Presentation: Jonathan Mattingly and Gregory Herschlag – Gerrymandering

Determining mathematical algorithms in partisan and racial gerrymandering

If you’re like me, and the title of Monday’s program made you tilt your head like a confused puppy, hang in there. After a visit to Wikipedia to refresh my eighth-grade Civics lessons, I recalled that gerrymandering is “a practice intended to establish a political advantage for a particular party or group by manipulating district boundaries.”

Rob Everett introducing our guest speakers

Actually, the title of Monday’s presentation by Duke mathematicians Jonathan Mattingly and Gregory Herschlag was “Sampling the Geopolitical Geometry of a State.” Rotarian and incoming board member Rob Everett introduced the pair, and if these are the types of guys Rob hangs out with, he’s clearly a pay grade or two ahead of me.

In the past two presidential elections, much ado has been made in our state accusing politicians of partisan and racial/socioeconomic gerrymandering. The US Constitution grants states and their legislatures primary authority in determining voting districts; however, it is silent on the issue of state legislative redistricting.

Congressional representatives are apportioned to the states on the basis of population. In 1960, the US Supreme Court issued a series of rulings in an effort to ensure equal legislative representation for all citizens. Most states are required to draw new district lines every 10 years following completion of the census.

In North Carolina, the State Legislature is responsible for drawing district lines, with no veto power for the governor. Our state is currently divided into 13 districts, each represented by a member of the US House of Representatives. By analyzing and changing geographic district boundaries, the balance can be dramatically changed to favor one party, one race, or one class of people.

Mattingly and Herschlag point out that it’s virtually impossible to manipulate one district without affecting its neighbors, thereby causing a cascade of change across a state. “Packing” – whereby a district has a large majority of a single party – is done to dilute the group’s effect in adjacent districts. “Cracking” refers to diluting the power of the opposing party’s supporters across many districts. Keep in mind that there are 3,000 voting precincts in North Carolina, and you begin to get an idea of how impactful gerrymandering can be in political outcomes.

Mattingly and Herschlag’s work contextualizes districts by relating the partisan preferences of a typical precinct’s voters to the preferences of the typical district. For example, a precinct’s voters may prefer one party, but may find themselves in a district in which the majority of votes cast are for the other party.

Quantifying Gerrymandering is a nonpartisan research group centered at Duke’s Mathematics department.  ( The project grew out of undergraduate work projects beginning in 2013 and has since been supported by graduate and undergraduate research; the current team is made up of graduate and post-doctoral students in addition to faculty members.

Mattingly and Herschlag comically subtitled their talk, “How a nice mathematician (like me) ended up in court,” as their research has been expanded to global, statewide analyses. When comparing a districting plan’s partisan makeup with a set of comparison maps, they can obtain a fundamentally global view useful in measuring gerrymandering. This data is becoming increasingly critical across our country in determining equitable outcomes by fairly districting voter characteristics so they are evenly distributed geographically.

Submitted by
Carver C. Weaver

Speak Your Mind