Do Appointers Replicate Themselves?
On a more basic level, we may think that a person who holds the appointment power is more likely to appoint a judge who thinks like her, all other things being equal. She may believe that such a choice is appropriate or may be following an unconscious bias toward individuals with similar views. In either case, judges appointed by people with similar views should decide cases in similar ways, and it is likely that the preferences of appointed judges will change as the preferences of the appointer changes (such as through a change in government). The appointer does not have to be political for this connection to hold—judges in India may, for example, tend to appoint judges with similar views to their own.
For this connection between the judge and who appoints her to hold, the appointer must be able to determine at the time of appointment how an individual will vote once appointed to the bench. Even if someone wishes to appoint individuals with preferences similar to her own, she may be unable to if she can’t tell how the individual will decide cases once on the court. Open hearings and questioning may provide greater evidence of the nominee’s views, but may still be somewhat unhelpful if nominees are unwilling to give fulsome or non-evasive answers during the process. Appointers can look at how candidates have acted in the past—how they decided cases on the court they are on, for example—but judges may act differently on a high court than on a lower court where they feel more bound by decisions of a higher court or be concerned about potential reversal by a higher court.
Despite these variables, as a starting point, we expect that:
H2: A judge’s voting will be positively related to the political views of whoever
Identifying the political views of the appointer may be quite simplistic such as coding the political party of the appointing party or parties. Although such a measure has some intuitive appeal, it also has several problems. One of the largest is that it may ignore or downplay differences across appointing parties. For example, different appointers (such as prime ministers or presidents) may be more or less conservative or liberal despite being from the same party. Although more sophisticated measures of ideology can be used, if how a judge votes is strongly related to the party of whoever appointed him in some countries rather than others, it provides an interesting opening to ask about the differences.
Further, despite the best efforts of some governments to influence high courts in their preferred ideological direction, these efforts are not always successful. One possible reason is that judges, once appointed, may change their judicial ideology. This is important, not least because following appointment it is not possible to directly influence decision-making by appointees. In most jurisdictions it is difficult legally, not to mention risky politically, to try to remove sitting high court judges. For example, in the united States, justices on the Supreme Court are entitled to hold office “during good behavior” and once a judge is confirmed, a president has no power to remove her from office. A judge’s voting may be most likely to correspond to the ideology of the appointing party in the initial period on the bench and may “drift” from that position over time. For example, looking at 23 US Supreme Court justices who served at least 15 terms since 1937, Epstein, Landes, and Posner found that eight of the justices became more conservative over their time on the bench, of whom Republican presidents appointed six. Democratic presidents appointed all four of the judges who became more liberal. Similarly on the Supreme Court of Canada some justices became more liberal, some more conservative, and some showed no change, but there was no apparent connection to the party of the appointing prime minister; both Conservative and Liberal appointees became more conservative or more liberal or stayed the same.
unsurprisingly, American presidents tend to appoint justices with political views similar to their own, particularly in recent years. Epstein, Landes, and Posner, for example, looking at non-unanimous uS Supreme Court decisions between 1937 and 2009 found that Republican appointees voted more conservatively overall than Democratic appointees. The difference was quite large in some areas of law such as in civil rights where the Republican appointees voted in a conservative fashion a little over half the time whereas Democratic appointees only about a third of the time. They point out that when using more sophisticated measures of policy views, Republican presidents can be seen to sometimes appoint liberal justices and Democratic presidents conservative justices. However, the overall relationship is strong. Similarly, Devins and Baum argue that although there was a connection between the party of the appointing president and the voting of justices in the past, it was never perfect—sometimes Republican presidents appointed justices who were more liberal than some Democratic appointees. However, they argue that the Court has been completely polarized since 2010 with Republican appointees as a bloc more conservative than Democratic appointees.
The relationship between the political views of the appointing party and the appointed judge is weaker in other countries. In Canada, the party of the appointing prime minister is positively related to voting by the appointed justice and the newspaper ideological scores for that justice, but the connection is not as strong as in the United States.132 On average over the 1982-2005 period, a justice appointed by a Liberal prime minister had a 5 percent higher liberal voting record than a Conservative appointee.133 The connection differs across areas of law—in some areas the party of the appointing prime minister has little or no connection with how the justices vote.134 Even in Charter of Rights and Freedoms cases, Liberal appointees were considerably more likely to vote liberally in equality cases than their Conservative counterparts, but not in other types of Charter appeals.135 Other indicators of personal views seem more closely related to how justices decide cases than the party of the appointing prime minister, and even when there is a connection, it is much more nuanced than in the United States.136
The justices on the High Court of Australia similarly seem to be less easily differentiated by the party that appointed them than in the United States.137 If and 1980s. Andrew D. Martin & Kevin M. Quinn, “Dynamic Ideal Point Estimation via Markov Chain Monte Carlo for the U.S. Supreme Court, 1953-1999” (2002) 10:2 Political Analysis 134.
- 132. Alarie & Green “Policy Preference,” supra note 68. Some early studies of the Supreme Court of Canada also found that Liberal and Conservative appointees voted differently: Michael Bader & Edward Burstein, “The Supreme Court of Canada 1892-1902: A Study of the Men and the Times” (1970) 8:3 Osgoode Hall LJ 503; Tate and Sittiwong, supra note 103 (examining Court decisions between 1949 and 1985).
- 133. Alarie & Green “Policy Preference,” supra note 68 at p 35 (note, however, they found a much smaller relationship between party of the appointing prime minister and ideal points for the judges).
- 134. See e.g. Ostberg & Wetstein, Attitudinal Decision Making, supra note 92 (finding that the party of the appointing prime minister was not significantly related to some criminal cases (right to counsel and search and seizure) or tax or union cases during the period 1984-2003); Songer, Judicial Behaviour, supra note 92 at 139-40 (finding that party of the appointing prime minister was weakly related in the criminal area but not for economic or civil liberties cases).
- 135. Benjamin Alarie & Andrew Green, “Charter Decisions in the McLachlin Era: Consensus and Ideology at the Supreme Court of Canada” (2009) 47 Supreme Court Law Review 475 at 491 [Alarie & Green, “McLachlin Era”]. But see Ostberg & Wetstein, Attitudinal Decision Making, supra note 92 at 133, 149 (finding that the party of the appointing prime minister was a weak predictor in free speech cases but not significant in equality cases).
- 136. Some argue that there is no connection between the party of the appointing prime minister and how judges ultimately vote: See Morton, “First One Hundred,” supra note 79; McCormick, “Selecting the Supremes,” supra note 13.
- 137. Smyth, “Explaining Historical Dissent,” supra note 84 at 108.
we return to Weiden’s attempt to compare justices in Canada, the United States, and Australia, we see he created a judicial politicization score for each country that measured the degree of correspondence between the political party of the appointing government and the newspaper ideology score of the appointed judge. He found that the Canadian Supreme Court had lower judicial politicization than both the uS Supreme Court and the Australian High Court during the period 1990-1999. A higher judicial politicization score indicates that the appointing government in a country tends more frequently to appoint ideologically like-minded judges. The US Supreme Court had the highest judicial politicization score at 0.917, which was much higher than that of the Canadian Supreme Court at 0.5. The Australian High Court was less politicized than the US Supreme Court, but more politicized than the Canadian Supreme Court.
Evidence of a tie to the politics of the appointing parties is also weak on the UK House of Lords, with studies coming to differing conclusions. Hanretty found the ideal points of Law Lords were weakly connected to the party who appointed the justice whereas Robertson found that previous partisanship affiliation was irrelevant to rates of agreements between judges. Also when the Lord Chancellor made the appointments (that is before the UK changed the committee process), a judge was more likely to be appointed to the House of Lords if his decisions as a lower court judge tended to be affirmed by the House of Lords and he was not extremely anti-government. However, evidence is mixed as to whether it mattered that the judge had been appointed to the Court of Appeal by the same party making the high court appointment—with Salzberger and Fenn finding it made no difference, and Hanretty finding that a judge was more likely to be appointed if he had been previously appointed by the same party.
Judges themselves play a greater role in the appointment of the high court in India and Israel. Israel provides an interesting opportunity to test the connection between candidates and the justices currently on the Supreme Court. As we saw, in Israel justices are typically appointed temporarily to the bench prior to a full-time appointment. Temporarily appointed judges who more frequently dissent (that is, disagree with their colleagues on the bench) were less likely to be permanently appointed to the High Court. Interestingly, temporary judges who tend to grant more applications by individuals against the state are more likely to be appointed to the bench.145
How does this connection between the appointer and the judge look when we use our data from the High Courts Judicial Database and the US database? It is difficult to compare across countries because of the different orientations of political parties in different countries and across time. However, Figure 3-8 shows what we find if we use a very blunt measure of “conservative” and “liberal” appointers across different countries (other than India as judges on the Indian Supreme Court appoint the new judges rather than a political party doing so).146 It shows the difference between the liberal voting rates for conservative and liberal appointees, controlling for the direction (liberal or conservative) of the lower court decision.147 The US Supreme Court has the highest difference with an almost 10 percent difference in liberal voting between liberal and conservative appointees in the period from 1970 to the early 2000s. Differences across judges on the Australian High Court and the Canadian Supreme Court were similar at about 5 percent with the difference for judges on the UK House of Lords the lowest at about 3 percent. These positive differences are statistically significant for Australia, Canada, and the UK, but likely as a function of the crudeness of the measure of ideology (party of the appointer) the 95 percent confidence intervals are particularly large for the United States.
in terms of background. Gadbois, “Background Characteristics”, supra note 88 (studying the court between 1950 and 1967); Abhinav Chandrachud, “An Empirical Study of the Supreme Court’s Composition” (2011) 46:1 Economic and Political Weekly 1 (studying judges on the court between 1985 and 2010). At least in the period 1950-1967, shared backgrounds, as opposed to ideology, did appear to be linked to greater agreement). Gadbois, “Background Characteristics”, supra note 88 at 336.
- 145. Eli M. Salzberger, “Temporary Appointments and Judicial Independence: Theoretical Analysis and Empirical Findings from the Supreme Court of Israel” (2001) 35 Israel Law Review 480 at 503 [Salzberger, “Temporary Appointments”]. For a discussion of temporary appointments, see also Blum, supra note 55.
- 146. See Appendix 1 for a classification of the parties in each country.
- 147. Based on logit regressions with vote (liberal versus conservative) as the dependent variable and party of the appointer and direction of the lower court decision as the independent variables, the coefficients on party of the appointer were significant at the .01 level for the Supreme Court of Canada and .05 level for Australia and the UK. The results were similar if the analysis also controlled for area of law. See Appendix 2 Table 2 for a summary of regression results.
Figure 3-8 Difference in liberal voting rates between judges appointed by conservative appointers and liberal appointers, 1970 to early 2000s. This figure shows the mean difference for each country, along with the 95 percent confidence interval, controlling for the direction (liberal or conservative) of the lower court decision. The difference for justices of the uS Supreme Court is almost twice as high as the Australian High Court or the Supreme Court of Canada, and about three times greater than for the uK. Note, however, the large 95 percent confidence interval for the uS Supreme Court and the fact that these confidence intervals overlap for all countries. India is not included as the judges are appointed by the chief justice rather than a political actor.
-  Some attempts to change the composition of a high court are more transparent than others.For example, in South Africa in the 1950s, the National Party government increased the sizeof the Appellate Division and appointed numerous Afrikaans-speaking judges to the Court inorder to move it in a more conservative direction. Judges who were appointed in the 1950s aspart of this court-packing plan voted for the underdogs at a much lower frequency than otherjudges, suggesting that the National Party’s faith in their conservatism was well placed. LoriHausegger & Stacia L Haynie, “Judicial Decision-Making and the Use of Panels in the CanadianSupreme Court and the South African Appellate Division” (2003) 37:3 Law and Society Review635 at 638.
-  See Epstein et al., “Ideology,” supra note 92 for an overview of different measures ofideology.
-  Epstein, Landes & Posner, Behavior, supra note 74 at 71-72 (noting that party of appointing president in the United States, for example, has several weaknesses, including ignoringdifferences in ideologies of presidents from the same party, assuming presidents always appointjudges for ideological reasons rather than also thinking of personal or other reasons, andneglecting the fact that presidents do not completely control appointments).
-  Lee Epstein et al., “Ideological Drift among Supreme Court Justices: Who, When andHow Important?” (2007) 101:4 Northwestern University Law Review 1483 at 1486 [Epstein et al.,“Ideological Drift”] (“[v]irtually every Justice serving since the 1930s has moved to the left orright or, in some cases, has switched directions several times”); see also Alarie & Green “PolicyPreference,” supra note 68 re: theory of drift.
-  United States Constitution, Article III, Section 1. The only mechanism through which ajustice on the US Supreme Court can be removed is through the process of impeachment. Onlyone justice has ever been impeached, and he was eventually acquitted by the Senate, whichallowed him to remain in office. Rutkus, “Appointment Process,” supra note 2 at 2.
-  See Epstein et al., “Ideological Drift,” supra note 124. See also Alarie & Green “PolicyPreference,” supra note 68.
-  Epstein, Landes & Posner, Behavior, supra note 74 at 116. See also Martin & Quinn,“Preference Change,” supra note 77; Epstein et al., “Ideological Drift,” supra note 124.
-  Alarie & Green “Policy Preference,” supra note 68 at 40-41. Ostberg & Wetstein, AttitudinalDecision Making, supra note 92 found that Supreme Court of Canada judges were more stableover time in some areas (criminal at 80) than others (civil liberties at 124).
-  Epstein, Landes & Posner, Behavior, supra note 74 at 112-13 (examining the fraction ofconservative votes in non-unanimous cases in various areas of law for the period 1937-2009).Schubert also looked at the US Supreme Court in the 1940s through the 1960s and found thatthe Court became more conservative in 1949-1952 when Harry Truman was in power, andthen gradually became more liberal during the terms of Presidents Eisenhower and Kennedybefore becoming more conservative again as a result of appointments by President Nixon. SeeSchubert, Attitudes and Ideologies, supra note 71 at 143; Schubert, Psychometric Analysis, supranote 71 at 89.
-  Devins & Baum, supra note 8.
-  Martin and Quinn used ideal point estimates to identify the median justices on the SupremeCourt from 1953 to 1999, and traced the changes in attitude of the court from a conservativeone in the 1950s, to a liberal one in the 1960s, and back to a conservative one during the 1970s
-  Weiden, “Judicial Politicization,” supra note 69 at 337.
-  Weiden, “Judicial Politicization,” supra note 69 at 338.
-  Weiden, “Judicial Politicization,” supra note 69 at 338.
-  David Robertson, “Judicial Ideology in the House of Lords: A Jurimetric Analysis” (1982)12 British Journal of Political Science 1.
-  Eli Salzberger & Paul Fenn, “Judicial Independence: Some Evidence from the EnglishCourt of Appeal” (1999) 42 Journal of Law and Economics 831 [Salzberger & Fenn].
-  Salzberger & Fenn, supra note 142; Hanretty, “Political Preferment”, supra note 42.
-  There have unfortunately been few studies from India of the connection between theappointing judges and those they appoint. The high court has tended to be fairly homogeneous