The combined talents of the team at Nate Silver's FiveThirtyEight and the Cook Political Report have performed a valuable service. With a study that is creative in design and rich in data, they have cast doubt on the idea that there is a straightforward solution to the controversy over the partisan gerrymandering of congressional district boundaries.
The gerrymandering issue is front and center in the courts.
The Supreme Court has taken two cases: one challenging a Republican-drawn map in Wisconsin, the other a Democratic-drawn map in Maryland. The justices will rule later this year on the constitutionality of district boundaries that appear to unduly favor the party that drew them.
Meanwhile, the Pennsylvania Supreme Court just ruled the congressional boundaries there violate the state constitution by giving Republicans more districts than the partisan split of the state should allow. The state court ordered the lines be redrawn by early next month. In North Carolina, a federal court tossed out the boundaries there for the same reason. That ruling was stayed by the Supreme Court as the justices deliberate over the cases already before it.
Partisan gerrymandering is often seen as the root of much of what is wrong with current politics, from the lack of competitive districts to the rise of ideological extremes, particularly on the right, and the ensuing gridlock in Washington. Most Americans recoil at the contorted shape of some districts and see malevolent hands at work.
The authors of the Gerrymandering Project at FiveThirtyEight acknowledge the problems. But they are candid — and back it up with multiple demonstrations — that if partisan gerrymandering is eliminated, the question of what comes next is not easy to answer.
"It's easy for opponents of gerrymandering . . . to argue what districts shouldn't look like," writes David Wasserman, who oversees House elections at the Cook Political Report. He adds, "But it's much more difficult to say what districts should look like, because reformers can disagree on what priorities should govern our political cartography."
Harry Enten of FiveThirtyEight, another member of the team that spent months on the project, writes that gerrymandering "is a far more complex topic than some analysts and partisans care to acknowledge. . . . There's no 'right' way to draw a district. Prioritize one goal — competitiveness or nonwhite representation, for example — and you have to sacrifice others."
The team set out to show how different goals produce dramatically different maps. In all they drew 258 different state congressional district maps and a total of 2,568 district maps, using a free, Web-based application. The results are all available for viewing at FiveThirtyEight.com.
They produced maps that heavily favored Democrats, with 263 districts where that party would have a clear advantage. They did another that heavily favored Republicans. In that case, there were 275 districts tilted toward the GOP. In both cases, fewer than 30 districts of the 435 were considered competitive.
But then they showed what the breakdown between the parties would look like if the mapmaking process sought to closely match the distribution of seats to the allegiances of the electorate. In another case, they drew boundaries to produce the maximum number of closely competitive districts; in another, they maximized the number of majority-minority districts. They produced one set of maps with districts that followed county boundaries as much as possible. In one case, they turned the process over to an algorithm designed to produce the most compact districts possible.
Redistricting is inherently a political process, and in most states it remains in the hands of politicians. After the 2010 elections, Republicans held the upper hand in far more state legislatures than the Democrats, which is one reason the GOP has the advantage in the House. But Democrats have not been exactly pure in their approach, either. In states where they controlled the process, they sometimes produced districts that were similarly designed to give themselves an advantage.
However, if Democrats were to control more state legislatures and governorships after the 2020 Census, and therefore have a freer hand in many states to draw new boundaries, they would face some obstacles that Republicans have not faced, according to the study.
Democratic voters are far more clustered in and around urban areas than are Republicans, making it easier to pack them into districts. Republicans are more spread out. This geography of politics favors Republicans in the construction of congressional districts.
Another factor is the role of the Voting Rights Act. Democrats can't maneuver their voters into particular districts so easily because that could reduce the number of majority-minority districts. In the past, Republicans have worked with minority (and Democratic) legislators to create more majority-minority districts, reaping the benefit of making surrounding districts more Republican in their leanings.
When the FiveThirtyEight team sought to make districts as compact as possible — and they note that there are scores of measures of what constitutes compactness — they ended up reducing the number of majority-minority districts. When they sought to draw maps proportionate to the breakdown of the electorate, they had to engage in some partisan gerrymandering to get there.
When they sought to produce the maximum number of competitive districts, geography again became a factor. "In an era when Democrats and Republicans are choosing to live next to like-minded neighbors, drawing lots of competitive districts can be tricky," Wasserman writes. "In fact, in some cases, it requires conscious, pro-competitive gerrymandering."
That particular effort produced a map with 242 competitive districts: a remarkable number that would put the political parties on the edge of breakdowns and enrich a new class of political operatives. It also could result, for better or worse, in a House whose members were reactive to the mood swings of an electorate influenced by 24/7 cable news and the power of social media. The "competitive" map also ended up with non-compact districts and a reduced number of majority-minority districts.
The lack of competitiveness in House districts has come gradually and steadily and not simply as a result of partisan gerrymandering. The Cook Political Report has calculated that the number of competitive districts declined from 164 after the 1996 election to just 72 after the 2016 election. Cook's team earlier concluded that less than a fifth of that decline was caused by redistricting.
One factor in reducing the number of competitive House districts is that fewer and fewer counties are balanced politically. Enten cites statistics showing that in the 1996 presidential election, 1,111 counties — about a third of the total nationally — were decided by margins of 10 percentage points or fewer. In 2016, "that number had plummeted to just 310." That makes it all the more difficult to produce competitive House districts without trampling on the goal of keeping counties and communities as intact as possible.
All of this may be a bit wonky for the average person. But there's a larger message in all the data. As people judge the current state of politics, there is more at fault than partisan gerrymandering, as distasteful as it might be. As Wasserman puts it: "Gerrymandering is a really easy practice to condemn and a really complex problem to solve. And just as there are not permanent majorities in American politics, there may never be such a thing as a perfect map."
Correction: An earlier version of this article misstated the number of U.S. House districts. There are 435, not 538.