Researchers at the Policy Institute at King’s College London, working with the charity Sense about Science and funded by Wellcome, surveyed 4,262 British adults about trust in scientists. Then they ran regression modelling to find out what actually drives it.
The answer was not expertise.
The two factors most strongly associated with trust were whether people felt scientists are on the side of people like them, and whether they thought scientists were open and transparent. Both outweighed perceptions of scientific competence. The headline number underneath is reassuring: 63 percent of respondents gave trust a score of seven or higher out of ten, and only five percent rated it three or below. Trust in British science is not collapsing.
What is happening is that it is sorting itself by identity. Among Green Party voters, 71 percent feel scientists are on their side. Labour 60, Conservatives 50, and Reform UK 32. Trust in climate scientists specifically sits at 54 percent overall, with Greens at 84 and Reform at 19. Reform supporters are roughly twice as likely as the public to believe scientists conspire with politicians and business. Forty percent of low-trusters feel scientists care more about social and political causes than about scientific progress.
Bobby Duffy, who directs the Policy Institute, framed the finding sharply. Trust is not just about whether people think scientists know their subject. It is about whether people feel scientists are doing good for them.
Now, the counter.
Two coffee shops, same machine, same roaster, same water filtration, same competence. In the first, the barista is technically flawless and will tell you, unprompted, that you have ordered wrong, that milk ruins that origin, that the refractometer says your palate is mistaken. Everything said is true. In the second, the barista asks how you like it, makes it that way, and mentions the new Ethiopian only when you seem curious. Same expertise. Utterly different business.
Everybody in coffee knows which of those two shops survives, and it is not a mystery or a tragedy. It is the same finding the King’s researchers just measured at national scale. People do not evaluate expertise directly, because they cannot. They evaluate whether the expert appears to be working on their behalf, and they use that as a proxy for everything else. It is a reasonable heuristic for anyone who is not themselves an expert, which is all of us about almost everything.
The uncomfortable implication for institutions is that being right is not a communications strategy. If the perception of shared interest is doing more work than the perception of competence, then correcting the record harder achieves nothing, and may actively confirm the suspicion that the expert is arguing rather than helping. The word woke appears in the data for a reason: low-trusters are nearly five times as likely as high-trusters to say scientists are too woke, which is not a scientific objection at all. It is a statement about which team someone thinks the scientist plays for.
That is the risk Duffy names, that polarisation turns trust in science into a marker of political identity rather than a judgement about knowledge, as it has in the United States over the past decade.
The practical version applies to anybody who sells anything on the strength of knowing more than the buyer. Lawyer, mechanic, doctor, developer, barista. Your credentials establish that you could help. Only your conduct establishes that you will.
Sixty-three percent still trust the scientists. Nobody asked whether the scientists trust them back.
Source: Nature, “People trust scientists seen to be ‘on their side’, UK study find”
