Our legislatures, both national and state-level, are supposed to be representative. So the whole process of divvying each state into districts should somehow result in a set of puzzle pieces that accurately represents the demographics of the whole. Any graduate of high school civics, of course, knows about gerrymandering. And the Supreme Court's Monday ruling in a North Carolina political representation case is a reminder of the complexities of redistricting: how, in fact, can you not gerrymander?
A 2002 story in The Economist has an elegant illustration of how powerful gerrymandering can be: Take a hypothetical state of 60 Democrats and 65 Republicans, 125 voters who are to be divided equally into five 25-voter districts. If the GOP has its way, they'll split the state so that in each district they hold a 13-12 majority; when the districts vote, the Republicans get five slim victories every time. But if the Dems get to make the map, they can concede one district and pack it red to the brim; the remaining districts would now vote 13-12 in Democrats' favor.
Dig a little deeper into the math and it's clear there is no way to make these districts truly fair: at best we'll give Republicans a 3-2 majority in the legislature, far overrepresenting their true majority. And Bartlett v. Strickland—that Monday SCOTUS ruling, which in legal maneuvers too complex to detail here altered the way "minority majority" districts are drawn—reminds us that redistricting is not just about red and blue. It's about black and white—and about urban and rural, college-educated and working-class, Northern and Southern, soccer moms and NASCAR dads. It's about all the complex demographics of the American voter.
Computers and algorithms now make gerrymandering easier: tweak what variables you want to adjust for and—voila!—the jigsaw is rendered. But algorithms can also drain the process of bias. The "shortest splitline" algorithm cares only for population, smashing states in panels of shattered glass:
Iowa, where a nonpartisan panel creates the district, uses a similar process, but one that also respects county and community boundaries.
Iowa, though, is 91 percent white. Minorities there simply aren't going to get a "minority-majority district." Arizona—the shattered glass I link to above—has one of the odder but more-often praised cases of gerrymandering in its 2nd Congressional District. In Arizona, the Hopi reservation is completely enclosed within the reservation of the Navajo, their ancestral and more numerous enemies. To keep the Hopi from losing their voice, their reservation is excised into a separate district that snakes along uninhabited riverland for hundreds of miles before joining a larger and more populated mass of land to the west:
Court cases have routinely affirmed legislators'—and the majority party's—right to lump their constituents as they see fit. So, a year out from the census, the redistricting battle is approaching. Politicians are gearing up for those races, tucked away at the state level, that will offer the power to rewrite the map; fair-vote evangilists are, no doubt, readying their crusade against those politicians.
But buried beneath all the statistics and legal arguments, is there any one method that makes sense? What do we choose: a system that preserves ethnic voices? One that devalues incumbency to keep races competitive? One that preserves the status quo? Or one that is blind to everything but mathematics? There's no easy way.