¹ Munger (1988: 60), for example, notes the general failure of early studies to find a simple direct relationship between social development and increasing rates of litigation; see Laurent (1959), Toharia (1974), Grossman and Sarat (1975), Friedman and Percival (1976a), and Daniels (1980, 1982). For criticisms of current theory see Munger (1986b), Daniels (1984), and Heydebrand and Seron (1986).

² For a brief biography of Durkheim and his work see Coser (1971).

³ This is not a novel observation. Several writers have commented upon the Durkheimian influence in this area. See, e.g., Daniels (1984).

⁴ For Durkheim's formal statement justifying his line of analysis see The Rules of the Sociological Method [1895] (1964b). I should note that in The Elementary Forms of Religious Life [1912] (1965), regarded by many as his single greatest work, Durheim finally began to develop an explicit macro-micro link that was absent from his earlier work (see Alexander, 1984; Alexander and Giesen, 1987). The link, however, was constructed by employing a more affective and less rational view of individual behavior than I propose in this essay.

⁵ One should note that many of these studies include more than one dimension. For instance, Grossman and Sarat (1975) are also interested in economic determinants of litigation.

⁶ Most of these remarks are addressed to the lack of micro-level theory. However, the lack of micro-level data can also cause difficulties. Within Durkheim, the absence of such data creates the potential for aggregation problems. Although the arguments suggest that certain people more than others will be influenced by their structural situation in society and thus will be more likely to commit suicide (or sue), we do not know if this is so because the analysis does not reach the individual level. For a discussion of aggregation problems in Durkheim, see Selvin (1965).

⁷ For a discussion of this problem in Durkheim, see Douglas (1967) and Hunt (1978).

⁸ In Division of Labor, the movement from repressive to restitutive law is the objective indicator of a moral change; the weakening of the collective conscience accompanying a societal movement from mechanical to organic solidarity. On this view see Merton (1965), Aron (1967), Lukes and Scull (1983), and Duncan (1984).

⁹ Durkheim's argument on the nature of law as an indicator is itself complex. Mechanical solidarity (the solidarity of similitude) creates a high level of collective conscience that is reflected in repressive laws that punish those who somehow violate this shared moral understanding. Organic solidarity (the solidarity of differences), however, has a more complex relationship to the legal indicator. Organic solidarity weakens the collective conscience and thus might lead to less repressive law, but it is not clear why this would cause restitutive law, rather than a simple diminution of law. Presumably restitutive law is a response to a social need to facilitate the increasing number of exchanges and interactions that typify an organically solid society. Related to this theoretical problem is the methodological problem of what rate we should be examining: restitutive law as a percentage of all law? Or restitutive law per capita? The choice is not unlike the choice between Friedman and Percival's (1976a) and Lempert's (1978) way of looking at the Alameda and San Benito counties data.

¹⁰ As I shall note below, a variety of rates are associated with disputing and litigation.

¹¹ Even with respect to suicide, however, complexities abound. Durkheim was plagued with the problem of different types of suicide, ultimately settling on three: egoistic, altruistic, and anomic. Egoistic and anomic suicide are not always easily distinguishable; for his effort to do so, see Durkheim ([1897] 1951: 258).

¹² The most consistent “finding” in longitudinal studies has been the shift in caseload away from property and contract cases and toward tort and family law. More recent studies usually focus upon one or two areas of law.

¹³ See Stookey (1986) for a distinction between what I am calling the dispute rate and the litigation rate. See Grossman et al. (1982: 97) for examples of what I have called the dispute rate (percentage of households reporting a dispute) and litigation rate (percentage of disputes filed in court).

¹⁴ One has to be careful about terms if confusion is to be avoided. I have used the terms “activity,” “disputes,” and “litigation.” Miller and Sarat (1980–81: 563) speak of “grievance,” “claim,” “dispute,” and “civil legal dispute.” For the sake of simplicity I have collapsed their first three stages into the term “dispute,” while their “civil legal dispute” is similar to my “litigation,” or at least the idea of legal activity. In many situations, however, it is worthwhile to think of grievance and claiming as stages between activity and disputes, especially when attempting to describe the micro process whereby one recognizes an event as something that raises a problem and begins to act on this recognition (see Felstiner et al., 1980–81; Mather and Yngvesson, 1981).

¹⁵ Munger (1988: 68) attributes this argument to Hurst (1956). Krislov (1983) attributes similar arguments to Fuller (1934). Those familiar with Krislov's article will find parallels in the following discussion, for I find myself in substantial agreement with his arguments.

¹⁶ This argument might also suggest a higher litigation rate. If, as Galanter (1983b) argues, court decisions have radiating effects, doctrinal and conceptual uncertainty may weaken the strength of the rays. If, as a result the parties are less certain of their rights, they may find it more difficult to settle a dispute without the intervention of the legal process. For a recent work arguing that litigation is a coping mechanism for a community under stress from social change, see McIntosh (1990).

¹⁷ This change might be in either direction, that is, depression or expansion (see Stookey, 1986; Munger, 1988).

¹⁸ See Friedman (1985). One might supplement this argument by adding that disputes in simplex and episodic relationships may involve a greater amount of relative conflict of interest and therefore would be more likely to be litigated rather than compromised (see Axelrod, 1970; Menkel-Meadow, 1983).

¹⁹ And, confoundingly, an increase in the number of disputes as activity increases.

²⁰ For such an argument see Toharia (1974), Grossman and Sarat (1975), and Kidder (1975).

²¹ For a more structural version of a cultural argument see Miller and Sarat (1980–81). The cultural argument plays an important role in cross-cultural analyses (see Miyazawa, 1987; Wagatsuma and Rossett, 1986).

²² In this passage Munger is criticizing what he calls the “normative effects model.” My sense, however, is that the comments, insofar as they suggest the uncertainty surrounding the predictions of various theories, are more generally applicable.

²³ For a further elaboration, see Coleman (1990).

²⁴ Not all theories have an implicit micro component. Some are built on a methodological holism in contrast to the methodological individualism that, as Coleman (1986) notes, grounds sociology in a theory of action. The first chapters of Suicide are an appeal for a type of methodological holism. Heydebrand and Seron (1986: 319) appear to join in such an appeal. They reject “an individualistic, reductionist bias ... focusing on disputing and litigious behavior [in which] the question of social and legal conflict is collapsed into one of individual motivation.” Of the current research, their prescriptions for progress are probably furthest from those presented here. As will become obvious below, however, this essay shares with Heydebrand and Seron's work a desire to move beyond rates as the dependent variable.

²⁵ The structure of this figure is borrowed from Coleman (1986: 1322). The examples have been changed, however, to reflect problems in the longitudinal study of courts.

²⁶ The tendency to treat such independent variables as “determinants” can in part be traced to the Durkheimian heritage. As Alexander and Giesen (1987: 8) note, “Durkheim's later theory became the ‘classical’ referent for sociologists who believed in the subjectivity of action but considered it to be ordered in a strictly macro, antivoluntaristic way.”

²⁷ See Priest and Klein (1984). For a review of attempts by economists to model litigation and settlement decisions and the empirical evidence bearing on these models see Cooter and Rubinfeld (1990).

²⁸ Here as elsewhere there is a trade-off between simpler models that are mathematically tractable and more complex models that are not. From this perspective there is no one right model of human action (Kahneman et al., 1987; Plott, 1987).

²⁹ “Sociologists have as their distinctive mission in the science of rational decision to analyze the social structural causes of rationality” (Stinchcombe, 1986: 8).

³⁰ Investigators who are in other ways rooted in different traditions are apparently coming to agree about the importance of legal culture. Thus Friedman (1985) explains the growth of law in twentieth-century America as the product of a developing legal culture of plural equality. Friedman (1987: 373) notes that in California in the last two decades of the nineteenth century, at least with respect to tort law, “The legal culture of the time was a culture of low expectations with regard to tort compensation. The rules of law suggest this; so does the behavior of juries. Munger in an article on tort litigation in West Virginia explains the relatively low level of industrial accident litigation in part by noting:

³¹ The exact content of the concept of legal culture is still in flux. Efforts to borrow directly from the concept of political culture have enjoyed limited success (Grossman et al., 1982). Friedman (1985) adopts a narrower definition based in part upon the idea that injuries will be compensated.

³² For a discussion of the idea of “rule logics,” see Lempert and Sanders (1986).

³³ For instance, the litigation rate for injury producing accidents is less than 1 percent in Japan, and lawyers are involved in out-of-court settlement in as little as 3 percent of the cases. American filings and lawyer participation are at least ten times higher.

³⁴ To be risk averse is to fail to assign a linear value function to risky choices. Most people would choose a sure $800 to an 85 percent chance at $1,000. Individuals also tend to be loss averse because the subjective value function is steeper for losses than for gains. Most people are reluctant to bet on a fair coin for equal stakes.

³⁵ Out-of-court settlements are sometimes thought of as situations in which repeat players may enjoy a bargaining advantage. Friedman (1987) discusses the accident compensation process in later nineteenth-century California in which employees typically bargained away their right to common law remedies in case of injury in exchange for employer relief plans. In Friedman's (1987: 371) words, this was “surely a bad bargain” for the employee. Companies also bargained with consumers (e.g., railroad passengers) in a way that is much like the way insurance claims adjusters respond to people harmed by their insureds (Ross, 1970). The implicit psychology embodied in such efforts is that a risk-averse individual will be unable to refuse such offers. Of course, arrangements such as the contingent fee may improve an individual's bargaining position, but as Friedman (1983: 24) points out, this system can only operate when legal rules create a surplus that can be used to pay the attorney.

³⁶ For a particular example of the effect of risk spreading, see Kagan (1984).

³⁷ A number of writers have called for efforts in this direction: see Lempert (1978), Friedman (1983), and Daniels (1986).

³⁸ Galanter (1990: 373–74) calls this theory that litigation patterns should “follow the ups and downs in the quantity of the underlying activity” the “underlying activity” hypothesis. Clearer insights often accompany the realization that rates cannot be understood as simple aggregations of the underlying activity as, for example, when we recognize that plea bargaining cannot be understood solely as a response to case pressure. See, for example, Padgett's (1990) insightful article on the history of plea bargaining. Padgett argues that long-range changes in plea bargaining discounts were influenced by, among other factors, improvements in the average strength of the state's case and the legal culture of judges.

³⁹ Coleman (1986: 1330) gives as another example of a competition effect the economic return to the individual of increased educational attainment. The return exists in the short run given a certain supply of educational attainment. If, however, overnight everyone had one more year of education, we would not expect to see this reflected in our individual paychecks. The macro effect is not a simple aggregation of individual effects.

⁴⁰ Of course it is possible that some landlords will not even conclude that a failure to pay rent in the full amount or on time constitutes a dispute. The economic climate might alter perceptions of when it is appropriate to claim.

⁴¹ Perhaps the “lumpy” nature of court supply is part of the reason we do not ordinarily hypothesize or test equilibrium models of litigation rates. Nevertheless, there are certain suggestions in the literature of some stability to litigation rates over time (McIntosh, 1983).

⁴² The court response need not necessarily be a product of conscious thought. Several economists have argued that selective litigation may produce more efficient laws (Priest, 1977; Goodman, 1978). As Cooter and Rubinfeld (1990) note, if a rule's efficiency is negatively correlated to the probability that litigants will test it in court and if efficiency is not negatively correlated to the probability of a rule surviving such a test before a judge, the law should at least weakly evolve toward efficiency.

⁴³ Cooter and Rubinfeld (1990) made this point in an early draft with respect to some law and economics analyses: “The law and economics literature on trial outcomes has recently begun to make use of principal-agent theory and game theory. The theories frequently lead to indeterminate predictions in part because the authors are unwilling or unable to put sufficient institutional structure on the problem to narrow down the theoretical possibilities.”

⁴⁴ As Coleman (1987) notes, Durkheim's description of the suicide rate as a social fact and his ongoing polemic against social psychology sometimes causes us to loose sight of the fact that he was engaged in the explanation of individual, micro behavior. Rates per se are not macro effects.

⁴⁵ Heydebrand and Seron have also been working on this aspect of litigation studies. Their work has in part been an effort to define and measure a new set of dependent variables for the longitudinal study of courts.

⁴⁶ For an example of a similar line of analysis see Burstein and Monaghan (1986) on populations of education cases over time.

⁴⁷ Galanter (1990: 372 n.1) is reluctant to call something as large and as sustained as automobile accidents a congregation, and yet is willing to entertain the idea that there is a congregation of malpractice cases.

⁴⁸ Other factors that Galanter (1990: 377–78) lists as influencing the emergence of a definable group of cases include the availability of legal services and the probability of success suggested by a set of legal rules.

⁴⁹ Administrative agency hearings, small claims courts, and juvenile courts all alter the landscape confronting a person with a dispute. While the sources of some of these changes may be outside the scope of the longitudinal study of courts, their consequences are not, for they directly influence the costs and benefits facing the individual with a dispute. Today this is an especially important issue as courts and lawyers confront and attempt to capture the alternative dispute resolution movement (see Seron, 1990). As Friedman (1983: 23) notes, “we could shed light on the inner history of courts if we knew more about the history of their rivals—arbitration for example.”

⁵⁰ The macro-micro interaction does not of course end here. We live in a world of nonrecursive relationships. As Galanter (1990) remarks, the response of the courts to a congregation of cases may either encourage or discourage other litigants or even other congregations.