James Lindgren (Northwestern University) has posted Ideology and Intelligence in Legal Scholarship (forthcoming Journal of Empirical Legal Scholarship, 2011) on SSRN. Here is the abstract:
-
A variety of commentators have suggested that there is positive correlation between the ideological valence of legal scholarship and the intellectual capacity of the author. In their most common form, these suggestions posit an association between "progressive" or "left leaning" political views, and "IQ" or "intelligence." Recent developments in content analysis and the measurement of inellectual capacity now permit empirical testing of these claims. In particular, sophisticated content analysis techniques enable researchers to provide an reliable intellectual capacity estimate based on the measurable characteristics of texts; these characteristics include sentence complexity, logical paragraph structure, and extensiveness of vocabulary.
-
For this study, 200 law review articles were randomly selected from a pool of 30 student-edited law reviews in the WESTLAW JLR database. Content-analysis software was used to generate Siegel-Spaeth ideology scores (normally used to estimate the ideological position of judges), and the Retiel technique was used to provide an estimate Stanford-Benet score for the author of each article. Standard techniques were applied to determine the statistical significance of the resulting correlations.
-
The results do not confirm the standard hypothesis. Siegel-Spaeth ideology scores that indicate "progressive" ideology were negatively correlated with the Stanford-Benet estimates. The Pearson product-moment correlation coefficient (-0.61) provides strong evidence of a negative relationship between progressive ideology and intelligence. This Pearson's R value included a small group of anomolous data points. 7 of the 200 articles had outlying Siegel-Spaeth scores; in layman's terms, these articles had scores that indicated extreme left-wing political ideology; these same articles had very high mean Stanford-Benet scores. If these outliers are excluded from the sample, the Pearson's R value decreases from -0.61 to -0.89 (a result that approaches linearity).
This is sure to be controversial. Highly recommended.
