Data scientists help courts grapple with increasingly divisive maps
Photo courtesy of Megan Gall
In some ways, the math community is responding, belatedly, to an invitation from Justice Anthony M. Kennedy to apply technical or scientific abilities to the issue of gerrymandering, more specifically partisan gerrymandering.
In Vieth, Kennedy joined with the court’s conservative bloc in rejecting a challenge to the Republican-controlled Pennsylvania General Assembly’s congressional remap as a partisan gerrymander.
The four-justice plurality opinion, written by Justice Antonin Scalia, questioned whether challenges to partisan gerrymandering were “justiciable,” or something that the courts have the authority to hear.
But Kennedy joined only in the result, and he wrote a solo concurring opinion that no “workable standard” yet existed for proving unconstitutional partisan gerrymandering.
“That no such standard has emerged in this case should not be taken to prove that none will emerge in the future,” Kennedy wrote. He pointed to computers, in the form of “computer assisted districting,” as among the possible paths to coming up with a workable standard.
“These new technologies may produce new methods of analysis that make more evident the precise nature of the burdens gerrymanders impose on the representational rights of voters and parties,” Kennedy wrote.
‘SOCIOLOGICAL GOBBLEDYGOOK’
Progressives think that the time predicted by Justice Kennedy has come, and that the new methods of analysis can identify partisan gerrymandering that goes too far. But the recent Supreme Court argument in Gill, the Wisconsin case, underscored just what data scientists are up against.
Chief Justice John G. Roberts Jr. scorned social scientists’ efforts to come up with a standard for measuring partisan gerrymandering.
Amid a discussion during the argument that touched on the efficiency gap, as well as some other social science tests for excessive partisan gerrymandering, Roberts said to Paul M. Smith, the lawyer defending the lower-court ruling striking down the Republican-drawn map, “The whole point is you’re taking these issues away from democracy and you’re throwing them into the courts pursuant to—and it may be simply my educational background—but [what] I can only describe as sociological gobbledygook.”
The comment drew laughter in the courtroom and plenty of press attention. Smith responded, “Your honor, ... this is not complicated. It is a measure of how unfair the map is.”
In the same session, Justice Alito practically mocked McGhee’s theory, even citing the researcher by name during oral argument.
“In 2014, a young researcher publishes a paper—Eric McGhee publishes a paper, in which he says that the measures that were … the leading measures previously, symmetry and responsiveness, are inadequate,” Alito said. “But I have discovered the key. I have discovered the Rosetta stone, and ... it is the efficiency gap.”
The theory is “full of questions,” Alito declared.
McGhee, the California political scientist who co-developed the efficiency gap theory and was a speaker at the regional MGGG workshops, was present in the courtroom that day.
“My view is when you deal with some complex topics, you’re going to have some gobbledygook,” he says. “There is a complexity here that I think people dismiss sometimes, and it’s that we have a geographic-based system of representation that has political parties layered on top of it.”
On his perception that the chief justice’s “sociological gobbledygook” comment was directed at least partly to him, McGhee says, “I don’t think we should dismiss out of hand Roberts’ concern about the legitimacy of the court” in gerrymandering cases.
Still, McGhee stands by the efficiency gap, saying it “is a solid-enough idea, and there is no contrary argument to validate the Wisconsin plan.”
IN THE TRENCHES
If the Supreme Court ends up recognizing partisan gerrymandering claims, it will take people such as Gall working in the trenches to bring successful challenges to politically drawn maps.
Gall, the summer workshop speaker who left her job as the staff social scientist with the Lawyers’ Committee to join the Legal Defense and Educational Fund in a similar role, is deeply involved in using computer tools to prove all forms of unconstitutional gerrymandering.
Gall has a master’s degree in geographic information systems and a PhD in political science and describes herself as a “quantitative political geographer.”
At the civil rights groups where she has worked, she has used her social science skills to help vet potential redistricting challenges.
“It’s a huge advantage for the organization to have someone in-house who can say, ‘There’s not enough quantitative evidence in this one to hang our hats on,’ or ‘There is enough evidence; let’s go,’ ” she says.
“One of the other big tasks is the data gathering,” Gall adds. “A good bit of my job is to work with the lawyers to pull these data sets together. And the data are a huge problem. Because we live in a federal system, every state does it differently. Counties and municipalities do it differently.”
She has participated in drawing remaps for about 40 to 50 jurisdictions, either as part of legal challenges or at the negotiation level.
Gall says the August workshop was “a really inspiring week because of the collaboration that came out of it. Take myself. I’m the only scientist where I am. I work in relative isolation. Then to go to a place where educators, mathematicians, computer scientists are all gathered; that really helped with the thinking outside the box.”
REDRAWING THE STATES
Gall points to Kevin Hayes Wilson as an example of a young mathematician who came to the Massachusetts workshop as a relative neophyte on gerrymandering but who became energized enough by the gathering to become an activist.
Wilson, 31, has a PhD in mathematics with a focus in number theory from Princeton University and was an academic in mathematics for several years before going to work in technology. He now works for the District of Columbia’s municipal government, applying number theory and his other math skills to problems such as rats.
On his own time, Wilson developed a tool after the 2016 election that allowed users to shift counties between states. If just three properly selected counties in the entire nation were shifted between states, Hillary Clinton would have won the White House against Donald Trump, the tool showed. It is called Redraw the States.
Wilson got some press for his endeavor, and he started learning about redistricting and gerrymandering—enough to sign up for the workshop at Tufts.
“I was curious about what math could actually bring to the gerrymandering debate besides these irreconcilable questions,” he says. “My perception of it is that you have a lot of people who see gerrymandering as a problem. You have a lot of approaches to fixing it. But gerrymandering has a lot of reasons for existing.”
Wilson attended the expert-witness track, and he says he could imagine himself taking the stand. “But it was also impressed upon us how much work and preparation go into that role,” he says.
Since the workshop, he has taken to reading up on the issue and has signed on to help Gall with one of her gerrymandering projects.
Wilson says, in somewhat of an understatement, “It turns out there is a bit of a translation issue between mathematicians and social scientists on the one hand and lawyers and judges on the other.”
Mark Walsh is a freelance writer based in Washington, D.C.
This article was published in the April 2018 issue of the ABA Journal with the title "Boundary Lines: Data scientists help courts grapple with increasingly divisive maps."
