I. Introduction

In 1991, Galanter and Palay set forth a model for understanding the organizing principle that frames the relationships between partners and associates in elite law firms (Reference Galanter and PalayGalanter & Palay 1991). Building from the economic tournament model developed by (Reference MalcomsonMalcomson 1984), whose work builds upon (Reference CarmichaelCarmichael's 1984) model of seniority systems, Galanter and Palay present a tournament model predicated on the exchange of surplus human capital for labor. They argue that the system of incentives established by the “up-or-out” tournament structure permits parties who operate in a system of imperfect information and distrust to enter into an economic arrangement that minimizes the long-term risk of either party cheating the other.

This article considers one aspect of this model: that the tournament works as a monitoring device to ensure that associates will not engage in opportunistic behavior by “shirking,” or failing to exert maximum effort or develop professionally; “grabbing,” by taking a partner's client; or “leaving,” by going somewhere else and taking the firm's investment of training with them (Reference Gilson and MnookinGilson & Mnookin 1985, Reference Gilson and Mnookin1989; Reference Galanter and PalayGalanter & Palay 1991). I present qualitative data from a study of all Silicon Valley law firms that have taken equity from their emerging growth company clients (dot-coms) as an integral part of their fee for representing these companies.

The tournament model is a widely used mechanism to control opportunistic behavior. However, this mechanism can only operate in certain economic (and social) circumstances. In the absence of those circumstances, the model breaks down, and with it the ability to control opportunism in the absence of some alternative mechanism. Specifically, the tournament model only works if associates are willing to participate. To participate, they must want the deferred prize. If they do not value partnership positively, the tournament model breaks down.

Prior research has not investigated whether the utilization of a tournament model prevents the opportunistic behaviors identified as grabbing, leaving, and shirking. The study tests this hypothesis by examining whether these opportunistic behaviors have occurred when external economic forces stressed the tournament model. The study provides initial evidence that shirking and leaving, though not grabbing, have become far more prevalent as many of the characteristics of the tournament model are blurred. I situate the findings of the study within the literature, applying tournament theory to understand the social processes that structure and govern relationships between partners and associates in law firms. The presentation and discussion of this study is intended to both examine the applicability of tournament theory to explain law firms' governance structures and also to further our understanding of tournament theory.

In Part II, I outline the Cravath system and Galanter and Palay's application of tournament theory, with particular attention to their thesis that the tournament works as a governance mechanism to prevent opportunistic behavior. In Part III, I discuss critiques raised by economists and legal scholars that pose important challenges to whether the tournament model is a useful analytical tool for evaluating law firm relationships or whether a tournament takes place at all. I discuss alternative explanatory models, as well as modifications to the model that consider it a worthwhile heuristic device that must take account of the varied motivations that players have within the tournament, as well as the ways in which the tournament is unfair. In Parts IV, V, and VI, I discuss this study of Silicon Valley law firms. In order to help understand the mechanisms behind firms' responses to opportunistic behavior, in Part VII I describe one firm's decisionmaking processes in confronting these perceived phenomena. Part VIII situates these findings within the legal tournament literature.

II. The Cravath System and Galanter and Palay's Tournament of Lawyers

Paul D. Cravath, founder of the New York firm Cravath, Swaine and Moore, is credited with developing the organizational structure and system of incentives currently known as the “Cravath system”Footnote ¹ (Reference SwaineSwaine 1946). This system involves hiring lawyers directly out of law school, only hiring the most gifted law students, paying them more than what they would be able to earn elsewhere, requiring that they devote all of their efforts to the firm's clients and not work for anyone else, and giving them graduated increases in responsibility through an intensive apprenticeship. At the end of the apprenticeship, five to nine years later, either the associate becomes a partner at the law firm or arrangements are made for the associate to leave the firm. Other firms copied the Cravath system, which quickly became the primary structural model for prestigious law firm development (Reference SwaineSwaine 1946).

Reference Galanter and PalayGalanter and Palay, in Tournament of Lawyers: The Transformation of the Big Law Firm (1991), trace the emergence of the elite large law firms to the turn-of-the-century emergence of office technologies (see also Reference HansmannHeydebrand 1989). As telephones, typewriters, an explosion of printed legal materials, and the need for legal research and information retrieval systems developed, these technologies transformed the legal environment in the direction of the business model, requiring higher costs for overhead, capital investment, and new economies of scale (Reference Galanter and PalayGalanter & Palay 1991). This transformation was also accompanied by a change in client relations, from the episodic representation of businesses and individuals to the continual representation of organizations capable of providing steady streams of revenue and requiring specialized services (Reference Galanter and PalayGalanter & Palay 1991).

Reference Galanter and PalayGalanter and Palay (1991) postulate that the experienced and successful lawyer has made a number of human capital investments, which the lawyer is concerned with protecting. For example, the lawyer has made human capital investments in developing skill and expertise as a lawyer in developing relationships with clients, in cultivating a reputation, and in developing new business. An experienced and successful attorney will have a surplus of this human capital, more than the lawyer has the capacity to convert into money through the lawyer's own efforts (Reference Galanter and PalayGalanter & Palay 1991).

The successful and experienced lawyer's surplus human capital is essentially nonrival; that is, the use of that surplus human capital by another does not diminish the value of the asset to the lender. For example, up to a certain point, the use of a park by others does not diminish the owner's enjoyment of the resource. Of course, certain aspects of the attorney's human capital cannot be shared, such as charisma or courtroom skills, but many productive aspects of the modern legal professional's life can be (Reference Galanter and PalayGalanter & Palay 1991). The attorneys who own the human capital would like to lend their surplus to another so as to maximize their return on the human capital they cannot realize. On the other side, some attorneys have little human capital investments. Instead, they have their productive labor capacity and, owing to the requirement of having to have a law degree, are part of a limited market with the skills to productively utilize the surplus human capital of another (Reference Galanter and PalayGalanter & Palay 1991). To illustrate the foundations for this exchange, the authors recount the first meeting between Thomas Shearman, founder of Shearman and Sterling, and David Dudley Field:

These two parties may wish to enter into a mutually beneficial economic agreement.

For Galanter and Palay, the law firm is “an institutional arrangement for conducting these activities as a governance mechanism” (Reference Galanter and PalayGalanter & Palay 1991:92). The organizing principle of the firm addresses the three economic concerns that an attorney with surplus human capital encounters owing to the opportunistic possibilities of the one to whom the capital is lent: leaving, grabbing, and shirking. The borrower attorney could leave, thereby taking the firm-specific skills that the lender attorney has invested in the borrower but on which the lender has not yet realized a return (Reference Galanter and PalayGalanter & Palay 1991:94). That is, the lender attorney invests time, energy, and information in training the borrower attorney, who may then take these skills elsewhere, including to a rival firm. The borrower attorney might grab by taking the clients of the lender attorney and appropriating them (Reference Galanter and PalayGalanter & Palay 1991:92). The lender attorney wants the borrower attorney to contract with the lender attorney to perform services, which the lender attorney will then resell to the client. The determinant of the profit to the lender attorney from entering into the transaction is the difference between the cost of the borrower attorney and the pay from the client (Reference Galanter and PalayGalanter & Palay 1991). The lender attorney does not want the borrower attorney to contract directly with the client.

Finally, the borrower attorney could shirk by failing to put in enough labor to maximize the human surplus capacity of the lender attorney or failing to make the human capital investments that the borrower attorney must make to further develop the future prospects of the lender attorney (Reference Galanter and PalayGalanter & Palay 1991:94). This fear is compounded by the realization that legal work can be difficult to quantify. It is not always clear what qualifies as a high level of output, given the subjective aspects of life as a professional. The doctor may do an exceptional job, but the patient may die anyway.

These concerns lead to the organizational structure of the firm: in “… a world with transaction costs, attorneys attempting to lend shareable human capital will require internal organization—that is, a firm—in order successfully to govern the transaction” (Reference Galanter and PalayGalanter & Palay 1991:92). The lender and the borrower agree that the borrower will devote 100% of the borrower's time and efforts to working for the lender attorney. “Branding provides a relatively inexpensive method of letting the world know who has the right to use her capital. But rather than burning her initials into A's forehead, P simply makes him part of an organization” (Reference Galanter and PalayGalanter & Palay 1991:98). Certainty is thereby provided to the lender's clients that the borrower attorney is authorized and skilled to handle their business, and potential counterfeiters are prevented from making similar claims.

Galanter and Palay argue that, having decided that the firm is the best economic model for governing the relationship between the parties, firms must still face the difficult opportunistic possibilities of leaving, grabbing, and shirking. So they “employ a complex monitoring scheme to protect themselves from opportunistic associates” (Reference Galanter and PalayGalanter & Palay 1991:99). An essential part of this scheme involves deferring pay to associates; that is, the firm pays them less than the resale value for the services they render. The firm then has the possibility of firing the associates if they engage in opportunistic behavior, thereby preventing them from recouping the deferred portion of their income.

To address the possibility of opportunistic behavior, the firm employs a “promotion-to-partner tournament” (Reference Galanter and PalayGalanter & Palay 1991:100). In this tournament, associates receive gradual incremental increases in pay to compensate them partially for deferred pay from previous years and to signal to them that the firm recognizes their increased skills. After a certain time, typically five to nine years, the firm will promote a percentage of each cohort to partner. At that time, they will be able to recoup their deferred income, as well as the deferred income of the associates that the firm has let go. The firm must promote a reasonably certain percentage of each cohort to partnership or associates will believe their lottery ticket to be worthless and will potentially engage in opportunistic behavior (Reference Galanter and PalayGalanter & Palay 1991). The firm promotes associates to partnership on the basis of merit to signal to associates that effort is rewarded and to avoid filling the partnership with unproductive lawyers. If a firm does not follow the rules of the tournament, it will be difficult to recruit associates, and the firm will no longer be able to maintain the monitoring mechanisms brought about by the participants' good faith playing in the tournament (Reference Galanter and PalayGalanter & Palay 1991).Footnote ²

III. Critiques of Tournament Theory as Applied to Law Firms

Since its introduction, Galanter and Palay's application of tournament theory has become the definitive model to explain the governance structure of law firms (Reference Samuelson and JaffeSamuelson & Jaffe 1990; Reference KronmanKronman 1993; Reference MalcomsonMalcomson 1984; Reference RibsteinRibstein 1998; Reference OrtsOrts 1998; Reference Sander and WilliamsSander & Williams 1992). However, the theory has faced powerful critiques, both from scholars who believe it requires significant modification and from those who believe there is no tournament operating (Reference KordanaKordana 1995; Reference Rutherglen and KordanaRutherglen & Kordana 1998; Reference Wilkins and GulatiWilkins & Gulati 1998; Reference JohnsonJohnson 1991; Reference LambertLambert 1992). A discussion of the continuing vitality of tournament theory as an analytic tool to understand the social processes governing the internal structures and relationships of law firms must take account of this literature.

Much of the criticism of tournament theory as applied to law firms centers on the question of the difficulty of monitoring the conduct and performance of associates in law firms (Reference KordanaKordana 1995; Reference Rutherglen and KordanaRutherglen & Kordana 1998; Reference MehtaMehta 1998; Reference Sander and WilliamsSander & Williams 1992; Reference HansmannHansmann 1996; Reference DemsetzDemsetz 1995; Reference Lazear and RosenLazear & Rosen 1981). For Galanter and Palay, the quality of an associate's output cannot be measured in terms of the quantity of the billed hour since (1) associates are reputed to greatly exaggerate their hours; (2) the quality of an hour spent looking for a file is unequal to the quality of that same hour spent on serious research; (3) associates often work in teams such that measuring individual input to the project is difficult; and (4) it is costly and time-consuming for partners to have to review, measure, and quantify this output (Reference Galanter and PalayGalanter & Palay 1991, Reference Galanter and Palay1998). The quality of the lawyer's input, measured in terms of the professional development of the lawyer, is likewise difficult to monitor or evaluate.

For Reference Galanter and PalayGalanter and Palay (1991, Reference Galanter and Palay1998), the tournament model provides a solution to the problem of monitoring. In essence, the tournament awards a deferred premium, or a super-bonus consisting of tenure, prestige, and the expectation of a high salary. Associates are motivated to remain employed by the firm, work hard, and develop the long-term professional qualities valued by the firm. The firm, in turn, is motivated to promote based on merit, as (1) the future profitability of the partnership is dependent upon the revenue-generating ability of the incoming partners and (2) to not do so would signal to associates that the firm is not playing by the rules, and they would lose their incentive to strive to win the tournament (Reference Galanter and PalayGalanter & Palay 1991; Reference Gilson and MnookinGilson & Mnookin 1989).

Reference KordanaKordana (1995) and Reference Rutherglen and KordanaRutherglen and Kordana (1998) challenge the basic underlying assumption that associate output is difficult to monitor. They point out that it is not difficult to assess how hard associates are working because associates meticulously bill the hours they have worked. Further, the quality of their work product is likewise not difficult to monitor as partners are expected to review and revise their work product (Reference KordanaKordana 1995; Reference Rutherglen and KordanaRutherglen & Kordana 1998; Reference Gilson and MnookinGilson & Mnookin 1985; Reference HansmannHansmann 1996). Further, even if many associates are working on a case, the writing of specific documents, or sections of documents, typically reflects individual effort and is easily separated from the work product of other associates (see also Reference Leibowitz and TollisonLeibowitz & Tollison 1980). They conclude that since the difficulty of monitoring is the basis of the tournament model, and the monitoring of associates' output is readily achievable in law firms, the tournament model is inapplicable to explain law firms' organizational structure.

In addition, the authors question why, if the costs of monitoring associates are so substantial, partners are not also subject to the tournament model (Reference KordanaKordana 1995; Reference Rutherglen and KordanaRutherglen & Kordana 1998). That is, if a tournament is necessary to prevent associates from grabbing, leaving, and shirking, partners, no longer subject to the tournament, should have an incentive to engage in these opportunistic behaviors. That partners are not subject to such a tournament indicates that the tournament model is inapplicable to explaining the governance structure of firms.

A further critique notes that associates join firms and make career track decisions for a variety of reasons that may be unrelated to the playing of a tournament for partnership (Reference KordanaKordana 1995; Reference Rutherglen and KordanaRutherglen & Kordana 1998; Reference Wilkins and GulatiWilkins & Gulati 1998). For example, associates might join a firm because of the high salaries available for associates relative to other job prospects, to gain training that will be useful in other occupations, or for the prestige that being an attorney with an elite firm might impart. None of these reasons for joining a firm are related to the playing of a tournament, and the tournament would not sufficiently inhibit opportunistic behaviors.

It has also been noted that the tournament model would be a highly unstable economic model for a firm to adopt. The standard tournament model assumes that a certain percentage of associates will be promoted to partnership each year. Yet this model ignores outside economic forces (Reference Rutherglen and KordanaRutherglen & Kordana 1998; Reference NelsonNelson 1992). During poor economic times, firms may not be able to promote many associates to partnership. During better times, they may promote more. Associates make decisions throughout their careers concerning the attractiveness—in terms of lifestyle, working conditions, and salary—of other options available to them. The standard tournament model does not take into account these outside economic forces. Kordana's quantitative research indicates that firms have not promoted a consistent percentage of each cohort to partnership. He notes that “careers within a firm do not occur against a stable economic background outside the firm. As the attractiveness of the other job opportunities available to associates outside of their current law firm fluctuates, the incentives provided by the tournament also change dramatically” (Reference KordanaKordana 1995:1919–20).

Reference KordanaKordana (1995) and Reference Rutherglen and KordanaRutherglen and Kordana (1998) suggest that a production-imperative model explains the relationships structuring law firm organization better than Galanter and Palay's application of tournament theory. Kordana suggests that firms hire associates in such numbers as they believe will be profitable in the long run, taking account of economic conditions and the demand for legal services (Reference KordanaKordana 1995; Reference Rutherglen and KordanaRutherglen & Kordana 1998). Associates come to the firms because they offer salaries that are higher than the other options available to the associate coming out of law school. Associates also know that some associates may be ultimately offered partnership, subject to economic and political conditions. In the meantime, associates use the firm to gain training and experience. They are also intensely aware of their marketability outside of the firm, either at other firms or in a wide variety of possible occupations. The complex social interplay of these factors determines the organizing structure of law firms and the course of associate career patterns, rather than the tournament (Reference KordanaKordana 1995; Reference Rutherglen and KordanaRutherglen & Kordana 1998).

Reference Wilkins and GulatiWilkins and Gulati (1998) accept that law firms do follow a form of a promotion-to-partnership tournament model and that this model provides some institutional incentive structure with regard to opportunistic behavior. However, they carefully note six ways in which the particular organizing structure of law firms differs from the standard economic model of a rank-order economic tournament: (1) associates have a variety of reasons for joining firms, some of which are unrelated or only tangentially related to playing in the tournament; (2) associates do not have an equal opportunity of winning the tournament due to a variety of discriminatory variables; (3) the interests of individual partners in monitoring, mentoring, evaluating, and advocating for associates may differ from those of the firm; (4) partners are not subject to a tournament, and a variety of mini-tournaments take place throughout an associate's career, rather than just a partnership decision; (5) the actual partnership decision is based, in large part, on factors not measured by past performance, such as “rainmaking” potential; and (6) the social processes through which the tournament operates are not transparent but rather are cloaked in secrecy (Reference Wilkins and GulatiWilkins & Gulati 1998).

Galanter and Palay see these distinctions as a “second-order phenomenon that might be expected whenever a recruitment device designed for institutional purposes provides opportunities for participants to pursue interests of their own” (1998:1687). They acknowledge that Wilkins and Gulati have enriched and furthered our understanding of the application of tournament theory to law firms. However, Galanter and Palay deny that they appropriated a static economic model and attempted to rigidly apply it to the myriad social interactions that operate to govern opportunistic behavior in firms. Rather, “our connection with the theory was one of affiliation rather than application” (Reference Galanter and PalayGalanter & Palay 1998:1684).

Wilkins and Gulati do not advocate abandoning tournament theory as an explanation of the internal labor market of law firms. Instead, they propose an alternate model in which the tournament operates as a useful heuristic device in which competition for partnership is one element of a complex incentive system. The goal of this incentive structure is to “motivate every associate to work hard with little supervision, while at the same time ensuring that the firm has a sufficient number of trained associates to satisfy its staffing and partnership needs” (Reference Wilkins and GulatiWilkins & Gulati 1998:1588). The tournament metaphor operates as a foundational building block through which the internal labor market of the firm can be elaborated. They defend the idea that monitoring and training are difficult and costly to measure. They also note that during Cravath's era, the internal governance structure of law firms more closely approximated the standard economic tournament model, and that this historical legacy still exerts an institutional force (Reference Wilkins and GulatiWilkins & Gulati 1998). Further, many associates, especially among the senior associate ranks, are motivated to win what they perceive to be a tournament by making partner.

Beyond the quest for partnership, Wilkins and Gulati elaborate three elements that provide incentives to associates to not engage in opportunistic behaviors. First, firms offer high “efficiency” wages for jobs that are relatively scarce (1998:1636). Associates know they can be replaced and will have difficulty finding similar employment opportunities elsewhere. There is thus a sizeable gap between the salary an associate can make at a firm and the salary an associate can make in other available jobs. Second, associates enjoy high levels of prestige in working for firms, and the fact of their employment operates as a market signal of the value of the associate as a lawyer. Associates do not want to risk losing these “reputational bonds” and the external labor market value that accompanies them (Reference Wilkins and GulatiWilkins & Gulati 1998:1638; see also Reference SpenceSpence 1974; Reference Baird, Gertner and PickerBaird, Gertner, & Picker 1994). Third, firms provide a very high level of general training in the skills necessary to be a good lawyer. Firms offer to provide this ongoing training, associates recognize the need for this mentoring, and this provides an incentive for associates to stay at firms, exert high levels of effort, and avoid the opportunistic behaviors that could endanger these prospects (Reference Wilkins and GulatiWilkins & Gulati 1998:1641).

Galanter and Palay respond to Reference Wilkins and GulatiWilkins and Gulati's (1998) critique as being a “question of how a theoretical construct like the tournament relates to the messy, diverse, changing reality of large law firms” (1998:1683). Reference Wilkins and GulatiWilkins and Gulati (1998) see the internal labor market of the firm as involving a complex interplay between a tournament, relatively high salaries, reputation bonds, and the opportunity for generalized training. These factors combine to structure and manage incentives in firms in such a way that associates do not engage in opportunistic behaviors. For Galanter and Palay, these variables “enlarge our understanding of how the tournament is contextually embedded in the American setting and provide a rich portrait of the micropolitics of the law firm” (1998:1690). However, for Galanter and Palay, their use of tournament theory encompasses most of the elaborations presented by Wilkins and Gulati. The tournament operates as a “cluster of devices” that includes elements of relational capital, signaling markers, and varying and diverse rationales for participation (Reference Galanter and PalayGalanter & Palay 1998:1691).

IV. Data and Methods

In order to test the limits of the tournament model, it is necessary to find particular historical moments when the economic environment has radically challenged assumptions/premises of the model. The dot-com bubble in Silicon Valley provides precisely such a time and place. The study therefore examines a universe comprising all law firms with eleven or more lawyers and with an office in Silicon Valley that were able to accept equity with respect to their representation of emerging growth companies.Footnote ³ I excluded the few firms that did not take equity as a matter of public policy (i.e., divorce lawyers, criminal defense lawyers, estate planning lawyers) or because the type of law practiced did not lend itself to taking equity (i.e., insurance defense or maritime law firms). I defined Silicon Valley broadly to include both Santa Clara and San Mateo counties in California.

I determined the size of a law firm based on the number of lawyers employed by the firm.Footnote ⁴ I sought to interview partners at firms that employed eleven or more lawyers. I was interested in the organizational dynamics, practices, policies, and culture of firms that accept equity. As such, I did not want to include solo practitioners or small groupings of lawyers who share office space and resources but lack a hierarchical organizational structure. Further, this decision was facilitated by the use of the Reference Martindale-HubbellMartindale-Hubbell Law Directory (1999), which separates law firms comprising more than ten lawyers from law firms of less than ten lawyers. I used the Martindale-Hubbell Law Directory as a primary source for size data on law firms. Martindale-Hubbell provides the most comprehensive list of law firms in the country. It publishes an annual directory of law firms, differentiated by number of attorneys.Footnote ⁵

To revise this listing of firms to include firms that could accept equity, I used the assistance of three informants: a retired judge in Silicon Valley, a partner in a San Francisco law firm, and an associate employed by a Silicon Valley law firm. These informants were knowledgeable with respect to the type of law practiced by firms in Silicon Valley. I asked the informants to review the list of firms and suggest any firms that they were aware of in Silicon Valley that were in a position to accept equity that were not on the list, and then to highlight any firms that they knew could not accept equity for public policy or legal specialty reasons. I removed a firm from the list if two or more informants independently stated that they had actual knowledge that the firm could not accept equity. I also removed a firm from the list if one informant noted that the firm did not or could not accept equity and I was able to confirm this information through either the firm's Web site or through a telephone call to the firm. I also asked the first ten respondents whom I interviewed to similarly review the list, and either added the firm or removed the firm if the information could be corroborated through the firm's Web site or through a telephone call.

The final list of firms that accepted equity was sixty-three. I interviewed at least one partner from each of these firms.Footnote ⁶ I contacted respondents by snowball sampling. After interviewing a partner, I would ask for a list of partners at other firms who would be the most knowledgeable with respect to the policies and practices of their firms. I would then contact those people and continue the process. Snowball sampling of organizations can introduce bias into a study in that the sample firms and the respondents within the firms may not be representative of the universe but rather of a selected group or network within a universe. These potential biases are always of concern. In this case, there are two reasons I believe bias through snowball sampling is less of a concern. First, I interviewed partners about matters that were within the knowledge of all partners. I was interviewing partners about the policies, practices, resource allocation, and strategic decisions of the firm. Partners (sometimes called shareholders or members, depending on how the firm has incorporated) are owners of the firm and are knowledgeable about the organizational behavior and decisionmaking processes of the firm. Partners generally attend all partnership meetings and participate in the making of decisions concerning firm governance issues. While it is possible that a particular partner's views could introduce bias, for the most part, any partner from the Silicon Valley office of a firm was in a position to discuss the policies, practices, and strategic decisionmaking processes of the firm with respect to taking equity. Further, I often interviewed more than one partner from a firm, and thus I was able to corroborate much of the information. And many of the firms' policies and practices were a matter of public record that I could verify independently. Second, there should be no bias in the sample firms, as I interviewed at least one partner from every firm in the universe. The sample in this case was the universe.

My unit of analysis was the organization. My focus in the study was on the ways in which the tournament model operates at law firms as a governance mechanism to prevent associates from engaging in theoretically predicted opportunistic behaviors. That is, I wanted to study the ways in which extreme economic forces stressed the governance structures of the firms to allow us to more closely examine the stability of these mechanisms and the response of firms to this perceived phenomenon. While it would have enriched the study to have the perspective of associates, time and resource constraints did not allow for this. Individual associates will have unique and personalized motivations for their actions and may not be aware of how the firms' policies and practices act as a governance mechanism to restrain them as a cohort from engaging in opportunistic behavior. The study should not be perceived as a labor study, but rather as an organizational study.

From November 2000 through November 2001, I conducted 105 semi-structured interviews, which were all taped and transcribed verbatim. I spoke with at least one partner from every law firm in the universe. The interviews were confidential. I identified six firms that were central in setting policies and procedures and whose practices were closely surveilled and copied by other firms in Silicon Valley. In five of the six “core” firms, I spoke with three to four partners. In one of the core firms, which I discuss in Part VII, I conducted six interviews. To add texture and a further level of understanding to my findings, I interviewed (1) four partners from law firms in the San Francisco Bay Area who do not have Silicon Valley offices but were identified by informants and respondents as playing a frequent role in equity deals; (2) eight principals from venture capital funds; and (3) two founders of dot-com companies.

V. Stresses to the Tournament

During the Internet bubble, demand for lawyers to represent dot-coms dramatically increased, as did opportunities for making riches far in excess of the former market rate for billable hours. Firms responded by devoting more and more resources into servicing the burgeoning dot-com startup companies who competed with one another for the attention of the elite firms. Firms positioned to service these startups, and especially the elite six core firms, responded internally by dramatically increasing their recruitment efforts and hiring associates. All of the firms stated that they were leveraged higher than they ever had been. While estimates varied, firms routinely said they had a ratio of 6:1 firmwide, with a core of 8–12:1 in departments that serviced dot-com clients.

The criteria for hiring also changed from the manner prescribed by the Cravath system.

Firms had little time to devote to the process of interviewing and vetting associates, and the opportunities they could leverage for hiring young lawyers, even if those lawyers were not a suitable match for the firm, were great. Firms also began taking associates who were in other divisions of the firm, or at other branch offices of the firm, and transitioning them into a type and kind of law foreign to their training. Finally, firms intensified their attempts to hire associates from other firms.

As the firms' associate-partner leverage increased, staffed based primarily on the criteria of whether the lawyer had a “butt to put in a chair” and firms plugged associates directly into deals, the character and quality of mentoring dramatically decreased.

Traditionally, associates spend one to two years in what could be called an intensive apprenticeship period. Cases are staffed by, typically, at least one partner, at least one junior to senior associate, and a new associate. Traditionally, firms do not expect to make a profit from their new associates. Many of the billable hours they accrue are “written off,” as it naturally takes them far longer to prepare documents than an experienced attorney should take, and more than can be reasonably billed to a client. Further, the documents prepared by new associates must be meticulously, repeatedly, and extensively redlined.

The skills involved in elite firm practice are largely acquired on the job, not though law school. Junior associates are also given little direct contact with clients, and will acquire much of their education in client relations and management through watching their mentors, who will often invite the junior associate to observe these interactions to assist them in their development.

Internally, firms will have new associates attend many training sessions, they will have one or two partners be responsible for their development, and senior associates are expected to play a role in integrating them within the firm and assisting in the socialization aspects of the firm. Critically, the firm, either through the mentoring partner or through attorneys specifically designated for these purposes, makes certain that the associate receives a variety of work experiences. The firm wants to make certain that the new associate learns the skills involved in being an elite law firm partner.

To be able to compete effectively in the tournament, associates need to develop skills in researching, effective writing, advocacy, client management, business development, time-effective management, and team participation, as well as internal management and socialization. While attorneys will specialize, even within their specialty they require skills and knowledge of all aspects of the practice relating to their specialty. For example, a partner specializing in the representation of emerging growth companies requires an extensive skill set, often acquired after more than a decade of continued and varied learning.

The skills of traditional lawyering, which partners would have acquired through their apprenticeship, are applicable to a wide variety of legal arenas. These same partners would be more than competent enough to serve a role in a litigation case if necessary, or to write an effective brief in a malpractice case, and at the very least have developed the judgment to know what they do not know and to leverage the most effective resources within the firm to respond to the issue. Associates who never acquired these skills have failed to develop professionally and were not in a position to meaningfully compete in the tournament.

During the dot-com frenzy, firms were hiring associates because of an immediate desire to use their labor to increase short-term profitability for the individual partners and for the firm. Partners involved in dot-com representation were too busy to mentor associates. Further, there were so many associates being hired that it would have been impossible to effectively mentor them. Firms reduced or cancelled training sessions. Instead, partners who specialized in the representation of emerging growth companies wanted new associates to work on their many cases immediately.

New associates were often given full case responsibility and did not have access to busy partners even to ask questions, have documents reviewed, or make strategic decisions.

Not only was their time not written off by the firms, but the firms also charged premiums for the representation of the clients.

Part of what permitted these changes to take place was the nature of the work being done and the climate in which it was taking place. One of the striking features of the law firms' representation of these dot-com companies is that there were few negative consequences to poorly done legal work.

The venture capital community dictated most of the essential terms of initial financing, and attempts at creativity were more than likely to be viewed as red flags or warning signs that something was not quite right with the transaction. A new associate could look at the ways in which the last deal was constructed, and deals before that, and use those documents as templates or forms for the current representation. It was not necessary that the associates understand the transaction or improve their legal skill set in order to continue to process the cases. However, it is necessary for them to develop these skills if they are to compete in the tournament.

The climate of Silicon Valley during this time also permitted and encouraged the use of junior associates on cases. The prevailing mood among the venture community, largely followed by the relatively unsophisticated dot-com companies, was that the quality of the legal documents was largely unimportant. Companies were being formed and either going public or failing within a year. The speed at which documents needed to be prepared, and the lack of experience and supervision of many of those preparing the documents, led to a climate in which speed prevailed over quality. The quality of the legal services provided was notoriously poor.

Firms immediately put associates to work on cases, filling out forms and putting together the same cookie-cutter documents in case after case.

As a result, the development of associates as attorneys was extremely limited. Many of them became skilled only at filling out the forms and repetitive documents that were required in the early stages of emerging company representation.

Further, as the quality of the documents was not stressed by clients or opposing counsel, and young lawyers did not receive supervision or training, many did not learn to think critically about the documents they were preparing over and over again.

These stresses in the Cravath system were also driven, and mutually influenced, by the expectations of associates. Associates informed hiring partners that their ambitions were not primarily focused on becoming partners in the firm, or even of enjoying long careers at the firm. They intended to come to the firm for a short time in order to work on emerging growth company IPO work.

Their career incentive was to either stay at the firm and realize sufficient stock option gains that they would be able to retire or transition out of law in a matter of a few years or, more often, to have access to emerging growth company clients that would enable them to develop the type of relationships that would allow them to leave the firm and go in-house to the company.

Since they did not view the practice of law, or their tenure in the firm, to include becoming a partner at the firm, their willingness to engage in or endure a long, traditional apprenticeship was reduced.

Associates had little incentive to seek professional development skills, as opposed to developing and furthering relationships with dot-com companies.

VI. Opportunistic Behavior

The radical changes in the economic environment threatened the assumptions of the tournament model in ways that would predict three deviations from the model: a greater probability of grabbing, leaving, and shirking. I discuss below the opportunistic behaviors that partners reported associates engaging in during this time period.

As previously mentioned, every Silicon Valley firm took equity unless they were either prohibited as a matter of public policy from doing so or their legal specialty did not lend itself to taking equity. Each of the firms discussed associates leaving and the difficulty that they encountered in recruiting new associates. Outside of the six core firms that did the vast majority of dot-com representation, all firms discussed associates leaving to work for one of the six core firms. They also discussed associates leaving to join dot-com companies that they knew through contacts or friends.

Shirking constituted opportunistic behaviors that, for the most part, these firms perceived to be a problem only in their limited dot-com representation departments. At the six core firms, the opportunistic behaviors of leaving and shirking were mentioned in every interview. Associates outside the six core firms were often leaving to join the six core firms and were shirking only with respect to the firms' limited dot-com representation clients or through haphazard occurrences. Associates at the core six firms were leaving and shirking with regard to the firms' dot-com clients.

No law firm reported associates grabbing clients in the way in which Reference Galanter and PalayGalanter and Palay (1991, Reference Galanter and Palay1998) and classical economists term grabbing.Footnote ⁸ When an associate left a firm to join a dot-com as in-house counsel, the firm still retained its equity and continued to represent the dot-com company in the expensive process of becoming a public company. For a dot-com company to be accepted as a client of one of the elite firms was a mark of status that signaled to potential venture investors and the market that the firm believed in its prospects for becoming successful and going public. The representation signaled that the dot-com company had made the cut, since for every client a firm was willing to represent, the vast majority of emerging growth companies that came to the firm were turned down. A firm would decline to represent the dot-com because either the firm did not have sufficient resources to staff its cases or because the firm did not think the company's prospects for going public were great enough to justify taking a chance on the company not going public. Correspondingly, a firm had a vast network of resources that could be mobilized to assist a start-up company in going public.

Grabbing in the classical sense did not occur because clients had a great interest in remaining with the firm, even after hiring away the associates. While some revenue might be lost from having the associate go in-house, this was insignificant compared to the overall revenue that the firm would generate from the representation. The reason firms were troubled by associates joining dot-coms was because they lost the ability to have these associates work on other cases. This opportunistic behavior is what Reference Galanter and PalayGalanter and Palay (1991) regard as leaving.

In the Silicon Valley case, dot-coms benefited from having one of the prestigious firms represent them and were not concerned about needing to maintain the relationship. Outside of the Silicon Valley example, firms may even encourage associates to go in-house to client companies. Having an associate in-house at a client company will help solidify and perhaps expand the relationship between the company and the firm. The in-house lawyer is usually in a position to direct or at least influence which firm a company uses. Further, the substantive work that the in-house lawyer performs will not meaningfully “steal” from the elite firm. The elite firm's revenues are most likely generated from large, complex, recurrent cases and issues, and not from the minor contract reviews or human resource policy issues that an in-house lawyer might have the time, training, and resources to perform. Any loss in revenue is minor compared with being able to solidify the relationship between the firm and the company.

In regard to leaving and shirking, I was not able to dis-aggregate leaving and shirking as variables. That is, I was not able to determine whether Firm A experienced leaving, but not shirking, nor the ways in which the attributes of Firm B had a different effect. Rather, these firms or departments spoke generally of the problem of leaving and shirking that they were experiencing, and what they perceived as the reasons for this behavior. I was also not able to collect data on the rates at which associates left firms to join other firms as opposed to joining dot-coms.

Junior associates had the majority of contact with the emerging growth clients. It was an extremely common experience for emerging growth companies to offer associates positions in their company as general counsel, chief operating officer, or a variety of other job titles and descriptions. Emerging growth companies believed that they could save a great deal of money by employing an associate full-time, rather than paying exorbitant legal fees and premiums to the firm for the associates' time.

Assuming they were funded, emerging growth companies could offer salaries competitive with what the associate was earning at the firm, along with a generous enough amount of stock options that the associate would be a paper millionaire, pending a liquidity event subsequent to a successful IPO.

Generally, partners who brought in the dot-com clients did not maintain a professional or personal relationship with the client. Partners were bringing in so many cases that their time was extremely limited and focused on those few clients whose com-panies were on the brink of going public. And associates could easily handle the myriad small, simple legal issues that confronted the relatively unsophisticated dot-com founders on a daily basis. For example, when a new dot-com company had to enter into leases, write employment contracts, obtain nondisclosure agreements, or figure out how often the board of directors meetings should occur, it would be left to the associates. As one dot-com founder told me, “When you meet one of the partners at these firms, take their picture, because it is probably the last time you'll see them unless you go public.” Had cases required greater de-grees of complexity or more technically specialized legal work, ju-nior associates could not have served this role. Companies also be-came very familiar with the associate who was assigned to their case.

Twenty-four-year-old associates were being given the opportunity to earn more in one year than lawyers who had followed the Cravath system and successfully won the tournament of lawyers might realize after twenty or more years of practice.

Firms came to regard themselves as primarily competing with their clients for their associates, rather than having associates compete with each other in the internal tournament of lawyers.

Founders of dot-com companies, and the majority of their employees, tended to be of the same age cohort as junior associates and tended to look to them as potential employees. This does not seem to be a case of “don't trust anyone over thirty.” Instead, associates, like the founders, were in a position to take a risk with their careers for the chances of short-term money.

The potential for short-term riches by leaving provided a career mobility incentive to junior associates that they would otherwise lack. Junior associates, often between twenty-five and thirty years old, had the chance to leave their firms and go in-house with dot-com companies in return for salaries comparable to what they were receiving at the firm, coupled with stock options sufficient to make them millionaires within a year or two (the time frame in which new companies were going from creation to IPO). Partners at Silicon Valley firms were earning annual draws of high six and seven figures and thus were out of reach of the dot-com companies. Further, partners were sharing more directly in the equity derived from the firm's representation of the client and thus stood to benefit far more directly than associates from companies going through liquidity events.

Associates also engaged in shirking. Partners likened the relationship between partners and associates to that between NBA coaches and their elite players.

If associates did not want to work for particular clients, or did not want to work for particular partners or departments, or were doing inadequate legal work, the firm would not take any sanctioning action against them.

Partners felt that they needed to compete with each other for the services of the associates, who felt free to pick and choose the opportunities that they wanted to pursue.

In some cases, associates who left firms to join dot-com companies would later be worth more money, at least on paper, than partners at the firm. Associates were in a position to refuse to work for partners whose cases did not involve the chance to work closely with dot-com companies. Associates could refuse to work on cases for clients they perceived as difficult to work with, or on matters that did not appeal to them for whatever reason.

Shirking, as it occurred in the case of Silicon Valley, seems to differ from the standard economic analysis. Shirking did not mean failing to work hard. There is no evidence that associates were not working long hours. Shirking, in this formulation, means failing to develop professionally by not seeking out the generalized training and skill set traditionally expected of attorneys at elite firms.

And shirking here implies failing to prioritize the cases, clients, and type of work that individual partners deemed important. While this is clearly opportunistic behavior, it should also be noted that the firms at which these behaviors occurred permitted and encouraged these behaviors. Firms were aware that these associates were not gaining generalized skills and that associates were picking and choosing which cases and partners they wanted to work with. The high demand for associate labor relative to the supply of associates, the high degree of leverage of the firms, and the very large returns that the firms were realizing from the sale of equity in companies that went public gave firms an incentive to eliminate the structures that would restrain these opportunistic behaviors.

VII. One Firm's Response to Opportunistic Behavior by Associates

In order to better understand the ways in which firms responded to the stresses on the tournament brought about by economic forces, I discuss one particular firm's decisionmaking processes. This firm was one of the six core firms I have identified. At this firm, I conducted six interviews with separate partners. I chose to discuss this firm for two reasons. First, I was able to interview six different partners. I was thus able to obtain a greater depth of information than in other firms, as well as to corroborate the information provided. Second, the firm's perception of the problem of opportunistic behavior and its responses to it implicate many of the mechanisms discussed by tournament theorists.

At a partners' meeting in late December 1999, the partners addressed what they saw as a crisis. The crisis had three interrelated characteristics. First, the firm had lost 10% of its lawyers during the course of the last year. These associates left to work for clients of the firm.

That is, they left their position as associates at the firm to become either general counsel or any of a variety of other positions at start-up companies. Start-ups, flush with venture capital, could match or beat the salaries the lawyers were receiving and in addition could offer the associates stock options that had the potential to make them millionaires when the company went public.

Second, the firm was forced to turn away potentially extraordinarily lucrative start-up representation because it did not have the manpower to staff the cases. Partners were having to turn away prospective clients whose companies stood good chances of going public and potentially making the partners and the firm millions of dollars in as short as eighteen months.

The growth of the firm came during the Internet start-up explosion of 1995–2000. In 1999 alone, the firm did 30 IPOs, 433 private and venture financings, and 47 mergers and acquisitions for high-tech companies.

Third, the associates at the firm were working harder than they had at the firm they broke away from. The firm had so strained its available resources that it was taking a big toll on morale.

The exhaustion, stress, and lack of morale, in turn, were fueling the defections. Each defection, in turn, increased the workload on the remaining lawyers.

At the meeting, partners discussed ways to resolve the situation. One issue was how to retain associates who were leaving to go to dot-com companies. The partners realized that they were competing with their clients for their key personnel, and frequently the clients were winning the recruiting war. The next issue was how to recruit more associates. It was felt that recruiting perhaps ten or fifteen more lawyers would reduce the stress level on the current lawyers and simultaneously help bring down the hours that the others were working, thereby increasing the morale of the remaining lawyers. The firm did not want to recruit more lawyers than this since the partners would be unable to mentor or supervise them. Adding these additional lawyers would also allow the firm to accept more promising clients rather than having to turn them away for staffing reasons.

However, it was discussed, the firm was having trouble recruiting talented associates compatible with its business model. At the partners' meeting, it was suggested that the firm increase associate pay to a far higher level than any other firm in the country. Overall, partners saw this as a way to distinguish the firm and permit them to hire a handful of new people, lower the hours their current associates were working, staunch the flow of associates to dot-com companies, boost the morale of current employees, and allow the firm to represent a greater number of dot-com companies that hopefully would go public.

On December 21, 1999, the partners assembled its associates and announced the firm's new compensation system. Salaries for their sixty-eight associates would be increased by approximately forty-five percent. Salaries for first-year associates were increased from $95,000 to $125,000, and they would be guaranteed a $20,000 bonus.Footnote ⁹ Fourth-year associates would make $165,000 with a $30,000 guaranteed bonus. It was crucial to the firm that the associates understand that the firm was raising salaries to hire new people to bring down the workload of associates and to boost morale. The firm was concerned that associates not think the firm was going to look at them as even more of an economic unit and expect them to work even harder.

Associates could still leave the firm to join dot-com companies for roughly the same pay, and still receive stock options that could make them millionaires if the companies went public.

Instead, while associates did receive the benefit of more money, the increase was primarily an attempt to decrease opportunistic behaviors by improving the lifestyle of the associates. The partners were hoping that paying their associates the highest salaries in the country, while reducing their hours through the increase in associates they would now be able to recruit, would serve to decrease opportunistic behaviors.Footnote ¹⁰

VIII. Discussion and Conclusion

Reference Wilkins and GulatiWilkins and Gulati (1998) claim that the complex interplay between the tournament, high salaries, reputational bond, and the opportunity to gain general training serves to inhibit associates from engaging in opportunistic behavior. In this study, each of the variables detailed by Reference Wilkins and GulatiWilkins and Gulati (1998) were implicated, and the overall effect was a partial breakdown in governance structures and the increased occurrence of opportunistic behaviors.Footnote ¹¹ Young associates, perhaps those least restrained by their prospects for partnership, were presented with opportunities to make more salary, and have the potential to make short-term riches, by leaving. The fear of losing the reputational bond developed through their law firm associations did not impede their engaging in opportunistic behavior, as they were less concerned with the market for their legal services. Similarly, the opportunity to remain at the firm in order to receive general legal training and mentoring did not prevent them from behaving opportunistically, as they were not primarily seeking to acquire legal skills and firms had dramatically cut back on such programs.

At the same time, it must be remembered that these law firms did not shut down due to a lack of associates. The vast majority of associates either did not leave or were replaced. In fact, given the high degree of leverage that many of the firms developed, and the lax hiring and mentoring, replacing more expensive older associates with a pool of younger associates was not always unwelcome. The majority of associates, especially older associates, did not abandon their specialized practices and did not leave the firm. The governance structures of the firm provided a sufficient incentive for them to not engage in opportunistic behavior.

Firms' leverage affected their capacity to restrain opportunistic behavior. This seems to provide support for Wilkins and Gulati's (1998) discussion of the ways in which general training and mentoring offer incentives to associates to not engage in opportunistic behavior. As described earlier, the departments in the firms that had dot-com clients all spoke of being so leveraged that mentoring and training suffered dramatically. Certain departments of firms were leveraged to such a degree that it was not possible to effectively monitor their performance, development, or client interaction. This failure in oversight, coupled with the shortage of partners' own time and exacerbated by the lax standards of hiring and the hiring of so many associates who could never become partner, led to a failure to effectively transmit organizational values against opportunistic behavior. Further, because of the quantity of repetitive and form-driven work that associates were assigned to perform, associates did not receive the more general legal training that over the long run would have increased their marketability. And because of the speed with which documents were expected to be finished, there was an acceptance of sloppy or error-filled work by the partners and the clients.

From the other side, many of the associates who gravitated to these firms in order to service dot-com companies were not inspired to gain these skills. Rather, they were focused on joining companies, and the firm was serving to bring them closer to these companies. This illustrates why receiving generalized legal training did not serve as a governance mechanism to prevent opportunistic behavior. For many of these associates, their reason for being at the firm was to engage in opportunistic behavior. The short time frame of the associates' expected tenure at the firm also explains why the element of reputational bond did not serve as a deterrent to engaging in opportunistic behaviors. For this element to serve as a governance mechanism, associates must view their employment at an elite firm as having a signaling role that indicates their quality as a lawyer. However, while being at these firms may have served as a signaling device that indicated to dot-com companies that these associates were of a high caliber, it would operate to speed the rate with which the associates left the firm. For example, employment by a first-year associate at an elite firm might signal to a dot-com company that the associate was someone they should seek to hire, based in part on the firm's reputation.

It is worth noting that the tournament seems to have worked as a disincentive to engage in opportunistic behaviors by the most senior associates and partners. From the literature on tournament theory, the cohort of associates most likely to be actively engaged in the tournament, and most disinclined to engage in opportunistic behaviors, is senior associates (Reference Wilkins and GulatiWilkins & Gulati 1998). In regard to engaging in opportunistic behaviors, it may be that the fact of successfully having gone through the tournament may partially act as a disincentive to engage in leaving and shirking. Firms frequently mentioned that they had sped up the rate of elevating certain key senior associates to partnership so as to deter them from leaving, as well as to assist them in cultivating new business prospects.Footnote ¹² While critics (Reference KordanaKordana 1995; Reference Rutherglen and KordanaRutherglen & Kordana 1998) have pointed out that a flaw in tournament theory is that only associates are subject to the tournament, it appears that the success of winning the tournament, and receiving the deferred compensation of past years' work, does seem to prevent partners from engaging in opportunistic behaviors. That is, the combination of high salaries, reputational bond, and the fact of partnership seem to have largely deterred partners from engaging in opportunistic behaviors during the dot-com boom.

The firm's decision to raise salaries does seem to provide some support for the position of Reference Wilkins and GulatiWilkins and Gulati (1998), Reference KordanaKordana (1995), and Reference Rutherglen and KordanaRutherglen and Kordana (1998) that money is an important element in restraining opportunistic behavior. It may also be that the notoriety of being the highest-paid associates increased the status and prestige of being associated with the firm, which decreased the incentive to behave opportunistically. However, it may also be the case that when the governance structures that restrain opportunistic behavior become stressed, increases in the quality of the work environment may become effective mechanisms to restrain opportunistic behavior.

This article has demonstrated limits to the applicability of tournament theory. Those limits are to be found in the economic environment in circumstances in which (1) exogenous reward structures offer many multiples of internal rewards; (2) demonstrably high short-term rewards outside the firm starkly contrast with the delayed long-term rewards inside the firm; (3) the managerial strata reduce their emphasis on long-term recruiting of potential partners in favor of short-term productivity by young associates; and (4) firms develop departmental leverage ratios in excess of their capacity to monitor, mentor, and train recruits. The ways in which these social forces stressed the governance structures of the firms allows us to more closely examine the stability of these mechanisms. The example of the opportunistic behaviors that took place during the dot-com frenzy reveals that the tournament model's ability to operate as a governance mechanism to restrain opportunistic behavior is vulnerable to economic forces outside of the legal system.Footnote ¹³