1 Two works that have attempted some explanation of recent trends are R. Tobin, Towards an Economic Perspective on the Legal Profession (1981), and Siow, “Occupational Choice Under Uncertainty,” 52 Econometrica 631 (1984). Some earlier works that applied economic analysis to the “market for lawyers” are discussed infra at note 54.Google Scholar

2 Nor is this growth over; a recent government report projects that law will be one of the ten fastest growing occupations in the nation between 1984 and 1995, and at least one analyst has projected a lawyer population of 930,000 by 1995. U. S. Bureau of Labor Statistics, Monthly Labor Rev., Nov. 1985, 45 table 2; B. Curran et al., The Lawyer Statistical Report: A Statistical Profile of the U. S. Legal Profession in the 1980s at 5 (1985). Given the recent surge in law school applications, even these estimates are probably too low.Google Scholar

3 See Galanter, , “Adjudication, Litigation, and Related Phenomena,” Law and the Social Sciences 166 (1987). Galanter gives statistics on “lawyers per capita” for the United States and 19 other major nations. The other nations have a combined lawyer population of about 475,000—only two-thirds of the U. S. lawyer population. However, two qualifications are in order: first, many of the statistics for other nations are from the 1970s, and there may be recent booms along American lines. Second, some countries—Japan, for instance—define the legal profession more narrowly than we do, classifying many of their trained legal workers in lower status occupations.Google Scholar

4 Between 1970 and 1985 the legal profession grew 141%, compared to 46% growth for the professions in general and 34% growth for the entire work force. Bureau of the Census, U. S. Dep't of Commerce, 1 Historical Statistics of the United States, 140 (1976); Bureau of the Census, U. S. Dep't of Commerce, Statistical Abstract of the United States: 1987 385 (1986) (“SAUS 1987”) (for the 1985 numbers, technicians were added to the population of professionals to maintain comparability with the 1970 figures).Google Scholar

5 The physician population rose from 146,000 in 1920 to 282,000 in 1970, a period during which the lawyer population rose from 123,000 to 271,000.Google Scholar

6 These measures are closely related to the “national product accounts,” which measure the annual growth of the economy in terms of output, or gross national product (GNP). National income, which is also measured by the Bureau of Economic Analysis in the Commerce Department, measures the “total factor costs of the goods and services produced by the economy.” In other words, it is a measure of economic inputs. SAUS 1987 at 413 (cited in note 4).Google Scholar

7 For example, if a law firm has $1,000,000 in receipts, and spends $100,000 on supplies, $100,000 on rent, $500,000 on salaries and nets a $300,000 profit for its partners, it has added $800,000 in value to the economy—the value of the supplies and the rent was generated by other sectors (e. g., the paper industry and the building industry).Google Scholar

8 This does not mean, however, that Americans spend more on lawyers than they do on cars. For one thing, nearly half of our cars are imported and do not show up in national production accounts; for another, the auto industry purchases a great many inputs, such as steel and rubber, which of course do not show up as value added by the auto manufacturers themselves.Google Scholar

9 The number is based on a 1981 National Law Journal survey of the in-house legal staffs of the 100 largest corporations. Nat. L. J., Mar. 30, 1981.Google Scholar

10 There are approximately 200,000 attorneys practicing law in corporate, public, and nonprofit settings. Assuming that the average income of those attorneys is $45,000 and the overhead costs of each attorney are, on average, equal to his or her salary, the value added by these other legal activities would be approximately $18 billion.Google Scholar

11 Friedman, L., A History of American Law 554 (1973). See also Galanter & Palay, “The Transformation of the Large Law Firm” (unpublished, 1988).Google Scholar

12 Bureau of the Census, U. S. Dep't of Commerce, County Business Patterns 7 (table 1, U. S. Summary) (1st Quarter Report, 1947) (“Census, County Business Patterns”). Classifications of firms by number of employees are far more readily available than (admittedly more interesting) data on the number of attorneys.Google Scholar

13 Employee figures are from Census, County Business Patterns U. S. Summary (1st Quarter reps., 1970, 1986). Judging from surveys of law firms (i. e., the annual survey of the National Law Journal) and from first-hand observation, nearly all of the firms with 250 or more employees had 140 or more attorneys.Google Scholar

14 See National Association for Law Placement, Directory of Legal Employers, an annual publication in which law firms across the country provide summary information for prospective new associates. Nearly all firms with more than 100 attorneys list over a half-dozen specialties covering most major areas of corporate law (although many are still best known for a single specialty).Google Scholar

15 Census, County Business Patterns (cited in note 12) (annual issues). The reports list the number of firms employing specified ranges of employees. Employees are rhose treated by the firm as employees for Social Security tax purposes. The estimates given here are based on interpolation of the data ranges given by the Census.Google Scholar

16 Id. Google Scholar

17 Id. It should be noted, however, that many of the smaller firms are new firms, so that employment at existing big firms may have grown as fast or faster than employment in other incumbent firms during this period. Moreover, we believe that a larger proportion of middle-sized firms (compared to the largest firms) became professional corporations during the 1970s, which partially accounts for their higher rate of employee growth (see note to table 3).Google Scholar

18 This figure was based on a comparison of the 1967 Census of Service Industries (which has some data on the number of attorneys by firm size) with a 1987 survey of large law firms by the National Law Journal. Because the 1967 census was incomplete and because the two data sources are based on such different sources, this conclusion is only a tentative one. 1967 Census of Business, Selected Services, Subject Reports 5–23 (Nov. 1970); ”The NLJ 250,” Nat. L. J. (Spec. Supp.). Sept. 28, 1987.Google Scholar

19 The total lawyer population grew an average of 5% to 6% during these same two decades—a significantly slower rate.Google Scholar

20 As early as 1944, a government study of eight major professions found that lawyers had the most unequal distribution of incomes. The study's data on 1941 incomes showed that the most affluent 5% of all lawyers accounted for 28% of all lawyer income. Among physicians, the second most unequal profession, the top 5% of earners accounted for only 18% of total physician income. U. S. Dep't of Commerce, Survey of Current Business 15, 18 (May 1944).Google Scholar

21 Heinz, J. & Laumann, E., Chicago Lawyers: The Social Structure of the Bar, ch. 10 (1982) (“Heinz & Laumann, Chicago Lawyers”).Google Scholar

22 Bureau of the Census, U. S. Dep't of Commerce, 1972 Census of Selected Service Industries, I Summary and Subject Statistics, ch. 4, table 4 (July 1976) (“1972 Service Census”).Google Scholar

23 The proportion of lawyers in private practice may be growing again. Among graduating law students who responded to a national survey, the proportion entering private practice increased from 52% in 1974 to 60% in 1983. Moreover, a 1985 update of the figures in table 6 found that the proportion of attorneys in private practice had increased to 74%. National Association of Law Placement, Class of 1975 Employment Report and Class of 1983 Report; B. Curran et al., Supplement to the Lawyer Statistical Report: The U. S. Legal Profession in 1985 (1986) (“Curran, Supplement”).Google Scholar

24 The Census of Service Industries reported around 14,000 paralegals in 1972, 32,000 in 1977, and 58,000 in 1982. These data are incomplete, however, since they only measure paralegal employment in law firms. By fitting the data with the Census data on the entire labor force (for 1979), we can estimate that slightly over half of all paralegals worked at law firms, and can adjust the service industry census figures accordingly. See 1972 Service Census, Table 2c (cited in note 22); Bureau of the Census, U. S. Dep't of Commerce, 1977 Census of Service Industries, I Subject Statistics ch. 5, table 6 (1981); Bureau of the Census, U. S. Dep't of Commerce, 1982 Census of Service Industries, Industry Series, Miscellaneous Subjects table 31 (1985).Google Scholar

25 Bureau of the Census, U. S. Dep't of Commerce, 1980 Census of Population, Subject Report 8B, Earnings by Occupation and Education 77 (May 1984).Google Scholar

26 U. S. Dep't of Labor, Employment and Earnings (Jan. 1987).Google Scholar

27 Another possibility is that big firms have been trying to cuttail their hiring of new associates, partly because associate salaries have escalated in recent years and partly because firms are uncertain whether the growth necessary to sustain a large staff of associates will last much longer. Paralegals offer firms a way of handling current demand without incurring great cost and long-run commitments.Google Scholar

28 A related issue is why law schools decided to expand so remarkably. Peter Pashigian, in a 1978 analysis of the market for lawyers, observed the slow growth of ABA-approved schools during the 1960s, and argued that since those schools were nonprofit institutions, they had no incentive to expand their enrollments in the face of rising applications. Pashigian, “The Number and Earnings of Lawyers: Some Recent Findings,” 1978 A. B. F. Res. J. 51 (1978).Google Scholar

29 Those without any background in economics may find our introduction of some economic concepts in part II heavy going. We would suggest that such readers simply skip those paragraphs that seem confusing and focus on the data and conclusions presented in this part. Parts III-VI are less technical.Google Scholar

30 Except where otherwise indicated, our references to “supply” refer to the “long run supply” of legal services. The “short-run supply” of legal services is the amount that can be supplied by existing lawyers (the entire short-run curve shifts every time new lawyers enter the market). Long-run supply takes into account the additional entry of new lawyers who are attracted by higher prices (i. e., salaries) in the market. Since we are interested in explaining developments over the past generation, the long-term focus is appropriate.Google Scholar

We will present the long-run supply curve as an upward-sloping line. Although the long-run function is arguably horizontal, we believe that the legal services industry has become such a large employer of talented college graduates that the increased entry into the legal profession since the 1960s has created economic “scarcity” in other markets, thus causing an upward-sloping curve. In any case, this distinction does not affect our analysis.Google Scholar

31 The Commerce Department's Bureau of Economic Statistics publishes detailed price indices for various economic sectors. These measures, as well as the consumer price index, attempt to measure the changing price of legal services by tracking the cost of a few basic legal transactions (e. g., a real estate closing) over time. Unfortunately, this unit is aggregated with several other components even in the Bureau's most detailed price indices. See Bureau of Economic Statistics, U. S. Dep't of Commerce, Survey of Current Business, July issues for each year.Google Scholar

32 Obviously, if a lawyer cuts his hours by 10% but charges the same prices, his income will fall 10%.Google Scholar

33 Although a digression from the purpose of this paper, it is interesting to note how little lawyer incomes suffered during the Great Depression (compare the 3% drop in mean lawyer incomes from 1929 to 1935 with the 15% drop in median family income). This suggests that much of what lawyers do is either immune to economic fluctuations or is, in fact, counter-cyclical (consider bankruptcy law, for example). On the other hand, data in table 11 suggest that lawyers did not weather the 1982 recession nearly as well.Google Scholar

34 It is important to keep in mind that part of this decline was due to the “cohort effects” of the lawyer boom itself. The proportion of young lawyers and female lawyers (both of whom earn less than other lawyers) was significantly higher in 1980 than it was in 1970. The larger proportion of low-earning groups pushed the median down. These cohort effects may explain as much as 60% of the drop in median income.Google Scholar

35 Comparing the earnings of lawyers with the incomes of families is in some ways a comparison of apples and oranges, since families often have multiple sources of income. Indeed, a major reason why median family income rose during the 1970s was the entry of many wives into the general work force (the earnings of males barely kept up with inflation during the decade; see Bureau of the Census, U. S. Dep't of Commerce, Statistical Abstract of the United States: 1981 at 407 (1981)). However, the comparison of lawyer and family incomes is appropriate here, because we are trying to say something about how affordable legal services were for American consumers. The sizable drop in lawyer incomes, combined with the small rise in median family income, suggests that legal services should have become more affordable to the typical American family during the 1970s.Google Scholar

36 See Mnookin, & Gilson, , “Sharing Among the Human Capitalists,” 37 Stan. L. R. 313 (1985).CrossRefGoogle Scholar

37 Again, however, part of the increased inequality was due to the cohort effect mentioned in note 34. A larger proportion of low-earning, young attorneys, for example, increased the apparent inequality of the profession. Limitations in age-specific income distributions make it difficult to assess how much of the increased inequality was due to this effect.Google Scholar

38 During this period many partnerships were converted to professional corporations and thus fell out of the IRS data. Because of this shift, the partner data should be interpreted cautiously.Google Scholar

39 For purposes of comparison with the census data, the IRS data show that between 1969 and 1979, sole proprietor incomes fell 28% and partner incomes fell 4%. The decreases were much smaller for this period because incomes actually rose significantly from 1969 to 1972.Google Scholar

40 See table 14 and the accompanying discussion in part V below for further analysis of trends at elite firms.Google Scholar

41 We are mindful in doing so of our earlier caveats about mixing price and quantity changes, as well as the fact that we are ignoring obvious distributional issues (big firm prices undoubtedly changed in different ways from sole proprietor prices). Notwithstanding this simplification, we think that the very general conclusions we can draw from this assumption are valid.Google Scholar

42 Although incomes may have declined 15–20%, rather than 10%, between 1970 and 1985, we use the lower figure to adjust for the cohort effect discussed in note 34. The median age of attorneys fell from 43 in 1970 to 37 in 1980. The lower incomes earned by younger attorneys (and lower prices charged for their services) did not reflect a change in the market as much as a proper depreciation for their lower level of experience.Google Scholar

43 The year 1970 seems like an appropriate starting point for analyzing the lawyer boom for two reasons. First, the most dramatic increases in first-year law school enrollments occurred in 1969 (a one-year jump of 23% at ABA-approved schools). Second, lawyer incomes began to increase sharply in 1969, after remaining stable from 1967 to 1969.Google Scholar

44 If prices fall, then supply must increase at least as much as output increases (regardless of one's assumptions about supply and demand elasticity). This implies that supply between the late 1960s and the late 1980s period must have increased at least 200%.Google Scholar

45 In economic terms, “price inelastic” means that a 10% decrease in price would lead to a less than 10% increase in purchases.Google Scholar

46 This would imply a price elasticity of about 20. In most markets it is rare to observe price elasticities greater than 2.0 unless there are very good substitutes for the product in question (which is certainly not the case with legal services).Google Scholar

47 Our claim that the demand for legal services is supported by an analysis of data on one small sector of the legal services industry: the market for wills. Data on the price of wills collected from a sample of firms indicates that the mean price of a will (on a $500,000 estate) rose about 10% from 1976 to 1980. Arguably, this is a good proxy of probate costs generally. The number of probate and other estate cases filed, however, declined less than 2% in a sample of eight states. This would correspond to a highly inelastic price tesponse of 0.2. Price information is from Altman & Weil, Inc., The Survey of Law Firm Economics (annual survey); data on probate cases is from the National Center for State Courts, State Court Caseload Statistics 68 (1977 ed.) and 91 (1981 ed.).Google Scholar

48 American Bar Association, Review of Legal Education (annual publication). It may seem misleading to use LSAT administrations as a proxy for applications, since some people take the exam more than once and others never actually apply. However, what little evidence is available on the volume of actual applications suggests that the LSAT is a good measure of the trend of applications. For example, an analysis of applications to 107 schools between 1967 and 1972 shows that applications to these schools increased by 158%, while LSAT administrations increased by 157%. Association of American Law Schools, Prelaw Handbook (1968 & 1973 eds.).Google Scholar

49 Between 1960 and 1970, the proportion of 18–24-year-olds attending college increased from 22% to 32%. Bureau of the Census, U. S. Dep't of Commerce, I Historical Statistics of the United States, 383 (1972). In addition to the factors mentioned above, the desire to avoid the draft may have also played a role in stimulating enrollments.Google Scholar

50 Warkov found that 70% of college students who planned to attend law school in both their freshman and senior years had fathers who were professionals or managers; 43% came from families with annual incomes over $15,000 ($54,000 in 1985 dollars); 60% came from families with annual incomes over $10,000 ($36,000 in 1985 dollars); 30% had fathers with graduate degrees. Those who actually ended up in law school had similar characteristics. S. Warkov, Lawyers in the Making tables 1.2, 1.3, 1.4 (1965). Heinz and Laumann, in a 1975 survey of almost 800 Chicago lawyers, found that 78% of the lawyers entering the bar from 1961 to 1975 had fathers from professional, technical, or managerial occupations; this represented an increase in the proportion of lawyers from society's upper strata compared to those admitted to the bar before 1960, despite the wider availability of higher education during the later period. Heinz & Laumann, Chicago Lawyers 190 (cited in note 21).Google Scholar

51 American Council of Education, National Norms for Entering College Freshmen, ACE Research Reports (annual series beginning in 1966). For example, from 1973 to 1983, students at schools with “very high” or “high” selectivity were consistently twice as likely to list “attorney” as their probable occupation as students at schools with “low” selectivity. However, the selective schools grew only 10% during this period, compared to a 35% growth at the least selective schools.Google Scholar

52 The 22-year-old male population is roughly estimated by taking a constant proportion of the total number of live births 22 years before the data date, adjusting for sex and mortality rates.Google Scholar

53 In analyzing the impact of the baby boom, we initially will only consider its impact on the volume of male attorneys. Although the population of women attorneys was also growing rapidly during this period, it will become clear below that purely demographic factors were of secondary importance in that phenomenon.Google Scholar

54 A few works in the 1970s attempted to analyze the legal profession as a labor market, making use of this relative earnings approach, including Freeman, ”Legal ‘Cobwebs’: A Recursive Model of the Market for New Lawyers,” 57 Rev. Econ. & Statistics 171 (1974); Pashigian, , “The Market for Lawyers: The Determinants of the Demand for and Supply of Lawyers,” 20 J. L. & Econ. 53 (1977) (for a less technical version of Pashigian's analysis, see Pashigian, 1978 A. B. F. Res. J. (cited in note 28). This approach was also followed in Siow, 52 Econometrica (cited in note t). For an analysis that includes the more recent period, see Williams, “An Economic Analysis of the Market for Lawyers” (dissertation in progress, 1989) (“Williams, ‘Market for Lawyers’”).Google Scholar

55 At the same time (that is, from 1969 to 1979) median family income was rising 7%, median wages were fairly stable, and the mean income in other professions, though falling in a number of cases, fell much less than lawyer incomes.Google Scholar

56 Pashigian, in his paper on the labor market for lawyers, used the IRS data (given here in table 11) as a proxy for mean lawyer income. See Pashigian, 57 Rev. Econ. & Statistics 53. However, these data only include sole proprietors and law firm partners–about half of all lawyers. Moreover, an increasing proportion of law firms during the 1970s converted to “professional corporations”–a legal change that excluded partners in those firms from the IRS data. Finally, we suspect that college students are more likely to have information about the starting salaries of lawyers than the year-to-year fluctuations of partner incomes.Google Scholar

57 Bureau of Labor Statistics, U. S. Dep't of Labor, National Survey of Professional, Administrative, Technical and Clerical Pay (annual report). Although the number of lawyers covered in this series is relatively small, we have reason to think it is representative of salaries among young lawyers. From 1967 to 1972, for example, this series shows a 58% rise in nominal salaries; the Census Bureau's quinquennial survey shows a 60% rise in nominal salaries of all associates over the same period.Google Scholar

58 Specifically, our index is the annual income of men aged 18–24 with four or more years of college.Google Scholar

59 American Council on Education, The American Freshman: Twenty Year Trends 74 (1987) (“ACE, Freshman”).Google Scholar

60 Bureau of the Census, U. S. Dep't of Commerce, Statistical Abstract of the United States: 1967 at 140 (1967).Google Scholar

61 Freeman, R., The Overeducated American 86 (1976).CrossRefGoogle Scholar

62 Bureau of the Census, U. S. Dep't of Commerce, Statistical Abstract of the United States: 1975 142 (1975), and Statistical Abstract of the United States: 1985 158 (1984). Of course, part of the drop in education degrees among women came from widening opportunities to enter other fields. But since men, too, entered teaching in rapidly declining numbers, it seems safe to assume that much of the decline reflected a tightening market for teachers. See, ACE, Freshman (proportion of male college freshmen planning to become teachers fell from 9.3% in 1970 to 2.6% in 1985).Google Scholar

63 This hypothesis is supported by the continued rapid rise of women MBAs after 1975, since there is reason to believe that places at business school were not as severely rationed. Pashigian, for example, presents data showing that the ratio of test-takers to actual admittees was, in 1973–74, 3.34 for law school but only 1.66 for business school. Pashigian, Rev. Econ. & Statistics at 76 (cited in note 54).Google Scholar

64 American Bar Association, Rev. of Legal Educ. (various issues). The eight prestige schools analyzed are the law schools at Harvard, Yale, Columbia, University of Pennsylvania, Stanford, Berkeley (Boalt), Michigan, and University of Chicago. Barron's Educational Series, Inc., Barron's Guide to Law Schools 34 (7th ed. 1986) (“Barron's Guide to Law Schools”). The original source for the rankings was the Ladd-Lipset Report, which asked faculty members of graduate schools across the country to rank top law schools. See Chronicle of Higher Educ., Jan. 15, 1979.Google Scholar

65 American Bar Association, Rev. of Legal Educ., various editions.Google Scholar

66 Id. at 66 (1986 ed.)Google Scholar

67 In 1967, when roughly one-third of those entering law school did not graduate, the average “academic attrition rate” announced by law schools was only 11%. Although this attrition rate did fall by half over the next decade, it seems that the proportion of law school exits explained by academic failure was only about one-third. It is possible, however, that more students in the 1960s and before had sufficiently poor performance in law school that they left voluntarily, even though they did not “flunk out.” Attrition figures are collected in the AALS, Prelaw Handbook, various issues (cited in note 48).Google Scholar

68 Specifically, the huge jump in applications from 1967 to 1971 seems to involve some additional dynamics that we have not discussed here. To assume that this jump was caused entirely by demographics and rising salaries would require a greater sensitivity to rising salaries than is reasonable. See Williams, “Market for Lawyers” (cited in note 54).Google Scholar

69 According to a recent survey, only 2.6% of law school graduates directly entered solo practice. National Association of Law Placement, 1986 Employment Report and Salary Survey. Google Scholar

70 See, e. g., “Surfeit of Lawyers,”Newsweek, Dec. 9, 1974, at 74; “Too Many Lawyers, Too Few Firms,”Business Week, June 2, 1972 at 21; “Top-Dollar Lawyers: Washington Lawers,”Newsweek, Mar. 14, 1977, at 97.Google Scholar

71 The monthly labor estimates for the lawyer population have a standard error of approximately 20,000 workers. We have converted the monthly figures into 12-month floating averages to eliminate seasonal fluctuations and smooth the trend. U. S. Dep't of Labor, Employment and Earnings, monthly issues.Google Scholar

72 Expected lawyer growth was computed by determining the number of new admissions to the bar (approximately 40,000 per year from 1980 to 1987) and subtracting net exits from the active bar. Net exits are calculated by computing the net exit rate during the 1960–80 period from two groups: lawyers over 55 years of age and lawyers under 55 years of age. The annual exit rate for the former group appeared to be 5%; for the latter group, 1.5%. This yielded estimated annual net exits of 10,500, for a net annual growth of 29,500.Google Scholar

73 See Posner, R., The Federal Courts: Crisis and Reform (1985) (“Posner, Federal Courts”); Bok, , “A Flawed System of Law Practice and Training,” 33 J. Legal Educ. 570 (1983).Google Scholar

74 Posner, Federal Courts 352 (app. B).Google Scholar

75 Id. Google Scholar

76 Galanter, , “Reading the Landscape of Disputes,” 31 UCLA L. Rev. 4 (1983).Google Scholar

77 Federal district court case filings rose 39% over the same period.Google Scholar

78 Some of these factors have been discussed by Richard Posner, Federal Courts 81–93; and by Marc Galanter, in ”The Day After the Litigation Explosion,” 46 Md. L. Rev. 3 (1986). Both authors give some quantitative assessment of how these factors have influenced court caseloads but do not attempt to measure the impact of these factors cumulatively on America's total output of legal services.Google Scholar

79 Although a few, such as John Kenneth Galbraith, have argued for some time that corporations engage in “demand creation” through advertising, this idea has never worked its way into conventional economic thought.Google Scholar

80 Moreover, it is difficult for clients to do comparative shopping for any but the most basic legal transactions. In a complex legal matter, an attorney usually cannot evaluate how much work will be required until a fair amount of research is done; thus, he is unlikely to give a definite cost estimate in advance.Google Scholar

81 Tullock, G., Trials on Trial: The Pure Theory of Legal Procedure (1980); Tullock, , “On the Efficient Organization of Trials,” 28 Kyklos 745 (1975); Tullock, “Efficient Rent Seeking,”Toward a Theory of the Rent-seeking Society 97 (Buchanan et al. ed. 1981).Google Scholar

82 In other words, if party A spends χ dollars on its case, and party B spends Y dollars, then the probability of A winning is χ/(χ+Y).Google Scholar

83 In the case described above, an equilibrium would occur when each party spends $250,000 on legal fees.Google Scholar

84 Tullock showed that where a case involves two parties, if marginal product is constant (i. e., increases directly with costs) then each party will spend twice as much as it would if marginal product increased only with the square root of costs.Google Scholar

85 Posner, Federal Courts 112 (cited in note 73). During the same period, the average length of U. S. Supreme Court opinions increased 88%. In a similar analysis, we also found that the average length of appellate opinions has been increasing over time, particularly since 1974.Google Scholar

86 The 1967 estimate is from the 1967 Census of Business, table 5 (cited in note 18) (firms with 50 or more lawyers). The current estimate is from Nat. L. J., Nov. 9, 1987.Google Scholar

87 The 1967 estimate of partner incomes is from the 1967 Census of Business, table 5 (55 firms with 50 or more lawyers). The current estimate of partner income is the median “profits per partner” from the largest 75 firms reported in ” The AM Law 100,” American Lawyer (spec. supp.) JulyAug. 1988.Google Scholar

88 However, since incomes went up in the 1967–72 period and in the mid-1980s. incomes for this group are probably now about the same as they were the late 1960s.Google Scholar

89 Curran, Supplement (cited in note 23).Google Scholar

90 The eight prestige schools are those given in note 64. Data on enrollments is from Rev. of Legal Educ., 1961 ed.Google Scholar

91 Id., 1981 ed.Google Scholar

92 The rising quality of young lawyers is a fact that has been overlooked by both the legal press and researchers. For example, see Gilson, & Mnookin, , “Coming of Age in a Corporate Law Firm: The Economics of Associate Career Patterns,” 41 Stan. L. Rev. 567, 589–92 (1989). They argue that the “up or out” system of promotion in law firms is becoming less universal because of a decline in the quality of associates.CrossRefGoogle Scholar

93 AALS, Prelaw Handbook, (1969 ed.) (cited in note 48); Baron's Guide to Law Schools, (1980 ed.) (cited in note 64).Google Scholar

94 Though other firms apparently still believe that quality has declined; see Gilson & Mnookin, 41 Stan. L. Rev. Google Scholar