1 Richard Abel, Comparative Sociology of Legal Professions: An Exploratory Essay, 1985 A.B.F. Res. J. 5.CrossRefGoogle Scholar

2 Id. at 6.Google Scholar

3 Id. at 8.Google Scholar

4 Id. at 57.Google Scholar

5 A substantial contribution. however, is the article by B. Peter Pashigian, The Number and Earnings of Lawyers: Some Recent Findings, 1978 A.B.F. Res. J. 5I. A more technical analysis was presented in his article The Marker for Lawyers: The Determinants of the Demand for and Supply of Lawyers, 20 J.L. & Econ. 53 (1977). Pashigian examined several factors in the demand for lawyers' services and the long-run supply of lawyers for an explanation of changes in lawyers' earnings. However, his study deals with the period 1920 to 1970 and therefore cannot be compared directly with the work reported here.Google Scholar

6 H.W. Arthurs, R. Weisman. & F.H. Zemans, The Canadian Legal Profession, 1986 A.B.F. Res. J. 447.CrossRefGoogle Scholar

7 Id. at 461.Google Scholar

8 Their article does, however, present a thorough description of the structure and behavior of the legal profession in Canada and provides a useful context for this paper.Google Scholar

9 This group is defined by code 2343, Lawyers and Notaries, in the 1980 Standard Occupational Classification. Notaries in Quebec (and in British Columbia in 1981) are included, but in other provinces notaries are in the paralegal group. (Quebec notaries are similar to solicitors in common law systems, with respect to their training and work, so are properly included with lawyers in this classification.) Some articling students, legal secretaries, and law clerks may have been incorrectly coded as lawyers. In an effort to exclude any miscoded paralegals, all respondents without a university degree were deleted from the data used in this study. The data for 1971 and 1981 are comparable because there were no changes in the occupational classification for lawyers over this period. Although the data include only those who worked full time, there may be considerable variation in the number of hours worked, or billed to clients.Google Scholar

10 The inflation factor based on the consumer price index between 1970 and 1980 was 2.168. The comparison of earnings in 1970 and 1980 should not be influenced by economic conditions in these years since each was a period of economic slowdown just prior to an expansionary phase.Google Scholar

11 Regression analyses of earnings have used either the logarithms of earnings or the actual levels as the dependent variable. For example, studies of earnings differentials (male/female, black/white, native/immigrant) are about equally divided between those that use actual levels and those that use logarithms. See R.A. Filer, Sexual Differences in Earnings: The Role of Individual Personalities and Tastes, 18 J. Hum. Resources 82, 86 n.5 (1983). This study uses actual earnings as the dependent variable, since the earnings are inflation-adjusted, so that the results can be understood by a wider readership.CrossRefGoogle Scholar

12 A trivariate classification is used for religion, with Jewish and Roman Catholic religions represented by separate dummy variables. All other religions were combined with Protestant for the reference group because most of this remainder were in the “no religion” category but could be considered to be “quasi-Protestant,” at least for the purpose of earnings analysis.Google Scholar

13 Home language is a better variable for ethnicity than the census question on ethnic origin because the latter allows for a multiplicity of origins in 1981 and differs from the 1971 question on ethnic origin.Google Scholar

14 A bivariate classification is used for birthplace, with foreign-born represented by a dummy variable and Canadian-born treated as the reference group. Foreign-born lawyers are not correlated with those preferring a third language. In the 1981 census, 71% of the foreign-born lawyers in law offices were anglophones and 6% were francophones. Only 23% of the foreign-born reported using a third language at home. Almost half of the foreign-born lawyers were born in the British Isles (27%) or the United States (18%). At least three-quarters of all immigrant lawyers came to Canada prior to attending law school (1981 Census, special tabulations).Google Scholar

15 To reduce the number of variables, widowed lawyers were included in the singles category, but they are less than 1% of the total.Google Scholar

16 The 1971 census provided only one category for advanced degrees, namely, a master's degree or doctorate. Three categories were used in 1981, but these are combined, for comparability with 1971. These latter categories revealed that a master's degree was held by 81% of those with advanced degrees; only 19% held a doctorate. The reference group for 1971 was those with a first professional degree, and for 1981 it was the combination of bachelor's degree (including LL.B.) and university certificate or diploma above bachelor level because the latter includes the notarial diploma in Quebec.Google Scholar

17 Four industry groups are used: law offices, public administration, private industry, and all other industries. In the following discussion, these are referred to as private practice, government service, private industry, and nonprofit organizations.Google Scholar

18 These are included to take account of the expected parabolic shape of the age-earnings profile.Google Scholar

19 An individual's workplace address was reported in the census and coded according to census tract. By comparing street addresses of the largest law firms with census tract definitions, it was confirmed that these firms were highly concentrated within a single census tract in each of the major cities. There are also medium and small firms in these census tracts, but they tend to share the same expensive locations and the higher-income clientele of the largest firms.Google Scholar

20 The total number of lawyers (full time plus part time) increased by 109% between the 1971 and 1981 censuses. By comparison, the increase was 42% for physicians; 49% for civil engineers; and 45% for accountants.Google Scholar

21 The lawyer growth rates for the 1970s were 50 to 60% in the United States, United Kingdom, Germany, and New Zealand; and 75 to 80% in the Netherlands and Belgium. See P.S.C. Lewis, A Comparative Perspective on Legal Professions in the 1980s, 20 Law & Soc'y Rev. 80 (1986).Google Scholar

22 Lawyers in each of these groups include only those whose primary job function is the practice of law.Google Scholar

23 Part of this change was due to the increasing proportion of females, of whom an increasing proportion were Catholic.Google Scholar

24 Appropriate tests revealed that multicollinearity was not a problem, with the possible exception of the age and age-squared variables. The robustness of the particular parameter estimates under alternative specifications added further confirmation to this conclusion. Testing with interactive variables yielded generally nonsignificant results, so these were not included in the final specification of the model.Google Scholar

25 The Chow test on the equality of the estimated coefficients indicated that there is a structural difference between the regressions for the two years, 1970 and 1980. The F-value was 4.6, which is above the 0.99 critical value of 1.8.Google Scholar

26 See Barry D. Adam & Douglas E. Baer, The Social Mobility of Women and Men in the Ontario Legal Profession, 21 Can. Rev. Soc. & Anrhrop. 21 (1984).CrossRefGoogle Scholar

27 F-tests showed that these industry variables as a group did not provide a significant contribution to the explanation of earnings in either 1970 or 1980.Google Scholar

28 City size is a proxy for several variables such as size of firm and field of law, but note that lawyers living in the large cities and working in the commercial core combined the effects of these two variables, and hence had much larger earnings.Google Scholar

29 This is approximately $14,500 when the 1970 negative and 1980 positive differentials are combined.Google Scholar

30 The real gross domestic product (GDP) for Ontario increased by 42% between 1970 and 1980, but for Alberta the increase was 166%. The GDP in Saskatchewan rose by 97%; in Manitoba the increase was a more modest 40% (Statistics Canada, No. 13–213).Google Scholar

31 Similar regressions for other selected professions showed that lawyers were not the only profession to encounter declining real earnings in the 1970s. The real earnings of architects and physicians also declined; but accountants, civil engineers, and dentists experienced higher average real earnings. Preliminary analysis indicates that no single set of factors can explain these differences between the professions.Google Scholar

32 Lower earnings could also be attributed to a declining rate of growth in demand. Preliminary evidence suggests, however, that this was of secondary importance. From Pashigian's finding that the major determinant of demand for lawyers' services in the United States was total real GNP (with income elasticity of about 1.8), one might conclude that the demand for lawyers' services in Canada would have increased by about 90% during the 1970s. Pashigian also notes that lawyers' real earnings in the United States doubled between 1950 and 1970, due to strong economic growth and the slow supply response of the law schools. See Pashigian's articles cited in note 5.Google Scholar

33 This approach is comparable to the decomposition methodology used to examine components of public/private and male/female wage differentials. For a detailed explanation, see R.L. Oaxaca, Male-Female Wage Differentials in Urban Labor Markets, 14 Int'I Econ. Rev. 693 (1973).CrossRefGoogle Scholar

34 The base for comparison is 1970, whereas for the cross-sectional decomposition of the male/female differential (in Oaxaca and others), males were used as the base. To see the similarity of the methodology, note thatGoogle Scholar

35 Mean values for all other variables were used in calculating these profilesGoogle Scholar