Measuring diversity: law faculties in 1997 and 2013.

• Author: Lindgren, James
• Position: IV. Race, Gender, Religion, and Party B. Overrepresented Groups Compared to the U.S. Full-Time Working Population and the General Population through VI. Afterword: Law Faculties in 2013, with footnotes and tables, p. 116-151 - Thirty-Fourth Annual Federalist Society National Student Symposium

2. Overrepresented Groups Compared to the U.S. Full-Time Working Population and the General Population

   Proportional representation is a zero-sum game. If someone is on the bottom, someone else must be on the top. Which groups are the most overrepresented in law teaching? Table 6 shows the forty significantly overrepresented groups in law teaching in absolute and relative terms. At the top of the list are white Democrats, the excess accounting for two-fifths of a faculty (40%). All of the thirty-four most overrepresented groups are white (or predominately white), most are Democratic, and most are either Jewish or nonreligious.

   If Table 6 were sorted according to a ratio of percentages, (69) all of the twenty-eight most overrepresented groups would be Jewish or nonreligious. Nearly half of these twenty-eight are also white and Democratic. The most overrepresented groups not defined by religion are white male Democrats, white Democrats, male Democrats, Democrats, and white female Democrats. By ratios, the most overrepresented group is white male Jewish Democrats. They are overrepresented by a ratio of nearly 28 to 1. Yet Jews were a traditionally locked-out group. In the 1930s, many law schools had no Jews or had a quota of one or two. (70) If one were serious about proportional representation, then one would wait to hire more white male Jewish Democratic law professors until 97% of them died out or resigned, perhaps taking as long as thirty years. Just to state this possibility shows that the proportional representation notion of diversity, taken seriously, is profoundly anti-Semitic in effect, if not in intent.

   Thus, the most overrepresented groups tend to be white or mostly white, just as the most underrepresented groups tend to be white or mostly white. The overrepresented groups tend to be Jewish or nonreligious; the underrepresented groups tend to be Republican, Independent, and Christian.

   One sees the same patterns of overrepresentation in Table 7, comparing law professors to the U.S. non-institutionalized general population ages 30-75. Ranked by percentage differences, the most overrepresented group is white Democrats, followed by Democrats. Ranked by ratios, all of the twenty-eight most overrepresented groups are Jewish or nonreligious. (71)

3. The Underrepresented and Overrepresented Groups Compared to Lawyers

   Evaluations of employment discrimination usually use the pool of qualified eligible workers, not the general population, as the proper comparison group. (72) Although the best comparison might be the makeup of the pool of those lawyers with top credentials over the last 40 years, information about this elite group is unavailable. All we have is information about those minimally qualified--lawyers, judges, and law professors in private or government service. In Tables 8 and 9 I compare the law professor population to the population of private lawyers, public lawyers, judges, and law professors of age 30-75. Current Population Survey data were used for race and gender, but the sample size of lawyers was only 413. (73) Even worse, the 1972-94 General Social Survey had only 129 lawyers aged 30-75. Thus, the estimates for lawyers, particularly their religion and party identification, are limited by the small samples. For this reason, the conclusions that one may legitimately draw from these data are tentative.

   Yet even by this excessively broad construction of the "pool," women and most minorities are either at parity or overrepresented in law teaching. Since on average African Americans receive lower grades in law school, (74) one would expect them to be underrepresented in law teaching compared to the lawyer population, even if there were no discrimination in hiring. Yet the data in Tables 8 and 9 show that most minorities and Democratic women are overrepresented in law teaching compared to lawyers more generally. The most obvious explanation--though not the only one--is that affirmative action is strong enough to overcome discrimination in the hiring decision itself.

   When one compares Table 8 with earlier Tables, some interesting patterns emerge. The race, party, and religion patterns are the same. Republicans and Christians are still underrepresented, and all of the most underrepresented groups are white or predominately white. However, the gender pattern is opposite. Compared to the lawyer population, more of the underrepresented groups in law teaching are male (eleven groups) than female (seven groups).

   If Table 8 were sorted to rank differences between law professors and lawyers by ratios, (75) ethnic minorities would still not be the most underrepresented groups. The top ten spots would be held by groups that are Republican and white (or predominantly white), as well as mostly female. Minorities would hold only two of the thirty-four spots on the list. For example, Latinas are 2.4% of lawyers but only 0.8% of law professors.

   When one examines the most overrepresented groups compared to lawyers in Table 9, the pattern is somewhat similar to that for the working population. The twenty most overrepresented groups are mostly white, Democratic, and Jewish or nonreligious--as well as more frequently male than female. If Table 9 were sorted by ratios, (76) one would see that three of the seven most overrepresented groups in law teaching are minorities. African Americans, for example, are overrepresented at 3.6 times their percentages in the lawyer population. (77) African-American females are significantly overrepresented at 4.6 times their percentages in the lawyer population. Further, the largest group recruited as diversity hires in the 1980s and 1990s, white female Democrats, is significantly overrepresented (21% of law professors versus only 9% of lawyers).

   These findings are consistent with the idea that, on balance, significant affirmative action has taken place for some groups, since they already more than reflect their numbers in the lawyer population. Other data, however, are better for determining how much affirmative action is occurring. Some law professors have mentioned the supposed increasing political diversity of law faculties, though their frame for comparison may have been the 1950s. (78)

4. The Core Diversity Groups and the Core Overrepresented Groups

   In the legal academy, whether a group or subgroup is underrepresented often turns on which comparison population is used as the benchmark. But some groups are significantly underrepresented, no matter which comparison population is used. These twenty-nine core diversity groups are those that are significantly underrepresented compared to the U.S. full-time working population, the U.S. non-institutionalized population, and the lawyer population.

   The last two columns in Table 10 are simple means of the percentages underrepresented and the ratios of representation for law professors compared to general, full-time working, and lawyer populations. As Table 10 shows, most of the core diversity groups are Republican and Christian, and all but one are white or mostly white. The lone exception is Latinas, who are significantly underrepresented compared to all three relevant populations.

   Sometimes a lack of diversity in a workforce is attributed to the "tyranny of the majority." Under this view, the largest groups in society exercise their power to ensure that they are overrepresented in good jobs, leading to the underrepresentation of minorities. Yet some of the core underrepresented groups are among the largest in the general population--Protestants (61.8%), white Protestants (48.5%), Republicans (38.6%), and white Republicans (34.5%).

   Table 11 presents the twenty-three core overrepresented groups, those that are significantly overrepresented across the board. The last two columns in Table 11 are simple means of the percentages overrepresented and the ratios of representation for law professors compared to general, fulltime working, and lawyer populations. What is most striking is that all of the twenty-three core overrepresented groups are white or mostly white, all but five are Jewish or have no religion, and a majority are Democratic.

   5. DISCUSSION AND POSSIBLE NORMATIVE IMPLICATIONS

      In this Article, I merely hint at the normative implications of these data. In part, my purpose is to show that sometimes facts are normatively stronger arguments than normative arguments themselves. (79) If instead of doing this study, I were to argue that affirmative action is a more coherent policy than diversity as proportional representation, I suspect that I would persuade almost no one. But by showing the implications of diversity--that in diversity hiring we should prefer white Protestants and white Republicans to African Americans--I show that diversity as proportional representation is not what many people really want.

      This study is consistent with the conclusion that affirmative action overcomes discrimination in the hiring process because African Americans are overrepresented in law teaching compared to the legal profession generally: all lawyers (in public and private practice), judges, and law professors. (80) Further, this comports with other research suggesting that affirmative action exists in faculty hiring. (81) For example, Deborah Jones Merritt and Barbara Reskin report that white women and minority men fare better than white men in getting jobs at elite institutions. However, they also report that minority women fare as badly as white men. (82) They also show that in the years 1986-91, women and minorities joined the tenure track at rates higher than their graduation rates in then-recent law school classes. (83)

      There is a widespread assumption that but for discrimination, the distribution of jobs in American society would be proportional. (84) Indeed, researchers often describe other patterns of job distribution as indicating "disparate impact," as if the hiring decision were the cause of differences in job distributions...