Printer Friendly

The rhetoric of restraint and the ideology of activism.

Criticism of judicial activism has become commonplace in political debate. In recent years it has been political conservatives who have most often sounded the alarm that unelected, activist judges are intruding on the prerogatives of the elected branches. This criticism traces to the Warren Court era, when conservatives called for "judicial restraint" or "strict constructionism" in place of liberal judicial activism, contending that "when liberal Courts overturn democratically enacted laws in favor of liberal, activist constitutionalism, they destroy citizens' rights to democratic participation and self-government." (1) According to these critics, "liberal, activist judges" substitute their personal preferences for the will of the people. (2)

The controversy over judicial activism has become an important issue in recent election campaigns, as well as a central concern in debates over Supreme Court nominations. In 2004, for example, former Attorney General John Ashcroft claimed that "intrusive judicial oversight and second-guessing of presidential determinations ... can put at risk the very security of our nation in a time of war." (3) While much of the attack on judicial activism is populist in nature, academics and judges--such as Robert Bork, Charles Fried, Antonin Scalia and Michael McConnell--make similar arguments. Professor McConnell, for example, has objected that "rule by Judges" is "inconsistent with the principles of self-government." (4) At the time of his appointment, Justice Scalia expressed concern about an "imperial judiciary." (5)

Yet the claim that conservative judges are more restrained than their liberal counterparts has also been challenged, especially in connection with decisions rendered by the Rehnquist Court. The National Director of the ACLU declared that the Rehnquist Court had become "one of the most activist courts in American history." (6) Critics of the Rehnquist Court have claimed that conservative justices were quite activist when the substantive outcome satisfied their ideological preferences.

Justices may find restraint more palatable when they agree with the policy consequences of the government action being evaluated. Indeed, the potential for conflict between the justices' substantive ideological orientations and their professed belief in judicial restraint may manifest itself along several institutional dimensions. Conservative justices' substantive preferences for conservative policies may conflict with legislative, gubernatorial or judicial policy choices made at the state level (see Bush v. Gore, for example), all of which may be produced by elected bodies in the states. In this situation, not only are conservative justices faced with a conflict that implicates their commitment to judicial restraint, but to federalism as well. Or the justices' preferences may conflict with outcomes produced by the coordinate branches in the federal government. Thus, the justices may be faced with constitutional challenges to federal legislation, or to constitutional or statutory challenges to federal administrative actions, that require them to evaluate the propriety of decisions rendered by the elected Congress and the President. In these situations, the justices must decide whether to defer to the elected branches or substitute their own judgments for those of elected officials. Yet in doing so, they may find they face a tension between their preference about the substantive law or policy at issue and their commitment to judicial restraint.

In this paper, we explore these various dimensions by focusing on the degree to which conservative rhetoric advocating judicial deference to the policy actions of the elected branches (at the federal and state levels) actually comports with reality. In particular, we focus on (1) the extent to which the rhetoric of judicial restraint characterizes the voting behavior of conservative justices faced with constitutional challenges to federal and state statutory law, and (2) the extent to which such rhetoric is consistent with conservative justices' voting behavior in cases raising challenges to the administrative actions of the executive branch at the federal level. In our analysis, we equate judicial restraint with a tendency to uphold the decisions of legislators and federal administrative agencies.

We find that, although in both instances conservative justices were somewhat more "restrainist" toward legislative and executive action, that restraint was contingent on the source of the law at issue. Thus, after controlling for ideological preferences regarding the substantive policies challenged in the cases, conservative justices were more deferential to state (as opposed to federal) legislation, and to action by executive branch (as opposed to independent regulatory) agencies. These findings suggest that countervailing considerations may actually explain the conservatives' "restraintist orientation": (1) their ideological commitment to state power and principles of federalism, and (2) their ideological commitment to a powerful executive branch. Indeed, when read in tandem, our results reflect a conservatism that is particularly critical of congressional power.


Activism may take several forms. In his seminal article, Brad Canon defines six dimensions of judicial activism: (1) the degree to which policies adopted through democratic processes are judicially invalidated, (2) the degree to which earlier court decisions, doctrines or interpretations are altered, (3) the degree to which constitutional provisions are interpreted contrary to clear language or original intent, (4) the degree to which judicial decisions make substantive policy rather than preserve democratic processes, (5) the degree to which the judiciary eliminates discretion of other governmental actors and makes policy itself, and (6) the degree to which judicial decisions preclude serious consideration of governmental problems by other political actors. (7)

Given the federal structure of our government, countermajoritarian activism itself has two dimensions. Judicial review may be invoked to invalidate federal or state laws; however, the invalidation of a federal law implicates somewhat different concerns than the invalidation of a state law or local ordinance. When the justices wield their power of judicial review to strike a federal statute, their action implicates separation of powers issues because the Court's decision challenges policy choices made by the coordinate branches. Similarly, when the justices invalidate decisions and actions of the federal administrative agencies, they encroach on the President's policy making authority within the executive branch. Such actions raise the potential for recriminations by Congress and the President, including reversal through constitutional amendment or restrictions on judicial budgets, among other things. (8) Moreover, in terms of majoritarianism, an act of Congress or the President at least arguably represents the will of the majority of U.S. citizens, rather than the majority of citizens within a particular state. Thus invalidation of congressional enactments and federal administrative decisions has more immediate national implications and thus broader countermajoritarian consequences. In contrast, Canon suggests that voiding state laws is "arguably less offensive in principle." (9) Practically, however, he recognizes that invalidation of a state or local law may also have national consequences to the extent such a ruling calls into question similar laws in other jurisdictions. Of course, such rulings are also "countermajoritarian" at the state or local level as well, and have the potential to alter the balance of power between the federal and state governments.


This section presents empirical data on the degree to which judges appear activist or restrained in reviewing legislation or administrative actions. We begin with a simple count of the relative frequency with which individual justices vote to strike a statute or decline to defer to an administrative agency decision, reported in Table 1.

A cursory look at these statistics provides some support for the assertion that conservative justices show more restraint. Justices White, Rehnquist, and Powell appear particularly deferential, while Marshall and Brennan were more aggressively activist. All the justices were more likely to strike legislation than agency action, although the relative difference varied widely among the justices. The association between conservatism and restraint is not strong, and the activism probabilities for Ginsberg and Breyer are not materially different from those for Thomas and Scalia.

Descriptive or bivariate statistics like those in Table 1 are suggestive, but they may generate false conclusions because of the absence of control variables that provide alternative explanations for the votes. For example, this frequency distribution fails to account for the source of the statute or agency decision and the ideological direction of the enactment or agency action under review.

In the following section, therefore, we use multivariate regression to compare the effects of several factors likely to influence the justices' votes in cases involving challenges to legislative enactments and agency action. These more comprehensive models include variables that account for the influence of ideology and the source of the policy at issue, while controlling for other factors that might influence judicial behavior in these important cases. We begin with our model of judicial review of legislation at the state and federal level, and then turn our attention to judicial review of administrative agencies.


Many of the most controversial Supreme Court decisions, and those most likely to be condemned as activist, involve invalidation of legislation on constitutional grounds. The central purpose of this article is to examine the historic claims of judicial restraint by conservatives. Consequently, our primary hypothesis is that more conservative justices will be less likely to vote to invalidate statutes on constitutional grounds. We recognize, however, that the decision to invalidate a statute may be shaped by federalism concerns. Indeed, one of the Rehnquist Court's most distinguishing characteristics is its preoccupation with federalism and states' rights,"' and most of the Court's pro-state-rights decisions may be attributed to the influence of the conservative justices on the Court. Our second hypothesis, therefore, is that conservative justices will be less likely to vote to strike state legislation. In addition, however, we also seek to evaluate whether the justices' alleged commitment to restraint is contingent on their attitudinal reaction to the legislative policy at issue. Without question, the justices' votes are often substantially influenced by their ideological preferences regarding the substance of individual cases. (11) Our third hypothesis is that justices will be more likely vote to uphold legislative enactments whose substantive policies conform to the justices' ideological preferences. In addition to these hypothesized influences on the justices' voting behavior, to control for their effects, we also included other independent variables that have a likely influence on the justices' votes. These control variables are explained below.

To evaluate our hypotheses, we coded the justices' votes in all cases heard by the Rehnquist Court during the 1986 through 2004 Terms in which the constitutionality of a federal, state or local statute was challenged. The dependent variable is dichotomous, coded as 1 if the justice voted to strike the statute and 0 otherwise. (12) Our data included 2465 votes for analysis (1600 in cases involving state legislation, and 865 in cases involving federal legislation). We present three models: (1) Model 1 is estimated based solely on the justices' votes in cases involving federal statutes; (2) Model 2 is estimated based on the justices' votes in response to challenges to state or local laws; and (3) Model 3 includes the votes of the justices in both federal and state/local cases. Because our dependent variable is dichotomous, we estimate the models using logistic regression with robust standard errors, clustering on the individual justice.

1. Variables and Measurement

Justice Ideology. To measure the justices' ideology, we relied on the Martin-Quinn ideal points, (13) arranged such that positive scores are associated with liberal justices, negative scores with conservative justices. Although the Martin Quinn scores are calculated based on the justices' voting behavior, and thus may raise endogeneity problems for some analyses, we are less concerned about that problem with respect to our model. In particular, we are not predicting the ideological direction of the justices' votes, but rather whether they voted to strike a challenged statute. Use of the Martin Quinn scores to construct an independent variable does not raise the same degree of concern.

According to the conventional wisdom, conservative justices on the Rehnquist Court are attitudinally predisposed to reduce federal power vis-a-vis the states and to enhance or protect state authority. (14) We thus test the relationship between ideology and the propensity to strike legislation (i.e. be "activist") in the following fashion. Because higher values of Justice Ideology indicate more liberal justices, a positive coefficient on this variable will indicate that conservative justices are less likely than liberal ones to vote to declare laws unconstitutional (therefore support the first hypothesis). In Model 1, a positive coefficient on the Justice Ideology variable will indicate that conservatives are less likely to vote to strike federal laws. In Model 2, a positive coefficient on the Justice Ideology variable will indicate that conservatives are less likely to vote to strike state/local laws. In Model 3, these relationships are subjected to an additional statistical test. In that model, the Justice Ideology measure is multiplied by a dummy variable indicating whether the statute is federal (coded 1) or state/local (coded 0). As a result, in Model 3, the coefficient on the ideology variable alone indicates the impact of justice ideology in cases involving state or local legislation. A significant coefficient on the interactive term (Ideology * Federal Statute), moreover, will indicate that the effect ideology is different depending on the source of the law being evaluated.

In addition to the justices' ideological response to the source of the statute at issue, we also hypothesized that justices' propensity to strike state or federal legislation will be shaped by their preferences for the policy embodied in the challenged law. The variable Alignment with Statutory Direction (hereafter Alignment) therefore measures the degree of ideological agreement between the justice and the statutory enactment. To construct this variable, we assigned an ideological score to the individual statutes using the directionality codes in the U.S. Supreme Court Database. Liberal statutes received a value of 1, and conservative statutes -1. To construct a variable measuring the consistency between the statute and the justice's ideologies, we simply multiplied the statute's ideology score and the individual justices' ideology score. Thus, the more extreme the justice's ideology (and therefore the more extreme the justice's Martin-Quinn Score), the more extreme the value of Alignment. Positive values on this variable indicate that the justice supports the substantive goals of the legislation at issue, negative values the opposite. We therefore expect its coefficient to be negative (where support is high, the justice will be less likely to vote to strike).

Age of Statute. The variables described above are of central relevance to our working hypothesis, but the literature also directs us to other control variables that may influence the justices' votes. First, the justices may also be influenced by a statute's age. In his early study, Dahl presented findings that the bulk of cases in which federal legislation were declared unconstitutional occurred more than four years after the legislation was enacted, and suggested that the Court was reluctant to invalidate legislation enacted by the "live" or current national majority. (15) Because we expect the justices to be more reluctant to invalidate laws from more recently elected legislative bodies, we included a variable in our models reflecting the age of the statutory enactment. We expect a positive coefficient on this variable.

Solicitor General The influence of the Solicitor General at the Supreme Court is well documented, (16) whether participating as amicus curaie or representing the federal government directly. (17) The Solicitor's high win rate reflects a unique level of deference that is due either to institutional respect for the Executive Branch or the Court's recognition of legal expertise of the office. Our models thus include three dummy variables (Solicitor involvement as amicus supporting the statute at issue, as amicus opposed to the statute, or as party representative) reflecting the Solicitor's Office presence in a given case.

Amicus Differential The Court may also be influenced by interest groups appearing via amicus briefs. Some research suggests that amicus curiae have an influence on outcomes before the Court. (18) We capture this possible influence through the differential support of amici for the parties challenging or supporting the statute, measured as the number of amicus briefs supporting the statute minus the number of amicus briefs opposing the statute. Larger positive values on this variable indicate greater support for the statute among the amici; negative values indicated that briefs opposed to a statute outnumbered those in favor. We thus expect a negative coefficient on this variable (the more amici in support, the less likely the justice will vote to strike the statute).

Civil Liberties. The nature of the legal issues presented to the Court may also affect the justices' decisionmaking. Historically, the Court has been particularly protective of civil liberties claims grounded in the Bill of Rights, and some of these claims invoke the "strict scrutiny" standard of review of legislation. (19) Hence, we expect such statutes to be more subject to invalidation and include the nature of the claim as a dummy variable, coded as 1 if a civil liberties challenge was brought to the statute, and 0 otherwise.

Lower Court Invalidation. The Supreme Court often selects cases in order to reverse the lower court. To control for this selection effect, our models include a dummy variable indicating whether the lower court (usually a Circuit Court of Appeals) invalidated the statute (coded as 1) or upheld it (coded as 0).

2. Results

The three models presented in Table 2 reflect a complex portrait of the justices' votes in judicial review cases involving constitutional challenges to legislative enactments. First, it is clear from the tables that judicial ideology is a critical determinant in the justices' voting behavior in judicial review cases. In Model 1, the Justice Ideology variable is statistically significant and negative. This indicates that liberals are less inclined to strike federal legislation and conservatives more likely to do so, directly contradicting our hypothesis that conservatives would show more judicial restraint than liberals. In Model 2, the opposite ideological influence is evident with respect to state and local legislation, with conservatives less likely and liberals more likely to vote to strike the law at issue. Finally in Model 3, the interactive term between ideology and federal laws is negative and significant, confirming conservatives are significantly less deferential toward federal laws compared to state laws.

The Alignment variable measures the effect of congruence between the justices' policy preferences and the substantive effects of the statute being reviewed. In all three models the Alignment variable is highly significant and in the expected direction. Thus, for all justices, activism is closely related to ideology, as justices are much less likely to invalidate a law that is ideologically aligned with their preferences and more likely to vote to invalidate when the law contradicts their policy preferences.

As for the control variables, most of them are statistically related to the dependent variable in the expected directions. The justices plainly respond to the position of the Solicitor General, whether the Solicitor participates as a party or amicus. In addition, with the exception of Model 1, the extent to which a positive differential exists between briefs filed in support of the statute and those filed in opposition to it affects the justices' willingness to strike legislation. At least in the case of state or local laws, therefore, the justices are sensitive to public opinion as expressed through amicus briefs filed by interest groups. Furthermore, the nature of the constitutional challenge matters: in civil liberties cases, the justices are more willing to vote to strike legislation, possibly because of the more stringent legal tests that are often applied in those cases. And the control for the nature of the Court's agenda is also statistically significant, yet the models reveal that the lower court variable operates differently in different contexts. Thus, in Model 1, if the lower court struck the statute below, the justices were also more likely to strike the statute. This outcome is different than what we expected, given the justices' propensity to reverse. But in the context of a challenge to a federal statute, it may be that lower court judges exercise caution in considering a constitutional challenge and choose to invalidate federal statutes only in extreme cases. In Model 2, however, the variable is in the expected direction (negative). At least in the case of state or local laws, the justices are likely to vote to reverse the lower court. Finally, the age of the statute at issue did not influence the justices' votes in any of the models.


Our second analysis involves judicial review of federal agency decisions. In the context of administrative action the connections between judicial restraint and the democratically elected branches are not as obvious as in the case of legislative enactments. Yet even though many administrative personnel are themselves unelected, agency officials, even those of independent regulatory commissions, are more accountable for their policymaking decisions than are judges. Judicial restraint implies deference to decisions of administrative agencies because agency officials are subject to the ongoing supervision of the President and Congress.

Presidents have many ways to influence administrative and regulatory policy. The President chooses the top levels of bureaucratic personnel, and through these appointments exerts control over bureaucratic policy. (20) Elena Kagan notes that every president since Ford has required proposed Federal Regulations to be approved by the White House before being issued, and can use executive orders and other tools to direct administrative actions. (21) Although administrative personnel are not elected, they are closely accountable to the president.

Congress controls the statutory authority and operating budgets which empower agencies. These tools give Congress a great deal of leverage over agency decisions. McCubbins and Schwartz, in their seminal paper likening congressional supervision of agency policymaking to "fire alarms" or "police patrols," explained that Congress has a myriad of indirect ways to monitor agency policymaking. (22) Even if Congress is not actively supervising the agencies, citizens and groups who oppose agency actions will bring them to Congress' attention. Congressional committees can then hold hearings or conduct investigations into agency actions. Weingast and Moran present empirical evidence that congressional preferences influenced the Federal Trade Commission in the 1970s and 1980s, even without visible intervention into the commission's activities. (23) More recently Balla and Wright have shown that Congress can ensure that its views are represented in agency policymaking processes by influencing the membership of agency advisory boards. (24) It is clear that Congress possesses significant tools to influence the actions of both executive and independent agencies, and that it has both the means and the incentive to exercise these tools. Because agencies are subject to continuous monitoring and control by both Congress and the President, their decisions are more closely connected to representative institutions than are court decisions.

1. Hypotheses

Given the agencies' ties to the elected branches, the same principles of judicial restraint should apply in this context as in the case of legislative enactments. Thus, if conservative justices are more committed to judicial restraint, this commitment should show up in a decreased likelihood of overturning the decisions of administrative agencies. The primary hypothesis with regard to the Court's administrative law jurisdiction is that conservative justices will be less likely than liberal justices to vote to overturn decisions made by administrative agencies.

The argument for judicial restraint may be stronger when the agency whose decision is being litigated is within the executive branch. Presidents exercise closer control over these agencies, primarily because they can fire the heads of such agencies. Therefore, the connection between the democratic process and agency decisions may be closer in the context of executive branch agencies. Therefore, our second hypothesis is that the tendency of conservative justices to be more deferential to the decisions of administrative agencies will be more pronounced when the decision was made by an executive branch agency.

As discussed above, there is reason to believe that the personal ideologies of justices will influence their votes. Sheehan's 1990 study shows that support for agency decisions is strongly related to their ideologies. (25) For example, Justice Douglas supported 85 percent of liberal agency decisions, but only 33 percent of conservative agency decisions. These findings lead to a third hypothesis: a justice will be more likely to support administrative action when the policy effects of that action are consistent with the justice's policy preferences.

To evaluate these hypotheses, we identified all cases heard by the Rehnquist Court during the 1986 through 2004 Terms in which the action of a federal administrative agency was challenged. Once we had identified the cases involving challenges to administrative action, we again used the Rehnquist Court Justice-Centered Database to construct our dependent variables. This coding process resulted in dichotomous dependent variables. In the context of challenges to administrative actions, the dependent variable is coded 1 if the justice voted to overturn the agency action and 0 otherwise. Our data included 2114 votes in cases involving challenges to administrative actions.

2. Model

With the above background and hypotheses, we operationalized variables to capture the qualities necessary for a test of the hypotheses, similar to the analysis of legislative invalidations. This section sets out and describes the independent variables and nature of the statistical test.

Justice Ideology. We hypothesize that conservative justices will be less likely to vote to overturn agency decisions. To indicate the justices' ideological predispositions we again use Martin and Quinn's Ideal Point Scores as discussed above. Alignment with Agency Action is calculated as in the analysis involving Alignment with Statutory Direction, based on the ideological direction of the agency decision being reviewed. We expect higher levels of alignment to lead to lower propensity to overturn the agency action.

Agency Policy Direction is a categorical variable coded 1 when the agency is arguing for a conservative policy position and -1 when the agency is advocating a liberal policy outcome. We computed this variable using the Spaeth database's variables indicating the ideological direction of the lower court decision and the appellant or respondent status of the government agency. If the lower court decision was liberal, the agency policy position is coded as conservative if the agency is the appellant and liberal if it is the respondent. If the lower court decision was conservative, the agency policy position is coded as liberal if the agency is the appellant and conservative if it is the respondent. Previous research has found that liberal agency decisions garner more support from Supreme Court Justices. (26)

Executive Branch Agency is a dummy variable coded 1 if the agency is within the executive branch and 0 if it is not. The inclusion of this variable is motivated by research investigating whether executive branch or independent agencies fared better in the Supreme Court. (27) It is used to evaluate whether agencies under the political control of the president are more or less successful at attracting support from the justices.

Party of President is a dummy variable coded 1 when the President is Republican and 0 when the President is a Democrat. In our data the only Democratic President is Clinton. This variable is included to control for the ideology of the sitting president. (28)

Lower Court Support for Agency Action is a dummy variable indicating that the lower court (usually a Circuit Court of Appeals) supported the agency action, introduced because of the Court's proclivity to reverse.

To evaluate our hypotheses, we present three models. We separately estimate the effects of the independent variables on the justices' decisions in disputes emerging from executive branch agencies and from independent agencies, and then combine these data to evaluate their effects in both types of cases. As noted, the dependent variable is dichotomous and so we utilize logistic regression to estimate our model; given the potential that the justices' individual votes would not be independent across cases, we again clustered on the individual justice. We present the results of our data collection process and model estimation below.

3. Results

The three columns of numbers in Table 3 relate to justices' votes in cases emerging out of independent agencies, executive branch agencies, and all agencies, respectively. Our first hypothesis, that conservative justices would be less likely to vote to overturn agency actions, suggests that Justice Ideology should have a positive and significant effect on the dependent variable. However, the negative, and very nearly significant (p=.100), coefficient for Justice Ideology in the first column shows that liberal justices showed more support for independent agencies than conservatives.

Moving to the second column of numbers, Justice Ideology does show the positive and significant relationship with the dependent variable predicted by the first hypothesis. With a coefficient (.105) nearly three times the size of its standard error (.039), this variable is highly significant and persuasively shows that conservative justices were more likely than liberal justices to defer to the decisions of executive branch agencies. The similarly positive and significant coefficient for Justice Ideology in the third column of numbers shows that this relationship persists when decisions from both types of agencies are considered. The negative and significant coefficient just below, for Justice Ideology * Independent Agency, reflects the fact that conservative justices are significantly less likely to support decisions by independent agencies compared to decisions by executive branch agencies. This result supports the second hypothesis, that conservative deference to administrative decisions will show up more in cases coming out of executive branch agencies than in cases coming out of independent agencies.

Our third hypothesis concerns the effect of alignment between the justices' assumed policy preferences and the effect of the agency decision being evaluated. The Alignment with Agency Action variable reflects this relationship. The negative and significant coefficients for this variable in all three columns indicate that justices are significantly less likely to vote to overturn decisions whose policy results they support. This result supports our third hypothesis, and shows that justices' ideologies are important in determining their stances toward agency decisions.

The results presented in Tables 2 and 3 show that, in certain situations and after controlling for other likely influences, conservative justices are indeed more reluctant to overturn laws and agency decisions. However, the highly significant effects of Alignment with Statute's Direction (Table 2) and Alignment with Agency Action (Table 3) show that policy agreement also influences these decisions. The regression results presented so far cannot tell us how much difference these factors make, but we can use these results to predict the probability that a particular justice will overturn a law or agency decision in particular circumstances. The next section does just that.


Table 4 shows the predicted probabilities that justices with the ideological scores of Scalia, O'Connor, and Stevens would vote to declare state or federal laws unconstitutional. (29) The top four lines in the table relate to a conservative justice with the same ideological score as Justice Scalia. These results suggest that the source of the law does make some difference to a Justice like Scalia. For example, Justice Scalia's likelihood of opposing a liberal state law is .330, or four percent lower than his probability of opposing a liberal federal law (.373). Likewise, Justice Scalia's probability of opposing a conservative state law (.121) is about two percent lower than his probability of opposing a conservative federal law (.143). However, the effects of the source of the law are dwarfed by the effects of the ideological nature of the law.

According to the model, the probability that Justice Scalia would vote to strike a liberal state law is .330, while the probability that he would vote to strike a conservative state law is only .121. This means that the change from a liberal to a conservative law lowers Justice Scalia's probability of voting against the law by about 21 percent. Similarly, these values show that Justice Scalia's probability of opposing a liberal federal law is .373, about 23 percent greater than his probability of opposing a conservative federal law (.143).

Moving to a more liberal Justice, such as Stevens, changes the results somewhat. Stevens' probability of voting to overturn a liberal state law is .290, while his probability of voting to overturn a liberal federal law is .162, showing that he is about 13 percent more likely to vote against a state law compared to a federal law. On the other hand, the differences in Stevens' predicted levels of support for liberal as opposed to conservative laws shows that the Stevens is about eight percent more likely to support liberal laws compared to conservative ones.

The bottom section of table 4 presents similar predictions for a Justice with Sandra Day O'Connor's ideology. As with Justice Scalia, the liberal or conservative nature of the law has a much greater impact on O'Connor's predicted votes compared with whether the law came from a state or the national government.

Moving to the predicted probabilities that justices will vote to overturn the actions of administrative agencies in Table 5, we see a similar story. (30) Justice Scalia is about 15 percent more likely to vote against liberal agency decisions compared to conservative ones, but only about 7 percent more likely to vote against independent agency decisions compared to executive branch decisions.

For Justice Stevens, the effect of the ideological consequences of the agency decision is striking. Stevens is nearly 30 percent more likely to vote to invalidate a conservative decision by an executive branch agency compared to a liberal decision by such an agency. He is about 15 percent more likely to vote to overturn a conservative decision by an independent agency, compared to a liberal decision by the same type of agency. Justice Justice Stevens is four percent less likely to vote to invalidate a liberal decision by an executive branch agency than a liberal decision by an independent agency (.264 versus .304). However, when it comes to conservative decisions, Justice Stevens is nearly 10 percent more likely to support executive branch decisions compared to decisions made by independent agencies.

The bottom four lines of Table 5 present the predicted probabilities using Justice O'Connor's ideological score. Note that for Justice O'Connor the model generates predictions that are affected only slightly by the liberal or conservative nature of the agency decision, or the source of the decision.

Overall, these predicted probabilities suggest that both the source of the law or action under consideration and its ideological implications can exert meaningful influence on the likelihood that a justice will vote to overturn it. However, in most cases the effect of ideology appears to matter more than whether the law was made by the state or federal government, or whether the agency decision came out of the executive branch or an independent agency.


Our findings yield some interesting conclusions about the association of conservatism and judicial restraint. The justices' votes in these cases are structured by their ideologies in two ways: ideology influences their responses to the source of the policy under review, as well as to the substantive content of the challenged policy. Conservative justices, like liberals, are ideological in their decision making but temper their ideologies with respect and deference for certain institutions. Conservatives show deference to state decisions and those of the executive branch, while not extending this deference to actions of the national legislature and the independent regulatory agencies often viewed as its agents. It appears that conservative justices do not show more judicial restraint across the board, but instead show deference to different institutions than do liberal justices.

The empirical results pertaining to constitutional review of laws suggest that conservative justices are more likely to show restraint when evaluating state laws, but liberals are more likely to show restraint when dealing with federal laws. Similarly, the results pertaining to review of administrative actions indicate that conservative justices show restraint (relative to liberal justices) when considering executive branch agency decisions, but not when considering decisions of independent agencies. It may be that, rather than a broad commitment to judicial restraint, these results indicate commitments to two different political principles: federalism and executive authority. Indeed, a comprehensive commitment to judicial restraint would presumably extend to the decisions of national lawmakers and independent agencies. It is difficult to see why members of Congress have less legitimacy than state legislators. And, whatever one may think of the constitutional status of independent regulatory and administrative agencies, if the justices are going to allow these agencies to exist then presumably their decisions should be accorded legitimacy. Thus, it is possible that instead of judicial restraint these results indicate a commitment to a particular understanding of the Constitution, one in which both states and the executive branch enjoy relatively great autonomy from legal restrictions (and the corresponding authority to put their preferences into policy).

The second common point that emerges from these two sets of analyses is that the justices' policy preferences are the dominant factor in their decisions. In all the analyses, a justice who disapproved of the policy consequences of the law or decision being challenged was much more likely to vote to overturn that law or decision. The predicted probabilities in Tables 4 and 5 make this point very clearly. Thus, even though conservative justices showed greater judicial restraint in some contexts, their decisions (along with those of liberal justices) were overwhelmingly the result of their views of good public policy.

(1.) Robert M. Howard & Jeffrey A. Segal, A Preference for Deference? The Supreme Court and Judicial Review, 57 POL. RES. Q., Mar. 2004, at 131, 132.

(2.) Lori A. Ringhand, Judicial Activism: An Empirical Examination of Voting Behavior on the Rehnquist Natural Court, 24 CONST. COMMENT. 143 (2007). The critique of activism has been wielded by conservatives, such that "whenever a politician uses the term 'activist judge,' the word liberal is sure to follow." Thomas Healey, The Rise of Unnecessary Constitutional Rulings, 83 N.C.L. REV. 847, 929 (2005).

(3.) See Jerry Seper, Ashcroft Rips Federal Judges on National Security, WASH. TIMES, Nov. 13, 2004.

(4.) Michael W. McConnell, The Role of Democratic Politics in Transforming Moral Convictions Into Law, 98 YALE L.J. 1501, 1538 (1989).

(5.) Jean Morgan Meaux, Justice Scalia and Judicial Restraint: A Conservative Resolution of Conflict Between Individual and State, 62 TUL. L. REV. 225, 227 (1987).

(6.) Edward Walsh, An Activist Court Mixes Its High-Profile Messages, WASH. POST, July 2, 2000, at A6.

(7.) Bradley C. Canon, Defining the Dimensions of Judicial Activism, 66 JUDICATURE 236, 239 (1983).

(8.) See, e.g., James Meernik & Joseph Ignagni, Judicial Review and Coordinate Construction of the Constitution., 41 AM. J. POL. SCI. 447 (1997). On the question whether the Court responds to potential retribution by Congress, see Jeffrey A. Segal, Separation of Power Games in the Positive Theory of Congress and Courts, 91 AM. POL. SCI. REV. 28 (1997); on Congressional responses to Supreme Court decisions on the constitutionality of federal statutes, see Meernik & Ignagni, supra.

(9.) See Canon, supra note 7, at 241.

(10.) See Richard H. Fallon, Jr., The "Conservative" Paths of the Rehnquist Court's Federalism Decisions, 69 U. CHI. L. REV. 429 (2002); Calvin Massey, Federalism and the Rehnquist Court, 53 HASTINGS L.J. 431 (2002); Keith E. Whittington, Taking What They Give Us: Explaining the Court's Federalism Offensive, 51 DUKE L.J. 477 (2001).


(12.) We are grateful to Rorie Spill Solberg, who was instrumental in collecting these data in connection with another co-authored project regarding judicial review.

(13.) Andrew D. Martin & Kevin M. Quinn. Dynamic hleal Point Estimation via Markov Chain Monte Carlo for the U.S. Supreme Court, 1953-1999, 10 POL. ANALYSIS 134 (2002).

(14.) See, e.g., Calvin Massey, Federalism and the Rehnquist Court, 53 HASTINGS L.J. 431 (2002); Ernest A. Young, The Rehnquist Court's Two Federalisms, 83 TEX. L. REV. 1 (2004). Some have questioned the sincerity of the Court's commitment to federalism, however, suggesting that it was a mere front for conservative policy preferences. See, e.g., Richard H. Fallon, Jr., The "Conservative" Paths of the Rehnquist Court's Federalism Decisions, 69 U. CHI. L. REV. 429 (2002); Peter J. Smith, Federalism, Instrumentalism, and the Legacy of the Rehnquist Court, 74 GEO. WASH. L. REV. 906 (2006).

(15.) Robert Dahl, Decision-making in a Democracy: The Supreme Court as a National Policy-maker, 6 J. PUB. L. 279 (1957).


(17.) Seth P. Waxman, Foreword: Does the Solicitor General Matter?, 53 STAN. L. REV. 1115, 1117 (2001); David A. Strauss, The Solicitor General and the Interests of the United States, 61 LAW & CONTEMP. PROBS., Winter/Spring 1998, at 165, 172 (discussing the Solicitor's reputation at the Court).

(18.) See, e.g., Joseph D. Kearney & Thomas W. Merrill, The Influence of Amicus Briefs on the Supreme Court, 148 U. PA. L. REV. 743 (2000).

(19.) See Gerald Gunther, The Supreme Court, 1971 Tern,--Foreword: In Search of Evolving Doctrine on a Changing Court: A Model for a Newer Equal Protection, 86 HARV. L. REV. 1, 8 (1972) (famously proclaiming that such review was strict in theory but fatal in fact).

(20.) See B. Dan Wood & Richard W. Waterman, The Dynamics of Political Control of the Bureaucracy, 85 AM. POE. SCI. REV. 801 (1991); Nolan McCarty & Rose Razaghian, Advice and Consent: Senate Responses to Executive Branch Nominations 1885-1996, 43 AM. J. POL. SCI. 1122 (1999).

(21.) See Elena Kagan, Presidential Administration, 114 HARV. L. REV. 2245 (2001).

(22.) Mathew D. McCubbins & Thomas Schwartz, Congressional Oversight Overlooked: Police Patrols versus Fire Alarms, 28 AM. J. POL. SCI. 165, 165 (1984).

(23.) Barry R. Weingast & Mark J. Moran, Bureaucratic Discretion or Congressional Control? Regulatory Policymaking by the Federal Trade Commission, 91 J. POL. ECON. 765 (1983)

(24.) Steven J. Balla & John R. Wright, Interest Groups, Advisory Committees, and Congressional Control of the Bureaucracy, 45 AM. J. POL. SCI. 799 (2001).

(25.) See Reginald S. Sheehan, Administrative Agencies and the Court: A Reexamination of the Impact of Agency Type on Decisional Outcomes, 43 W. POL. Q. 875 (1990).

(26.) See Jeff Yates, Presidential Bureaucratic Power and Supreme Court Justice Voting, 21 POL. BEHAV. 349 (1999).

(27.) See Donald W. Crowley, Judicial Review of Administrative Agencies: Does the Type of Agency Matter?, 31 W. POL. Q. 265 (1987); Sheehan, supra note 25.

(28.) This political relationship has been suggested by research. See Jeffrey Yates, supra note 26.

(29.) The values in the table are calculated under the assumption that there are five more amicus briefs in favor of striking the laws than in favor of upholding the laws, that the law in question does not concern civil liberties, that the Solicitor General has submitted an amicus brief in support of the law, and with the other independent variables set at their means.

(30.) These probabilities were calculated with all independent variables except Justice Ideology, Alignment with Agency Decision, and Agency Policy Position set at their means.

Stefanie A. Lindquist*

Joseph L. Smith **

Frank B. Cross ***

* Associate Professor of Political Science and Law, Vanderbilt University.

** Assistant Professor of Political Science, University of Alabama.

*** Herbert D. Kelleher Centennial Professor of Business Law, McCombs School; Professor of Law, University of Texas Law School; Professor of Government, University of Texas at Austin.
Table 1: Propoertion of Votes Invalidating
Legislation or Agency Action, By Justice

Justice Votes to Invalidate Votes to Invalidate
 Statutory Enactments Agency Action

Blackmun .544 .367
Brennan .617 .480
Breyer .441 .298
Ginsburg .468 .310
Kennedy .481 .358
Marshall .572 .488
O'Connor .435 .412
Powell .310 .300
Rehnquist .283 .280
Scalia .415 .368
Souter .527 .370
Stevens .534 .391
Thomas .439 .431
White .316 .183

Table 2: Logit Model of Justices' Voting Behavior in Cases
Involving Judicial Review of Legislative Enactments, 1986-2004
(with hypothesized direction of coefficients)

Variable Model 1 Model 2 Model 3
 Federal Laws State Laws All Laws

 -.117 .223 .113
Justice Ideology (+/-) (.070) * (.078) ** (.065) **

Alignment with -.352 -.207 -.260
Statutory Direction (-) (.056) *** (.061) *** (.047) ***

Ideology * Federal -.316
Statute (-) -- -- (.070) ***

Federal Statute (-) -- -- (.218) **

 .002 -.001 -.001
Age of Statute (+/-) (.005) (.001) (.001)

 -.869 -.802 -.817
SG Support--Amicus (-) (.265) *** (.119) *** (.116) ***

 -.249 -.529
SG Support--Party (-) (.208) -- (.149) **

 .661 .529
SG Oppose--Amicus (+) -- (.179) *** (.149) ***

 -.020 -.019 -.023
Amicus Differential (-) (.030) (.014) (.013) *

Civil Liberties .574 0.338 .420
Challenge (+) (.227) ** (.121)*** (.063) ***

Lower Court .184 -.422 -.222
Invalidation (-) (.136) * (.089) *** (.075) **

 -.788 .341 .360
Constant (.277) ** (.125) ** (.183) *

Pseudo R-Square .074 .089 .082

 1082.17 **
Wald Chi-Square 158.78 *** * 838.57 ***

N 865 1600 2465

* p<.10, p<.05, ** p<.01, *** p<.001. All estimated using Stata 8.2
with robust standard errors clustered on the individual justice.
Significance tests are one-tailed.

Table 3: Logit Model of Justices' Voting Behavior in Cases
Involving Judicial Review of Administrative Agency Decisions,

 Model 1 Model 2 Model 3
Variable Independent Executive All
 Agencies Agencies Agencies

 -.089 .105 .104
Justice Ideology (+) (.070) (.039)*** (.039) **

Justice Ideology * -.184
Independent Agency (-) -- -- (.049) **

Alignment with Agency -.399 -.203 -.255
Action (-) (.026) *** (.022) *** (.020) **

Agency Policy .299 .173 .204
Direction (+) (.085) *** (.043) *** (.042) **

Party of President at -.382 .280 .085
Time of Court Decision (.174) ** (.135) * (.126)

Independent Agency -- -- (.108)

Lower Court Support .684 .359 .473
for Agency Action (+) (.154) *** (.099) *** (.104) **

 -.733 -.916 -.927
Constant (.203) *** (.113) *** (.164) ***

Pseudo R-Square .126 .051 .065

Wald Chi-Square 241.13 *** 225.73 *** 365.61 ***

N 645 1469 2114

* p<05, ** p<01, *** p<.001. All estimated using Stata 8.2 with robust
standard errors clustered on the individual justice. Significance tests
are one-tailed.

Table 4: Predicted Probabilities of Vote to Strike Legislation

Justice Source of Statute Probability of
Simulated Statute Direction Vote to Strike

Scalia State Liberal .330
Scalia State Conservative .121
Scalia Federal Liberal .373
Scalia Federal Conservative .143

Stevens State Liberal .292
Stevens State Conservative .356
Stevens Federal Liberal .162
Stevens Federal Conservative .206

O'Connor State Liberal .316
O'Connor State Conservative .188
O'Connor Federal Liberal .281
O'Connor Federal Conservative .164

Table 5: Predicted Probabilities of Vote to
Invalidate Agency Action

 Probability of
 Agency Vote to
Justice Type of Policy Invalidate
Simulated Agency Direction Agency Action

Scalia Executive Liberal .390
Scalia Executive Conservative .237
Scalia Independent Liberal .466
Scalia Independent Conservative .297

Stevens Executive Liberal .264
Stevens Executive Conservative .552
Stevens Independent Liberal .304
Stevens Independent Conservative .453

O'Connor Executive Liberal .347
O'Connor Executive Conservative .327
O'Connor Independent Liberal .364
O'Connor Independent Conservative .345
COPYRIGHT 2007 Constitutional Commentary, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2007 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Symposium: The Rehnquist Court in Empirical and Statistical Retrospective
Author:Lindquist, Stefanie A.; Smith, Joseph L.; Cross, Frank B.
Publication:Constitutional Commentary
Date:Mar 22, 2007
Previous Article:Judicial activism: an empirical examination of voting behavior on the Rehnquist natural court.
Next Article:An empirical analysis of the confirmation hearings of the justices of the Rehnquist natural Court.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters