A JUDGE’S ROLE
A judge may act differently depending on her formal or informal role on the court or on a specific panel. Take, for example, what could be seen as the two opposite ends of the spectrum—a newly appointed judge and the chief justice. Are newly appointed judges less willing to express themselves either through dissents or more extreme voting? Does a chief justice dissent less than other judges, perhaps because she seeks out consensus on the court or because she is able to stack panels in her favor? We will examine both these roles in turn.
Consider a judge who is a “freshman”—newly appointed to the bench. Judges may go through an acclimation period while they adjust to a new set of institutional norms. During this phase, he may be more equivocal and less decisive than his more experienced colleagues, displaying less extreme preferences. He may take time to become comfortable with the type of work and authority that comes with the role. He may also face internal norms of deference to more senior colleagues, as in India. How much he is influenced by this freshman effect may depend on other features of the court such as whether he was appointed by the judges on the court itself, which may at least initially make him less willing to conflict with their opinions. This analysis leads to the following hypothesis:
H2: Freshman judges are likely to dissent less than their more experienced colleagues and to be more ideologically moderate.
As with most of these empirical questions, the so-called “freshman effect” has received most scholarly attention in the united States but this attention has provided only mixed evidence of such an effect. For example, a freshman judge on the uS Supreme Court does not appear to be any more likely to remain unaligned with a voting coalition (either liberal or conservative or moderate) than their more experienced colleagues. Further, freshman judges may be subject to greater “vote fluidity”—that is, changing views between those initially expressed in the judges’ conference and their final vote—than more senior judges, but the evidence is mixed. 55
However, a different effect of longer time on the high court has received greater support.56 Many judges change their ideological positions or vote differently after their first two terms. Freshman judges may be more moderate than judges who are more senior.57 Further, Hagle found that a substantial number of judges appointed to the US Supreme Court between 1953 and 1989 showed initial signs of voting instability—that is, their ideological positions on a number of issue areas changed significantly after their first two terms.58 More recently, Epstein et al. found that the vast majority of US Supreme Court justices exhibit “ideological drift”—a significant change between their first term voting preferences and the ideological positions they adopt later in their careers.59 They claim that their results provide “strong corroborating evidence” for Hagle’s research
- 55. Maltzman & Wahlbeck, “Burger Court,” supra note 13 at 589 (finding freshman had higher rates of voting fluidity on the Burger Court); Dorff & Brenner, supra note 13, n. 22. See also Brenner, “Another Look,” supra note 53 at 324, n. 35 (finding weak evidence that freshman judges were more likely to change their views between the conference and the final votes than judges who had been on the bench for more than three years). But see Timothy M. Hagle & Harold J. Spaeth, “Voting Fluidity and the Attitudinal Model of Supreme Court Decision Making” (1991) 44:1 Political Research Quarterly 119 (finding that freshman had higher rates of voting fluidity than their colleagues on the Burger Court).
- 56. Freshman justices on the US Supreme Court may also write fewer opinions and rely more heavily on precedents. On opinion writing see Saul Brenner & Timothy M. Hagle, “Opinion Writing and Acclimation Effect” (1996) 18:3 Political Behaviour 235 at 254, n 104 [Brenner & Hagle, “Opinion Writing”]; Terry Bowen, “Consensual Norms and the Freshman Effect on the United States Supreme Court” (1995) 76:1 Social Science Quarterly 222; but see Forrest Maltzman, James F. Spriggs II & Paul J. Wahlbeck, “The Politics of Dissent and Concurrences on the US Supreme Court” (1999) 27:4 American Politics Research 488 at 503-504, n. 103 [Maltzman et al., “Politics of Dissent”] (finding freshman judges are no less likely to join or author a concurring or dissenting opinion). For precedents, see Mark S. Hurwitz & Joseph V. Stefko, “Acclimation and Attitudes: ‘Newcomer’ Justices and Precedent Conformance on the Supreme Court” (2004) 57:1 Political Research Quarterly 121.
- 57. Brenner, “Another Look,” supra note 53 at 324, n. 35.
- 58. Hagle, “Freshmen,” supra note 53 at 1147, n. 114 for Supreme Court justices. Wood et al. and Shippen found freshman do appear to undergo voting instability, though conclude that the behavior is not widespread enough to confirm Hagle’s hypothesis. Sandra L. Wood et al., “‘Acclimation Effects’ for Supreme Court Justices: A Cross-Validation, 1888-1940” (1998) 42:2 American Journal of Political Science 690; Charles R. Shipan, “Acclimation Effects Revisited” (2000) 40:2 Jurimetrics 243 at 255 (“[t]he results indicate that some new Justices exhibit acclimation effects, but most do not ... . Thus, this study demonstrates that no general effect exists for newcomers to the Court”).
- 59. Lee Epstein et al., “On the Perils of Drawing Inferences about Supreme Court Justices from Their First Few Years of Service” (2008) 91:4 Judicature 168 [Epstein et al., “Perils”].
on voting instability; however, they also note that some justices continue to change their preferences after this initial acclimation period, and that neither the direction nor significance of these changes is predictable.
When we look at our data for high courts in the period 1970 through the 2000s, there is little evidence of a freshman effect in the other countries we are studying. Take, for example, the claim that freshman judges may be more moderate than senior colleagues. In Canada, judges who were on the Court for less than three years did not vote differently overall than more senior judges— judges who were seen as “liberal” were not more likely to vote in a liberal direction than their more senior liberal colleagues, nor were “conservative” freshman more likely to vote in a conservative direction than senior conservatives. The same was true for the Australian High Court in this period as well as the Indian Supreme Court.
The lack of a voting difference for freshman on the Supreme Court of Canada is consistent with the limited study by other researchers of this issue. For example, although there is a positive correlation between a judge’s ideology scores based on newspaper articles and how he voted in criminal cases in his first year on the bench, there is no significant overall correlation between how a first year judge votes and these scores. Further, as in the United States, many Supreme
Court of Canada judges change how they vote over time; there is no predictable pattern as to whether a judge will become more conservative, more liberal, or stay the same. Ostberg, Wetstein, and Ducat did find one significant difference between freshmen and non-freshmen—freshmen judges participated in fewer panels. They argue that “modern Chief Justices of the Canadian Court have utilized a powerful acclimation tool to help ease the transition for justices in the initial stages of their tenures. This is an important finding because it suggests that an institutional feature such as panel assignment can have a profound effect on a justice’s acclimation to a Court"
Although there was no evidence of freshman voting differently on an ideological basis than senior colleagues for Canada, Australia, and India, the UK House of Lords was different. Freshman at the House of Lords actually appear to vote more “extremely” than those who have been on the Court longer. Freshman judges appointed under a more conservative government were about 3 percent less likely to vote in a liberal direction than more senior conservative appointees. Similarly, freshman liberal appointees were about 2 percent more likely to vote in a liberal direction than more senior liberal appointees. As we discuss in Chapter 8, the UK House of Lords may have had a fairly strong norm of consensus, which could dampen the tendency to vote in particular directions over time.
The story is the same if we look not at how extreme judges are but at whether freshman judges are less willing to dissent than their more senior colleagues because, for example, they need time to become accustomed to the bench. When we look at our data for the 1970s through the 2000s, there is no evidence of freshman (again defined as being on the bench for less than three years) being less likely to dissent than their senior colleagues except new justices on the US Supreme Court who were slightly less willing to dissent (by about 5 percent) than more senior judges. Similarly, Ostberg, Wetstein, and Ducat found newly appointed justices on the Supreme Court of Canada were no more likely to join or write the majority opinion.69 Further, for India, a strong norm of seniority may similarly reduce the applicability of the freshman theory. In his analysis of the Supreme Court of India, Robinson notes that split decisions are rare, despite the frequent use of two-judge panels, because “tradition dictates that the junior judge generally defer to the opinion of the senior"70
There is therefore limited evidence of a freshman effect outside the United States. Recall, of course, that the assignment of judges to panels outside the United States is not necessarily random, which makes it difficult to interpret the results. The freshman judges may be appointed to different types of cases than more senior judges (such as being less likely to be placed on contentious cases than their more senior colleagues). However, it is interesting that freshman judges do not appear to vote differently or dissent less frequently than their more senior colleagues except potentially in the UK.
-  Saul Brenner, “Another Look at Freshman Indecisiveness on the United States SupremeCourt” (1983) 16:2 Polity 320 [Brenner, “Another Look”]; Maltzman & Wahlbeck, “BurgerCourt,” supra note 13 at 589, n. 22; Dorff & Brenner, supra note 13 at 770, n. 22; Timothy MHagle, “‘Freshman Effects’ for Supreme Court Justices” (1993) 37:4 American Journal of PoliticalScience 1142 at 1145 [Hagle, “Freshmen”].
-  Snyder’s 1958 investigation of the US Supreme Court constitutional cases found that freshman justices were more likely to align with a moderate voting bloc than more senior justices.Eloise C. Snyder, “The Supreme Court as a Small Group” (1958) 36 Social Forces 232 at 237238. However, more recent studies have identified weaknesses with this study, and Heck andHall found no evidence of such an effect on the Warren Court and the Burger Court. TerryBowen & John M. Scheb, “Reassessing the ‘Freshman Effect’: The Voting Bloc Alignment ofNew Justices on the United States Supreme Court, 1921-90” (1993) 15:1 Political Behaviour 1;Robert L. Dudley, “The Freshman Effect and Voting Alignments: A Reexamination of JudicialFolklore” (1993) 21:3 American Politics Quarterly 360. See e.g. Edward V Heck & Melinda GannHall, “Bloc Voting and the Freshman Justice Revisited” (1981) 43:3 Journal of Politics 852.
-  Epstein et al., “Perils,” supra note 59 at 179, n. 126.
-  To study whether freshman judges are more moderate, we ran logit regressions with thedependent variable being whether the judge voted in a liberal direction and the independentvariables being whether the judge was a “freshman” (had been on the court for less than threeyears) along with the party of the appointer, the direction (liberal or conservative) of the lowercourt decision, the area of law, a time trend, and both with and without an interaction termbetween the party of the appointer and whether the judge was a freshman.
-  Both the freshman coefficient and the interaction term were insignificant at the .05 level.For the United States, when the regression included whether the judge had been on the Courtfor less than three years but not the interaction term, the coefficient on the freshman variable was significant at the .05 level and negative (-0.095), but when the interaction term wasincluded in the regression neither the freshman nor the interaction term was significant.
-  For Australia, the freshman coefficient and the interaction term were insignificant at the .05level. For India, as there is no “party” of the appointer, we used the lifetime liberal voting ratein criminal cases and the freshman variable as independent variables along with all the othercontrol variables except the interaction term and ran the regression on non-criminal cases. Thecoefficient on the freshman variable was not significant at the .05 level.
-  Benjamin Alarie & Andrew Green, “Policy Preference Change and Appointments to theSupreme Court of Canada” (2009) 47 Osgoode Hall LJ 47 1 at 7-8. See also C.L. Ostberg, MatthewE. Wetstein & Craig R. Ducat, “Acclimation Effects on the Supreme Court of Canada: A CrossCultural Examination of Judicial Folklore” (2003) 84:3 Social Science Quarterly 704 at 714[Ostberg et al., “Acclimation Effects”] (finding that freshman judges did not dramaticallychange their ideological position).
-  Alarie & Green “Policy Preference,” supra note 64.
-  Ostberg et al., “Acclimation Effects,” supra note 64.
-  The coefficients on the freshman and the interaction terms were significant at the .05 level.
-  We ran logit regressions with the dependent variable being whether the judge dissented andthe independent variables being whether the judge was a freshman (less than three years on thecourt), the ideology of the judge (either the party of the appointer or the lifetime liberal votingrecord in criminal cases), the direction (liberal or conservative) of the lower court decision, thearea of the law, and a time trend variable. The freshmen coefficients were not significant at the.05 level for the UK, India, or Australia. The freshman coefficient was significant at the .1 levelin Canada and was negative as predicted but small (-1.85%). For the United States, the freshmancoefficient was significant at the 0.05 level and showed about a 5.5 percent lower probability ofdissenting for freshman. We also ran these regressions using liberal voting rates in criminal