¹ The available materials amount to more than four hundred bound folio volumes with approximately two hundred thousand handwritten pages, most of which are barely legible. Because archival care of the documents has been less than perfect, their content had to be determined by individual inspection. The changing system of recording in the courts had to be reconstructed to avoid double counting, for in certain periods each hearing was reported in five documents. Because of the mass of materials, sampling techniques were required to establish the general trends. After frenetic annual fluctuations in the caseload were discovered, a full count of consecutive years was made for some periods (1569–94 and 1640–73). To further reduce this considerable source of error (cf. Daniels, 1984: 766), the quantity of paper used for the court records was measured to obtain an approximation of the caseload. The results, reported in Figure 1, shows the risk involved in the unavoidable sampling by single years, since it is never clear whether the probe hit a minimum or a maximum of the fluctuations.

² The rural areas had their own courts. In 1811, these areas were included in the district of the lower court, whereby its population rose by one-third (see Fig. 2). This may have increased judicial business by about 10 percent but no more, because we can safely assume that the litigation rate was essentially lower in rural areas than in urban centers. The earliest available statistics that allow a comparison between the city proper and rural districts with their own courts show that an additional 32 percent of the whole population produced only 12 percent more law suits (calculated from the sources for 1865 for Amt Vegesack and Amt Bremerhaven, two districts that adjoin Bremen).

³ Outbreaks of the plague seriously reduced the population; the worst took about four thousand lives. Such losses make the calculation of early modern litigation rates even more uncertain. Since they tend to underestimate the rate, however, they do not affect one principal finding of the study, namely, that before 1710 there were at least as many lawsuits per capita as today.

⁴ The upper court (Obergericht) consisted of twelve members of the Bremen city council, in which the city's political and financial aristocracy was represented. This court decided all cases involving claims over about $15,000 and also heard appeals against decisions of the lower court (Niedergericht). The latter was composed of one presiding member of the city council and two assessors, none of whom were required to be professional jurists. Small cases were heard by the two assessors sitting as single judges (Gastgericht). During the first 150 years of the time series, the archbishop still formally exercised his judicial powers through his governor (Stadtvogt), but there are strong reasons to believe that the services of this weak official were neglected by the citizens, who preferred the municipal courts for their speed and efficiency (see Hiemsch, 1964: 18–24). Court reforms in the nineteenth century separated the judiciary from administrative offices and required judges to have formal legal training. But the two basic levels of the court system remained unchanged. A commercial court (Handelsgericht), added in 1845, attracted new lawsuits from the merchants. By the end of the century, a labor court (Gewerbegericht) was established; this took a negligible part of the caseload away from the regular courts (3.6 percent in 1900).

⁵ For a full discussion and further evidence, see Wollschläger (1989).

⁶ The extraordinary range of the data is the main reason that considerable sources of error seem to be tolerable for the prestatistical period. Due to the different bases of estimations and the aforementioned changes in the legal categories of the cases, the margin of error may be as high as 30 percent for a single year. This seems to be acceptable, however, because research may still document the large-scale changes in the caseload that establish the long-term trend.

⁷ In these figures, except in absolute caseload data, the decimal digit is important, because it can mean the difference between “normal” or rapid growth, stagnation, or decline.

⁸ Since the physical qualities of paper and the habits of the secretaries seem to be invariant over short periods, this method should reveal short-term fluctuations.

⁹ This results from the semilogarithmic scale that, on the y-axis, transforms all values into powers of ten. On a linear scale, the exponential trend would have the shape of a ski jump; see the curve for U.S. divorce rates in Friedman and Percival (1976b: 64, Fig. 1).

¹⁰ This is the principal difference between Bremen and those rural regions in Europe where litigation declined during the age of industrialization (e.g., Scandinavia, France, Italy, and parts of Germany). In these areas economic growth also had an effect, but its dominant influence must have been to improve the payment of debts, thereby reducing defaults and other breaches of contracts. Judicial business decreased because there was less need to sue.

¹¹ These parallels suggest that economic factors were dominant internationally. While in Nazi Germany collectivist ideology and the persecution of Jewish lawyers may have added to the decline of private lawsuits after the Great Depression, political factors cannot explain the general pattern.