Work and Honor in the Law: Prestige and the Division of Lawyers' Labor

by Rebecca L. Sandefur
Work and Honor in the Law: Prestige and the Division of Lawyers' Labor
Rebecca L. Sandefur
American Sociological Review
Start Page: 
End Page: 
Select license: 
Select License


University of Chicago

Two theories compete to explain the distribution of professional prestige among lawyers. Each theory arises out of a general theory of the social organization of the legal profession. Their disagreement centers on the source of the values that under- lie lawyers'prestige order and on the aspects of the division of labor that are conse- quently salient in determining prestige. The client-type thesis holds that the profes- sion is organized largely by its relationships with its clients, and so lawyers esteem service to socially poweiful clients. The professional purity thesis holds that the profession is organized around a core of abstract knowledge, and so lawyers esteem work that is "professionally pure," in the sense of being free of non-legal consider- ations. Analyses of prestige judgments from a contemporary sample of lawyers in the city of Chicago reveal that some of lawyers'professional values are intrinsic to the profession, as predicted by the professional purity thesis. Lawyers' esteem for pro- fessionally pure work complements their tendency to derogate service to people-

especially those with little wealth or power-and ful, and wealthy organizations.

DISCUSSIONS of the American legal profession return perennially to the is- sue of lawyers' independence of legal judg- ment and professional action-independence from the demands of their clients, from mar- ket pressures, and from the influence of other lawyers or groups of lawyers (Auer- bach 1976; Carlin 1962; Derber 1982; Der- ber and Schwartz 1991; Freidson 1994; Gor- don 1988; Heinz and Laumann 1982; Nelson and Nielson 2000; Rosen 1989; Smigel

Direct all correspondence to Department of Sociology, University of Chicago, 1126 E. 59th Street, Chicago, IL 60637, (r-sandefur@ A version of this paper was pre- sented at the annual meetings of the American Sociological Association, August 21, 1998, San Francisco, CA. Research was supported in part by grants from the American Bar Foundation and the National Science Foundation (grant no. SBR- 941 1515). I thank Andrew Abbott, John P. Heinz, Edward 0.Laumann, Heather MacIndoe, Jenni- fer Pashup, Christopher Swanson, and the ASR Editors and reviewers for helpful comments. Any errors remain my responsibility.

to esteem service to large, power-

1964). Interest in independence persists, in

part, because lawyers, like doctors, are ex-

emplars of a classic "free" profession. Fur-

ther, recent changes in the organization of

legal work and in the markets for certain

kinds of legal services may have curtailed

historic freedoms by placing some lawyers

under bureaucratic control or increasing

their economic dependence on particular cli-

ents. Lawyers' independence is also central

to their role as "officers of the court" (Gaet-

ke 1989). Lawyers are supposed not only to

advocate zealously for their clients' inter-

ests, but also to work in the broader inter-

ests of justice: "Where . . . [interests] con-

flict, the [lawyer's] duty to the system of jus-

tice must transcend the [lawyer's] duty to the

client," as well as pressure from colleagues

and hopes for personal gain (American Bar

Association Commission on Professionalism


The integrity and commitment of indi-

vidual lawyers certainly plays a role in en-

suring their independence, but patterns of

social organization within the profession

REVIEW,2001, VOL.66W~~~:382-403)

also play a role by making some courses of action and kinds of work easier to pursue or more rewarding than others. One powerful force of social control within any group is the esteem granted to exemplary individuals and the prestige that attends certain social roles (Goode 1978). In any given group, who is esteemed and for what reveals something about that group's values and indicates spe- cific acts and social roles its members wish to reinforce. The characteristics of presti- gious roles-their authority to direct others, their power to affect others, their autonomy from others' control, their cognitive sophis- tication, the reserves of compassion and em- pathy required to perform them well-are empirical traces of what work is considered important and unimportant and of what are considered exemplary and inadequate per- formances.

Two theories compete to explain the dis- tribution of esteem among lawyers. Each theory arises out of a general theory of the social organization of the legal profession. Their disagreement centers on the source of the values that underlie lawyers' prestige or- der and on the aspects of the division of la- bor among lawyers that are consequently sa- lient in determining prestige. The client-type thesis holds that external social hierarchies and values penetrate the bar in such a way that lawyers most esteem work for clients who enjoy the highest regard in society (Heinz and Laumann 1982; Laumann and Heinz 1977). The professional purity thesis holds that lawyers' value system is intrinsic to the profession: Lawyers' regard for an area of legal practice reflects the degree to which its practitioners are able to exclude nonprofessional tasks from their work (Ab- bott 198 1, 1988). It is only because different kinds of clients require tasks that are more or less professional that clients' characteris- tics have any relationship to prestige.

The relative honor accorded to someone by virtue of a position in a certain role or as the practitioner of particular tasks is commonly called prestige. Prestige is an entitlement to

' In a given situation, additional factors con- tribute to the esteem enjoyed by an individual.


deference (Shils [I9681 1994), and as such it has an inherently hierarchical quality. To have prestige is to receive a kind of respect and admiration from others that requires the making of "invidious comparisons" (Veblen [I8991 1994)-without the judgment of "greater" and "lesser" there can be no defer- ence. The entitlement to deference arises out of respect and admiration that are deserved in the sense that they are grounded in the values of the group in which the person is esteemed or the role is considered presti- gious.* This sense of "value" is both a quali- tative statement about what is valued and a quantitative or ordinal statement about what things are valued above other things.

Prestige is "granted by an individual or a collectivity for performances they consider above the average" (Goode 1978:7, empha- sis omitted). For the granting of deference, "above average" may mean better than most others at a given role or task, or it may mean that the role or task itself most nearly ap- proaches the fulfillment of commonly held ideals. Thus, for example, in groups that value scholarship, the learned will be presti- gious and the erudite will be among the most prestigious. In this understanding, those tasks are "most esteemed . . . which are in their internal structure and their function closest to the centres" of the group that is the admir- ing or disapproving audience of their perfor- mance (Shils [I9681 1994: 195, emphasis in original). Put in the terms of personal honor, "the qualities enjoined by honor provide the link. . . between self and the idealized norms of the community" (Berger, Berger, and

Competence at accustomed tasks, charm, intelli- gence, beauty, or noble birth are among the many qualities of a person that might confer esteem. Here, I am concerned with the prestige that ac- crues to roles, positions, and tasks, not to indi- vidual people per se.

Different authors give different emphasis to the degree to which these values tend to be the values of some dominant group, or to reflect a widely recognized, supposedly objective func- tional import of given tasks or types of work (Bourdieu 1984; Davis and Moore 1945; Tumin 1953). In any specific case, the degree of con- sensus is an empirical question, analytically dis- tinct from questions of the predictors of prestige (Wegener 1992). I address the question of con- sensus below.

Kellner 1974:86). In this scheme, the objec- tive functional importance of a role or type of work in a division of labor may or may not be relevant to the esteem it enjoys; the point is that the role or type of work corresponds to legitimating values.

Prestige is not only an effect, as it were, of the values of some collectivity; it is also a basis for power on the part of the es- teemed. Because of its genesis in the under- lying values of a shared system of meaning, the granting of deference is in some sense involuntary. "[Wle cannot truly promise a certain amount of respect or esteem to some- one because we cannot force ourselves to feel it or to avoid feeling it" (Goode 1978: xi). Prestige is a particularly potent basis for the form of power that resides in influence, or the ability to get others to do what one wants because others already want to com- ply with one's wishes, even before those wishes are expressed (Parsons 1963). Pres- tige legitimates authority in situations in which the role or activity that accords pres- tige is deemed relevant by the interactants.


The division of labor in the bar is structured along multiple lines. The task for an analy- sis of prestige is to discover which principles that divide work also serve as grounds for invidious comparisons. Fields of legal prac- tice are differentiated in a variety of ways: by the legal doctrine of the matters, by the kinds of clients served and the social dis- tance between clients and their lawyers, and by the specific activities and tasks that the fields of practice require of the lawyers who work in them. Lawyers perform many dif- ferent kinds of work, ranging from the simple and accessible to the highly special- ized, intellectually challenging, and posi- tively arcane. Legal work includes routine tasks that can be done by relatively unso- phisticated computer programs, such as the preparation of simple tax returns and the drafting of standard-form wills; yet, the same areas of law can present legal problems of intriguing complexity.

Within a given doctrinal area of law, law- yers may serve different types of clients. The most crucial distinction among types of cli- ents is probably whether they are large orga- nizations or individual people and the small businesses they own. For example, a tax lawyer may advise on the income tax returns of Philip Moms, Oscar Mayer, or Beatrice, or on the income tax returns of Bob Smith, Julia Jones, or The Village Bar. The two classes of tax work involve not only substan- tively different bodies of law, but tend to in- volve different kinds of problems that vary substantially in their legal complexity. A cor- porate tax lawyer will advise Philip Morris about how it may appropriately take advan- tage of expenses incurred in purchasing Os- car Mayer, while a personal tax lawyer will consider such questions as how much of the cost of someone's blood pressure medication is tax-deductible. The different kinds of problems likewise involve a lawyer's contact with different kinds of people. In Philip Moms's case, the lawyer will encounter the in-house legal staff and high-level firm man- agers and executives of a powerful multina- tional corporation. A small-time personal tax lawyer will work with such clients as bar- tenders, teachers, and the proprietors of businesses like taverns and independent sec- ondhand bookshops.

One theory of the social organization of the legal profession holds that the division of labor among lawyers and the allocation of lawyers of different social origins across that division are driven largely by differentiation in the types of clients that lawyers serve. In a study of the legal profession in Chicago in the mid-1970s, Heinz and Laumann (1982) present a greatly overdetermined social sys- tem in which external social hierarchies, such as those of ethnoreligious heritage and class, overlaid and corresponded to the divi- sion of legal work into two "hemispheres," one serving largely corporate clients and one serving natural persons and lesser economic players such as small businesses. Although most lawyers practiced in a variety of areas of law, few lawyers were found working in both "hemispheres." Lawyers who had higher social standing outside the bar on ac- count of their ethnoreligious heritage and class background-that is, white Northern and Western European men3 from profes-

Women composed just 3 percent of practicing lawyers in the sample (Heinz and Laumann 1982).

sional families, whose religious affiliation was mainline Protestant-tended to serve corporate clients. Lawyers who were mem- bers of socially less-esteemed groups, such as Jews and Southern and Eastern Europe- ans and those from families of lower socio- economic status, tended to provide legal ser- vices to individuals and to small businesses.

External social hierarchies and values pen- etrated the bar as well as overlaid it. As part of the general theory of two hemispheres, Heinz and Laumann (1982) argued that the esteem that lawyers gave to different fields of legal work was determined largely by the kinds of clients served in that work. The last column of Table 1 reports the rank ordering by prestige of the 30 fields of law Heinz and Laumann assessed in their study. As the table shows, work for corporations was the most estimable work, while work performed largely for personal clients had lower pres- tige in the bar. Furthermore, the more a field of law served persons of low social stand- ing, the lower still was its prestige among lawyers.

In Heinz and Laumann's explanation, a single principle underlies this distinction by client type: Values whose sources lay outside the legal profession in large part determined what lawyers esteemed. The lawyers who served corporations served "the core eco- nomic values of our society. . . . [Tlhe more a field of law serves these values, the higher its prestige will be within the profession" (Heinz and Laumann 1982: 130). In general, lawyers more highly esteemed work for "es- tablishment" clients (Heinz and Laumann 1982:128). As Table 1 demonstrates, Chi- cago lawyers esteemed working for the state against criminal deviance above working to defend those accused of crime, and defend- ing corporations was more highly regarded than was suing or prosecuting them. Other characteristics of a field of law-the intellectual challenge of its work, in particular- did affect the relative prestige of fields within their "hemisphere." Thus, the high value lawyers placed on intellectual chal- lenge explains why personal client fields that served larger proportions of clients of high social standing had higher prestige: Rich clients were argued to have more inter- esting and challenging legal problems (Heinz and Laumann 1982: 129). Overall,

however, core values of U.S. society under- girded the prestige order of the bar.

Heinz and Laumann's (1982:385) theory of intralegal prestige derives from their gen- eral observation that the Chicago bar of the mid-1970s was riven by a divide between corporate and personal clients into "two le- gal professions . . . , suggest[ing] a corre- sponding stratification of law into two sys- tems of justice, separate and unequal." Later work observes that the finding of high pres- tige for corporate work is "ironic" (Nelson and Trubeck 1992: 182). It contradicts the re- ceived wisdom that professionals value au- tonomy from powerful clients. The corre- sponding finding that work for individual people, particularly those of low socioeco- nomic status, is especially derogated is a sad comment on a profession that is supposed to value public service. It appears that classic values of professionalism such as service, expert autonomy, and intellectual challenge are of secondary importance. Client type is the salient divisive principle both of prestige and of work because lawyers value service to wealth and power.

Abbott (1981, 1988) suggests an alterna- tive explanation for the apparent structuring of prestige along the lines of client type-the professional purity thesis. Abbott (1988: 52-58) holds that professions are character- ized by their organization around cores, or "centers," of abstract knowledge. Deference is due to that work which most closely ap- proaches this core, for that work is the most professionally pure. Its purity obtains be- cause nonprofessional considerations have been removed. Abbott identifies a tricho- tomy of professional work that describes the division of legal labor into three bundles of tasks varying in professional purity: diagno- sis, inference, and treatment. Diagnosis de- notes the translation of the client's problem into professional terms, which permits infer- ence. In inference, the nonprofessional as- pects of the client's problem are least present: The problem has been reformulated in terms of the abstract knowledge and con- ceptual apparatus of the profession. Profes- sional knowledge is then applied or manipu- lated in order to solve the problem now de- fined in professional terms; thus, the work of inference is the most professionally pure (Abbott 1988:40, 47, 48). Treatment is the

Table 1. Prestige and Size of 42 Fields of Law: Chicago Lawyers Survey, 1995
Field of Law     Score in 1995 a
Securities     84
Trademark and Copyright     75
International Law, Private     74
International Law, Public     7 1
Patents     70
Civil Litigation, Business     66
General Corporate     66
Tax, Corporate Federal Income     65
Anti-Trust, Defense     65

Real Estate, Finance and Development 62 Banking 58 Anti-Trust, Plaintiffs 58 Tax, Estate and Gift 50 Environmental Law, Plaintiffs 48 Environmental Law, Defense 47 Civil Rights and Civil Liberties 44 Employment, Management 38 Municipal Law 36 Criminal Prosecution 33 Bankruptcy, Business 3 3 Public Utilities and Administrative Law 33 Tax, Personal Federal Income 32 Civil Litigation, Personal 28 Real Estate, Zoning and Eminent Domain 27 Uniform Commercial Code 26 Admiralty 26

provision to the client of the solution, in- volving a translation out of professional terms.

Rather than focusing on whom a profes- sion serves, the professional purity thesis fo- cuses on the concrete tasks professionals perform. As an illustration, Abbott (1981) presents the example of the courtroom, in which judges are more professionally pure than the lawyers who argue before them. The work of the judge consists in dealing with issues that the lawyers have purified by de- fining them in specifically legal terms-by constructing arguments that apply the rel- evant statutes, precedents, and principles to
Percentage     Prestige
Number of     of Total     Rank in
Practitioners     Hours Worked     197Sc
(Table I continued on next page)

the particular client's particular problems. The judge, presented with a dispute defined wholly in the conceptual categories of the profession, infers the answer to what has been construed as a truly legal question.

Lawyers' work for different types of cli- ents, according to Abbott, exhibits similar variation in professional purity. When work- ing for people, a lawyer's client may be an angry ex-spouse, or a dispossessed child, or a distressed victim of malpractice or shoddy manufacture whom the lawyer must comfort or console, whom the lawyer must direct away from personal experience back to the legal matter at hand, or whose grief and pain
(Table I continued)

Field of Law

Probate Tax, State and Local Employment, Labor Personal Injury, Defense Criminal Defense Personal Injury, Plaintiffs Bankruptcy, Personal Real Estate, Residential Transfers Consumer Law, Seller/Creditor

Prestige         Percentage     Prestige
Increment     Number of     of Total     Rank in
Score in 1995 a     Practitioners     Hours Worked     1975

General Family Law, Paying Clients General Family Law, Poverty Clients Real Estate, LandlordJTenant

Consumer Law, Consumer/Debtor Juvenile Law Immigration Divorce

Notes: Fields are listed in rank-order of prestige in 1995; total number of practitioners is 674. a This score measures the percentage of respondents who think the specialty enjoys better-than-average

prestige (rated as "outstanding" or "above average"). The number of lawyers working at least 5 to 24 percent of their time in each listed field. These data come from Heinz and Laumann (1982, table 4.1). In the 1975 Chicago Lawyers Survey, the following fields were undifferentiated: Real Estate, Civil Liti-

gation, and Tax. In the 1975 Chicago Lawyers Survey, "Real Estate, Zoning and Eminent Domain" was listed as "Con- demnations."'In the 1975 Chicago Lawyers Survey, "Real Estate, Landlord/TenantW was listed as "Landlord/Tenant."

must be exploited to sway a jury. Work for corporations is more prestigious than work for people because the corporation itself is "the lawyers' creation. The muck of feelings and will is omitted from it ab initio" (Abbott 1981: 824). Professional purity also explains why legal work performed for personal cli- ents of high social status is more prestigious. Highly educated clients are often knowl- edgeable about the law, and they often re- move professionally irrelevant aspects from matters before presenting them to their law- yers, saving their lawyers the work of diag- nosing their problems and translating them into legal terms (Abbott 1981). In other words, work for high status clients is more professionally pure because the challenge it presents is generally a specifically legal challenge, demanding peculiarly legal skills and knowledge (Abbott 1981925, note 6). The professional purity thesis thus resolves the apparent irony in the finding that work for more powerful clients is more prestigious work: The crucial aspect of autonomy is not autonomy from client control of lawyers' ac- tions; it is freedom from client "dirtying" of the legal matter with nonlegal consider- ations. It is quintessentially legal autonomy.

Three implications of Abbott's theory are important for present purposes. First, the principles that divide labor and prestige are intrinsic to the profession: It is the degree to which tasks are defined and comprehensible within the knowledge of a profession that determines their prestige within that profes- sion. Second, and consequent to the first, the apparent relationship between client type and prestige is spurious. The fact that clients of different types present problems of vary- ing degrees of professional purity accounts for the finding that work for different kinds of clients differs in prestige. Third and cor- ollary to the first, external social hierarchies and values may overlay the bar in the ways described above, for instance by allocating persons of high social standing to positions of high prestige within the bar, but they do not penetrate the bar in the sense of affect- ing professional values about work.

The difference between these two theories of intraprofessional prestige thus resolves it- self into a simple empirical question, though not necessarily one that is easily answered in practice. If the relationship between cli- ent type and prestige can be substantially explained away by considering characteris- tics of work that adequately index profes- sional purity, then Abbott is correct to ob- serve that the relationship between the sta- tus of the client and the status of the work is spurious. However, if much of the relation- ship between client type and prestige re- mains once theoretically salient work char- acteristics have been taken into account, one of two interpretations is suggested. One in- terpretation would be that purity is not re- lated to the organization of work into fields of practice, but rather to the organization of work by some other principle, such as orga- nizational rank or seniority. The other inter- pretation would be simply that the profes- sional purity thesis is false.


The data I use were collected in a 1995 survey of the Chicago bar. These data provide a unique, rich body of information about the experiences and beliefs of contemporary ur- ban American lawyers. Respondents were selected in a random sampling of lawyers in good standing with the Illinois Attorney Registration and Disciplinary Commission who had office addresses in the City of Chi- ~ago.~

The response rate among contacted

To practice law in the State of Illinois, a law- yer must be in good standing with the Illinois

lawyers was 82 percent, resulting in a com- pleted sample size of 788.5 In face-to-face interviews averaging a bit over an hour in length, the lawyers were asked about their employment and educational histories, po- litical affiliation and social participation, family background, and aspects of their cur- rent work and current job, as well as about their social, political, and professional val- ues. They were also asked to report on the prestige among lawyers of 42 areas of legal practice. Analyses here employ data col- lected from lawyers actively engaged in the practice of law as private practitioners work- ing alone or in law firms, or as lawyer-em- ployees of government or of private for- profit or not-for-profit corporations; 674 such lawyers constitute the analysis sample. I limit these analyses to practicing lawyers in order to use information provided by law- yers closest to and, hopefully, most knowl- edgeable about the actual work of their prac- ticing colleagues and the esteem in which those colleagues hold that work.

Each respondent was asked how much time was spent working in each of 42 areas of practice. The lawyer was handed a booklet with the fields of law listed in alphabetical order and asked to choose from five prede- termined categories of response for each area, indicating the percentage of working time devoted to each field of law. The re- sponse categories consisted of intervals: 1 to 4 percent, 5 to 24 percent, 25 to 49 percent, 50 to 99 percent, and 100 percent. Most law- yers practice in more than one area of law.6

Attorney Registration and Disciplinary Commis- sion, an agency under the supervision of the state's Supreme Court. Good standing requires that the lawyer has not been disbarred or sus- pended, and has registered with the Commission and paid a registration fee.

Comparison of the collected sample with other sources of data on the Chicago bar suggests that the sample represents well the population from which it was drawn. Solo practitioners and lawyers from local law schools appear to be mildly underrepresented in the sample (Heinz et al. 1998).

The survey collected time-allocation informa- tion for a total of 53 areas of legal practice. Re- spondents reported 11 additional areas of prac- tice in the course of the survey. Twenty-three percent of lawyers report practicing in only 1 of the 53 areas.

Respondents were instructed not to concern themselves with whether their responses added to 100 percent.

Reports of the distribution of lawyers' time across the areas of practice become the basis for the analyses of the relationship between the prestige of a field of law and its charac- teristics. I construct measures of the charac- teristics of each field by aggregating across lawyers who reported working at least 5 to 24 percent of their time in that field. Each lawyer's contribution to the field is estimated based on his or her reports about the distri- bution of work time and the number of hours he or she reports working in a typical week.

To construct a measure of each lawyer's contribution to each field, I assigned each lawyer's proportionate time allocations a nu- merical value equal to the midpoint of the interval indicated. Each lawyer's time allo- cations were then normed to sum to 100 per- cent. Each lawyer's hours were then allo- cated across the reported fields in accordance with his or her time allocations. For example, a lawyer might report that he or she typically worked 48 hours each week, dividing the time worked across three fields of law: "Pat- ents" and "Trademark and Copyright" law, to each of which he or she devoted 5 to 24 percent of work time, and "General Corpo- rate" law, to which he or she devoted 55 to 99 percent of work time. For this lawyer, as- signing the percentage intervals their mid- points results in a total time allocation of

108.5 percent (14.5% + 14.5% + 79.5%). The inverse of the summed time allocations was used as a scaling factor. The sum of the prod- ucts of the scaling factor and each time allo- cation equals 100 percent by construction (e.g., (14.5)(.9217) + (14.5)(.9217) + (79.5)(.9217) = 100). The product of the normed time allocation for a given field and the lawyer's reported hours then provides an estimate of the number of hours per week the lawyer works in the given field of law. In the present example, the lawyer devotes an esti- mated 6.4 hours per week to each of Patents and TrademarkICopyright law, and 35.2 hours per week to practice in General Corpo- rate law. Table 1 lists the number of respon- dents among the 674 practicing lawyers who report investing at least 5 to 24 percent of their time in each field of law and the per- centage of total reported hours invested in each field of law.

Because the fields of practice share per- sonnel and it is the reports of personnel that I aggregate to characterize the fields, the characteristics of the different fields-type of client, frequency of specific work tasks, and so forth-appear less differentiated than they actually are. The measures used here thus present a conservative picture of differ- entiation among the fields of law.7 These measures also contain a certain amount of error that arises from their construction. For example, I do not have measures of the types of clients who go to lawyers for securities work; I instead have measures of the types of clients served by lawyers who do securi- ties work. I do not have measures of the fre- quency with which lawyers go to appellate courts for securities cases; I instead have measures of the frequency with which law- yers who do securities work appear in appel- late courts. All fields are subject to this kind of measurement error to some degree, al- though fields in which more practitioners are dilettantes are probably more strongly con- taminated than fields that require a greater share of a lawyer's time. Exploratory analy- ses found no relationship between prestige and a measure of dilettantism. To the extent that measurement error is random with re- spect to the variables of theoretical interest, it will tend to weaken equally the effects of all variables on prestige.

As part of the survey, respondents were asked to rate each field of law in an alpha-

Any method of estimating lawyers' time con- tributions to each field using these data will re- sult in some imprecision. Although the method I have chosen cannot provide precise estimates of the absolute amount of time invested in each field by each lawyer, the estimates probably reflect well lawyers' relative contributions to each field. An ASR reviewer noted that some fields on the list are not mutually exclusive (e.g., General Cor- porate and Anti-Trust Defense), suggesting that my method for estimating time allocations may result in overestimation of the amount of law- yers' time devoted to the more general fields and underestimation of the amount of time devoted to more specific fields. Readers should be mind- ful of this when inspecting the distribution of to- tal legal effort presented in Table 1.

betically ordered listing of fields in terms of its "general prestige . . . within the legal pro- fession at large." The rating scale ranged from 1 to 5, where 1 indicated that the re- spondent believed the field was "poor" in prestige and 5 indicated that the respondent believed the field was "outstanding." For each respondent who took the question seri- ously, I then have a report on the profession as a whole by an active member of the pro- fession. Forty-seven of the 674 practicing lawyers did not respond to the question; it was near the end of the interview. Overall, each of the 42 specialties was rated by be- tween 85 percent and 92 percent of the prac- ticing lawyers who responded to the survey; some respondents did not rate all of the spe- cialties presented to them. Sixty-eight per- cent of the practicing lawyers who answered the question rated all the fields, 95 percent rated 35 or more fields, and 97 percent rated 25 or more fields. No characteristics of law- yers strongly pattern their propensity to rate more or fewer specialties. (Details of these analyses are available from the author on re- quest.)

The prestige of a field of law is measured by a "prestige increment" score: the percent- age of ratings that indicate the field is "out- standing" or "above average" in prestige. This score has certain benefits over the more common procedure of simply averaging re- spondents' ratings. First, the interpretation of the prestige increment score is straight- forward: It is the percentage of respondents who think the specialty enjoys better-than- average esteem in the bar. Second, such a score avoids potentially misleading aspects of average rating scores that arise whenever one tries to characterize a nonnormal distri- bution by its central tendency.

As a tool for measuring prestige, the field rating item has clear face validity: It asks di- rectly about the prestige of the specialty among lawyers. Goldthorpe and Hope (1974), writing of occupational prestige, as- sert that

. . . a measure of occupational prestige is to
be regarded as valid to the extent that one
may correctly infer from it the relative
chance that a member of an occupational
category has of experiencing deference, ac-
ceptance, or derogation in his relations with
members of other categories. (P. 5)

A measure of prestige is valid, then, if the respondent behaves as though it means what the sociologist thinks it means, viz. the defi- nition of prestige as deference entitlement outlined above. The Chicago Lawyers data provide no direct indications of how the re- spondents interpreted the question about prestige. Research on populations much more heterogeneous who were asked to rate the "social standing" or "prestige" of occu- pations much further removed from their own experience suggests that different ques- tions result in similar occupational rankings (Treiman 1977).

Investigations of the effects of occupa- tional standing on behavior have focused more on preferences for higher- or equal-sta- tus contact than on deference. In studies of occupational prestige and patterns of differ- ential association such as assortative mating, friendship choices, and residential settle- ment, scholars have found a notable ten- dency for people to prefer association with others of superior occupational standing and to wish to avoid association with others of inferior occupational standing (Laumann 1965, 1966, 1973; also see the research cited in Treiman [1977:28-291). Treiman (1977) argues that findings of preferences for equal- or higher-status contact indicate the validity of standard prestige questions because be- haviors argued to be associated with relative prestige differences are strongly correlated with responses to such questions.

A small amount of investigation into def- erence-inducing properties of occupational prestige has been pursued in work on status generalization (Berger, Rosenholtz, and Zel- ditch 1980; Webster and Driskell 1978). For example, studies of mock jury deliberations have found that jurors' occupational stand- ing affected their likelihood of being chosen as foreman and the influence they had in de- liberations. Jurors from more prestigious oc- cupations were more likely to be chosen as foreman and were more influential than were jurors with humbler jobs (Strodtbeck, James, and Hawkins 1957). Considered together, social distance and status generalization studies of prestige suggest that a broad range of respondents respond similarly to prestige items and that the items behave as predicted by theories of prestige. For the present case, face validity is the strongest indicator of the validity of the item used to measure the pres- tige of the fields of law, independent of the measure's tendency to behave as predicted by theories of prestige.

The prestige item asked lawyers about their perception of the prevailing values of the profession, rather than about their own values (cf. Bourdieu 1984). That is, lawyers are reporting on their colleagues, and not necessarily on themselves. In such a case, systematic disagreement about prestige among respondents could arise when lawyers report on distinct reference groups whose members make substantially different pres- tige judgments. In the terms of the theories to be tested, this would indicate the presence of multiple cores or centers within the pro- fession. To investigate this question, I ana- lyzed the relationship between prestige rat- ings and lawyers' race, gender, age, political affiliation, the location of the respondent's own fieldls of practice within the prestige order, the types of client the lawyer served, and the lawyer's score on a measure of eco- nomic liberalism (Payne and Nelson 2000). I found no systematic differences among the prestige orders reported by lawyers differen- tiated by these factors. Correlations of pres- tige ratings across the groups of lawyers ex- amined ranged from .96 to .99. There were no instances of inversions or other signifi- cant reordering of the fields. Examination of differences in the ratings of specific fields revealed that such differences generally ex- hibited a kind of "occupational egoism" (Kahl 1957), the tendency to think one's own work is more highly regarded at large than others' ratings reveal it to be. (Details of these analyses are available from the author on request.) Thus, while some Chicago law- yers may disagree with the values of the pro- fession "at large," they appear to be in rela- tive agreement about what those values are.


THECLIENT-TYPE THESIS. The main con- tention of the client-type thesis is that fields of law that serve business organizations serve the "core economic values of society" and are prestigious for doing so. In their ex- ploratory analyses, Heinz and Laumann (1982) also found effects on prestige of the

social status of personal clients served in a field of law and of the intellectual challenge of work in the field. They argued that the lat- ter pair of findings was consistent with the client-type thesis, since high status personal clients presented lawyers with more interest- ing and challenging legal work. The client- type thesis thus makes the following three predictions about the effects of specific vari- ables on prestige:

Hypothesis 1:The greater the proportion of businesses among the clients of a field of law, the higher its prestige will be.

Hypothesis 2: The greater the proportion of persons of low social status among the personal clients of a field of law, the lower its prestige will be.

Hypothesis 3: The greater the intellectual challenge of a field of law, the higher its prestige will be. The inclusion of a mea- sure of this challenge should mediate a substantial portion of the effect of per- sonal-client type on prestige.

Following Heinz and Laumann (1982, chap. 4), measures of client type are con- structed from respondents' reports of the percentage of their total clients that were businesses and the percentage of their per- sonal clients who were of low social status. In the analyses that follow, the percentage of low-status clients is measured as the per- centage of all personal clients who are sales and clerical workers, blue-collar workers, or who are unemployed. The measures of cli- ent type are the weighted average of the re- ports of the lawyers who practice in each field, where each lawyer's weight is the pro- portion of total hours in the field contributed by that lawyer.

The intellectual challenge of the work, as indicated by the level of specifically legal skill it requires, is measured by the percent- age of hours worked in the field by respon- dents who report that their work requires so much specialized skill and knowledge that it could not be done by an educated layman. That is, it is the percentage of hours in each field reported by respondents who circle "1" or "2" in response to the question shown in Figure 1.

THEPROFESSIONAL PURITY THESIS. The thesis of professional purity suggests that the

Different kinds of law require different kinds of professional activities. [Below are] a

series of paired statements that describe different demands made on the lawyer. These

are presented as polar opposites. Please circle the number that best represents your

position in relation to the two opposites. If the situation in your practice is midway

between poles, circle code 3; if your situation is at one or the other extreme, circle 1 or

5; if your position leans somewhat to either pole, circle 2 or 4.

A The type and content of my practice is such that even an educated layman couldn't stand really understand or prepare the documents.

1 2 3

Figure 1. Item Measuring Intellectual Challenge

more time lawyers in a given field devote to the work of inference, in favor of treatment or diagnosis, the greater will be the field's prestige among lawyers. The thesis posits that purity varies across fields of law, but professional purity will also vary within fields of law, along the lines of lawyers' or- ganizational rank, and along other aspects of organizations' internal divisions of labor. For example, senior lawyers who work in large law firms are likely spared the tedium of sifting through documents in discovery because young associate lawyers do such work for them. For variation in professional purity to affect the prestige of the fields of law, such variation must arise from the fact that different areas of law require lawyers to perform bundles of tasks that vary substan- tially in their professional purity.

The specific tasks performed in different fields of law and the way those tasks are or- ganized are conditioned by aspects of the le- gal matters themselves (Kordana 1995; Nelson 1988:177). Describing the legal work performed for businesses, Kordana (1995) notes:

A large corporate matter (e.g., a securities
offering or a merger) or a large litigation
matter requires a great deal of what can be
termed "paperwork." For a corporate matter
these tasks include "due diligence" . . . and
drafting and proofreading a variety of agree-
ments. For a litigation matter the paperwork
includes document discovery . . . and ex-
haustive legal research on the myriad issues
involved in or potentially relevant to the dis-

A relatively smaller proportion of the
work to be done by a large firm in a corpo-

A para-professional could be trained to handle many of the procedures and under- documents in my area of law.

4 5

rate transaction or litigation matter consists of client interaction and strategic, complex legal work. This includes synthesizing and applying legal research results to the facts at hand, using discovery for strategic advan- tage over the opponents in a litigation mat- ter, ensuring that agreements or settlements successfully meet client objectives, and keeping clients apprised of developments. (Pp. 1924-25, notes suppressed)

To the extent that "paperwork" is profes- sional at all, it falls in the realm of diagno- sis-the work of formulating the client's problem as a legal matter and casting that problem in terms of other matters that have set legal precedents. The work of managing relationships with clients consists, in part, of treatment-of translating the legal under- standing of the client's problem back into lay terms so that the client may comprehend and evaluate the progress of the case-and, in part, of work that is entirely nonprofes- sional in Abbott's terms, the "business" work of a salesperson dealing with a cus- tomer. "Strategic, complex legal work" is in- ference, the work that is most professionally pure.

This description of how the tasks involved in large corporate and litigation matters are organized reveals that professionally pure work often confers professional authority. Professional authority involves power, as well as purity, but it differs from managerial authority in important ways. Managerial au- thority in law firms-the power to hire and fire and responsibility for other administra- tive work of the firm-is formally shared by the managing partner and other equity part-

ners or subcommittees of those partners (Nelson 1988; Smigel 1964). While such au- thority may or may not confer professional pre~tige,~

the distribution of managerial au- thority bears no necessary relationship to lawyers' substantive areas of practice. Pecu- liarly professional authority, on the other hand, arises because some people "in the normal course of work involving coopera- tion with others in a division of labor, . .. because of experience or specialized knowl- edge, direct those others" (Freidson 1986: 143). Fields of law in which work is orga- nized in the pyramidal division of labor that Kordana (1995) describes provide an oppor- tunity for some lawyers to specialize in "strategic, complex legal work" by delegat- ing simpler and relatively impure work to other lawyers.

As a measure of the degree to which work in a field of law requires professionally pure "strategic, complex legal work," I calculate the percentage of hours in each field of law contributed by lawyers who report that they "frequently" or "very frequently" supervise and review the work of other lawyers. Not all of the work of supervising and reviewing involves legal inference. Some of it consists of the mentoring of young lawyers by older lawyers and the administrative work of di- viding up tasks within a group of lawyers working together. However, despite its con- tamination by other factors, this measure is the best available of the degree to which the division of labor within a field of law per- mits some lawyers to pursue professionally pure work by allowing them to delegate much of the work of diagnosis to other law- yers.

To specify as precisely as possible a rela- tionship between purity and prestige, I esti- mate the effect of supervising and reviewing other lawyers' work net of both lawyers' se- niority and their managerial authority. As a measure of managerial authority, I calculate the percentage of hours in each field of law contributed by lawyers who report that they "frequently" or "very frequently" make "de- cisions regarding the management of [their] organization (such as recruitment, assign- ments, etc.)." To control for purity that may

Smigel (1964:238) and Abbott (1981) argue that it does not.

arise from a preponderance of senior lawyers within a given field rather than from the work demands of the field of law itself, I control for the age distribution of the fields of law with two terms-the share of lawyers who are younger than 35, and the share who are older than 55.9 As before, each lawyer's contribution to a field's characteristics is weighted by the share of total hours worked in the field contributed by that lawyer.

All fields of law require some amount of office practice, but many fields also require lawyers to appear in court. All courtroom work is not the same, however. In trial courts, lawyers must establish facts. Wit- nesses must be called and examined and cross-examined; evidence must be presented and supported and contested. Lawyers diag- nose facts by presenting arguments to the judge or jury that delineate how those facts are comprehended by the law. The judge then performs legal inference based on the lawyers' arguments, either making judg- ments of both fact and law him- or herself, or instructing the jury in how law applies to the case at hand and leaving them to be the judges of fact. In appellate courts, matters of fact are generally accepted as settled-ap- peals center on the question of whether the proper legal inference was made. In appel- late courts, lawyers' arguments center on matters of law and are therefore relatively professionally pure.

As a measure of the degree to which a field of law demands that lawyers engage in professionally impure trial court work, I calculate the average number of days per month that lawyers in a field report appear- ing in state and federal trial courts. As a measure of the degree to which a field of law demands that lawyers engage in the relatively professionally pure work of ap- peal, I calculate the average number of days per month that lawyers in a field report ap- pearing in state and federal appellate courts. As with other measures of fields' character- istics, each lawyer's contribution is

These ages represent the 25th and 90th per- centiles, respectively, of the age distribution of practicing lawyers in the sample. This relatively young age distribution is characteristic of the contemporary legal profession nationally (Galanter 1999).

weighted based upon his or her time invest- ment in the field.

Abbott (198 1 :825) has argued that lawyers esteem legally challenging work not so much for its intellectual challenge, but be- cause that challenge is specifically legal and, therefore, professionally pure. Understood in the conceptual framework of professional purity, the fact that work requires specialized skills and knowledge not accessible to an educated layman is indicative of its proxim- ity to the core of abstract knowledge at the center of the profession. Thus, the two theo- ries make the same prediction about one variable's effect, but differ in their under- standing of that variable's meaning.

The hypotheses specified in the profes- sional purity model of prestige then are:

Hypothesis 4:The degree to which lawyers' work requires directing and reviewing the work of other lawyers will vary positively with prestige. This effect will remain after controlling for lawyers' se- niority and for lawyers' involvement in the management of their organizations.

Hypothesis 5:The greater the demand for le- gal skill and knowledge in a field of law, the higher its prestige will be.

Hypothesis 6:The more frequently a field of law requires its practitioners to appear in trial courts, the lower its prestige will be.

Hypothesis 7:The more frequently a field of law requires its practitioners to appear in appellate courts, the higher its pres- tige will be.

Hypothesis 8: The inclusion of measures of professional purity will account for the effects of client type on prestige.


Table 1 (see pp. 386-87) presents a prestige hierarchy of the fields of law that is strongly patterned along the lines of client type. With the exception of Estate and Gift Tax and Civil Rightsfcivil Liberties, all of the fields in the top half of the status order are fields in which the work is done almost exclusively for business corporations or for government agencies. Estate and Gift Tax is a service provided primarily to well-to-do individuals who have considerable assets. Civil Rights1 Civil Liberties law is performed both for in- dividuals and for organizations-defense against discrimination claims is included in this field, as well as more activist legal work pursuing those claims. Most of the fields with low prestige serve primarily personal clients and much of this work is generally unsavory: clients include accused criminals, divorcing couples, juvenile delinquents, people facing financial troubles. Among the fields at the bottom of the status order, the dichotomy of client type seems to provide little differentiating power. Personal injury work for plaintiffs, mostly individuals, and for defendants, mostly businesses, are of es- sentially equivalent low prestige. Both sides of consumer law-work for sellers and creditors and for consumers and debtors- are of roughly equal low prestige.

Table 2 presents estimates from maximum likelihood (ML) multiple regressions of the logit of the prestige increment score on the client-type and professional purity vari- ables.1° Models 1 and 2 present the results from the client-type model. As predicted, the full client-type model (Model 2) reveals that the percentage of business clients is posi- tively associated with prestige (P = .33, p < .05), while the percentage of personal clients who are of lower social status is negatively associated with prestige (P = -.31, p c .01). The intellectual (or legal) challenge of a field of law is also positively associated with prestige, net of client-type (P = .28,p < .05). However, contrary to the predictions of both the client-type and professional purity the- ses, intellectual (or legal) challenge is asso- ciated with service to business clients rather

lo The logit of the prestige increment score is In [pl(l-p)], where p is the proportion of re- spondents indicating that the field enjoys "out- standing" or "above average" prestige in the bar. All analyses that follow use this measure as the dependent variable and are computed for 40 fields of law. Admiralty Law and Public Interna- tional Law, each with only two practitioners, are excluded because the number of practitioners is too small to provide reliable measures of field characteristics. Appendix Table A presents the correlations and covariances among all variables used in the analyses, as well as the mean for each.

Table 2. Maximum Likelihood Coefficients, Standard Errors, and Standardized Coefficients from Regressions of Prestige on Selected Characteristics of 40 Fields of Law: Chicago Lawyers Survey, 1995

Client-Type Thesis     Professional Purity Thesis
Independent Variable     Model 1     Model 2     Model 3     Model 4 a
Intercept     -1.65***     -2.28***     -2.91"     -3.10***
    (.48)    (.56)     (.98)     (.83)
Percentage of business clients     2.34***     1.51*     -    -
    (.57)    (.69)         
    ~521    L.331         
Percentage of personal clients of     -2.90**     -2.85**     -    -
low social status     (1.17)     (1.12)         
    [-.31]    [-.31]         
High intellectual (legal) challenge     -    1.93'     2.93***     2.59**
        (1.01)    (37)     (33)
        L.281    L.421     L.371
Days per month in trial courts     -    -    -.09*     -. 1 0*
            (.05) [-.29]    (.05) [-.33]
Days per month in appellate courts     -    -    2.63"     2.63"
            (1.02)    (1.04)
            ~291    ~291
Frequent supervision and review     -    -    2.90**     2.79**
of other lawyers' work             (1.00)     (1.00)
            L.361    L.341
Frequent decision-making in     -    -    .05     {o}
management of one's             (1.36)     
organization             [.ool     
Percentage of practitioners     -    -    -1.04     I01
younger than 35             (1.04)     
        [-. 111
Percentage of practitioners older than 55     -    -    -2.44 (2.14)     {O}
        [-. 141
R2     .49     .54     .61     .59

Notes: Numbers in parentheses are standard errors. Numbers in square brackets are standardized coeffi- cients (P).Public International Law and Admiralty Law are excluded. Total number of practitioners is 674; the number of cases is 40.

a Coefficients in curly brackets are constrained to equal zero in an overidentified model; x2 = 1.69,

d.f. = 3, p = .64. 'p < .05 **p< .O1 ***p< .001 (one-tailed tests)

than service to high-status personal clients: thority and terms indicating the age distri- Intellectual (or legal) challenge accounts for bution of the field bear no significant rela- about 37 percent of the effect of the percent- tionship to prestige, net of other measures, age business clients on prestige, but none of and their exclusion from the model does not the effect of personal client-type. The full result in notable changes in the other coeffi- client-type model explains 54 percent of the cients or in a significant decrease in fit. As variance in prestige. predicted, intellectual (legal) challenge, fre-

Models 3 and 4 of Table 2 present the re- quency of appearance in appellate courts, sults from ML multiple regressions for the and frequency of supervision and review of professional purity model. Managerial au-other lawyers' work are positively and sig-




Figure 2. Path Model of the Prestige of the Fields of Law

nificantly associated with prestige, while frequency of appearance in trial courts is significantly and negatively associated with prestige. The professional purity model ex- plains 59 percent of the variance in prestige.

The professional purity thesis holds that client type affects professional prestige only because work for different kinds of clients requires tasks that vary in professional pu- rity. In a path analytic or structural equation modeling framework, the professional purity thesis predicts that direct effects of the cli- ent-type variables on prestige will be negli- gible because effects of client type will be mediated though characteristics of work. Figure 2 presents this model graphically. The client-type thesis holds that the paths labeled A and B are the most important determinants of legal prestige. The professional purity thesis holds that the inclusion of information about the specific tasks lawyers do as part of their work for different kinds of clients- the paths labeled C, D, E, and F-should account for the effects hypothesized in the cli- ent-type thesis.

Table 3 presents an explicit test of the relative explanatory strength of the two theories. The results come from maximum likelihood (ML) estimation of two models, each of which includes five regression equa- tions. Prestige is modeled as an effect of work tasks and client type. As specified by the professional purity thesis, each work task is modeled as a function of client type. Be- cause theory does not give guidance on the causal relationships among the tasks, I have not specified causal relationships among them. However, I have modeled these tasks

Table 3. Maximum Likelihood Coefficients, Standard Errors, and Standardized Coefficients from Two Models of Prestige: Chicago Lawyers Survey, 1995
Dependent Variable
High     Days per     Days per     Supervision
Intellectual (Legal)     Month in Trial     Month in Appellate     and Review of Other     Prestige
Independent Variable     Challenge     Courts     Courts     Lawyers' Work     Model la     Model 2b

Percentage of business clients

Percentage of personal

clients of low social


High intellectual (legal) challenge

Days per month in trial courts

Days per month in appellate courts

Frequent supervision andreview of other lawyers' work

Notes: Numbers in parentheses are standard errors. Numbers in square brackets are standardized coeffi- cients (P)Public International Law and Admiralty Law are excluded. Number of respondents = 674; number of cases = 40.

Residual correlations for both models of prestige are r(C,D) = -.25, r(C,E) = .13, r(C,F) = -.18, r(D,E) = .24, r(D,F) = -.24, r(E,F) is constrained to equal zero.

a x2

= .40, d.f. = 1, p = .53. Coefficients in curly brackets are constrained to equal zero in Model 2; x2 = 7.41, d.f. = 3, p = .06. *p< .05 '*p< .O1 "'p < .001 (one-tailed tests)

as distinct outcomes of a common process: in Table 3. The estimated error structure is differentiation along the lines of client type presented in the notes to Table 3. and whatever omitted factors are included in The first four columns of Table 3 present the residual terms. It seems reasonable to the results of estimation of the four equa- suspect that some of the omitted factors rep- tions predicting the professional purity resented in the residual terms may be com- measures as effects of client-type. These es- mon across the work characteristics. Experi- timates are identical for the two models mentation with various specifications of the presented in the table; only the equation for error structure suggested that no correlation prestige differs between the two models. existed between the residuals for appellate The fifth column of Table 3 presents the re- court appearances and supervising and re- sults of estimation of the prestige equation viewing other lawyers. In the interest of par- for the first model, in which prestige is an simony, this residual correlation is con- effect of both client-type and professional strained to be zero in the models presented purity, and purity itself is caused by client type. In other words, these are the results from the estimation of all the paths in Fig- ure 2.

The first model in Table 3 fits the data well: The null hypothesis that the specified model generated the covariances observed among the variables is not rejected (x2 = .40,

d.f. = 1, p = .53). Three measures of profes- sional purity-high intellectual (legal) chal- lenge, days per month in trial court, and fre- quent supervision and review of other law- yers-are significantly related to client-type; measures of client type explain between 34 and 49 percent of the variance in these mea- sures. The frequency with which lawyers ap- pear in appellate courts, however, is not sig- nificantly affected by the types of clients served in a field of law. In this model, fre- quency of appearance in appellate courts and supervision and review of other lawyers' work are significantly and positively associ- ated with prestige. The effect of high legal or intellectual challenge on prestige is posi- tive but does not quite reach conventional significance levels (p = .06, one-tailed test). Net of other work characteristics, appellate court appearance is positively associated with prestige (P = .33, p c .01, one-tailed test)-fields that require lawyers to engage in more professionally pure appellate work, including personal client fields, have higher prestige, net of other characteristics.

Measures of professional purity account for 46 percent of the direct effect on pres- tige of the percentage of clients who are businesses, and 29 percent of the direct ef- fect on prestige of the percentage of personal clients who are of low social status." Fields that serve lower percentages of business cli- ents lose prestige, in part, because they are less intellectually and legally challenging and do not evidence a division of labor that permits some lawyers to specialize in the work of inference. As the share of personal clients who are of low social status in- creases, fields of law lose additional pres- tige, because work for such clients is less likely to exhibit a division of labor among lawyers that involves some lawyers special-

l1In both cases, the calculation of effect size reduction results from comparing the client-type coefficients in Model 1 of Table 2 to the coeffi- cients in the fifth column of Table 3.

izing in inference: Each 10 percentage-point increase in the share of personal clients who are of lower social status leads to a 3.9 per- centage-point reduction, on average, in the share of hours contributed by practitioners who supervise and review other lawyers' work.

In Model 2 of Table 3, the direct effects of client type on prestige are constrained to equal zero, as indicated by the curly-brack- eted zeros, while the effects of the other fac- tors are allowed to vary as in the previous model. The estimates for the coefficients and error structure of the equations for the field characteristics are identical to those for the first model. In other words, these are the re- sults of estimating the model in Figure 2 when the paths labeled A and B are con- strained to equal zero. The chi-square statis- tic for the second model is a test of the null hypothesis that the observed covariances were generated by the process described in the professional purity thesis. Put differ- ently, the test statistic is for the null hypoth- esis that the professional purity model is a "correct" model of the social process that leads to the observed pattern of relationships among the variables. The chi-square statis- tic for this model equals 7.41 (d.f. = 3, p = .06). At a conventional alpha level of p < .05, this result suggests marginal evidence for rejecting the hypothesis that the profes- sional purity model generated the data. Comparing the chi-square statistics and de- grees of freedom for the two models pro- vides a test of whether including direct ef- fects of client type on prestige significantly improves the fit of the model. The chi- square for this test equals 7.01 (d.f. = 2); the fit of the model improves significantly (p = .03) when direct effects of client type on prestige are permitted.12

l2 The small number of observations in the models means that one or two particularly influ- ential observations may have substantial effects on the estimates. To investigate this, I reestimated the models on two separate covariance matrices, each of which was calculated after de- letion of the four most influential observations (as measured by Cook's distance) in the profes- sional purity and client-type models. Both sets of estimates suggested notable reductions in the ef- fects of client type when purity is controlled, and both sets of estimates presented marginal evi-


The results of these analyses suggest support for both theories of the sources of legal pres- tige. Variation across the fields of law in measures of the purity of lawyers' tasks be- haves, to some extent, as predicted by the thesis of professional purity. Available mea- sures of professional purity account for a substantial portion of the effects of client type on prestige. At the same time, available evidence also suggests that an effect of cli- ent type on prestige remains after profes- sional purity is controlled. It is possible that the inclusion of additional or more refined measures might permit a more definitive ad- judication between the two theories. Such an investigation awaits more detailed data than those which, to my knowledge, are currently available. For the present, both theories re- ceive some support.

Though Abbott (198 1, 1988) presents pu- rity as a matter of the peculiarly profes- sional, some of the findings in support of the purity thesis are consistent with research on prestige across the occupational structures of modern complex societies. A lawyer's power to supervise and review the work of other lawyers is indicative of his or her authority, a factor known to be positively associated with prestige across occupations (Spaeth 1979). In the present analysis, the measure of authority is professional, but this quality arises in part because it is reported by pro- fessionals in reference to their work as pro- fessionals. Further, in the division of labor in law firms, the lawyers who have the pro- fessional authority that permits professional purity may also tend to have organizational power that comes from managing the organ- ization's relationships with valuable clients (Nelson 1988). Power confers prestige (Len-ski [I9661 1984; Treiman 1977), and the fact that many prestigious fields of law provide both opportunities to do professionally pure work and opportunities to manage relation- ships with such clients suggests that lawyers may value more than quintessentially legal work. Future research should attend to ques-

dence for including direct effects of client type on prestige. These analyses indicate that the sub- stantive findings are robust to deletion of the most influential observations.


tions of the relationship between profes- sional authority, influence that comes from the control of important relationships with clients, and professional purity. As a grow- ing share of lawyers comes to work in larger and larger organizations (Galanter and Palay 1991; Heinz, Nelson, and Laumann forth- coming), understanding the role of organizations' divisions of labor in creating the distribution of authority, influence, and pu- rity within the profession will become in- creasingly important.

As is the case with most studies of pres- tige pursued in the last 50 years, the present one has relied on information provided by questions that elicit respondents' reports of others' prestige judgments. Findings of rela- tive consensus in such reports suggest that the values revealed are the prevailing values of the profession. These values may prevail because they are truly generally shared, or they may prevail because they are merely dominant and dissent is relatively hidden. Future work on prestige would benefit from new data, of a type seldom collected, that permits researchers to distinguish between respondents' own beliefs and their reports of the beliefs of others.


During the past few decades, most prestige research has focused narrowly on occupa- tional prestige, and debates have centered on measurement and scaling techniques (e.g., see Coxon, Davies, and Jones 1986; Hauser and Warren 1997; Nakao and Treas 1994; Nam and Terrie 1982; Powers 1982). This work is important, but it has diverted atten- tion from consideration of what prestige means and what its distribution reveals about the group in which judgments are held. The present investigation has returned to a fun- damental question in the study of prestige: What are the relevant dimensions of the di- vision of labor for a group that is an apprais- ing audience of that labor.

In returning to that question, this investi- gation has enlarged our understanding of the social organization of the legal profession. The problems presented by different kinds of clients demand different kinds of work. Cli- ent type clearly structures the division of tasks among lawyers: Whom a field serves strongly conditions the amount of diagnosis, (Galanter 1974; Hadfield 2000). Legal work inference, and treatment in lawyers' work. for corporations is more likely to hold out The service of corporate clients, especially, the rewards not only of high pay and the appears to permit the emergence of profes- chance to use complex legal skills, but also sionally pure work. No doubt this happens of the chance to do work that is regarded as in part because such clients present "purer" truly "lawyerly." Hence, it is not surprising problems, but the greater economic re-that such work tends to attract more talented sources of such clients-their ability to fund lawyers (Hadfield 2000), even those not large teams of lawyers to work on their prob- strongly motivated by a desire for wealth or lems for long periods of time-are also im- for contact with the wealthy. Heinz and plicated. At least some of the values of the Laumann's (1982:385) conclusion appears legal profession appear to be intrinsic to the correct: There do exist "two legal profes- profession, in that work that approaches the sions . . . two systems of justice, separate center or core of the profession is highly re- and unequal." The findings here suggest that garded. This finding suggests that, at least this inequality is reinforced by those legiti- to some extent, the profession is organized mating values of the profession that are in- around a core of abstract knowledge, as trinsically professional. Abbott (198 1, 1988) has proposed.

This aspect of the profession's social or- Rebecca L. Sandefur is completing her Ph.D in ganization and the attendant orientation to- Sociology at the University of Chicago. In fall 2001 she begins her appointment as Assistant

ward professional purity leave lawyers dou-

Professor of Sociology at Stanford University.

bly at risk of co-optation. Studies of lawyers

Her publications include "A Paradigm for Social

advocating for the poor have found that, de-

Capital" (with Edward 0.Laumann, Rationality

spite these lawyers' firm rejection of the

and Society, 1998, vol. 10, pp. 481-501) and

"core economic values of society," they are

"The Changing Value of Social Capital in an Ex- quickly bored and discouraged by the rou- panding Social System" (with Edward 0. tine nature of much of their work serving Laumann and John P. Heinz, pp. 21 7-33 in Cor-this population (Katz 1982). These frustra- porate Social Capital and Liability, edited by R. tions lead such lawyers to seek work that is Th. A. J. Leenders and S. M. Gabbay, Norwell, MA: Kluwer Academic Publishers). Her disser-

more "professionally significant," either in-

tation investigates change in American lawyers'

volving "comple[x] . . . legal issues or . . .

careers over the course of the second half of the

[large] political or fiscal stakes" (Nelson and

twentieth century. In addition to preparing this

Trubek 1992:210). Opportunities for such

work for publication, she is beginning work on a significant, complex, and respected work are comparative study of labor markets and a study much scarcer in personal legal services than of historical changes in the life-course timing of they are in work for large organizations educational attainment.

Appendix Table A. Correlations, Covariances, and Means for Variables Used in the Analyses, 40 Fields of Law: Chicago Lawyers Survey, 1995
Variable     (1)     (2)     (3)     (4)     (5)     (6)     (7)     (8)     (9)     (10)
(1) Percentage of     1.00     -    -    -    -    -    -    -    -    -
business clients     (.077)
(2) Percentage of     -.42     1.00     -    -    -    -    -    -    -    -
personal clients of     (-,015)     (.018)                                 
low social status         [.I751                                 
(3) Days per month     -.55     .63     1.00     -    -    -    -    -    -    -
in trial courts     (-.626)     (.341)(16.841)                         


(Appendix Table A continued on next page)

(Appendix Table A continued)


Variable (1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

    Days per month in -.27 .34 .42 1.00 -----appellate courts (-,010) (.006) (.241) (.019)

    Frequent supervision .49 -.47 -.55 -.28 1.00 -----
    and review of other (.021) (-,010) (-.354) (-,006) (.024)
    lawyers' work [.387]
    Frequent decision- .10 .24 .10 -.01 .O1 1.00 ---making in the (.003) (.003) (.044) (-.OOO) (.OW) (.OlO) management of one's [.361] organization

    High intellectual .67 -.30 -.50 -.07 .21 -.07 1.00 --(legal) challenge (.033) (-,007) (-,373) (-,002) (.006) (-,001) (.033)

    Percentage of -.05 -.26 -.02 -.05 .13 -.31 -.09 1.00 -practitioners (-,002) (-,005) (.Oll) (-.OOl) (.003) (-.004) (-.002) (.019) younger than 35 [.294]
    Percentage of .28 .18 .02 .07 -.06 .27 .33 -.41 1.00 practitionersolder (.005) (.002) (.006) (.001) (-,001) (.002) (.004) (-,004) (.005) than 55 [.084]
    Prestige .64 -.53 -.58 .02 .52 -.06 .59 -.06 .03 1.00

(.224) (-.088)(-2.996) (.004) (.102) (-,008) (.134) (-,011) (.003) (1.578) [-.797]

Note: Numbers in parentheses are covariances; means are presented in square brackets. Public International Law and Admiralty Law are excluded.


Abbott, Andrew. 1981. "Status and Strain in the bridge, MA: Harvard University Press. Professions." American Journal of Sociology

Carlin, Jerome E. 1962. Lawyers on Their Own: 86:819-33. A Study of Individual Practitioners in Chicago. . 1988. The System ofProfessions: An ~s- New Brunswick, NJ: Rutgers University Press. say on the Division of Expert Labor. Chicago,

Coxon, A. P. M. and P. M. Davies, with C. L. IL: University of Chicago Press. Jones. 1986. Images of Social Stratification:

American Bar Association Commission on Pro- Occupational Structures and Class. London, fessionalism. 1986. In the Spirit of Public Ser- England: Sage. vice: A Blueprint for Rekindling Lawyer Pro- Davis, Kingsley and Wilbert E. Moore. 1945.

fessionalism. Chicago, IL: American Bar As- "Some Principles of Stratification." American sociation. Sociological Review 10:24249. Auerbach, Jerold S. 1976. Unequal Justice: Law-

Derber, Charles. 1982. "The Proletarianization of

the Professional: A Review Essay." Pp. 13-34 New York: Oxford University Press. in Professionals as Workers: Mental Labor in Berger, Joseph, Susan J. Rosenholtz, and Morris

yers and Social Change in Modern America.

Advanced Capitalism, edited by C. Derber.

Zelditch, Jr. 1980. ''Status Organizing Pro- Boston, MA: G. K. Hall.
cesses." Annual Review of Sociology 6:479-Derber, Charles and William A. Schwartz. 1991.

"New Mandarins or New Proletariat?: Profes- Berger, Peter L., Brigitte Berger, and Hansfried sional Power at Work." Research in the Soci- Kellner. 1974. "Excursus: On the Obsoles- ology of Organizations 8:71-96. cence of the Concept of Honor." pp. 83-96 in Freidson, Eliot. 1986. Professional Powers: A The Homeless Mind: Modernization and Con- Study in the Institutionalization of Formal sciousness. New York: Vintage. Knowledge. Chicago, IL: University of Chi- Bourdieu, Pierre. 1984. Distinction: A Social cago Press. Critique of the Judgement of Taste. Cam


-. 1994. Professionalism Reborn: Theory,

Prophesy, and Policy. Chicago, IL: University of Chicago Press.

Gaetke, Eugene R. 1989. "Lawyers as Officers of the Court." Vanderbilt Law Review 42:39


Galanter, Marc. 1974. "Why the 'Haves' Come Out Ahead: Speculations on the Limits of Le- gal Change." Law and Society Review 20 (Fall):95-160. . 1999. "'Old and in the Way': The Com- ing Demographic Transformation of the Legal Profession and Its Implications for the Provi- sion of Legal Services." Wisconsin Law Re- view 1999:1081-1113.

Galanter, Marc and Thomas Palay. 1991. Tournament of Lawyers: The Transformation of the Big Law Firm. Chicago, IL: University of Chi- cago Press.

Goldethorpe, John H. and Keith Hope. 1974. The Social Grading of Occupations: A New Ap- proach and Scale. Oxford, England: Clarendon.

Goode, William J. 1978. The Celebration of He- roes: Prestige as a Control System. Berkeley, CA: University of California Press.

Gordon, Robert W. 1988. "The Independence of Lawyers." Boston University Law Review 68: 1-83.

Hadfield, Gillian K. 2000."The Price of Law: How the Market for Lawyers Distorts the Jus- tice System." University of Michigan Law Re- view 98:953-1006.

Hauser, Robert M. and John Robert Warren. 1997. "Socioeconomic Indexes for Occupa- tions: A Review, Update, and Critique." Sociological Methodology 27: 177-298.

Heinz, John P. and Edward 0. Laumann. 1982. Chicago Lawyers: The Social Structure of the Bar. New York and Chicago, IL: Russell Sage Foundation and American Bar Foundation.

Heinz, John P., Robert L. Nelson, and Edward 0. Laumann. Forthcoming. "The Scale of Justice: Observations on the Transformation of Urban Law Practice." Annual Review of Sociology.

Heinz, John P., Robert L. Nelson, Edward 0. Laumann, and Ethan Michelson. 1998. "The Changing Character of Lawyers' Work: Chi- cago in 1975 and 1995." Law and Society Re- view 32:751-75.

Kahl, Joseph A. 1957. The American Class Struc- ture. New York: Rinehart.

Katz, Jack. 1982. Poor People's Lawyers in Transition. New Brunswick, NJ: Rutgers Uni- versity Press.

Kordana, Kevin. 1995. "Law Firms and Associ- ate Careers: Tournament Theory versus the Production-Imperative Model." Yale Law Journal 104: 1907-33.

Laumann, Edward 0. 1965. "Subjective Social Distance and Urban Occupational Stratifica-

tion." American Journal of Sociology 71:26


. 1966. Prestige and Association in an American Community: An Analysis of an Ur- ban Stratification System. Indianapolis, IN: Bobbs-Merrill.

. 1973. Bonds of Pluralism: The Form and Substance of Urban Social Networks. New York: Wiley.

Laumann, Edward 0. and John P. Heinz. 1977. "Specialization and Prestige in the Legal Pro- fession." American Bar Foundation Research Journal 1977: 155-216.

Lenski, Gerhard. [I9661 1984. Power and Privi- lege: A Theory of Social Stratification. Chapel Hill, NC: University of North Carolina Press.

Nakao, Keiko and Judith Treas. 1994. "Updating Occupational Prestige and Socioeconomic Scores: How the New Scores Measure Up." Sociological Methodology 24: 1-72.

Nam, Charles B. and E. Walter Terrie. 1982. "Measurement of Socioeconomic Status from United States Census Data." Pp. 29-42 in Measures of Socioeconomic Status: Current Issues, edited by M. G. Powers. Boulder, CO: Westview.

Nelson, Robert L. 1988. Partners with Power: The Social Transformation of the Large Law Firm. Berkeley and Los Angeles, CA: Univer- sity of California Press.

Nelson, Robert L. and Laura Beth Nielsen. 2000. "Cops, Counsel, and Entrepreneurs: Construct- ing the Role of Inside Counsel in Large Corpo- rations." Law and Society Review 34:457-94.

Nelson, Robert L. and David M. Trubek. 1992. "Arenas of Professionalism: The Professional Ideologies of Lawyers in Context." Pp. 177- 217 in Lawyers Ideas/Lawyers Practices: Transformations in the American Legal Pro- fession, edited by R. L. Nelson, D. M. Trubek and R. L. Soloman. Ithaca, NY: Cornell Uni- versity Press.

Parsons, Talcott. 1963. "On the Concept of In- fluence." Public Opinion Quarterly 27(Spring):37-62.

Payne, Monique R. and Robert L. Nelson. 2000. "Divided Opinions: The Political and Eco- nomic Values of Chicago Lawyers, 1975 and 1995." Paper presented at the annual meeting of the Law and Society Association, May 2000, Miami, FL.

Powers, Mary G. 1982. Measures of Socioeco- nomic Status: Current Issues. Boulder, CO: Westview.

Rosen, Robert Eli. 1989. "The Inside Counsel Movement, Professional Judgement and Orga- nizational Representation." Indiana Law Jour- nal 64:479-553.

Shils, Edward. [I9681 1994. "Deference." Pp. 197-203 in Social Stratification: Class, Race, and Gender in Sociological Perspective, edited by D. B. Grusky. Boulder, CO: Westview.

Smigel, Erwin 0. 1964. The Wall Street Lawyer: Professional Organization Man? Glencoe, IL: Free Press.

Spaeth, Joe L. 1979. "Vertical Differentiation among Occupations." American Sociological Review 44:746-62.

Strodtbeck, Fred L., Rita M. James, and Charles Hawkins. 1957. "Social Status in Jury Delib- erations." American Sociological Review 22: 713-19.

Treiman, Donald J. 1977. Occupational Prestige in Comparative Perspective. New York: Aca-

demic Press.

Tumin, Melvin. 1953. "Some Principles of Strati- fication: A Critical Analysis." American So- ciological Review 18:387-94.

Webster, Murray, Jr. and James E. Driskell Jr. 1978. "Status Generalization: A Review and Some New Data." American Sociological Re- view 43:220-36.

Wegener, Bernd. 1992. "Concepts and Measure- ment of Prestige." Annual Review of Sociology 18:253-80.

Veblen, Thorstein. [I8991 1994. The Theory of the Leisure Class: An Economic Study of Insti- tutions. New York: Penguin Books.

  • Recommend Us