Indefensible: Why Trust Is a Systems Outcome, Not a Promise

Inspired by: Indefensible, by Ben Taub (The New Yorker, January 27, 2020)

PART I — WHY TRUST COLLAPSES

The Rhetorical Question That Won’t Go Away

Every few years, a government agency, a congressional panel, or a national security office enters the spotlight and produces a statement that is — at a minimum — puzzling, and at worst, irreconcilable with observable reality. When this happens, ordinary people ask a question that sounds angry but is actually deeply rational: How can anyone trust what the government says? The question gets dismissed as cynical, yet it is the opposite. It is an intuitive public demand for coherence — a basic feature of any functioning system of truth. Trust does not collapse because citizens are skeptical; trust collapses because official claims fail elementary tests of consistency, evidence, and transparency.

In “Indefensible,” Ben Taub illustrates this problem with clarity: government statements contradict each other; evidence is withheld; explanations shift depending on the audience; and the actors inside the system are often shielded from meaningful accountability. The details differ from case to case, but the pattern is familiar. Wherever the incentives inside a public institution reward secrecy, the public’s trust — rightly — evaporates. The collapse of trust is not a cultural failure. It is a systems failure.

The Indefensible Pattern: Contradictions, Silences, and Systems That Reward Opacity

Taub’s article does more than recount an episode of institutional incoherence. It shows how contradictions become the predictable output of a system designed to protect itself from scrutiny. When an official narrative is offered without access to underlying evidence, and when contradictory explanations persist long after they are exposed, trust cannot survive. People think they are reacting to a shocking story. They are actually reacting to a structural design.

In this sense, the problems Taub describes aren’t moral failures of individuals. They are failures of architecture. The system itself creates incentives that make deception easier than transparency, ambiguity safer than clarity, and contradiction less costly than accountability. Once these incentives take hold, they are remarkably durable. Systems that reward opacity generate more opacity. Systems that shield decision-makers from scrutiny encourage bolder forms of contradiction. And systems that restrict access to evidence create the illusion of certainty while eliminating the conditions that produce real certainty.

Trust collapses not because people lack faith, but because the structure of the institution makes reliable information practically impossible to obtain. When the public encounters contradictions, shifting explanations, or withheld evidence, they are observing the logical outcome of a system that was never engineered for truth in the first place.

Historical Lens: When Trust Failed for Structural Reasons, Not Cultural Ones

It is tempting to think distrust is a modern phenomenon — a product of social media, polarization, or a vaguely defined sense that “institutions don’t work like they used to.” But a closer historical lens shows something subtler and more revealing. Moments of profound public distrust have almost always arisen from structural conditions that rewarded secrecy, insulated decision-makers from accountability, or prevented independent verification of official claims.

The Vietnam era is a classic example. The Pentagon Papers did not erode public trust because they were shocking; they eroded trust because they exposed how thoroughly the government had controlled information to maintain a narrative detached from internal evidence. The public’s reaction was not hysteria or naïveté — it was a rational systems-level response to learning that the architecture of information flow had been weaponized.

The Iraq WMD intelligence failure followed a similar pattern. Competing claims existed, but the internal system was designed to favor confirming evidence and suppress contradictory analysis. This wasn’t a cultural failure of optimism or patriotic blindness. It was a structural failure of information vetting, incentive alignment, and independent review.

The Flint water crisis, too, was not fundamentally a story about individuals making bad choices. It was about a system in which the cost of acknowledging contamination was borne by the government, while the cost of not acknowledging contamination was borne by the public — a perfect inversion of aligned incentives. Predictably, trust collapsed.

And long before any of these, the Tuskegee syphilis study revealed the catastrophic consequences of information monopolies: deception that could flourish only because participants had no access to external verification, and the public had no visibility into internal decision-making. Trust did not die because the public was fragile. Trust died because the system structurally forbade truth.

Across every example, a pattern appears: Whenever public institutions create or maintain monopolies on information, trust collapses as soon as the monopoly cracks.

Information Monopolies: The Root of Institutional Contradiction

A system fails when one entity controls what facts can be known, who can verify them, and how they can be challenged. This is the definition of an information monopoly, and it is the single most important predictor of institutional distrust.

Information monopolies produce several predictable outcomes:

• Contradictions: Because internal incentives encourage coherence to power, not coherence to evidence. • Silences: Because revealing information increases risk for insiders and rarely yields institutional benefit. • Shifting explanations: Because narratives must adapt to new leaks or external scrutiny rather than to internally accepted facts. • Detached certainty: Because claims cannot be meaningfully challenged, so certainty becomes a rhetorical performance rather than a measurable property.

It is here that Taub’s “Indefensible” becomes more than a troubling episode. It becomes a case study in how monopolies on information distort reality and undermine the very conditions that make trust possible.

Trust cannot coexist with information monopolies. Not because people are suspicious — but because the system itself has removed the mechanisms by which trust could be rationally earned.

Why Trust Fails as a System

Trust is not the default state of human institutions. It is the emergent outcome of coherence, verification, and accountable incentives. When those foundations erode, trust doesn’t fade slowly — it collapses abruptly, like a bridge whose load-bearing structure has been invisibly compromised.

A system cannot produce trust when:

• Claims cannot be externally validated • Evidence is withheld or selectively released • Accountability mechanisms lack independence • Incentives reward loyalty over truth • Corrections are costly, embarrassing, or institutionally dangerous • Errors have no clear pathway to acknowledgment or repair

In such systems, distrust is not a mood. It is a mathematically predictable output — as inevitable as rust in an unprotected metal beam.

The public’s skepticism, then, is not evidence of social decay. It is evidence of people intuitively responding to systems that no longer obey the logic of truth.

The Science Analogy: Trust Emerges From Replication, Not Rhetoric

If you want to understand why trust collapses in opaque government systems, compare them with systems that reliably produce truth. Science is not perfect, but it has one structural advantage that governments often lack: no single actor gets to monopolize evidence. Claims must be replicable. Methods must be transparent. Data must be available for independent inspection. And when findings cannot be reproduced, the claim does not become more eloquent — it simply becomes less credible.

Scientific truth is not enforced by authority; it is enforced by conflict. Researchers question each other. Labs attempt to replicate results. Journals require methods sections detailed enough for reproduction. And when fraud occurs, it is typically exposed from outside the originating team.

In other words, trust in science is an emergent property of distributed verification.

Government systems often operate in the opposite direction. Instead of encouraging external replication of their claims, they frequently limit access to the underlying evidence. Instead of distributing the tools of verification, they sometimes withhold data, redact key details, or classify the very information that could resolve contradictions. Instead of rewarding those who find errors, they can punish or marginalize them.

No matter how sincere a government actor may be, the system cannot produce trust if it forbids the conditions required for truth. This is why scientific trust tends to be self-repairing — and institutional trust, once damaged, is so hard to restore. The architecture is different.

The Paradox: Trust Requires Making Deception Impossible

The deeper lesson is counterintuitive. A trustworthy system is not one that assumes good intentions. A trustworthy system is one that makes deception structurally impossible.

That means transparency loops. It means independent review. It means competing centers of evidence that no single actor controls. It means incentives that reward correction rather than concealment.

Sincerity cannot substitute for structure. Goodwill cannot substitute for constraints. And hope cannot substitute for verification.

The paradox is that trust — something we imagine as delicate, emotional, and easily bruised — actually emerges from rigor and resistance. Systems build trust not by softening scrutiny but by hardening it.

Reframing the Collapse

When people ask, “How can anyone trust the government?”, the question often sounds cynical because we frame trust as a matter of virtue. But trust is not about virtue. It is about engineering. It is about whether systems allow contradictions to stand unchallenged or force them into the open. Whether errors quietly compound or become visible early. Whether those who uncover the truth are empowered or silenced.

The public’s distrust is not a flaw in civic culture. It is an observable systems response to architectures that do not permit truth to accumulate.

Transition to Part II

Part I showed why trust collapses: opaque incentives, information monopolies, and incompatible relationships between evidence and authority. Part II will show the other half of the equation: how trust is engineered — how systems can be designed to produce transparency, distribute verification, and make coherence possible again.

HOW TRUST IS ENGINEERED

Trust Is Not a Sentiment. It Is a Designed System Property.

If Part I shows why trust collapses, Part II shows how trust is built. The central insight is simple but transformative: trust is not created by personalities; it is engineered through constraints. A trustworthy institution is not one filled with virtuous people. It is one in which even flawed people would find it difficult to lie, conceal, or distort without detection.

This distinction matters. A system that relies on virtue will eventually collapse under the weight of ordinary human incentives. A system that relies on structural transparency—distributed evidence, independent review, and consequences for distortion—can produce trust even when humans inside the system are imperfect.

Trust, then, is a systems output: the predictable behavior of an institution whose design makes truth more stable than deception.

Engineering Principle 1: Transparency Loops

A transparency loop is a structural feature that forces information to flow outward rather than upward. Examples include public reporting requirements, external audits, mandatory data releases, and time limits on classification.

What these loops have in common is that they prevent any single actor from becoming the sole interpreter of the truth. When information automatically flows to the public or to independent reviewers, institutions cannot rely on narrative control. Claims must be anchored in evidence because the evidence is visible.

Transparency is not a cultural value; it is a mechanical one. A transparency loop works whether or not the people inside the system wish it did.

Engineering Principle 2: Distributed Verification

One of the strongest trust-building mechanisms is structural redundancy. When multiple independent entities have access to the same evidence—scientific boards, journalists, courts, inspectors general—no single actor can dominate the narrative.

Distributed verification means that truth does not have to be believed. It has to survive scrutiny.

This is why democratic systems rely on separation of powers, why scientific research requires replication, and why public data portals matter. The moment verification becomes distributed, it becomes difficult for contradictions to persist. The architecture collaborates with the truth.

Engineering Principle 3: Oversight That Cannot Be Turned Off

Some accountability mechanisms exist only when convenient. They can be paused, bypassed, ignored, or politically weakened. These do not build trust; they build the appearance of oversight.

Real trust emerges only when oversight cannot be suppressed. Independent inspectors general must retain subpoena power even when their findings embarrass the people who appointed them. Auditors must be protected from retaliation. Scientific advisory boards must have access to unfiltered data. Courts must be able to demand evidence, not assurances.

If an oversight system can be disabled at the moment it is needed most, it is not a system of trust. It is a system of optics.

Engineering Principle 4: Incentives That Reward Truth and Penalize Distortion

No system can rely on the goodwill of its participants. Incentives must be engineered so that accuracy is in the institution’s interest.

This can take many forms:

• Rewarding early disclosure of errors • Penalizing the suppression of contradictory evidence • Protecting whistleblowers with meaningful, enforceable guarantees • Making transparency less costly than concealment • Structuring promotions around reliability rather than loyalty

Institutions follow their incentives, not their aspirations. When incentives align with truth, trust follows.

How These Mechanisms Work in Practice: Historical Cases Where Trust Rose

Although failure dominates the headlines, history shows moments when structural reforms dramatically increased public confidence. The pattern is consistent: trust increases when systems distribute verification, create transparency loops, and embed oversight that no actor can disable.

After Watergate: The creation of new disclosure rules, campaign finance transparency, and inspector general frameworks did not merely punish wrongdoing. They installed structural mechanisms that made future deception harder. Public trust rose not because the public became optimistic but because the system became more verifiable.

The National Transportation Safety Board (NTSB): Its investigations are widely regarded as the gold standard of transparency. Findings are public, evidence is accessible, and hearings are open. The trust people place in the NTSB does not depend on liking its members. It depends on knowing that the evidence speaks louder than the institution.

Auto Safety in the 1970s–1980s: The widespread release of crash-test data, mandatory reporting systems, and independent verification by consumer groups created an ecosystem in which truth was distributed. Fatalities fell dramatically, not because manufacturers suddenly became benevolent, but because hiding unsafe design became structurally impossible.

GAO Audits: When the Government Accountability Office releases findings based on independently verifiable evidence, its credibility stems from architecture: the GAO reports to the legislative branch, not the executive, and its findings are public by design. The system creates trust, not the agency’s personality.

These examples show that trust does not return through speeches or appeals to unity. Trust returns when systems regain the mechanical ability to make truth visible.

The Engineering Metaphor: Systems That Make Truth the Path of Least Resistance

Engineering offers a surprisingly accurate way to understand institutional trust. In well-designed systems, safety and reliability are not ideals; they are emergent behaviors created by constraints. Bridges don’t stay up because the steel is virtuous. Airplanes don’t remain aloft because the pilots are noble. These systems are engineered so that even when something goes wrong, the design channels failure into detectable, correctable forms.

Trustworthy institutions work the same way. They do not rely on the integrity of their leaders; they rely on the integrity of their architecture.

An institution cannot ask for trust; it must behave its way into trust by making honesty the easiest choice, transparency the default state, and verification, the shared expectation. When the structure forces information into the open—and when independent actors can inspect, challenge, and replicate that information—trust becomes the natural by-product.

Whereas Part I showed why trust collapses—opaque incentives, information monopolies, and contradictions that cannot be resolved—Part II shows something more hopeful: trust is not mysterious. It can be deliberately constructed.

Sidebar: Five Engineering Principles of Public Trust

1. Transparency Information must escape the institution automatically. Leaks should be unnecessary in a healthy system.

2. Distributed Verification Multiple independent actors must have access to evidence. No single entity can be the sole arbiter of truth.

3. Irreversible Oversight Accountability mechanisms must remain active even when they threaten those in power.

4. Incentive Alignment The system must reward accuracy, early correction, and evidence-based decision-making—and penalize obfuscation.

5. Feedback Loops Errors must not only be detectable but correctable, with clear pathways to public acknowledgment and systemic repair.

These principles are not moral suggestions. They are mechanical requirements, as essential to trustworthy governance as load calculations are to bridge design.

Conclusion: Reframing the Original Question

The essay began with a rhetorical question that still hangs in the air: How can anyone trust what the government says? The power of that question lies not in its frustration but in its precision. Distrust is not emotional overreaction. It is a systems diagnosis.

When institutions contradict themselves, hide evidence, or prevent independent verification, distrust is not just reasonable—it is mathematically inevitable. Trust is not what people feel when they are inspired. Trust is what people observe when systems behave predictably under scrutiny.

The better question, then, is not “How can anyone trust the government?” but “What structural conditions make trust a rational response?”

The answer is clear: Systems must be designed so that truth is easier than deception, transparency is harder to suppress than reveal, and verification is distributed rather than monopolized. When the architecture changes, the trust will follow.

This shift in perspective is not cynical; it is constructive. It shows that trust is neither naïve nor nostalgic. It is an engineering challenge—and engineering challenges can be solved.

Sources (Annotated)

Taub, Ben. “Indefensible.” The New Yorker, January 27, 2020. Investigates contradictions and withheld evidence within U.S. national-security processes, providing the central case study for how information monopolies and opaque incentives undermine institutional trust.

Ellsberg, Daniel. The Pentagon Papers. A foundational example of how public trust collapses when internal evidence contradicts official narratives—and how distributed information can change a nation’s understanding of its own institutions.

United States Government Accountability Office (GAO). Selected audit reports. Demonstrate how independent, irreversible oversight mechanisms increase institutional trust by making evidence and analysis publicly available regardless of political convenience.

National Transportation Safety Board (NTSB). Accident Investigation Reports. Gold-standard models of transparent, public-facing evidence review. Illustrate how trust emerges when systems distribute verification and make all findings openly accessible.

Markowitz, Gerald & Rosner, David. Deceit and Denial: The Deadly Politics of Industrial Pollution. A historical documentation of cases such as the lead industry and chemical pollution, showing how information concealment by institutions predictably erodes public confidence.

Oreskes, Naomi & Conway, Erik M. Merchants of Doubt. Explains how manufactured uncertainty exploits structural weaknesses in verification systems. Useful for understanding why institutions that control information can distort public understanding for long periods.

Gawande, Atul. “The Checklist Manifesto” (selected chapters). Although focused on medicine and aviation, its core insight—that structured transparency and reproducible process create reliability—applies directly to how trustworthy systems are engineered.

© 2025 Michael A. Pink