Skip to Main Site Navigation Skip to Content Skip to Footer
Back To Top

Eastern student wins math conference prize for research on gerrymandering

Published on May 15, 2023

Eastern student wins math conference prize for research on gerrymandering

liam with map
Liam Hemingway with a map showing how he redrew Connecticut's congressional districts using 2022 census data

Eastern Connecticut State University senior Liam Hemingway won a top prize at a national conference for his applied math research that looked for a way to draw unbiased congressional districts. In the process, he learned that gerrymandering, or drawing partisan districts, can happen even when you divide a state solely by population.

Hemingway won the spring 2023 Council on Undergraduate Research Mathematical, Computing and Statistical Division student award at the recent National Conference on Undergraduate Research (NCUR) at the University of Wisconsin-Eau Claire. His research, presented there and at Eastern’s CREATE conference and at a Northeastern Mathematical Association of America meeting in New Hampshire, used graph theory to create fair congressional districts.

Liam developed new maps for two states, Georgia, which is heavily gerrymandered, and Connecticut, which is not.

A drawing from the 1800s, left side of slide, depicts the salamander-like pattern that gave gerrymandering its name for distorting congressional districts in Massachusetts

Connecticut's current map, upper left, and Liam's reconstructed map

Hemingway looked at two states: Connecticut and Georgia. Connecticut is considered fairly free of gerrymandering, while Georgia’s districting map is considered to be biased in favor of Republicans. One of the goals of his study was to make “compact” districts, as the law recommends, so that the maps would look more “blobby” and less “snaky” than highly gerrymandered maps. 

“They don’t want people to make these snakelike maneuvers,” he said, referring to the distorted districts that gerrymandering can produce. 

In Connecticut’s case, a couple of sharp hooks were rounded out, but the new map didn’t change the political balance of the state’s five congressional districts. In Georgia, however, the result was an even more partisan map. The algorithm that Hemingway used compacted districts in a way that consolidated Democrats in urban areas, resulting in less representation than they have now, even though the parties aren’t that far apart in how many people claim affiliation with them statewide.

“You could unintentionally make something that’s more biased,” said his faculty mentor, Professor Megan Heenehan, acting department chair for mathematical sciences.

Professor Megan Heenehan and Liam Hemingway with a slide from his research

Adjustments would be needed in further research to account for urban areas attracting more Democrats and rural areas being more Republican, Hemingway noted.  “It was beyond the scope of this project,” he said, but “it gives you ideas of what you could change to take it further,” he said.

Hemingway’s work in math and statistics has gained him admission to a top statistics graduate program at Iowa State University, where he will have a teaching assistantship and will work toward a Ph.D. He’s not sure of his research area there yet, he said, “but this is definitely an option.”

Written by Lucinda Weiss