Over 90 Percent of All Social Science Studies Are Leftist

James Manzi, a doctoral candidate in sociology at the University of Oxford, analyzed nearly 600,000 English-language abstracts from social science publications spanning from 1960 to 2024, according to Nius. His findings, published in the academic journal “Theory and Society,” paint a stark picture of ideological homogeneity in academia.

Manzi employed a large language model artificial intelligence system, training it on texts from American politicians, media outlets, and think tanks. He developed an ideological scale ranging from 0 to 10 to differentiate between conservative and progressive positions. Conservative sources such as Ron DeSantis, the National Review, and the Heritage Foundation anchored the value of 3, while progressive sources including Elizabeth Warren, the New York Times, and the Center for American Progress were assigned a value of 7. The AI then evaluated the academic abstracts along this defined scale.

Overwhelming Leftward Bias Across All Disciplines

The results were unambiguous. Approximately thirty percent of the abstracts were classified as politically relevant by the AI system. Of these, ninety percent fell within the left spectrum, with only five percent categorized as right-leaning. Every single one of the eleven disciplines examined showed a leftward tendency.

Economics demonstrated the weakest leftward pull with an average score of 5.7. Sociology registered significantly higher at 6.9, while Gender Studies topped the chart at 7.6. The analysis also revealed that this tendency has intensified over the decades, with only policy-adjacent fields such as criminology and political science showing brief countermovements during the 1970s and 1980s.

The most left-leaning disciplines also proved to be the most ideologically homogeneous, displaying the smallest variation in political viewpoints.

Survey Data Confirms the Pattern

While Manzi acknowledged limitations in his methodology, including the inability to trace precisely why the AI positioned certain works at specific points on the political scale and the absence of human reviewers to verify results, his findings align with previous research. The Higher Education Research Institute at the University of California, Los Angeles, has documented a clear majority of left and left-liberal academics in its Faculty Surveys over decades, with a ratio of approximately six to one compared to conservatives. In humanities and social sciences, the disparity is even greater.

Political Homogeneity Distorts Research Quality

The consequences of such political uniformity in academia are not merely theoretical. Psychologists Anthony N. Washburn and Linda J. Skitka demonstrated in an experiment that both conservatives and left-liberals were significantly more willing to undertake rigorous analysis when superficial results contradicted their preexisting assumptions. Their paper, published in 2018 under the title “Science Denial Across the Political Divide,” revealed motivated reasoning affects researchers across the political spectrum.

Migration researcher Alexander Kustov reported from his own discipline that findings which cannot readily be interpreted as arguments for positive effects of immigration are quietly downplayed or omitted from publications. Sociologist Neil Gross described a self-reinforcing mechanism whereby left-leaning disciplines primarily attract left-leaning recruits, making these fields progressively more homogeneous over time.

The Oxford study raises fundamental questions about the integrity of social science research and its ability to provide balanced, objective analysis on politically contentious topics. With such overwhelming ideological uniformity, the capacity of academia to challenge prevailing assumptions or provide genuine intellectual diversity appears severely compromised.

With information from Nius