The rules of the game are set. Now the scoring begins.
In a 4-3 decision, the Wisconsin Supreme Court recently laid down two key guidelines that are already having a major impact on the outcome of the contest over redrawing the state’s legislative and congressional maps.
The high court’s Nov. 30 opinion dealt a one-two punch to gerrymandering opponents. In drawing new district lines, justices said they would not consider what impact those lines would have on the balance of power between the two major political parties. They also said they would make as few changes as possible to the current maps — maps that have given Republicans an almost-unbreakable hold on the state Legislature for the past decade.
“Partisan fairness” has become a rallying cry for activists pushing to eliminate gerrymandering. In many states, including Wisconsin, those activists are progressives who were energized by maps drawn to cement GOP majorities after the “red wave” elections of 2010 gave Republicans control of 2011 redistricting. But in some blue states, such as Illinois, the script flips, and political conservatives complain about Democratic gerrymandering.
In key court cases in recent years, the activists argued that gerrymandering violated their rights to free expression and equal protection of the laws under the U.S. Constitution. However, in its 2019 Rucho v. Common Cause decision, the U.S. Supreme Court ruled that federal courts have no business deciding partisan-gerrymandering claims. The justices left open the possibility that state courts could act on those cases. But in its Nov. 30 opinion, the closely divided Wisconsin Supreme Court decided that such claims are out of bounds for this state’s courts as well.
“Claims of political unfairness in the maps present political questions, not legal ones,” Justice Rebecca Bradley wrote for the high court’s majority. “Such claims have no basis in the constitution or any other law and therefore must be resolved through the political process and not by the judiciary.”
In her dissent, Justice Rebecca Dallet termed the majority decision “an unnecessary and sweeping overreach” that “likely has far-reaching consequences” for the future of representative government in Wisconsin. Writing for the minority, Dallet warned that “willfully blinding the court to the partisan makeup of districts increases the risk that we will adopt a partisan gerrymander.”
The majority also endorsed the idea that a court-drawn map should revise the current map only as much as legally necessary—the least-change position advocated by Republican leaders and the Wisconsin Institute for Law and Liberty, the conservative legal organization that represented clients who brought the state case. Bradley characterized this as an exercise of judicial restraint, writing, “Because the judiciary lacks the lawmaking power constitutionally conferred on the Legislature, we will limit our remedy to achieving compliance with the law rather than imposing policy choices.”
Dallet contended that applying a least-change approach to the “sharply partisan” 2011 maps amounted to “inserting the court directly into politics by ratifying outdated partisan political choices. . . . Try as it might, the majority is fooling no one by proclaiming its decision is neutral and apolitical.” She also accused the majority of inaccurately claiming the approach was commonly used when it has no precedent in similar cases.
Writing separately, Justice Brian Hagedorn concurred with almost all of Bradley’s opinion, but suggested a bit more flexibility in the least-change approach. In deciding between proposed maps that meet all legal requirements, he said the justices could consider other traditional redistricting criteria, such as keeping communities of interest intact. “While a remedy must be tailored to curing legal violations, a court is not necessarily limited to considering legal rights and requirements alone when formulating a remedy,” Hagedorn wrote.
Because Hagedorn has been a swing vote in some high-profile cases, his position opens a narrow window for compromise on the sharply divided court.
Still, the combination of Rucho and the latest Wisconsin ruling box in opponents of political gerrymandering, leaving them no judicial recourse, in either state or federal courts. If judges at either level ultimately draw the maps, as seems likely at the moment, partisan fairness won’t be a factor.
But that won’t end the political debate, and even the courtroom battle may be far from over.
In proposed maps submitted to the court Dec. 15, Democratic Gov. Tony Evers, progressive forces, voting rights groups, and scholars offered options that would reduce the size of Republican majorities while, they contended, remaining within the court’s least-change parameters. Meanwhile, GOP leaders offered maps that would solidify their hold on power — the same maps that they had pushed through the Legislature, only to be vetoed by Evers.
In drawing district lines, “partisan fairness” itself can be defined in different ways, but it generally involves estimating what percentage of seats each party would win in a hypothetical 50-50 statewide vote at the top of the ticket, and how those numbers would change in response to shifting statewide vote outcomes.
With legislative and congressional redistricting underway in statehouses around the country, several national websites offer online ratings of various maps—or tools to come up with your own rating—based on the degree of partisan fairness or bias in each proposal.
Among the ratings websites is the Princeton Gerrymandering Project, which handed out “A”s for partisan fairness to the Wisconsin Assembly, state Senate, and U.S. House redistricting plans proposed by Evers’ People’s Maps Commission but rejected by the Legislature.
That advisory commission is a new addition to the state’s redistricting scene, with no official standing beyond the executive order that Evers used to create it. He advocated its maps in court until the Nov. 30 ruling made that position untenable, leading him to submit his own maps.
The Princeton site gives “F”s to maps drawn up by GOP leaders. Those maps are largely based on current district lines.
Another website, Dave’s Redistricting, offers ratings but not letter grades, while a third site, PlanScore, allows users to upload proposed maps in order to generate their own ratings. Like the Princeton site, both Dave’s Redistricting and PlanScore find strong pro-GOP biases in the Republican-drawn maps and greater partisan fairness in the advisory commission’s maps, although they use significantly different methodologies that produce different results.
While calculating partisan leanings in a district map involves a lot of math, it’s hardly an exact science, according to John D. Johnson, research fellow in the Lubar Center for Public Policy Research and Civic Education at Marquette Law School. Data issues and subjective judgments can produce significant differences between different approaches, Johnson says.
All such analyses involve reorganizing past election results into proposed new districts. Municipalities typically tally votes by “reporting units,” which group one or more voting wards together, often because they share a polling place. But unlike legislative and congressional districts, ward lines can change over the course of a decade, usually as a result of municipal annexations, and reporting units can change from one election to the next.
Wisconsin’s Legislative Technology Services Bureau tries to keep that all straight by updating statewide ward maps twice a year and reallocating past election results from the old wards to the new wards. However, the difficulty of that process results in some approximation in the ward-level estimates where boundaries changed.
Once the data are in hand, a redistricting analyst must choose which data to use. That involves deciding which past elections can best predict future trends.
“All measurements of ‘partisan lean’ stem from the idea that the future electorate will behave like the past electorate,” Johnson says. “But we know that the electorate changes from one election to another.”
Those changes can reflect the differences between high-turnout presidential elections and lower-turnout midterms, or long-term shifts in regional political preferences, or strong feelings for or against particular candidates, Johnson says.
For example, Democratic U.S. Sen. Tammy Baldwin’s 2018 re-election by a near 11-point margin “was a remarkable outlier in Wisconsin’s recent political history” of very close statewide elections, Johnson says. When the Princeton analysts averaged that race’s results with those of the 2018 gubernatorial and 2020 presidential contests, they calculated that Wisconsin leans Democratic by 4.2 percentage points statewide. But when they changed their formula to exclude the 2018 senatorial election, they found the state leaning blue by just 0.9 percentage points, a more accurate reflection of its battleground status.
Johnson’s own Lubar Center analysis averages the results of three elections—2016 presidential, 2018 gubernatorial, and 2020 presidential—to produce an even narrower split of only 0.3 points in Democrats’ favor, virtually a tie with Republicans.
By contrast, Dave’s Redistricting, developed by Seattle software engineer Dave Bradlee, averages results of six elections—2016 and 2020 presidential, 2016 and 2018 senatorial, 2018 gubernatorial, and 2018 attorney general—to come up with a statewide lean of 1.5 points blue.
PlanScore, a project of the Campaign Legal Center, uses what Johnson calls “by far the most complex methodology,” with “significantly different results than the other methods.” It deploys higher mathematics to factor in presidential, congressional, and legislative election results over the past 10 years, producing a statewide lean of 1.7 points red.
The PlanScore model incorporates data from the past decade of elections, but this causes it to less accurately reflect the current political geography. That is because rural areas have become more Republican and suburbs have become more Democratic in recent elections, Johnson says. The website is testing a new model to compensate for that issue.
That political geography also means that statewide leanings don’t translate directly into a proportionate number of legislative seats for each party, Johnson has noted. With Democrats packed into larger cities, the more widely dispersed Republicans have an advantage in winning more districts, he has written.
Even with the baked-in GOP advantage, all four methods (including Johnson’s) found that the Republican-drawn map would skew much more red than inevitable—and perhaps even more than the current map, which Democrats have denounced as an extreme partisan gerrymander.
Republicans currently hold 61 of 99 Assembly seats, 21 of 33 Senate seats, and 5 of the state’s 8 U.S. House seats. To demonstrate how the four methods differ in projecting the future partisan breakdown, Johnson applied all of them to maps proposed for Wisconsin. His analysis uses the original PlanScore method.
All four methods find that Democrats would have an edge in just two congressional districts under the Republican map. But for the advisory commission’s map, three methods show at least a slight Democratic lean in four districts, while PlanScore finds only two blue districts.
Of the congressional maps submitted Dec. 15, three methods find two blue seats each in Evers‘ map and the map from progressive voters represented by Democratic attorney Marc Elias’ firm, while Dave’s Redistricting projects three Democratic seats in each of those maps. PlanScore predicts two blue seats would result from a map submitted by a group of professors who call themselves the Citizen Mathematicians and Scientists. while the Lubar Center and Princeton project three and Dave’s Redistricting finds four.
Similar differences show up in analyzing the legislative maps drawn by GOP leaders; the governor’s advisory commission; Evers himself; Elias’ clients; voting rights groups represented by the progressive legal organization Law Forward; the professors’ group; and the conservative group Common Sense Wisconsin (which did not submit its map in court). The original PlanScore method consistently finds fewer Democratic seats under each Assembly plan than do the other rating sites.
For the Assembly:
- The Republicans’ map would create 34 Democratic-leaning seats, according to PlanScore, while the other ratings methods project a number from 36 to 38.
- Evers’ map would yield 36 blue-leaning districts, PlanScore estimates, while the other three methods predict 39 to 44.
- The advisory commission’s map would produce 40 blue-leaning seats in PlanScore’s analysis, compared with projections of 44 to 47 under the other three ratings systems.
- The Common Sense map would draw 35 Democratic-leaning districts, to judge from PlanScore, while the other ratings methods project it to yield 39 to 42.
- PlanScore also predicts the Elias map would create 35 blue-leaning seats, compared with projections of 39 to 43 from the other three methods.
- The Law Forward map also would produce 35 Democratic-leaning districts by the PlanScore calculation, while the other three methods all agree it would create 42.
- PlanScore also finds 35 blue-leaning seats in the scholars’ map, while the other three methods projected 40 to 43.
For the Senate:
- The advisory commission’s Senate map would create 10 blue-leaning seats, according to PlanScore, compared with 15 or 16 projected by the other three methods.
- PlanScore also predicts the scholars’ map would produce 10 Democratic-leaning seats, while the other three methods project 11 to 14.
- PlanScore agrees with the Princeton and Lubar Center analyses that Democrats would have an edge for just 10 Senate seats under both the GOP and Common Sense maps. By contrast, the Dave’s Redistricting method projects 11 blue-leaning Senate seats under the Republican map and 12 under the Common Sense map.
- The Lubar Center, PlanScore and Princeton analyses agree that maps from Evers, Law Forward and Elias all would draw 11 Democratic-leaning districts. Dave’s Redistricting concurs with that prediction for the Elias map, but projects 12 for the Law Forward map and 13 for Evers’ map.
“PlanScore (in its original form) mainly just projects the map as shifting several shades more red,” Johnson says. “In any case, using the past to make predictions about the future is always rife with uncertainty.”
One certainty is that the map drawn by the state Supreme Court will look a lot like the current map—but maybe not as much as Republicans want.
Although Republican lawmakers claimed their own new map embodied a least-change approach, they appear to have made some changes specifically to protect incumbents rather than to comply with legal requirements. The justices rejected a GOP request to defer to the legislatively approved map, noting that Evers’ veto had prevented it from becoming law.
All parties will get a chance to present their cases in a hearing that would start Jan. 18 and could last up to four days—potentially the high court’s longest hearing ever.