Its results so far suggest the district lines North Carolina used in its U.S. House races in 2012 and 2016 were “highly atypical” in how they performed on the way to giving Republicans a lopsided share of the state’s delegation, said Justin Luo, a student who’s working on the project with math professor Jonathan Mattingly.
The team’s work has progressed since an initial paper in 2014 and now addresses elections in other states.
So far, Mattingly and his colleagues have looked at eight states, and the district lines in Maryland also stand out as lopsided, favoring Democrats who’d likely hold a majority of the state’s delegation in any case but because of them have avoided facing a tough race or two.
“Our tool is completely nonpartisan,” said Greg Herschlag, a visiting professor who’s also on Mattingly’s team. “We have ability to make statements about whether something is gerrymandered in favor of the Republicans, or in favor of the Democrats. We can go both ways.”
Luo, Herschlag, Mattingly and other team members staged a public briefing on their work Wednesday, at the Sanford School of Public Policy.
The move was timely as redistricting is one of the state’s and country’s hottest political topics.
Congressional or state legislative, sometimes both, face court challenges in at least three states, North Carolina included. This week saw a three-judge panel order new elections next year for the N.C. General Assembly, on the grounds the district boundaries used since 2012 are a “racial gerrymander” that violates the U.S. Constitution.
Another case, from Wisconsin, recently saw another three-judge panel toss a district map for the state legislature that a majority said had obviously stacked the deck for Republicans “under any likely future electoral scenario for the remainder of the decade.”
The Wisconsin case leaned heavily on mathematical modeling, similarly to the Mattingly team’s, that advocates for striking the boundaries argued show how skewed they are.
Mattingly said his group’s model is different, but could be adjusted to consider whatever factors researchers think important.
What matters is that it’s possible to develop tools for gauging skews, something “that’s been a struggle for the courts,” said Tom Ross, a Sanford School fellow, former state judge and former president of the UNC system.
Mattingly’s first article on the researcher, the one in 2014, was co-authored with former student Christy Vaughn.
They developed a model that relies on computers to generate random district maps, subject to the proviso the districts be relatively compact and that they each include as near as possible to 1/13th of the state’s population.
Now at Princeton, Vaughn and her boyfriend “stayed up a week to hand-code” the original precinct data, Mattingly said.
Mattingly and Vaughn then used the computer-generated districts and 2012’s precinct-level election results to see how many races Democrats and Republicans would have won, on average. They reckoned the Democrats’ statewide majority of the popular vote that year should have yielded the party’s candidates anywhere from six to nine seats in the U.S. House.
In real life, the lines the GOP-controlled General Assembly drew for Congressional races wound up giving Democrats four seats in 2012.
The subsequent refinements to the Duke team’s project suggest there’s little surprise in that. The GOP so heavily packed Democrats into three districts that U.S. Reps. David Price, D-4th, G.K. Butterfield, D-1st, and Mel Watt, D-12th won in runaways. A fourth Democrat, former U.S. Rep. Mike McIntyre, barely hung on that year in the 7th District, only for the party to lose it in subsequent elections.
The lines used in House races this year in North Carolina, redrawn on court order, are somewhat less one-sided but still skewed, said Herschlag, who earned his doctorate in math at UNC in 2013.
The team has posted a lay-language explanation of its latest work online.
It said there’s a common feature to the skews Mattingly, Herschlag and the students have found so far.
“A partisan legislature was in charge of redistricting in five of the eight states we have studied,” it said. “In each case, nearly every district that had the potential to be competitive went to the party in control of the legislature. The party in power achieved this by packing the opposing party's voters into as few districts as possible.”
In two of the three, Iowa and Arizona, district-drawing decisions are in the hands of commissions independent of state legislatures.
The third, New York, had a between-chambers split in party control of its state assembly and came up with a plan that allowed “the most potentially competitive districts” to play out that way in real life, while awarding the dominant party in the less-competitive ones easier rides than could be the case.
Ross’s fellowship at the Sanford School has served to highlight the utility of commission-driven redistricting.
He lined up a group of former judges to game out how the process might work. They came up with a map that, given recent voting patterns, would favor Republicans somewhat but likely generate closely-fought battles in three of the state’s 13 Congressional districts.
The Mattingly team said the judges’ map doesn’t appear skewed.
But in its web article, it cautions the redistricting process isn’t a panacea for the country’s polarization.
“The urban-rural divide in the American electorate means that in most states, there can only be a handful of districts that can have the potential to be competitive, regardless of how [they’re] are drawn,” it said.
In North Carolina, the key decision is how to split Mecklenburg and Wake counties, both of which have bigger populations than can fit in a single Congressional district.
“Given that they have to be split, how they’re split is an open question,” Herschlag said, observing that they “also tend to be heavily Democratic.”
©2016 The Herald-Sun (Durham, N.C.) Distributed by Tribune Content Agency, LLC.
