The Brennan Center for Justice published a major new report last week that uses multiple statistical measures to examine how congressional maps around the country mete out partisan advantages to one party or the other. The study provides detailed mathematical evidence for what redistricting-watchers have long known: The redistricting plans passed in the wake of the 2010 census give Republicans a monumental and consistent advantage nationwide.
Although these tests don’t necessarily determine the root causes of these advantages, states with single-party control over the redistricting process stand out as having the worst disparities between the popular vote and seat counts. (For instance, Donald Trump won Michigan by 0.2 percentage points, but Republicans hold 64 percent of the state’s congressional districts.)
Indeed, Daily Kos Elections’ own past work has used hypothetical nonpartisan maps to demonstrate that intentional gerrymandering is likely responsible for the vast majority of this GOP edge. Other factors, such as the geographic “clustering” of Democrats into cities while Republicans are more efficiently spread out over wider territory, do also likely play a role.
The Brennan report’s three statistical tests include a novel approach known as the “efficiency gap.” Proponents of this approach hope it will undergird a new effort asking the Supreme Court to start striking down partisan gerrymanders as unconstitutional, which it just might do in an upcoming Wisconsin case that’s likely headed before the high court this fall over the GOP’s state Assembly gerrymander shown at the top of this post. So how does the efficiency gap work? We’ll explain it and much more below.
The efficiency gap aims to measure how many votes for each party are “wasted” in legislative or congressional races by counting up all the votes for a losing candidate and for a winning candidate that were in excess of a majority of the major-party vote, which is the minimum needed to win a race. It then subtracts the wasted Republican votes from the wasted Democratic votes and divides that remainder by the total number of major-party votes statewide. Consequently, a positive proportion indicates a gap favoring Republicans, while a negative one gives Democrats an advantage.
Here’s a simplified example. While the efficiency gap tends to be most meaningful with larger numbers of districts, we’ll imagine a state with three congressional districts, since the math behind the concept is the same either way. The combined vote for each party’s three candidates across all three districts is exactly 150 votes for each side, or 300 total statewide—in other words, this is a perfectly divided state. However, in two of these hypothetical seats, Republicans win 60 votes to 40 for the Democrat, meaning they “wasted” 9 votes since it only would have taken 51 to win, for a total of 18 wasted votes. Meanwhile, one seat goes 70-30 for the Democrat, so 19 Democratic votes are wasted here.
The efficiency gap also adds votes for losing candidates to the “wasted voted” pile. In this case, 80 Democratic votes were wasted in the two races Republicans won (40 in each case), for a total of 99 wasted votes. Republicans, meanwhile, only wasted 30 votes in the race they lost to the Democrat, giving them a final tally of 48 wasted votes. Subtracting the GOP wasted votes from the Democratic total gives us a remainder of 51 votes. Remembering that there were 300 votes overall, 51 divided by 300 equals an efficiency gap of 17 percent, marking a very large penalty against Democrats.
In our hypothetical example, one party easily wins a two-thirds majority of seats even though the statewide support for both parties was exactly the same—an ideal outcome for anyone wishing to put in place a partisan gerrymander, but not a positive result for democracy. The creators of the efficiency gap contend that a gap of 7 percent or more in favor of one party is so historically atypical that it amounts to evidence of a partisan disparity so extreme as to potentially be unconstitutional if there aren’t mitigating factors like the state’s underlying political geography.
The Brennan report also explores two other tests. One, called the “seats-to-votes curve,” compares the share of seats a party wins with its statewide share of the two-party vote over time to create a graph based on historical data. If a party consistently wins far fewer seats than its statewide popular vote share, particularly if the popular vote is relatively evenly divided, it’s a sign that the map significantly favors the other party. And if such a curve is relatively neutral between the parties historically but a new map produces a result that’s far out of line with past expectations, it could be a sign of nascent gerrymandering.
The third approach is known as the “mean-median” test, which uses a statewide election instead of legislative or congressional results. As Daily Kos Elections has previously outlined in this explainer post, this test would, for example, rank every district from Hillary Clinton’s biggest victory margin to Trump’s biggest victory margin. The district that falls in the middle of that ranking is thus the median district. In our approach, which is similar to Brennan’s, we compare that median district to the overall statewide margin, which can reveal a partisan edge. For example, even though Clinton won Virginia by more than five points, the median seat favored Trump by more than three, giving the GOP an almost nine-point advantage.
Applying these three tests, the Brennan Center finds that Republican-drawn states have by far the largest partisan advantages, particularly in heavily populated states like Florida, Michigan, North Carolina, Ohio, Pennsylvania, Texas, and Virginia. By contrast, the rare Democratic-drawn states such as Maryland or Illinois have much smaller or even negligible Democratic advantages. All three tests have their drawbacks, which the report itself details, but the totality of evidence leads to only one conclusion: Republicans enjoy a significant partisan advantage thanks to the way they have drawn congressional maps across the country.
