¹ Krislov recently discussed the problem of definitions in terms of nominalism versus essentialism, a dichotomy that generated such heat in medieval universities that it split faculties and led to the exodus of both dissident students and professors (Krislov, 1983: 163–65). Friedman makes a similar distinction between a functional definition and an institutional definition (1976: 25–27). Applying this distinction to courts and greatly oversimplifying, essentialists look for the essence of courts of law and then create a standard model that defines what is a court. Much of the debate over whether trial judges today adjudicate or administer their caseloads is premised with an essentialist definition that courts should in fact adjudicate disputes to some meaningful extent.

² Although the population of adults might make more theoretical sense, I am unable to obtain those statistics (see Lempert, 1978: 93–97). In looking at specific categories of cases, more narrowly defined bases would be preferable, such as the coal production rate (Munger, 1988: 80–82) or the mining accident rate (Stookey, 1986: 299–300).

³ For a description of the SLADE project, which collected most of the data for 1945–70 used here see Merryman et al., (1979: v–ix, 21–29).

⁴ For political units used in the regional breakdown, by country see Merryman et al., (1979: 33–35).

⁵ The capital region in Spain, therefore, includes only the province of Madrid. This differs from Spain's regional treatment in Merryman et al. (1979) and in Toharia (1974), where the capital region included Madrid plus three other urbanized provinces. These three provinces here are included in the urban region.

⁶ The difference in Chile is not made up by mining. The percentage of economically active population (EAP) in Chilean mining and quarrying declined from 3.9 percent in 1945 to 2.5 percent in 1970 (Merryman et al., 1979: 531).

⁷ It may be useful to examine the exceptions to this general pattern for clarification. In Italy (Figure 1) the semi-agrarian region had a higher litigation rate than predicted by the socioeconomic region hypothesis. This means (putting aside for the moment the possibility that the hypothesis is wrong) that some other independent variable beyond the socioeconomic variable was more influential. It is unlikely, however, to be either a legal norm variable (e.g., procedural or jurisdictional rule) or a political variable, since these were mostly uniform throughout the country at a given point in time. Alternatively, it might be a legal structure variable (e.g., number of judges or lawyers per capita), a legal process variable (e.g., court delay), a cultural variable (e.g., attitudes toward formal courts), or a social variable (e.g., existence of alternative informal dispute resolution fora). In the case of Italy the semi-agrarian region—the least developed economically—consists of most of the area south of Rome plus the islands of Sardinia and Sicily. Further disaggregation of the data into political regions, such as Sicily or Campania (including Naples), might reveal the key to why the litigation rate was so high in this area (see Ruffini, 1978: 212–14, 223–32). The most likely explanation probably lies in the fact that there were many more judges and lawyers per capita in the semi-agrarian region than in the less litigating semi-urban region throughout the period 1945 to 1950. Lawyers may stimulate demand to file civil cases, while more judges increase the supply of third party agents ready to resolve disputes, which lowers the “cost” in terms of access to prospective plaintiffs.

The ratio of judges and lawyers per 100,000 population in the two regions was as follows (Merryman et al., 1979: 448, 470):

In Spain (Fig. 2) the socioeconomic region hypothesis seems verified. Nevertheless, notice that after 1970 the litigation rates for the urban and semi-agrarian regions run very close together. It is probable that at some point of high socioeconomic development the differences in economic structure between regions are reduced so that the variable no longer has any explanatory power. By 1986, for instance, the percentage of EAP in agriculture had decreased to 14 percent in the Spanish semi-agrarian region from 23 percent in 1970, down from 38 percent in 1960 (Merryman et al., 1979: 530; Instituto Nacional de Estadística, Anuario, Año 1987 (Manual), 1987: 269–70, 643–44). The Spanish urban region during the same period saw almost all workers leave farming, as the proportion of EAP in agriculture dropped from 15 percent in 1960 to 7 percent in 1970 to 3 percent in 1986 (ibid.).

Friedman and Percival (1976a: 277, 300–301) found that civil litigation rates from their urban and rural counties in California converged after 1950 as the rural county became more urbanized and took part more in American mass culture.

⁸ “I suggest that we adopt, more or less, a ‘conceptual agnosticism’ as a research strategy in this area, assuming little more than the idea that environmental factors may affect patterns and changes in court activities and that they will be channeled through certain structural constraints (which have to do with the court system itself)” (Daniels, 1984: 754).

Wolf Heydebrand has successfully used a variant of this approach—organization theory—to study U.S. district courts (1977: 771–85; see Heydebrand and Seron, 1986: 313–20). I have tried the conceptual agnosticism that Daniels mentions in my research on federal courts (Clark, 1981: 96–148).

⁹ Daniels drew on the idea of punctuated equilibria from Stephen Jay Gould, who attacked notions in evolutionary biology, paleontology, and geology that pictured change as linear, progressive, and gradual (e.g., Gould, 1982: 138; Gould and Eldredge, 1977: 145).

¹⁰ After reflecting on Daniels's critique I am left with four general hypotheses to use in viewing the Italian and Spanish data or to use for longitudinal trial court studies in general. I will refer to these as the random, linear, curvilinear, and cyclical hypotheses. First, the random hypothesis, encompassing the belief that civil filing rates from one year to the next cannot be predicted (or explained), seems wrong. Trends—up, down, or flat—continue for several years. Moreover, the regions within Italy or within Spain follow roughly the same patterns from 1945 to the present.

Second, Daniels (and Munger) successfully refute the linear social development model as a general strategy for caseload studies. Nevertheless, certain linear model formulations, emphasizing a particular type of social change for a limited period of time, might have some explanatory power. Take, for instance, the situation where a category of cases (e.g., automobile accident cases) is examined in conjunction with social change—economic growth or depression—related to that category (see Stookey, 1986: 297–99).

Third, most curvilinear models adopt a social development hypothesis with the same limitations as the linear variation. José Juan Toharia proposed an explanation for the relationship between social change and civil litigation in Spain after examining caseloads in various regions from 1960 to 1967. Prior to rapid economic development in a region, holding all other factors constant, the civil litigation rate is stable. As a region industrializes and workers leave agriculture, the number of civil filings increases relative to the population. Finally, once the region modernizes to the point of urban Spain in 1967 (defined as 10 percent or less of the working population in agriculture), citizens use courts for civil matters less than before and filing rates begin to fall. We are left, then, with a curvilinear pattern reflecting the impact of economic development on civil caseloads (Toharia, 1974: 187–90). Toharia's original thesis, however, does not fit Spanish reality since 1968, as shown in Fig. 2.

Fourth, a cyclical hypothesis appears to offer the most promise for long-term diachronic studies. It would incorporate the advantages Daniels sees in investigating local environmental conditions and structural constraints, with peaks and valleys resulting from punctuated equilibria. Taking this perspective, one might view the plateau for Spanish regions from 1945 to 1967 or for Italian regions from 1952 to 1968 as static equilibria, adjustments to environmental conditions, that were disturbed at the end of the 1960s. The continuous upward slopes in both Italy and Spain during the 1970s could then be analyzed as dynamic equilibria, involving constant growth rates. Furthermore, cyclical models could draw on the extensive economics literature on business cycles (Stookey, 1986: 285–86) or on the attractive idea that civil caseloads are comprised of categories of cases—responding to social demand or government policies—that live out life cycles of birth, growth, and death (Friedman, 1967: 798–810).

The cyclical hypothesis does not depend on a capitalist economy, as civil caseload data for the Soviet Union reveal. Beginning in 1920, from a low of 300 civil claims per 100,000 inhabitants, the number of filings grew to 2,500 in 1928 as life normalized after the revolution and civil war. At that time the revival of comrades' courts, collectivization in the countryside, and nihilistic attitudes toward formal law caused by 1930 a sharp decrease in the litigation rate to 800 claims. During the 1930s civil filings again increased as regular courts regained some of their subject-matter jurisdiction, reaching 2,700 filings per 100,000 population in 1940. This collapsed to a rate of 90 claims during World War II in 1943, rebounding to a postwar high in 1952 of 2,100 claims. By 1956 this rate had fallen to 1,300 filings, stabilizing since 1957 to between 630 and 900 claims per 100,000 inhabitants (Van den Berg, 1985: 143–47).

¹¹ The statistics for Peruvian regions outside the capital consist of samples gathered from one province in each region. This may account for the smoothness in trend lines for the semi-urban, agrarian, and traditional regions.

¹² For an excellent study of the Costa Rican situation, see Gutiérrez (1979).

¹³ The proportion of labor cases may have been higher in earlier years. For example, 21 percent of the civil cases “decided with sentences” in 1964 concerned labor issues (Merryman et al., 1979: 173, 227).

¹⁴ Some labor cases, about ten per 100,000 population, were filed annually between 1945 and 1959 in Colombian civil courts where no labor court existed in the judicial district (Merryman et al., 1979: 215). In 1981 about 4 percent of the labor cases were filed in ordinary civil courts (Departamiento Administrativo Nacional de Estadística, 1983: 429).

¹⁵ Per 100,000 inhabitants, labor cases filed in Costa Rican first instance courts would add the following to Figure 7: 444 in 1945; 341 in 1950; 504 in 1960; 556 in 1970; and 667 in 1982 (Dirección General de Estadística y Censos, Anuario 1982, 1985: 94; Merryman et al., 1979: 218).

¹⁶ Reform in 1984 raised the jurisdictional minimum to 1,000,000 lire (about $750), which has increased the usefulness of these courts to persons with small claims. Preliminary 1985 statistics show 150 filings per 100,000 inhabitants. Concurrently the reform has reduced the caseloads of first instance civil courts with professional judges (Istituto Centrale di Statistica, Annuario Italiano, Edizione 1986, 1987: 203).

¹⁷ Voluntary jurisdiction cases concern a mixed bag of matters including declaration of death, authorization of wills, adoption, guardianship, sale of a minor's or incompetent's assets, recognition of bailments, rejection of letters of credit or checks, and certain maritime, partnership, and insurance disputes (Ramos, 1986: 2:1300–1370). It is likely that most voluntary cases are not actively contested by a defendant, since only 5 percent of these cases were resolved in favor of defendants in 1982 (Instituto Nacional de Estadística, Anuario, 1986, 1987: 410).

¹⁸ Jurisdictional amounts were revised in a major 1984 reform (Ramos, 1986: 1:34; see Ley no. 34, 6 August 1984). For the divisions used in 1982, see Miguel y Romero and Miguel y Alonso (1967: 61–65).