Math professor helps lead geometry-based efforts to combat gerrymandering

11/10/2015 - Medford/Somerville, MA - Mathematics department tenure recipient Moon Duchin poses for a photo at the Faculty Inspirations event in the Hirsch Reading Room of Tisch Library. Evan Sayles for Tufts University

Gerrymandering, the redrawing of political districts to favor a particular group or agenda, is an increasingly relevant phenomenon in American politics. Common forms of gerrymandering include partisan gerrymandering, in which a political party attempts to redraw district boundaries in order to favor their party over another, and racial gerrymandering, in which the party in power manipulates district lines to either disenfranchise or support voters of certain races.

The term gerrymandering has a local origin, taking its name from Massachusetts Governor Elbridge Gerry’s redrawing of districts around Boston in 1812 to favor a political ally. A satirical cartoon in a local newspaper depicted the elaborate and unorthodox shape of the district as resembling a salamander, hence the “mander” half of the phrase. 

The practice continues today, despite its controversial nature, but one Tufts professor recently decided to address the issue through her discipline. Professor of Mathematics Moon Duchin founded the Metric Geometry and Gerrymandering Group (MGGG) earlier this semester with the goal of combatting gerrymandering.

As evidenced by its name, MGGG uses metric geometry to identify U.S. districts that may have been gerrymandered. Using a combination of legal knowledge and mathematics, the group hopes to educate others about the issue of gerrymandering as well as the mathematics behind what constitutes a fair district and help create solutions prior to the 2020 U.S. Census, in which congressional redistricting will almost certainly occur.

Duchin explained that MGGG is specifically invested in the prevention of negative racial gerrymandering, citing a gap in mathematical research on that specific topic.

“We’re really interested in racial fairness,” Duchin said. “We think that’s been the aspect of redistricting law that’s received the least mathematical attention.”

Duchin’s interest in gerrymandering began when she decided to discuss the mathematics of redistricting while teaching the class Mathematics of Social Choice

Upon looking into the issue of gerrymandering, she found that the field was not as fully explored as she had thought. For example, she realized that a measure used to police redistricting attempts known as compactness was often only described in nebulous terms that obscured the motivations behind redistricting.

“Only a few states have any kind of definition of compactness, yet about half of the states require it explicitly,” Duchin said.

After Kristen Clarke, an old friend of Duchin’s and president of the Lawyer’s Committee for Civil Rights Under Law (LCCRUL), spoke to Duchin’s class about issues regarding redistricting and gerrymandering, Duchin saw an opportunity.

“I realized that the problem was still interesting for mathematicians and there was a need out in the field for mathematicians to work on court cases,” Duchin said.

More specifically, Duchin said that litigators have noticed lack of available expert witnesses available to testify in court cases regarding gerrymandering.

Shortly after Duchin started the MGGG, it gained traction on Facebook and other social media platforms, which led to the group being featured in media outlets such as the Chronicle of Higher Education and NPR’s Science Friday.

“It’s been amazing,” Duchin said. “We set up a gift account so that people could give money to the project, and the donations have been rolling in from people who want to help out what we’re doing.”

One way to detect which districts are engaged in gerrymandering is to look at how people in those districts voted and check if it clearly benefitted a certain group or agenda. However, this can only be accomplished after an election has already happened, likely as a result of a lawsuit that is calling for fairer redistricting, according to Duchin.

Visiting Assistant Professor Mira Bernstein for the program in Science, Technology and Society (STS), a longtime friend of Duchin and another founding member of MGGG, said the group is therefore basing its findings off of the geometric shapes of districts.

“Our group is not focusing on statistics of which voters voted how, it’s focusing on shapes of districts,” Bernstein said. 

MGGG seeks to compile guidelines for determining whether or not a district has biased boundaries based on its shape.

“The focus is on achieving a systematic and fair understanding of geometric issues related to gerrymandering, as well as how mathematical structures affect how to design fair districting plans,”Justin Solomon, another member of the group and an assistant professor at MIT in electrical engineering and computer science, told the Daily in an email.

Duchin pointed to North Carolina’s 12th congressional district as an example of a district whose shape, which resembles a river, indicates boundaries that were created with biased motives.

Ironically, some of the congressional redistricting that occurs today is facilitated by the Voting Rights Act of 1965, which gave states the power to create majority-minority districts in order to better represent minority groups. However, Duchin said this provision can be manipulated by representatives to actually drown out minority votes, because it allows legislators to create unfairly shaped districts under the pretense of creating majority-minority districts.

“Everybody uses it to cheat,” Duchin said.

A primary goal of the MGGG is to educate mathematicians so that they may serve as expert witnesses in gerrymandering cases. Bernstein said the work done by Clarke and others at the LCCRUL has aided them in these efforts. 

“They [LCCRUL] are our anchor to the real world, where this is necessary,” Bernstein said.

One way in which MGGG hopes to educate both the public and expert witnesses about gerrymandering is through the Geometry of Redistricting summer school program, which will be held at Tufts from Aug. 7 to 11. The workshop will begin with three days of talks given by professors and experts from fields ranging from political science, law, mathematics and computer science, highlighting the interdisciplinary nature of MGGG and its goals.

The final two days will be geared specifically toward expert witnesses who plan to testify in court about gerrymandering and toward high school teachers and college professors who want to incorporate aspects of the project into their own curriculums, according to Duchin.

The group also plans to hold similar programs later in the year in states that have experienced gerrymandering extensively, such as Wisconsin, North Carolina and Texas.

Looking forward, Duchin said she has been contacted by graduate students who want to study what the group has been working on at Tufts.

“I’d like to build something long-term, because I think this project has legs,” Duchin said.

Bernstein elaborated on the group’s long-term goals.

“Our long-term plan is to serve as a hub for people approaching these types of issues … as well as connect people with scientific expertise to lawyers and redistricting committees,” Bernstein said.