Judicial Politicization, Ideology, and Activism at the High Courts of the United States, Canada, and Australia

Article excerpt

Abstract

This article proposes a new cross-national thesis for judicial decision making. The judicial politicization theory posits that judges on highly politicized high courts will be more likely to decide cases using ideological and attitudinal factors than judges at less politicized courts. The theory holds that informal norms regarding judicial appointment by the executive are more important than the formal selection mechanism in determining whether a judiciary is highly or less politicized. The results show significant attitudinal judicial voting at each high court and strong support for the contention that judges on highly politicized courts are more likely to decide cases ideologically.

Keywords

law and courts, comparative judicial politics, high courts, United States, Canada, Australia

This article examines two related and highly salient issues: judicial decision making and judicial activism. Judicial decision making refers to the processes and factors that influence judges to decide cases as they do. In other words, why does a judge vote for one litigant instead of the other in a particular case? Does the judge base his or her decision on the legal principles governing the dispute, or is he or she affected by ideological, policy, or strategic considerations? For more than fifty years, public law scholars have attempted to unravel the mysteries of judicial decision making, and a variety of theories and models have been propounded and tested.

The other, related topic examined in this article is the concept of judicial activism. Judicial activism, as typically defined, occurs when a court strikes down or invalidates an act of the legislature or the executive. This simple definition, though, excludes the deeper issues and disagreements surrounding the concept. Critics of judicial activism assert that judges should seldom invalidate legislation or executive actions, while supporters hold that the ability of judges to overturn unconstitutional and unjust laws is a necessary check in a political system. Thus, judicial activism remains controversial, both in the trenches of academia and on the battlefields of politics. Nearly every election cycle in the United States finds conservative politicians decrying "activist judges" who "write laws from the bench," while defenders point out that it is these same judges who protect unpopular minorities from oppression through judicial rulings. In academe, normative commentators have described judicial activism as a form of "creative constitutional development" (Lewis 1999, 3) or as an "unfortunate phenomenon" (Wolfe 1997, x).

In recent decades, judicial activism has been recognized as not an exclusively American phenomenon, but, indeed, present in many democracies worldwide, to a greater or lesser degree (Holland 1991, 2000; Tushnet 2003). The influence of high courts throughout the world of modern democracies increased dramatically in the latter half of the twentieth century, as supreme courts expanded their role beyond dispute resolution to the creation of public policy (Tate and Vallinder 1995). But despite evidence that judicial activism has been increasing in many industrialized democracies, there have been few empirical, cross-national studies of the phenomenon, and few scholars have proposed theoretical frameworks to explain and predict judicial activism. Similarly, there has been little cross-national research into the dynamics of judicial decision making in comparative perspective. Much of the cross-national work that has been done has been primarily qualitative in nature (see, e.g., Guarnieri and Pederzoli 2002; Koopmans 2003; Tate and Vallinder 1995), most likely due to the difficulty in obtaining empirical data across states.

This article will attempt to remedy that deficiency, at least in part, by examining judicial decision making and judicial activism in the 1990s in the high courts of the United States, Canada, and Australia. Clearly, the process of judicial decision making influences the practice of judicial activism; thus these related phenomena are examined in tandem. However, this project will eschew analysis of the normative dimensions of judicial decision making and judicial activism, and instead approach these issues from an empirical and positive perspective.

This article proposes a new thesis for cross-national judicial decision making in industrialized democratic societies: the judicial politicization theory. Briefly, the theory posits that judges at a highly politicized high court in an established democracy are more likely to decide cases according to ideological/attitudinal factors, and will correspondingly be more likely to engage in judicial activism and strike down acts of the legislature. The theory posits, as a corollary, that judges at a high court that is less politicized will be more likely to decide cases based on legal factors, and will be less likely to invalidate laws enacted by the political branch. Stated another way, the full version of the theory proposes that the attitudinal model of judicial decision making will predominate in politicized judiciaries, while the legal model will be more likely to prevail in less politicized courts.

The modified or reduced version of the judicial politicization theory omits the corollary and focuses solely on ideological and attitudinal factors in judicial decision making. The results of the analyses in this study suggest that the modified version of the theory is best supported by the data.

The Judicial Politicization Theory

The conventional wisdom on judicial selection systems among American public law scholars is that the process used to choose the judges will significantly affect the judicial decision making by those judges (see, e.g., Brace and Hall 1997; Dubois 1980; Hall 2001; Hanssen 1999; Kritzer 2006; Sheldon and Lovrich 1991). For example, Gates and Johnson (1991, 159) state, "The systems themselves may influence the decision making of judges and their relationship to the political system." Much research has been conducted in the American states to determine whether appointive systems (wherein the executive chooses the judge, sometimes after a commission has produced a short list of names) or elective systems (where the legislature or voters directly choose the judges) have the greatest impact on judicial decision making (see, generally, Baum 1998; Bonneau and Hall 2009; Hanssen 2004). In other words, the scholarly consensus suggests that formal selection mechanisms are a predominant influence on judicial behavior.

However, the theory proposed in this article challenges the conventional wisdom and asserts that in cross-national context, informal norms are the dominant factor.1 While there appears to be no question that judicial appointment systems are a highly significant influence on American state court judges, this article argues that in cross-national perspective, formal selection mechanisms are less important than the "selection culture" inherent in a modern democratic political system. The selection culture refers to whether the appointing executive typically relies on ideological and partisan factors to choose judges or whether other factors, such as qualifications and merit, are the most important criteria. Stated another way, a country's judiciary is highly politicized if the judges are chosen by the executive based on partisan grounds, while another nation's judiciary is less politicized if the magistrates are selected on nonpartisan factors. The logical extension of this observation is that judges in a highly politicized judiciary will tend to engage in a greater degree of ideological decision making, and judges in a less politicized judiciary will be less likely to decide cases on ideological or attitudinal grounds. The causal mechanism is straightforward: A politicized judicial selection culture results in a greater tendency for judges to be chosen based on partisan and ideological grounds; after appointment, these magistrates will have a greater likelihood of deciding cases in accordance with their personal ideological preferences and thus tending to overturn oppositional legislation. Therefore, the judicial selection culture of a country, not the formal appointment mechanism, is the critical factor in determining whether its supreme court is highly politicized, and whether its judges will tend to be driven by political and ideological concerns.

Thus, the central premise of the judicial politicization theory is the contention that judges on a highly politicized high court in an established democracy are more likely to decide cases according to ideological/attitudinal factors, and will correspondingly be more likely to engage in judicial activism and strike down acts of the legislature when the challenged law is oppositional to the judge's ideology. The corollary to the theory is that judges on a high court that is less politicized will be more likely to decide cases based on legal factors, and will be less likely to invalidate laws enacted by the political branch. In other words, the attitudinal model of judicial decision making will predominate in politicized judiciaries, while the legal model will be more likely to prevail in less politicized courts.2 It is important to note that judicial activism is not tied to a particular ideological orientation; that is, both liberal and conservative judges may be likely to overturn oppositional legislation.3

To be sure, there is no question that under certain circumstances, the formal judicial selection process could moderate the degree of judicial politicization. For example, transparency (through open hearings) could moderate, to a greater or lesser degree, the influence of ideology in judicial selection. Thus, an open system, with multiple actors/veto points and open hearings, could lessen (though probably not eliminate) the tendency of the executive to select high court judges based on ideological congruence. The United States has a much more open process than Australia or Canada, as U.S. Supreme Court nominees receive intense scrutiny by interest groups and the oppositional political party, followed by confirmation hearings in the Senate Judiciary Committee and a full floor vote in the Senate. By contrast, high court judicial selection in Canada and Australia has involved little or no scrutiny by interest groups and virtually no transparency. This example serves to illustrate the basic contention of the judicial politicization theory. There are virtually no checks on the executive in high court judicial selection in Canada and Australia, yet the process there is consistently less politicized than in the United States. However, if the selection culture in Canada or Australia should change, then the prime minister would be able to make highly ideological judicial appointments without difficulty.

There are several hypotheses that can be derived from the full judicial politicization theory; these are tested below and discussed more fully in the Results section.

Hypothesis 1: Attitudinal voting in judicial review cases is likely to be more significant in politicized judicial systems. Thus, high court judges in the United States are more likely to vote according to attitudinal preferences in judicial review cases than are Canadian or Australian judges.

Hypothesis 2: In less politicized high courts, the presence of a lower court dissent will increase the likelihood of a vote for the invalidation of a law. In highly politicized high courts, the presence of a lower court dissent will not result in a greater tendency to vote for the nullification of a statute.

Hypothesis 3: In less politicized high courts, the fact that the lower court struck down the challenged law will increase the likelihood of a vote for the invalidation of the law. In highly politicized high courts, the fact that the lower court invalidated the challenged law will not result in a greater tendency to vote to strike the law.

Quantifying Judicial Politicization

It is difficult to ascertain whether a country's judiciary is politicized or not, because there has been no empirical or quantitative method yet proposed to measure this concept. One method to determine the degree of judicial politicization is to conduct a review of the existing qualitative literature and the thick descriptions that often exist in those studies. This is of course a crucial introductory step, but the lack of an empirical measure in these studies prohibits a researcher from making definitive conclusions about the level of judicial politicization in a particular apex court. Therefore, this project proposes and conducts an empirical test for the degree of judicial politicization in a high court.

The judicial politicization indicator is operationalized by analyzing partisan and nonpartisan appointments to the high court and simply calculating the percentage of nonpartisan judicial assignments. A nonpartisan appointment is defined as a judicial appointment in which the political ideology of the judge differs from that of the executive. To illustrate, imagine that former President George W. Bush had nominated the legal director of the American Civil Liberties Union (ACLU) to serve on the U.S. Supreme Court. This admittedly extreme example would provide an illustration of a nonpartisan appointment to the U.S. high court. Clearly, the highly conservative Bush would not have nominated the ACLU legal director to the court based on ideological congruence; rather, the appointment would have presumably been made on the basis of other factors, such as experience, qualifications, merit, or some other nonideological basis.

To quantify nonpartisan appointments, I compared the ideology of the judge with the ideology of the appointing executive. If the two ideological values did not match, this was coded as a nonpartisan appointment. For the judicial ideology scores of the American Supreme Court justices, I used the well-known Segal and Cover (1989) scores. These values were derived through a content analysis of editorials regarding judicial appointments in leading American newspapers. The Segal and Cover scores, updated by Epstein and Segal (2005), exist on a scale from most conservative (scores closer to 0) to most liberal (scores closer to 1).

For the Supreme Court of Canada ideology scores, another set of researchers (Songer and Johnson 2002; Ostberg and Wetstein 2007) conducted a content analysis (consistent with Segal and Cover 1989) of Canadian newspapers to derive individual ideology scores for members of the Canadian Supreme Court.4 The Canadian ideology values were transformed to match the Segal and Cover scale, which is a scale from 0 to 1.

For the High Court of Australia, I conducted an original content analysis of Australian newspaper editorials using the Segal and Cover (1989) procedures and developed an ideology scale for each judge sitting on the High Court in the 1990s. These scores are the first ideology values to date for the members of the High Court of Australia using the same scale as the Segal and Cover values. Thus, the values for the American, Canadian, and Australian judges were all coded using the same procedures and can be viewed comparatively. The initial ideology values for the judges on each high court are shown in Table 1. In addition, an extensive discussion of the methodological issues regarding the coding process is found in the appendix, which is located in the Supplemental Materials section for this article at the Political Research Quarterly Web site, http://prq.sagepub.com/supplemental/.

The next step was to compare judicial ideology with executive ideology. I inferred executive ideology from party identification. Thus, if the executive was a member of the Republican Party in the United States, a member of the Liberal Party in Australia, or a member of the Progressive Conservatives in Canada, he or she was coded as having a conservative ideology (value of 0 to .499). Conversely, if the executive was a member of the Democratic Party in the United States, Australian Labor Party in Australia, or Liberal Party in Canada, he or she was coded as having a liberal ideology (value of .501 to 1.00).5 So, if the ideology of the executive and the ideology of the judge did not correspond, this was coded as a nonpartisan appointment. The mean for nonpartisan appointments was calculated, and then subtracted from 1 to provide an intuitive measure of judicial politicization. It should be noted that any judge receiving an ideology score of .500 was classified as a neutral appointment and was not included in the judicial politicization calculation, because that value is the exact midpoint of the scale that runs from 0.000 to 1.000.6

Note that both associate justice appointments (associate justices are called puisne justices in Australia and Canada) and chief justice appointments or promotions are included. However, if only initial appointments to the high court (thus excluding chief justice promotions) are used in the calculations, the overall judicial politicization scores do not vary significantly.7

The details for the U.S. Supreme Court are found in Table A1, Australian results are shown in Table A2, and Canadian high court data are shown in Table A3. The judicial politicization index for each court is graphically displayed in Figure A1. Observe that the higher the value for each high court shown in Figure A1, the higher the degree of judicial politicization. Tables A1 through A3 and Figure A1 are located in the online Supplemental Materials section for this article at http://prq.sagepub .com/supplemental/.

Figure A1 shows that the United States has the most politicized judicial system with an index score of .917, followed by Australia with a score of .8 and Canada with an index score of .5. These scores comport with the scholarly consensus regarding the supreme courts of the United States, Canada, and Australia. Researchers have found that the judiciary in the United States has become exceptionally politicized in the past twenty-five years, while the apex court of Canada has remained relatively nonpoliticized, as compared with the U.S. high court.

So, judicial selection of Supreme Court judges in the United States is most likely to be based on ideological congruence, while the selection of high court judges in (especially) Canada and Australia is less likely to be based on this factor. In other words, in the United States, there is a very strong correlation between the ideological preferences of the nominating president and the ideology of the judge, as measured by Segal and Cover (1989) judicial ideology scores, whereas high court judges selected in Canada and Australia and serving in the 1990s sometimes did not share the ideology of the appointing executive. Thus, it appears that these Canadian and Australian judges were selected less on ideology and more on other factors, such as qualifications and experience.

However, it must be noted that the time frame for this study, 1990-99, inclusive, is relatively short in the broad scope of judicial behavior in the twentieth and twenty-first centuries. Indeed, the decade of the 1990s was somewhat anomalous in Canadian politics because the Progressive Conservative Party was in power from 1984 to 1993, while the Liberal Party dominated federal politics for most of the latter half of the twentieth century. Thus, the majority of the high court judges on the Canadian Supreme Court in the 1990s were appointed by prime ministers belonging to the Progressive Conservative Party. Therefore, the quantification of judicial politicization is necessarily time bound to some degree. An expansion of these analyses using a much longer time period would shed greater light on the degree of judicial politicization in each of these courts.

The analysis and quantification of judicial politicization is just the first part of the theory. The thesis also holds that judges at a highly politicized high court will tend to decide cases based on attitudinal/ideological factors, while the corollary posits that legal factors should be more influential for judges at less politicized courts.

Testing the Theory

In this section, the research design and operationalization of the variables are discussed. Judicial review cases were used in the analyses because of the inherent significance of these cases: the decision to overturn an act of the political branch is a very important one, and one that can be assumed to be taken very seriously by the judges. Also, the database that was created to conduct the analyses combined nonunanimous and unanimous court cases, so as to provide the most rigorous test of the theory.8 More information on case selection and coding is provided in the methodological appendix, located in the Supplemental Materials section for this article at http://prq.sagepub.com/supplemental/.

Overview of the Model

Here, the model incorporating attitudinal, legal, and control variables for voting in judicial activism cases in the 1990s in the high courts of the United States, Canada, and Australia is presented. As discussed in the appendix, a generalized estimating equation (GEE) logistic regression is used to estimate the model because the dependent variable here-a judge's vote to strike down a law-is dichotomous. The judges' ideology is measured by using ideological scores derived from content analysis of newspaper editorials (again, existing values were used in the American and Canadian cases; newly derived scores for Australian judges are presented here for the first time). These individual judicial ideology scores are then adjusted (following Lindquist and Klein 2006) for statute direction so that higher positive values indicate agreement with the law, while higher negative values indicate that there is ideological inconsistency with the challenged statute.9 Therefore, it is expected that this variable will have a negative coefficient, because the greater the degree of ideological disagreement with a statute, the greater the likelihood of a vote to strike. The size of the marginal effect for this coefficient will indicate how strongly judicial ideology influences judicial decision making.

The legal variables seek to measure, first, whether individual judges are influenced by the jurisprudential factors of a lower court dissent and, second, the nullification of the challenged law by the immediate lower court. Thus, the legal variables seek to measure nonideological factors that may influence a judge when considering the constitutionality of a statute. The first legal variable is the presence of dissenting opinions in the lower court, which should signal to the judges that there are multiple interpretations of the constitutionality of a statute or, alternatively, that there may be a problem with the majority opinion's legal reasoning. Therefore, in nonpoliticized high courts, where legal factors are more influential, it would be expected that the presence of a lower court dissent would increase the likelihood of a vote for judicial activism.

The second legal variable deals with a lower court invalidation of the challenged law. Here, it is expected that in less politicized supreme courts, the fact of a lower court nullification should increase the likelihood of a judge to vote to strike down a law because of the tendency to agree with the lower court's legal reasoning supporting the invalidation of the challenged statute. In other words, the fact that a lower court has reviewed the constitutionality of a law and found a basis for nullification should be persuasive for those judges who are influenced by jurisprudential considerations.

Finally, there are a number of control variables in the model, including the presence of a state/province or local law and whether the solicitor general (in the United States and Australia) or attorney general (in Canada) supports striking or retaining the law. Another control variable is the saliency of the individual court case to the judges, measured by the total number of interveners/amici curiae multiplied by the total number of nongovernmental interveners. The fourth control variable measures whether the federal government is a direct party in the case. The fifth control variable is the amount of disparity in resources between the litigants (the coding for this variable is discussed in the methodological appendix). The final control variable is the issue of the case, operationalized as a series of dummy variables.

Results

Here, the estimates for each high court analysis are presented, and the central premise of the judicial politicization theory-that judicial attitudes tend to strongly influence judicial decision making in highly politicized courts-is assessed. Stated another way, the primary prediction of the judicial politicization thesis is that the more highly politicized the high court, the more likely the judges will decide judicial review cases on the basis of their attitudes and ideological preferences. The corollary to the theory- that legal factors are more likely to be influential in less politicized courts-is also examined. Again, hypothesis 1 states,

Hypothesis 1: Attitudinal voting in judicial review cases is likely to be more significant in politicized judicial systems. Thus, high court judges in the United States are more likely to vote according to attitudinal preferences in judicial review cases than are Canadian or Australian judges.

As shown in Figure A1, the United States is by far the most highly politicized high court, while Australia is the second most politicized, and Canada is the least politicized of all three courts. So, if the theory is correct, we would expect to see that the independent variable judicial ideology has negative coefficients that are significant for each of the three high courts, and that the effect size is largest for the U.S. Supreme Court, followed by smaller effect sizes for the Australian and Canadian high courts.10

Examining Tables 2 and 3 reveals just that.11 For U.S. Supreme Court judges, judicial ideology is significant at the 99 percent level. For Australian high court judges, ideology is significant at the 95 percent level, and ideology is significant at the 90 percent level for Canadian justices. Table 3 provides estimates of the marginal effects of the independent variables on the dependent variable (vote to strike) for each court. The data in Table 3 show that American high court judges are more influenced by ideology than either Canadian or Australian justices, just as the theory predicted. The marginal effect of judicial attitudes for American judges is -.31 (standard error of .07) or -31 percent. For Australian justices, the marginal effect is -.18 (SE = .09) or -18 percent. For judges at the Supreme Court of Canada, the marginal effect size is -.13 (SE = .07) or -13 percent. These results show that in judicial review cases in the 1990s, judicial ideology is more than twice as influential for American than for Canadian judges. Australian high court judges are situated between the United States and Canada: judicial attitudes account for an estimated -18 percent effect size. Thus, because of the significance of the attitudinal variables in each case as well as the effect sizes, hypothesis 1 can be tentatively confirmed: at least for these three tribunals in the 1990s, attitudinal voting is more likely to be observed in judicial review cases in highly politicized high courts. To be sure, further research replicating this analysis in other high courts and different time periods would provide greater support for this conclusion.

Influence of the Law

Having found evidence that attitudinal voting is more likely to be influential in highly politicized judiciaries, the influence of legal variables in these courts can be assessed. Again, the corollary to the theory posits that legal factors are more likely to influence judicial decision making in less politicized high courts. The first legal variable is the presence of a lower court dissenting opinion, which can signal to the high court judges that there are multiple interpretations of the constitutionality of a statute or that there may be a problem with the majority opinion's legal reasoning. So, it is expected that in less politicized courts, such as in Canada and Australia, the presence of a lower court dissent would increase the likelihood of a vote for judicial activism and a positive coefficient will be observed. Hypothesis 2 states more formally,

Hypothesis 2: In less politicized high courts, the presence of a lower court dissent will increase the likelihood of a vote for the invalidation of a law. In highly politicized high courts, the presence of a lower court dissent will not result in a greater tendency to vote for the nullification of a statute.

On the basis of the data from Tables 2 and 3, hypothesis 2 cannot be confirmed. As predicted, a lower court dissent was not significant in the American case, but was highly significant in both Australia and Canada. The data from Table 3 indicate that in Australia, the presence of a lower court dissent accounts for a .34 (SE = .10) or 34 percent marginal effect change in the dependent variable. This is a quite large effect size, and suggests that this legal factor is very influential for judges at the High Court of Australia, as predicted. However, at the Canadian Supreme Court, lower court dissent is significant, but the coefficient is in the opposite direction than predicted and the effect size is small: -.04 (SE = .01) or -4 percent. In other words, the presence of a lower court dissent in Canada slightly decreases the likelihood of a vote to nullify a statute, contrary to the expectations of the full judicial politicization thesis.

The second legal variable is whether the lower court struck down the law in question. Again, it is expected by the corollary to the theory that in less politicized high courts, a lower court invalidation should increase the tendency to vote to strike down a challenged law, because the fact that a lower court has reviewed the constitutionality of a statute and found a basis for nullification should be persuasive for those judges who are influenced by jurisprudential considerations. Stated more formally,

Hypothesis 3: In less politicized high courts, the fact that the lower court struck down the challenged law will increase the likelihood of a vote for the invalidation of the law. In highly politicized high courts, the fact that the lower court invalidated the challenged law will not result in a greater tendency to vote to strike the law.

On the basis of the data in Tables 2 and 3, hypothesis 3 cannot be confirmed. As predicted, this variable is not significant for the U.S. high court but is highly significant for the Canadian and Australia supreme courts. In Canada, the presence of a lower court invalidation of the challenged statute accounts for an estimated marginal effect size of .03 (SE = .01) or 3 percent increase in the dependent variable (vote to overturn the law). This is a fairly small effect but is in the predicted direction and thus consistent with the full judicial politicization thesis. However, in Australia, the coefficient is in the opposite direction than predicted, and the estimated effect size is large. At the High Court of Australia, the presence of a lower court nullification accounts for a -.27 (SE = .06) or -27 percent change in probability in the likelihood of a vote to strike at the High Court of Australia. In other words, the fact that the lower court overturned the challenged law decreased, not increased, the likelihood of a vote to strike the challenged law by 27 percent by the judges at the High Court of Australia, contrary to the expectations of the theory.

Overall, the two legal variables provide mixed support for the proposition that less politicized high courts will have a greater likelihood to be influenced by legal factors in judicial decision making. As predicted by the corollary, neither legal variable was significant for the U.S. Supreme Court. Lower court dissent was highly significant in Australia, and lower court statutory invalidation was significant in Canada. But the presence of a lower court dissent had the opposite effect in Canada than predicted (albeit with a small marginal effect size), while the fact that the lower court struck down the challenged law had the opposite effect than predicted at the Australian high court (and a large effect size). Therefore, hypotheses 2 and 3 cannot be fully confirmed. These data suggest that consistent with the full theory, legal factors may have some influence in judicial decision making in less politicized high courts, but this influence is complex, and may be court bound and not generalizable.

The Modified Judicial Politicization Theory

As noted, the results of the analyses indicate that hypotheses 2 and 3 cannot be fully substantiated, casting doubt on the validity of the full judicial politicization thesis. It is possible that additional research that operationalizes legal influence in an alternative manner may supply different results consonant with the corollary to the theory. Nonetheless, the results of this study suggest that the corollary to the theory cannot be confirmed and so should be rejected. Thus, given the results of the data analyses herein, the modified judicial politicization theory states that, judges at a highly politicized high court in an established democracy are more likely to decide cases according to ideological/attitudinal factors, and will correspondingly be more likely to engage in judicial activism and strike down acts of the legislature.

When the judicial politicization thesis is modified in this way, eliminating the corollary and considering hypothesis 1 exclusively, then the theory can be tentatively confirmed, at least for these cases. Again, the data indicated that attitudinal voting was significantly more likely to be observed in judicial review cases in highly politicized high courts in the 1990s. To be sure, the influence of the law should not be discounted: the models presented herein suggest that legal factors are important within a high court, but may not be generalizable across national courts. Stated another way, the law and its effects on judicial decision making appear to be highly idiosyncratic and court bound. At least for now, a universal theory incorporating the legal model and demonstrating consistent cross-national results must wait.

Control Variables

The model also includes six control variables, some of which yielded interesting results. First, the variable measuring case salience was not significant for the Canadian or Australian high courts, but was significant at the 95 percent level for the American court. However, the marginal effect for this variable was .00 for the U.S. Supreme Court; thus, the effect of case salience on the decision to strike a law appears to be very low or nonexistent. To review, case salience was operationalized by counting the total number of interveners/amici in each case and multiplying that number by the total number of nongovernmental interveners/amici. So, this variable sought to measure cases containing legal and political issues that would attract both governmental interveners (such as a state/province attorney general) and also private groups. The assumption is that individual cases attracting a significant amount of governmental and nongovernmental interveners/amici (and corresponding legal briefs or appearances from these parties) will signal the importance of the case to the justices. The results showed that highly salient cases were not influential in the decision to vote to strike down a law in any of the high courts. This result comports with previous studies (see, e.g., Hagle and Spaeth 1991) of the U.S. Supreme Court that found that judicial fluidity is less likely to occur in highly salient cases, because a judge's decision in these disputes is more likely to be predetermined and unlikely to change on the basis of information found in amicus briefs. However, these results are some of the first to extend this finding to the high courts of Canada and Australia.

The next control variable measured whether the solicitor general/attorney general supported invalidation of the challenged law. There is a wealth of research indicating that the U.S. Supreme Court is strongly influenced by the position that the solicitor general takes in a particular case (see, e.g., Johnson 2003; but see McGuire 1998). It would be expected in all three countries that when the solicitor general (or attorney general in Canada) takes a position that a law should be overturned, the judges on the high court would be more likely to vote to strike that statute. This expectation was confirmed at the U.S. and Australian high courts, but not at the Canadian Supreme Court. In Australia, the fact that the solicitor general supports legal nullification accounts for a marginal effect size of .12 (SE = .04) or 12 percent increase in the probability of a vote to strike a law. At the U.S. Supreme Court, the marginal effect for the influence of the solicitor general was only .04 (SE = .02) or 4 percent, much lower than in Australia. In Canada, the influence of the attorney general was not significant.

Thus, it appears from these data that justices at the Supreme Court of the United States and the High Court of Australia are moderately influenced by the positions taken by the solicitor general, but that Canadian high court judges are not influenced by their attorney general in judicial review cases. Again, these results for the U.S. high court are consistent with prior research, but this is a new finding for the Australian court. The results underscore some of the similarities between the American and Australian apex courts. The fact that the judges on the Canadian Supreme Court were not influenced by the attorney general is consistent with Roach (2000; see also Edwards 1987), who notes that the modern attorney general is compelled to "articulate their own independent vision of the public interest," even though that vision may be in conflict with the courts (Roach 2000, 39). In other words, the Canadian attorney general occupies a different legal and jurisprudential space in Canadian jurisprudence than the American and Australian solicitors general. Further empirical research may shed greater light on the role of the Canadian attorney general in the Charter era.

Another control variable is whether the national government is a direct party in the case. Across all systems, it would be expected that a high court, acting strategically, would be less likely to strike down a statute when the government is a party. This variable is highly significant for the Australian court, moderately significant for the U.S. high court, and significant at the 90 percent level for the Canadian judges. However, the coefficients are in opposite directions for the American and Australian cases. At the High Court of Australia, the presence of a governmental party decreases the likelihood of a vote to strike by a marginal effect of -.11 or -11 percent (SE = .03). But at the U.S. Supreme Court, the presence of a governmental party increases the tendency for a vote to strike by 11% (.11; SE = .04). At the Canadian Supreme Court, the presence of a governmental litigant also increases the likelihood of a vote to strike by .05 (SE = .03) or 5 percent. The results for the Australian high court are in accordance with expectations, but the data for the American and Canadian courts may indicate that these judges are less likely to defer to the governmental party.

A related control variable is party resource disparity. This variable measures the status of each litigant (such as individual in a civil case, business, state or provincial government, etc.) and assigns a value to that status. Higher values indicate parties with more resources at their disposal. Then, the lower value is subtracted from the higher value to indicate the disparity (if one exists) in resources between the parties. So, higher numbers indicate larger litigant resource disparities.

The party resource disparity variable is not significant for the American court, but is highly significant for the high courts of Canada and Australia, although the coefficients are in different directions. In Canada, the greater the disparity between the parties, the greater the likelihood of a vote to strike the law, with an estimated marginal effect of .06 or 6 percent (SE = .01). In Australia, litigant resource disparity has the opposite effect: a greater disparity in litigant resources tends to increase the probability of a vote to strike by an estimated -.05 or -5 percent (SE = .01). Clearly, there is a different dynamic in the high courts of Canada and Australia regarding party resource differential. It is tempting to invoke Galanter's (1974) thesis demonstrating the likelihood of those parties with superior resources to prevail in litigation; however, the Galanter findings are inapplicable in these analyses. This is because the model used herein does not specify whether the party with superior resources sought to overturn or sustain the challenged law. Still, the findings for this control variable are intriguing; further research may clarify whether "repeat players" have greater success in judicial review cases in Canada and Australia.

The presence of a state/province or local law was also included in each model as a control variable. The data reveal that the presence of a state/province or local law is highly significant in the United States and Australia, and moderately significant in Canada. However, the data show that the direction for this variable differs for the high courts of Australia and the United States. At the U.S. Supreme Court, the presence of a state or local law increases the likelihood of a vote to strike by an estimated .19 (SE = .06) or 19 percent. But at the High Court of Australia, the presence of a state or local law decreases the likelihood of a vote in favor of invalidation by an estimated -.30 (SE = .05) or -30 percent. Canada sits within these extremes, and a challenge to a province or local law increases the marginal effect of a likelihood of a vote to strike by .06 (SE = .02) or 6 percent. These data indicate that when controlling for other variables, judges at the U.S. high court are far more likely to nullify state laws in judicial review cases than are justices at the High Court of Australia. The results for the U.S. high court are consistent with prior research, which has indicated that the court is far more likely to defer to the Congress than to state legislatures in certain circumstances (see, e.g., Lindquist and Solberg 2007; Solberg and Lindquist 2006). However, the data for the Australian High Court reveal the opposite: that the justices of that tribunal may be willing to give a substantial amount of deference to state legislatures in accordance with principles of federalism. This is a very preliminary, but potentially significant, finding. Further analysis and replication in other time periods may confirm these results.

The final control variable is the substantive issue of the individual court case. This variable was operationalized as a series of dummy variables so that the influence, if any, of a particular case issue could be assessed comparatively. 12 There was only one issue area which was significant for each high court and in the same direction: civil rights. A challenge to a civil rights law tended to increase the estimated likelihood of a vote to strike in each court: 13 percent (SE = .04) in Canada, 13 percent in the United States (SE = .03), and 27 percent (SE = .07) in Australia. This suggests that laws restricting civil rights are particularly salient to judges in each system, and are thus more likely to be overturned.

Similarities between the supreme courts of the United States and Australia are also present for federalism and speech/religion cases. Cases involving federalism issues were highly significant for the high courts of the United States and Australia, and tended to increase the likelihood of a vote to strike by an estimated 21 percent (SE = .06) and 25 percent (SE = .07), respectively. Speech and religion statutes were highly significant at the U.S. Supreme Court, but only moderately significant at the High Court of Australia. A speech or religion law tended to increase the tendency to invalidate by an estimated marginal effect of 25 percent (SE = .03) at the U.S. apex court and by 18 percent (SE = .08) at the Australian high court.

Economic laws were significant for the Canadian and Australian high court judges, although in the opposite directions. The estimated marginal effect of an economic law was -3 percent (SE = .01) for the Canadian court and 11 percent (SE = .05) for the Australian high court. Due process statutes were significant only at the Canadian Supreme Court. Laws restricting due process rights were highly significant in Canada, and decreased the likelihood of a vote to strike by 32 percent (SE = .05). This result is surprising, given that the Canadian high court is dominated by liberal justices. However, this result is consistent with the criminal law issue variable, which was also highly significant in Canada, and resulted in an estimated 19 percent decrease in the likelihood of a vote to overturn the challenged law (SE = .02).

Taken together, the due process and criminal law issue variables suggest that the Supreme Court of Canada, although dominated by liberal judges, was highly deferential to legislators in the 1990s when dealing with public order laws. Conversely, the high courts of the United States and Australia, both dominated by conservative judges in that decade, tended to be particularly concerned by laws involving speech, religion, or federalism. For the American case, this analysis supports the work of Keck (2004) regarding judicial activism and the Rehnquist Court.

The analyses regarding the substantive case issue variables resist broad synthesis, due to the conflicting direction of the results. However, several broad generalizations may emerge and help to explain these conflicting results. First, as previously noted, it appears that laws restricting civil rights are very salient for judges in each of the three high courts, and are likely to invoke a vote to overturn. One potential hypothesis stemming from these results is that high court judges in industrialized democracies may view their role as that of guardian of civil rights; additional research in other apex courts may lend further support to this proposition.

Regarding some of the differing results between the supreme courts of Canada and Australia, another potential explanation may be the differences between the Australian Constitution and the Canadian Charter of Rights and Freedoms. Specifically, the Australian Constitution does not contain a bill of rights or other explicit guarantee of freedoms (except for freedom of religion), but does contain a right to property and commerce. By contrast, the Charter of Rights and Freedoms does contain extensive protections of civil rights and liberties but deliberately excludes rights to trade and property. These differences may very well help to explain why the judges at the High Court of Australia are more likely to strike down laws restricting economic activities than the Canadian justices but are also less likely to strike down state/province laws than their Canadian counterparts.13 Again, additional research in different time periods may shed light on these findings.

Conclusion

In addition to the judicial politicization theory and related analyses, there are a number of potentially significant results from this project. One of the strongest findings from this study is that attitudinal voting by supreme court judges was found to be present in judicial review cases in the 1990s at the high courts of the United States, Australia, and Canada. From the results of these analyses, it appears that, at a minimum, the attitudinal model of judicial decision making can apply beyond the United States to the high courts of Australia and Canada. Furthermore, although the presence of attitudinal voting has been well documented at the U.S. Supreme Court, many scholars have found that attitudinal voting is most likely to be observed in certain issue areas, such as civil liberties cases (Epstein and Mershon 1996) and in nonunanimous cases only. So, the strong presence of ideological voting at the U.S. Supreme Court in a variety of issue areas and in combined unanimous and nonunanimous cases may be an addition to the literature regarding the American high court.

The finding of attitudinal voting by judges at the High Court of Australia is one of the first results of this kind. There has been only one recent attempt to measure attitudinal voting at the High Court (Smyth 2005; see also Schubert 1968, 1969a, 1969b),14 and that study used political party as a proxy for judicial preferences, rather than developing a set of judicial attitudes in the manner of the Segal and Cover (1989) measurement standards. The finding of ideological voting by Australian High Court judges found herein runs counter to the assertions of numerous Australian legal commentators (see, e.g., Heydon 2003). Clearly, further research on the High Court can shed more light on the question of whether Australian apex court judges are indeed substantially influenced by ideology in decision making, or whether these results are time bound and limited to the 1990s.

This study is also one of the first to test attitudinal voting in exclusively judicial review cases at the Supreme Court of Canada. There have been several previous studies that found substantial attitudinal voting by judges in specific types of cases at the Canadian high court: Ostberg and Wetstein (2004b) discovered that there is a strong tendency toward ideological voting in equality cases, Ostberg and Wetstein (2004a, 2007) found strong attitudinal voting in criminal and economic cases, and Songer and Johnson (2002) and Ostberg, Wetstein, and Ducat (2002) found a significant likelihood for judges to vote their attitudes in nonunanimous cases. In contrast, this study finds a significantly weaker influence for attitudinal voting at the Canadian Supreme Court in the 1990s when all cases- with varying issue areas and unanimous and nonunanimous results-are examined. Thus, the results from this study differ somewhat from the conventional wisdom on attitudinal voting at the Canadian high court. On the basis of the results herein, the judges at the Canadian Supreme Court do not appear to be as driven by ideology as the high court judges from Australia or the United States, at least in judicial review cases in the 1990s. Again, further research that expands the universe of cases analyzed at the Supreme Court of Canada can provide further confirmation of this finding.

Finally, this project proposed and tested a new crossnational theory of judicial decision making, the judicial politicization thesis. The comparative analyses of attitudinal voting for judges at the supreme courts in this study reveal that ideology plays the largest role at the U.S. high court, with an estimated marginal effect size of -31 percent; followed by the Australian supreme court, with an effect size of -18 percent; and then the Canadian high court, with an estimated effect size of -13 percent. Taken by themselves, the Canadian and Australian results are surprising, because most comparative courts researchers would likely have predicted that the Supreme Court of Canada would be most similar to the American high court. However, these results support the central premise of the judicial politicization thesis, which posits that the greater the politicization of the court, the more likely those judges will tend to vote ideologically. Thus, the core element of the judicial politicization theory is supported by the data from the high courts of the United States, Canada, and Australia in the 1990s. Replication of these results in additional high courts in established democracies would provide greater support for the generalizability of the theory.

The corollary to the judicial politicization theory-that legal factors will tend to be more influential for judges in less politicized courts-could not be confirmed, and thus the modified version of the theory omits it. It may be that the model is underspecified and requires further factbased legal variables to be developed and operationalized. Hopefully, the research in this project regarding legal factors and comparative judicial decision making has indicated a direction for other scholars to explore.

In summary, this analysis of both the attitudinal and legal models in cross-national perspective has supplied intriguing results that both confirm and challenge the conventional wisdom regarding judicial decision making. Additionally, a new theoretical framework has been proposed to explain the variation in judicial behavior between high courts of differing nations. While the judicial politicization theory must await substantial additional testing in other high courts before it can be fully accepted, it may provide a suggestive blueprint for future research.

Acknowledgments

Thanks to H. W. Perry Jr., Tse-min Lin, Sanford Levinson, Bruce Buchanan, and Gary Freeman as well as the anonymous reviewers at PRQ for their thoughtful and constructive comments and assistance. The author also thanks the participants at the 2007 Midwest Political Science Association panel, where an earlier version of this article was presented.

Declaration of Conflicting Interests

The author declared no conflicts of interest with respect to the authorship and/or publication of this article.

Funding

The author received no financial support for the research and/or authorship of this article.

Notes

1. Sheldon and Lovrich (1991; see also Miller 1998; Richardson and Vines 1970) discuss informal judicial selection norms through their "articulation" model, which posits that there are three crucial stages to judicial appointment in the American states: initiation, screening, and affirmation.

2. It is consistent with the theory to speculate that a highly politicized high court will be more likely to engage in strategic behavior. However, that question is beyond the scope of this project, but does provide an interesting avenue for future research.

3. It is theoretically possible that there could be certain circumstances in which a judge may have a preference to nullify the challenged statute on the basis of his or her ideology but not vote to strike the challenged law. This could be due to the wording of the law, or to the judge's vote to dismiss the case on the basis of a related legal doctrine, such as standing. While these cases are certainly possible, they are fairly rare, and any case containing justiciability issues (standing, mootness, ripeness) was excluded from the database used for this project. Furthermore, if the high court did not take a firm position on the constitutionality or unconstitutionality of the challenged law and remanded the case back to a lower court, the case was also excluded.

4. I made one modification: the original paper containing the judges' values (Ostberg et al. 2004; see also Ostberg and Wetstein 2007) separated Justice L'Heureux-Dube's scores into two categories: criminal and civil rights. In order to obtain a single score representing L'Heureux-Dube, these two values were combined.

5. Admittedly, using party as a proxy for ideology is a fairly crude measure. However, this is a very common practice among social scientists when no better measures are available (see, e.g., Smyth 2005). In the United States, the NOMINATE scores (Poole and Rosenthal 1997; Poole 1998) are an excellent indicator of ideology; however, no comparable measures exist for Canada or Australia. Thus, the decision was made to be consistent and use political party for all three cases.

6. An alternative method for calculating judicial politicization is to eliminate all neutral appointments (ideology scores of .500) and instead use the conventional rounding rule so that an ideology score of .500 is considered to have a liberal ideology. If this method is used, then the overall judicial politicization scores would be as follows: United States = .923; Australia = .714; Canada = .529.

7. If only initial high court appointments are used, the judicial politicization scores would be as follows: United States = .909; Australia = .889; Canada = .500.

8. The analyses in this article were also conducted using a database containing only nonunanimous cases. In the interest of brevity, those results are not included here. However, the results from the nonunanimous case analyses did not differ significantly from the combined database analyses, and were even more significant for certain variables.

9. As discussed in the methodological appendix (available online at http://.prq.sagepub.com/supplemental/), the judicial attitude variable was constructed as follows. Liberal or conservative statute direction was coded as 1 for liberal laws and -1 for conservative laws. Then, the judges' ideology scores were transformed into a scale with positive and negative values. Then, the judge's score was multiplied by the statute direction. Thus, each judicial ideology score reflects the degree of agreement with the law (positive value) or disagreement (negative value). For example, in the American case, Justice Scalia (who is a strong conservative) has a rescaled score of -.501; thus, for a conservative law (-1), his value would be .501, indicating ideological congruity.

10. Again, it is expected that the judicial ideology variable will have a negative coefficient, because the greater the degree of ideological disagreement with a statute, the greater the likelihood of a vote to strike.

11. The full model is estimated for each high court and presented in Table 2. The data in Table 2 show that each of the models is highly significant. Also, because the coefficients for logistic regression models are not easily interpretable, Table 3 shows the marginal effect for each independent variable for the likelihood of a vote in favor of judicial activism. For the dichotomous independent variables, the marginal effect indicates the result of changing the variable from 0 to 1. Thus, the values shown in Table 3 allow the effect size of each independent variable to be seen. Those independent variables that were significant at a 90 percent confidence interval or greater are indicated by boldface.

12. Following the Spaeth (2001) coding, there were thirteen different issues areas: criminal procedure, civil rights, speech or religion, due process, privacy, attorneys, unions, economic activity, judicial power, federalism, interstate/ interprovince relations, taxation, and miscellaneous. However, not all of these issues were included in the model because of collinearity issues.

13. I am indebted to an anonymous reviewer for this point.

14. Additionally, the Smyth (2005) article varied from the present study in that Smyth sought to explain dissent rates, not decisions on the merits.

[Reference]

References

Baum, Lawrence. 1998. American courts: Process and policy. 4th ed. New York: Houghton Mifflin.

Bonneau, Chris W., and Melinda Gann Hall. 2009. In defense of judicial elections. New York: Routledge.

Brace, Paul R., and Melinda Gann Hall. 1997. The interplay of preferences, case facts, context, and rules in the politics of judicial choice. Journal of Politics 59:1206-31.

Dubois, Philip L. 1980. From ballot to bench: Judicial elections and the quest for accountability. Austin: University of Texas Press.

Edwards, John Ll. J. 1987. The attorney general and the charter of rights. In Charter litigation, ed. Robert J. Sharpe, 45-68. Toronto, ON: Butterworth.

Epstein, Lee, and Carol Mershon. 1996. Measuring political preferences. American Journal of Political Science 40:261-94.

Epstein, Lee, and Jeffrey A. Segal. 2005. Advice and consent: The politics of judicial appointments. New York: Oxford University Press.

Galanter, Marc. 1974. Why the "haves" come out ahead: Speculations on the limits of social change. Law and Society Review 9:95-160.

Gates, John B., and Charles A. Johnson, eds. 1991. The American courts: A critical assessment. Washington, DC: CQ Press.

Guarnieri, Carlo, and Patrizia Pederzoli. 2002. The power of judges: A comparative study of courts and democracy. Oxford, UK: Oxford University Press.

Hagle, Timothy M., and Harold J. Spaeth. 1991. Voting fluidity and the attitudinal model of Supreme Court decision making. Western Political Quarterly 44:119-28.

Hall, Melinda Gann. 2001. State supreme courts in American democracy: Probing the myths of judicial reform. American Political Science Review 95:315-30.

Hanssen, F. Andrew. 1999. The effect of judicial institutions on uncertainty and the rate of litigation: The election versus the appointment of state judges. Journal of Legal Studies 28:205-32.

Hanssen, F. Andrew. 2004. Learning about judicial independence: Institutional change in the state courts. Journal of Legal Studies 33:431-73.

Heydon, Dyson. 2003. Judicial activism and the death of the rule of law. Australian Bar Review 23:110-33.

Holland, Kenneth M., ed. 1991. Judicial activism in comparative perspective. New York: St. Martin's.

Holland, Kenneth M. 2000. Judicial activism in western Europe. In Handbook of global legal policy, ed. Stuart S. Nagel, 179-208. New York: Marcel Dekker.

Johnson, Timothy R. 2003. The Supreme Court, the solicitor general, and the separation of powers. American Politics Research 31:426-51.

Keck, Thomas M. 2004. The most activist Supreme Court in history: The road to modern judicial conservatism. Chicago: University of Chicago Press.

Koopmans, Tim. 2003. Courts and political institutions: A comparative view. Cambridge, UK: Cambridge University Press.

Kritzer, Herbert M. 2006. Law is the mere continuation of politics by different means: American judicial selection in the twenty-first century. Available from SSRN: http://ssrn.com/ abstract=949946

Lewis, Frederick P. 1999. The context of judicial activism: The endurance of the Warren Court legacy in a conservative age. Lanham, MD: Rowman and Littlefield.

Lindquist, Stefanie A., and David E. Klein. 2006. The influence of jurisprudential considerations on Supreme Court decisionmaking: A study of conflict cases. Law and Society Review 40:135-62.

Lindquist, Stefanie A., and Rorie Spill Solberg. 2007. Judicial review by the Burger and Rehnquist Courts. Political Research Quarterly 60:71-90.

McGuire, Kevin. 1998. Explaining executive success in the U.S. Supreme Court. Political Research Quarterly 51:505-26.

Miller, Mark C. 1998. A comparison of the judicial role in the United States and Canada. Suffolk Transnational Law Review 22:1-25.

Ostberg, C. L., and Matthew E. Wetstein. 2004a. Economic cases and the attitudinal model in the Canadian Supreme Court. Paper presented at the annual meeting of the Midwest Political Science Association, Chicago.

Ostberg, C. L., and Matthew E. Wetstein. 2004b. Equality cases and the attitudinal model in the Supreme Court of Canada. Paper presented at the annual meeting of the Canadian Political Science Association, Winnipeg, Manitoba.

Ostberg, C. L., and Matthew E. Wetstein. 2007. Attitudinal decision making in the Supreme Court of Canada. Vancouver: University of British Columbia Press.

Ostberg, C. L., Susan W. Johnson, Donald R. Songer, and Matthew Wetstein. 2004. The nature and extent of attitudinal decision making in the Supreme Court of Canada. Paper presented at the annual meeting of the American Political Science Association, Chicago.

Ostberg, C. L., Matthew E. Wetstein, and Craig Ducat. 2002. Attitudinal dimensions of Supreme Court decision making in Canada: The Lamer Court, 1991-1995. Political Research Quarterly 55:235-56.

Poole, Keith T. 1998. Estimating a basic space from a set of issue scales. American Journal of Political Science 42:954-93.

Poole, Keith T., and Howard Rosenthal. 1997. Congress: A political-economic history of roll-call voting. Oxford, UK: Oxford University Press.

Richardson, Richard J., and Kenneth Vines. 1970. The politics of federal courts. Boston: Little, Brown.

Roach, Kent. 2000. The attorney general and the charter revisited. University of Toronto Law Journal 50:1-40.

Schubert, Glendon. 1968. Political ideology on the High Court. Politics 3:21-40.

Schubert, Glendon. 1969a. The dimensions of decisional response: Opinion and voting behavior of the Australian High Court. In Frontiers of judicial research, ed. Joel B. Grossman and Joseph Tanenhaus, 163-95. New York: Wiley.

Schubert, Glendon. 1969b. Judicial attitudes and policy-making in the Dixon Court. Osgoode Hall Law Journal 7:1-29.

Segal, Jeffrey, and Albert Cover. 1989. Ideological values and the votes of Supreme Court justices. American Political Science Review 83:557-65.

Sheldon, Charles H., and Nicholas P. Lovrich Jr. 1991. State judicial recruitment. In The American courts: A critical assessment, ed. John B. Gates and Charles A. Johnson, 166-85. Washington, DC: CQ Press.

Smyth, Russell. 2005. The role of attitudinal, institutional and environmental factors in explaining variations in the dissent rate on the High Court of Australia. Australian Journal of Political Science 40:519-40.

Solberg, Rorie Spill, and Stefanie A. Lindquist. 2006. Activism, ideology and federalism: Judicial behavior in constitutional challenges before the Rehnquist Court, 1986-2000. Journal of Empirical Legal Studies 3:237-61.

Songer, Donald R., and Susan W. Johnson. 2002. Attitudinal decision making in the Supreme Court of Canada. Paper presented at the annual meeting of the Midwest Political Science Association, Chicago.

Spaeth, Harold J. 2001. United States Supreme Court judicial database: 1953-2000 terms. 16th ICPSR version. East Lansing: Michigan State University, Department of Political Science; Ann Arbor, MI: Inter-University Consortium for Political and Social Research.

Tate, C. Neal, and Torbjorn Vallinder, eds. 1995. The global expansion of judicial power. New York: New York University Press.

Tushnet, Mark. 2003. Judicial review of legislation. In The Oxford handbook of legal studies, ed. Peter Cane and Mark Tushnet, 164-82. New York: Oxford University Press.

Wolfe, Christopher. 1997. Judicial activism: Bulwark of freedom or precarious security? Rev. ed. Lanham, MD: Rowman and Littlefield.

[Author Affiliation]

David L. Weiden1

1 Indiana University-Purdue University Indianapolis, IN, USA

Corresponding Author:

David L. Weiden, Department of Political Science, Cavanaugh

Hall 503E, Indiana University-Purdue University Indianapolis,

425 University Blvd., Indianapolis, IN 46202-5140

Email: dweiden@iupui.edu