550
Opinion of the Court
examined the district's entire border—all 234 boundary segments. See id., at 92a. He recognized "a strong correlation between racial composition and party preference" so that "in precincts with high black representation, there is a correspondingly high tendency for voters to favor the Democratic Party" but that "[i]n precincts with low black representation, there is much more variation in party preference, and the fraction of registered voters favoring Democrats is substantially lower." Id., at 91a. Because of this significant correlation, the data tended to support both a political and racial hypothesis. Therefore, Peterson focused on "divergent boundary segments," those where blacks were greater inside District 12 but Democrats were greater outside and those where blacks were greater outside the district but Democrats were greater inside. He concluded that the State included the more heavily Democratic precinct much more often than the more heavily black precinct, and therefore, that the data as a whole supported a political explanation at least as well as, and somewhat better than, a racial explanation. Id., at 98a; see also id., at 87a ("[T]here is at least one other explanation that fits the data as well as or better than race, and that explanation is political identification").
Peterson's analysis of District 12's divergent boundary segments and his affidavit testimony that District 12 displays a high correlation between race and partisanship support an inference that the General Assembly did no more than create a district of strong partisan Democrats. His affidavit is also significant in that it weakens the probative value of appellees' boundary segment evidence, which the District Court appeared to give significant weight. See id., at 20a-21a. Appellees' evidence was limited to a few select precincts, see App. 253-276, whereas Peterson analyzed all 234 boundary segments. Moreover, appellees' maps reported only party registration figures. Peterson again was more thorough, looking also at actual voting re-
Page: Index Previous 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 NextLast modified: October 4, 2007