A statistical approach to law school citation rankings

Citation rankings have emerged as a popular approach to ranking the scholarly impact of law faculties. This paper develops a statistical approach for inferring faculty quality from citation counts and determining when differences among law schools are significant. Statistical tests demonstrate that the distribution of citations within faculties closely follows the lognormal distribution, subject to small adjustments. This suggests a simple test for comparing faculties: whether they could be drawn from lognormal distributions with the same log mean. Under this approach, the geometric mean of citations is the most efficient measure for summarizing faculty quality. Using citation data collected from HeinOnline, this article provides a citation ranking for 195 law schools in the United States. Most differences between peer schools are statistically insignificant, and confidence intervals on citation ranks are extremely wide. Except for the highest-ranked faculties, citation rankings provide little information on the relative quality of faculties.