib/rthg-study-flags-s6.ts. Running the rules on the lock cohort yields N = 400.
2. Citizenship filter. The S5 instrument did not ask citizenship directly; the strongest available signal is the 2024-vote demographics question, where “I was not eligible to vote” is the only answer that is definitionally inconsistent with being a U.S. citizen-voter. Those 8 respondents are excluded. That leaves N = 392 strict-clean and American, the cohort behind every percentage on this page unless otherwise noted.
On “Prefer not to say”. A small number of respondents picked “Prefer not to say” on the 2024-vote question. Excluding them as a sensitivity check leaves N = 383; the headline yes-rate and principle-agreement rate are unchanged within rounding. See the unit tests at lib/study-tests for the exclusion rules expressed as runnable test cases.