Gerrymandering on the decline

Partial successes will affect this November's election - and beyond.

From press coverage, you might think that partisan gerrymandering is the worst it’s ever been. But a look at the data suggests that at a federal level, the problem is actually getting better - enough to make a difference in Congress. 

The U.S. Supreme Court has recognized partisan gerrymandering, the purposeful drawing of a statewide plan to favor one party in an extreme manner, as an offense against the Constitution. But perversely, in that same decision, Rucho v. Common Cause, they declined to do anything about it. Since my first time writing on the topic nine years ago, press coverage of the topic has skyrocketed. Now I see several dozen news stories every week. Recently, Jane Mayer published an excellent article about the consequences of a gerrymandered legislature for Ohio citizens. So when gerrymandering occurs, a bright light shines on it. And there is plenty to cover. But all that coverage tends to overlook the gains that reformers have made in the last decade at the level of individual states.

Partisan gerrymandering peaked in 2012

In 2012, partisan gerrymandering had grown to its highest level in the modern era of voting rights. Nine states controlled by Republicans had engineered their Congressional district boundaries to eliminate competition and guarantee dozens of seats for their own side. The distortion in these states was so large that even when combined with the other 43 states (including two Democratic gerrymanders in Illinois and Maryland), Republicans could lose the national vote by over 7 percentage points and still maintain control of Congress. Indeed, in 2012 they retained control despite winning 1.3 million fewer votes than Democrats nationally.

Competition also reached a low point in 2012. Lack of competitiveness arose from two causes: first, geographic sorting of citizens of reliable partisan preference, making all but the most closely divided districts a sure thing for one party or the other; and second, gerrymandering to reduce competition even further. Defining a competitive district as one with a difference of seven points or less between Democratic and Republican candidates, in 2012 only 33 out of 435 Congressional districts were competitive.

One decade later, the most recent redistricting cycle has generated 47 such competitive districts. How did competition increase, and why aren’t you hearing about it?

Split-party control, citizen-driven reforms, and state courts

Like a crime procedural, committing a gerrymander requires three factors: motive, means, and opportunity. Many states in 2021 removed the means and the opportunity. Whether through split-party control of government (Pennsylvania), removal of redistricting power from legislatures by citizen initiative (Michigan, Colorado, and Virginia), or state court action (North Carolina and Maryland), many gerrymanders of the previous decade were not repeated. And when a citizen commission or a court was left in charge of redistricting, partisan balance and increased competition ensued.

As to why we’re not hearing about it, I suspect that the pro-reform movement is a victim of its own success. A whole community of activists has grown, ready to call out offenses - of which there are still plenty. Bad news is more interesting than good news, so we don’t hear too much about successes like Michigan.

Based on strict statistical criteria for detecting the worst outliers, here is a graph showing the number of extreme partisan gerrymanders, decade by decade:

At first, the number of new partisan gerrymanders matched 2012. This was not unexpected. In arguments in Rucho v. Common Cause, attorney Paul Smith warned of an impending “festival of gerrymandering.“ The festival was bipartisan. Republican gerrymanders in Texas, Florida, Ohio, and North Carolina were matched by Democratic gerrymanders in Maryland, Illinois, and New York. (Note that the chart leaves out smaller offenses, for example by Republicans in Georgia and Utah, and by Democrats in Oregon.)

But that was not the end of the story. State court cases were brought successfully in North Carolina, New York, and Maryland, and maps were redrawn. As my co-authors and I have written, legal principles to fight partisan gerrymandering can be found in the state laws and constitutions of all 50 states. Some courts are willing to apply them. (For this and other reasons, this fall’s state Supreme Court races take on outsized importance in Wisconsin, Ohio, and North Carolina. Those are important targets for reformers’ energies in the current campaign season.)

A two-thirds reduction in partisan bias

Because both parties committed offenses, some of which were turned back, the House is now actually more majoritarian than the Senate or the Presidency. In 2012, Republicans could win control of the House with as little as 46.3% of the vote, i.e. they could do it even with a 7.4-point loss to Democrats. Using Dave’s Redistricting data, I calculate that this year, Republicans need 48.9% of the vote, i.e. they can only sustain a 2.2-point loss. Going from a 7.4% to a 2.2% handicap means that over two-thirds of the partisan asymmetry that existed in 2012 is now gone.

As a consequence, the risk of an anti-majoritarian House is reduced. In 2012, a national margin between 0.1% to 7.3% of Democratic-over-Republican votes (“D+0.1% to D+7.3%”) would likely have elected a Republican chamber. This year, that window is considerably smaller, D+0.1% to D+2.1%. (As of the end of August, national surveys are showing a median of D+2.5%.) A mismatch, in the form of a Democratic national vote and a Republican legislative majority, can still happen, but it’s less likely than it was a decade ago. In other words, Democrats have to reach a lower threshold of vote than they did in 2012 to maintain control of the House.

If control of Congress becomes competitive, thank redistricting reform

Recent polls suggest that this year’s Congressional elections might end up on either side of that mismatch zone. As recently as two months ago, a Republican House in 2023 was a foregone conclusion. But after the June 24 Dobbs decision overturning Roe v. Wade, Democratic retention of Congress has come into the realm of possibility.

The above graph is a beta-test of the 2022 tracker for the Princeton Election Consortium (data from FiveThirtyEight). It shows the generic Congressional preference, calculated as a moving median (black trace). The values of this statistic are on the right axis. The orange line is a prediction for what the national margin will be, based on special elections in 2022. Both have predictive power for the November election.

Because of districting patterns including gerrymandering, the margin necessary to take control of the House is different. That’s plotted on the left axis. It’s shifted by 2 points (rounded from the 2.2-point handicap I mentioned above). The straight horizontal black line corresponds to the threshold for which party controls the House. The location of that black line is the net outcome of redistricting.

As of today, the black trace and the orange line are both above that black-line threshold. So to the extent that Democratic retention of the House is even possible this year, a balanced outcome from redistricting played a major role.

From historic lows, increased competition

A second positive outcome in this year’s redistricting is an increase in competition compared with the last decade. Dave’s Redistricting has helpfully tabulated voting tendency of the new districts using precinct-by-precinct voting data from the last several years. And based on those calculations, I count 47 competitive Congressional districts. Compare that with 33 such districts in 2012.

Half of the newly-drawn competitive districts are concentrated in a handful of states: 5 each in Michigan and Ohio, 4 each in Pennsylvania and Florida, and 3 each in California and New York. That doesn’t speak well for other states…but it’s an improvement over 2012.

Sure, competition is likely to decrease from 2020’s 54 competitive races, which is what some analysts have flagged. But that’s to be expected in the first election after redistricting. Legislators like to make districts safer for themselves and their friends. The decline is modest, which is a relatively good development.

Competitive districts were drawn by independent commissions (for example Michigan and California) and by courts (for example Pennsylvania and New York). Surprisingly, competition was also drawn in two heavily gerrymandered states, Ohio and Florida. Here, in search of the maximum number of seats, Republican legislators appear to have cut their margins pretty close. If as a consequence they end up losing districts, that’s what election scholars Bernie Grofman and Thomas Brunell call a dummymander. Pigs get fat, hogs get slaughtered.

A path to future reform - if the Supreme Court allows it

These gains in competition and partisan balance are tenuous. As I wrote several weeks ago, aggressive application of the independent state legislature doctrine can potentially do away with both state courts and independent commissions as checks on legislative power. Indeed, if arguments raised in regard to the Electoral College were taken to their extreme logical conclusion, even the governor could not stop a legislature from redistricting on its own. However, I’m not sure that will happen, on the grounds that it would potentially benefit Democrats, who are more often hemmed in by the status quo. Considering the Supreme Court’s recent tendency to advance partisan interests, the 6-3 conservative majority may waver at ruling in a way that might reduce Republican representation in Congress.

If the Supreme Court lets these paths to reform stand, there is a lot of work to do in coming years. Half of the states have an initiative process that allows them to take redistricting power away from the legislature. This includes Arkansas, Nevada, and Oregon. In a few more states, the governor and legislature may come from opposite parties. Finally, there is one more path: fear. A partisan legislature may lose its grip in power due to demographic changes, triggering a dummymander. At that point, the other party gains control of redistricting. Such a threat can lead legislatures to enact reforms that take redistricting power away from either party - including the opposition. In 2018, the Virginia General Assembly did exactly this. In coming years, Georgia could do the same.

Also, note that even the worst-case Supreme Court ruling only affects federal offices. Redistricting reform at the state and local level is still possible. These are critical to break the loop of self-dealing by legislators. That’s good, because unlike Congressional gerrymandering, state-level offenses do not offset one another. Voters on one side targeted in Georgia and Texas can gain no comfort from other voters targeted in Illinois or Maryland. All these states need reform - separately.

Not all states are covered by state-level reforms. A truly national solution will have to wait for national legislation, a less reactionary Supreme Court, or both. That may take a little while. In the meantime, I encourage you to cultivate your inner federalism.

In the long term, these redistricting battles have one more consequence: they set the stage for future reforms. Different types of progress are possible, depending on whether a state underwent reform, or whether it’s still gerrymandered. These future reforms can build on the new districts to improve competition and representational fairness. That’s a big topic - for another day.