Andrea Rosso spent years building a study of how life in divested, majority-Black neighbourhoods shapes the risk of Alzheimer's disease. This summer the money stopped, months after it had already been delayed once, on the grounds that her research no longer aligned with federal priorities. Her colleague at Emory, Negar Fani, was midway through mapping how self-reported experiences of racial discrimination correspond to measurable changes in the brain's white matter. Her grant went the same way. The letters from the National Institutes of Health made the objection explicit: work built on surveys and self-report was judged insufficiently objective to count as science.

That is a strange claim to make about a field that has always depended on people describing their own pain, their own sleep, their own trauma. It is also, on its face, a domestic American story, playing out through a federal funding agency and a Supreme Court that has already allowed hundreds of millions in similar cancellations to proceed. But the argument being deployed here, that measuring the health effects of discrimination is inherently unscientific, does not belong to any one funding system. It is a transferable claim, and Britain is already rehearsing a milder version of the fight it emerges from.

Reform UK's chairman has talked openly about sending "task forces" into the councils the party now controls to strip out diversity and inclusion spending, and the same rhetoric has been aimed at similar roles within the NHS. So far the argument has stayed at the level of job titles and headcount: whether an NHS trust needs a named diversity lead, whether a council post is worth its salary. The NHS Confederation has pushed back, framing such roles as investment rather than waste. It is, in other words, a fight about personnel, and personnel fights are recoverable. A post can be reinstated. A programme can be relaunched.

What the American cancellations show is what happens once the argument moves past headcount and reaches the evidence base itself. Nobody in Washington fired a diversity officer this summer. They defunded the actual research documenting how discrimination shapes long-term health outcomes, using a methodological objection dressed up as scientific rigour. That is a harder thing to reverse than a job title, because once a study stops, its data collection stops with it, and the years already invested in cohorts and follow-up cannot simply be restarted when political winds shift.

Britain has its own smaller version of the infrastructure that would be vulnerable to that argument. The NHS Race and Health Observatory, still a relatively young body, has just launched new community grant programmes and renewed funding to study ethnic disparities in maternal and neonatal outcomes, among other things. Its work rests on exactly the kind of evidence the American cancellations single out: surveys, self-report, community testimony about how discrimination is experienced. None of that funding is under serious threat today. But the rhetorical tools now being tested in Washington, the claim that lived experience cannot count as data, are cheap to import and do not require any particular electoral outcome to start circulating in policy debate.

There is a version of fiscal pressure heading toward the NHS this autumn that makes this worth naming now rather than after the fact. A tight Budget in October will sharpen every argument about which programmes are essential and which are discretionary, and small, politically exposed research bodies are precisely the sort of spending that gets reclassified as discretionary when departments are asked to find savings. The American case offers British policymakers something more useful than a warning about American politics. It offers a preview of the specific argument that will be used, should anyone here choose to make it: not that diversity work costs too much, but that the underlying research was never rigorous to begin with. NHS leaders and life sciences funders would do well to have an answer to that argument ready before it is asked of them, rather than after the grants have already stopped.