Trump's science purge and the cost of ideology

The Trump administration’s decision to cut dozens of federal science advisory committees — panels staffed by external researchers who provide independent technical guidance to agencies ranging from the FDA to the EPA — has now been running long enough to produce its first documented consequences. NPR reported this week on the accumulated effects: regulatory decisions delayed for want of technical expertise, agency staff relying on in-house analysis that lacks the peer-validation that outside committees provided, and a growing number of research universities that have quietly advised scientists not to accept federal advisory roles, anticipating further political interference. The cuts were part of a broader effort to reduce the influence of what administration insiders characterise as a “permanent scientific establishment” hostile to the administration’s priorities. The committees affected include bodies advising on pharmaceutical approvals, chemical safety, environmental monitoring, and pandemic preparedness — a range that spans the full geography of federal regulatory activity.

The received wisdom

The progressive framing of this story is, frankly, largely correct in its factual content, even if its political inference is sometimes overdrawn. Science advisory committees exist precisely to insulate regulatory decisions from short-term political pressure; dismantling them concentrates decision-making authority inside agencies whose leadership serves at the president’s pleasure. The affected panels were cheap relative to their function — most operate on budgets of hundreds of thousands of dollars annually, not billions — and their elimination saves almost nothing in fiscal terms while removing a check on regulatory quality that took decades to build. The editorial consensus among science journals and public health institutions is that the cuts represent a degradation of regulatory infrastructure for ideological reasons, and on the narrow technical question of whether these committees produced value, that consensus is defensible.

Canada’s simultaneous decision to actively recruit top American researchers — dozens of scientists have accepted positions at Canadian universities, with more in the pipeline — is the market’s verdict on what the U.S. is doing to its research environment. When talent chooses to leave, the argument that the pruning is merely cosmetic becomes hard to sustain.

A different read

And yet the more interesting question is not whether the cuts were wise — they were not — but why the administration made them, and what that tells us about a recurring failure mode in American conservatism.

The appeal of dismantling scientific advisory structures is not primarily fiscal. It is political. Over several decades, the federal advisory committee system expanded substantially, and with it expanded a class of experts whose professional identity became intertwined with regulatory activism. A number of environmental advisory panels, in particular, took positions that went well beyond technical guidance into advocacy — recommending regulatory outcomes that reflected particular policy values rather than simply describing the scientific state of play. The blurring of the line between technical expertise and political preference gave fuel to critics who argued, not without some justification, that “the science” was being deployed as a conversation-stopper to foreclose democratic deliberation on questions that actually involve value trade-offs, not just empirical facts.

This is a real problem. The politicisation of expertise is not a conservative invention. The credentialism that treats expert consensus as beyond the reach of democratic accountability has produced genuine failures: the confident assertions of epidemiological models in 2020-21 that proved substantially wrong in their predictions; the long-running suppression of heterodox views on dietary fat, on the origins of COVID-19, and on the benefits and risks of various pharmaceuticals. Scientists who raised these questions through legitimate channels were sometimes marginalised not on scientific grounds but on political ones. The conservative intuition that the federal advisory structure had captured some of this institutionalised groupthink is not entirely paranoid.

But — and this is the critical point — the appropriate response to the politicisation of expert bodies is reform, not demolition. The relevant historical model is the Thatcher government’s approach to the British civil service in the 1980s: a genuine effort to break up institutional sclerosis that, in several cases, went too far and damaged long-built institutional capacity that proved expensive to rebuild. The Thatcher government’s abolition of the Greater London Council, for instance, was politically satisfying but created a governance vacuum in London that lasted nearly twenty years until the GLA was created. Institutions that are abolished tend not to be cleanly replaced; the muscle memory, the informal networks, the institutional knowledge — these do not survive reorganisation charts.

The administration’s retaliation against Anthropic, which a judge ruled was illegal, is a useful companion story. There, too, the pattern is one of using administrative power against entities perceived as insufficiently loyal rather than engaging the underlying policy question — in that case, the appropriate regulatory framework for frontier AI — on its merits. A genuinely conservative approach to AI regulation would involve building expert advisory capacity, not dismissing it. What the administration is engaged in instead is closer to what might be called reflexive anti-institutionalism: a disposition in which any established body, any accumulated expertise, any independent voice is presumptively suspect, and the appropriate response is removal rather than engagement.

The irony is that this approach undermines conservative goals rather than advancing them. Regulatory agencies with less external oversight are not less powerful; they are more powerful, and less accountable. The advisory committees that were cut served as a constraint on agency discretion — which means their removal leaves agencies freer to act without external validation. That is not a conservative outcome. It is an administrative-state outcome with a different set of political preferences at the top.

What to watch

The most significant near-term test will be the FDA’s handling of pharmaceutical approvals in the absence of several key advisory panels. If approval timelines lengthen — which is the counterintuitive but historically documented result of removing external peer review — the administration will face an awkward choice between acknowledging the costs of its decision and doubling down. Watch also for whether Canada’s recruitment of American scientists produces a measurable research output gap within the next two years: citation rates, patent applications, and clinical trial initiations are the leading indicators of research ecosystem health. And watch Congress: Republican appropriators have historically been more protective of federal scientific infrastructure than the executive branch, and a bipartisan push to restore some advisory functions through legislative mandate is not implausible in the medium term.

— J