If you want to draw a congressional map in Utah, it now has to pass three tests: ensemble analysis, the median-mean difference test and partisan bias.
All three are ways of measuring a map’s fairness, but the last has gotten the most attention.
Utah lawmakers codified those standards into law during an Oct. 6 special session, alongside a new map they say passed all three measures.
But whether the partisan bias test is a good barometer in the Beehive State is up for debate.
“We gave the example of Utah as the state where it would be most crazy to try to use this idea,” said professor Moon Duchin, who teaches computer and data science at the University of Chicago.
She co-authored a 2021 academic study that found the partisan bias and other partisan symmetry tests have flaws when applied in a state like Utah.
“We in our research group spent well over a year trying to figure out how to make that idea work in a way that's manageable for a court or a commission,” she said. “We found that it's just full of problems and paradoxes.”
That’s for two main reasons: one party — Republicans, in Utah’s case — is clearly favored in statewide elections, plus the state has only four congressional districts.
“So you put those two things together, few districts, significant skew, and the metrics really backfire when you try to use them in a simple way,” she said.
“What I mean in the simple way is creating an election index, which is what the law now calls for. As I understand it, you mix together a bunch of elections into one, and you call that an index of all the recent elections, and then you just plug that into these formulas. I wouldn't recommend doing that anywhere.”
The new legislative tests stem from an August court ruling, which declared the 2021 congressional map unconstitutional and forced the Legislature’s hand to redraw the boundaries. Additionally, the ruling left it up to lawmakers to come up with tests and standards to judge how fair any future map is.
The Legislature, wrote Judge Dianna Gibson, “retains discretion in determining what judicial standards are applicable and they retain discretion to determine the ‘best available data and scientific and statistical methods’” to use in the map-making process.
Vassar College assistant professor of statistics Daryl DeFord, another co-author of the 2021 report, said setting metrics for any test can be thorny.
“If you say, well, a map is fair if it passes some threshold, then that sort of incentivizes people to go look for maps that pass that threshold but are advantageous to their party or their group,” he said. “And that's part of where the measure really falls down in a state like Utah.”
According to the report, the “most extreme partisan outcomes” are still a possibility under the partisan bias test and would still receive a passing score. But in Utah, something even more interesting happens.
“A plan that, for example, at the congressional level has four expected Republican seats and no expected Democratic seats gets a good score under these measures,” said DeFord. “But a plan that has, say, three expected Republican seats and one expected Democratic seat is actually … going to look like it's Republican favoring even though it sort of had one more seat going to the other party.
The researchers called it the “Utah Paradox.”
“It's not that anything is being computed wrong, but it shows you that maybe the system shouldn't be used,” Duchin said.
The model tends to work better in a larger state with a more equal partisan makeup, according to Duchin, such as North Carolina.
For DeFord, narrowing the concept of fairness to a specific formula or test is oversimplifying a complex issue.
“The problem is A, these unintended consequences of specifying a particular formula, but also B, this broader question of, like, does that formula actually capture what you mean by fairness?”
Defining fairness in a law also presents its own problem. It can mean something different for everyone. It’s something University of Utah student and new voter Joselin Flores is mindful of.
“I think some people would say that [a certain map] is fair, but I think they would say that because they have a better advantage with their representation, but with other people that are displaced and just not looked at, it's obviously not fair,” she said. “Fair would be subjective to everybody, because everyone has their own definition of fair, and everyone is treated in a different way.”
It’s a feeling shared by 28-year-old Salt Laker and water engineer Evan Watts. He mused about a way to introduce some more checks and balances to the process.
“I'm uncomfortable with the idea that, you know, whoever holds the political majority gets to write the rules that benefit their people, let's say their tribe,” he said. “I don't know what fair looks like, necessarily. I can imagine it's a very difficult job. There's a lot more nuance than just saying, you know, X number of people get X number of representation.”
The fight over Utah’s congressional map will continue in court. In the hours after Utah’s fairness tests became law, attorneys representing plaintiffs in the case against the Legislature filed a complaint arguing that the new law violates the plaintiffs’ constitutional rights.
The next deadline in the redistricting timeline is Oct. 17 for supporting submissions from both parties. A new map must be in place by Nov. 10, according to Lt. Gov. Deidre Henderson’s office, to be used in the 2026 midterms.
