SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Judicial Independence
  5. Recess Appointees and Constrained Decision Making
  6. Data, Model, and Analyses
  7. Conclusion
  8. References
  9. Appendix
  10. Biographies

In this article, we directly test the presence of judicial independence by examining judicial recess appointees who have later been confirmed by the Senate to full-time Article III judicial positions. Specifically, we compare the votes of recess-appointed courts of appeals judges during their temporary appointment tenure with a similar period following Senate confirmation. We find substantial differences in pre- and postconfirmation voting, suggesting that the structural protections of the Constitution provide judges a certain amount of independence.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Judicial Independence
  5. Recess Appointees and Constrained Decision Making
  6. Data, Model, and Analyses
  7. Conclusion
  8. References
  9. Appendix
  10. Biographies

Judicial independence is an essential component of our legal system. For example, former Supreme Court Justice Sandra Day O'Connor made an impassioned case for the importance of judicial independence when she argued “[l]aws properly adopted by a country must be applied, interpreted and enforced by an independent judiciary” (2006). To promote independence, Article III of the Constitution guarantees judges life tenure and an undiminished salary. As Hamilton explained in Federalist 78:

Periodical appointments, however regulated, or by whomsoever made, would, in some way or other, be fatal to [the courts'] necessary independence. If the power of making them was committed either to the executive or legislature, there would be danger of an improper complaisance to the branch which possessed it; if to both, there would be an unwillingness to hazard the displeasure of either. (Rossiter 1961, 471)

Thus, these structural protections were intended to insulate judges from the political process and allow them to make decisions “without regard to the preferences of politically accountable officials” (Rosenberg 1992, 371), and the assumption is that these provisions create an independent judiciary. To date, however, no one has empirically tested this assumption because most of the previous literature on the decision making of the federal judiciary examines the behavior of judges who already have Article III protections (see, e.g., Sala and Spriggs 2004; Bergara, Richman, and Spiller 2003; Cross and Nelson 2001; Ferejohn 1999; Segal 1997; Spiller and Gely 1992). Thus, previous studies have not been able to ascertain whether Article III does, in fact, provide judges with any degree of independence, instead assuming judicial independence exists and then proceeding to test under what conditions Article III judges are constrained. To remedy this state of affairs, we provide an innovative test of judicial independence. Specifically, we compare the votes of recess-appointed courts of appeals judges during their temporary appointment tenure with a similar period following Senate confirmation. Examining recess appointees who have ultimately been confirmed offers a unique opportunity to isolate decision making with constitutional protections and decision making without. A recess appointee who has not been confirmed is operating under conditions of complete (or almost complete) constraint. If there are differences between their voting behavior before and after confirmation, this supports the assertion of judicial independence.

While there have been previous examinations of recess appointments and voting (Graves, Howard, and Corley 2009), as we explain in more detail later on in this manuscript, among other things, this manuscript is the first to present a theoretical justification for differences in voting pre- and postconfirmation of recess appointees and the first to offer a control model to test for acclimation effects.

Judicial Independence

  1. Top of page
  2. Abstract
  3. Introduction
  4. Judicial Independence
  5. Recess Appointees and Constrained Decision Making
  6. Data, Model, and Analyses
  7. Conclusion
  8. References
  9. Appendix
  10. Biographies

We do not argue that judges are completely independent after they have been confirmed. The other branches have control over jurisdiction, court creation, appointment, enforcement of court rulings, appropriations for the operation of the courts, and can impeach judges. Thus, there are many ways in which Congress and the president can exert influence over the federal judiciary. This has led many scholars to test the separation of powers (SOP) model, which holds that judges are constrained in their decisions by the policy preferences of Congress and/or the president (see Bergara, Richman, and Spiller 2003; Cross and Nelson 2001; Ferejohn 1999; Spiller and Gely 1992, finding support for the SOP model; but see Sala and Spriggs 2004; Segal 1997, finding no support for the SOP model). While avoiding this debate, we do argue that the structural protections of the Constitution provide some degree of judicial independence to the federal judiciary. Specifically, we argue that while there might be constraint imposed by the other branches of government, independence means that there is at least some freedom to impose voting preferences inconsistent with the preferences of the electoral branches of government.

We recognize that there are different conceptions of judicial independence. For example, one particularly expansive conception of judicial independence claims that not only are judges able to make decisions free of influence from coordinate branches of government, but they are free also from the private sector and even from within the judiciary (see Kornhauser 2002). Let us emphasize that our definition of judicial independence in the context of our examination of votes is consistent with SOP expectations. We mean “judges who are free from potential domination by other branches of government” (Rosenberg 1992, 370). Specifically, judicial independence means judges can make decisions that are not affected by political pressure from outside the judiciary (Clark 2011). Since Article III provides protections from undue external pressure, we define judicial independence as a judge being independent of the executive and the legislative branches, meaning that their decisions must be reached without regard for the political preferences of the members of these branches. Thus, we argue that an independent judge will vote attitudinally, that is, resolve cases by the policy preferences of the individual judge rather than the preferences of Congress or the president.

According to the attitudinal model of judicial decision making, judges make decisions according to the facts of the case vis-à-vis the ideological attitudes and values of the judges (Segal and Spaeth 2002). Although the attitudinal model originally applied only to the U.S. Supreme Court, systematic evidence demonstrates that ideological preferences of judges influence the behavior on the U.S. courts of appeals as well (see, e.g., Hettinger, Lindquist, and Martinek 2006; Benesh 2002; Songer, Segal, and Cameron 1994). If judges are truly independent, they will vote according to their own preferences rather than the preferences of the Senate or the president.

We recognize that judicial independence is a continuous (“more-or-less”) concept rather than a dichotomous (“yes-or-no”) concept. Accordingly, we do not argue that before confirmation recess appointees have absolutely no independence and after confirmation they have complete independence. In the U.S. system of separation of powers, there are inherent limits to the true independence of courts. We do argue, however, since recess appointees do not yet have the protections of Article III, and thus are subject to more constraints in their decision making than judges with Article III protections, predictable changes in voting behavior after confirmation is evidence of a certain amount of judicial independence.

Recess Appointees and Constrained Decision Making

  1. Top of page
  2. Abstract
  3. Introduction
  4. Judicial Independence
  5. Recess Appointees and Constrained Decision Making
  6. Data, Model, and Analyses
  7. Conclusion
  8. References
  9. Appendix
  10. Biographies

When a recess appointee hears a case, he does not have a permanent appointment, and whether the appointee receives tenure is still contingent upon renomination by the president and confirmation by the Senate (Mayton 2004). Thus, a recess appointee does not have the structural protections provided by Article III, and, consequently, he is theoretically constrained in his ability to render decisions “unaffected, or at least minimally affected, by the strength of partisan positions among members of the other branches” (Rosenberg 1992, 371).

Jefferson B. Fordham, dean of the University of Pennsylvania Law School, remarked that a recess appointee “is serving under the overhang of Senate consideration of a nomination, which is not in harmony with the constitutional policy of judiciary independence” (quoted in Fisher 2002, 125). The Senate, of course, is not the only actor the recess-appointed judge might have to please. In a report written for the Congressional Research Service, Fisher (2005, 4) notes “(a) recess judge might also have to keep one eye out for the reaction of the White House, which would review decisions issued during the recess period to determine whether they justified nomination of the judge to a lifetime appointment.” Such concerns led the House Judiciary Committee to issue a report in 1959 questioning the independence of judges sitting via recess appointment from political influence (Report 1959).

Herz (2005) employs an analogy to illustrate the difference between the situation of a judge sitting temporarily via recess appointment and that of a judge holding permanent commission. “These circumstances” he argues, “put the recess appointee in something of the same position as a law professor on a ‘look-see visit’; his or her job becomes one extended interview. These circumstances are utterly at odds with the commitment to judicial independence reflected in Article III's good behavior clause and salary protections” (Herz 2005, 450).

In fact, a recess appointee is analogous to a judge who is facing reelection. Studies have found that, under certain conditions, elected judges vote strategically to minimize electoral opposition (see Hall 1992, 1987). Specifically, Hall found that “state supreme court justices who have views contrary to those of the voters and the court majority, and who face competitive electoral conditions will vote with the court majority instead of dissenting on politically volatile issues” (Hall 1992, 428).

While recess-appointed judges will not have to confront the voting public electorate, existing research on court nominations, judicial behavior, and judicial voting provides insight into how a recess appointee will behave. We assume that a recess appointee wants to be confirmed, and, accordingly, the “electorate” that the nominee would care about is the president and the Senate, since the president must still renominate the recess appointee, and the Senate must confirm him. Research has shown that the median ideology of the Senate (Moraski and Shipan 1999) or even the partisan make up of the Senate (Epstein and Segal 2005) can be critical to confirmation as the president and Senate clash over the nominee, each seeking an ideological advantage in separation of powers struggles (Yates and Whitford 1998). Thus, it is likely that the more liberal or Democratic the Senate, the greater the likelihood that a recess appointee's votes would strategically bend in a liberal direction. Moreover, we expect that a recess appointee would be more likely to vote in a liberal direction if appointed by a liberal president, even controlling for the judge's own ideology, and to cast conservative votes if appointed by a conservative president. In contrast, we anticipate that a confirmed judge would be freer to depart ideologically from the president and the Senate.

A recent book by Graves and Howard (2009) examined all federal judicial recess appointees including appellate (Graves, Howard, and Corley 2009) and Supreme Court recess appointments. While much of the book is beyond the scope of this article, a few of their observations provide emphasis for this project. First, their data show that for both appellate court (Chapter Four) and Supreme Court recess nominees (Chapter Five), the workload for similar periods of time pre- and postconfirmation stays roughly the same. In other words, there is no reason to expect any systematic difference in case assignment and workload. This is confirmed to an extent by Abraham in his classic Justices and Presidents (1992). In his description of the recess appointment of Earl Warren, for example, Abraham notes that Eisenhower expected Warren to be confirmed and that Warren, immediately upon his appointment, took a leading role in the Court. He led discussions, made assignments, and voted with as much frequency before his confirmation as he did after confirmation.

Graves, Howard, and Corley (2009) do find evidence of differences in appellate court voting pre- and postconfirmation. However, given that the focus was on the differences in voting by the judges' pre- and postconfirmation, the authors did not focus on judicial independence and, accordingly, did not develop a theory to explain why one would expect vote differences pre- and postconfirmation. That is, the authors failed to explain the meaning of independence and how it relates to the confirmation process controlled by the president and senate, and how the constraints imposed by elected officials would lead to voting difference between the independent judge and the constrained judge.

This change in emphasis also leads to modeling changes. Unlike the Graves, Howard, and Corley (2009) chapter which used three atheoretical models, this article, because of the theory development and hypotheses, includes a new key variable that interacts partisan senate control with judicial ideology. This interactive variable tests a specific form of political constraint on recess-appointed judges with greater nuance than in the chapter, which does not address the mechanisms of partisan effects absent tenure protection.

Finally, in this article we present the results of a control model to examine the possibility that the interactive model results are a function of acclimation differences or a freshman or “acclimation” effect of voting. To determine if this is so, we collected data for a control group of appellate court judges chosen to match the recess-appointed judges as closely as possible, including same appointing presidents, same or similar circuit seats, and party identification, as well as time period. The data cover an equivalent number of days on the bench as the comparable recessed judges, and we estimated the same interactive model.

Based on our theory of judicial independence, we argue that while there might be constraint imposed by the other branches of government, independence means that there is at least some freedom to impose voting preferences inconsistent with the preferences of the electoral branches of government. Thus, we focus on the voting behavior of the judges. This leads to the following hypotheses:

  • Hypothesis 1: Judicial votes made after confirmation, with the protections of Article III in place, are more likely to be guided by the ideology of the judge.
  • Hypothesis 2: Judicial ideology will matter more after confirmation than before, while Senate ideology/control will matter more before confirmation than after. Thus, votes cast by a judge without the protection of Article III are more likely to be liberal when the Senate is controlled by Democrats, especially if the judge is conservative.
  • Hypothesis 3: Votes cast by a judge with Article III protections are more likely to depart from the ideological position of the president.

Although circuit court judges do enjoy lifetime tenure and protection from diminution in their salaries, their decisions are subject to review by a higher court. Evidence suggests that lower court decisions—specifically those of the federal circuit courts of appeals—are highly, although not perfectly, congruent with decisions of the U.S. Supreme Court (Brent 1999; Songer, Segal, and Cameron, 1994). Therefore, ideological voting could be affected by the preferences of the judicial hierarchy as well as those of the judge. Whether compliance of lower court judges with higher court decisions results from aversion to reversal (McNollgast 1995) or from a legal preference for adherence to doctrine (Feeley and Rubin 1998), we control for this potential influence.

Similarly, cases appealed to the courts of appeals are generally heard by a three-judge panel. The panel is charged with deciding the case before it in accordance with the law of its circuit. Thus, we would expect the votes of the judges on the panel to be the same as the votes of the full circuit. Although there may be outlier panels that decide cases out of line with the full court's position, en banc review provides a form of monitoring and there is some evidence that a panel ruling out of line with the full court is more likely to be reviewed en banc (see Clark 2008; George 1998; Note 1989). Thus, we control for the ideology of the circuit.

Data, Model, and Analyses

  1. Top of page
  2. Abstract
  3. Introduction
  4. Judicial Independence
  5. Recess Appointees and Constrained Decision Making
  6. Data, Model, and Analyses
  7. Conclusion
  8. References
  9. Appendix
  10. Biographies

Data

Our first step was to select a collection of votes for valid comparison. We gathered data on each of the judges' reported decisions from the date they arrived on the bench to the date they were confirmed by the Senate, decisions that could potentially have been affected by the temporary nature of their appointment. We also collected data for a similar number of case-votes from the same judges for the same amount of time following their confirmation. That is, if the judge sat as a recess appointee for 100 days prior to confirmation, we gathered vote data from the date of the recess appointment to 100 days following confirmation. Votes were considered “before” or “after” confirmation by comparing the report date of the case with the date of Senate confirmation.1 The average number of preconfirmation days was 197 days, or about 6.5 months, with a standard deviation of ninety-nine days.

In the postwar era, sixteen judges have received recess appointments to the circuit appellate courts. For our purposes, we modeled the votes of the fourteen appointees for whom we have voting data and ideological measures.2 Excluding cases for which data were not available resulted in a data set of 996 votes, 375 cast while recess appointed and 621 made while sitting as permanent Article III judges.

In addition to the pre- and postconfirmation votes of the judges, we collected data to test our hypotheses on the nature and direction of the voting. Specifically, our dependent variable was the ideological direction of the vote. Following the coding scheme developed by Songer and Kuersten and Haire (Kuersten and Haire 2008; Songer 1998) for the U.S. Courts of Appeals Database, we coded a vote as liberal (1) or conservative (0). Votes that could not be classified as either conservative or liberal under the Songer rules (Kuersten and Haire 2008; Songer 1998) were excluded from the analysis. We collected case-level data for several independent variables, including whether or not a case was subsequently petitioned for certiorari review (1, 0 otherwise) and whether the case was a Civil Liberties/Civil Rights case (1, 0 otherwise) since there is evidence that highly salient cases affect the extent to which judges vote ideologically (see Spaeth and Segal 1999; Segal 1986).

We then collected ideology measures for relevant actors and institutions. The standard contemporary method of identifying political actors in the same ideological space, allowing cardinal distance comparisons, is the Basic or Common Space introduced by Poole (2003), extended to include federal lower court judges by Giles, Hettinger, and Peppers (2002, hereafter GHP) and to the U.S. Supreme Court by Epstein et al. (2007). We used these scores to place the president, home state senators, judicial circuits, and Supreme Court in comparable ideological space.

Placing the recess-appointed judges in that space proved a more difficult challenge. As noted above, GHP (2001, 2002) introduce a method for measuring the ideology of lower federal court judges that makes use of the norm of senatorial courtesy. GHP attribute to lower federal court judges the ideology of the political actors involved in the judicial selection process. For judges selected under conditions when senatorial courtesy would apply, when the state in which the judge would sit has one or more senators of the president's party, they impute the ideology score, or the mean if more than one, of the relevant senator(s). If senatorial courtesy does not apply, the ideology score of the president is given to the judge. Although GHP (2002) confirm the validity and performance of their measure for lower court judges generally, they do not investigate whether the scores are valid for judges selected under a heterodox method such as recess appointment. If the utility of the GHP scores lies in large part with their use of the appointment process to identify the operation of political choice on judicial ideology, then we must question whether the scores will be as accurate for judges selected under a different process. In addition, the GHP procedure results for many cases in attributing the ideology of the appointing president to the nominee, which makes it difficult to assess whether judges depart ideologically from the president.

Fortunately, another method is available to place recess appointees in the Common Space without relying on the selection process. Many scholars have attempted to place actors across institutions in shared ideological space with the use of “bridging observations,” actors serving in both institutions. Poole (2003) for instance, uses instances of members of the House serving in the Senate and presidents serving in the House, Senate, or both to create the comparable Common Space measures of members of Congress and presidents. Likewise, Nixon (2004) uses legislators with Common Space scores who served in administrative agencies to place Congress and agency appointees in comparable ideological space. Howard and Nixon (2003) exploit the fact that many members of Congress subsequently became federal court judges (sixty-three judges in all) to estimate a predictive model of first dimension Common Space scores using background characteristics of the judges (see also Howard 2007, 2008). Initial evidence for consistency can be shown from Segal-Cover scores for Supreme Court justices.3 While the number of Supreme Court justices with prior congressional experience is limited (four—Minton, Burton, Vinson, and Black), the correlation between their floor voting and Segal-Cover's ideology measure is extremely strong (Pearson = −.946***). For the thirteen circuit court judges for whom a legislative career exists, the correlation between their floor voting and bench behavior (based on Songer's [1998] and Kuersten and Haire's [2008] appeals courts databases) is also very strong (Pearson = −.621, significant at an alpha level of .001). This is the method employed to generate ideology scores for the recess-appointed appeals court judges. The resulting scores correlate with the Judicial Common Space scores (those devised by the GHP method) at 0.84, but several of the scores are considerably different, typically less extreme, than those imputed by the GHP method.

With ideology scores for all relevant actors, we then calculated the ideological difference between each recess appointee and the median of the circuit within which he sat and the contemporaneous president and Supreme Court. The Common Space scores increase in conservatism—movement from negative to positive on the real line indicates increasingly conservative ideology—and the difference scores were calculated by subtracting the circuit, president, or Supreme Court from the appointee's ideology. Thus, the differences measure how conservative (+) or liberal (-) the judge is relative to the actor in question.

We present descriptive statistics in Table 1.

Table 1. Descriptive Statistics, and Expected Effects—Pooled, Before and After Confirmation
VariableExpected EffectsPooledBefore ConfirmationAfter Confirmation
Effect BeforeEffect AfterMeanStd. Dev.MeanStd. Dev.MeanStd. Dev.
Liberal VoteNANA0.4560.500.4880.500.4340.50
Ideology of Judge (Conservatism)None−0.0050.29−0.0280.290.0010.28
Democratic Senate (Separation of Powers)+None0.7370.440.7520.430.7240.45
Democratic Senate * Ideology (SOP)+None−0.1040.18−0.1200.18−0.0940.18
Ideological Difference from CircuitNone+0.0160.24−0.0010.230.0260.24
Ideological Difference from Supreme CourtNone+−0.0320.33−0.0580.33−0.0130.33
Ideological Difference from PresidentNone0.1190.210.1260.200.1140.22
Petitioned for CertiorariNone0.0680.250.0640.250.0710.26
Civil Liberties / Rights IssueNone0.3590.480.3070.460.3910.49
 N = 996N = 375N = 621

Table 1 splits the data into pre- and postconfirmation samples but also shows the central tendency and variability for all of the observations combined. The proportion of liberal votes cast by these judges actually decreases slightly among the postconfirmation decisions, but the means and standard deviations of most other variables remains remarkably constant across periods. One notable difference is between the proportion of civil liberties and rights cases decided before and after confirmation.

The only interaction included in the basic specification, Democratic Senate multiplied by ideology, also yields an interesting change. The average ideology for all votes—pooled, before and after confirmation—is close to zero, indicating that the votes included are roughly balanced among liberal and conservative judges. However, when interacted with the indicator for Democratic control of the Senate, the average ideology of votes drops by about a tenth of a unit. Thus, judges' casting votes while the Democrats control the Senate are, on average, more liberal than the overall average. This is interesting because it belies the expectation that judges are more likely to be recess appointed when they are ideologically distant from the Senate. However, it also urges caution in the interpretation of party effects, because support for the situation in Hypothesis 2, conservative judges facing a liberal Senate and vice-versa, is small.

Another notable difference is that the number of votes within a similar period after confirmation is larger overall than the number of recess-period votes. However, reducing the postconfirmation periods to include only as many votes as were cast preconfirmation has no effect on the substance or significance of the results presented here. We include the reestimated model in the Appendix.

Model

The dependent variable is whether each recess-appointed judge's vote4 in each case is liberal (1) or conservative (0). Since our dependent variable is dichotomous, we estimate a logit model.

Four variables are included making use of the Common Space ideology scores described above. The ideologies of the recess appointees, estimated using the Howard and Nixon technique (2003), are entered directly. We also include ideological differences between the judge's score and the median of the circuit in which the judge is sitting, the contemporaneous Supreme Court median, and the contemporaneous president. As described above, these variables indicate how conservative the judge is compared to the other actor.

In addition, the model specification contains a variable indicating whether (1) or not (0) the Democratic Party controlled the Senate at the time of the judge's vote, and to detect differences in ideological voting based on the party whose favor they seek, judges' ideologies are conditioned on party control of the Senate. Finally, in order to capture the salience of certain cases, we included an indicator variable identifying cases that were petitioned to the Supreme Court after the court of appeals' decision. Although the certiorari petition follows the judge's vote, the presence or absence of such a petition can be seen as an indication of whether the litigants in the case were resourceful and sophisticated enough to pursue further appeals by seek review from the highest court, something a judge might discern while casting the vote. The last variable captures cases that raise civil liberties or rights issues. Both of these control variables should capture highly salient cases.

Analyses

Preconfirmation, we hypothesize that these judges will conform to the preferences of those who must act on that nomination, being more likely to vote liberally when the Senate is controlled by Democrats and when the sitting president is more liberal than the judge and less likely to vote according to their own personal policy preferences. Postconfirmation, we hypothesize that judges will be free of the constraint to conform to the preferences of the Senate and the ideology of the president, and instead vote in accordance with their own personal policy preferences.

Since our hypotheses are conditioned on whether or not the judge is voting preconfirmation or postconfirmation, we split the variables into two groups using a dichotomous “regime” variable indicating whether the case was before or after confirmation.5

The results of our analysis are presented in Table 26 and show dramatic evidence of pre- and postconfirmation voting. The most pronounced effect is the dramatic change in the relationship between ideology and vote choice before and after a judge is confirmed to a permanent judicial seat. Before confirmation, ideology has no discernible impact on votes, regardless of which party controls the Senate. After confirmation, however, the effect of ideology is substantially large and statistically significant regardless of which party has a Senate majority. In short, judges' votes exhibit much more independence after confirmation than before—to the extent that personal policy preferences are followed after confirmation and not before.

Table 2. Regime Logit Model of Liberal Vote—Recess Appointed JudgesThumbnail image of

We can learn more about the results graphically than through coefficients and test statistics. A graph of the effect across the range of ideology with a Democratic Senate is produced in Figure 1. The x-axis indicates the conservatism of the judge, from left to right. All of the remaining variables are fixed at their central values, means for continuous variables, and modes for binary variables. The graph presents the probabilities across the range of ideology with a 95 percent confidence interval after confirmation.7 As expected, after confirmation, the probability of a liberal vote decreases significantly when the judge is more conservative. Thus, after confirmation, the judge votes in accordance with his own personal policy preferences, whereas before confirmation, ideology evinces no discernible effect.

figure

Figure 1. Predicted Probability of Liberal Vote—by Judge's Ideology, Democratic Senate.

Download figure to PowerPoint

We find no evidence that Democratic control of the Senate held any influence on liberal voting by recess appointees before confirmation, however. Thus, Hypothesis 2 receives no support from the results. Democratic control of the Senate does have a statistically significant effect, however, after confirmation, and it is negative. Interpreted directly, the result suggests that judges are less likely to vote liberally after confirmation when the Democratic Party has a majority in the Senate. This is a curious result, not anticipated either by any expectations drawn from our theory of how recess appointees will behave or from any separation of powers literature, which would predict that any influence Congress may have on a judge would be to vote more consistently with the ideological direction of the party in power. A closer look, though, suggests that the result may be an artifact of the data-generating process. The Republican Party held a majority in the Senate in only ten of the Congresses between the end of WWII and 2006. As a result, only a handful of votes (six) in our data set were cast by a judge recess appointed by a Democratic president while the Republicans held the Senate, compared to 289 votes by Republican appointees facing a GOP Senate. The split of votes between Democratic and Republican appointees with a Democratic Senate is far better balanced, so if the ideology measures underestimate the ideological voting of appointees, it would result in overstating the conservatism of voting during Republican Senates and be captured by the variable for Democratic Senates.

The difference of the judge ideologically from the president postconfirmation produces a negative coefficient, as hypothesized, and is statistically significant. This means that judges are more likely to depart ideologically from the president after confirmation—they are less likely to vote liberally the more conservative they are than the president. Thus, judges are constrained by the preferences of the president preconfirmation (and their votes should conform more to the ideology of the president), but after confirmation, judges vote in accordance with their own personal policy preferences.

Interestingly, recess appointees do not appear to follow hierarchically superior precedent. There is no statistically significant effect on the votes of judges sitting via recess appointment relative to the ideology of the Supreme Court. However, the Supreme Court constrains judges' vote choices after they are confirmed by the Senate.

Specifically, ideological difference from the Supreme Court has a conditional positive effect, only postconfirmation, suggesting that after confirmation judges are more likely to vote liberally as the Supreme Court's relative liberalism (the degree to which the judge is more conservative than the Court) is increased, holding the ideology of the judge and other conditions constant. A chi-square test concludes that the postconfirmation coefficient is significantly different from its preconfirmation analog as well. This suggests that, after confirmation, these circuit court judges respond to the direction of the Supreme Court, whereas, before confirmation, the judges are indifferent to such judicial directives, perhaps because the judges are instead paying attention to the president and the Senate. The equivalent measure for the circuit median is not significant before or after confirmation, however, nor are they statistically distinct from each other.

Arguably, the “ideology” of the circuit and the Supreme Court indicate the state of the law that the judges are bound to apply. Thus, these judges seem unaffected by the law until they receive Senate confirmation. Another interpretation is that these results indicate the influences of the principal-agency relationships between appeals court judges, the circuits within which they sit, and the Supreme Court above them. Those influences are less pressing on the judges when the fate of their appointment is still in the hands of the president and Senate, but relevant after they gain the protections of Article III.

Finally, we examined the possibility that the model results are a function of acclimation differences. Literature shows that a “freshman effect” for justices of the Supreme Court (Hurwitz and Stefko 2004; Hagle 1993; Rubin and Melone 1988; Scheb and Ailshie 1985) might exist and that justices may tend to moderate their votes in their first few years on the court. Thus, what we see as evidence of independence might be nothing more than a freshman or “acclimation” effect of voting.

To determine if this is so, we collected data for a control group of appellate court judges. That is, we collected comparable data for five judges nominated, confirmed and appointed via the ordinary or nonrecess process: John C. Pickett (confirmation date, October 12, 1949), James Alger Fee (April 23, 1954), John Hastings (August 22, 1957), Jesse S. Bell (September 23, 1961) and Terrence L. O'Brien (April 15, 2002). These judges were chosen to match the recess-appointed judges as closely as possible, including same appointing presidents, same or similar circuit seats, and party identification, as well as time period. The data cover an equivalent number of days on the bench as the comparable recessed judges.

In Figure 2, the votes of these judges are plotted (markers are jittered to illustrate density) along with a local regression line to illustrate the trend in liberal voting. Although two of these judges demonstrate some volatility in the direction of their votes during these initial days, patterns are neither consistent nor shared. Two judges, Fee and O'Brien, experienced a change in control of the Senate during their first year on the bench, identified by the vertical line on their graphs, and in neither case is a change in their voting balance observed. Thus, there is no evidence of a freshmen effect.

figure

Figure 2. Liberal Voting by Judges Affirmed before Appointment—Contemporaneous with Recess-Appointed Judges.

Download figure to PowerPoint

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Judicial Independence
  5. Recess Appointees and Constrained Decision Making
  6. Data, Model, and Analyses
  7. Conclusion
  8. References
  9. Appendix
  10. Biographies

We recognize that in the modern era recess appointments are relatively few. The examination of recess appointee voting pre- and postconfirmation is interesting in and of itself, but the critical issue is the meaning of judicial independence. These recess appointees allow us to examine judicial independence via this natural experiment. Judicial independence is widely cited as a normative good and a normative goal in designing a judicial system. Judicial independence protects judges from undue interference. However, despite all the normative assertions and all the statements from scholars and judges calling for judicial independence, there has been, to date, little empirical proof of what that actually means.

In large part that is because it has been difficult to find direct evidence showing voting by the same judge who has been both independent and constrained by other actors in the political system. That is what we have attempted to do with this study. Although we recognize that our study is limited by the small number of recess-appointed judges, it is only by examining judicial recess appointees who have later been confirmed by the Senate to full-time Article III judicial positions that we can directly test the presence of judicial independence. Additionally, we did test whether the results are a function of acclimation differences, and we found that they were not.8 Thus, although we do acknowledge the limits of our study, we believe that the results allow us to gain insight into the structural protections of Article III.

Do the protections of lifetime tenure and an undiminished salary provide judges with independence? The results demonstrate that these judges did alter their voting behavior after confirmation. Judges sitting by temporary recess appointment do not vote according to their ideological predispositions. We contend these judges are concerned with the impression that their actions on the bench make on the president who must renominate and support them and the Senate that must vote to confirm them.

In contrast, we found that these same judges' voting behavior was very consistent with their ideology, measured independently of their own voting records, after they were confirmed. We also find that the judges were much more responsive in their voting behavior to the ideological direction of the circuit in which they sat and the U.S. Supreme Court following their permanent appointment than they were beforehand. In a sense, judges sitting temporarily, awaiting the approval of the political branches to keep their jobs, are not full members of the judicial branch during that time, but visiting jurists on a “look-see” extended interview, as Herz (2005) characterizes the situation. The important distinction, however, between these situations is that their colleagues have no role at all in deciding whether the visiting judge will receive an offer of permanent employment.

After confirmation, however, when the judge has the protections of Article III, the judge behaves independently, voting in accordance with his own personal policy preferences. Thus, we find that the structural protections of the Constitution provide judges a certain amount of independence.

The implications are clear for judicial design, both for individual states and other countries. Judges who do not enjoy the protections offered by something similar to Article III of the U.S. Constitution will tailor rulings, to some extent, to the preferences of the electorate, whether it is an electorate of voters or an electorate of the legislative branch of government.

However, the normative implications of our findings do not necessarily suggest that judicial independence is always a positive development. Support for judicial independence is usually premised on the idea that judges must rule free from political independence based on the law and facts of the case. Independence in our finding means that an unconstrained judge is free to follow personal policy preferences. This was the explicit concern of Justice O'Connor in the quote we cited to begin this article.

Based on our findings, judicial independence is not necessarily the same as ruling based on the law and facts. Independence means that judges are free to make opinions premised on ideology free from interference from the political system. It was precisely this concern that underscored the findings of Bonneau and Hall in their recent book on state courts, In Defense of Judicial Elections (2009). The authors argue that lack of independence in the form of judicial elections is not necessarily a normative wrong because it prevents a lack of accountability. As Hall notes in a later article, “we have not achieved an acceptable level of conceptual clarity in discussing the central values of accountability and independence” (2009, 284).

We take no position on this issue of whether or not judicial independence is a good or a bad thing. We do show, however, that a lifetime appointment does matter to a judge and does help determine the voting outcomes.

Notes
  1. 1

    Certainly, some votes classified as postconfirmation by this rule were actually cast before the Senate vote, but because the conference dates are not known for these cases and judges are always free to change their votes up until the report date, any earlier demarcation would be arbitrary and possibly wrong. Furthermore, a judge casting a vote shortly before the Senate is scheduled to vote, knowing that the case would not be reported until afterward, could easily be free of whatever pressures he had felt when he knew that his votes could be scrutinized by the president or Senate before receiving a permanent commission.

  2. 2

    They are David Bazelon, H. Nathan Swaim, David Fahy, George T. Washington, William H. Hastie, John A. Danaher, Carroll Hincks, Leonard P. Moore, Griffin Bell, Walter P. Gewin, Paul Hays, Thurgood Marshall, Roger L. Gregory, and William H. Pryor. The ideology metric used for this analysis, the Poole Basic or Common Space (1998), does not provide ideology scores for presidents before Eisenhower, but we extended our analysis back to the late 1940s by using Truman's Senate score for his ideology as president (Sala and Spriggs 2004). Elmer Schnackenberg and Walter Bastian were excluded because their short recess periods (eighty-four and seventy-three days, respectively) did not produce any preconfirmation votes. Charles Pickering, never confirmed by the Senate, could not be analyzed for the opposite reason. We recognize that this is a relatively small number of justices. However, examining some fourteen justices and their votes is not very different from many published Supreme Court analyses. There have been only thirty-seven members to the Supreme Court since 1937, and scholars often analyze the voting behavior of subsets such as the Warren Court or the Roberts Court. Thus, so much of successful Supreme Court research relies on a smaller number of justices than we do in our sample of fourteen.

  3. 3

    Segal-Cover scores are measures of the ideology of Supreme Court justices that are based on analysis of preconfirmation newspaper editorials regarding the nominations of each justice (Segal and Cover 1989).

  4. 4

    Our unit of analysis is the individual vote, and not the judge, due to differences in cases that will cause some votes to be more or less vulnerable to external influences. Differences between judges that are not accounted for by variables in the model could still be addressed as unmodeled heterogeneity using fixed effects. However, fixed effects are inappropriate here because ideology is constant within judges (and thus perfectly collinear with the fixed effects). Random effects lack this restriction, but results from the random effects estimator applied to the same model revert to the logit estimates because the variance of the random effects is virtually zero.

  5. 5

    In this formulation, cases have their observed values only for the covariates of the “regime” into which they fall; they are zero otherwise. So, a case that occurs before confirmation for a judge would be coded for the preconfirmation variables (the top half of the table) and the variables below would all be coded zero, while a case voted on after confirmation is coded zero for the first set of covariates and coded for the proper values in the second set. This “regime” model allows us to estimate separate marginal effects on the outcome for the covariates across conditions (before and after confirmation) without the complications of the traditional interaction model, because the modifying variable (confirmation) is discrete (Brambor, Clark, and Golder 2006; Wright 1976).

  6. 6

    In this model, hypothesis tests for individual variables compare the coefficient value to the null (zero) rather than to the complementary coefficient in the other regime. Thus, for example, the insignificant z-score for judge's ideology, Republican Senate preconfirmation, while the postconfirmation variable is significant, merely means that the former value cannot be distinguished from zero, but the latter can (at a conventional significance level). Fortunately, a chi-square test can compare the two estimated coefficients to each other.

  7. 7

    A similar graph, showing the probabilities when the Senate is controlled by Republicans, evinces essentially no effect for the ideology of the judge before confirmation (although the slope is negative, unlike in Figure 1) but a pronounced impact afterward. We include this figure in the Appendix.

  8. 8

    We could not include district court judges in this analysis because the number of preconfirmation votes for district court judges was too small.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Judicial Independence
  5. Recess Appointees and Constrained Decision Making
  6. Data, Model, and Analyses
  7. Conclusion
  8. References
  9. Appendix
  10. Biographies
  • Abraham, Henry. 1992. Apolitical History of Appointments to the Supreme Court. Oxford: Oxford Univ. Press.
  • Benesh, Sara C. 2002. The U.S. Courts of Appeals and the Law of Confessions: Perspectives on the Hierarchy of Justice. New York: LFB Scholarly Publishing.
  • Bergara, M., B. Richman, and P. T. Spiller. 2003. “Modeling Supreme Court Strategic Decision Making: The Congressional Constraint,” Legislative Studies Quarterly 28: 247280.
  • Bonneau, Chris W., and Melinda Gann Hall. 2009. In Defense of Judicial Elections. New York: Routledge.
  • Brambor, Thomas, William Clark, and Matt Golder. 2006. “Understanding Interaction Models: Improving Empirical Analyses,” Political Analysis 14: 6382.
  • Brent, James C. 1999. “An Agent and Two Principals: U.S. Court of Appeals Responses to Employment Division, Department of Human Resources v. Smith and the Religious Freedom Restoration Act,” American Politics Quarterly 27: 236268.
  • Clark, Tom S. 2008. “A Principal-Agent Theory of En Banc Review,” Journal of Law, Economics, & Organization 25: 5579.
  • Clark, Tom S. 2011. The Limits of Judicial Independence. New York: Cambridge Univ. Press.
  • Cross, Frank B., and Blake J. Nelson. 2001. “Strategic Institutional Effects on Supreme Court Decisionmaking,” Northwestern University Law Review 95: 14371493.
  • Epstein, Lee, Andrew D. Martin, Jeffrey A. Segal, and Chad Westerland. 2007. “The Judicial Common Space,” Journal of Law, Economics, & Organization 23: 303325.
  • Epstein, Lee, and Jeffrey A. Segal. 2005. Advice and Consent: The Politics of Judicial Appointments. Oxford: Oxford Univ. Press.
  • Feeley, Malcolm M., and Edward L. Rubin. 1998. Judicial Policy Making and the Modern State: How the Courts Reformed America's Prisons. New York: Cambridge Univ. Press.
  • Ferejohn, John. 1999. “Independent Judges, Dependent Judiciary: Explaining Judicial Independence,” Southern California Law Review 72: 353384.
  • Fisher, Louis. 2002. “Recess Appointments of Federal Judges.” In The Supreme Court and the Federal Judiciary, edited by Steven C. Caldwell . New York: Nova Science Publishers.
  • Fisher, Louis. 2005. CRS Report for Congress, Federal Recess Judges. http://www.senate.gov/reference/resources/pdf/RS22039.pdf (accessed October 3, 2013).
  • George, Tracey E. 1998. “Developing a Positive Theory of Decisionmaking on U.S. Courts of Appeals,” Ohio State Law Journal 58: 16351696.
  • Giles, Micheal W., Virgina A. Hettinger, and Todd Peppers (GHP). 2001. “Picking Federal Judges: A Note on Policy and Partisan Selection Agendas,” Political Research Quarterly 54: 623641.
  • Giles, Micheal W., Virgina A. Hettinger, and Todd Peppers (GHP). 2002. “Measuring the Preferences of Federal Judges: Alternatives to Party of the Appointing President.” Emory Univ. Typescript.
  • Graves, Scott E., and Robert M. Howard. 2009. Justice Takes a Recess: Judicial Recess Appointments from George Washington to George W. Bush. Lanham, MD: Rowman & Littlefield, Lexington Books.
  • Graves, Scott E., Robert M. Howard, and Pamela C. Corley. 2009. “Appellate Courts Recess and Voting.” In Justice Takes a Recess: Judicial Recess Appointments from George Washington to George W. Bush, edited by Scott E. Graves and Robert M. Howard. Lanham, MD : Rowman & Littlefield, Lexington Books.
  • Hagle, Timothy M. 1993. “Freshman Effects' for Supreme Court Justices,” American Journal of Political Science 37: 11421157.
  • Hall, Melinda Gann. 1987. “Constituent Influence in State Supreme Courts: Conceptual Notes and a Case Study,” Journal of Politics 49: 11171124.
  • Hall, Melinda Gann. 1992. “Electoral Politics and Strategic Voting in State Supreme Courts,” Journal of Politics 54: 427446.
  • Hall, Melinda Gann. 2009. “The Controversy over Electing Judges and Advocacy in Political Science,” Justice System Journal 30: 284291.
  • Herz, Michael. 2005. “Abandoning Recess Appointments?: A Comment on Hartnett (and Others),” Cardozo Law Review 26: 443461.
  • Hettinger, Virginia A., Stefanie Lindquist, and Wendy L. Martinek. 2006. Judging On a Collegial Court: Influences on Federal Appellate Decision Making. Charlottesville, VA: Univ. of Virginia Press.
  • Howard, Robert M. 2007. “Controlling Forum Choice and Controlling Policy: Congress, Courts and the IRS,” Policy Studies Journal 35: 109123.
  • Howard, Robert M. 2008. “Courts, Justice and Governing Coalitions and the Audits of Low-Income Taxpayers,” Journal of Theoretical Politics 20: 181200.
  • Howard, Robert M., and David C. Nixon. 2003. “Local Control of the Bureaucracy: Federal Appeals Courts, Ideology, and the Internal Revenue Service,” Washington University Journal of Law & Policy 13: 233256.
  • Hurwitz, Mark S., and Joseph V. Stefko. 2004. “Acclimation and Attitudes: ‘Newcomer’ Justices and Precedent Conformance on the Supreme Court,” Political Research Quarterly 57: 121129.
  • Kornhauser, Lewis A. 2002. “Is Judicial Independence a Useful Concept?” In Judicial Independence at the Crossroads: An Interdisciplinary Approach 45, edited by Stephen Burbank and Barry Friedman . Thousand Oaks, CA: Sage.
  • Kuersten, Ashlyn K., and Susan B. Haire. 2008. Update to the Court of Appeals Database 1997–2002. http://artsandsciences.sc.edu/poli/juri/appct.htm (accessed October 7, 2013).
  • McNollgast (McCubbins, Matthew D., Roger G. Noll, and Barry R. Weingast). 1995. “Politics and the Courts: A Positive Theory of Judicial Doctrine and the Rule of Law,” Southern California Law Review 68: 16311683.
  • Mayton, William Ty. 2004. “Recess Appointments and an Independent Judiciary,” Constitutional Commentary 20: 515555.
  • Moraski, Bryan, and Charles R. Shipan. 1999. “The Politics of Supreme Court Nominations: A Theory of Institutional Constraints and Choices,” American Journal of Political Science 43: 10691095.
  • Nixon, David C. 2004. “Separation of Powers and Appointee Ideology,” Journal of Law, Economics, and Organization 20: 438457.
  • Note. 1989. “The Politics of En Banc Review,” Harvard Law Review 102: 864884.
  • O'Connor, Sandra Day. 2006. “Judicial Independence and Rule of Law Essential Judges Say,” Remarks on 50th Anniversary of Pakistani Supreme Court. State Department Release http://dhaka.usembassy.gov/uploads/images/gl-VcF8S9FA-yaqZsDBwoQ/pre1aug21_06.pdf (accessed October 3, 2013).
  • Poole, Keith. 2003. Changing Minds? Not in Congress! GSIA Working Paper No. 1997–22, updated January 15. http://voteview.com/chminds/chminds2.htm (accessed October 7, 2013).
  • Report. 1959. “Recess Appointments of Federal Judges.” House Committee of the Judiciary, 86th Cong. 1st Sess.
  • Rosenberg, Gerald N. 1992. “Judicial Independence and the Reality of Political Power,” Review of Politics 54: 369398.
  • Rossiter, Clinton (ed.). 1961. The Federalist Papers: Alexander Hamilton, James Madison, and John Jay. New York: Mentor Books.
  • Rubin, Thea F., and Albert P. Melone. 1988. “Justice Antonin Scalia: A First Year Freshman Effect?Judicature 72: 98102.
  • Sala, Brian R., and James F. Spriggs II. 2004. “Designing Tests of the Supreme Court and the Separation of Powers,” Political Research Quarterly 57: 197208.
  • Scheb, John M., II, and Lee W. Ailshie. 1985. “Justice Sandra Day O'Connor and the ‘Freshman Effect.’ ” Judicature 69: 912.
  • Segal, Jeffrey A. 1986. “Supreme Court Justices as Human Decision Makers: An Individual-Level Analysis of the Search and Seizure Cases,” Journal of Politics 48: 938955.
  • Segal, Jeffrey A. 1997. “Separation-of-powers Games in the Positive Theory of Congress and Courts.” American Political Science Review 91: 2844.
  • Segal, Jeffrey A., and Albert Cover. 1989. “Ideological Values and the Votes of Supreme Court Justices,” American Political Science Review 83: 557565.
  • Segal, Jeffrey A., and Harold J. Spaeth. 2002. The Supreme Court and the Attitudinal Model Revisited. New York: Cambridge Univ. Press.
  • Songer, Donald. 1998. The Multi-User Database on the United States Courts of Appeals, 1925–1996. http://artsandsciences.sc.edu/poli/juri/appct.htm (accessed October 7, 2013).
  • Songer, Donald R., Jeffrey A. Segal, and Charles M. Cameron. 1994. “The Hierarchy of Justice: Testing a Principal-Agent Model of Supreme Court-Circuit Court Interactions,” American Journal of Political Science 38: 673696.
  • Spaeth, Harold J., and Jeffrey A. Segal. 1999. Majority Rule or Minority Will. Cambridge: Cambridge Univ. Press.
  • Spiller, P. T., and R. Gely. 1992. “Congressional Control or Judicial Independence: The Determinants of U.S. Supreme Court Labor Relations Decisions, 1949–1988,” Rand Journal of Economics 23: 463492.
  • Wright, Gerald C. 1976. “Linear Models for Evaluating Conditional Relationships,” American Journal of Political Science 20: 349373.
  • Yates, Jeff, and Andrew Whitford. 1998. “Presidential Power and the United States Supreme Court,” Political Research Quarterly 51: 539550.

Appendix

  1. Top of page
  2. Abstract
  3. Introduction
  4. Judicial Independence
  5. Recess Appointees and Constrained Decision Making
  6. Data, Model, and Analyses
  7. Conclusion
  8. References
  9. Appendix
  10. Biographies

Alternative Table. Interactive Logit Model of Liberal Vote

figure
figure

Predicted Probability of Liberal Vote—by Judge's Ideology, Republican Senate.

Biographies

  1. Top of page
  2. Abstract
  3. Introduction
  4. Judicial Independence
  5. Recess Appointees and Constrained Decision Making
  6. Data, Model, and Analyses
  7. Conclusion
  8. References
  9. Appendix
  10. Biographies
  • scott e. gravesis a research associate at the National Center for State Courts in Williamsburg, Virginia. Dr. Graves' work focuses on the assessment of trial and appellate judicial processes with an emphasis on impact evaluation and forecasting. Before joining the National Center, Dr. Graves served on the faculty of the political science department of Georgia State University, with research interests including judicial politics and methodology. He holds a PhD from the State University of New York at Stony Brook.

  • robert m. howardis a professor of political science at Georgia State University in Atlanta, Georgia. Dr. Howard received his bachelor's degree from Union College in Schenectady, New York, earned a law degree from Suffolk University in Boston, Massachusetts, and a Masters and a PhD in American politics and public law from the State University of New York at Stony Brook. His main research interests are the impact of courts and litigation on social and public policy, and judicial decision making. He is the author of numerous articles and books, including Judging Law and Policy: Courts and Policymaking in the American Political System and Getting a Poor Return: Courts, Justice and Taxes.

  • pamela c. corleyis an associate professor in the Political Science Department at Southern Methodist University. She received her JD and PhD from Georgia State University. She specializes in judicial politics and American political institutions. She is the author of Concurring Opinion Writing on the U.S. Supreme Court (SUNY Press, 2010), which is the first systematic examination of the content of Supreme Court concurrences. She is also the coauthor of The Puzzle of Unanimity: Consensus on the United States Supreme Court (Stanford University Press, 2013), which provides the first comprehensive account of how the Court reaches consensus.