Trust in scientists and institutions has not vanished. It has relocated, and the criteria for earning it have quietly changed.
The Credential Collapsed, But Not the Appetite for Knowledge
American trust in scientists fell from 87% at the start of the pandemic to 73% within a few years, while the share reporting little or no trust nearly tripled, climbing from 12% to 27%. That decline was sharp enough to count as a hinge point rather than a slow drift, and it has not meaningfully recovered since.
The shift was not a retreat from knowledge itself. Demand for medical guidance, economic forecasts, and scientific explanation did not fall during this period; in many categories it rose. What changed was narrower and more specific: a title, a lab coat, or a government seal stopped functioning as an automatic guarantee of accuracy. For decades, credentials operated as a shortcut. A person did not need to evaluate a claim on its merits if the claim came from someone whose institution had already done the vetting. That shortcut required two things to hold: confidence that the institution was competent, and confidence that the institution was not compromised by politics, profit, or error. Both assumptions took damage during the pandemic, as guidance changed, predictions failed, and institutions that once spoke with single-source authority were seen disagreeing with each other in public for the first time in most people's adult lives.
Once that happened, the shortcut stopped working, but the underlying need for answers did not disappear. People still wanted to know what was true. They simply could no longer outsource the question of who to believe to a credential alone, and they began applying a different test instead.
The New Gatekeepers Are Likable, Not Credentialed
That different test is built on familiarity rather than certification, and the data on how it actually functions is counterintuitive. Research on what makes a recommendation persuasive found that survey respondents trusted amateur runners more than Olympic athletes, because relatable, visible experience outweighed formal achievement in the eyes of the audience. The Olympian's expertise is real, but it is also remote. The amateur's expertise is smaller, but it is legible. The audience can trace exactly how the amateur arrived at their knowledge, because the audience can imagine arriving there themselves.
This single finding explains a pattern that shows up across unrelated fields. A doctor's medical degree signals years of training the average person cannot evaluate directly. A wellness creator explaining their own diagnosis, treatment, and recovery signals something the audience can verify against their own experience step by step. The credential asks for trust based on process the audience cannot see. The personal account asks for trust based on a story the audience can follow start to finish. When both are offered side by side, a measurable share of people now choose the second, not because they have concluded the doctor is wrong, but because the story format gives them a way to check the claim that the credential format does not.
This is the actual mechanism driving what gets called a crisis of expertise. It is not that audiences became hostile to expertise. It is that a new, competing form of authority emerged that satisfies a need the old form never had to address: the need to feel that the path to the conclusion is visible and walkable, not just the conclusion itself.
The Trust Didn't Vanish. It Migrated.
The scale of that migration is large enough to function as a parallel information economy rather than a niche alternative. Roughly half of consumers say they depend on influencer recommendations to make decisions, and close to seven in ten say they trust what influencers tell them, figures that match or exceed measured trust in traditional institutions across several categories.
This extends well past lifestyle content and product reviews. Two in five American adults now get health and wellness information from social media influencers or podcasts, a category that a generation ago belonged almost exclusively to clinicians with medical degrees and the institutions that trained them. The expert has not been removed from that conversation. They have been repositioned as one voice competing among many, on terms set by visibility and relatability rather than by rank or credential. A board-certified physician posting a careful, hedged explanation competes directly against a creator with no medical background offering a confident, simplified one, and confidence plus clarity frequently wins the engagement that determines who gets heard at all.
This is not a failure of the audience's intelligence. It is the predictable output of a system that rewards attention held, not accuracy delivered. Trust did not collapse inside that system. It became a side effect of whoever held attention long enough to be believed.
The Honest Counterpoint
Not every measure supports a sweeping crisis narrative, and the strongest objection to this argument comes from researchers rather than commentators. A recent review in a peer reviewed psychology journal found little evidence of a broad public crisis of trust in science, noting that the polarization that does exist is concentrated in specific countries and specific topics rather than universal across the public.
That distinction matters and should narrow the claim rather than soften it. The accurate version of this argument is not that institutional trust collapsed everywhere, for everyone, on every subject. It fractured unevenly, concentrated heavily around politically charged topics like climate science and vaccines while remaining comparatively stable in areas with less political stakes. What did shift uniformly, even where overall trust held steady, is the criteria audiences now apply before extending it. Trust became conditional on perceived transparency and relatability in a way it was not before, and that conditional opening created space for anyone visible enough to occupy it, regardless of whether they had earned the position through expertise.
The Question
If trust now depends on familiarity rather than credentials, what happens the first time a familiar voice is wrong, and no credentialed check remains in the loop to catch it?
0 Comments