1 433 U.S. 350 (1977).Google Scholar

2 Id. at 384. Before Bates, virtually all state bars regulated the advertising of their members by adopting the rules of the American Bar Association's Code of Professional Responsibility. Canon 27 of the ABA's former Canons of Profession Ethics began the nationwide ban on lawyer advertising, although many state and local bars had outlawed advertising earlier. See Julius Henry Cohen, The Law: Business or Profession? (rev. ed. New York: G. A. Hennings Co., 1924). The following version of Disciplinary Rule DR 2-101(B), enacted in substantially the same form in all jurisdictions, was in effect immediately before the 1976 and 1977 amendments:Google Scholar

A lawyer shall not publicize himself, or his partner, or associate, or any other lawyer affiliated with him or his firm, as a lawyer through newspaper or magazine advertisements, radio or television announcements, display advertisements in city or telephone directories, or other means of commercial publicity, nor shall he authorize or permit others to do so in his behalf.Google Scholar

Despite consumer group efforts, some pressure within the bar to modify the ban, and legal precedent indicating that the ban was in serious question (see Goldfarb v. Virginia State Bar, 421 U.S. 773 (1975), and Virginia Pharmacy Bd. v. Virginia Consumer Council, 425 U.S. 748 (1976)), the organized bar maintained the ban. In early 1976, the ABA approved only a modest change in the Code of Professional Responsibility to allow listing of “brief biographical and other informative data” in the yellow pages of the phone directory or in “reputable” directories. Prior to Bates, few states adopted even this “reform.”Google Scholar

3 433 U.S. at 366-67. Although the Court suggested that regulation of the forms of advertising and other promotion listed in the text accompanying this footnote may be permissible, it also indicated that the legality of these forms was beyond the holding of its opinion. Last term, the Court ruled that the bar could, with state authorization, constitutionally prohibit personal solicitation of clients for pecuniary gain, even in circumstances where there was no proof of actual harm to the client. Ohralik v. Ohio State Bar Ass'n, 436 U.S. 447 (1978). Solicitation by attorneys for nonprofit organizations, however, enjoys greater constitutional protection and can only be restricted upon an actual showing of harm. In re Primus, 436 U.S. 412 (1978). For the bar's response to Bates, see notes 4, 6 infra. Google Scholar

4 Responding quickly to Bates, on August 10, 1977, the ABA House of Delegates approved the report of the Board of Governors concerning lawyer advertising. The report recommended that the Code of Professional Responsibility be amended as provided in proposal A (explained below) and that proposals A and B be circulated to the states for their consideration. Proposal A is “regulatory” in nature, specifically authorizing only certain prescribed forms of lawyer advertising. In contrast, proposal B is “directive,” allowing publication of all information not “false, fraudulent, misleading or deceptive.” Proposal A retains many of the present disciplinary rules that specify the categories of information that can be published, such as name, field of law practice concentration, education, and client references. Proposal B does not list specific items approved or disapproved for advertising, instead adopting the general antifraud standard. Neither proposal permitted television advertising unless “the appropriate state authorities [determine] that it is necessary to provide adequate information to consumers of legal services.” But see note 6 infra. American Bar Association, Task Force on Lawyer Advertising, Board of Governors, Report 177B (Aug. 4, 1977). For a reprint of the report and the proposals, see 827 Antitrust Trade Reg. Rep. (BNA) §F (Aug. 1977).Google Scholar

As to the reaction of the states:Google Scholar

As of July 15, 1978, 29 jurisdictions have made some kind of changes to their Code provisions relating to lawyer advertising. Some permit print and newspaper advertising only, such as: Connecticut, Delaware, Idaho, Iowa, Louisiana, Missouri, New Mexico, Oklahoma, Rhode Island, Utah, Vermont; some include radio along with print media and newspapers, such as: Indiana, Nebraska and Wyoming; and some include television, with a variety of specific limitations on what or who can be shown or who and how the announcing can be done, such as: Colorado, Georgia, Kentucky, New York, North Carolina, Ohio and Tennessee. Then there are the extremes: Mississippi and Alabama allow newspaper advertising only, whereas, D.C., Maryland, Michigan and Minnesota merely prohibit false, misleading, deceptive or fraudulent advertising. And the Wisconsin and New Jersey Supreme Courts have suspended provisions which would unduly curtail or limit truthful advertising by lawyers. The District of Columbia's rules specifically deal with the question of solicitation.Google Scholar

Informational Report of ABA Commission, at 6 (Aug. 1978).Google Scholar

5 BNA, supra note 4, at §A. See also the brief of the Justice Department's Antitrust Division opposing the proposed Florida advertising regulations, discussed in Fla. B. News, Oct. 25, 1978, at 1, cols. 3-5. The FTC is currently investigating certain activities of the legal profession, including efforts to restrict advertising and the growth of legal clinics.Google Scholar

A legal standard permitting “full-scale advertising” refers to one that treats legal advertising by the standards applied to advertising of other businesses. If it permitted mass media advertising, proposal B in note 4 supra would approximate this standard.Google Scholar

6 Further, only 22 states permit radio advertising. See ABA Commission on Advertising (Feb. 1978). At the August 1978 ABA meeting, the ABA Commission proposed, and the ABA House of Delegates approved, television advertising. No other changes were made, however, leaving the restrictive proposal A (see note 4 supra) as the officially approved regulation.Google Scholar

7 For example, a common justification for keeping fees high through minimum fee schedules was avoiding poor quality. See Note, A Critical Analysis of Bar Association Minimum Fee Schedules, 85 Harv. L. Rev. 971, 982-85 (1972). The legal literature contains numerous other statements anticipating loss in quality from the lower prices that may result from advertising. See, e.g., ABA Commission on Professional Ethics, Op. No. 302 (1961); Transcript of Bates hearing at 378-79; Brief of ABA in Bates, at 17; and Terence A. Shimp, Ohio Lawyers' Attitudes Toward Legal Service Advertising, 4 Ohio N.U.L. Rev. 576 (1977). There is less agreement on the likely effect of advertising on the prices of legal services. A survey of Ohio lawyers revealed that those attorneys expect advertising to lead to higher prices. Id. On the other hand, in a national survey, 61 percent of respondents disagreed with the statement that advertising would raise prices. See Law-Poll: Is Advertising Laying an Egg? Lawyers May Be More Interested in Solicitation, 64 A.B.A.J. 673 (1978).Google Scholar

8 This article focuses on the sale of services to individuals, not corporations. Attorneys who provide services to corporations, particularly to large ones, often do not compete with those selling services to individuals. In addition, the arguments herein are not generally applicable to services for those who are eligible for free services.Google Scholar

9 See Bureau of Census, U.S. Department of Commerce, 1972 Census of Selected Service Industries: Legal Services [hereinafter cited as 1972 Census].Google Scholar

10 In 1972 sole practitioners received about $2.1 billion from individuals while partnerships received about $2.6 billion. Id. Google Scholar

11 See Bureau of Census, U.S. Department of Commerce, 1967 Census of Business, Selected Services: Law Firms.Google Scholar

12 See 1972 Census, supra note 9. Specialist was not defined; respondents were simply asked whether they considered themselves specialists.Google Scholar

13 Id. This use seems “relatively” small when compared with the use of paralegals by the legal clinic studied in sec. II infra. As with specialist, paralegal was not defined. For a definition of paralegal, see note 22 infra. Further, it is not possible to tell from this data whether some attorneys use secretaries as paralegals but did not report such use.Google Scholar

14 See, e.g., Edward Hastings Chamberlin, The Theory of Monopolistic Competition (Cambridge, Mass.: Harvard University Press, 1933).Google Scholar

15 Since many forms of advertising do not require the personal appearance of the attorney, such advertising allows the firm to reach more consumers than would personal solicitation. Of course, advertising is not necessarily more cost efficient to attorneys than personal solicitation. See notes 36-37 infra and accompanying text.Google Scholar

16 All else may not be equal, as we discuss at text accompanying notes 26-31 infra. Google Scholar

17 See Armen Alchian, Costs and Outputs, in William Breit & Harold Hochman, eds., Readings in Microeconomics 159 (2d ed. New York: Holt, Rinehart & Winston, 1971); Jack Hirshleifer, The Firm's Cost Function: A Successful Reconstruction? 35 J. Bus. 235 (1962); Walter Y. Oi, The Neoclassical Foundations of Progress Functions, 77 Econ. J. 579 (1967).Google Scholar

18 Hirshleifer, supra note 17, at 239.Google Scholar

19 Further, the increase in planned volume may be for one service only, allowing the absolute volume and size of the firm to decrease. For example, if a firm specializes in simple divorces, with a sufficient planned volume of such divorces it may have lower costs than firms with larger volumes consisting of a multitude of different cases.Google Scholar

20 See Lester Brickman, Expansion of the Lawyering Process Through a New Delivery System: The Emergence and State of Legal Paraprofessionalism, 71 Colum. L. Rev. 1153, 1211 (1971).Google Scholar

21 Id. at 1213.Google Scholar

22 A paralegal works under the supervision of a lawyer, performing routine legal duties that do not require full legal training.Google Scholar

23 See Altman & Weil, Inc., The 1976 Survey of Law Firm Economics 9 (Ardmore, Pa.: Altman & Weil, Inc., 1976).Google Scholar

24 The data in the text are derived from estimates in Project, The Unauthorized Practice of Law & Pro Se Divorce: An Empirical Analysis, 86 Yale L.J. 104 (1976) [hereinafter cited as Unauthorized Practice]. The data are suggestive only; they do not represent a systematic survey of the profession. For very simple cases (involving almost no negotiation), some lawyers spend as much as 55 percent of their time in client activity and form preparation. Id. For additional data on the time lawyers spend on a divorce, see George F. Cole & Howard L. Greenberger, Staff Attorneys vs. Judicare: A Cost Analysis, 50 J. Urban L. 705 (1973); Richard J. Arnould & Robert N. Corley, Fee Schedules Should Be Abolished, 57 A.B.A.J. 655 (1971) [hereinafter cited as Fee Schedules].Google Scholar

The divorce business is not trivial: in the 12 months ending on June 30, 1976, there were over 1,000,000 divorces in the United States. See Unauthorized Practice, supra, at 110 n.26. Further, one study revealed that fully one-third of the divorces sampled did not involve children or awards of property or alimony. Id. at 160. In these divorces, some lawyers spend over one-half of their time in activity that could be delegated to paralegals, as the text following this footnote indicates.Google Scholar

25 This summary is drawn from David U. Strawn, Legal Aid for Fun and Profit, 12 Law Off. Econ. & Management 23 (June 1971).Google Scholar

Of course, if lawyers or clients prefer, some client contacts will involve lawyer time and thus not be delegated. Assuming that the lawyer could productively spend time elsewhere, the more time the lawyer spends with the client, the higher the fee will have to be to meet expenses.Google Scholar

26 See sources in note 17 supra. Again, all else must be constant.Google Scholar

27 See sources cited in note 17 supra. Google Scholar

28 See Louis De Alessi, Costs 5-6 (working paper, Aug. 1978).Google Scholar

29 This quotation is usually attributed to Chesterfield Smith, former president of the American Bar Association.Google Scholar

30 As applied to legal services, the possibility of having to use lower quality inputs is a major source of potential increased costs from a faster production rate in adopting the different production techniques discussed in text accompanying notes 20-25 supra. For example, a faster rate of production may require employees who are tired and hence less productive to work longer hours. Further, congestion costs could increase because in the short run the firm has a relatively fixed entrepreneurial capacity with which to manage the increased output. Although the headaches of managing and controlling a greater output are not trivial, as the text notes they may well be insignificant when compared with the cost savings available. Again, the ultimate weighing of the different effects on cost are only performed empirically. We present some empirical evidence on this issue at text accompanying notes 74-75 infra. Google Scholar

As an additional element of cost, one might inquire whether advertising raises per-unit costs. In some cases, however, advertising may be a lower cost (per unit sold) form of solicitation than methods currently used. See text accompanying notes 36-37 infra. Google Scholar

31 This is true only if all else is equal. Among those things that must be held equal is the risk that some output may not be sold. The probable increase in the expected sales may be accompanied by an increased variance in sales, such that lower production costs are offset by higher uncertainty costs. Advertising, however, has in general been found to be a useful way of reducing risk. See, e.g., S. A. Ozga, Imperfect Markets Through Lack of Knowledge, 74 Q.J. Econ. 29 (1960).Google Scholar

32 Again, all else must be equal. For the effect of possible shifts in demand, see note 34 infra. As explained there, even with demand shifts, consumers will still be better off with the lower costs.Google Scholar

33 This proposition is demonstrated in nearly every introductory economics textbook.Google Scholar

34 See George J. Stigler, The Economics of Information, 69 J. Pol. Econ. 123 (1961). This proposition is necessarily true only for the full price of a commodity. The full price equals the cost expended in searching out the item purchased plus the actual purchase price. A reduction in search costs may increase the demand for the product at the current purchase price, tending to raise the purchase price (unless per-unit costs do not rise as quantity increases, as is certainly possible). Even if the purchase price rises, however, the consumer will pay a lower full price if advertising yields a more than offsetting reduction in the cost of search.Google Scholar

To illustrate these points, consider a commodity whose full price to an individual consists of $1.00 spent on search plus a $3.00 transaction price. If advertising reduces the search cost of finding that item to $0.20, the consumer could pay up to $3.79 for the item and still be better off. Thus, the purchase price could increase, and the consumer could still prefer this situation to the previous state.Google Scholar

35 These methods are not always “ethical.” See, e.g., Jerome E. Carlin, Lawyers' Ethics (New York: Russell Sage Foundation, 1966), and his Lawyers on Their Own (New Brunswick, N.J.: Rutgers University Press, 1962).Google Scholar

36 Another reason why price might drop is that the previous price was a cartel price and advertising allows firms to inform consumers that prices will be lower. Thus, the price would drop because it was set above cost, not because costs dropped. This explanation, however, does not appear to capture the effects of the advertising that lawyers in fact are employing. See note 73 infra. It is possible that the effect of any cartel is to force competition into less efficient means. Thus, costs would drop with the elimination of the enforcement mechanism of limits on advertising. For example, advertising may be a more efficient form of publicity, as explained in text accompanying note 35 supra. Google Scholar

37 See e.g., Lee Benham, The Effect of Advertising on the Price of Eyeglasses, 15 J. Law & Econ. 337 (1972). John F. Cady, Drugs on the Market: The Impact of Public Policy on the Retail Market for Prescription Drugs (Lexington, Mass.: D.C. Heath & Co., 1975).Google Scholar

38 See note 7 supra. Google Scholar

39 See Robert B. Yegge & Eli Jarmel, New Careers in Law 13 (Chicago: ABA Special Committee on Legal Assistants, 1972).Google Scholar

Economists have also suggested that advertising will increase quality because it establishes “brand name” capital. Advertising increases consumer identification of the firm name with its products, and thereby penalizes substandard products of a firm that advertises more heavily than those of a nonadvertising firm because customers more easily remember the shortfall in quality. Accordingly, heavily advertised products should be of superior quality. See Phillip Nelson, Information and Consumer Behavior, 78 J. Pol. Econ. 311 (1970), and his Advertising as Information, 82 J. Pol. Econ. 729 (1974).Google Scholar

The only data and they are sparse on quality in cases of advertising bans indicate that there are no systematic quality differences. See John F. Cady, Restricted Advertising and Competition: The Case of Retail Drugs 11 (Washington, D.C.: American Enterprise Institute, 1976); Alex Maurizi & Thom Kelly, Prices and Consumer Information: The Benefits from Posting Retail Gasoline Prices 5-6 (Washington, D.C.: American Enterprise Institute, 1978) (summarizing evidence on eyeglasses).Google Scholar

40 Another reason is the constraint on prepaid legal services, also a source of larger planned volume. The organized bar has long sought to restrain prepaid services. The earliest opposition came from state bar associations' prohibitions against the “unauthorized” practice of law. This line of defense collapsed after a series of Supreme Court decisions holding these prohibitions unconstitutional to the extent that they acted against legal referral services or prepaid plans. See United Transp. Union v. State Bar, 401 U.S. 576 (1971), and cases discussed therein. Following these decisions, the bar's tactics appeared to switch to favoring open-panel plans (which are more like traditional firms) over closed-panel ones (many of which resemble the legal clinic form discussed in sec. II). See Hearings on Recent Developments in Prepaid Legal Insurance Plans Before the Subcommittee on Representation of the Senate Committee on the Judiciary, 93d Cong., 2d Sess. (1974). In the past few years, most, although not all, barriers to closed-panel plans have been removed. Except for those established by unions and employers, however, these plans still face considerable legal uncertainty.Google Scholar

41 Mayer notes of such routine divorce cases that “some lawyers handle them on a volume basis, and have a couple of hundred cases pending all the time.” Martin Mayer, The Lawyers 49 (New York: Dell, 1968).Google Scholar

42 For a description of lawyers' activities in more complex divorce cases, see Carlin, Lawyers on Their Own, supra note 35, at 91-101.Google Scholar

43 David M. Engel, The Standardization of Lawyers' Services, 1977 A.B.F. Res. J. 817, 819-26.CrossRefGoogle Scholar

44 Id. at 825-26 states that the use of fixed fees is only a weak indicator of standardization. His conclusion is questionable, in part because of his assertion that fees may vary for reasons other than economic ones. If fees do vary for such reasons, then this would indicate only that services for which fees are not fixed might also be standardized, not that fixed fees are inconsistent with standardization. Engel also argues that even if fees are fixed, “the possibility would still exist that some clients who demanded little time and effort from their lawyers were, in effect, subsidizing more complicated tasks performed for other clients.”Id. at 826. It is dubious that many clients are routinely provided services at a price set to subsidize the provision of services to others. Most economists disagree with the argument that there is systematic cross-subsidization, since the disfavored consumers could simply shop for a lower price elsewhere, destroying the ability to cross-subsidize. Although advertising will facilitate such destruction, even without advertising systematic cross-subsidization will be difficult to sustain. (Of course, lawyers may sometimes reduce their fees, for example in pro bono work or to clients from whom they expect future business. As with other businessmen, lawyers may use lower prices as a form of promotion.)Google Scholar

45 It might be suggested that the bar associations' minimum fee schedules (struck down in Goldfarb v. Virginia State Bar, 421 U.S. 773 (1973)) support the hypothesis that individual services are more standardized, since services subject to minimum fees were invariably those sold to individuals. This test of standardization relies on the assumption that the minimum fee was, in fact, the charge made for the service, an assumption that was not always true. See Fee Schedules, supra note 24, and discussion in note 57 infra. Where minimum fees were mandated, some lawyers actually charged prices above or below the minimum, reflectiing, inter alia, differences in the service provided. A better test of standardization would examine the services that lawyers themselves felt to be routine, as evidenced by their willingness actually to quote a fixed fee in advance of their performance.Google Scholar

46 A final reason that the low-cost techniques could not be employed is that the volume required is not obtainable even by advertising. Although this is an empirical issue to be addressed in sec. II infra, the argument appears meritless. For example, all that would be required to employ a paralegal specializing in divorce would be a guarantee of 40 or so hours a week of routine divorce work, hardly an unobtainable amount. Although, as the evidence in sec. II infra indicates, the increased volume from advertising far exceeds that necessary to use a single paralegal, even such a small change in the techniques of providing divorce services could significantly reduce costs, given the large number of divorces anually. See note 24 and accompanying text supra. Google Scholar

It is worth noting that this article does not attempt to specify the precise volume that an individual law firm must attain to adopt all, or some combination, of the cost-saving techniques discussed at text accompanying notes 20-25 supra. Instead, we are presenting (inter alia) (a) the theoretical explanation as to why advertising will lead to the adoption of such changes and (b) empirical evidence testing that theory.Google Scholar

47 Although the clinic did not legally advertise, it engaged in activities that had an identical purpose and effect. When the clinic opened in Los Angeles in 1972 (five years before Bates) Leonard Jacoby and Stephen Meyers staged an “open house” at the request of, and in conjunction with, the Consumer Legal Action Council. Members of the news media were present and received information concerning the clinic, including fees to be charged. As a result, the clinic received extensive radio, newspaper, and television coverage, which helped generate a substantial amount of business. In March 1973, the state bar accused Messrs. Jacoby and Meyers of unprofessional conduct both for advertising and for using the term “clinic.” In response, Jacoby and Meyers organized and distributed a press kit with their rebuttal to the charge. Again, the clinic recevied considerable publicity from the media, in effect obtaining more free advertising. At the administrative hearing, Jacoby and Meyers were found to have advertised, and the use of the word “clinic” was held to be misleading. Finding that a substantial portion of their conduct was principally directed toward generating business, the board concluded that Jacoby and Meyers could be disciplined. The California Supreme Court reversed the board, holding that the clinic could not be constitutionally prohibited from cooperating in the publication of a newsworthy article even when the attorneys themselves were the subject. It further held that the use of “clinic” was not misleading. See Jacoby v. State Bar, 19 Cal. 3d 359, 138 Cal. Rptr. 77, 562 P.2d 1326 (1977).Google Scholar

48 Luck, Timing and People Create Legal Clinic Success, 2 New Directions in Legal Services 115, 118 (July-Aug. 1977). Mr. Sharrow has since left the clinic to work in prepaid legal services.Google Scholar

49 All figures for Jacoby & Meyers are from a list of questions that the clinic answered in the summer of 1978. For data on receipts per lawyer for traditional firms, see note 72 infra. Google Scholar

50 For a report on the television advertising campaign of the clinic, which will cost about $300,000 in 1978, see 13 Media Decisions 66 (No. 7, 1978). Clinics in Baltimore, Cleveland, Cincinnati, Houston, and Phoenix also advertise on television. Id. at 96.Google Scholar

51 For traditional firm data, see note 13 and accompanying text supra. Although traditional firm data come from an earlier year than the clinic data, there is no reason to believe that in the interim traditional firms have had a tenfold increase in use of paralegals per attorney.Google Scholar

52 See note 48 supra. Google Scholar

53 Even the average sole practitioner, who relies on individuals more than other traditional firms, receives only 71 percent of his receipts from individuals. See note 9 and accompanying text supra. Google Scholar

54 Stuart Auerbach, The Case for Lawyers' Advertising: It Wins Clients, Washington Post, June 20, 1978, at 1, col. 2.Google Scholar

55 See 1975 Economic Survey of the Maryland State Bar Association, Md. B.J. (June 1976).Google Scholar

56 In 1970, the average minimum fee for an uncontested divorce in seven northern California counties was $270. (We omitted Sacramento figures because the fee of $995 was triple that of any other city. Inclusion of Sacramento would of course only raise the average.) See Ann Nichols, The Pricing of Physician and Lawyers Services 386 (Ph.D. diss., Colorado State University, 1975). In that same year, the minimum fee in Inglewood, the city in Los Angeles County where Jacoby & Meyers has an office, and in which the clinic competes, was $350. See ABA Standing Committee on the Economics of Law Practice, Minimum Fee Schedules, 36 (appendix C) (1970). Over two years later, the clinic's fee was only $100. Although, as explained in note 45 supra, not all attorneys followed the fee schedule, one survey indicated that 31.3 percent always used the divorce fee, and 31.7 percent sometimes used it. See Fee Schedules, supra note 24, at 660.Google Scholar

57 The sample was chosen from the mailing list of Jacoby & Meyers. About 50 percent of those on the list have used the clinic only to consult a lawyer briefly for a small fee (usually $15) and thus did not use the clinic for any specific matter such as a divorce, will, or real estate trans-action. This method of selecting the sample should not bias the results. Consumers were unaware of the specific reason for the survey and were not asked to distinguish between traditional firms and the legal clinic. Instead they simply identified the lawyer who had provided the service, allowing us to separate clinic clients from customers of traditional firms. We prepared the questionnaires, but the firm that maintains the clinic's mailing list mailed them to every twentieth name on the alphabetical master list. We thus maintained client confidentiality, since we never learned the names or addresses of those to whom questionnaires were mailed.Google Scholar

Of about 650 questionnaires received (700 were mailed, but 42 were returned as undeliverable), 94 (about 14 percent) were returned completed. Of these, 74 were usable, 52 from users of traditional firms and 22 from clinic clients. The main reason for responses being unusable was failure to provide the name of the attorney, thus not allowing us to distinguish whether the lawyers in those cases were from the clinic or from traditional firms.Google Scholar

Since respondents to mail surveys are often not representative of the population as a whole, see, e.g., Paul L. Erdos, The Nonresponse Problem, ch. 15 in Professional Mail Surveys (New York: McGraw-Hill Book Co., 1970), the question arises as to whether the survey technique biases the results. There are at least two reasons that the results obtained are still valid for considering whether the clinic has better or poorer quality, subjectively measured. First, the survey compares the two groups surveyed (clinic consumers and traditional firm consumers) and does not just attempt to generalize from those surveyed to the population. Since there appears to be no bias due to differences between the groups surveyed, see notes 62-64 and accompanying text infra, the question then becomes whether any differences between those sampled and the actual population are relevant. Assuming that such differences exist and that a survey would show a majority of the population preferring traditional firms to the clinic, this would not necessarily support banning full-scale advertising. If only some members of the population prefer the clinic's services to those of traditional firms, then consumers as a class are better off with the addition of the clinic alternative. Since the clinic requires advertising to provide its low-cost, high-volume service, the case for advertising would be made even under the restrictive assumptions.Google Scholar

Second, the respondents to our survey may be representative of the appropriate population. The national population of users of legal services is identified in the Legal Needs study report (see Barbara A. Curran, The Legal Needs of the Public: The Final Report of a National Survey (Chicago: American Bar Foundation, 1977) [hereinafter cited as Legal Needs study]. Although it is impossible to tell for sure without the underlying Legal Needs study data, which are as yet unavailable, it appears that individuals in our sample are more heavily concentrated in the 25-44 age bracket, better educated, and wealthier than the national population of users of legal services. The appropriate population for our survey differs, however, from the national population in that (a) our sample was drawn from heavily urban Los Angeles and (b) the relevant population has a higher percentage of cases dealing with routine services than the national population since clinics concentrate on such services as simple divorce. (For example, at least half of clinic work involves divorces, see note 53, supra, while roughly 20 percent of those surveyed in the national survey purchased divorces, see Curran, supra, at 196-99.) Since divorce consumers appear to be more concentrated in the 25-44 age group than spread throughout the population (id. at 195), and since individuals in Los Angeles are better educated and have higher incomes than the nation as a whole, our survey may not differ significantly from the population appropriate to our more limited inquiry. See U.S. Dept. of Commerce Bureau of the Census, Statistical Abstract of the United States 123 (1976), and County & City Data Book: A Statistical Abstract Supplement 44-52, 233-34 (1977).Google Scholar

58 For the results of the national survey and the full questionnaire, see Curran, supra note 57. Because the questionnaire used in our survey was based on the related parts of the form used in the Legal Needs study, it is interesting to compare the results of the two inquiries. Our respondents as a group rate traditional firms (as well as clinics) from good to fair, while the more comprehensive Legal Needs study found law firms generally rated as excellent to good. Curran, supra note 57, at 210.Google Scholar

Although again it is impossible to know for certain without the underlying data, two reasons appear to explain the difference. For one, our sample contained a higher percentage of divorce and tort cases than the national sample, and consumers in these cases were more critical of their attorneys than consumers in other cases. Id. at 213. (Nevertheless, the divorce respondents in the Legal Needs study rated their attorneys more favorably than did our divorce respondents.) Second, and more importantly, we used a mail survey, while the Legal Needs study report involved a personal interview. Mail surveys differ from personal interviews in that mail respondents feel more strongly about the questions than do nonrespondents and accordingly our respondents may have been more critical. See Erdos, supra note 57. (See note 57 supra as to why the mail survey technique does not bias the results.)Google Scholar

Further, there were several similarities between the Legal Needs study and our survey. Our results exhibited characteristics similar to those of the national sample. For example, within our responses, divorce and tort answers were less favorable, just as in the national survey, and the consumers whom we surveyed gave lowest ratings to lawyers for “keeping them informed” of the matter and the highest for “honesty,” again just as in the national survey.Google Scholar

59 All tests for significance are one tailed. When using t-tests, the researcher must choose between one-tailed and two-tailed tests. When the researcher has a theoretical reason for hypothesizing the direction of any difference, one-tailed tests are appropriate; when he is testing for a difference without hypothesizing a priori that one mean will be larger or smaller, two-tailed tests are appropriate. Since the theory of sec. I indicates that clinics have, if anything, better quality, one-tailed tests are used. A t-statistic that is significant at a particular level in a two-tailed test will be significant in a one-tailed test at a level one-half of the two-tailed level, Thus, a number significant at .10 on a one-tailed test is significant at .20 on a two-tailed test.Google Scholar

60 Most statisticians adopt the convention that rejection of a null hypothesis (here, that the clinic and traditional firms are of equal quality) is warranted when the significance level is no higher than .05. In other words, one would accept the alternative hypothesis that clinics are of superior quality when the probability of a difference caused by chance is less than or equal to I in 20 (.05).Google Scholar

A potential problem in interpreting our results is whether consumers rated the clinic better on all categories solely because the clinic charges lower fees. This does not appear to be the case. First, since only one of the clinic respondents had seen more than one lawyer concerning the type of service for which he rated the clinic and only three had ever seen any lawyer other than one in the clinic, most clinic consumers did not have experience with traditional firms. Second, the consumer did not know that the questionnaire would be used to compare clinics with traditional firms. Third, the questions asked for absolute (“Was he honest in dealing with you?”) rather than relative responses.Google Scholar

61 The binomial test is appropriate when the population is perceived as divided into two classes according to hypothesized probabilities. The test tells whether it is reasonable to believe a particular sample was drawn from the population as hypothesized. See Sidney Siegel, Nonparametric Statistics 36-42 (New York: McGraw-Hill Book Co., 1956). Here, the binomial test suggests that it is highly unlikely that the clinic and traditional firms sampled come from a population where the two types of firms provide equal quality. The binomial test and the t-test do not analyze identical phenomena. The former is a “goodness-of-fit” comparison between the sampling distribution and the hypothesized distribution of the population. The latter measures possible differences in the means of two sampling distributions.Google Scholar

62 The Legal Needs study survey indicated that racial minorities tended to rate lawyers less favorably than did nonminorities, and that women rated lawyers more favorably than did men. We use the income and education variables as rough approximations of minority status.Google Scholar

63 To control for the four variables, the sample was divided into two groups for each variable: (1) service divorce or tort and real estate or will; (2) sex male and female; (3) income less than $15,000 per year and $15,000 or greater per year; and (4) education high school diploma or less and junior college degree or higher. Of these eight groups, the clinic's overall rating was better with seven. Only those with less than $15,000 income rated traditional firms higher than clinics, with the difference significant at only .40, or, in other words, insignificant by the convention of statisticians. On the other hand, of the seven groups that preferred the clinic, one (those with more education) had a difference significant at less than .10, one (males) a difference significant at less than .05, and one (higher incomes) a difference significant at less than .01.Google Scholar

64 Lacking precise knowledge of the characteristics of the underlying population for consumers of individual services in Los Angeles, we used equal samples as an approximation.Google Scholar

Besides sex, the other distributions were 50 percent with incomes of $15,000 or more and 50 percent below $15,000; 50 percent with more than a high school diploma and 50 percent with a high school diploma or less; and 50 percent divorce or tort and 50 percent other legal matters. Equal samples appear to be a correct approximation as to sex and education. See Curran, supra note 57, at 186-89, and the discussion in note 57 supra comparing our sample to the appropriate population. With the income variable, to the extent that equal samples do not reflect the appropriate population, any bias appears to understate the clinic's superiority. In the national population, although it is impossible to state precisely without the underlying data from the Legal Needs study survey, it appears that more than 50 percent have family incomes in excess of $15,000 per year (measured in 1978 dollars). See Curran, supra note 57, at 70-72, 189. (On possible differences between the national population and the population from which the Los Angeles clinic draws clients, see note 57 supra.) Only as to the service variable is the clinic's superiority perhaps overstated. Since clinics concentrate on divorce work, more than 50 percent of the relevant population may see the clinic for divorce or tort. However, even if 75 percent of the consumers see the clinic for this work, the clinic would still be rated superior on six of the seven questions.Google Scholar

65 We focused on so-called initial orders to show cause. Based on our conversation with both traditional firm and clinic lawyers and our review of the records in the cases that we used, there are several relevant facts about the hearings. First, besides child support, spousal support is frequently at issue. Second, disposal of disputed property is not normally at issue, although the judge will often know that the parties before him have property that must be disposed of at a later proceeding which will officially dissolve the marriage. Third, although visitation and child custody may be at issue, we attempted to avoid these disputes by taking only cases where the wife retained custody and visitation did not appear to be at issue. Fourth, the attorneys and the judge may examine and cross-examine the parties. Fifth, although the parties are allowed to file briefs, they apparently rarely do. Finally, as discussed more fully in note 67 infra, the court publishes a schedule of recommended awards for initial orders to show cause. Presumably those cases that are actually litigated, rather than settled using the guidelines, involve what one of the parties considers to be unusual financial requirements. The guidelines state that such requirements will be considered in determining the proper award.Google Scholar

66 A random sample was taken of the public records in divorce cases filed in the downtown branch of the Superior Court of Los Angeles County, yielding a list of divorces for which only lawyers from traditional firms were retained. Jacoby & Meyers provided a list of file numbers for their cases in which support was litigated in the Los Angeles court. Information was then gathered for each case from court records on the variables mentioned in the previous paragraph of the text, all of which arguably influence the size of the judge's award of child support. In this manner we developed a sample of 58 cases, of which 17 involved the clinic (with the clinic representing the wife in 11 cases). In one case in our sample, the husband's financial form was officially submitted a few days after the adjudication, which presumably was based on an informal submission of the same information concerning the husband's finances.Google Scholar

67 According to the recommended awards of the Superior Court of Los Angeles County that became effective July 1, 1977, if the husband's monthly income (net of deductions) is $1,000, his wife should receive $450 with one child, $475 with two children, and $500 with three children. If net monthly income is $2,000, a spouse with one child should receive $800, with two children $900, and with a third child $1,000. Because this relationship between the number of children and the award is nonlinear, for this variable we used the logarithm (base 10) of the number of children.Google Scholar

68 Weighted least squares were used to correct for heteroscedasticity. See J. Johnston, Econometric Methods 214-21 (2d ed. New York: McGraw-Hill Book Co., 1972).Google Scholar

69 In other words, results as extreme as those observed could have occurred by chance (rather than because of any systematic relationship between the particular independent variable and the monthly child support award) with a probability of less than 0.5 in 100, 3 in 100, and 5 in 100 for the three variables mentioned.Google Scholar

The insignificance of the husband's expenses and wife's income variables is not surprising, particularly given that the guidelines for initial orders to show cause imply that the number of children and the husband's income will be the most important determinants. As to the husband's expenses, there are at least two reasons why judges might not consider them to be crucial. First, the expenses may not reflect the husband's current expenses since the husband will often have recently left home and may not have had sufficient time to determine his new pattern of expenses. Second, in many of the cases that we studied, the husband included expenses that the wife also included. Presumably, many of these expenses would not continue, for example, those involving the children who no longer reside with him. As to the wife's income, the guidelines indicate a desire for not discouraging wives from working, and this may influence the importance of the income variable for determining child support.Google Scholar

70 See note 60 supra. Google Scholar

71 It must also be considered whether there are other explanations for the results than chance or the superiority of the clinic. One such explanation is that judges feel “sorry” for clinic consumers because they cannot “afford” other attorneys. Even if true, this would indicate that consumers would be better off with clinic representation. Further, there is no reason for the judges to regard clinic consumers as downtrodden. Our sample of clinic cases involved family incomes only about $250 a month less than the $1,800 monthly total of those using traditional firms. Finally, given that the bar has investigated and prosecuted the clinic, one could argue that judges might be prone to be more hostile toward the clinic, not less.Google Scholar

Another possible explanation for the results is that we did not account for the trade off that judges may make between child support and other contested items in the hearing. See note 65 supra. Spousal support is one such item. If clinic cases did not involve spousal support awards while traditional firm cases did, our results might be misleading if judges systematically award less child support when they also award spousal support. To test for this possibility, we added a variable as to whether or not spousal support was contested. This new variable was insignificant, while the clinic variables retained the same signs and significance as in table 2. Further, in cases where spousal support is sought, perhaps judges may systematically reduce child support awards in traditional firm cases more than in clinic cases. We conducted a separate test of those cases in our sample where spousal support was not at issue, again finding that the variable for clinic representation of the wife was statistically significant. The variable for clinic representation of the husband was so insignificant that it did not affect the child support award sufficiently for its influence to be calculated. Finally, we considered whether knowledge that the court must eventually allocate property at issue between the parties would affect the child support award. A variable to indicate whether such knowledge existed was insignificant, with the significance of clinic variables again unaffected.Google Scholar

72 See text accompanying note 49 supra. Adjusted for an annual inflation rate of 7 percent, the receipts-per-lawyer ratio of traditional firms that sell predominantly individual services would now be $77,000. See Altman & Weil, supra note 23. The ratio for the clinic is over $80,000, despite the fact that the clinic's fees appear to be at most one-half of those of its traditional competitors producing the same services (see notes 55 and 56 supra and accompanying text). If we assume that the two types of lawyers work the same number of hours, the clinic has apparently succeeded in considerably reducing lawyer time per case as compared with the time expended in traditional firms.Google Scholar

73 Of the reasons given in sec. I as to why prices might drop, only a drop in costs would seem to explain both better quality and increased specialization. If prices are said to drop simply because of cheating on a cartel (see note 36 supra), this does not explain the increased specialization of the clinic.Google Scholar

74 See note 29 supra.Google Scholar