holding 6.2% absolute disparity between Hispanic representation in jury pool and age-eligible citizen population insufficient for prima facie fair cross section claim
How later courts described this case
- holding 6.2% absolute disparity between Hispanic representation in jury pool and age-eligible citizen population insufficient for prima facie fair cross section claim
Written by the judges who cited it.
The opinion
Chief Judge KOZINSKI,
with whom Judge WATFORD joins, concurring:
I join Judge Hurwitz’s opinion (except footnote one) because it faithfully applies the law of our circuit. See United States v. Rodriguez-Lara, 421 F.3d 932, 943 (9th Cir.2005). But I do so without enthusiasm because the rule we are bound to apply is clearly wrong. It makes no sense to measure disparity for fair cross section purposes by looking at absolute disparity, and accept up to 7.7 percent of the total jury pool as a permissible deviation. See United States v. Suttiswad, 696 F.2d 645, 649 (9th Cir.1982) (holding that 7.7 percent absolute disparity is acceptable). The absurdity of this number is brought home by observing that a group that is less than 7.7 percent of the total population can never be underrepresented, no matter how far the jury pool percentage deviates from that in the total population. See Rodriguez-Lara, 421 F.3d at 943 n. 10.
This anomaly disappears for larger groups. A group that is 75 percent of the total population could register a cognizable disparity if it were only 67 percent of the jury pool, while a group that is 7.5 percent could never register a disparity, even if entirely absent from the pool. I have a hard time accepting a rule that favors larger groups and ignores smaller groups altogether.
Our cases have referred to the “main alternative” to absolute disparity as “comparative disparity.” Id. We rejected that approach because we thought it was unworkable with small numbers: “[I]f Hispanics are 2% of the community and 1% of the jury pool, the comparative disparity is 50%, but for every 100 jurors, there is only one fewer Hispanic than would be propor *1026 tional.” Id.; see also United States v. Sanchez-Lopez, 879 F.2d 541, 547-48 (9th Cir.1989). But we’re not dealing here with a hundred people; we’re dealing with a jury pool of over 40,000. For a group that size, there are statistical methods that can easily tell us whether a sub-group that is 5.2 percent of the population (blacks in the Southern District) is underrepresented if it makes up only 3.5 percent of the jury pool.
In the equal protection context, the Supreme Court has used “standard deviation analysis,” see, e.g., Castaneda v. Partida, 430 U.S. 482 , 496 n. 17, 97 S.Ct. 1272 , 51 L.Ed.2d 498 (1977), which “seeks to determine the probability that the disparity between a group’s jury-eligible population and the group’s percentage in the qualified jury pool is attributable to random chance,” Berghuis v. Smith, 559 U.S. 314, 130 S.Ct. 1382 , 1390 n. 1, 176 L.Ed.2d 249 (2010). More than two or three standard deviations means that “the hypothesis that the jury drawing was random would be suspect to a social scientist.” Castaneda, 430 U.S. at 496 n. 17, 97 S.Ct. 1272 . Doing some quick math, I calculate that the disparity between 5.2 percent and 3.5 percent in this case is more than 14 standard deviations. See id. (detailing the formula). So there’s cause to worry.
I’m not sure whether standard deviation analysis is appropriate here, but I suspect that a statistician would laugh at our current methodology. As a three-judge panel, we’re not free to depart from Rodriguez-Lara , but an en banc court could, and perhaps should, take a fresh look at the issue.