social-politics education bullshit

Liberal Bias

"He who knows only his side of the case, knows little of that."

John Stuart Mill

[ This is inspired (and to some extent based on) a comment by Scott Siskind Before You Get Too Excited About That Github Study.... He has since deleted the comment, but it can still be found on archive.today. It was also copied to reddit, where it was also deleted, but (again) can still be found on archive.today. This page overstates my conviction in the thesis: that researchers and the media are biased towards liberal ideas. I'm personally quite liberal on most issues. ]

Faculty are Liberal

From The social and political views of American professors we see that among full-time college faculty,

  1. 44% of respondents are liberal, 47% are moderate, and 9% are conservative. That is, liberals outnumber conservatives nearly 5-to-1.
  2. In 1972, those numbers were 46%, 27%, and 28%, respectively. This suggests the main shift has been from conservative to moderate rather than from moderate to liberal.
  3. 58% social sciences faculty are liberal.
  4. 57% of faculty are liberal at "elite" schools compared to 37% at community colleges.

If we switch to "Democrat", "Independent, and "Republican", we find these tendencies depend a lot on the academic discipline. For instance, economics is fairly balanced, with 34% Democrat and 29% Republican. Sociology strongly skews left with 49% Democrat and 6% Republican. Psychology skews even more strongly with 78% Democrat and 7% Republican.

In other words, faculty tend to be liberal, especially in fields relevant to politics and policy. This tendency has been getting stronger over time and is especially strong at elite research universities - the ones with the greatest impact on social discourse and government policy.

The "liberal explanation" is that, as Stephan Colbert said "reality has a well known liberal bias" and so those who must study fields relevant to policy learn that liberal policies are best. Unfortunately, its hard to systematically evaluate this claim.

The conservative explanation is that liberal people are much more likely to go into academia in the first place, which is true Political liberalism and graduate school attendance: A longitudinal analysis and they then do biased research that justifies their pre-existing beliefs, which as we shall soon see is also true to at least some extent.

Both the liberal and conservative explanations are almost certainly true, and the real question is one of extent.

Ideology Affects Research Conclusions

A meta-analysis of personality differences between races found evidence of a pro-black publishing bias Tate.

A meta-analysis of economists finds that different sub-fields have different average political leanings and that even within these sub-fields political leanings are associated with statistical estimates Jelveh:

Using our estimates from column 1 row 2 of table 8, moving from the most left wing to the most right wing within this sample would change the elasticity by .35 points, changing the optimal top tax rate from 84% to 58%. Extrapolating to the most liberal ideology of -1 to the most conservative ideology of 1, we end up with optimal tax rates from 96% to 52%. While 52% is still a high tax rate (resulting from the small elasticities uniformly found in the literature, even by conservatives), this result shows that same standard optimal taxation formula may yield quite different prescriptions depending on the ideology of the researcher producing the elasticity.

A psychology study submitted the same paper with just words and numbers swapped to flip the conclusion from pro-liberal to pro-conservative and found that reviewer-ideology significantly affected how publishable the referee rated the paper Abramowitz.

More generally researcher expectations skew research results in a wide variety of psychology studies Rosenthal, which is in line with the general idea that a researcher's preconceived expectations skews their results.

The political parties legal scholars donate to correlates with the bias of their writings Chilton (meanwhile most are liberal McGinnis).

And finally, remember that there is a large replication crisis, which suggests researchers aren't idealized truth-seekers limited only by funding, but that large bias (political or otherwise) exists in research De Menard.

Bias Against Conservatives

[ More generally, see also Compromising scholarship: Religious and political bias in American higher education and Recalibrating academic bias. ]

Law schools appear to discriminate against conservative professors Phillips and similar evidence exists for professors at colleges and universities Rothman (Democrats will rightly point out that correlation ≠ causation Ames and then promptly forget this when its inconvenient for them, but c'est la vie).

29% of sociologists would disfavor hiring a Republican; 39% and 41% would be "less likely to hire" evangelical Christians or fundamentalists Compromising scholarship: Religious and political bias in American higher education, which generally supports the claim that the bias is strongest against cultural/social conservatives rather than economic ones Recalibrating academic bias.

Evidence of political bias in college grading is mixed Fair and balanced POLITICAL OUTCOME BIAS IN GRADING.

Wikipedia

Wikipedia (a) is likely liberally biased Greenstein Tzuka and (b) influences the scientific literature Thompson.

The Media is Liberal

News outlets are generally more liberal than members of congress Groseclose.

The people working there are also skew left quite hard, with Democrat journalists outnumbering Republicans ones more than 7 to 1. Similar discrepancies are seen in other media too: with editors, screen writers, and actors all skew 10+ to 1 towards the Democrats Democratic vs. Republican occupations.

Alexander, S. (2016). Before You Get Too Excited About That Github Study.... Slate Star Codex. https://slatestarcodex.com/2016/02/12/before-you-get-too-excited-about-that-github-study/ Jelveh, Z., Kogut, B., & Naidu, S. (2018). Political language in economics. Columbia Business School Research Paper, (14-57). https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2535453 Tate, B. W., & Mcdaniel, M. A. (2008). Race differences in personality: An evaluation of moderators and publication bias. http://www.people.vcu.edu/~mamcdani/Publications/AOM_as_of_8%201%202008.pdf Gross, N., & Simmons, S. (2007, October). The social and political views of American professors. In Working Paper presented at a Harvard University Symposium on Professors and Their Politics. http://www.conservativecriminology.com/uploads/5/6/1/7/56173731/lounsbery_9-25.pdf Fosse, E., Freese, J., & Gross, N. (2011). Political liberalism and graduate school attendance: A longitudinal analysis. Unpublished manuscript. https://scholar.harvard.edu/ethanfosse/publications/political-liberalism-and-graduate-school-attendance-longitudinal-analysis Groseclose, T., & Milyo, J. (2005). A measure of media bias. The Quarterly Journal of Economics, 120(4), 1191-1237. https://doi.org/10.1162/003355305775097542 Abramowitz, S. I., Gomes, B., & Abramowitz, C. V. (1975). Publish or politic: Referee bias in manuscript review 1. Journal of Applied Social Psychology, 5(3), 187-200. https://doi.org/10.1111/j.1559-1816.1975.tb00675.x Rosenthal, R. (1994). Interpersonal expectancy effects: A 30-year perspective. Current directions in psychological science, 3(6), 176-179. https://doi.org/10.1111/1467-8721.ep10770698 Democratic vs. Republican occupations. Verdant Labs. http://verdantlabs.com/politics_of_professions/ Chilton, A. S., & Posner, E. A. (2015). An Empirical Study of Political Bias in Legal Scholarship. The Journal of Legal Studies, 44(2), 277-314. http://doi.org/10.1086/684302 Phillips, J. C. (2018). Political Discrimination and Law Professor Hiring. NYUJL & Liberty, 12, 560. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3224508 McGinnis, J. O., Schwartz, M. A., & Tisdell, B. (2004). The Patterns and Implications of Political Contributions by Elite Law School Faculty. Geo. LJ, 93, 1167. Yancey, G. (2011). Compromising scholarship: Religious and political bias in American higher education. Baylor University Press. http://isbn.nu/160258477X Yancey, G. (2012). Recalibrating academic bias. Academic Questions, 25(2), 267-278. http://doi.org/10.1007/s12129-012-9282-y Fosse, E., Gross, N., & Ma, J. (2011). Political bias in the graduate admissions process: A field experiment. Working Paper, Harvard University. https://scholar.harvard.edu/ethanfosse/publications/political-bias-graduate-admissions-process-field-experiment Musgrave, P., & Rom, M. (2015). Fair and balanced? Experimental evidence on partisan bias in grading. American Politics Research, 43(3), 536-554. https://doi.org/10.1177/1532673X14561655 Rom, M. C., & Musgrave, P. POLITICAL OUTCOME BIAS IN GRADING. https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.708.2014&rep=rep1&type=pdf Rothman, S., Lichter, S. R., & Nevitte, N. (2005, March). Politics and professional advancement among college faculty. In The Forum (Vol. 3, No. 1, pp. 1-16). De Gruyter. https://studentsforacademicfreedom.org/wp-content/uploads/2020/05/Politics_Faculty.pdf Ames, B., Barker, D. C., Bonneau, C. W., & Carman, C. J. (2005). Hide the Republicans, the Christians, and the Women: A Response to. In The Forum (Vol. 3, No. 2). https://www.pitt.edu/~cwb7/assets/papers/Forum%2005%20article.pdf De Menard, A. (2020). What's Wrong with Social Science and How to Fix It: Reflections After Reading 2578 Papers. Fantastic Anachronism. https://fantasticanachronism.com/2020/09/11/whats-wrong-with-social-science-and-how-to-fix-it/# Thompson, N., & Hanley, D. (2018). Science Is Shaped by Wikipedia: Evidence From a Randomized Control Trial. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3039505 Greenstein, S., & Zhu, F. (2018). Do experts or crowd-based models produce more bias? Evidence from Encyclopædia Britannica and Wikipedia. Mis Quarterly. https://doi.org/10.25300/MISQ/2018/14084 Tzuka, S. Ashtear, L. A. (2020). The left-wing bias of Wikipedia: Is Wikipedia’s neutral point of view truly dead?. The Critic. https://thecritic.co.uk/the-left-wing-bias-of-wikipedia/ Cofnas, N., & Carl, N. (2018). Does activism in social science explain conservatives’ distrust of scientists?. The American Sociologist, 49(1), 135-148. https://doi.org/10.1007/s12108-017-9362-0 Haidt, J., Jussim, L., & Martin, C. The Problem. Heterodox Academy. https://web.archive.org/web/20180126015622/https://heterodoxacademy.org/problems/