Politico, June 30, 2015
Headline: Gerrymandering Isn’t Evil
Byline: John Sides and Eric McGhee
Few aspects of American elections inspire as much hand-wringing as gerrymandering. Its “distorting effect” has “increasingly polarized our politics,” Elizabeth Drew writes in the New York Review of Books. It’s another example of “what voters loathe about Washington,” writes veteran journalist Ron Fournier. The reporter Robert Draper calls it “the most insidious practice in American politics.” And you know something in politics is truly hated when it earns its own special report on The Daily Show.
The logic of this critique is straightforward: Gerrymandering results in districts that are dominated by one party, which makes elected legislators beholden only to their party’s base, which then gives them the incentive to be hardcore ideologues, which in turn makes politics so polarized. This logic is one reason voters in some states have taken the power to draw districts away from state legislators and put it in the hands of independent commissions. In an Arizona case, the Supreme Court yesterday gave explicit constitutional permission to set up such commissions in other states, so the reform might catch on.
But the notion that so many of our political ills stem from gerrymandering is, in fact, a bad idea that simply will not die—what we call a Zombie Myth. And when it comes to Zombie Myths in American politics, gerrymandering remains one of the most persistent. Actual evidence from political science research shows only weak correlations between gerrymandering and both polarization and electoral competitiveness. So why does the Zombie Myth persist? We think three major misconceptions are to blame.
Myth No. 1: Gerrymandering a district always makes it safe for incumbents.
According to the logic of the Zombie Myth, gerrymandering produces uncompetitive, safely red or blue districts that create gridlock in Congress. Conventional wisdom further dictates that the point of gerrymandering is to create many safe seats for your party and only a few safe seats for the opposing party. Or, as Draper puts it, redistricting allows “our elected leaders to entrench themselves in 435 impregnable garrisons from which they can maintain political power while avoiding demographic realities.”
But this thinking confuses two different kinds of gerrymandering. The first, what we call a “partisan” gerrymander, redraws district lines with the intention of increasing the number of seats for one party; the second, a “bipartisan” gerrymander, seeks to protect incumbents of both parties by making every district safer for either a Democrat or a Republican. It may seem like an academic distinction, but it’s an important one. Advocates of competitive elections often think they’re attacking partisan gerrymandering, when their real target is the bipartisan kind.
Consider the logic of a partisan gerrymander. A party needs only slightly more than 50 percent of the vote to win a given seat, so anything beyond that is technically irrelevant. “Impregnable garrisons” make for bad partisan gerrymanders because they waste a party’s voters on big wins. For example, a district where 80 percent of the voters belong to your party would have far more of your party's voters than necessary to actually win the seat. It would be better to move some of those supporters to an adjacent district you’re currently losing. You would still win the first district—now with, say, 55 percent instead of 80 percent—but you’d also have a better shot at winning the second. One victory has now potentially become two. In other words, if you’re a legislator who’s in charge of gerrymandering your state, your goal is to win many districts by a relatively narrow margin, and lose the remaining few by a landslide. So partisan gerrymanders don’t produce only safe, uncompetitive districts, or “impregnable garrisons.” In fact, they typically produce at least a few marginally competitive districts—like that 55 percent district in the example above—and often even more such districts than the plans they replace.
Take, for example, North Carolina. During redistricting in 2011, state Republicans created a more favorable playing field for Republican candidates running in North Carolina’s 13 U.S. House districts. They redrew the district boundaries to make the three most Democratic districts even more Democratic, and to spread Republican voters across the other 10 districts, which Mitt Romney later won by margins of between 55 and 62 percent. Clearly those 10 districts tilt Republican, but no one would call all of them “impregnable” or extremely safe.
Partisan gerrymanders can even backfire—something that critics of gerrymandering sometimes ignore. This happens when parties spread their voters just a little too thin, turning a gerrymander into a “dummymander.” When an unfavorable political tide sweeps through, dummymandered districts switch parties, undoing the advantage the gerrymandering party had supposedly engineered for itself. Political scientists Thomas Brunell and Bernard Grofman have documented how, after the 1990 census, Democratic redistricting schemes in several Southern states actually made the Democrats vulnerable to Republican gains. In the 2000s, a similar phenomenon occurred after a Republican gerrymander in Pennsylvania.
To be sure, gerrymandering schemes rarely create a statewide plan that is as competitive as it could be; the risk of dummymandering is too high. But in a study by one of us (McGhee) that examined state legislative elections over a 30-year period, the results showed that as a party wins more districts without earning more total votes—which is, after all, the point of a partisan gerrymander—the competition in those districts was slightly higher, rather than lower. That means partisan gerrymanders are often at least a little more competitive than what they replaced—and they certainly don’t try to minimize competition as one of their goals.
The bottom line: We can complain about making seats safer, and we can complain about capturing new seats by conquest—but we shouldn’t pretend they’re the same.
Myth No. 2: Partisan gerrymanders are ubiquitous.
If the assumption of many commentators is that the party in control of redistricting will automatically extract as much advantage as possible, it stands to reason that partisan gerrymanders must be everywhere.
But historically speaking, the opposite is true: Fair plans have been the rule, not the exception—even in the era of computer-assisted redistricting. A 2014 study by McGhee and Nicholas Stephanopolous tallied the “wasted” votes for each party in each redistricting plan over the past 40 years. Wasted votes are those votes cast that didn’t help a party win a seat—votes in excess of the 51 percent the party needed to win, plus all the votes it received in the districts it lost. If partisan gerrymandering were ubiquitous, you would expect to find an imbalance in wasted votes statewide, with the gerrymandering party wasting far fewer votes than the other party. This is because the gerrymandering party would create districts with narrower, but potentially decisive, advantages (as in the hypothetical 55-45 district we’ve discussed), while the other party would be forced to waste a lot of votes packed into a smaller number of lopsided districts.
Instead, the study found that most redistricting plans over this period had a roughly even balance of wasted votes throughout the state. That is, for every 55-percent Republican district, there was usually a roughly 55-percent Democratic district to match it. And even most of the plans that were not balanced nonetheless achieved balance at some point in the redistricting cycle, largely because of shifting political tides.
Of course, there have also been some very imbalanced maps, but they have been few and far between. The stereotype of redistricting—that it almost always entails a partisan gerrymander with some very competitive districts and some lopsided ones—therefore does not receive much support.
Myth No. 3: Strangely-shaped districts are bad.
This leads us to the third, and arguably most pernicious, misconception: Any weird-looking district must be gerrymandered, and therefore is inherently bad. The Internet is full of pictures of “bad” districts. As the political scientist Jonathan Ladd has noted, what is missing is a sophisticated discussion about what defines a district as “good.” For example: Should districts be geographically compact—simple shapes, like circles or squares? Should they try to represent an existing geographic community or respect geographic boundaries, like those of cities or counties? Should at least some districts include enough members of a racial or ethnic minority that they could elect a representative from their group? Should districts have an evenly balanced electorate to be politically competitive? Or should districts help ensure that the percentage of votes each party receives in the state as a whole would be roughly proportional to the percentage of legislative seats that it wins?
Perhaps unsurprisingly, these different criteria can require significant trade-offs. “Almost all the generally accepted principles of redistricting can come into conflict with each other,” write political scientists David Butler and Bruce Cain. Districts in which racial minorities can elect a representative from their group tend to favor Democrats and are therefore not politically competitive. Some simple-shaped districts might not be competitive either, if they’re drawn around a politically homogeneous community, such as Berkeley, California. Odd shapes [can help] create proportionality between votes and seats. And that’s the point—“weird” or “ugly” districts are a tool to accomplish some of these goals. If you want districts to respect any given value, you’ll often end up with something that looks funny.
A strangely shaped district can certainly be the start of a conversation about gerrymandering, but the shape itself tells us very little about why a district was drawn that way, or whether its shape may actually serve a defensible, or even laudable, purpose. Simply drawing districts with straight lines—as some have advocated—is pathological from the standpoint of representation. Good districts are about people, not polygons.
Despite highlighting these myths, we don’t see ourselves as defenders of the status quo. It’s entirely possible to believe that gerrymandering is not a major cause of our political dysfunction, or that weirdly shaped districts can be good, while also believing that the process needs reform. For example, independent redistricting commissions do appear to increase competition, assuming that’s what we want. And there’s rarely any defensible public interest in partisan gerrymanders. Empowering bipartisan commissions like the ones the Supreme Court has now blessed in Arizona might help to reduce the frequency of such partisan power grabs.
In fact, the simple reality that severe partisan gerrymandering isn’t that common might strengthen the case for change. Skeptical politicians could pursue reform with the knowledge that it would alter the status quo in only a handful of egregious cases. In those cases, the consequences could be significant. But we needn’t fear that reform would turn our politics upside-down overnight.
If all this is true, why do so many well-intentioned reformers take it as an article of faith that gerrymandering is the diabolical villain of modern American politics? Perhaps it’s just tempting—even consoling—to think there’s a simple explanation for our bitter partisan battles. But silver bullets are usually an illusion, and gerrymandering is no different. We need to have a much more sophisticated conversation about what gerrymandering is and what consequences it has. Perhaps then we’ll stop pointing fingers at districts that look weird, and start talking more about what good districts should look like.
Questions:
Who do you believe should draw electoral districts? Explain why. What criteria would you want them to use? What is your rationale for selecting these criteria?