If insanity is doing the same thing over and over and expecting a different result, then litigants who challenge gerrymandering must be mad. Last month, a federal court threw out the Texas Democratic Party’s claim that the state’s new congressional and state house districts are unlawful. This was the twelfth time in a row that this sort of claim has failed in the current cycle. Plaintiffs’ record of futility now spans at least three dozen cases over four decades.

It doesn’t have to be this way. Litigants keep losing these lawsuits because they keep proposing standards the courts have already rejected (such as partisan intent). They’re failing to capitalize on encouraging comments by the Supreme Court, which show that it’s open to a test based on partisan symmetry—the idea that district plans should treat the parties equally. In a forthcoming law review article, Eric McGhee and I lay out just such a test. If plaintiffs were to use it in litigation, they’d have a fighting chance at winning. And if they were to win, then the whole landscape of redistricting in America would be transformed.

First, some background. States redraw their districts every ten years in order to equalize their populations. When this redrawing occurs, the party in power often tries to disadvantage (or gerrymander against) the opposing party. Gerrymandering takes the form of packing the opposition’s supporters into a handful of districts (where they win in landslides) or cracking them among multiple districts (where they lose by slim margins).

Gerrymandering is an age-old practice, but in a 1986 case, the Supreme Court held that it can be unconstitutional if it’s severe enough. The Court confirmed this holding in a pair of decisions in 2004 and 2006. But these cases rejected just about every proposal for identifying unlawful plans. “Predominant partisan intent,” “disregard for traditional districting criteria,” “minority party entrenchment”—none of these were workable standards, in the Court’s view. But these are precisely the standards that plaintiffs have continued to suggest.

While the tone of the 2004 and 2006 decisions was mostly dour, there were glimmers of hope in the latter case. Justice Stevens wrote excitedly about the concept of partisan symmetry: the idea that “the electoral system [should] treat similarly-situated parties equally” in terms of the conversion of votes to seats. Symmetry isn’t the same as proportionality (long the Court’s bête noire), because it asks only that the parties’ votes translate into seats with equal efficiency, not that their shares of votes and seats be identical. Justice Souter commented that “[i]nterest in exploring this notion [of symmetry] is evident.” And the Court’s enigmatic swing voter, Justice Kennedy, added that he was not “discounting its utility in redistricting planning and litigation.”

No litigants have seized this opportunity yet, but they should. To assist them, McGhee and I have devised a new metric of partisan symmetry called the efficiency gap. The efficiency gap is simply the difference between the parties’ respective wasted votes in an election, divided by the total number of votes cast. Wasted votes are ballots that don’t contribute to victory for candidates, and they come in two forms: lost votes cast for candidates who are defeated, and surplus votes cast for winning candidates but in excess of what they needed to prevail. When a party gerrymanders a state, it tries to maximize the wasted votes for the opposing party while minimizing its own, thus producing a large efficiency gap. In a state with perfect partisan symmetry, both parties would have the same number of wasted votes.

Suppose, for example, that a state has five districts with 100 voters each, and two parties, Party A and Party B. Suppose also that Party A wins four of the seats 53 to 47, and Party B wins one of them 85 to 15. Then in each of the four seats that Party A wins, it has 2 surplus votes (53 minus the 51 needed to win), and Party B has 47 lost votes. And in the lone district that Party A loses, it has 15 lost votes, and Party B has 34 surplus votes (85 minus the 51 needed to win). In sum, Party A wastes 23 votes and Party B wastes 222 votes. Subtracting one figure from the other and dividing by the 500 votes cast produces an efficiency gap of 40 percent in Party A’s favor.

The efficiency gap has several properties that make it ideal for measuring the extent of gerrymandering. First, it directly captures the packing and cracking that are at the heart of every biased plan. Surplus votes for winning candidates are the definition of packing, and lost votes for defeated candidates the essence of cracking. All a gerrymander is, in fact, is a plan that results in one party wasting many more votes than its opponent. The efficiency gap tells us exactly how big the difference between the parties’ wasted votes is.

Second, as an arithmetical matter, the efficiency gap represents a party’s undeserved seat share: the extra fraction of seats a party wins relative to a neutral plan. Above, for instance, if Party A and Party B had each wasted the same number of votes, Party A would have won two seats and Party B three. Instead, Party A won four seats, or 40 percent (two out of five) more than it should have. This is precisely what the efficiency gap reveals.

Third, the efficiency gap can be calculated for any election, no matter how uncompetitive. This is not the case for other partisan symmetry metrics (which only work for close races). And fourth, the gap is computed using actual rather than hypothetical election results. Again, this is not true for other metrics (which ask what would happen in a fictional tied election).

But enough of this abstract talk. What are the actual efficiency gaps of past and present plans? In our article, we discuss both congressional and state plans, but here we cover only the former. From 1972 to 2010, then, the level of gerrymandering of the typical large-state House plan held roughly constant. (This is the absolute gap trend line in the chart.) But it then spiked in 2012 to the highest peak (1.6 seats) recorded in the modern era. Over this 40-year period, the typical large-state House plan also became steadily more pro-Republican. (This is the net gap trend line in the chart.) In 2012, the typical plan was more skewed in Republicans’ favor (1.2 seats) than ever before.

Average Net and Absolute Efficiency Gaps for House Plans, 1972-2012
Chart includes only House district plans with at least eight districts. Efficiency gap is measured in seats for greater comparability.

At the state level, seven large states had pro-Republican efficiency gaps of at least two seats in 2012: Florida, Michigan, North Carolina, Ohio, Pennsylvania, Texas, and Virginia. Not a single large state had a pro-Democratic gap this large. But in three of the states with big pro-Republican gaps (Michigan, North Carolina, and Texas), the Republican advantage might not last if public opinion shifts modestly. (The horizontal lines show how each state’s gap would change given plausible swings in the parties’ performances.) Also notably, plaintiffs did not challenge any of this cycle’s most severe and durable gerrymanders. (Plans in red were sued on gerrymandering grounds.) Instead, they oddly focused their efforts on less objectionable maps.

Efficiency Gaps for House Plans by State, 2012
Chart includes only House district plans with at least eight districts. Efficiency gap is measured in seats for greater comparability. Plans in red were challenged on gerrymandering grounds. Horizontal lines show range of values that efficiency gap could take given plausible swings in parties' performances.

This data is more than merely interesting. It also lays the groundwork for a manageable standard that courts could use to curb gerrymandering. Specifically, we recommend the following approach. First, the plaintiff must show that a plan has an efficiency gap that is above a given threshold—two seats for House plans—and that is unlikely to fade away given plausible vote swings. If the plaintiff makes this showing, then the burden shifts to the state to prove that its plan’s large gap is the unavoidable result of a legitimate policy: complying with the Voting Rights Act, for instance, or drawing compact districts. If a map with a smaller gap could achieve the state’s goals at least as well, then its plan would be invalid.

This approach is workable because it perfectly parallels the Court’s test in one person, one vote cases. Substitute “population deviation above ten percent” for “efficiency gap above two seats” and you have exactly the standard the Court has used for decades to decide how unequal districts’ populations may be. The approach also is workable because it answers, for the first time, the Court’s nagging question about gerrymandering: “[H]ow much political ... effect is too much?” An efficiency gap of two seats for House plans is too much, because it means the plans are extreme outliers in the overall distribution.

We’re under no illusion that the Court is in any rush to adopt a gerrymandering test with real teeth. But the Court wouldn’t have to do anything for our approach to take root. All that’s needed is for a lower court to decide that the efficiency gap is consistent with the Court’s 2006 comments. Then the metric could spread from lower court to lower court, much as many doctrines spread without Court intervention. The result would not be a national bar on gerrymandering, at least not at first, but it would still amount to a redistricting revolution.