A large LLM‑based content analysis of ~600,000 abstracts shows the social sciences moved sharply left in the 1960s, stabilized in the 1970s–80s, rose again after 1990, and surged in the mid‑2010s. About 90% of politically relevant articles lean left across 1960–2024, with policy‑proximal fields (economics, political science) less extreme than feeling‑focused fields (sociology, gender studies).
— If academic research is systematically ideologically skewed, it affects public policy, trust in expertise, and how evidence is interpreted in politics and culture.
Steve Sailer
2026.03.26
100% relevant
James Manzi’s study (Theory and Society) applying 2025‑anchored LLM coding to ~600,000 abstracts and reporting a ~90% left lean and the three‑phase timeline (1960s, post‑1990 rise, mid‑2010s explosion).
← Back to All Ideas