The idea of using mathematical algorithms to determine whether electoral districts are fair has gained surprising traction in the past year, including a Jan. 9 federal court ruling that used math to call North Carolina congressional districts biased. Now a lawmaker wants to bring the process to New Hampshire.
Under a proposed bill, HB 1666, a process known as efficiency gap analysis would be applied to statewide districts in New Hampshire after the next redistricting in 2021. If the analysis finds problems, “the redistricting for that elected body shall be deemed to be gerrymandered and therefore not valid” and the districts redrawn before the next election.
The bill is sponsored by Re. Jerry Knirk, D-Freedom, who argues that the method would help all elected officials.
“Partisan gerrymandering decreases competitiveness in the districts leading to disagreements and lawsuits every ten years. The lawsuits are costly and the partisan appearance erodes voter’s faith in the process. This is a bipartisan problem, with both Republicans and Democrats guilty when they are in power,” his prepared statement says. “We do not know which party will actually be in power in 2021 when the next redistricting occurs. That is why we should pass a plan to prevent partisan gerrymandering now when it is in the interest of both Democrats and Republicans.”
The bill is slated for a hearing Jan. 18 in front of the House Election Law Committee. It would apply to state House and Senate districts and those of the executive council and the state’s two members of the U.S. House of Representatives.
Last year Knirk proposed a similar bill to remove redistricting from control of the Legislature and have it done by mathematical algorithm; it died in committee. This year’s bill is more specific, citing a particular method to be used.
In 2016, New Hampshire Public Radio used thuis method, efficiency gap analysis, to look at state senate results since 1994. It found that in 8 of the past 11 elections, “Republicans have enjoyed a disproportionate number of Senate seats compared to actual vote tallies. In two years, the math swung in the Democrats’ direction, and one year was a wash for both parties.”
New Hampshire will be redistricted after the 2020 Census.
In 2016 a state Senate bill sought to create a “non-partisan” committee to do that redistricting, but it was killed.
The idea of using abstract mathematics to draw or test political districts, rather than the judgment of lawmakers or regulators, has drawn attention in the past year, fueled in part by concerns that gerrymandered districts are sidestepping voters’ wishes and undermining confidence in democracy.
Moon Duchin, a mathematics professor at Tufts University, has established programs for mathematicians exploring methods of analyzing the fairness of electoral districts with an eye toward creating more expert witnesses who can testify in trails challenging political boundaries. She was one of the featured speakers at the recent Joint Mathematics Meeting in San Diego, probably the premier gathering of professional mathematicians in the country, and she wasn’t the only person taking this approach. Another session, by two professors from Florida State University, was titled “Monte Carlo measurement of gerrymandering in Pennsylvania through random tessellation,” a description that means little to most non-mathematicians.
More significantly, a federal court on Jan. 9 cited the work of mathematicians including Jonathan Mattingly, an expert in mathematical modeling at Duke University, when ruling that the Congressional districts in the state of North Carolina gave an unfair advantage to Republican candidates.
That state’s GOP-controlled General Assembly has filed an appeal, and the case is likely to end up in the U.S. Supreme Court.
The court heard arguments in October about a similar case brought by Wisconsin Democrats, who say that state’s GOP-drawn districts are illegally gerrymandered.
During oral arguments, justices indicated they don’t think there is an objective way to measure gerrymandering. Mathematicians are trying to develop scientifically objective ways to do that.
The NHPR analysis, like the North Carolina and Wisconsin cases, reflect a political obstacle in this push. The Republican Party dominates control of state legislatures around the country, and since state legislators usually draw political districts any push for taking redistricting away from lawmakers is sometimes seen as an attack on the GOP.
One exception is California, which is politically dominated by Democrats. In that state, voters in 2010 passed a ballot measure that took the process out of the hands of legislators and gave it to an independent redistricting commission.
Another obstacle to mathematization is transparency. Mathematical analysis of political districts is not always easy to understand, leading to suspicions that the methods might actually hide a bias.
The efficiency gap requires counting each party’s “wasted” votes — votes for losing candidates and votes for winning candidates beyond what the candidate needed to win — and dividing them by the total number of votes cast. It attempts to uncover such things as cases where voters from one party are packed into a single district, limiting their ability to elect many people.
Knirk’s bill sets an 8 percent efficiency gap limit for an outside group to examine the boundaries, and a 50 percent limit to draw new boundaries.