Log in / Register
Home arrow Economics arrow Commitment and cooperation on high courts : a cross-country examination of institutional constraints on judges


The last category of factors that we will consider as affecting rates of disagreement is norms. As we saw in Chapter 2, a judge may follow a norm because of an internal or personal sense of guilt or shame at not meeting the norms of the court. Alternatively, he may face external penalties for not following a norm, such as esteem or disesteem of colleagues or, more directly, “sanctioning” by the chief justice, such as being left off important panels. norms themselves may be hard to identify directly. Part of our analysis will rely on the notion of norms as the residual once the other factors are taken into account—that is, the level of consensus on the court, controlling for differences in policy preferences and workload.

A norm of consensus may overcome or mitigate differences in policy preferences judges may have in a particular outcome. The Israeli Supreme Court presents a good example of how a strong norm of consensus can overcome ideological division. Despite the ideological split on many contentious issues before the court, the ISC boasts a 95 percent unanimity rate.[1] Similarly, in the uK, the House of Lords appears to have had an overriding consensual norm such that the law lords went “to great lengths to attempt to ensure that they speak with a single voice”[2] and are also “less likely to reverse colleagues with whom they will have to work in the near future”[3]

In addition, however, it may be possible to analyze other factors directly related to norms. In particular, the identity of the chief justice is one of the most important factors when considering norms of consensus. A chief justice can affect the cohesion of the court in a number of different ways.[4] Chief justices may have varying skills in social leadership (leading through force of personality) and task leadership (such as requiring formal meetings and debates and encouraging collaboration).[5] She may also have different mechanisms for enforcing a norm of consensus.[6] A chief justice’s previous political experience may be helpful in becoming an effective leader, and it may be that “the ideal Chief Justice is a persuasive, esteemed, able, and well-liked judge who perceives, fulfills, and even expands his role as head of the Court"[7] Alternatively, a chief justice may use sanctions, such as leaving an individual who continually dissents off important panels or assigning him to difficult or uninteresting appeals.

In terms of task leadership, the chief justice may establish or encourage formal decision-making processes. There may be differences in procedures not only between courts but on the same court under different chief justices. For example, post-hearing conferences can be an important part of building consensus. In Canada, the justices attend a post-hearing conference during which they announce, in reverse order of seniority, how they would decide the case. The length of the conferences varies depending on the chief justice. Some chief justices allow for a substantive discussion of issues, whereas others prefer a brief statement of opinion.[8] Chief Justice McLachlin has even reconvened additional conferences, with the express goal of building consensus and reducing the number of written judgments.[9] The system in the United States is much the same, although the justices state their views in order of seniority.[10] The

Supreme Court of Israel, which tends to sit in panels of three, convenes two conferences, one before the hearing and one after.[11] At the other end of the spectrum is Australia, where there is no conference procedure and judges are markedly individualistic in their judgment writing.[12] Certain chief justices have tried to introduce a judicial conference procedure in Australia, but none has succeeded.[13]

There is an aspect of norms that may relate to workload. We had noted that a judge may be more likely to dissent on larger panels as there is a greater likelihood that she will differ from the majority in terms of policy preferences and the other judges may be less likely to take any disagreement personally.[14] Relatedly, if judges on larger courts are less frequently assigned to the same panels with other judges, the result may be weaker collegial norms. Examining this effect requires controlling for panel size because larger supreme courts, such as the Supreme Court of India, also happen to hear cases in smaller panels, and smaller panels may result in greater unanimity. It is therefore the interaction between the number of judges and likelihood of being placed on a panel with another judge repeatedly that may be important.

Although the impact of the chief justice will vary across courts, we hypothesize:

H5: The dissent rate will vary under the leadership of different chief justices, once we control for differences such as in panel size.

There is a wide variance in the tenure of chief justices across the high courts. Since 1970, the uS Supreme Court has had 3 chief justices, and Australia and Canada 6 each. India, on the other hand, has had 32 chief justices and the uK has had 11. Figure 8-7 shows the tenure of chief justices for each high court from 1970 to the present (it includes the chief justice of the court in 1970 (and their time even if before 1970) and the current chief justice for each country whose tenure is therefore truncated). Related to the number of chief justices in each country, chief justices spent very different times leading the courts. uS Supreme Court chief justices held the position for on average 12 years, whereas in India chief justices averaged less than 2 years in the role. Chief justices

Tenure of chief justices over the period from 1970 to the early 2000s, average tenure and range

Figure 8-7 Tenure of chief justices over the period from 1970 to the early 2000s, average tenure and range (years). Chief justices on the uS Supreme Court on average stay in office longest whereas those on the Indian Supreme Court by far stay the shortest.

would likely have much greater ability to influence norms the longer they lead the court.

In the united States, the chief justice’s leadership has figured prominently among the factors affecting consensual norms.[15] For example, the drastic rise in dissensus in the 1940s, seen in a rapid rise in both dissents and concurrences, occurred during the Stone Court. Marshall (1801-1835), Taft (1921-1930), and Hughes (1930-1941) were seen as actively seeking to minimize dissent whereas conversely, Stone (1941-1946), generally characterized as an ineffective leader, presided over an increasingly fragmented court.[16] A norm of consensus existed in the earlier time period (Waite Court in the 1880s) as evidenced by a high level of disagreements in conference that did not manifest in dissents in the final dispositions.[17] Stone was responsible for “changing the operational norms of the Court from institutional unity to permitting free expression of individual views"[18] Further, Chief Justice Hughes (Stone’s predecessor) may have played a role in the demise of consensus, as the rate of concurrence rose under Hughes, paving the way for more open disagreement in the form of dissent under Stone.[19] More recently, the Rehnquist Court displayed high rates of dissent and concurrence—higher, in fact, than the lower appellate courts at the time.[20]

Similarly, dissent rates have varied under Canadian chief justices. As can be seen in Figure 8-8, dissents rates dropped to their lowest levels toward the end of Laskin’s chief justiceship in 1984, but rose thereafter. This drop in unanimity (and along with a rise in the total number of dissenting votes) may be due more to the introduction of the Charter of Rights and Freedoms in the early 1980s than less successful leadership or a change in norms.[21] Looking across chief justices (and controlling for a range of factors including whether the cases were constitutional) we see the dissent rates were similar for the Court under chief justices Laskin (1973-1984), Dickson (1984-1990), and McLachlin (2000 to the end of the study in 2005). Only during the leadership of Chief Justice Lamer (1990-2000) was there a statistically higher rate of dissent, which may in part be due to the presence of Justice L’Heureux-Dube.[22] However, Lamer has been argued to be an effective task leader, in the sense that his dissent rate was relatively low and he was frequently able to convince his colleagues to join him.[23]

Percent of cases with at least one dissent, 1970 to early 2000s

Figure 8-8 Percent of cases with at least one dissent, 1970 to early 2000s. The dissent rate on the uS Supreme Court was fairly uniformly high and on the Indian Supreme Court uniformly low over the whole period. In Canada, the dissent rate peaked in 1984, at the end of Chief Justice Laskin’s tenure and right before the influx of cases under the new Canadian Charter of Rights and Freedoms. The Australian dissent rate rose over the period while the UK House of Lords saw a decline in dissents followed by an increase to prior levels.

Leadership may provide at least a partial explanation for dissent rates on the Australian High Court, as is evidenced by the changes in dissent patterns that coincide with the appointment of new chief justices.[24] Chief justices vary in their ability to build consensus, just as on the U S Supreme Court.[25] For instance, chief justices Knox, Gavan-Duffy, Isaacs, Dixon, and Barwick did not have a statistically significant effect on dissent rates.[26] On the other hand, Latham had a negative effect, which is consistent with his characterization as a consensusbuilding judge.[27] Contrary to the norm at the US Supreme Court, the justices of the Australian High Court adhere to a deep-seated seriatim tradition, and it is not felt that formal conferences boost cohesion.[28] One interesting finding, which shows the power of the chief justice, is that the Australian High Court, under certain chief justices, saw a rise in both unanimity and dissent in constitutional cases. It was the middle ground of concurrence that became less common. In other words, constitutional cases had a polarizing effect.[29]

As we can see in Figure 8-8, dissent rates on the Indian Supreme Court were uniformly low over the period 1970 to 2000. Indian judicial leadership has not often been studied, though there is reason to believe that chief justices provide at least a partial explanation for dissent rates. Between 1950 and 2010, individualistic behavior was relatively rare on the Supreme Court but rates dropped as time went on. Chandrachud argues that the early disagreement was attributable to the Court’s atmosphere under Chief Justice Kania, whereas the more recent positive consensual norm grew under S.R. Das’s leadership.[30] The chief justice may have an impact when he sits on a constitutional bench, as in such cases unanimity is marginally more likely given their larger size and more controversial nature. Of course, the chief justice may merely select constitutional benches that are more likely to be unanimous from the start.[31]

In order to determine whether the dissent rates over the period from 1970 to the early 2000s are correlated with changes in the leadership of the court, we need to control for the caseload, the ideological makeup of the court at any given time, and the mix of cases. Once we do so, the level of dissent, as measured by the fraction of opinions with at least one dissent, was not consistently correlated across courts with the identity of the chief justice.[32]

Perhaps not surprisingly given the high caseload and the high rate of turnover of chief justices in India, the identity of the chief justice was not related to the fraction of opinions with at least one dissent over the period of the 1970s through 2000s. Interestingly at the other end of the spectrum, in the united States there were only two chief justice during this period, and they also did not appear to be correlated with differences in the fraction of opinions with at least one dissent.

For the other countries, the results were more mixed. Figure 8-9 shows the change in the fraction of cases with at least one dissent across chief justices relative to some baseline chief justice (for Australia, Chief Justice Gibbs, for Canada Chief Justice Laskin, and for the uK, Lord Reid). For Australia, the two most recent chief justices (Gleeson and Brennan) led courts with greater dissension than Chief Justice Gibbs, with a rate that was higher by 15 percent. In Canada, only the Lamer Court had a statistically different fraction dissenting than the Laskin Court. A higher dissent rate for the Laskin Court has been attributed at least in part to the presence of Justice L’Heureux-Dube, who had a very high dissent rate. Finally in the UK, the fraction of dissenting opinions varies considerably across leaders of the House of Lords, with Lord Reid appearing to preside over a high point in agreement in the early 1970s, not seen again until Lord Goff and Lord Browne-Wilkinson in the late 1990s.

Interestingly, the results are slightly different when we look in Figure 8-10 at what makes individual judges dissent when we can control for a different set of factors. When we examine the probability that a judge will dissent and control for the party appointing the judge, the lower court direction, the area of law, and the panel size, there is still no statistically significant difference at the US Supreme Court or the Indian Supreme Court across chief justices. However, the patterns are different than when looking at the annual data in Figure 8-8 (note the different scales in the two figures). The probability that an individual judge dissents is no longer statistically different across chief justices in Australia. In Canada, the Lamer Court no longer has a statistically significant higher dissent rate. Instead once we control for the characteristics of the individual judge, the case mix, and the panel size, judges under the earlier chief justices are slightly more likely to dissent than later judges. Finally, the pattern in the UK becomes much more varied when we look at individual judges. Dissenting rates under Lords Goff and Browne-Wilkinson remain similar to those under Lord Reid but when looking at individual judges, we see they are slightly higher under Lords Diplock, Scarman, and Keith.

variables for chief justices. We also ran these regressions with fraction of dissenting opinions as the dependent variable and the results were similar, though the results for Australia were less significant, and the size of the impact amplified in Canada and the UK. See Appendix 3.

The change in the percentage of cases with at least one dissent for each chief justice, 1970 to early 2000s

Figure 8-9 The change in the percentage of cases with at least one dissent for each chief justice, 1970 to early 2000s. For Australia the change is relative to the percentage of cases with at least one dissent during Chief Justice Gibbs’s tenure (1981-1987), for Canada relative to Chief Justice Laskin (1973-1984), and for the UK relative to Lord Reid (1969-1975). The chief justices for each country are ordered from furthest in the past at the top to most recent at the bottom. These differences were found using annual data, controlling for total cases per year, the ideological split on the court (measured by the standard deviation in the liberal voting rate on the court in the year), and the fraction of cases in the year that involved civil liberties issues. The solid bars are significant at at least the 0.5 level and the non-solid bars are not statistically significant at the 0.5 level. Each court has at least one chief justice with a statistically different rate of disagreement from the baseline chief justice, with the greatest differences for the earlier periods for the UK House of Lords.

Difference in the probability that an individual judge will dissent, 1970 to early 2000s

Figure 8-10 Difference in the probability that an individual judge will dissent, 1970 to early 2000s. For Australia the change is relative to dissenting during Chief Justice Gibbs’s tenure (1981-1987), for Canada relative to Chief Justice Laskin (1973-1984), and for the UK relative to Lord Reid (1969-1975). The chief justices for each country are ordered from furthest in the past at the top to most recent at the bottom. These differences were found using judge- specific data, controlling for the party appointing the judge, the lower court direction (liberal or conservative), the area of law, and the panel size. The solid bars are significant at at least the 0.05 level, and the non-solid bars are not statistically significant at the 0.05 level. The differences across chief justices on an individual judge level are not statistically significant for the Australian High Court. The probability of dissenting seems to have dropped under more recent chief justices in Canada, whereas the rates vary in the UK.

  • [1] Weinshall-Margel, “Attitudinal” supra note 40.
  • [2] T.T. Arvind & Lindsay Stirton, “Lawyers and the Legal Model: Judicial Ideology, JudicialProfessionalism and Institutional Strategy among the Law Lords” (2012) at 39, online:
  • [3] Chris Hanretty, “The Decisions and Ideal Points of British Law Lords” (2012) 43 BritishJournal of Political Science 703 at 715.
  • [4] These theories about the role of the chief justices are summarized in Jason Pierce,“Institutional Cohesion in the High Court of Australia: Do American Theories Travel WellDown Under?” (2008) 46:3 Commonwealth & Comparative Politics 318 at 324-325 [Pierce].
  • [5] Rice & Zorn, supra note 26; Ostberg et al., “Leaders,” supra note 28.
  • [6] Caldeira & Zorn, “Of Time,” supra note 4 at 878 (discussing the use of sanctions andrewards by chief justices to enforce norms); Walker, Epstein & Dixon, supra note 8 at 381.
  • [7] David J. Danelski, “The Influence of the Chief Justice in the Decisional Process of theSupreme Court” in Sheldon Goldman & Austin Sarat, eds, American Court Systems: Readingsin Judicial Process and Behavior (San Francisco: W.H. Freeman, 1978) at 507.
  • [8] Macfarlane, “Consensus,” supra note 19 at 388.
  • [9] Macfarlane, “Consensus,” supra note 19 at 409.
  • [10] Bertha Wilson, “Decision-Making in the Supreme Court” (1986) 36 University of TorontoLaw Journal 227 at 236. As in Canada, the length of the conference in the UK varied under different chief justices. Alan Paterson, Final Judgment: The Last Law Lords and the Supreme Court(Oxford: Hart, 2013) at 86 [Paterson] (noting that the presiding law lord directed the processof the conference).
  • [11] Yoram Shachar, “Inside the Supreme Court of Israel—Decision-Making in Small Panels”(Paper delivered at the IDC Law Review conference, March 19, 2014) [unpublished].
  • [12] Smyth, “Some Are More equal,” supra note 22 at 201.
  • [13] Dixon had limited success in the 1960s, but ultimately fared no better than Latham,Barwick, Mason, and Gleeson: Narayan & Smyth, “What explains Dissent,” supra note 75at 405-406.
  • [14] Edelman et al., supra note 25 at 142.
  • [15] Caldeira & Zorn, “Of Time,” supra note 4 at 900 (arguing that there is a consensual normon the US Supreme Court that changed over time influencing both dissents and concurrencesand partly though not wholly related to changes in the chief justice).
  • [16] Caldeira & Zorn, “Of Time,” supra note 4 at 878. See also Sunstein, “Unanimity”’ supranote 3.
  • [17] Epstein et al., “Norm of Consensus," supra note 36 at 376.
  • [18] Walker, Epstein & Dixon, supra note 8 at 384-385 (although they note that other factorswere also influential such as a change in the court’s docket control, the high turnover rate, andexperience of the judges on the Court at the time).
  • [19] Haynie, “Leadership," supra note 16 at 1164.
  • [20] Scott D. Gerber & Keeok Park, “The Quixotic Search for Consensus on the U.S. SupremeCourt: A Cross-Judicial Empirical Analysis of the Rehnquist Court Justices" (1997) 91:2American Political Science Review 390 at 405.
  • [21] Ostberg et al., “Leaders," supra note 28 at 523-524.
  • [22] Songer et al., “Explaining Dissent," supra note 22 at 404-405.
  • [23] Ostberg et al., “Leaders," supra note 28 at 526. Peter McCormick, “Follow the Leader: JudicialPower and Judicial Leadership on the Laskin Court, 1973-1984" (1998) 24 Queen’s LJ 237 at 276(using Laskin’s own dissenting votes as a measure and finding that Laskin became less of a dissenter as time went on—an indication of a “strong and decisive leadership").
  • [24] Russell Smyth & Paresh Kumar Narayan, “Hail to the Chief! Leadership and StructuralChange in the Level of Consensus on the High Court of Australia” (2004) 1:2 Journal ofEmpirical Legal Studies 399 at 423.
  • [25] Smyth, “Historical Consensual,” supra note 8 at 259.
  • [26] Smyth, “Explaining Historical Dissent,” supra note 50 at 102; Smyth, “Variations in Dissent,”supra note 75 at 239; Smyth, “Role of Attitudinal,” supra note 51.
  • [27] Smyth, “Explaining Historical Dissent,” supra note 50 at 102. But see Smyth, “Role ofAttitudinal,” supra note 51.
  • [28] Pierce, supra note 96 at 333.
  • [29] Andrew Lynch, “Does the High Court Disagree More Often in Constitutional Cases?A Statistical Study of Judgment Delivery 1981-2003” (2005) 33 Federal Law Review 485at 519.
  • [30] Abhinav Chandrachud, “Speech, Structure and Behaviour on the Supreme Court of India”(2012) 25:2 Columbia Journal of Asian Law 222 at 252.
  • [31] Nick Robinson et al., “Interpreting the Constitution: Supreme Court Constitution Benchessince Independence” (2011) 41 Economic & Political Weekly 27 at 31. Similarly, in the UKHanretty found that the presence of the president or deputy president of the Supreme Courtincreases the likelihood of agreement: Chris Hanretty, “Lumpers and Splitters on the unitedKingdom Supreme Court” (2013) American Political Science Association 2013 Annual MeetingPaper, at 19.
  • [32] See Appendix 3 for regression results relating to the fraction of dissenting opinions. Theseresults used annual data for an OLS regression with the fraction of opinions in the year with atleast one dissent as the dependent variable. The independent variables were the total cases inthe year, the standard deviation of the lifetime liberal voting rates of the judges on the court inthat year, the fraction of cases that involved civil liberties issues that year, and a series of dummy
Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
Business & Finance
Computer Science
Language & Literature
Political science