What are the odds?

October 14th, 2014

Last week, I wrote that Ninth Circuit Judge Reinhardt has the “uncanny ability to be on the right panels.” Based on nothing more than intuition, I recalled that Reinhard somehow made it onto the panels of the three most important gay-rights cases in recent years: Prop 8 panel (confirming my 2010 prediction), SmithKline v. Abbott, and the Idaho/Nevada Same-Sex Marriage case. Derek Muller did some back-of-the-envelope calculations and figure that the odds Reinhardt was on all three panels was roughly 1:1,000. I was content to leave the thought experiment at just that.

Apparently I was not the only one who had the same idea. Monte Stewart, who represents several of the groups defending same-sex marriage bans, filed a petition for rehearing en banc in the 9th Circuit. I have no interest in most of his arguments, where he asserts that the 9th Circuit got the caselaw wrong with respect to equal protection and due process. The most significant aspect of the motion is Part 4. Stewart alleges that “The appearance is strong and inescapable that the assignment of this case to this three-judge panel was not done through a neutral process but rather was done in order to influence the outcome in favor of the plaintiffs.”

This claim is based on a statistical analysis performed by Texas A&M Statistics Professor Emeritus James H. Mantis, described in this appendix. In short, he argues that the odds that Judges Reinhardt and Berzon were consistently appointed to many panels concerning gay-rights cases is not consistent with a neutral selection process. The petition identifies 11 cases (“the relevant cases”) decided since 2010 that implicate some gay rights issue:

  1. Kemp v. Ryan (Rymer, Callahan, Ikuta)
  2. U.S. v. Osazuwa (Reinhardt, Wardlaw, Berzon)
  3. Diaz v. Brewer (Schroeder, Thomas, Bennett)
  4. Log Cabin Republicans v. United States (Alarcon, O’Scannlain, Silverman)
  5. Ward v. Carr (Wallace, Noon, M. Smith)
  6. Perry v. Brown (Prop 8) (Reinhardt, Hawkins, N.R. Smith)
  7. Barnes-Wallace v. City of San Diego (Canby, Kleinfeld, Berzon)
  8. Galario v. Adewundmi (Farris, D.W. Nelson, Nguyen)
  9. Golinski v. OPM (Alarcon, Thomas, Berzon)
  10. SmithKline Beecham Corp v. Abbott Labs (Schroeder, Reinhardt, Berzon)
  11. Jackson v. Abercrombine (Idaho/Nevada SSM) (Reinhardt, Gould, Berzon)

Of these 11 cases, Judge Berzon was on five panels, and Judge Reinhardt was on four panels. Three other judges were on two of these panels, and eighteen other active status judges were on none. The attached statistical analysis concludes that such a selection was not the result of a “neutral judge-assignment process.”

As an aside, not all of these cases generated a ruling favorable to gay-rights. This cuts against the argument that the mere presence of Reinhardt and Berzon implies an unfair draw. In my mind, the most significant cases from the lot are Perry v. Brown, SmithKline Beecham Corp v. Abbott, and Jackson v. Abercrombine. All three with Reinhardt, the latter two with Reinhardt and Berzon.

Accepting the 11 cases as a representative sample, depending on the methodology used, the professor calculates different probabilities of Reinhardt, Berzon, or both being on the same panel. The odds of the Reinhardt/Berzon selections range from 61:1 to 444:1.

table-1

table-2

table-3
To put this in terms that may be more understandable, in Poker, the odds of getting three of a kind are 47:1,  a straight 255:1, a flush 508:1, and a full house 694:1.

I haven’t done an independent verification of these numbers, but the high likelihood that these specific judges appear consistently on certain types of cases does raise a red flag. And it would not be unprecedented.

I write this as someone who clerked for a judge that made an allegation concerning non-neutral panel selections. In the University of Michigan Affirmative Action case, Judge Danny Boggs wrote in a “procedural appendix,” that then-Chief Judge Boyce Martin (now retired) deliberately delayed the circulation of a petition for rehearing en banc. He allegedly did this with the goal of waiting for two conservative judges to take senior status, making them ineligible to vote for rehearing. Boggs also asserted that Martin improperly placed himself on the three-judge panel.

A subsequent investigation found that Martin’s actions “raise an inference that misconduct has occurred.” Because corrective actions were taken, notwithstanding these findings of wrongdoing, Acting Chief Judge Batchelder deemed the matters concluded. No disciplinary action was taken against Judge Martin (although he would later be placed under investigation, and resign abruptly, due to allegations about improper reimbursements).

Part of me likes to think that federal judges are beyond interfering with the process of panel assignments, but Judge Martin’s chicanery in the 6th Circuit has disabused me of such optimism.