Sophisticated Ideology: Can I Win?
Unfortunately there is a problem at the core of the error correction view of how judges decide which appeals to hear. For the most part, a judge does not advance his interests if he votes for the court to hear an appeal, but then cannot get a majority of judges lined up in favor of his preferred outcome. In deciding whether the Court should hear an appeal, he may therefore consider not only what outcome he would prefer but attempt to predict whether he would be in the majority if the appeal is heard on the merits. This is known as “sophisticated ideological voting” or an “outcome prediction strategy” Being in the majority allows the judge to be in a position to set an important precedent and influence the evolution of legal doctrine.
Judges who wish to affirm a lower court ruling (that is, agree with the result) may be more likely to follow the outcome prediction strategy. They will evaluate the potential costs and benefits of hearing an appeal differently than those who wish to reverse a decision. Although affirming the lower court’s decision has the benefit of extending its precedential influence, voting to grant leave to appeal also creates the risk that the court will overturn a favorable result. Affirm-minded justices therefore may choose to strategically make either “aggressive grants” (voting to grant leave in order to affirm the decision) or “defensive denials” (voting not to hear an appeal rather than risk reversal of an agreeable result). For instance, in Bowers v. Hardwick, an appeal from a ruling that a law criminalizing sodomy was unconstitutional, the US Supreme Court initially did not have enough votes to grant certiorari, and Justice White wrote a dissent from the denial. Two liberal judges, Justices Brennan and Marshall, then joined White, apparently believing that White wished to affirm the decision to spread its precedential value (an aggressive grant). However, when Justice Rehnquist, a conservative, also joined them, Justice Brennan changed his vote, “presumably because Rehnquist’s ‘join’ made him aware that White was actually pursuing a reversal strategy" Conversely, as the risks of reviewing a disagreeable outcome are less significant, judges who wish to reverse a lower court decision are more likely to follow a simple error-correction strategy. On this theory, then, judges making aggressive grants are most likely to join the majority during the final vote, as they perform the most thorough risk assessments before granting leave to appeal.
For the outcome prediction strategy to work, a judge must be able to make some reasonable prediction of how the court will vote on the merits. She will be better able to predict the outcome where the membership on the court is stable so she is able to build a reasonable assessment of the voting by the other judges. Further, a judge may find the outcome prediction strategy more successful if she is ideologically aligned with the majority of her colleagues and be more willing to grant certiorari than a judge whose preferences conflict with the majority. The existence of stable ideological coalitions may then increase the accuracy of outcome predictions, leading judges to use it more consistently than other strategies.
In addition, to predict outcomes, a judge needs to be able to have a good sense of which judges will hear the appeal, such as on the US Supreme Court, which hears all appeals en banc. Judges in other countries may have more difficulty. For example, Flemming and Krutz have argued that strategic concerns do not play a significant role in leave decisions at the Supreme Court of Canada, largely because for each appeal the chief justice decides not only the size of the panel that will hear the appeal but also who will sit on the panel. A judge on the Supreme Court of Canada hearing a petition for leave to appeal will have difficulty predicting how a case will be decided during a full hearing.
However, hearing appeals in panels may not mean the outcome prediction strategy is invalid for a court. For the Israeli Supreme Court, final outcomes seem to reflect the same preferences as leave-to-appeal decisions, despite the Court’s use of small panels, as the desire to reverse a lower court decision is an important factor in granting of leave.
Judges’ votes during leave decisions may therefore rest on strategic factors that reflect their unique institutional constraints. We will therefore examine the following hypothesis:
H7: Judges who vote to grant review are more likely to be in the majority during a hearing on the merits, particularly if the lower court decision accords with their personal views.
On this view, a justice first determines her preferred outcome for a given case, then the probability of such a resolution carrying the day, given the composition of the court as a whole, and only then decides whether to vote to grant or to deny certiorari—voting to grant only when the justice is reasonably confident that her preferred outcome will be the resolution on the merits. For this strategy to work, the justices must have sufficient familiarity with their colleagues’ inclinations to enable them to predict how they will vote on the merits.
Although justices of the US Supreme Court do pay attention to non-strategic cues or signals, whether or not they agree to hear an appeal is related to outcome on the merits. Caldeira, Wright, and Zorn, for example, examining the 1968, 1982, and 1990 terms found that if a judge voted to hear an appeal, he would be significantly more likely to be in the majority. Further, Brenner and Krol found that justices who both voted to reverse and won on the merits voted for cert almost 80 percent of the time, compared to about 40 percent for those who lost on the merits and voted to affirm, whereas no such pattern occurred when the grant rates of reverse-minded judges who won were compared with those who lost.
Another way to test whether a judge is predicting outcomes when she is deciding whether the court should hear the appeal is to look at how she fits in with the rest of the court. Caldeira, Wright, and Zorn, for example, found that on a conservative court, the more liberal a given justice is, the less likely he is to vote to grant certiorari. Conversely, on a liberal court, increasing liberalism of a justice is positively correlated with the net probability of voting to grant certiorari. Thus a conservative judge on a conservative court votes to hear appeals at a much higher rate than when on a very liberal court—as would a liberal judge on a liberal court. They argued that these findings supported the existence of sophisticated voting amongst justices, be they liberal or conservative.
Although justices on the uS Supreme Court seem to act strategically when deciding whether to hear an appeal, Canadian Supreme Court justices do not. If we look at the Supreme Court of Canada in the period 1984 to 2011, justices on the panel deciding whether to grant leave were no more likely to be in the majority on the merits than justices who were not on the leave panel. Further, it makes no difference if a judge on the three-justice leave panel is sitting with other justices with the same preferences or with different preferences—a conservative justice, for example, is no more likely to be in the majority on the merits decision than a justice not on the leave panel, whether or not the panel included one or two other conservative justices or two liberal justices. As a result, Canadian justices do not appear to decide whether to grant leave based on their assessment of whether their preferred outcome will win on the merits.
To summarize, US Supreme Court justices appear to base their decision on whether the Court should hear an appeal at least in part on whether they feel their views will win when the case is heard, particularly if they want to affirm the lower court decision. Affirm-minded justices risk much more than reverse-minded justices when voting to grant cert, and thus have more incentive to employ strategic voting. Canadian Supreme Court justices, on the other hand, do not appear to vote strategically. A key difference between the Canadian and American Supreme Courts may lie in institutional design choices. It is more difficult to game the system in Canada. unlike in the united States where all justices hear the merits decision, a justice on the Supreme Court of Canada has to predict not only how particular justices will vote, but also the chief justice’s choice of both the size and composition of the panel that will ultimately hear the decision on the merits.
-  One of the first scholars to explore certiorari strategies was Glendon Shubert, whoseresearch established that judges vote to grant or deny leave to appeal based on the likelinessof achieving their preferred results at a hearing on the merits. Glendon Schubert, “The Studyof Judicial Decision Making as an Aspect of Political Behaviour” (1958) 52:4 American PoliticalScience Review 1007 [Schubert, “Study of Judicial”]; see also, Glendon Schubert, “The CertiorariGame” in Quantitative Analysis of Judicial Behaviour (Glencoe, Illinois: The Free Press, 1959);Glendon Schubert, “Policy Without Law: An Extension of the Certiorari Game” (1962) 14:2Stanford Law Review 284.
-  A judge may also want to minimize conflict with colleagues, as a more united court may beperceived as more legitimate. We will consider this point in more detail in Chapter 8, as docketcontrol can influence consensual norms.
-  See e.g. Saul Brenner, “The New Certiorari Game” (1979) 41:2 Journal of Politics 649 [Brenner,“New Certiorari Game”]; Boucher & Segal, supra note 78; Sara C. Benesh, Saul Brenner &Harold J. Spaeth, “Aggressive Grants by Affirm-Minded Justices” (2002) 30:3 American PoliticsResearch 219, n. 26 [Benesh et al., “Aggressive Grants”]; see also Gregory Caldeira, John Wright& Christopher Zorn, “Sophisticated Voting and Gate-Keeping in the Supreme Court” (1999)15:3 Journal of Law, Economics and Organizations 549 [Caldeira et al., “Sophisticated Voting”].
-  Benesh et al., “Aggressive Grants,” supra note 91, n. 26; Boucher & Segal, supra note 78,n. 32; Brenner, “New Certiorari Game,” supra note 91, n. 32.
-  Lee Epstein & Jack Knight, The Choices Justices Make (Washington, DC: CQ Press, 1998) at61, 122 (a fourth judge eventually joined in, and the Supreme Court went on to reverse the lowercourt’s decision). In another case, Miller v. California, 413 U.S. 15 (1973), the US Supreme Courtupheld laws banning certain sexually explicit works. For a subsequent case, Lilies v. Oregon, 425U.S. 963 (1976) (in which a lower court followed Miller, Justice Stevens voted to deny certiorari,saying that “[r]egardless of how I might vote on the merits after full argument, it would bepointless to grant certiorari in case after case of this character only to have Miller reaffirmedtime after time").
-  Caldeira et al., “Sophisticated Voting,” supra note 91 at 566, n. 32.
-  Flemming & Krutz, “Selecting Appeals,” supra note 42 at 235, n. 16.
-  Eisenberg et al., “Case Selection,” supra note 14, n. 8; Eisenberg et al., “Case Selection,”supra note 14 (“[w]ith respect to both criminal and civil cases, the table offers evidence thatdisagreement with the ruling in the district court is an important criterion in referring casesfor all justices. Full reversal occurs in over 65 percent of both criminal and civil cases reviewedon the merits” at 196).
-  Caldeira, Wright & Zorn, “Unified Model,” supra note 62 at 4, note (“for justices to votestrategically on certiorari, they require some reasonably good idea of the likely outcome of aparticular case should that case reach the merits. If this is the case, and if justices both have anduse this information in making their certiorari decisions, then justices who vote for certiorarishould ceteris paribus, be more likely to be on the winning side on the merits”).
-  Caldeira, Wright & Zorn, “Unified Model,” supra note 62 at 5 (this effect remains even whencontrolling for the ideology of the judges). See also Schubert, “Study of Judicial,” supra note 89;Palmer, supra note 78 at 396.
-  See also Brenner & Krol, supra note 77 at 833; Brenner, “New Certiorari Game,” supra note91 at 653-654; Boucher & Segal, supra note 78 at 831 (finding each justice on the Vinson Courtwas significantly less likely to vote for cert when he eventually voted to affirm the lower court’sdecision, although there was evidence that justices would take a chance and grant cert if theyhad a high level of support from other members of the Court).
-  Caldeira et al., “Sophisticated Voting,” supra note 91 at 564.
-  Caldeira et al., “Sophisticated Voting,” supra note 91 at 565.
-  Caldeira et al., “Sophisticated Voting,” supra note 91 at 566.
-  Alarie & Green, “Docket Control,” supra note 9 at 20.
-  Alarie & Green, “Docket Control,” supra note 9 at 24.
-  Flemming & Krutz, “Selecting Appeals,” supra note 42 at 235 argue that Canadian SupremeCourt justices cannot act strategically because this prediction is too difficult. See also Alarie& Green, “Docket Control,” supra note 9 at 25; Flemming, Tournament of Appeals, supra note13 at 85.