Drawing on personal experience, the skeptical visitor would tell the story of how some lawmaker was able to persuade friends in the state legislature to fashion a safer seat, or the woeful tale of someone who lost a seat because of the gerrymanderers’ creative handiwork. The net result of these machinations, they insisted, was a decline in electoral competition and accountability, and a rise in partisanship and polarization.
Thus, the belief in the pernicious effects of gerrymandering is not confined to the general public and the news media. It is widely accepted by experienced, professional politicians. There are good reasons to oppose the absurdly shaped districts that gerrymandering often produces (more on this later). But before succumbing to the notion that jiggered legislative districts are at the root of America’s gridlock and divisiveness, it is worth considering the proposition that I, my co-authors and the many political scientists who have studied the effect of gerrymandering on polarization are not nuts.
Arguments about the negative effects of gerrymandering usually take the following form. First, by using the latest sophisticated software, state legislatures can carve out districts that guarantee electoral victory for one party or the other, often in the service of protecting incumbents. Generally, the gerrymanderers accomplish this by packing Republican voters into Republican districts and Democratic voters into Democratic districts. Second, because such gerrymandering makes the districts less competitive, candidates are freed to pander to their bases while ignoring moderate and independent voters. Moreover, politicians who do not pander face primary challenges from ideologically purer candidates.
The result often is that only conservative Republicans can win in districts designed to elect Republicans, just as liberal Democrats usually dominate Democratic districts. Because redistricting no longer produces moderate, bipartisan or heterogeneous districts, moderates have trouble winning election to the House.
These arguments suggest a straightforward solution to excessive partisanship and polarization: Take the politics out of redistricting (a process that occurs every 10 years, as required by the Constitution) by drawing districts that are heterogeneous with respect to voter ideology and partisanship. Appealing to independents would become the key to winning election, and polarization would become a thing of the past.
While there might be other reasons for favoring such heterogeneous districts, there is little systematic evidence to support the claim that gerrymandering has had a substantial effect on polarization. In fact, there is considerable evidence that it has played at most a tiny role.
The first and most obvious counterpoint is found in the Senate, which of course is not subject to redistricting. Senators must run statewide, and it’s hard to win without going beyond your party’s base. But the consensus among political scientists is that the House and the Senate have closely tracked each other in terms of polarization over time. Although the Senate may be a bit less partisan than the House, it has become a very polarized body without the aid of any gerrymandering.
Similarly, in small states with only one House district, gerrymandering can have no effect. Yet those states don’t have any better records of sending moderate legislators to Congress. Wyoming’s at-large seat has been held by a succession of very conservative lawmakers, including Dick Cheney, while socialist Bernie Sanders had a long run as Vermont’s at-large representative and is now one of its senators.
The gerrymandering-leads-to-polarization argument also suggests that partisanship should primarily be reflected in an increasing number of safe conservative and safe liberal seats. But our research suggests that the main cause of political division is the behavior of Democratic and Republican legislators representing similar districts, not how the lines are drawn. In other words, polarization has grown because Democrats and Republicans are representing moderate districts in increasingly extreme ways. So even if the number of safe conservative and liberal seats had not risen, the U.S. House and state legislatures would have become nearly as polarized.
Some of the increase in polarization is related to the greater numbers of solidly liberal Democratic and solidly conservative Republican districts. Although this is consistent with the effects of gerrymandering, the evidence suggests that the increase in safe districts reflects many other factors as well. Political scientists date the decline in competitive congressional districts to the 1960s and 1970s, well before the growth of sharp political division. The most prominent explanation focuses on the regional realignments that moved the South solidly into the Republican column and the Northeast into Democratic control. Increasing partisan gaps between urban and rural voters have also exacerbated polarization. Most important, however, is our finding that the increase in the number of safe districts directly associated with redistricting is not much larger than the rise associated with the long-term trends.
So, given these facts, how much of Congress’s polarization can be attributed to gerrymandering? To get at this, my collaborators, Keith Poole of the University of Georgia and Howard Rosenthal of New York University, and I decided to try our hand at gerrymandering by using our computers to draw our own districts. For each map, we used the characteristics of the districts — partisanship, average income, racial and ethnic composition — to predict how liberal or conservative their representatives might be. But even when we tried our best to create as many heterogeneous and competitive districts as possible, the predicted level of polarization was only slightly below what we observed in the real Congress. So even if there were a radical transformation of how legislative districts are drawn, the effects on polarization would be minimal.
Still, isn’t it just common sense to redraw districts to maximize competition? My answer is probably not. A system in which all districts elected middle-of-the-road candidates would deprive many groups of congressional representation. Not only would African Americans, Latinos and other racial and ethnic groups be underrepresented, so would liberal and conservative voters. Representative institutions such as Congress should have a membership that fully reflects the diverse interests and views of the public.
But wouldn’t creating districts that reflect the views of these groups harm voters by eliminating partisan competition? Wouldn’t this lack of competition produce lawmakers who are low-quality, ineffective and possibly corrupt? Not necessarily. Political scientists Shigeo Hirano of Columbia University and James Snyder of Harvard University have shown that in districts dominated by a single party, the competition within primary elections does as good a job of selecting effective, talented legislators and tossing out ineffective, scandal-ridden incumbents as the general election does in districts with party competition. Reforming the redistricting process to favor party competition could reduce the representative nature of legislatures without increasing any of the supposed benefits of competition.
Should Americans stop worrying, then, and learn to love the gerrymander? No. Even if its effects on polarization are as small as I believe them to be, the practice of elected politicians drawing districts for themselves and their political allies is an invitation to overt corruption. A key to any successful democracy is a widespread belief in the fairness and impartiality of elections. Having incumbents participate in designing districts promoting their job security does little to enhance the legitimacy of American democracy. But even if we take the politics out of drawing the maps, we shouldn’t expect the divisiveness and polarization of our current politics to wither away.
That’s a deeper problem than mapmaking can solve.
Read more from Outlook, friend us on Facebook, and follow us on Twitter.