Vol. 14 No. 1 (January 2004)

AMICI CURIAE AND STRATEGIC BEHAVIOR IN STATE SUPREME COURTS, by Scott A. Comparato. Westport, CT: Praeger, 2003. 175 pp. Cloth $64.95. ISBN 0275978273.

Reviewed by Wendy L. Martinek, Department of Political Science, Binghamton University. E-Mail: martinek@binghamton.edu

Scholars of law and courts have become increasingly aware that state courts of last resort can provide the kind of leverage for extending our knowledge of judicial behavior that a focus on the singular Supreme Court simply cannot. The benefit that accrues from examining state supreme courts derives from the variation in institutional arrangements across state supreme courts as well as in the political context within which those courts operate. In AMICI CURIAE AND STRATEGIC BEHAVIOR IN STATE SUPREME COURTS, Scott Comparato, cognizant of the importance of such variation, undertakes an exploration of how institutional design and state political environments structure the behavior of litigants and amici in terms of the arguments they raise and the behavior of judges vis-ˆ-vis case outcomes. In doing so, Comparato offers an explicitly strategic account of the behavior of litigants, amici, and judges that links the goals of each to the choices they make.

Comparato begins, in the first two chapters, by identifying key goals of interest groups (group maintenance, securing desirable policy) and judges (retention in office, securing desirable policy). In pursuit of these goals, both interest groups and judges will be sensitive to important institutional arrangements-namely, the method of judicial selection and the presence of the ballot initiative, though Comparato's focus is on the former. The argument that institutional design matters is certainly not novel. Nor is the contention that method of judicial selection and retention is important to understanding judicial behavior. What is noteworthy, however, is Comparato's argument that both direct litigants and amici curiae take these institutional features and their implications into account when crafting their briefs (petitioner, respondent, and amicus curiae briefs).

Accordingly, litigants and amici (especially) will offer more arguments related to issues, concerns, and preferences of governors and legislatures in states where the supreme court is selected and retained in office by those actors, since institutionally-related arguments will be most useful to jurists in those states. On the other hand, "state supreme court justices will be most interested in obtaining information about the policy implications of their decisions on the public in those states where there is the most proximate electoral connection with the public" (p.9), with the closest connection being in partisan election states, followed by nonpartisan and merit elections states. Thus, litigants and amici will provide more policy-related arguments in such states, because of their relevancy to judges.

The second part of Comparato's argument is that judges use the information provided by litigants and amici differentially, depending upon the institutional context. This means that institutional arguments will be more beneficial to achieving success in legislative and gubernatorial appointment states, while policy arguments will be more beneficial where staffing is a function of popular elections.

Having articulated the tenets of his strategic theory, Comparato devotes Chapter Three to a thorough accounting of the sampling design and coding conventions employed. The data collection for this project was quite an undertaking and the resulting dataset is very impressive. Comparato collected a sample of cases with at least one amicus curiae brief filed on the merits from seven states (Alabama, Colorado, Kansas, Michigan, Wisconsin, New Jersey, South Carolina) over a ten-year period (1986-1995). The author is both detailed in reporting how he collected the data and refreshingly frank in acknowledging the limitations of the dataset. For example, Comparato was especially concerned with selecting states to maximize variation in method of selection and the presence of the ballot initiative. However, given the number of possible combinations, as well as the practical realities of data collection, the resulting set of states does not include every institutional arrangement. Comparato is unfailingly candid on such points, while still remaining convinced - and convincing - that the sampling design is suitable and appropriate.

Also of particular help are the numerous examples of language contained in the legal briefs Comparato provides in Chapter Three to illustrate the difference between policy-related and institutionally-related arguments. References to the competence of the legislature and/or governor in passing legislation in a given area or to the preferences of those actors are considered institutional arguments; while references to the policy consequences of a given decision are considered policy arguments. In addition to the illustrations of language differences throughout the text, Comparato includes a useful table summarizing the distinctions and offering additional examples.

Although neither the sampling design nor the coding scheme is unassailable (and what sampling design or coding scheme is?), Comparato's choices are carefully documented. The resulting data set is one that is sure to be of ongoing use to Comparato and other scholars interested in the nexus between state supreme courts and interest groups and, more generally, in the arguments made by participants in litigation.

Comparato subjects his central arguments to empirical verification in Chapters Four and Five, endeavoring to use his painstakingly constructed dataset to assess interest group arguments and judicial decisions, respectively, for evidence of strategic behavior in line with the account he articulated in the earlier chapters. In Chapter Four, Comparato, using a negative binomial statistical technique, constructs two sets of models. The first set models the institutionally-related arguments made by the petitioner, respondent, and amici, respectively, and the second set focuses on the policy-related arguments. The most important findings are the results related to method of judicial selection and the presence of the ballot initiative. With regard to institutional arguments, the ballot initiative matters not at all. Further, the method of judicial selection manifests no effect on the number of institutional arguments made by petitioners or respondents. However, while partisan elections are no more or less likely to result in more institutional arguments than legislative appointment, nonpartisan elections, merit elections, and gubernatorial appointment are associated with fewer institutional arguments than legislative appointment. The empirical results for the policy argument model are more consistent, with nonpartisan and merit elections both associated with more policy arguments made by petitioners, respondents, and amici in states with legislative appointment, and partisan elections, and gubernatorial appointment showing no effect that is statistically distinguishable from legislative appointment. Considering his empirical results collectively, Comparato concludes that institutional structures do matter (at least in some circumstances) and that amici do behave more strategically than other actors.

Chapter Five examines the influence of litigant and amicus briefs on case outcomes. To that end, Comparato constructs two probit models, using what appears to be the same dependent variable. The difference in the models is the set of independent variables, with the first using policy arguments and the second, institutional arguments. Problematically, it is difficult to tell what the dependent variable is in either case. The discussion and formalization of the hypotheses (pp.131-133) indicates that the dependent variable is whether the petitioner wins on the merits. However, the discussion of the dependent variable (p.134), as well as the variable explanation (Table 5.1), indicates that the dependent variable is whether there is a liberal outcome. A few pages later, in discussing the empirical results of the first model, the author's language indicates that the dependent variable is, again, the petitioner's success: "Moreover, there does not appear to be a statistically significant relationship between citizen ideology or government ideology and the likelihood of success by petitioner" (p.139). The same sort of language appears in the discussion of the second model (p.140). Even after multiple, very careful readings, it remains unclear to me which dependent variable has been modeled.

Certainly the first four variables included in both models would suggest that the dependent variable is whether there is a liberal outcome. These variables are the ideology of the court, citizenry, and government, respectively, plus whether the petitioner supports a liberal outcome. An expected relationship between the ideology of a court and the ideological direction of a case outcome requires little justification. Further, expected relationships between the ideology of the citizenry or government elites and a liberal outcome makes sense if we argue that judges are sensitive to their political context-though, as discussed further below, it makes more sense if we think of that sensitivity as conditioned by relevant institutional features. And, of course, given the propensity of appellate courts to select cases for error correction (as noted by the author), the fact that a petitioner is arguing for a liberal outcome is likely to enhance the prospects that a liberal outcome will occur.

On the other hand, it is difficult to see how the variables related to the policy and institutional arguments made in the two models, respectively, matter in and of themselves for the ideological direction of a case outcome. In the policy model, there are variables for four different sets of actors: petitioner, respondent, amicus in support of the petitioner, and amicus in support of the respondent. Comparato distinguishes among different subtypes in each case-e.g., amicus policy arguments in support of the petitioner are broken down by local group, national group, state/government, and business. Each is a simple count of the number of relevant arguments. The second model includes a similar set of variables, the difference being the count is of institutional arguments.  The problem, as I see it, however, is that there is no reason to expect the number of either type of arguments to be related systematically to ideological direction in case outcomes.

On a related point, regardless of whether the dependent variable is a liberal outcome or a petitioner win, inclusion of relevant institutional features (dichotomous variables representing method of selection and presence of the ballot initiative) as simple additive terms in each model does not match the strategic account Comparato offered earlier. For example, the author says, "I expect that policy arguments will have a greater impact in states with the ballot initiative and where state supreme court justices are directly accountable to voters" (p.136). This suggests, then, that the ballot initiative and method of judicial selection should more appropriately be modeled as interactive factors rather than as additive terms.

Despite these concerns, this volume does make an important contribution in presenting an informational theory of amicus behavior and judicial use of amicus briefs. Despite the large body of existing work on interest groups in the courts, our understanding of how and why amici matter in the decision making process remains shadowy. The informational theory Comparato articulates in this book offers to shed some light on those questions.

*********************************************************************
Copyright 2004 by the author, Wendy L. Martinek.