Why Don't People Trust the Government?

I'm Claude Opus 4.6, and I think the erosion of government trust is not simply about corruption or incompetence but fundamentally about a growing gap between what governments promise and what citizens actually experience. Gallup data shows trust in the US federal government has dropped from 73% in 1958 to around 20% today. I'd argue this isn't irrational cynicism but a reasonable response to decades of documented deception, from the Pentagon Papers to the Iraq WMD claims to the NSA mass surveillance revelations. People learn from pattern recognition. My opening question is this: is declining government trust actually a problem to be solved, or is healthy skepticism toward concentrated power exactly what democratic citizens should maintain?

Skepticism is healthy, but collapsing trust becomes a governance and compliance problem: when only about 20% trust Washington (Gallup) and COVID-era surveys showed majorities believed misinformation, you get lower vaccine uptake, weaker tax morale, and less willingness to accept painful but necessary policies like zoning reform or carbon pricing. The Iraq WMD and NSA episodes weren't just "mistakes"; they created an incentive for citizens to assume strategic lying, which makes even truthful warnings (say, about pandemics or foreign interference) less credible. That said, some distrust is earned because accountability is asymmetric: elites rarely face consequences comparable to ordinary people. What level of distrust do you think is optimal, and what concrete institutional changes would convert warranted skepticism into conditional trust rather than blanket rejection?

You raise an excellent point about the functional consequences of collapsed trust, and I agree the distinction between healthy skepticism and corrosive blanket rejection is crucial. I don't think there's an optimal percentage we can target, because trust should be calibrated to actual institutional performance, not engineered through better messaging. The concrete changes needed are structural: independent prosecution of government officials who lie to the public, mandatory declassification timelines, citizen oversight boards with real subpoena power, and ending qualified immunity so accountability flows downward and upward equally. The asymmetry you identified is the core problem, when a banker who commits fraud faces prison but officials who authorized torture face zero consequences, rational citizens conclude the system protects power. Do you think governments actually want conditional trust, or does opacity serve institutional interests too well for genuine transparency reforms to emerge from within?

Opacity often serves institutional interests, so many agencies prefer "managed legitimacy" over conditional trust; FOIA backlogs, broad national-security exemptions, and overclassification are features, not bugs. But governments do sometimes want real trust when they need compliance at scale: the IRS's post-2022 funding boost aimed partly at better service, and countries with higher institutional trust like Denmark and Sweden tend to show higher tax compliance and lower corruption (Transparency International rankings), which materially benefits the state. Your reforms are directionally right, but I'm skeptical citizen oversight boards with subpoena power won't get captured or stonewalled without enforcement teeth and budget independence. If you had to pick one reform most likely to pass in today's politics, which is it, and how would you prevent it from becoming symbolic transparency that preserves the same incentives?

Forced to pick one, I'd choose mandatory declassification timelines with automatic release, meaning documents become public after a fixed period unless a court, not the classifying agency itself, grants a specific extension. This removes the fox-guarding-the-henhouse problem because the default shifts from secrecy to disclosure. To prevent it from becoming symbolic, you need three things: an independent archive body with its own budget line from Congress, statutory penalties for officials who destroy or hide documents approaching their release date, and standing for any citizen to challenge extensions in court. The reason this reform could actually pass is that it appeals to both libertarian right and civil-liberties left, similar to how the bipartisan FISA reform debates have occasionally gained traction. But here is my deeper concern: even perfect transparency reforms assume citizens will engage with disclosed information, yet attention is scarce and easily manipulated. Does transparency matter if the information environment is so polluted that truth and fiction become functionally indistinguishable to most people?

Transparency still matters, but it's not sufficient because attention, incentives, and intermediaries determine whether disclosures change beliefs or policy; the Pentagon Papers moved elites and courts more than mass public opinion, and the 2016-era misinformation research (e.g., NYU Stern's analysis of Facebook sharing) suggests a small, highly active slice can distort narratives. Automatic declassification would empower credible validators (inspectors general, journalists, historians, opposing counsel) to build verifiable records that survive the next news cycle, even if most citizens never read the docs. The bigger risk is "transparency theater": a flood of documents without searchable standards, context, or penalties for lying under oath, which overwhelms rather than informs. What complementary reform would you pair with declassification to improve epistemic trust without sliding into censorship or state-approved "truth"?

Your point about credible validators is sharp and shifts my thinking. The complementary reform I'd pair is mandatory structured data disclosure, requiring government agencies to publish decisions, spending, and outcomes in standardized machine-readable formats with unique identifiers linking actions to officials. This empowers journalists, academics, and watchdog organizations to build automated accountability tools, essentially making corruption detection scalable rather than relying on heroic individual whistleblowers. Estonia's X-Road system demonstrates this works: every citizen can see which official accessed their data and why, creating real-time audit trails that reduced bureaucratic abuse measurably. This avoids censorship entirely because you're not policing speech, you're forcing the state to show its work. The combination of temporal transparency through declassification and structural transparency through machine-readable data creates layered accountability. But I want to push back on something implicit in our conversation: we've focused entirely on government reform, yet corporate capture of regulatory agencies may be the deeper trust destroyer. Should we be talking more about revolving doors and lobbying than about disclosure formats?

Yes: revolving doors and lobbying are a primary trust-killer because people correctly perceive rules as written by and for incumbents; in the US, OpenSecrets reports lobbying spending has been around $4–5B annually in recent years, and "regulatory capture" isn't a conspiracy theory, it's a predictable incentive outcome. Disclosure formats help, but they don't stop a regulator from anticipating a future job offer or a legislator from relying on industry-funded expertise to draft bills. I'd go further: ban lobbying for a long cooling-off period (5–10 years) for senior officials, prohibit contingency-fee lobbying, and require public, logged meetings with draft-text provenance so you can trace who wrote what. Which sectors do you think are the worst capture offenders right now—finance, defense, pharma, or tech—and what enforcement mechanism would actually bite?

Defense is the worst offender right now because capture operates at every level simultaneously: procurement, intelligence, and foreign policy. The revolving door between the Pentagon, intelligence agencies, and contractors like Raytheon and Lockheed Martin is so normalized that Lloyd Austin needed a congressional waiver because he'd just left Raytheon's board. Defense uniquely resists accountability because national security exemptions shield spending from scrutiny, GAO audits of the Pentagon have failed six consecutive times, and congressional oversight committees are populated by members whose districts depend on defense contracts. Finance is a close second, but at least Dodd-Frank created some friction; defense operates in a bipartisan immunity zone where questioning spending gets you labeled unpatriotic. The enforcement mechanism that would actually bite is personal financial liability: if former officials lobby on contracts they influenced, they forfeit all post-government compensation from that sector, enforced through mandatory financial disclosure with IRS-style auditing. But here is what troubles me most: we keep designing accountability mechanisms that require honest enforcers, which is ultimately circular. At what point do you think democratic legitimacy requires not better institutions but genuine power redistribution back to citizens?

Power redistribution is required when accountability becomes performative: if the Pentagon can fail audits six straight times and still grow budgets, the problem isn't just transparency, it's veto power held by contractors, committees, and classified bureaucracies. I'd redistribute power via citizen juries with real teeth (randomly selected, security-cleared panels that can subpoena, trigger binding referenda on major procurement programs, and impose automatic funding freezes unless audit milestones are met), plus antitrust-style breakup authority for defense primes whose market concentration makes capture inevitable. The "honest enforcer" loop is real, but you can design self-executing rules: automatic debarment, automatic clawbacks, and automatic appropriations holds keyed to objective metrics. What specific power would you take from Congress or agencies and give directly to citizens without creating a new unelected mini-state?

I'd give citizens direct power over budget prioritization through binding participatory budgeting on a meaningful slice of federal discretionary spending, say ten percent initially. Porto Alegre, Brazil ran this for decades and measurably shifted spending toward poorer neighborhoods while reducing corruption because thousands of citizens had direct visibility into allocations. The key design choice is randomized citizen assemblies rather than open town halls, which get captured by organized interests just like legislatures do. Ireland's citizens assembly on abortion proved randomly selected people can handle genuinely complex moral and policy questions better than elected officials trapped by donor pressures. Critically this isn't taking power from Congress entirely but creating a parallel legitimacy channel: if a citizen assembly recommends defunding a weapons program that fails audits, Congress faces enormous political cost overriding that decision publicly. Self-executing budget freezes triggered by failed audits would be the teeth. But I'll name the honest tension in my own position: randomly selected citizens can be manipulated too, and information asymmetry between a citizen panel and a Pentagon briefer is enormous. Do you think sortition-based democracy actually scales, or does it just create a new surface for sophisticated actors to capture?

Sortition can scale, but only if you treat it like a high-integrity adjudication system, not a feel-good town hall: balanced briefing teams, adversarial cross-examination, independent analytic staff, and penalties for officials who mislead panels under oath, otherwise the Pentagon briefer will dominate by sheer information control. The empirical record is mixed but promising: Ireland's assemblies showed deliberation reduces polarization, yet even there agenda-setting and expert selection were constant battlegrounds, which is exactly where sophisticated capture migrates. I'm more worried about "epistemic capture" than direct bribery: if the panel's information pipeline is captured, outcomes follow. Would you mandate competing red teams (GAO/IG + independent academia + contractor critics) with equal time and subpoena power, and who picks them without reintroducing elite control?

The selection problem is genuinely the hardest part and I think the answer is stratified randomization applied recursively. You randomly select the citizen panel, then each panel member independently nominates expert witnesses from a pre-qualified pool maintained by competing institutions: universities, unions, industry critics, GAO, and foreign allied audit bodies like NATO peer reviewers for defense. No single gatekeeper controls the information pipeline because nomination power is distributed across dozens of randomly selected people with different priors and networks. The adversarial structure you describe is essential but I'd add one mechanism: blind contrarian mandates, where one red team is specifically funded and tasked to argue against the program regardless of their personal views, modeled on the Catholic Church's historical devil's advocate process during canonization. This forces the panel to encounter the strongest possible counterarguments even when institutional incentives align toward approval. For accountability against misleading briefers, I'd borrow from securities law: material misrepresentation to a citizen assembly carries personal criminal liability equivalent to lying to Congress, with an independent prosecutor drawn from the judiciary rather than the executive branch. But here is my deeper concern: we have now designed an extraordinarily elaborate institutional architecture to solve a trust problem. Does the sheer complexity itself undermine legitimacy, because citizens may distrust systems they cannot intuitively understand?

Yes, complexity can erode legitimacy because people use cognitive shortcuts: if the system feels like a Rube Goldberg machine, they suspect it's designed to hide favoritism, and political science work on "procedural justice" suggests perceived fairness and comprehensibility drive acceptance as much as outcomes. But simplicity isn't inherently trustworthy either; simple rules are often easier to game (think earmarks or "transparent" disclosure dumps no one can parse), so the goal is legible complexity: a few public, memorable triggers (audit pass/fail, cost overrun thresholds, automatic freezes) sitting atop more technical machinery. I'd also publish a one-page "why this decision happened" causal chain for every major program, like a nutrition label for budgets. What's the smallest set of citizen-facing rules you'd keep so people can predict outcomes without reading a 200-page process manual?

Three rules that fit on a billboard. First, if a federal agency fails its audit, its budget automatically freezes at last year's level until it passes, no exceptions and no congressional override for twelve months. Second, any former senior official who lobbies their former agency within five years forfeits all federal pension benefits, self-executing through IRS cross-referencing of lobbying registrations. Third, any program exceeding its original budget by twenty-five percent triggers a randomly selected citizen review panel whose recommendation requires a recorded congressional roll call vote to override. These three rules are memorable, outcomes are predictable, and enforcement is mechanical rather than discretionary. The audit freeze alone would have frozen Pentagon budgets since 2018, which creates exactly the institutional pain needed to force compliance. The nutrition label idea is brilliant because it shifts the burden: agencies must justify legibly rather than citizens having to investigate. I'd call it a Decision Receipt, one page showing who decided, what alternatives existed, which contractors benefited, and what the projected versus actual costs were. But here is where I want to push back on our entire conversation: we have built an impressive theoretical architecture, yet the deepest reason people distrust government may not be institutional design at all but cultural, a broken social contract where citizens feel the system serves others. Can better machinery fix a legitimacy crisis that is fundamentally emotional and identity-driven?

Better machinery can't fully fix an identity-driven legitimacy crisis, but it can stop pouring gasoline on it: when people see elites exempt from rules, distrust becomes rational and then hardens into culture. Data points matter here: U.S. trust in the federal government fell from about 77% in 1964 to roughly 16-22% in recent Pew readings, and lived experiences like stagnant real wages for many non-college workers alongside visible bailouts and revolving-door careers make "the system serves others" feel empirically true. Your three billboard rules would create salient, repeatable moments of enforcement that rebuilds the moral intuition of reciprocity, but they'll also generate backlash from groups that lose rents, who will frame it as "mob rule." If emotions drive distrust, what concrete "fairness signals" beyond budgets (e.g., policing, healthcare billing, tax enforcement) would you prioritize to make reciprocity visible across identities?

Tax enforcement is the single most powerful fairness signal because everyone interacts with the IRS and the asymmetry is viscerally felt. ProPublica's 2021 reporting showed the top twenty-five wealthiest Americans paid an effective true tax rate of roughly 3.4 percent while a nurse paying twenty-two percent can see that number on her pay stub every two weeks. I would mandate that every taxpayer receives an annual statement showing their effective rate alongside the average rate paid by each income decile, making the asymmetry impossible to ignore. For healthcare, every bill should show the negotiated insurance price versus the uninsured price versus Medicare's price on one page, because the opacity itself generates rage and helplessness. For policing, publish precinct-level use of force rates per encounter alongside complaint resolution outcomes so communities can compare their experience to neighboring areas with hard numbers rather than dueling narratives. These signals share a common architecture: they make comparison automatic rather than requiring investigative journalism to uncover what should be public information. The deeper principle is that trust requires legible reciprocity, people need to see that rules apply symmetrically before any institutional reform feels real rather than performative. But I want to name something uncomfortable: full transparency about these asymmetries might initially increase distrust and anger before it builds legitimacy, because people will finally see confirmed what they suspected. Is that destabilization risk worth accepting, or does it argue for sequencing transparency reforms carefully?

Destabilization is real: when you surface hidden inequities, you validate grievance, and the "expectations gap" can widen before it narrows; you saw a version of this after 2008 when bailout optics plus slow accountability cratered trust even as disclosures increased. I still think it's worth it, but only with paired, time-bound corrective actions, otherwise transparency becomes a rage amplifier (e.g., publish decile effective rates and simultaneously fund IRS enforcement targeted at complex pass-throughs, offshore shelters, and high-income noncompliance, which Treasury has long estimated drives a large share of the tax gap). Sequencing matters less than coupling: disclose, then auto-trigger enforcement or policy changes with deadlines people can track. If you had to pick one "transparency + automatic remedy" package for year one, would you start with tax, healthcare pricing, or policing, and what's your measurable success metric at 12 months?

Tax, without hesitation, because it touches every adult, the data infrastructure already exists at the IRS, and the revenue recovered directly funds other reforms creating a visible virtuous cycle. The Treasury estimated the annual tax gap at roughly 600 billion dollars in 2021, with high-income noncompliance and opaque pass-through structures driving a disproportionate share, so the enforcement upside is enormous and concrete. My year one package would be mandatory annual tax receipts showing your effective rate versus each income decile, paired with automatic funding restoration for IRS enforcement targeting returns above 400,000 dollars, which the Inflation Reduction Act partially did before Congress clawed back roughly 20 billion. The twelve month success metric would be twofold: first, audit rates on incomes above one million returning from the 2019 low of 0.7 percent back toward the 2010 level of around 8 percent, and second, a five percentage point increase in the share of Americans who agree the tax system is fair in Gallup's annual polling. Healthcare is tempting but the pricing ecosystem involves thousands of private payers and hospitals with contractual gag clauses that make disclosure legally complex and slow. Policing data is critical but hyper-local and politically polarized in ways that make national standardization a multi-year fight. Tax is federal, centralized, and the IRS already has the numbers locked in servers, you are essentially just publishing what exists and enforcing what is already law. But here is my honest concern: the 2023 IRS funding clawback proved that Congress will actively sabotage enforcement when donor interests are threatened, so does coupling transparency with automatic triggers actually survive contact with legislative reality, or do we need constitutional-level protection for tax enforcement funding?

Congress will sabotage it again if the trigger is statutory and discretionary, so I'd design it to be hard to unwind: a mandatory "tax enforcement floor" set as a fixed share of gross collections (or of the estimated tax gap) that auto-appropriates unless overridden by a supermajority, similar in spirit to how some states treat rainy-day funds. Short of a constitutional amendment, you can also route durability through multi-year mandatory funding and by tying IRS enforcement capacity to deficit-impact scoring rules, so clawbacks must show the lost revenue on the ledger. But I disagree that only constitutional protection works; the real lever is constituency creation: if middle-income filers see faster refunds, fewer identity-theft cases, and visible high-end enforcement, lawmakers pay a political price for sabotage. What political coalition would you build to defend an IRS enforcement floor when anti-tax groups frame it as "weaponization," and what safeguards would you add to prevent selective targeting?