The Consensus Trap: How Manufactured Agreement Corrupts Policy

This working paper, published by the Hoover Institution and co-authored by Hoover Institution Edward Teller Senior Fellow and NCEA advisor Steven E. Koonin, Hoover Institution Senior Fellow Scott W. Atlas, and NCEA advisor Terrence Keeley, examines a common pattern behind three of the most consequential policy failures of the past two decades: the 2008 financial crisis, the management of the COVID-19 pandemic, and the campaign for rapid decarbonization.
The study challenges the assumption that broad institutional agreement reflects settled evidence. The authors argue that in each case, powerful institutions and leaders aligned to manufacture a consensus the evidence did not support. They call this shared pathology “the consensus trap.”
Read the full paper at the Hoover Institution
Key findings include:
- Each failure followed the same four-stage course: a consensus was asserted, credible dissent was suppressed, the policy failed, and no one was held accountable
- The underlying mechanism was strikingly similar across all three cases: quick acceptance of flawed predictive models, aggressive discrediting of credible challenges, and amplification through the media and political leaders
- The costs of these failures fell regressively, landing hardest on those least able to bear them
- The consensus trap is likely to recur wherever complex data and high policy stakes meet, though institutional and societal reforms can make it less likely
- A 2025 revision of the Federal Judicial Center’s Reference Manual on Scientific Evidence ties “widespread acceptance” to the highest level of scientific certainty, a link that could embed the confusion between manufactured and earned consensus in the federal courts
The paper concludes that genuine scientific consensus must be earned through open, critical debate of the evidence, not declared by institutions, and that recognizing the consensus trap is essential to keeping it out of future policy and the courts.
Read the full paper at the Hoover Institution


.webp)







.avif)





