Volume 2, No. 1 (January, 1992) pp. 6-10

TOURNAMENT OF LAWYERS: THE TRANSFORMATION OF THE BIG LAW FIRM by Marc Galanter and Thomas Palay. Chicago and London: Univ. of Chicago Press, 1991. 197 pp. Cloth $27.50.

Reviewed by John P. Heinz, American Bar Foundation and Northwestern University School of Law.

Big law firms are now in big trouble. They are getting rid of considerable numbers of both partners and associates, and some are going out of business. The firms have been hurt by the decline in the economy -- especially, by the slump in real estate development and by the cooling of the 1980s enthusiasm for corporate takeovers (mergers and acquisitions, leveraged buyouts) -- but the impact of the decline may be greater because the firms' growth in recent decades was based on a pyramid or Ponzi scheme that had to keep expanding in order to survive. The basic thesis of Galanter and Palay's book, Tournament of Lawyers, is that characteristics of the internal labor markets of large law firms made it efficient for the firms to promote to partnership a fixed (or, perhaps, increasing) percentage of their associates, thereby causing the firms to grow at an exponential rate. Thus, "a firm that began with one person in 1920 and grew by 10 percent per year would still have only six lawyers in 1940... [but the] same firm that grew to six attorneys prior to World War II would have forty-five attorneys in 1960 and three hundred five lawyers a mere twenty years after that" (p. 79). When the level of demand is no longer sufficient to sustain that growth, however, a crisis will occur and the system may collapse. That may be what we are seeing now.

There is more to be said about the dynamics of these internal labor markets, but let me first comment on the book more generally. This is an exceptionally well-written piece of work. The arguments are clearly and simply stated, and on the whole are very persuasive. The work is informed by broad reading in the literature on lawyers and law firms, as well as by a sophisticated grasp of economic theory. In sum, this book deserves to take its place alongside Erwin Smigel's The Wall Street Lawyer and Robert Nelson's Partners With Power on the short shelf of essential works on the large law firm.

The book includes a concise and informative chapter on the history of the creation of the large firms. It presents several examples of the point, which has been well-documented by Robert Gordon, that the commercialization of law practice has now been decried for at least a century. Mature lawyers of every generation seem to have felt that professionalism was going to hell in a handbasket. But it is fun to know that Learned Hand said in 1925:

... in my own city the best minds of the profession are scarcely lawyers at all. They may be something much better, or much worse; but they are not that. With courts they have no dealings whatever, and would hardly know what to do in one if they came there. (Quoted at p. 7.) This was written at a time when such advocates as John W. Davis were in their prime. The book also includes a chapter summarizing the nature of the large law firm in its "golden age," circa 1960. This is presented as a

Page 7 follows:

point of comparison and contrast to the "transformation" of the large firm into what it had become in the late 1980s. All of this is very nicely done and is well-documented.

For the most part, Galanter and Palay are appropriately cautious about their generalizations and conclusions. In one instance, however, wishful thinking may have led to overstatement. They assert (at p. 57):

Barriers against Catholics, Jews, women and Blacks have been swept away. The social exclusiveness in hiring that was still a feature of elite law practice in 1960 has receded into insignificance. Performance in law school and in the office counts for more and social connections for less.

One would like to believe that this were true. And perhaps it is, but Galanter and Palay do not present proof of the proposition. Almost every other factual observation in the book is meticulously documented with either statistical evidence or a citation of authority. It is significant, I think, that the footnote to the passage quoted above cites only sources that make the contrary point -- i.e., that social exclusion exists -- and the footnote includes the observation that "even without deliberate exclusion, selection on the basis of educational credentials and the candidates' social affiliations, personal preferences, and career expectations will maintain some degree of association between legal roles and the social origins of lawyers. (footnote 131 at p. 57)"

The main story in this book, however, is the story of the growth of the large firms, and what drove it. Galanter and Palay plot the actual growth in the size of several firms from the 1920s until the mid-80s, and fit those data to three alternative models of the growth patterns. They observe that the firms display a steeper rate of growth after 1970 than before. One of their models of the growth patterns, therefore, is a "kinked linear function" -- growth by a constant number of partners per year, but with a "kink" or change in the number at 1970 so that there is a steeper ascent after that point. They also observe, however, that the rates of growth increase both before and after 1970. They therefore assess the fit of an exponential growth function -- i.e., growth by a constant percentage each year. But the exponential function does not account for the sharp change in the rates of growth in 1970. Finally, therefore, they fit a "kinked exponential function" -- an exponential function with a change in 1970 to a higher rate -- and they conclude that this last model most closely matches the actual patterns of growth.

It is important to note the nature of the sample of firms that Galanter and Palay use. They start with fifty of the firms that were among the largest in 1986, and then trace the growth histories of those firms. This may be, of course, quite different from starting with the fifty firms that were largest in 1930, 1940, or 1950, and then tracing the histories of those firms up to the mid-80s. That is, by starting with fifty that were among the largest in 1986, they focus on cases of success. They exclude the firms that fell by the wayside over the years or that failed to grow. What we have, then, is largely a story of how the successful firms got where they are. We need to keep this firmly in mind.

Page 8 follows:

But what accounts for the growth of these successful firms? Why do they grow at an exponential rate rather than by some fixed number of lawyers per year or in a more irregular pattern? The authors' thesis rests on the premise that the partners in these firms have more human capital than they can exploit by themselves, and that this capital (or, at least, some part of it) is transferable to other lawyers. Human capital consists of intelligence and skills, education and experience, professional reputation, and relationships with clients. Some kinds of lawyers can share this capital with juniors -- e.g., they can give them useful information and can attract clients -- while other assets cannot be readily transferred (trial lawyers cannot give someone else their courtroom presence). The nature of most of the work done by large firms permits this sharing of assets. By using the labor of junior lawyers, then, the partners can more fully exploit the value of their own capital, and thus increase their profits. The amount of such labor that the partners can use depends upon the amount and kind of their assets.

Once the associates are hired, the partners then need some mechanism to assure that these junior lawyers exploit the partners' capital properly and productively. The partners need to be sure that the associates will work hard, will refrain from stealing the partners' clients or trade secrets, will not damage the firm's reputation, and will not leave the firm prematurely (i.e., before the partners have recovered the full cost of the training that the firm has given the associate). But monitoring the work of the associates is difficult and costly. The level of supervision that would be required to protect the partners' interests would take too much of the partners' time. Therefore, the partners need some mechanism to supplant or supplement monitoring. Galanter and Palay argue that the mechanism used by these firms is what they call the "promotion-to-partner tournament." Essentially, the firm promises to promote the best of the associates after a period of service to the firm, usually about six to ten years. During those years, the associates compete for the prize of partnership. But the associates also need some assurance. They need to know that the firm will not cheat them -- i.e., that it will not benefit from their labor and then decline to promote them. The associates will want the firm to take in enough new partners each year to give the associates a reasonable chance of becoming a partner (i.e., the odds against them are not unacceptably large). To provide this assurance, the firm implicitly promises the associates that a more-or-less fixed percentage of them will become partners after the requisite period of years. It is this promise that creates the exponential growth rate.

The associates are also reassured by two additional assumptions about the interests and behavior of the firms: First, since it would be irrational for the firm to promote the less able and hard-working associates because that would diminish the quality of the partnership, the associates can assume that the tournament will be conducted fairly and the race will go to the swift. Second, the promise to create a substantial number of new partners gains credibility because of the firm's need to hire new associates. If the firm breaches its promise, the associates will go elsewhere. And then the firm would no longer have enough labor to exploit its human capital:

Growth occurs because, at the end of the tournament, the firm must replace not only the losing associates who depart, but also all those who win and are promoted. If the firm did not hire associates

Page 9 follows:

to replace its newly promoted partners, then the pretournament partners would share their surplus human capital with fewer associates and, therefore, make less money (at p. 102).

Galanter and Palay's argument is more subtle and more elegantly detailed than my brief summary can suggest, but I have tried to capture the essence of it.

Their thesis explains the pattern of exponential growth, but it does not explain why the rate of growth accelerates in 1970 -- i.e., the 1970 "kink" or "shock." Galanter and Palay say that their "overall objective has been to explain the portion of growth not accounted for by this shock" (at p. 110). But this leaves about half of the post-1970 growth unexplained. Growth by a fixed percentage (i.e., the exponential rate) accounts for only 50 percent of the growth after 1970. The authors briefly survey an inventory of some explanations that have been suggested, but none of these satisfactorily accounts for the sudden (and quite uniform) change in 1970. It is the suddenness of the change that is so puzzling. If half of the growth is accounted for by the requirements of the firm's internal labor markets (the promotion-to-partner tournament), the other half might well be explained by factors external to the firms, especially perhaps by increases in demand for the firms' services. But what was so unusual about 1970? What happened then to cause such a sharp change in the rate of increase rather than a more gradual, phased change? It is the kink in the growth curves that remains unexplained.

The amount of growth in recent years is, in any event, quite substantial. The authors report:

Between 1960 and 1985 the average size of New York's biggest firms increased almost 375 percent, from an average of 45 lawyers to 214 lawyers. The increase of the [firms outside New York] was somewhat greater, growing 425 percent, from an average of 50 to 261 lawyers (at p. 61).

But remember that the sample includes only the firms that were largest in 1986. This report, therefore, refers to the growth rates of the most successful firms.

In spite of this substantial growth, however, the share of the market for legal services that is controlled by the largest law firms is still relatively modest, by comparison to other professions. Galanter and Palay note, for example, that in 1982 the eight largest law firms received only 1.7 percent of receipts for legal work, while the big eight accounting firms received 28.3 percent and the eight largest engineering, architecture, and surveying firms received 16.4 percent of receipts in their fields. Even the fifty largest law firms held only a 7.1 percent share of the market while fifty accounting firms held a 36.3 percent share and the fifty biggest engineering, architecture, and surveying firms held 34.6 percent of that market. (At Table 7.)

Will the big law firms continue their rapid exponential growth until they finally achieve a degree of market concentration that is comparable to that of the accounting or the engineering/architecture firms? Perhaps that question should now be put in the past perfect conditional: Would the firms have continued to

Page 10 follows:

grow but for the decline in demand for their services? The authors point out that the internal labor markets of the big accounting firms are very different in structure from those in the large law firms. While the associate-to-partner ratios in the law firms are on the order of one or two to one, the ratios in the Big Six accounting firms range from about ten to one to seventeen to one. Therefore, if the law firms were to become more like the big accounting firms, the law firms would need to have far more associates per partner and thus far more surplus human capital in order to make those associates productive: "To increase the associate-to-partner ratio the firm would have to dramatically increase the value of its human capital per partner" (p. 123). But what is the meaning or cause of the difference between law and accounting firms in their associate-to-partner ratios? Are we to conclude that the partners in accounting firms possess, on the average, about ten times as much transferable human capital as do partners in big law firms? This would mean that the ability of the accountants to attract business or to impart specialized knowledge to their subordinates is ten times that of the law partners. Is this plausible? If it is not, then the number of associates per partner would not appear to be dependent upon the amount of human capital possessed by the partners. And, if that is the case, a fundamental premise of the Galanter and Palay thesis is in doubt. If the number of associates hired by the firms is not determined by the level of the partners' excess human assets, then the firms' rates of growth may not be driven by the same sort of inevitable need. It would not be so necessary, then, for the firm to replace every associate promoted to partner, or the firm might decide to add many more associates even though the human capital of the partnership had not increased.

Perhaps the principal determinant of the structure of the internal labor markets of the varying types of firms, and thus of their rates of growth, is not the amount of their surplus capital but another factor identified by Galanter and Palay (as pointed out above), the cost of monitoring the work of subordinates. It may be that the cost of monitoring accounting subordinates, on the average, is substantially lower than the cost of monitoring the work of law associates. If the work of the accountants were more routine, for example, that might well be the case. The marginal return to an accounting firm from hiring another associate might then be much greater than the marginal return to a law firm.

Is the amount of human capital that is possessed by a firm ascertainable apart from the firm's behavior in the hiring of associates? That is, is this element of Galanter and Palay's theory verifiable? How are we to operationalize the value of the firm's reputation, or of its established relationships with clients, or of its ability to transfer specialized information or skills? Galanter and Palay do not, of course, attempt this, nor do they tell us how it might be done. But it would certainly be interesting to have some evidence on the point.

This is a thoughtful and well-argued book. It gives us a great deal of information about the innards of large law firms, and it stimulates musings about future paths for research. This is a very substantial accomplishment.