Symposium: Mind the gap? The efficiency gap, its failures and the “problem” of geography and choice in redistricting

Chris Winkelman is general counsel to the National Republican Congressional Committee, which filed an amicus brief in support of the state appellants in Gill v. Whitford. Phillip Gordon is an associate at Holtzman Vogel Josefiak Torchinsky PLLC and contributed to the NRCC’s brief.

In the dark recesses of single-origin-coffee shops, natural grocery stores, microbreweries and free-range-egg-and-eight-dollar-mimosa brunches in places as far flung as San Francisco, Brooklyn and Washington, D.C., people are coming to a profound, and to them, disturbing revelation: The United States is a republic. As much of a shock as this must be to them, the Constitution spells out how one of our country’s most fundamental acts of representative government is undertaken — choosing the “Times, Places, and Manner of holding Elections.” This is a power the Constitution gives to state legislatures in Article 1, Section IV, and like most other parts of the Constitution, it is subject to certain constraints.

First, the same section of the Constitution gives Congress the power to “make or alter” election regulations. Congress in fact used that power on multiple occasions when it enacted the Voting Rights Act, the various Reapportionment Acts and the requirement of single-member districts for Congressional seats. Second, the Supreme Court imposed requirements such as “one-person, one-vote” to comply with the equal protection clause of the Constitution. Third, many state constitutions or statutes impose various criteria on redistricting, such as single-member districts, compactness and contiguity. Finally, the framers and the Supreme Court understood that redistricting, as Justice Byron White put it, is “intended to have substantial political consequences.” The history of redistricting in the United States, if it teaches us anything, teaches us that redistricting is a political act, was intended to be a political act, and has always been a political act. This simple truth has been difficult for the plaintiffs in Gill v. Whitford to accept, and has proven even more difficult for the Supreme Court.

The Supreme Court has wrestled with the concept of how much partisanship is too much since first finding partisan-gerrymandering claims justiciable in Davis v. Bandemer, although the court sidestepped an earlier opportunity to address the question in 1932 in Wood v. Broom. Various tests have been offered by litigants in the hopes of finding a judicially manageable standard, as yet to no avail. It has been over 30 years since the court’s decision in Bandemer and the court has still not found the long sought-after desiderata of partisan gerrymandering. Just when plaintiffs, distraught that their failures at the ballot box cannot be saved by wins in the courtroom, had given up hope of ever finding a standard that would meet with the approval of five justices, come the plaintiffs in this case with a “scientific” method of determining impermissible partisan gerrymandering: the so-called “efficiency gap.”

The scientific-sounding efficiency gap simply measures the difference in parties’ “wasted” votes. The efficiency gap counts any vote as wasted if that vote was for a losing candidate or was more than what the prevailing candidate needed to win a given election (i.e., 50 percent of the vote plus one in a two-party election). These supposedly wasted votes are then divided by the total number of votes in an election, and the resulting number is the misleadingly named efficiency gap. However, even a cursory inspection of this so-called methodology reveals analytical flaws and partisan skullduggery too blatant to pass constitutional muster or stand up to common sense.

Voters vote!

A fundamental problem with the efficiency gap is that it treats voters as monolithic blocs who vote party above all else. This assumption is contrary to reality. The efficiency gap, much like most statistical election models, attempts to predict the future. The efficiency gap is particularly bad at predicting the future because it relies on the results of a single statewide election for its calculation, aggregating a series of district-by-district elections. As recent elections have laid bare, the assertions that voters 1) will never change their mind, and 2) vote for the party only and not the candidate, are not supported by actual election outcomes. The efficiency gap does not account for vote switchers or split-ballot voting. In fact, the authors of the efficiency gap state that a gap of eight percent ought to be sufficient to render a legislative reapportionment a justiciable partisan gerrymander.

On the nationwide level one need only look at the difference between the 2012 and 2016 presidential elections to see that voters vote based on candidate, not party. In the wake of the 2016 presidential election, multiple studies were conducted to determine how many people voted for President Barack Obama and then voted for President Donald Trump. It has been determined that anywhere from 11 percent to 15 percent of voters switched their votes.

Things do not improve for the efficiency gap at the congressional level either. In the 2016 general election, there were 12 Democratic members of Congress who won election in districts where Trump won the vote. Similarly, there were 23 Republican members of Congress who won in districts where Hillary Clinton won the popular vote. Having watched these races develop, I can tell you — and the results bear out — that candidates matter, issues matter and campaigns matter. While the Democrats clung to a strategy that asked voters to mostly ignore their congressional candidates in favor of focusing on the Republican presidential candidate, Republicans focused on the candidates and the issues district by district.

In a more recent example, Secretary of Health and Human Services Tom Price had consistently won the Georgia 6th congressional district by 20 to 30 percent. The voters of the Georgia 6th elected Karen Handel in a 2017 special election by less than a four percent margin. One could simply chalk this up to a lack of incumbency, but that ignores something very interesting about elections in the United States: If incumbents matter as much as they do, then so does a person’s choice to vote for the incumbent. To put it another way, if 20 percent fewer people vote for a candidate of the same party as the previous incumbent, that seems to reinforce rather than erode the idea that people vote for the person and not the party.

As long as Democrats live in “The Bubble,” they will continue to be at a disadvantage in redistricting 

The efficiency gap is mired in a plethora of problems, both methodological and quantitative. There is not nearly enough time or space to fully document them all here. The biggest single problem with the efficiency gap is that it assumes that political populations are relatively evenly dispersed geographically. Scientific literature and common-sense experience do not support that assumption. Currently, Democrats in the United States are mostly clustered in urban areas, while Republicans tend to inhabit more suburban and rural areas. “Saturday Night Live,” in its skit called “The Bubble,” captured this perfectly when they asked, “What if there was a place where the unthinkable (President Trump’s election) didn’t happen and life could continue for progressive Americans just as before?” Their answer was as funny as it was unsurprising, “Well, now there is… The Bubble is a community of likeminded freethinkers, and no one else.” The skit ends with the discovery that “The Bubble” is just Brooklyn, New York, “with a bubble on it.”

While amusing, this skit highlights what we all know and see when we look at a county-by-county map after a presidential election. There is a sea of red counties interspersed with pinpoints of blue. This feature of the United States’ political geography is well documented in political-science literature and applies equally to Wisconsin, the state at issue in this case, where Democrats cluster in and around the major cities of Madison and Milwaukee and Republicans inhabit the remainder of the state. Justices Anthony Kennedy and Antonin Scalia both noted this clustering effect in Vieth v. Jubelirer, as did Justice Sandra Day O’Connor in Bandemer.

This asymmetrical grouping of voters has real-world consequences on attempts to form legislative districts using traditional districting criteria (compactness, contiguity, equal population etc.). Traditional districting criteria exist, at least in part, to give courts and map makers some guidelines for evaluating maps to ensure compliance with the equal protection clause of the Constitution. Given the focus that the Supreme Court has placed on the shapes of districts, the lower court could be forgiven for rejecting the challenged maps in Wisconsin’s Act 43 if the shapes of the district boundaries were particularly egregious. They were not. The plaintiffs in this case even conceded that the challenged districts were relatively compact and contiguous and that they met the requirements of “one-person one-vote.” I urge readers to compare Wisconsin’s Act 43 maps to maps in states like Maryland and Illinois. Wisconsin’s maps look pedestrian by comparison. To put it another way, if the maps were a coloring book, Wisconsin’s would be for toddlers with crayons and Maryland’s would be for adults with very sharp pencils. However, the two-judge district court majority, in a novel approach, eschewed traditional districting criteria in favor of the efficiency-gap test in order to rule that Republicans had given themselves an overwhelming unconstitutional electoral advantage over the life of Act 43.

The district court ignored the actual impact that its decision will have. Compactness, as Kennedy said in Vieth, helps Republicans because of the effect of political geography. In order to comply with this new efficiency-gap standard, the Wisconsin General Assembly would have to create maps that are less compact and contiguous. The Supreme Court has long lamented the snakes, “sacred Mayan bird[s],” “Rorschach ink-blot test[s]” and “uncouth twenty-eight-sided figure[s]” that creative cartographers have made into legislative districts. Yet, in this case, the challengers are asking the court to force state legislatures across the country to fix the Democrats’ political geography problem by ignoring years of precedent to make less compact and contiguous maps. The court should roundly reject this invitation.

Given all of the flaws inherent in the efficiency-gap theory, and the tenuous nature of Supreme Court jurisprudence in this area, we can only hope that, after 30 years of uncertainty, the court will finally decide what the Constitution has always demanded: In a constitutional republic, it must be the people’s representatives who draw districts, not the courts and certainly not unelected academics promoting flawed and biased notions like the “efficiency gap.”