Anti-Intellectualism & Trust in Science

Search a term like “anti-intellectualism” and an AI-generated summary will hand you a dictionary definition in under a second: distrust of experts, intellectuals, or established knowledge. That definition is accurate and it is also close to useless for anyone actually responsible for a research organization’s public standing. The operational questions are different: What does the evidence actually say about trust in science, as opposed to the anecdote of a hostile comment thread? Which mechanisms are driving change, and which are just noise? And what, concretely, should a researcher, a communications office, or a research integrity officer do differently because of it? This guide is written for that audience, not as a general essay on the phenomenon.

What the evidence actually shows about trust in science

Public trust in science is not a single number that is either “up” or “down” – it is a set of related measures that move differently depending on what is asked, who is asked, and about which domain of science. Three sources anchor most of what is reliably known:

Pew Research Center has tracked Americans’ confidence in scientists to act in the public interest on a roughly annual basis since 2016-2019. Its reporting shows a clear pattern: confidence was comparatively high before 2020, dropped notably around the COVID-19 pandemic period as science became entangled with contested public-health policy, and has not fully recovered – Pew’s November 2024 report on the topic describes trust as “slightly higher than it was” the prior year but “remains lower than before the COVID-19 pandemic.” A consistent, separate finding across Pew’s surveys is a partisan gap in confidence that widened over the same period, and trust levels that vary by scientific field (medical scientists, for instance, are typically rated differently than researchers working on climate or contested policy topics). See Pew’s ongoing trust-in-science research hub for the current reports rather than a single cited figure, since Pew re-surveys and republishes this regularly.
Wellcome Global Monitor (fielded via the Gallup World Poll across well over 100 countries) is the largest cross-national study of public attitudes toward science and health, and its design lets researchers compare trust levels across very different political and economic contexts rather than within a single country. It consistently finds that trust in scientists is not a Western-vs-non-Western or rich-vs-poor-country story in any simple sense – it correlates more with local trust in institutions generally, and with whether people feel science “is for people like them,” than with national income.
Eurobarometer (the European Commission’s recurring survey programme) periodically measures EU citizens’ knowledge of and attitudes toward science and technology, and – like Pew – tends to find that trust in scientists as a profession remains comparatively high even where trust in specific claims, institutions, or policy responses is contested.

The consistent thread across all three: generalized distrust of science as an enterprise is a minority position in most measured populations. What has genuinely shifted is something narrower and more consequential for research organizations – trust has become more domain-specific and more politically sorted. People increasingly evaluate a scientific claim partly by who is making it and what it implies for a policy position they already hold, not only by its methodological merits. That distinction matters because it changes what a research office should actually respond to: not “is the public turning against science,” which the aggregate data does not support, but “which specific findings, in which specific political contexts, are going to be read adversarially.”

Four mechanisms behind the shift, not one

“Anti-intellectualism” is often used as a single explanatory label, but the erosion in domain-specific trust described above is better understood as the product of several distinct, partly independent mechanisms.

1. Politicization

When a scientific finding has a direct, visible implication for a live policy fight – vaccine mandates, climate regulation, gun research – public reaction increasingly tracks political identity more than it tracks the underlying evidence. This is well documented in the survey literature above as a widening partisan gap in trust, concentrated in specific fields rather than science overall. It is a genuine mechanism, not a media artifact, but it is also domain-limited: it explains division over climate science far better than it explains division over, say, materials engineering or astronomy.

2. The replication crisis and its aftermath

Beginning most visibly with psychology in the early 2010s, a wave of high-profile failures to replicate published findings gave critics of science a concrete, citable grievance rather than an abstract one: “even scientists can’t reproduce their own results.” CASRAI’s guide to the replication crisis covers the origins and the open-science response in depth; the point relevant here is that the field’s response – preregistration, open data, reproducibility standards, registered reports – is itself a trust-repair mechanism, and one worth citing directly when a research office is asked “how do we know this isn’t another one of those.” A defensive answer (“that was a different field”) lands worse than pointing to the concrete methodological reforms that followed.

3. Misinformation dynamics

Social platforms did not invent scientific misinformation, but they materially changed its economics: a false or oversimplified claim about a study can now reach more people, faster, and with less friction than a careful caveated summary of the same study – especially once it is stripped of the uncertainty language the original paper actually used. Researchers who skip the “what this study does and does not show” framing when a finding is newsworthy are, in effect, leaving that framing to be supplied by whoever amplifies it next, which is rarely careful.

4. Institutional distrust as the deeper current

In much of the data above, distrust of scientists correlates more strongly with general distrust of institutions – government, media, large organizations – than with anything specific to science. This is arguably the least tractable mechanism for an individual researcher or university communications office to address directly, because it did not originate in science and will not be resolved by better science communication alone. It is, however, the reason a purely “explain it better” response consistently underperforms – see below.

Why “just explain the facts better” keeps failing: the deficit model problem

The intuitive response to public skepticism – give people more accurate information, more clearly – is what science communication researchers call the deficit model: the assumption that distrust results from a gap in public knowledge, and that closing the gap closes the trust deficit. Decades of science communication research have repeatedly found this model incomplete at best. People with more scientific literacy are not uniformly more trusting of contested findings; in some documented cases, higher literacy combined with strong prior political identity produces more sophisticated motivated reasoning against a finding, not less resistance to it. Trust responds more reliably to perceived motives, transparency about uncertainty and conflicts of interest, and whether the communicator is seen as sharing the audience’s values, than to the sheer volume or clarity of facts presented.

This does not mean clear communication is worthless – it means clarity is necessary but not sufficient, and a strategy built entirely on “correct the misunderstanding” will underperform a strategy that also addresses source credibility, two-way engagement, and honest uncertainty. This is the operating assumption behind current public engagement practice (see below), not a fringe position within the field.

What this means operationally

Public engagement is a funder obligation, not a soft skill

For US-funded research, public engagement and scientific literacy is one of the explicitly listed example categories under NSF’s Broader Impacts merit review criterion (NSF PAPPG Chapter III.A) – alongside STEM education, workforce development, and societal well-being. Broader Impacts is not optional add-on framing; it is one of two merit review criteria applied to every proposal, weighted alongside Intellectual Merit. UKRI and Wellcome run comparable, if differently structured, public engagement funding requirements and standalone schemes. CASRAI’s guide to science communication and public engagement grants covers the funder-specific mechanics (NSF, UKRI, Wellcome) in detail and is the better resource if you are drafting a Broader Impacts section rather than trying to understand the trust landscape behind it; the two pages are meant to be read together, not as substitutes for each other.

Communicating uncertainty without sounding evasive

A recurring failure mode: researchers either overstate certainty to sound authoritative, or hedge so heavily that a lay reader hears “we don’t actually know anything,” both of which cost credibility when the story develops further. The more defensible pattern, consistent with the deficit-model critique above, is to state plainly what is known, what is not yet known, and what would change the conclusion – framed as normal scientific practice rather than as a weakness being admitted under pressure. Naming the limitations of a single study proactively, before a critic does it for you, is a credibility signal, not a liability, in most of the survey evidence on source trust.

Handling hostile or bad-faith coverage

Not every critical response is a good-faith request for clarification, and treating all of it as such wastes limited communications capacity. A workable operational split: respond substantively and promptly to good-faith misunderstandings (these are winnable and consistent with the evidence that transparency builds trust); do not attempt to “win” an exchange with a source arguing in bad faith, since engagement itself can amplify a fringe claim’s visibility – a well-documented dynamic in misinformation research generally. Research integrity offices and communications teams are better served by a pre-agreed threshold for which responses get a public reply and which get logged and monitored instead of engaged.

Where the evidence is genuinely contested

In the interest of treating this even-handedly: not every claim in this space is settled. The size of the partisan trust gap, how much of it is durable versus event-driven (a spike around a specific controversy that partially recedes), and how much responsibility platform algorithms bear versus pre-existing political sorting, are all active areas of disagreement among researchers who study this professionally. Readers should treat any single statistic – including the ones cited above – as a snapshot from a specific survey wave and methodology, not a fixed fact about “the public,” and should check the primary source’s current release before citing a number in an institutional communication.

Frequently asked questions

Is public trust in science actually declining, or does it just feel that way?

The honest answer is “it depends which measure and which domain.” Aggregate confidence in scientists generally has not collapsed in the datasets above, but trust has become more politically sorted and more field-specific since around 2020, which is a real and consequential change even though it is not the same as a uniform collapse in trust.

Does better science communication actually reduce anti-intellectualism?

Clear communication helps but is not, on its own, sufficient – see the deficit-model discussion above. Source credibility, transparency about uncertainty and conflicts of interest, and two-way engagement consistently outperform one-way “explain the facts” messaging in the science communication literature.

Is anti-intellectualism the same thing as science denial?

No. Anti-intellectualism is a broader disposition of distrust toward experts and intellectual authority generally; science denial refers to rejection of a specific, well-established scientific consensus (evolution, vaccine safety, anthropogenic climate change are the most studied examples). The two overlap but are not synonymous – someone can hold one without the other.

What should a research integrity office do differently because of this?

Treat public-facing transparency as part of the integrity function, not a separate communications problem: proactively disclosing limitations, funding sources, and conflicts of interest in public-facing summaries does real trust-building work, independent of and in addition to whatever it does for internal compliance.

Anti-Intellectualism and Public Trust in Science: What the Evidence Actually Shows

Ask about Anti-Intellectualism and Public Trust in Science: What the Evidence Actually Shows