The Supreme Court is weighing the question of whether voting districts can be drawn in ways that give an advantage to one party, thereby violating the principle of one person, one vote. In Harris v Arizona Independent Redistricting Commission, a group of Republican voters argue that the districting commission redrew the boundaries in 2011 such that, as the Tucson Sentinel put it, “almost all of Arizona’s Republican-leaning districts are overpopulated, and almost all of the state’s Democratic-leaning districts are underpopulated.” The US constitution requires every state to reevaluate the boundaries of voting districts after each national census, taken every ten years, and to redraw those boundaries to take into account changes in population. But did Arizona’s redrawing amount to gerrymandering—the deliberate manipulation of voting district boundaries to give Democrats an advantage? Or was the commission simply trying to comply with the Voting Rights Act amendments requiring that districts should be drawn so as to maximize minority voters?
The impact of partisan gerrymandering can be significant, leading to a majority-vote party in a state (Democrats) obtaining minority status in the legislature in both North Carolina and Pennsylvania in 2012. Only the state of Florida presently outlaws partisan gerrymandering, which means that the Supreme Court’s decision could have far reaching political impact. But can statistics provide a non-partisan way to figure out how much imbalance constitutes a problem?
Yes it can, said Sam Wang, a professor of neuroscience at Princeton, in a recent opinion piece for the New York Times. The problem, as Wang points out, is that “Partisan redistricters stuff voters of the opposing party into a smaller number of districts, while spreading their own voters over a larger number of districts to eke out as many bare wins as possible.” This is the argument before the court in Harris, where conservative voters in Arizona claim they have been overstuffed into fewer districts and therefore have less political voice than they should.
Certainly, statistical descriptions and criteria about districting can add some objectivity to the process—but only if we carefully lay out what these statistics mean, and what they do not. Though statistics can guide us, there is a decidedly human element to statistical decision-making.
Full Article: Can statistics save us from gerrymandering? – STATS.