Decision Making on State Courts of Last Resort

As Brace and Hall (Reference Brace and Gann Hall1997, 1207) state, “To understand judicial behavior, the pivotal role of institutions must be taken into account.” To that end, scholars have significantly advanced our understanding of judicial behavior by examining decision making under the disparate institutional designs of state courts (e.g., Hall Reference Hall1995; Brace and Hall Reference Brace and Gann Hall1997; Langer and Wilhelm Reference Langer and Wilhelm2005). Across states, judges face different electoral pressures owing to varied state judicial selection and retention procedures. Generally speaking, sitting judges on state courts of last resort face either reelection or reappointment considerations.

For judges facing reelection, success is far from ensured (Hall Reference Hall2001). Instead, these elections introduce popular accountability. For instance, scholars have found evidence that state supreme court justices are held accountable for criminal issues, notably increases in the murder rate (Hall Reference Hall2001; Brace and Boyea Reference Brace and Boyea2008). Moreover, with “new-style” judicial campaigns, judges subject to popular vote are potentially held accountable for any of a variety of decisions (Canes-Wrone, Clark, and Park Reference Canes-Wrone and Clark2012; Canes-Wrone et al. Reference Canes-Wrone, Clark and Kelly2014). All the while, quality challengers take into account the vulnerability of incumbent judges in choosing whether to run (Hall and Bonneau Reference Hall and Bonneau2006).

Given the electoral vulnerability and accountability of these elected judges, it is perhaps not surprising that, in noneconomic issue areas, scholars have consistently found evidence that elected judges react to public opinion within the state (Hall Reference Hall1995, Reference Hall2001; Brace and Hall Reference Brace and Gann Hall1997). Not only do these judges respond to public opinion, but their behavior varies across their electoral vulnerability (Hall Reference Hall2001; Huber and Gordon Reference Huber and Gordon2004; Brace and Boyea Reference Brace and Boyea2008). For judges facing reelection, factors beyond ideology and legal constraints exert influence on judicial decision making.

This extensive array of high-quality research has clear implications for the rule of law and conceptions of justice and suggests the critical importance of institutional designs. But while much has been learned, no attention has been given to the effect of the economy. The concern restraining scholarly inquiry is the visibility, or lack thereof, of state supreme court economic cases. To wit, Hall (2001) postulates that justices would not be held accountable for economic decisions because citizens are not aware of state supreme court decisions in economic cases. In other electoral contexts, numerous studies have documented that the effect of the economy on electoral outcomes is contingent on voters attributing blame to the incumbent party (e.g., Powell and Whitten 1993). Thus, this argument holds that if voters are unaware of the courts’ role in economic cases, it may be unlikely for voters to deem the courts responsible for economic conditions and thus to vote them out. In all, state court decisions lack sufficient visibility to encourage any change in voting behavior across electoral contexts for state supreme court judges.

Responding to Changes in the Economy

The dearth of attention paid to economic factors influencing judges stands in stark contrast to research in other venues. A rich empirical research tradition has documented the relationship between the state of the economy and the way political actors behave. There is a general consensus among scholars and political commentators that the economy matters for citizens when they step into the voting booth. With changes in the economic health of the country come shifts in support for Congress (Kramer Reference Kramer1971; Kinder and Kiewiet Reference Kinder and Kiewiet1981; Conover Reference Conover1985; Lockerbie Reference Lockerbie1991) and shifts in support for the president and the president’s party (Kernell Reference Kernell1978; Hibbs Reference Hibbs1982; Markus Reference Markus1988, Reference Markus1992; Alvarez and Nagler Reference Alvarez and Nagler1995, Reference Alvarez and Nagler1998). In all, there is little debate that the economy matters in the voting decisions of citizens.

These same dynamics have been observed at the state level. Building on insights gleaned at the national level, research by Peltzman (1987) suggested that voters also held governors accountable for changes in the economic climate. While there has been debate over whether or not national or local economic factors matter (e.g., Atkeson and Partin 1995, 1998; Niemi, Stanley, and Vogel 1995; Carsey and Wright 1998a, 1998b), scholars are generally in agreement that, at the state level, economic factors matter in citizen voting decisions.

It is unsurprising, therefore, that elected officials are attuned to the state of the economy and adjust their behavior in light of the sentiment of their constituents. Extensive evidence supports the responsiveness of elected officials to the concerns and ideology of voters (Miller and Stokes 1963; Erikson, Wright, and McIver 1993; Stimson, Mackuen, and Erikson 1995; Erikson, Mackuen, and Stimson 2002; Jones, Larsen-Price, and Wilkerson 2009). Elected officials, driven by a desire for reelection (Downs 1957), are therefore attentive to the state of the economy. Depending on their proximity to elections, elected officials may respond to the economy differently (Persson and Tabellini 2000). Overall, for nearly all actors involved in governance, there is evidence of changes in behavior in response to changes in the economy.

Despite the breadth of research suggesting that the economy matters for political actors, little attention had been paid to the potential influence of the economy on judges. By incorporating the state of the economy, Brennan et al. (2009a, 2009b) propose an approach to judicial decision making for a specific category of cases, those dealing with economic issues, that postulates a theoretically motivated avenue for the economy to influence judicial decisions. This economic theory of judging begins with the straightforward claim that justices prefer “national prosperity to an economy plagued by high unemployment, high inflation, and low productivity” (Staudt and He 2010, 91). As rational actors, justices “seek to advance their economic preferences through their Article III decision-making power” (92). In other words, the theory makes the relatively uncontroversial assumption that justices prefer a prosperous national economy to the alternative. Accordingly, while justices cannot “create” economic policy, they can use their privileged position as the decision maker in cases challenging various provisions of law to promote policies that they believe enhance growth while deterring policy when it appears that policy makers have ineptly handled the economy.

Under this macroeconomic theory of the Court, justices do not need to be skilled in economics. Rather, they need only have a preference for a prosperous economy, be able to interpret whether the economy is expanding or contracting, and have some conception of whether a downturn is typical, such as a momentary slowing in the business cycle, or atypical, such as a deep recession or depression (Staudt and He 2010). Insofar as this is true, then when the economy is expanding, the Court believes that policy makers are performing well and will vote in favor of the government. When the economy contracts, however, the justices will have one of two reactions. If the contraction is short-term or minor, it is interpreted by the Court as reflective of poor policy-making choices. Thus, the Court will vote more often against the government’s policy position. However, when the economy enters an atypical recession or depression, the Court will seek to work with policy makers to encourage an economic turnaround.Footnote ¹ In sum, the macroeconomic theory of judicial behavior holds that unless there is a depression or recession, justices of the US Supreme Court use the economy as a signal of economic competence and enable or stymie policy makers according to their perceived success or failure.

The theory and empirical support for the theory have faced a myriad of criticisms (Baker, Feibelman, and Marshall 2009; Young and Blondel 2009). To start, critics argue that bias could arise in these models from selection effects, as during different periods of economic success or decline there may be different pools of cases before the Supreme Court. This selection bias could be driven either by changes in the litigants bringing the cases or by changes in decision making on certiorari. In either case, the justices may be hearing fundamentally different cases during different economic periods, a process that could drive the observed change in voting patterns. Beyond selection effects, the empirical evidence provided for the theory has come under criticism for not conclusively demonstrating whether the Court’s response is deliberately intended to affect the economy (Young and Blondel 2009).Footnote ² As one alternative explanation of the observed behavior, critics point out that the state of the economy may actually be legally relevant to the decision at hand in regulation cases. Finally, critics have challenged the underlying theory that judicial behavior will vary like that of voters or elected officials. In suggesting future refinements, these critics suggest looking in other venues for evidence that macroeconomic decision making occurs, particularly in lower federal courts and in state courts.

Analysis

To assemble the data, I begin with the State Supreme Court Data Project, a collection of variables gathered on a sample of state supreme court cases decided from 1995 through 1998 across all 50 states (Brace 2007). In order to test for macroeconomic judicial behavior, the data set was reduced to solely cases dealing with economic issues in which the government was a party.Footnote ³ The final database consists of 201 cases, featuring 435 unique justices and covering all 50 state supreme courts. The database is organized into a cross-sectional data set with each observation representing an individual justice vote.

The dependent variable is a justice vote, with 1 indicating that the state supreme court justice voted for the state government in an economic case and votes not in favor of the government coded as 0. According to the macroeconomic theory, the probability of a state supreme court justice voting in favor of the government’s position in this subset of cases dealing with economic issues should decrease as the economy worsens, with the status of the economy serving as a signal of the policy-making competence of the incumbent government.

To test whether state court judges respond to the economy, I include an indicator of the status of the state economy. To begin, I use the change in the employment rate, which translates nicely to the state level, where state employment varies greatly from state to state and where the data are widely available through the US Department of Labor’s Bureau of Labor Statistics.Footnote ⁴ Because a month-to-month change in the employment rate may ignore shifts in the state of the economy across periods beyond 1 month, I estimate models with five different indicators of shifts in the employment rate, based on changes in the monthly state employment across different periods of time. Specifically, I estimated the difference in the employment rate at time t and time t − i, for lags of 1, 2, 3, 6, and 12 months. Because the value is calculated as a difference, interpretation requires a brief explanation. Specifically, positive values indicate that the economy has worsened (e.g., if the employment rate has decreased from last month to this month), while negative values indicate that the economy has improved (the employment rate has increased).

The variety of selection mechanisms used to determine the makeup of state supreme courts means that state supreme court justices face varying levels of accountability with voters. This variation, as detailed above, matters for the decision making of state supreme court justices on criminal issues but has not been documented as mattering for decision making on cases involving government regulation of the economy. With the importance of the economy on reelection decisions in other political spheres, there is considerable reason to suspect that electoral vulnerability also matters to judges on state courts of last resort.

Therefore, to capture the electoral accountability of state supreme court justices, I include a dichotomous variable indicating whether or not the judge is subject to any election. This variable takes on the value of 1 if the judge is subject to a retention, nonpartisan, or partisan election. All judges subject to election are coded as such, in accordance with previous research showing that all elected judges are subject to relatively high levels of electoral defeat and partisan pressures (Hall 2001).Footnote ⁵ If the judge is subject instead to solely reconfirmation via either the governor, the legislature, or some combination, this variable takes on the value 0.

Most importantly, I include an interaction term for the change in the employment rate (Δunemployment) and selection systems in order to account for differences in responsiveness conditional on the electoral system. The evidence above indicated that unelected justices of the US Supreme Court were more likely to oppose the government when the economy was performing poorly during typical downturns. With the economy of the nation not in any atypical cycle during the period under study (1995–98), Supreme Court–level findings suggest that unelected state supreme court justices would vote more (less) often against the government during downturns (upswings).

I expect that elected justices, because of their concerns over electoral accountability, are even more vociferous in their opposition to the perceived failures of the government. Elected judges have the greatest incentive to encourage economic growth, insofar as they believe that economic downturns decrease their likelihood of reelection. In other words, any downturn in the economy is worthy of an atypical reaction for the elected state supreme court justice. Judges not subject to election, on the other hand, are still likely to be subject to gubernatorial or legislative reconfirmation. They are most likely, therefore, to support the policies of the incumbent government in an effort to secure reconfirmation.

While the macroeconomic theory suggests that such a dynamic may exist in the context of state courts, the theory, as detailed above, is also subject to continued criticism. An additional benefit of this analysis stems directly from the inclusion of the interaction term for judicial selection systems. To wit, if selection effects or legal relevance drives the macroeconomic behavior of judges, as some critics hypothesize, or if elected judges simply perceive no electoral accountability for their actions, then we are unlikely to observe a difference across electoral contexts. If selection effects or the legal relevance of the economy drives the empirical results supporting the macroeconomic behavior, as critics have persuasively argued, there is no reason to expect an observable difference between states using different judicial selection mechanisms. In other words, to the extent that these are biasing results, the bias induced by selection effects or legal relevancy would need to correlate with both electoral contexts and economic indicators and would need to do so in ways also consistent with the macroeconomic theory of judicial behavior.

With decades of research suggesting the importance of ideological factors for judicial behavior (e.g., Pritchett 1948; Segal and Spaeth 2002), I include an independent variable accounting for the ideological predispositions of the justices. Of the many options, I use the Brace, Langer, and Hall (2000) measure of state supreme court justice ideology, PAJID.Footnote ⁶ The authors derive PAJID ideology scores by accounting for the party affiliation of state elites and the ideology of the electorate at the time the judge assumes his or her seat. Scores on the measure range from 1 to 100, with 1 indicating conservative and 100 indicating liberal. Here, I expect that Republicans will vote less often in favor of the government position on economic regulation than Democrats. I also account for the differential success rate for the government between cases in which they are petitioner and in which they are appellant by including a dichotomous variable indicating if the government was the petitioner. For this variable, 1 indicates that the government is the petitioner; I expect the coefficient estimate to be positive in line with prior research on the US Supreme Court (Brennan et al. 2009b). Finally, I include a variable to account for party agreement between the judge and the government involved in the litigation. State supreme court justices addressing cases featuring policies supported by members of their own party are likely to be more receptive to those arguments. To address this, I include a dichotomous variable, with 1 indicating that the voting judge is of the same party as the state governor and 0 indicating otherwise.

With the variables and expectations in place, I turn to the estimation. The voting variable exhibits a large amount of variance across state contexts, necessitating statistical consideration but also allowing for analysis of state-specific influences on the likelihood of a progovernment vote. In order to account for this variance, the model estimation again utilizes a mixed-effects approach that allows for state-level variation but borrows on the observations available in other states in order to generate estimates near a standard level.

I estimate a mixed-effects logit model (see Bailey and Alimadhi 2007; Imai, King, and Lau 2007), which, in this analysis, allows me to group effects by state and by individual court cases. I allow random-effect intercepts across both of these groupings, which accounts for the variation in state-specific receptivity to government regulation and case-specific factors, which may affect justices’ votes. My approach avoids wildly inconsistent and different coefficient estimates that would arise from no-pooling estimates while also accounting for state-level influences on judicial decisions not included in the model. For these very reasons, estimation of mixed-effects models is optimal in analyses with small sample sizes grouped by common variables, such as the states or cases used in this study (Gelman and Hill 2007; Lax and Phillips 2009). Mixed-effects logit thus offers a robust opportunity for estimating the relationship between the independent variables and the likelihood of a vote in favor of the government.

Discussion and Concluding Thoughts

Taking up the call of some legal scholars for analysis of judicial behavior in the states (Hall 1995) and the call of other legal scholars (Young and Blondel 2009) for greater and more rigorous analysis of the macroeconomic theory of judging, this article has sought to understand if and how macroeconomic judicial behavior operates in the varied institutional contexts of state courts. The analysis above, documenting evidence of variations in the macroeconomic theory of judicial behavior across state contexts, adds to the extensive and ongoing literature (e.g., Brace and Boyea 2008; Canes-Wrone et al. 2012) documenting the importance of state judicial selection and retention mechanisms. As state economies grow or struggle, judges respond differently according to their election or appointment status.

The research above thus offers substantial enhancements to our understanding of judicial decision making in economic cases, the macroeconomic theory of judicial behavior, and judicial selection. As in prior research in other issue areas before the courts, I find that the electoral vulnerability of state court judges matters for their decision making. Unelected state supreme court justices defer more often, on average, to the government when the economy is struggling and less often when it is doing well, while elected judges in state courts respond to short-term downward shifts in the economy by opposing the government more frequently. Additional corroborative support is seen as more long-term declines in the economy correlate with increased deference to the government in economic court cases, consistent with the macroeconomic theory’s holding that the severity of crisis could lead to additional teamwork.

The above findings do, of course, pose the question of why elected state supreme court justices respond to short-term changes in the economy in the same way as the unelected US Supreme Court justices. This “why?” question is especially difficult to answer but may speak to initial conjectures in Brennan et al. (2009b) that Supreme Court justices feel a responsibility, or accountability to the public, for the state of the economy. Across state court analyses, I have documented evidence that judicial support for government regulation varies according to whom the judges believe they are dependent on for continued tenure. Judges on state courts of last resort who are facing elections, and are thus accountable to the public, respond to economic declines by supporting the government less. On the other hand, judges who face reappointment are accountable to other institutions and thus support the government’s position more often as the economy struggles. This suggests that Supreme Court justices, insulated from institutional reappointment, lean toward accountability to the public without the direct election mechanism seen in state courts.

Moving forward, future research could explore the role that within-court panel dynamics have in the decision-making process in these cases. In this study, I have controlled for specific panels by utilizing mixed effects across cases. However, a large body of literature within judicial decision making suggests that panel-specific voting patterns may shift according to the other members of the court, their positions and expertise, and strategic concerns (Caldeira and Zorn 1998; Kastellec 2007; Boyd, Epstein, and Martin 2010). With judges facing different times to retention decisions within panels, there is the possibility of panel-specific dynamics not covered here.

In all, though, this analysis provides evidence of new influences on the voting of judges on state courts of last resort and offers substantial encouragement for scholars interested in macroeconomic explanations for judicial behavior. Moreover, the research suggests additional evidence for the importance of judicial selection mechanisms on judicial decision making in the area of economic regulation. With ongoing debate on the viability of different selection mechanisms, the research suggests the important influences of those decisions on judicial outcomes.