1 McKirdy, Charles R., The Lawyer as Apprentice: Legal Education in Eighteenth Century Massachusetts, 28 J. Legal Educ. 124 (1976).Google Scholar

2 Gottlieb, Suzanne, The Feasibility of Part Time Day Legal Education, 30 J. Legal Educ. 291 (1979); Rickson, Roy E., Faculty Control and the Structure of Student Competition: An Analysis of the Law Student Role, 25 J. Legal Educ. 47 (1973).Google Scholar

3 For some this metaphor may be close to home:. The student life is … a treadmill, class and books all day, closed up in my study briefing cases every evening. And never enough sleep. The only time I get to see Annette [wife], during the week, is over dinner. On the weekends, … I've managed to take off Friday and Saturday nights for movies, music, restaurants; but even then I'm sure I'm not the best of company. Right now, law's my greatest enthusiasm, yet after losing me to the legal world all week, it's the last thing Annette wants to hear about…. To A [nnette], I'm sure it all seems a jumble and a bore. I continually make resolutions to talk about other subjects, which I somehow never keep. For Annette, there is no easy solution, and she's often left frustrated and alone. During the day time on weekends she has taken to heading off for outings in the city by herself—to museums, to exhibits, on shopping expeditions. But there are instants when it is plain that she is bitter at how unavailable I am. Last week she told me that I am more intimate with the law than with her, a not-so-subtle reference to the fact that we have been getting together in the bedroom infrequently. (From the drift of conversations around school, I take it that that problem is not unique to us.) All of it leaves me feeling toweringly guilty at moments, and also helpless, since I cannot make the law school and its demands dissolve. Scott Turow, One L, at 81, 141 (New York: G. P. Putnam's Sons, 1977). Also see Field, Mona, Bar-Cross'd Lovers, Student Law., Nov. 1980, at 45.Google Scholar

4 Lewis A. Coser, Greedy Institutions: Patterns of Undivided Commitment 1–8 (New York: Free Press, 1974).Google Scholar

5 Stephen, R. Marks, Multiple Roles and Role Strain: Some Notes on Human Energy, Time and Commitment, 42 Am. Soc. Rev. 921 (1977).Google Scholar

6 Interpretation, Council of the Section of Legal Education and Admissions to the Bar of the American Bar Association, in Memorandum 7879–11 (to Deans of ABA Approved Law Schools, from James P. White, Consultant on Legal Education to the ABA), Sept. 25, 1978, at 16.Google Scholar

7 Approval of Law Schools: American Bar Association Standards and Rules of Procedure, As Amended—1979, at 10 (Chicago: American Bar Assocation, 1979) (emphasis in original).Google Scholar

8 Interpretation of Standard 305(a)(iii), 12 Legal Educ. Newsletter, Oct. 1980, at 4. This interpretation was enacted in February 1980 and reinstates the hour limit imposed prior to 1975. During the intervening years, the limit was 15 hours. The ABA authority in this matter rests on the fact that most states require applicants for the bar to be graduates of an approved law school.Google Scholar

9 See Charles D. Kelso, The AALS Study of Part-Time Legal Education: Final Report, in Association of American Law Schools 1972 Annual Meeting Proceedings, pt. I, § II, at 1–33. (Washington, D.C.: Association of American Law Schools, 1972); Nahstoll, Richard W., Current Dilemmas in Law-School Accreditation, 32 J. Legal Educ. 236, 254–55 (1982).Google Scholar

10 Reed noted that the original distinction between day and evening law school instruction developed because of differing law school locations:. In law schools situated in small towns, such as Cambridge, instruction was naturally offered from the beginning during the regular working hours of the day, as in other departments of the university. Students had no access to a sufficient supply of law offices…. In the larger cities, there was an equally natural tendency for instruction to be given at irregular hours, for the convenience primarily of the instructors who, as judges and practitioners, had other things to do. Alfred Z. Reed, Training for the Public Profession of the Law 394 (New York: Charles Scribner's Sons, 1921) (prepared for the Carnegie Foundation for the Advancement of Teaching). There are a large number of accounts of this crucial period in the history of legal education. See Jerold S. Auerbach, Enmity and Amity: Law Teachers and Practitioners, 1900–1922, in Donald Fleming & Bernard Bailyn, eds., Law in American History (Boston: Little, Brown & Co., 1971); Jerold S. Auerbach, Unequal Justice: Lawyers and Social Change in Modern America 96–101 (New York: Oxford University Press, 1976); Albert P. Blaustein & Charles O. Porter, The American Lawyer: A Summary of the Survey of the Legal Profession 181–84 (Chicago: University of Chicago Press, 1954); Cox, Michael P., Part-Time Legal Education: The Kelso Report and More, 27 J. Legal Educ. 473 (1976); First, Harry, Competition in the Legal Education Industry, 53 N. Y. U. L. Rev. 311 (1978); id., The Business of Legal Education, 32 J. Legal Educ. 201 (1982); Gee, E. Gordon & Jackson, Donald W., Bridging the Gap: Legal Education and Lawyer Competency, 1977 B. Y. U. L. Rev. 695, 739–43, 754–55; Albert J. Harno, Legal Education in the United States: A Report Prepared for the Survey of the Legal Profession 76–80 (San Francisco: Bancroft-Whitney, 1953); Herbert L. Packer & Thomas Ehrlich, eds., New Directions in Legal Education: A Report Prepared for the Carnegie Commission on Higher Education 24–28 (New York: McGraw-Hill Book Co., 1972); Stevens, Robert B., Law Schools and Legal Education, 1879–1979: Lectures in Honor of 100 Years of the Valparaiso Law School, 14 Val. U.L. Rev. 179, 199–209 (1980); Robert Stevens, Two Cheers for 1870: The American Law School, in Fleming & Bailyn, supra; Preble Stolz, Training for the Public Profession of Law (1921): A Contemporary Review, in Packer & Ehrlich, supra.Google Scholar

11 The distinction between three- and four-year law school programs is not simply one of length of time or of day versus evening classes; rather it is one of status. Reed described evening division law instruction as “cheapened copies of the full-time model.” Alfred Z. Reed, Present-Day Law Schools in the United States and Canada 289 (New York: Carnegie Foundation for the Advancement of Teaching, 1928). While the stigma attached to part-time law programs may be undeserved at present (see Cox, supra note 10, and Kelso, supra note 9), it has continued to linger from its early days. Thus, the implications of the sanction attached to the accreditation standard is to consign students subject to its invocation to a stigmatized status in this highly stratified profession of law. See Jerome Carlin, Lawyers on Their Own: A Study of Individual Practitioners in Chicago (New Brunswick, N.J.: Rutgers University Press, 1962); Erlanger, Howard S., The Allocation of Status Within Occupations: The Case of the Legal Profession, 58 Soc. Forces 882 (1980); Ladinsky, Jack, Careers of Lawyers, Law Practice, and Legal Institutions, 28 Am. Soc. Rev. 47 (1963); Laumann, Edward O. & Heinz, John P., Specialization and Prestige in the Legal Profession: The Structure of Deference, 1977 A.B.F. Res. J. 155; Frances Kahn Zemans & Victor G. Rosenblum, The Making of a Public Profession (Chicago: American Bar Foundation, 1981). The issue is made more complex by the fact that few law schools, and none at the highest status, have both full- and part-time divisions. Thus, a student denied admission to a particular school's full-time program on these grounds may be thereby excluded from the school altogether. Kelso, supra note 9, at 2, 281–84, noted that the distinction between part-time and full-time legal education is a structural and programmatic one and not one based on the description of students' concentration of activities. He found that many students in the full-time programs he studied worked a substantial number of hours in part-time jobs and therefore were, in effect, part-time students. This is, of course, the subject of the present study as well.Google Scholar

12 “The AALS and ABA define a full-time law student as one whose scheduled studies require substantially all his working hours. Neither organization has told schools precisely when to classify students as part time. Each school has been left free to determine what is substantially the full working time of its students.” Kelso, supra note 9, at 2; also see at 283.Google Scholar

13 Peter deL. Swords & Frank K. Walwer, The Costs and Resources of Legal Education: A Study in the Management of Educational Resources (New York: Council on Legal Education for Professional Responsibility, 1974); id., Financing Legal Education, 64 A.B.A.J. 1880 (1978).Google Scholar

14 Boyer, Barry B. & Cramton, Roger C., American Legal Education: An Agenda for Research and Reform, 59 Cornell L. Rev. 221, 276–78 (1974); Gellhorn, Walter, The Second and Third Years of Law Study, 17 J. Legal Educ. 1 (1964); Pipkin, Ronald M., Legal Education: The Consumers' Perspective, 1976 A.B.F. Res. J. 1161; Stevens, Robert B., Law Schools and Law Students, 59 Va. L. Rev. 551, 652–59 (1973).Google Scholar

15 American Bar Association, Section of Legal Education and Admissions to the Bar, Lawyer Competency: The Role of the Law Schools (Report and Recommendations of the Task Force on Lawyer Competency: The Role of the Law Schools) (Chicago: American Bar Association, 1979); Burger, Warren, The Special Skills of Advocacy: Are Specialized Training and Certification of Advocates Essential to Our System of Justice? 42 Fordham L. Rev. 227 (1973) (4th Annual John F. Sonnett Memorial Lecture, Nov. 26, 1973, at Fordham Law School); Cramton, Roger C., Rising Expectations in Law Practice and Legal Education, 7 N. Ky. L. Rev. 159 (1980); id., Lawyer Competence and the Law Schools, 4 U. Ark. Little Rock L.J. 1 (1981); Cramton, Roger C. & Jensen, Erik M., The State of Trial Advocacy and Legal Education: Three New Studies, 30 J. Legal Educ. 253 (1979); Gee & Jackson, supra note 10.Google Scholar

16 Law Schools, Students Face New Loan Plan, 67 A.B.A.J. 551 (1981); Financial Aid Cuts Threaten Law Students, Student Law., Apr. 1982, at 47.Google Scholar

17 Carrington, Paul D. & Conley, James J., The Alienation of Law Students, 75 Mich. L. Rev. 887 (1977); id., Correspondence, Negative Attitudes of Law Students: A Replication of the Alienation and Dissatisfaction Factors, 76 Mich. L. Rev. 1036 (1978); Hedegard, James M., The Impact of Legal Education: An In-Depth Examination of Career-relevant Interests, Attitudes, and Personality Traits Among First-Year Law Students, 1979 A.B.F. Res. J. 791; Kimball, Edward L., Farmer, Larry C. & Monson, D. Glade, Ability, Effort, and Performance Among First-Year Law Students at Brigham Young University, 1981 A.B.F. Res. J. 671; Loftman, Guy R., Study Habits and Their Effectiveness in Legal Education, 27 J. Legal Educ. 418 (1976); Rathjen, Gregory J., The Impact of Legal Education on the Beliefs, Attitudes and Values of Law Students, 44 Tenn. L. Rev. 85 (1976); references cited in note 14 supra.Google Scholar

18 Quoted in Lester Brickman, Is Law School a Full Time Enterprise? Part Time Students and Part Time Teachers, 10 Council Legal Educ. Prof. Resp. Newsletter, May, 1978, at 1 (reporting a panel presentation by Roger C. Cramton, Ronald M. Pipkin, Norman Redlich, & Robert B. Stevens).Google Scholar

19 Cramton, Roger C., Change and Continuity in Legal Education, 79 Mich. L. Rev. 460, 464 (1981), labeled student part-time employment “the new apprenticeship.” Also see Spector, Bruce R., A Year Clerking Before the Third Year—A Way to Give Intelligent Direction to a Career and the Third Year of Law School, 30 J. Legal Educ. 521 (1980); Gee & Jackson, supra note 10; Pipkin, supra note 14; Gellhorn, supra note 14.Google Scholar

20 The research literature on law students is remarkably barren on the topic of part-time employment. Stevens briefly mentioned part-time employment in comparisons between the classes of 1960 and 1970 in the schools he surveyed:. The issue of part-time versus full-time study pervades the history of American legal education. Our preliminary data suggest that the question has not been resolved. With the exception of Yale, a majority of students in our sample schools held full-time or part-time jobs in 1960. By 1970 the numbers of those working twenty hours or more had declined at Iowa, Pennsylvania, and U.S.C. Although only at Pennsylvania did the percentage dip to under 50 percent. At Yale over two-thirds of the students in 1970 had outside employment compared with almost 50 percent of the students ten years earlier. Stevens, supra note 14, at 589–90. In an analysis of first-year grades and data from the admissions files of students entering the University of Washington Law School between 1956 and 1959, Lunneborg and Lunneborg reported:. Students were asked to indicate the number of hours they felt they would have to be employed while attending law school. Those who said they would not work had significantly better first year grades, e.g. 17 per cent of them failed compared with 44 per cent failure among those who anticipated working essentially full time. Lunneborg, Clifford E. & Lunneborg, Patricia W., Relations of Background Characteristics to Success in the First Year of Law School, 18 J. Legal Educ. 425, 433 (1966). In a recently published study, Kimball et al., supra note 17, at 689, reported that part-time employment of first-year students at Brigham Young University's law school did not produce a significant differential in grades when ability and effort (class attendance and study time) were controlled. Kelso, supra note 9, at 281–82, 468–73, although focusing on evening division part-time legal education, provided data on the employment of full-time day students. He noted that the incidence of part-time employment among third-year students varied between 38 and 60 percent at the schools studied. The rate of employment was inversely associated with levels of school resources. Similarly, a greater proportion of employed students exceeded 21 working hours per week at the lower-resource schools than at the higher-resource schools.Google Scholar

21 Other studies from these data were reported in Pipkin, Ronald M., Law School Instruction in Professional Responsibility: A Curricular Paradox, 1979 A.B.F. Res. J. 247; id., Commentary, Entropy and Skewness in the Allocations of Students to Law Schools, 59 Wash. U.L.Q. 901 (1981); id., supra note 14; Spangler, Eve, Gordon, Marsha A. & Pipkin, Ronald M., Token Women: An Empirical Test of Ranter's Hypothesis, 84 Am. J. Soc. 160 (1978).Google Scholar

22 A primary benefit of such a research design is to provide breadth to the data by sampling many different settings. One of the limitations, however, in contrast to a longitudinal design, is that the sample is not followed through time. Thus, as in this study, there were no data to ascertain longer-term consequences of part-time employment. It is commonly believed that part-time employment has risks for future academic performance. That idea could not be tested here.Google Scholar

23 Two of the schools also had a part-time evening division. For cross-school comparability only students in the full-time day programs were included in the samples.Google Scholar

24 If, for some reason, respondents could not be reached in time for participation during the week of the time log, then they were asked only to complete the questionnaire. It is largely for this reason that the two sample sizes differ, although a part of the difference also results from the neglect of a few respondents to provide a usable time log after completing the questionnaire.Google Scholar

25 Another 8 percent indicated that they were looking for a job but had not found one at the time of the survey. No measurement was made of whether students had been previously employed while in school and were not currently working, nor, if they were first- or second-year students, whether they thought they would take a part-time job during the following year or two years. Therefore, this statistic substantially understates the percentage of students likely to take a part-time job at some point during their legal educations. Also see Stevens as quoted in note 20 supra.Google Scholar

26 Later in the analysis these variables were taken as control variables to account for the temporal-spatial-social location of respondent's attitudes and activities so that individual-level relationships could be freed from contextually specific stimuli. It could be argued that year in school is actually an individual-level variable. It measures the amount of students' progress in the academic program and could operationalize differing degrees of socialization into the law student role and the legal profession. While that individual-level meaning is probably involved in the way the variable is used here as well, what is intended is that it describe variance in the structure of opportunities in the academic and professional environments tied to duration in the student role. Among these are an increasing freedom of action permitted by movement into the elective curriculum and an increasing acceptance on the part of local legal employers of a student's commitment and expertise that might translate into a marketable skill.Google Scholar

27 The literature in social science is in dispute about structural effects. Some argue that they are merely poorly measured individual-level effects or that if they exist, they cannot be isolated from individual-level effects. See Hauser, Robert M., Context and Consex: A Cautionary Tale, 75 Am. J. Soc. 645 (1970); Firebaugh, Glenn, A Rule for Inferring Individual-Level Relationships from Aggregate Data, 43 Am. Soc. Rev. 557 (1978). Others hold that social climate does exert a real influence on individual behaviors and that it can be identified and measured. See Tannenbaum, Arnold S. & Bachman, Jerald G., Structural Versus Individual Effects, 69 Am. J. Soc. 585 (1964); Lincoln, James R. & Zeitz, Gerald, Organizational Properties from Aggregate Data: Separating Individual and Structural Effects, 45 Am. Soc. Rev. 391 (1980). While I am cognizant of the former view and concerned about the issues raised by it, my present study is in the tradition of the latter view.Google Scholar

28 For an analysis of the role of parental assistance in funding students in legal education, see Swords & Walwer, Costs, supra note 13, at 281–86.Google Scholar

29 Explanations for the phenomena under study, part-time employment and its correlates, may be (a) at the contextual/structural level, (b) at the individual level, (c) at both levels, or (d) an interaction between the two levels. The policy implications of findings vary substantially depending on the level or levels of explanation. E.g., assume the finding of a higher incidence of part-time employment at law school A than at law school B. A contextual explanation would be supported by the collateral finding that students at the two schools who otherwise had comparable attributes worked in part-time jobs at different rates. Presumably, if paired students were assigned to the two schools, those attending school A would be more likely to take on part-time employment than would their counterparts at school B. The reasons would be found in differing school policies, reward systems, control mechanisms, cultural milieus, etc. In other words, the explanation would be structural: characteristics of students would be secondary to those of the organizations in affecting this behavior. Structural explanations would not be applicable to the hypothetical, however, if an analysis of individual-level variables revealed that students with certain specified characteristics were more likely than others without those attributes to take part-time jobs and that such students made up a larger proportion of the student body at school A than at school B so that when individual-level effects were controlled, structural effects fell to zero. In this situation, the reason for differential employment patterns at the two schools could probably be a result of differing recruitment practices or recruitment pools. Thus, the finding would direct the focus more on the attributes of individual students and less on the organizations. Explanation would be at both the contextual and individual levels if school effects did not disappear after controlling significant individual-level relationships with part-time employment. Such would be the case when structural and individual-level variables each made independent contributions to patterns of part-time employment. In the hypothetical, findings supporting this explanation would be when school A had a larger proportion of students than school B with characteristics inclining them toward part-time jobs and the environment in school A was niore conducive to working outside the law school than in school B so that a greater proportion of the student body, irrespective of individual characteristics, were inclined toward part-time employment. The higher incidence of part-time employment at school A could be said to be a consequence of both recruitment and school milieu. Lastly, the explanation may require an understanding of an interaction between variables at the two levels. For example, in one setting certain characteristics of individuals may enhance the probability that they will work part time and in the other setting these same characteristics appear to suppress that probability. In such a case, part-time employment would be tied to qualitative differences in the institutional settings that differentially affected certain types of students.Google Scholar

30 Goodman's Everyman's Contingency Table Analysis (ECTA) program was used. For a discussion of loglinear analysis, see Stephen Elliott Fienberg, The Analysis of Cross-classified Categorical Data (2d ed. Cambridge, Mass.: MIT Press, 1980); Goodman, Leo A., A Modified Multiple Regression Approach to the Analysis of Dichotomous Variables, 37 Am. Soc. Rev. 28 (1972); id., Causal Analysis of Data from Panel Studies and Other Kinds of Surveys, 78 Am. J. Soc. 1135 (1973); id., Multiplicative Models for the Analysis of Occupational Mobility Tables and Other Kinds of Cross-Classification Tables, 84 Am. J. Soc. 804 (1979); id., A Brief Guide to the Causal Analysis of Data from Surveys, 84 Am. J. Soc. 1078 (1979); David Knoke & Peter J. Burke, Log-linear Models (Quantitative Applications in the Social Sciences No. 20) (Beverly Hills, Cal.: Sage Publications, 1980); Page, William F., Interpretation of Goodman's Log-linear Model Effects, 5 Soc. Methods & Res. 419 (1977).Google Scholar

31 The attempt here was to find an “objective” measure of financial need that would permit comparisons between employed and nonemployed students. Clearly, subjective assessments of financial need were relevant (see § V infra), but as they are likely to be part of the legitimizing rationalizations for the activity, to use them risked tautology.Google Scholar

32 Income was reported to be as of the time of the survey, 1975–76.Google Scholar

33 An excessive number of zero cells inhibits the usefulness of the loglinear technique.Google Scholar

34 Readers familiar only with the traditional use of chi-square as a test of independence between variables in cross-tabulated categorical data should note that in loglinear analysis the statistic is used somewhat differently. For model testing, as in table 2, the chi-square assists in identifying models where expected cell frequencies are reasonably close to the observed frequencies. Therefore, “successful” tests are those in which the chi-square values are greater than the critical level, usually set at .05. When models have low chi-square values they must be rejected for misspecifying the structure in the observed data. In table 3, where the significance of individual parameters are tested, the likelihood ratio chi-square statistics are used to determine whether the terms significantly contribute to the model under test. Here, as in traditional readings of chi-squares, a parameter is held to be significant—not equal zero—if the chi-square value is at the .05 level or lower. In selecting the “best-fitting” model, all statistically significant parameters must be included. The most parsimonious models usually have a chi-square value in the .20 to .30 range of statistical significance. Models with chi-square values with a probability greater than .5 are typically to be rejected for overspecifying the relationships; i.e., more information than necessary is being used to generate a fairly accurate prediction.Google Scholar

35 The model in H₁₃ posits that part-time employment is a function of year in school, school type, and whether parents assisted in financing the student's education. Parents' income, under the model, is held to be unrelated to student part-time employment. The exclusion of the parameter {IW} has a reasonable justification. It could be that it did not matter how affluent the parents; if they gave financial assistance to their children, the pressure to take a part-time job may have been lessened. While this hypothesis proved not to be entirely correct, it did receive some support from the fact that the {IW} parameter contributed much less than the {HW} parameter to the reduction in unexplained chi-square (see table 3).Google Scholar

36 ECTA uses hierarchical modeling, which sets interaction terms to include the marginal effects of all lower-order terms. Thus, for example, the third-order interaction term of {YSW} also incorporates the second-level terms of {YW} and {SW}, as well as the first-order marginals of {Y}, {W}, and {S}.Google Scholar

37 This is condition (c) in note 29 supra.Google Scholar

38 This is computed by the ratio of chi-square for H₁ to the others: i.e., H₁LRχ²–H₁LRχ²/H₁LRχ², where i= models 2, 3, 4 …n.Google Scholar

39 At one point the variable sex was included in this model, but it was found to have no effect. Rates of pan-time employment and the influences of year, school, and parents' help with finances did not differ for male and female students.Google Scholar

40 An odds is the ratio between the frequencies of two categories. When odds are used to describe a marginal distribution, as here, an odds of 1.00 means that an individual selected at random from the sample has a 50/50 chance of falling in either group. More information is needed to make the prediction more accurate. When odds ratios are computed to express a relationship between two variables, then the value of 1.00 indicates no relationship. Odds take on only positive values and have no upper limit. Odds ratios smaller than 1.00 indicate an inverse relationship between the considered variables.Google Scholar

41 Table 4 can be used to target where (and when) student part-time employment was a prevalent activity in the population sampled. (If these schools could be taken as representative of law school types generally, then such a matrix of odds could be used by those concerned with general educational policies to identify where the activity was likely to be beyond some significant level.) E.g., if a critical level of odds were set at .9 or higher, meaning that a student drawn at random from any subgroup would be almost as likely to be employed as not, then 10 of the 36 subgroups could be identified through table 4 as exceeding the threshold value. Excluded are first-year students at all three types of law schools and all students at the elite schools. The 10 included groups would be all second- and third-year students at the public law schools, except those from the higher-income families whose parents provided financial support, and second- and third-year students at the private regional schools who did not receive parents' support, regardless of the level of parent's income.Google Scholar

42 The odds of employment are fixed and independent in the expected frequencies under the model. Consequently, odds ratios can be computed across levels of one variable within any level of another and will always produce the same ratio value.Google Scholar

43 While support from spouses would seem to offset reductions from savings and parents' support, it may also carry an understanding that the student share the burden of expenses by working part time. As well, being married rather than single often increases, rather than decreases, expenses. Therefore, the net effect of marriage may not alter levels of financial need. Indeed, married students reported difficulty in financing their legal educations at the same rates as unmarried students (see at p. 1135 infra).Google Scholar

44 Richard L. Meile, Performance and Adjustment of First Year Law Students (Ph.D. diss., University of Washington, 1961); references cited in note 14 supra.Google Scholar

45 York, John C. & Hale, Rosemary D., Too Many Lawyers? The Legal Services Industry: Its Structure and Outlook, 26 J. Legal Educ. 1 (1973).Google Scholar

46 Joel Seligman, The High Citadel: The Influence of Harvard Law School 185–200 (Boston: Houghton-Mifflin, 1978).Google Scholar

47 David F. Cavers, Legal Education in Forward-looking Perspective, in Geoffrey C. Hazard, Jr., ed., Law in a Changing America 139 (Englewood Cliffs, N.J.: Prentice-Hall, 1968); Johnstone, Quintin, Student Discontent and Educational Reform in the Law Schools, 23 J. Legal Educ. 255 (1970); Zemans & Rosenblum, supra note 11, at 47–55.Google Scholar

48 Sums irrespective of sign.Google Scholar

49 Emile Durkheim, The Elementary Forms of the Religious Life 345–47 (New York: Free Press, 1965).Google Scholar

50 Eviatar Zerubavel, Hidden Rhythms: Schedules and Calendars in Social Life 138–66 (Chicago: University of Chicago Press, 1981).Google Scholar

51 Sebastian De Grazia, Of Time, Work, and Leisure (Garden City, N.Y.: Doubleday & Co., Anchor Books, 1964).Google Scholar

52 John P. Robinson, How Americans Use Time: A Social-psychological Analysis of Everyday Behavior (New York: Praeger, 1977).Google Scholar

53 Goode, William J., A Theory of Role Strain, 25 Am. Soc. Rev. 483 (1960); Komarovsky, Mirra, Presidential Address: Some Problems in Role Analysis, 38 Am. Soc. Rev. 649 (1973); Marks, supra note 5.Google Scholar

54 See note 7 supra.Google Scholar

55 By defining their agenda and organizational activities as student's “work,” law schools create a cultural incongruity whereby work for pay in part-time employment is defined as leisure and whereby activities for which money is paid, including legal education, is defined as work.Google Scholar

56 The correlation is r= .614, p < .001 (N=374).Google Scholar

57 At the time of the survey, 15 hours per week in part-time employment was the maximum permitted under interpretations of the accreditation standard. See note 8 supra.Google Scholar

58 This measure was used in place of parents' income and assistance because it had been shown to be a proxy for them and it simplified the analysis.Google Scholar

59 F = 1.61, df = 1,331, p= n.s.Google Scholar

60 If the assumption can be made that those students in greater financial need would feel compelled to work a greater amount of time than those with less need, then this finding suggests that employment time may have been scheduled largely by employers rather than as a matter of individual students' discretion. Neither the assumption nor the speculation could be tested in the data.Google Scholar

61 Adjusted means show the effect of one factor with the others controlled. This adjustment is important when the effects are moderately correlated. Here, however, as there was no correlation or interaction between the effects of school and year, the variables operated independently. Therefore the unadjusted means best described their effects.Google Scholar

62 A few of the sample schools had some portion of their day curriculum meeting in the evening. A test of whether employed students were more likely than others to choose such courses showed that after controlling for school and semester, that was not the case. Employed and nonemployed students did not differ in enrollment in evening courses. The presence of individual flexibility for many in scheduling both class and working times was also supported by the responses to the following questionnaire item: “It is difficult to schedule classes so that I can have blocks of time available during the day for other activities”; 47 percent of the sample disagreed. After controlling for the effects of school and semester, part-time employment was not found to have influenced the responses.Google Scholar

63 Weekday employment was the most clearly defined. The other two were unambiguous, but weekend employment appeared to extend into Monday and Friday nights, thereby blurring slightly its distinction from night employment.Google Scholar

64 The distributions could not be statistically tested because of the small cell Ns.Google Scholar

65 A loglinear analysis of the weekday/not weekday employment patterns showed that a saturated model, one including all marginals, was required to account for the distributions. In a saturated model the observed and expected odds are the same.Google Scholar

66 F = 11.94, df = 3,332, p < .001.Google Scholar

67 F = 294.0, df = 1,289, p < .001.Google Scholar

68 The F for school effects was 2.18 and for year, .25. With df= 1,289, p= n.s.Google Scholar

69 The analogy here is to elasticity models in economics. Szalai has cautioned that the analogy not be pressed too far. “Time can only be spent, not ‘earned’; therefore time-budgets have no income side and no budgeting in the strict sense of the word is involved in establishing them.” Alexander Szalai, Differential Evaluation of Time-Budgets for Comparative Purposes, in R. Merritt & S. Rokkan, eds., Comparing Nations: The Use of Quantitative Data in Cross National Research 257 (New Haven, Conn.: Yale University Press, 1966), cited in Philip J. Stone, Models of Everyday Time Allocations, in Alexander Szalai, ed., The Use of Time: Daily Activities of Urban and Suburban Populations in Twelve Countries 179 (The Hague: Mouton, 1972). However, as Stone notes “[allocations of time, from this point of view, can be used to infer priorities of values or strengths of motivations.”Id. at 181 (emphasis in original).Google Scholar

70 Law school legend has it that the appropriate ratio is three hours of study for each hour of class time, but this is undermined by the credo that there is always more to be learned.Google Scholar

71 For this analysis and the balance following in this section, the contextual variables were not collapsed to increase their precision.Google Scholar

72 The control variables were entered into the analysis prior to the covariate so that the effects of work duration could be free of context. An earlier ANCOVA was run on the study and leisure time data using work patterns, school, and semester as main effects with work duration as the covariate. Work patterns proved to have no significant independent effect on the duration of study or leisure time after the effects of work duration were considered. The F value for work patterns and study = .80, df = 3,289, p= n.s., and for work patterns and leisure, F = 2.10, p= n.s.Google Scholar

73 Of course, there is no way to know the exact trade-off for any one individual. The prediction that respondents would study 30 minutes more a week if they worked 1 hour less is based on estimates derived from comparing their average study time with that of persons who did work 1 hour less than them.Google Scholar

74 The other 22 minutes in each hour worked “came from” sleeping, eating, personal and family time, or other “school time.”.Google Scholar

75 The questionnaire included a number of items concerning study and leisure time use. Responses were strongly influenced by year in school and school attended, but in contrast to the findings from the behavioral data on employment and study time, for these attitudinal items part-time employment had no independent effect. The items were:.Google Scholar

1. (strongly agree to strongly disagree) (a) “In order to get my studying done, I have had to organize my time to a greater degree in law school than in college,” (b) “When I am not studying I often have a nagging feeling that I should be,” (c) “Most students here seem to have more free time than 1 have.”.Google Scholar

2. (very frequently to very infrequently) (d) “How frequently have you come to the law school intending to study but have gotten diverted into other activities like card games, general rapping with other students, TV watching, etc., and have studied less than you intended?”.Google Scholar

3. (substantially more to substantially less) (e) “In comparison to the amount of time you spent in leisure and recreation activities prior to entering law school, how much time do you now spend in these activities?”.Google Scholar

4. (Yes/No) (f) “Do you have sufficient time for leisure and recreation?”.Google Scholar

76 Trumbull, William M., Part-Time Legal Education, 14 J. Legal Educ. 349–50 (1962), listed distinctions he believed to exist between evening and day division law students. As they were based on comparisons between the employed and nonemployed they may be considered applicable here as well.Google Scholar

1. Part-time law students evidenced a greater degree of fatigue during class preparation, class attention/discussion, and examination studying;.Google Scholar

2. Part-time law students were not able to carry over their cultural development from college to law school because of a lack of time;.Google Scholar

3. Part-time law students did not have adequate time or outside sources to prepare for class as well as full-time law students (as well as too little time to reflect on, review, and refresh their recollections of the material presented);.Google Scholar

4. Part-time law students did not perform as well academically as did full-time students;.Google Scholar

5. Part-time law students did not have sufficient time or opportunity for independent research; and.Google Scholar

6. Part-time law students were not able to particiate fully in student activities and organizations, e.g. student bar association, moot court, and law review.Google Scholar

77 F = 1.07, df = 1, 1134, p= n.s.Google Scholar

An analysis of covariance was also done using only employed students to determine whether there was a relationship between course load and the weekly duration of working time. In other words, were those students who work longer hours able to do so because they carried lighter course loads? The answer was no. Quantity of working time, when entered as a covariate after the effects of school and semester, showed no relationship to course load (F = .005, df = 1,332, p= n.s.). The same result was found in a separate analysis using the variable of work patterns.Google Scholar

78 Partial betas show the magnitude of influence of each independent variable on the dependent variables net of the effects of the others. The greater the beta value, the greater the independent explanatory power of the variable. In general, semester effects were linear toward increased disengagement by semester. The school effects tended to distinguish between the elite and nonelite law schools in that students at the latter indicated less engagement.Google Scholar

79 E.g., see Cramton as reported in Brickman, supra note 18.Google Scholar

80 Supra notes 14 & 17. Also see Stone, Alan A., Legal Education on the Couch, 85 Harv. L. Rev. 392 (1971).Google Scholar

81 Carrington and Conley implied such a finding. They concluded from their study of Michigan law students that “roughly one in seven … drops out emotionally and intellectually, without formally withdrawing from the school.” These students, identified by an attitudinal inventory, were labeled by the researchers as “alienated.” The authors also reported that the only correlate of “alienation” in their data was with the experience of working while in undergraduate school. They raised a query: “The working student has long been admired as the model of dedication to high purpose; can it be true that such persons are unusually cynical?… Perhaps the findings … suggest that the student who works … is manifesting indifference to the academic enterprise as frequently as he is fulfilling a real economic need. Carrington & Conley, Alienation, supra note 17, at 887, 891.Google Scholar

82 Each item had a loading for every factor which was used in the generation of factor scales, but to simplify the table only the significant single factor loadings are given.Google Scholar

83 The two control variables, semester and school, were entered without collapsed categories.Google Scholar

84 In general, semester effects were linear in the direction of increased negativity by length of time in school. Students in the elite schools tended to be more negative on most scales than those at the nonelite schools. (These data were previously presented in a somewhat different form, see Pipkin, supra note 14.).Google Scholar

85 A separate analysis of covariance was done on the subsample of employed students from which time use data were available (N= 374). The amount of time worked during the surveyed week was entered as a covariate after the main effects of semester and school. The hypothesis tested was whether students who worked greater amounts of time may have been distinguished in attitudes from those working lesser amounts. The covariate had no statistically significant effect on any of the factor scales.Google Scholar

86 Pipkin, supra note 14.Google Scholar

87 Cavers, supra note 47; Zemans & Rosenblum, supra note 11.Google Scholar

88 York & Hale, supra note 45.Google Scholar

89 Pipkin, supra note 14, at 1183–89.Google Scholar

90 Respondents placed their jobs in the law-related category without any inquiry made about job tasks or job titles.Google Scholar

91 It produced a reduction in unexplained LRχ² of 6.90, with a difference of 4 degrees of freedom, p= n.s.Google Scholar

92 To include law school grades in the analysis, first-year students had to be excluded, as they had no or few grades at the time of the survey. The grades variable was constructed by standardizing the grade distributions from each sample within each school and then classifying respondents as either above or below the mean.Google Scholar

93 The deletion of first-year students weakened the year effect substantially and caused the model including year to over-fit the data {FGYS}{YJ} (LRχ²= 21.36, df = 22, p= .499). Although {YJ} had to be included because it was statistically significant, the model of independence {FGYS}{J} fit reasonably well (LRχ²= 27.72, df = 23, p= .227), reflecting the deletion of first-year students from the analysis and the fact that the major shift in job setting type occurred between the first and second years.Google Scholar

94 This, of course, contrasts with the finding from the “objective” data of a limited influence from financial need. See note 31 supra.Google Scholar

95 Note that less than 15 percent saw their job as a leg up into a career, and less than 10 percent saw the job as résumé building. While the limit of two response choices may have constrained the explanation of motives, from these responses it appears that students working in law settings were more likely to have selected their part-time jobs with the idea that it might supplement their education rather than be an early start in a career.Google Scholar

96 F = 70.18, df = 2,428, p < .001.Google Scholar

97 As indicated by this response, working students were aware of the work and study time trades evidenced in the analysis of the time use data.Google Scholar

98 Zemans & Rosenblum, supra note 11, at 137; Baird, Leonard L., A Survey of the Relevance of Legal Training to Law School Graduates, 29 J. Legal Educ. 264, 273 table 3 (1978).Google Scholar

99 For study, F = .154, df = 3,364, p= n.s.; for leisure, F = .025, df= 3,364, p= n.s.Google Scholar

100 This interpretation is necessarily tautological as no separate measures of institutional policies on part-time employment were collected. However, it is supported by Kelso's findings, supra note 9, at 282–83, on differences in faculty's and deans' attitudes toward student part-time employment. Using his classification of law schools by resource levels (which has some parallel in this study)—“A” through “C” identifying schools with descending levels of resource—he found that levels of employment were much higher at “B” and “C” schools than at “A” schools (supra note 20). The degree to which faculty and deans saw positive consequences from student part-time employment varied directly by levels of school resources and thus by levels of student employment.Google Scholar

101 Had this study not included the elite law schools, which are clearly atypical of law schools in genera], the independent effects of school type on part-time employment would have been less clear. The contrasts between the public and private regional schools, while present, were not as sharp.Google Scholar