When is distrusting institutions the rational move?: The Story

The vial, the pill, the priest

February 5, 2003. Colin Powell held up a vial of white powder at the United Nations and told the world Iraq possessed weapons of mass destruction. The New York Times published it. Congress authorized the invasion. The weapons were never found. Four thousand American soldiers and several hundred thousand Iraqi civilians were dead, and the apparatus that produced the lie was still operational.

Twelve months earlier, a different institution was failing at a different speed. Purdue Pharma’s internal data showed OxyContin was addictive at rates far exceeding its marketing. The FDA had approved it. Doctors prescribed it 7.2 million times in 2002 alone. By the time the deception became public, over 500,000 Americans had died.

Underneath both, a slower catastrophe. The Catholic Church knew about predator priests for decades. The Boston Globe’s Spotlight investigation documented a system of concealment spanning continents and centuries. Bishops reassigned known abusers. Vatican officials suppressed investigations.

The dataset

Iraq. OxyContin. The Church. Each failure alone might be absorbed as aberration. Together, they constitute a dataset. The dataset says: the institutions demanding default trust have, on the questions where trust mattered most, earned default suspicion.

For some, the topsoil has washed away — institutional credibility does not regenerate on a human timescale, and the trust defaulters have stopped waiting. The Bayesian math is not much kinder: by the time sub-prime mortgage ratings collapsed in 2008, the rational skeptics had run out of prior to give. But treating the CDC’s vaccine data and the Church’s internal investigations as the same kind of claim is its own error — what does trust even mean when applied identically to both? The deeper question may not be whether to trust but whether the structures worth trusting can be rebuilt at all.

The sharpest version

Is institutional trust a technology that broke, or a technology that was always mis-described to the people using it? If it broke, it can be fixed. If it was always a sales pitch dressed as social infrastructure, the fix is not repair but replacement. The 5:1 liberal-to-conservative ratio among journalists is a real structural bias — but it tells you something specific about political coverage and nothing about NOAA’s hurricane tracking.

The institutions that lied about Iraqi weapons of mass destruction, approved an opioid that killed half a million Americans, and sheltered predator priests for decades still issue guidance every morning and still expect compliance. Each failure alone might be absorbed as aberration. Together they constitute a dataset, and the dataset says something specific: the institutions demanding default trust have, on the questions where trust mattered most, earned default suspicion. The question is what to do with that information. Some treat every institutional claim the way a burned investor treats a pitch deck. Some run the Bayesian math and find the posterior probability of reliability too low to justify deference. Some insist the question is never whether to trust institutions but which institution, on which topic, with what track record. The disagreement matters because the person caught in the middle — the parent deciding about a vaccine, the homeowner reading a water-quality report — cannot wait for resolution.

Perspectives:
- Trust defaulters
- Rational skeptics
- Selective trust
- Repair advocates