Reflections on class in American legal education

Denver University Law Review

Volumn 88, Issue 4, 2011, Pages 683-717

  Lempert, Richard ^a  

^a UNIVERSITY OF MICHIGAN   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 84866441485     PISSN: 08839409     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Conference Paper

References (212)

1. Richard H. Sander, Class in American Legal Education, 88 Denv. U. L. Rev. 631 (2011).
2. ABRAHAM FLEXNER, MEDICAL EDUCATION IN THE UNITED STATES AND CANADA: A REPORT TO THE CARNEGIE FOUNDATION FOR THE ADVANCEMENT OF TEACHING (1910).
3. The Carnegie Foundation supplied the law school world with its own version of the Flexner Report, the Reed Report, but its impact was quite different from that of the Carnegie critique of medical education, which in less than a decade resulted in more than half the nation's medical schools closing, the consolidation of medical education within universities, and the disappearance of proprietary medical education. This may have been because Reed, to the disappointment of the law's professional establishment, saw law as two-tiered rather than unitary profession and supported rather than called for the termination of part-time legal education.
4. ALFRED ZANTZINGER REED, TRAINING FOR THE PUBLIC PROFESSION OF THE LAW: HISTORICAL DEVELOPMENT AND PRINCIPAL CONTEMPORARY PROBLEMS OF LEGAL EDUCATION IN THE UNITED STATES WITH SOME ACCOUNT OF CONDITIONS IN ENGLAND AND CANADA 55-56 (1921).
5. Moreover, the timing of Reed's report undercut the more Flexner-like recommendations of an American Bar Association Committee chaired by Elihu Root, a distinguished corporate lawyer, diplomat and Nobel Peace Prize Winner.
6. ELIHU ROOT, REPORT OF THE SPECIAL COMMITTEE TO THE SECTION OF LEGAL EDUCATORS AND ADMISSIONS TO THE BAR OF THE AMERICAN BAR ASSOCIATION (1921). Training for the bar through apprenticeships did, however, almost entirely disappear during the 20th century.
7. JEROME E. CARLIN, LAWYERS ON THEIR OWN: THE SOLO PRACTITIONER IN AN URBAN SETTING 168 (1994).
8. These graduates were overwhelmingly male.
9. Judith S. Kaye, Women Chiefs: Shaping the Third Branch, 36 U. TOL. L. REV. 899, 899 (2005) (noting that before and during the 1960s, women's presence in law schools was miniscule).
10. David Segal, Is Law School a Losing Game?, N.Y. TIMES, January 9, 2011, at BU1, available at http://www.nytimes.com/2011/01/09/business/09law.html?src= me&ref=general.
11. Those who graduate near the bottom of higher prestige law schools are also reporting difficulties in finding law jobs, though there have been reports that some law schools are serving as employers of last resort for hard to place students until the time for reporting law graduate employment rates to U.S. News has passed.
12. PIERRE BOURDIEU & JEAN-CLAUDE PASSERON, REPRODUCTION: IN EDUCATION, SOCIETY AND CULTURE 160, 209 (Richard Nice trans., 1977).
13. In addition, a student's identity may affect how his or contribution to a discussion is received.
14. The same anti-war statement will be received differently if it comes from an Iraqi war veteran rather than an anti-war activist, and the impact of a statement questioning or supporting racebased affirmative action will similarly differ depending on whether the speaker is black or white.
15. These conclusions have not to my knowledge been the subject of rigorous empirical examination, although anecdotal evidence, such as the plethora of journals reflecting gender or ethnic themes and reports by professors of how their classrooms have been affected by diversity, can be easily found.
16. A study of the University of Michigan Law School alumni found that lawyers tended disproportionately to serve clients of their own ethnic group and that Michigan graduates did substantial pro bono work and often occupied leadership positions in communal or political settings, with minority law school graduates being more involved in pro bono and leadership activities than whites.
17. Richard O. Lempert, David L. Chambers & Terry K. Adams, Michigan's Minority Graduates in Practice: The River Runs Through Law School, 25 LAW & SOC. INQUIRY 395, 436-37, 440, 453-58 (2000).
18. Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 301-02 (1978)
19. Stating that preferential classifications have never been approved absent a determination of past discrimination.
20. Although the Court has not directly addressed the issue, it is likely that the Civil Rights Act will be read to prevent private schools from engaging in affirmative action programs that would be unconstitutional if done by state schools.
21. Grutter v. Bollinger, 539 U.S. 306 (2003)
22. Reaffirming Bakke, might be read as opening the door to the last of these justifications since Justice O'Connor, who wrote the opinion, specifically mentioned the special value that minority lawyers have for the military and business.
23. However, as Professor Sander notes, the continued constitutionality of educational affirmative action is itself precarious.
24. This is an issue on which the pre-existing views of the Justices seem more influential than any legal or empirical arguments lawyers might bring to bear.
25. Four of the current Justices would most likely welcome the opportunity to ban affirmative action while a fifth, Justice Kennedy, has a position whose nuances are hard to discern.
26. Except for the bluest of bloods everyone may, of course, rise in social status, but for those who come from well-educated, well-off families a professional education does not without much more mean a rise in social status
27. SES scores may also mislead when they place rural and urban families in the same social class.
28. For example, assume a school had 200 applicants, 40 of whom came from families whose SES scores placed them in the bottom quarter of all American families and that of these 40, ten came from families that were intergenerationally lower class while 30 were immigrants' children whose families, despite occupying low status and low paying jobs in the United States, had backgrounds of privilege quite distinct from the backgrounds of families we think of as lower class.
29. If the school admitted 100 students, 10 of whom were immigrants children and 6 of whom were intergenerationally lower class, it would appear from the SES measure that 16% of the entering class came from lower class origins when, I would argue, only 6% did.
30. At the same time, relative to their representation in the applicant pool, it would appear that only 37.5% of applicants with lower class origins were admitted compared to 52.5% of students from more advantaged backgrounds.
31. However, 60% of the intergenerationally lower class gained acceptance under this hypothetical scenario.
32. I am not saying these kinds of effects will occur, but the data are such that we cannot exclude this possibility.
33. Also missing is information on family wealth, which is not included in most studies that use SES as a variable not because it is conceptually unimportant but because reliable wealth data is hard to come by.
34. Indeed, conceptually wealth may be the most important indicator of a family's social class.
35. In its absence, occupation is generally taken to be the best single measure of social class among the SES index variables because occupational prestige is thought to do most to locate a person's position in the social hierarchy. I discuss why the absence of income and wealth data is especially unfortunate given all that Professor Sander seeks to accomplish in the text of note 23 infra.
36. Professor Sander reports that about a quarter of his sample cases lacked information on either three or all four of the indicators and so were excluded from his analysis.
37. Note also that the parental SES data are frozen at a point in time, but people's occupation and educational achievements change over time.
38. To the extent these SES variables reflect class differences by more than definition, a law student's class roots may be different than a current status report may make them appear.
39. A person working as a retail clerk may have been running a successful business until an economic down turn when his child was a junior in college, or a mother who worked as a teacher's aide most of her adult life may recently have completed a bachelor's degree and been hired as a regular teacher.
40. Other categories that are almost empty in the usable sample of almost 3000 are cases where we know both parents' occupations and one parent's education
41. Cases where we know one parent's occupation and the other parent's education
42. Particularly puzzling is that although there were 375 respondents who chose to report only their father's and mother's education or about half the number who reported only their two parents' occupations, there was no respondent who reported just one parent's education although there were 534 respondents who reported just one parents' occupation.
43. Something seems wrong here.
44. I am grateful to Professor Sander for providing me with the detailed breakdowns regarding variable availability that I report in this paragraph.
45. I should add that he recognizes in his paper the possible overstatement of black SES that could result from the absence of usable data from students raised in single parent households.
46. Responses were received from about 51% of those in the nationally representative sample and from about 43% of those in the minority oversample, where one might expect lower SES attorneys to be disproportionately represented.
47. Professor Sander recognizes this and notes that in the Census PUMS data an index reproducing his measure of SES correlates somewhere between .4 and .45 with household income.
48. Although this correlation is, no doubt, highly significant in the statistical sense, it does not denote a particularly close relationship or one which justifies dismissing concerns regarding the implications of the absence of income information for the validity of operationalized SES.
49. A correlation between the study's SES index and household income of between .4 and .45 means that the SES index explains only sixteen to twenty percent of the variance in household incomes in Professor Sander's PUMS subsample.
50. Moreover, this may be an overestimate of the correlation in the AJD data because the PUMS on spousal occupation and education is likely to have been more complete than it is in the AJD sample.
51. GITA Z. WILDER, RACE AND ETHNICITY IN THE LEGAL PROFESSION: FINDINGS FROM THE FIRST WAVE OF THE AFTER THE JD STUDY 58 tbl.36 (2008), available at http://www.americanbarfoundation.org/uploads/cms/documents/race-and-ethnicity- monograph.pdf.
52. JOHN P. HEINZ & EDWARD O. LAUMANN, CHICAGO LAWYERS: THE SOCIAL STRUCTURE OF THE BAR (1982) (analyzing the social differentiation among different types of lawyers).
53. There are other problematic aspects but a detailed discussion would take us far afield and soon, I expect, pass the limits of my knowledge. Simply put CAMSIS coding is a novel way of assigning status values to occupations based on interaction patterns of people who, more often than not, have different occupations. Theoretically these patterns should be based on occupation pairs involving friends, neighbors and relatives, but often data constraints mean that the only pairs that can be created are those of husbands and wives.
54. Code creation becomes problematic when data sources include many couples where only one spouse is working or where there exist what are called "pseudo diagonals" (strong but misleading husband-wife associations as when a husband may be categorized as an agricultural proprietor and a wife as an agricultural laborer.) CAMSIS codes, which were originally established for occupations in the U.K., can and have been established for different countries as in the U.S. Census data based rankings that Professor Sander employs, but they can also be created for particular data sets.
55. Moreover, they are claimed by their creators to represent occupational status directly and not be limited to specifying the relative status of different occupations.
56. An additional complication exists in combining husband and wife occupations into a common index, since they are not on the same scale.
57. Thus those who adapted the scale to the 2000 census data which Professor Sander uses caution: [I]t is a little misleading, albeit a commonly made mistake, to analyse [sic] a mixed gender population through CAMSIS scores which are the male scale scores for the men and the female scale scores for the women. The occupational scale indexing used for men and women is invariably the same, further giving the impression of equivalent meanings.
58. However the CAMSIS methodology assumes different systems of relative positions prevail within the male and female occupational structures, and hence implicitly that equivalent titles are not necessarily the same between genders.
59. Accessing and Using CAMSIS Scale Scores, CAMSIS: SOCIAL INTERACTION AND STRATIFICATION SCALE, http://www.camsis.stir.ac.uk/useofscores.html (last visited May 6, 2011) (emphasis in red in original).
60. Professor Sander converted CAMSIS codes by gender to percentiles as normalized against the 5% PUMS subsample, perhaps to deal with this issue, but using standard scores or the husband's score is the recommended procedure.
61. I am unclear how Professor Sander assigned his percentile scores or what the implications of his assignment across genders are.
62. He reports in a methodological appendix that for women he assigned a 99th percentile rankings to codes of 75 and above, yet in the CAMIS ranking data he kindly provided me, a CMASIS score of 75 for women seems to be at the 94th percentile.
63. For a useful summary of how CAMSIS scores are assigned and cautions in using them, see id.
64. Richard H. Sander, A Systemic Analysis of Affirmative Action in American Law Schools, 57 STAN. L. REV. 367, 436 (2004)
65. David L. Chambers, Timothy T. Clydesdale, William C. Kidder & Richard O. Lempert, The Real Impact of Eliminating Affirmative Action in American Law Schools: An Empirical Critique of Richard Sander's Study, 57 STAN. L. REV. 1855, 1886 n.111 (2005).
66. Not everyone who graduates law school takes the bar exam. Some move directly into positions, like teaching or business management, where they can take advantage of their legal education without having to qualify for legal practice.
67. Yakowitz estimates that 150,000 people have taken the bar and never passed, but her estimate is admittedly crude.
68. Jane Yakowitz, Marooned: An Empirical Investigation of Law School Graduates Who Fail the Bar Exam, 60 J. LEGAL EDUC. 3, 15-17 (2010).
69. She also drops a gratuitous footnote calling into question data that I and two coauthors published, which indicated that 94% of University of Michigan Law School graduates (a claim based on respondents to a survey we conducted which we believe with little loss of accuracy can be generalized to the population of Michigan Law School graduates had passed at least one bar exam,), and suggests the truer figure is closer to 85%.
70. Richard O. Lempert, David L. Chambers & Terry K. Adams, Michigan's Minority Graduates in Practice: The River Runs Through Law School, 25 LAW & SOC. INQUIRY 395 (2000)).
71. I do not know how Yakowitz arrived at her number, but it is wrong. Reports by state bars back to Michigan for the period 2006-2008 indicate that the overall bar passage rates of Michigan students during these years was a bit more than 96%, and this is an underestimate since some bar takers were on their second or third attempt.
72. Moreover, bar passage standards have tightened since the years when we conducted our study. I expect Yakowitz may have been misled by Michigan's bar passage rates in California, reputedly the nation's most difficult bar.
73. These rates during the period for which I saw official data are closer to her estimate, but only a small fraction of Michigan's graduates ever attempt the California Bar.
74. I also know there were some earlier years when Michigan's California bar passage rates were at or near the best of any law school, no doubt a function of the students who took jobs in California.
75. RONIT DINOVITZER ET AL., AFTER THE J.D.: FIRST RESULTS OF A NATIONAL STUDY OF LEGAL CAREERS 90 (Janet E. Smith et al., eds., 2004).
76. To get the most accurate estimations the sample should be adjusted with weights the project provides.
77. Professor Sander's data is unweighted, and while this might introduce a bit more noise, it appears that using unweighted data changes little.
78. GABRIELE PLICKERT & RONIT DINOVITZER, AFTER THE J.D: FIRST RESULTS REPORT, TECHNICAL ADDENDUM 1 ( 2007), available at http://www. americanbarfoundation.org/uploads/cms/documents/ weighted-ajdreport-9.6.07.pdf.
79. PLICKERT & DINOVITZER, at 4
80. Indicating that the border state of Tennessee, Florida and the cities of Houston and Atlanta were chosen for inclusion.
81. Sarah Zearfoss, A Snapshot of the Entering Class: 5 Things About the New 1Ls, LAW QUADRANGLE, Fall 2010, at 1.
82. DENNIS GILBERT, THE AMERICAN CLASS STRUCTURE IN AN AGE OF GROWING INEQUALITY 13 (7th ed. 2008).
83. WILLIAM E. THOMPSON & JOSEPH V. HICKEY, SOCIETY IN FOCUS: AN INTRODUCTION TO SOCIOLOGY 216-17 (1994).
84. LEONARD BEEGHLEY, THE STRUCTURE OF SOCIAL STRATIFICATION IN THE UNITED STATES 239 (4th ed. 2005).
85. GORDON MARSHALL, IN PRAISE OF SOCIOLOGY 16 (1990).
86. Higher-grade professionals, administrators, and officials; managers in large industrial establishments; large proprietors
87. Lower-grade professionals, administrators, and officials, higher-grade technicians; managers in small industrial establishments; supervisors of non-manual employees
88. Routine non-manual employees, higher grade (administration and commerce).
89. Routine nonmanual employees, lower grade (sales and services)
90. Small proprietors, artisans, etc., with employees
91. Small proprietors, artisans, etc., without employees
92. Farmers and smallholders; other self-employed workers in primary production
93. Lower-grade technicians; supervisors of manual workers
94. Skilled manual workers
95. Semi-skilled and unskilled manual workers (not in agriculture, etc.)
96. Agricultural and other workers in primary production.
97. Jack Metzgar, Politics and the American Class Vernacular, WORKINGUSA, June 2003.
98. Most class researchers would, however, acknowledge that any categorical breakdown of class is imperfect, and there can be considerable heterogeneity among those placed in the same social class by whatever coding scheme is used.
99. In Professor Sander's classification scheme, unlike some other possible schemes, it is obvious that those who place near the bottom of his third quartile are closer to those in the bottom quartile than they are to those at the top of their class quartile.
100. Although this information is lost when categorical classifications by quartile are used, it would not be lost in other forms of quantitative analysis.
101. Fact, there is no single best way to assign people to positions in a class structure.
102. Most appropriate assignment depends on the purpose behind the assignment.
103. If, for example, the classic Marxian view of class pertained, and class consciousness was defined by relationship to means of production, then heterogeneity within a class on dimensions like education and income would not matter if class consciousness was the key variable. Despite the heterogeneity on other dimensions, people who stood in the same relationship to means of production would share the same, common class consciousness. However, even for the classic Marxist theorist the world is seldom so neatly organized.
104. Professor Sander is far from alone in eliding this difference.
105. Researchers, including text book writers and top scholars, often treat SES as if it exhausted the meaning of social class, perhaps because SES more than class lends itself to continuous measurement and percentile distributions.
106. The CAMSIS coding methodology was designed to tighten the link between occupational position and social status by using associational information to create its operational prestige scores, reflecting the notion that class identities are reflected by and rooted in relationships.
107. As I described above, there exist limitations to the CAMSIS codes, including limitations that emerge when the only information on personal associations that is available is for husband-wife pairs, as well as the counterintuitive nature of a scale in which the status of an occupation can vary substantially depending on the gender of the holder. See supra note 31 and accompanying text. (This is not always problematic since occupational status, however measured, may be differently sorted by gender, but the differential sorting is also likely to be more than occasionally in error.)
108. Nevertheless, I regard Professor Sander's use of CAMSIS codes as one of a number of ways in which he has attempted to make the best of what, from a data quality/availability standpoint, is a bad situation.
109. Only one of five differences tested by Professor Sander would, by convention, be considered even marginally significant, but the finding of marginal significance must be discounted when there have been five independent tests.
110. Thus the best conclusion to draw from Sander's data is that, in the sample he examined, there is no substantial evidence of reliable differences in standardized index scores associated with parental background.
111. If the slight advantage suggested in the data for those whose parent possesses a professional or doctoral degree is real, it may reflect the ability of a student with a lawyer parent to get letters of recommendation from school alumni or other influentials, and what may be a greater likelihood among students from families with doctoral degrees to themselves pursue advanced academic degrees before applying to law school.
112. Sander draws his data from the book Lawyers in the Making by Seymour Warkov and Joseph Zelan published in 1965.
113. I could not acquire a copy of this book but have drawn on the data analysis that forms the basis for the book
114. SEYMOUR WARKOV, LAWYERS IN THE MAKING: THE 1961 ENTRANTS TO AMERICAN LAW SCHOOLS (1963), available at http://www.norc.uchicago.edu/NR/rdonlyres/7D0DE824- FD6E-46CA-B72D-AEFF1DEBB6F1/0/ NORCRpt-96.pdf.
115. This analysis may contain more tables than what were published so I do not know if the information that follows was available to Professor Sander.
116. Stuart Rojstaczer and Christopher Healy (whose article Professor Sander cites at his footnote 75) find reliable distinctions between average grade inflation in schools of different types (e.g. public-private), but the amount of inflation varies within school types, so that a satellite public school may have grades that are more inflated than those at an elite private college or flagship public university.
117. Stuart Rojstaczer & Christopher Healy, Grading in American Colleges and Universities, TCHRS. C. REC., Mar. 4, 2010, at 2-3.
118. Additionally, grade inflation varies by fields within universities and these differences will often be greater than grade inflation variation across college and university types.
119. Moreover, if this is not complex enough, it could be that within fields broadly defined grade inflation varies by major. For example, Rojstaczer and Healy's data indicate that grade inflation is greatest in the humanities where after controlling for likely student ability grades tend on average to be .4 higher than in the natural sciences and .2 higher than in the social sciences.
120. A Classics major in a particular school may, however, have grades that are less inflated than the grades of most other humanities majors and of some or all natural science majors depending on the school and its professors.
121. Admissions officers whom I have known have had a sense of the degree of grade inflation by school and by major, or at least of the law school relevant abilities that grades reflect.
122. Indeed, where admissions officers have dealt with numerous students from a handful of feeder schools over a sufficiently long period of time, some have developed a sense of grade inflation not just by school and by major but sometimes also by professor, along with a professorspecific sense of "letter of recommendation inflation," or, on occasion, deflation.
123. My recollection is that when I first started teaching, most law school financial aid, to the extent it existed at all, was need-based and the expectation of repayment was presented as a moral rather than a legal obligation.
124. Later competition for the most able minority students resulted in financial aid packages for the apparently most able that had a significant non-need component.
125. Still later, competition for students with the kinds of credentials that boosted U.S. News rankings led to a broadening of non-need-based awards to all students.
126. These shifts were also supported and perhaps fostered by the increasing availability of student loans, which meant that students willing to take on debt could attend law school even without scholarship aid.
127. At some elite schools, like my home school the University of Michigan, recognition that whether or not a loan was easily met depended on the career path a law student chose or was forced into, meant that some of what might have been scholarship aid was channeled to loan forgiveness programs that evaluated need for assistance as it in fact existed after graduation.
128. Originally, this was done to enable law school graduates to take relatively low paying public interest jobs. However, since the need for loan repayment assistance depends on income level and not employment type and from a reluctance to evaluate different careers by reference to their social value post law school, earned income became the key to postgraduation financial assistance. I believe much the same trajectory characterizes the history of financial support at many of the nation's wealthiest and most selective law schools.
129. There is also the question of whether we would be doing lower SES students any favors by encouraging more to attend law school.
130. Noting the financial difficulties faced by law students graduating with high debt and low job prospects).
131. Part of the excess was due to the fact that the school was still enrolling better off white students through the summer to keep its U.S. News rankings up.
132. The five black students were originally spread across UCLA's four first year sections.
133. The students, I have been told, came to the dean the day before classes began and asked to be placed in the same section since they felt that the pressures of being the only black in a sea of White and Asian faces would put too much pressure on them.
134. Their request was granted, meaning that one first year section had meaningful black representation and the other three sections had none.
135. I agree with several of the points that Professor Sander makes in comparing class-based and race-based affirmative action.
136. I regard as most likely true his suggestions that to attain similar "minority" representation, when the minority is those of low SES, preferences would not need to be as steep as they are when the preferred group is a racial or ethnic minority.
137. I use the hedging words most likely only because I am unsure of the implications of class-based differences in law school attendance and because much depends on how one defines class.
138. I don't think I would as readily assign a place in the lower class to the children of immigrants from many Asian and some other countries even if by the SES measures Professor Sander uses they are in the lower ranks.
139. If instead of this measure we were considering students from families with deeply sunk lower SES or class roots, like Appalachian whites or urban unskilled laborers
140. I am not at all certain that the preferences needed to admit a representative proportion to law schools of varying strata would be any less than they are for other minorities, and I would not be surprised if they were in fact greater.)
141. I also agree that in the current political climate class-based preferences would be better received by the public than race or ethnicity-based preferences and that since class is not a suspect classification, classbased affirmative action would, at least given the current Supreme Court, rest on firmer constitutional grounds.
142. I also agree that in today's increasingly multi-racial nation there are challenges in identifying who is Hispanic, Native American or black, for purposes of affirmative action, but I believe that class too has its ambiguities and problems of definition and that we are fooling ourselves if we think that crude SES measures define America's lower classes.
143. Where I completely part company from Professor Sander is on his claims of black disadvantage resulting from affirmative action by the nation's more elite law schools and in his suggestion that mismatch is at the core of the problem.
144. Professor Sander and I (and coauthors) along with others who have looked at a range of data have gone back and forth for some years now on these issues.
145. I remain convinced that Professor Sander's mismatch thesis is largely if not entirely unsupported and is, if anything, least applicable to the nation's most elite law schools.
146. I have no desire to rehearse this particular dispute here, but refer the reader to our mutual contributions to the debate and to works cited therein.
147. Richard H. Sander, A Reply to Critics, 57 STAN. L. REV. 1963 (2005)
148. Richard Lempert, William Kidder, Timothy T. Clydesdale & David L. Chambers, Affirmative Action in American Law Schools: A Critical Response to Richard Sander's - A Reply to Critics (Univ. of Mich. Law & Econ., Olin Working Paper No. 06-001, 2006), available at http://papers.ssrn.com/sol3/ papers.cfm?abstract-id=886382##.
149. Black and Hispanic law students are similarly underrepresented relative to their population proportions in almost all the nation's law schools, but a "critical mass" for diversity purposes has never been defined as a proportional presence equal to the group's population proportion.
150. There are actions schools could take which would have as an expected outcome an increase in lower SES representation, but they do not involve preferences, in the sense of admitting a student of lower SES ahead of a higher SES student with a higher LSAT/GPA index score.
151. Professor Sander has identified some of them. One example is eliminating legacy advantages.
152. A second, which would require funding at a level that I think no law school can afford but which a few wealthy undergraduate colleges manage: it is to run a needs blind admissions system and guarantee students that if admitted their needs will be met mainly with scholarship aid. A third is to broaden recruitment to include active outreach not just to a group of select feeder schools but to undergraduate schools which those from lower SES backgrounds are disproportionately likely to attend. I think each of these approaches has much to commend it, but each has financial costs that must be recognized.
153. At Michigan, for example, the list of such students could go on and on. It would include an Olympic gold medalist, a physician in his 50s who was a leader of the AMA, a concert pianist, a top chess player, and the like.
154. By this definition beneficiaries of affirmative action include in many schools people of a certain race or ethnic background, most commonly blacks, Hispanics and Native Americans, some legacy admits and in the case of public colleges and universities some in-state residents.
155. It would also not surprise me, although I have no data, if some religiously supported institutions give preferences to applicant's who share the institution's religious views.
156. The possibility of religiously based affirmative action has not been examined because the schools that may practice this are private, and data relating to their applicant and admissions pool is not available for study, and we don't speak of affirmative action in the case of legacy admits or in-state residents because the preferential admission of students in these categories has long been regarded as non-problematic.
157. This view is being challenged with respect to legacy admits.)
158. Also it is generally believed that preferences accorded legacies and in-state applicants are not as great as those enjoyed by the beneficiaries of race or ethnicity-based affirmative action.
159. This is, I am sure, true on the average, but there are cases where some legacies, or in-state students with powerful governmental backers, have enjoyed preferences as large as those enjoyed by most minority students.
160. Moreover, at Michigan which is the school I know best, the distance between the index credentials of the average in-state admit and those of the bottom 10th percentile non-resident admits has often, and perhaps over the past two decades always, been greater than the distance between the bottom 10th percentile resident admit and the average minority admit.
161. The open question is what does "to some degree" mean. I can't state this more precisely but I can say that I don't find the fact that the UCLA experiment which Professor Sander describes advantaged those defined as low SES applicants by the equivalent of about 40 LSAT index points troublesome.
162. This is particularly likely if the SES index, like the one Professor Sander uses in his study, omits information on family wealth and parental income.
163. The problem of overinclusiveness is common to all sorts of affirmative action but can only be determined with reference to specific justifications.
164. To take an innocuous example, affirmative action at state schools for state residents can be justified on the ground that the parents of such applicants have been paying state taxes in support of their colleges and universities for years in order to be able to better educate their children.
165. But an in state preference will be granted to a person who moved to the state eighteen months before applying to law school and who has paid little if any state taxes.
166. Alternatively the justification could be that the state needs a highly educated work force to prosper and state residents educated in state are more likely to remain residents after graduation than those who move to the state solely to get an education.
167. If only tax equity justified residency preferences including the recent mover in the applicant pool benefiting from residence-related affirmative action would be an example of overinclusion.
168. If the sole justification were the "stay and work" justification and if recent movers are as likely to remain residents after graduation as those raised in state, then including the recent mover in the group eligible for residency-based affirmative action would be consistent with the affirmative action justification.
169. Professor Sander at several points in is article suggests that race and ethnicity-based affirmative action programs suffer from serious problems of overinclusion.
170. He points out, for example, that minority admits at elite law schools have SES credential distributions that are relatively close to those of the average white
171. He also faults programs at schools like Harvard for treating as black for affirmative action purposes applicants from the West Indies or of West Indian parentage and students who may call themselves black but have one or more white grandparents.
172. With respect to the latter groups, I would argue that so long as society characterizes such students as black regardless of their personal histories or how they racially self- identify, the equity justification for affirmative action cannot be totally rejected, and the two other affirmative action justifications I discuss below remain.
173. The AJD data indicate that Asian students received 28% of their support while attending law school from their families.
174. Whites on average received only 19% of their support while in law school from their families and blacks only 9%. WILDER, supra note 25, at 59 tbl.37.
175. In making this point, I have focused on students of Asian heritage since they were the group predominantly benefited by the short-lived UCLA experiment, but not all whites who are in the country's lower SES ranks at the time they apply to law school come from families that have occupied these ranks all their lives. For example, while a student is in college one parent may have become unemployed and the other may have been laid off from a well-paying highly skilled position and been only able to find work as a low paid unskilled clerk.
176. Professor Sander's indicators, with income and perhaps family wealth added, might be the only reliable indicators of social class that a law school could acquire. Professor Sander also used census tract data in his UCLA experiment, but low average income census tracts may have pockets of better off residents.
177. One can still cite equity considerations to argue that students disadvantaged by their family's low SES status deserve a social mobility boost even if their parent's low SES does not closely relate to what one might regard as class-linked perspectives and experiences.
178. But refusing to entertain affirmative action as a mobility booster does not necessarily thwart social mobility; it simply extends and delays it.
179. Thus many Jews of my parents' generation went to law school, often at night, and became the lawyers who populated the lower ranks of the bar.
180. Despite their professional degrees many fared poorly in economic and other ways.
181. But their children were often able to attend better law schools or follow other entrepreneurial and professional paths that enabled them to move into careers that placed them well within the ranks of America's upper middle class.
182. To look at the first generation only, upward mobility attributable to professional training was for many, except to the extent mobility was defined by professional degrees, not much greater than that enjoyed by the children of push cart fathers who opened their own shops.
183. Over two generations, however, mobility was substantial and lower SES origins were left far behind.
184. I do not think there is one set of experience or viewpoints that characterizes all or even most members of a particular social class or of a racial or ethnic group for that matter.
185. Nor do people who fall into the same social class as defined by SES necessarily have the same interests much less political, religious or other preferences.
186. It is possible, however, to link statistically attitudes and experiences with class location and to find systematic differences between classes in how these are distributed.
187. I know of no good current data that would shed reliable empirical light on this matter.
188. The Warkov data discussed at note 58 supra suggests this supposition is reasonable.
189. I do not mean to suggest that all differences between lower class and more privileged law students will be wiped out by a shared elite education.
190. Commenting on an earlier version of this piece, Deborah Malamud pointed out that the family situation of the low-SES law student will not rapidly change and that a student's continued involvement with family may shape her attitudes and behavior.
191. I am sure she is right and that this example is not unique. Moreover, people differ.
192. I have no doubt that some students from lower class backgrounds with elite undergraduate educations make distinct contributions to a law school's education environment which students from more advantaged backgrounds could not or would not make.
193. Similarly, to qualify an argument below, there are no doubt students from low SES backgrounds who go on to high paying, high status careers but who because of their own backgrounds not only remain concerned with the situations of low SES individuals but also work to better their conditions. Nevertheless, I still maintain that with respect to diversity standpoint an elite law school is likely to get fewer benefits from admitting more lower SES students than one might expect and fewer benefits than those gained by ensuring the presence of a critical mass of minority law students.
194. I believe the same will be true of post-graduation societal benefits. Moreover, I would not be surprised if a large proportion of low SES students bring nothing in the way of an educational or societal diversity payoff.
195. I recognize, however, that these are empirical claims, and we lack empirical evidence.
196. Barack Obama's, Dreams from My Father, which chronicles his development of a black identity provides as good an example of the push toward blackness as I can think of. Recently, however, there has been some pushback, as more young people of mixed heritage are asserting a multiracial identity.
197. Susan Saulny, Black? White? Asian? More Young Americans Choose All of the Above, N.Y. TIMES, Jan. 30, 2011, at A1.
198. For example the views of a black student who thinks affirmative action for blacks is unconstitutional will contribute uniquely to a discussion because unlike the same views expressed by whites, supporters of affirmative action will be unable to dismiss the black student's views as just racism.
199. If black and Hispanic students from low SES backgrounds count for class diversity as well as racial/ethnic diversity I expect there has often been greater numbers from the nation's lowest SES quartile.
200. It is possible that making social-economic status salient by adopting an affirmative action program for students in the bottom quartile of the SES scale would by making SES salient lead to the creation of groups that would organize to achieve these ends, but unlike the situation with blacks, Hispanics or Native Americans there has been no serious call for the creation of such preferences either within or external to the law school community.
201. I am, of course, here talking personally and anecdotally.
202. No doubt there are occasions where students reference personal experiences stemming from an impoverished background.
203. Perhaps had I taught welfare law rather than evidence I would have heard such stories, but even then I expect a good portion of them would have come from minority students.
204. Let me remind the reader that I am talking about moral/policy justifications for affirmative action and not simply those justifications that the law as currently interpreted recognizes as compelling state interests.
205. Some who oppose race-based affirmative action think that discrimination and its effects are a thing of the past.
206. This is not so. For a summary of recent findings and data
207. Richard Lempert, A Personal Odyssey Toward a Theme: Race and Equality in the United States: 1948-2009, 44 LAW & SOC'Y REV. 431, 440-55 (2010).
208. If there are differences, they will not necessarily mean that using an SES index composed of only two measures distorts a true picture, for it could be that the number of available measures is itself an indicator of SES.
209. If this were the case, one might, for example, expect those whose scores were based on only two variables to contain a higher proportion of the respondents of low SES than one finds when four measures are available.
210. Still, it would be a comfort if the relationships Professor Sander reports were robust to differences in index construction.
211. There are some data sources that might be explored to see if they offer anything of value.
212. The most prominent are the Panel Study of Income Dynamics, the Wisconsin Longitudinal Study and the Longitudinal Study of Youth.