Of all the good-government obsessions that keep people focused on process instead of substance, one of the very worst-and I know that lots of you reading this share it-is over the "unfairness" of how congressional district lines are drawn. Within that overrated problem, there's nothing worse than the obsession with pretty and ugly districts. Really: let it go. If you care about politics and public policy, find something else to worry about.
This paragraph of polite rage is brought to you by a current feature over at Slate ridiculing funny shaped House districts-such as one in Maryland which, as Chris Kirk tells us, has been called "'a crazy quilt,' 'a blood spatter from a crime scene,' and a 'broken-winged pterodactyl, lying prostrate across the center of the state.'"
Kirk brings us through the most successful partisan gerrymanders of the last cycle-Michigan, North Carolina, Ohio, and a few more. And it's true: In a handful of states, partisan gerrymanders really did cost Democrats (or, in a few others, Republicans) a few seats in the House.
Overall, however, political scientists like John Sides and Eric McGhee have found that gerrymandering is not a big deal, at least as far as partisan effect is concerned. It is true that current districts help Republicans, but that appears to be primarily because of where Democrats and Republicans live, not because of cleverly drawn lines. That is, because Democrats tend to clump together more closely, it's extremely hard in many states to draw any lines (and particularly lines for compact districts) which do not give Republicans a little edge.
Why don't Republicans take those biases and make them much worse? In part it's because, even in this very partisan era, many states feature divided government and therefore must have both parties sign off on any legislative plan.
In large part, however, small partisan effects happen because the incentives of individual legislators conflict with the interest of their party. The party doesn't always win out. The key to this is the basic math and logic of any partisan gerrymander. In order for a party to win the maximum number of seats in a state, it's necessary to stick as many of the other party's voters into a small number of very lopsided seats. As a result, the victim party "wastes" votes in those seats, since in first-past-the-post elections there's no bonus for winning by a large margin. Meanwhile, in the rest of a state, the party tries to be "efficient" by winning with relatively small margins.
The problem? No incumbent wants to win by a relatively narrow margin, even if it's good for the party. After all, a district that gives Republicans a 5 percent head start can easily produce a Democratic win if it's a good Democratic year overall. Or a district with a 5 percent edge in 2012 can drift to even or worse by 2020. Or-well, no incumbent wants to win by five percentage points, anyway; they want to have districts with 30 or 40 point margins so that they don't have to worry at all about re-election.
In other words, because incumbents often think mainly about their own careers, many states produce bipartisan gerrymanders-maps with only lopsided districts that give incumbents from both parties easy re-elections. That was the outcome, for example, in California's post-2000 redistricting, which meant that seriously contested House races in that state were rare for a decade.
In short, partisan gerrymandering is less of a problem than most people seem to think.
But yes, we concede, it matters sometimes. There's also a legitimate argument that district lines that help one party are unfair, and are bad for democracy. Those who accept those arguments, however, should be the last ones to care about funny district shapes-what those who do and study redistricting call "ugly" districts, such as the "flexing bodybuilder" in Pennsylvania's 16th district.
Here's the thing: There are a whole lot of potential criteria for carving districts out of a state. There are also, however, reasons to keep together "communities of interest"-put rural voters with other rural voters, or people in one cultural part of the state in with others who share some cultural background, or perhaps to keep economic regions together so that mining folks are in one district and farming in another. One might also want to keep together various ethnic or other demographic groups or, perhaps, to create districts which give some of those groups a good chance of being elected. Then there are other political boundaries, such as city and county lines; those in county government, for example, might prefer not to have their jurisdiction unnecessarily split between multiple congressional districts, so that they won't have any specific advocate within the federal government. And it might even make sense to follow geographic natural divisions such as rivers or mountain ranges. What's more, some continuity might seem appropriate; district lines must change over time as the population changes, but starting from scratch every ten years means that fewer voters will get to know their member of Congress over the long haul.
All of those seem like reasonable criteria to me; the trick is balancing them when they conflict, as they almost always do. But where, I would ask, do pretty, compact, boxy, districts fit in? I don't see it. It doesn't guarantee that communities of interest stay together. It doesn't actually, serve any purpose at all. It just seems "right" to lots of people. Even though very few of us believe that there's something better about Colorado and Wyoming than there is about, say, Massachusetts and West Virginia, to pick a couple of the less boxier states.
For that matter, thinking about the states can remind us that very compact, rational-looking shapes-such as North and South Dakota-can be used for perfectly partisan gerrymandering, as Republicans successfully got away with when they created those two states in order to get themselves two extra Senators. Who could have guessed that a straight line could be so nefarious?
So have fun with Slate's little game, but remember: there's nothing wrong with funny district shapes. They tell us exactly zilch about whether it's a good or bad districting job. District shapes are the absolute last thing anyone should be worrying about.
