SMU researchers say Texas redistricting maps show bias
Math for Unbiased Maps created 1.5 million random remaps and none were as biased as the one approved by lawmakers
DALLAS — Researchers at Southern Methodist University say the new congressional maps passed by Texas lawmakers are significantly biased and designed to create a Republican advantage.
Andrea Barreiro is an Associate Professor of Mathematics at SMU and part of the Math for Unbiased Maps group. She says the group produced 1.5 million random maps based on redistricting rules like similar populations and keeping districts contiguous.
The map adopted by the legislature, however, was far beyond even the most grossly biased map of 1.5 million.
“We were actually a bit shocked and felt a bit angry, I think,” Barreiro said. “It was very clearly biased.”
Math for Unbiased Maps started last spring to study rescaling using what’s called the Markov Chain Monte Carlo (MCMC) method to guide a random search. The goal was to see how gerrymandered the cards were, where they are deliberately set to give a party an edge.
“We wanted fair cards as voters and as citizens,” she said.
Redistricting only takes place every 10 years when new census data is released. This year, Texas added two congressional seats, and Barreiro says there are likely billions of ways to carve up the state, even bound by legal mapping guidelines.
Census data shows that 95% of Texas’ growth is due to communities of color that vote Democratic. This includes 1.9 million new Hispanic residents, nearly 10 times the number of non-Hispanic white residents added to the state.
“That will determine the shape of power for the entire decade,” Michael Li, senior attorney at NYU’s Brennan Center, told the WFAA last fall. “It’s something that’s difficult, but I like to call the redrawing of the Olympics every 10 years.”
Li said Texas has already been flagged and faced legal challenges for intentional discrimination.
They tested both the new map of state senate districts and congressional districts. The 1.5 million maps created gave them a working basis.
The median model would give Democrats a two-district advantage, but the approved maps instead gave Republicans a 12-district advantage.
“Excessive political gerrymandering hurts the voter,” Barreiro said. “It makes racing less competitive. He favors one political party over another. He may favor one racial group over another.
This is done in two ways. Either the districts are lumped together where all of a party’s voters are placed in the same district so their voters have less impact, or they are “split” meaning they are diluted with so many voters of the other party that their votes count less. Barreiro says Texas in North Texas, Beth Van Duyne’s district has turned redder and Collin Allred’s district has turned bluer, meaning primaries matter more than election days.
“Essentially, the outcome is predetermined almost before you show up at the voting booth in November,” she said. “These quarters swirl around each other like a cinnamon roll, like what’s the logic in that.”
The group also reviewed the new Dallas City Council map. The city is heavily Democratic, so partisan division isn’t an issue, but some options had carved out different demographic segments before a final map solved the problem according to Math for Unbiased Maps.
Better technology is making the problem worse than it was ten years ago.
“The problem got worse from 2010 to 2020 and even if you look at the maps you’ll see they’re drawn on a much finer scale,” Barreiro said. “It hurts voters on both sides, and we can demonstrate that it happened in the Texas legislature this year.”