Ever since the district boundaries for the assembly and state Senate districts were devised in 2011, Wisconsin Democrats have been complaining about the unfairness of the district boundaries and their influence on the makeup of the state assembly and senate. Prior to that republicans complained about the reverse unfairness imposed by Democrats. Clearly, the state needs a fair procedure; Step one is an examination of key terms.
Influence of District Boundaries
The district boundaries have great influence over not one but two outcomes. First, they influence the competitiveness within a district, and can even be crafted to make a district "safe" for a candidate of the favored party. Second, they influence the statewide array of wins and losses, which determines how many seats each party wins within the 99 seats of the Assembly and the 33 seats in the State Senate. Shifts in district boundaries can influence which party will gain a majority in the Assembly or the Senate or both. In turn, majority status provides access to the tools of legislative power, i.e., committee assignments, the flow of legislation, as well as whether to open or close a legislative session and to permanently "table" legislation. The majority can also determine what can go on the advisory ballots so the voters can express a more direct opinion.
In the current case, the legislature seat count is far less than proportional to the Democratic vote count in statewide elections. The Democrats hold 36% of the Assembly seats while in state-wide elections -- i.e., those that cannot be gerrymandered -- the voters expressed a preference for Democrats by over 50%.
Non-representative government has led to unrepresentative policy: Post-Dobbs reversion to the antiquated 1849 state law on reproductive rights; blocking a referendum to let the voters express their preferences on reproductive law as did the voters in Kansas and Ohio; reducing state funding per student in the UW system to 46th in the nation, including refusal to fund the much-needed Engineering building for the UW-Madison campus; and continuing to reject federal money for railroads and Medicaid, thereby costing Wisconsin taxpayers billions.
Standard requirements of fairness
Modern design of district boundaries must meet a complex array of requirements. The districts must have roughly the same population; they must be compact (not stretched out like a salamander); they must be contiguous; and they must adhere to the requirements of the federal voting rights act. To simultaneously meet all these requirements requires use of computer algorithms. Fortunately, there are experts in the design and use of these algorithms at Duke, Princeton, and University of Wisconsin - Milwaukee where Professor Matthew Petering has established District Solutions LLC to house his efforts to create fair maps through his FASTMAP algorithm.
Tools of Gerrymandering. Packing and Cracking
Fair maps would reverse the anti-representative effect of gerrymandering. Therefore, to understand fair maps requires understanding the two key tools of gerrymandering: packing and cracking. Both of these strategies involve shifting district boundaries that affect the margin of victory within a district, and consequently affect the proportion of legislative seats for each party.
Packing refers to opportunistically shifting some district boundaries to corral an excess of Democratic voters, ensuring that within that district the Democratic candidate will win far more votes than needed to win that seat, thereby "wasting" the unnecessary votes. Cracking refers to shifting some Democratic voters into districts where Republican candidates typically win by large margins. The idea here is to subtract Dem votes from a district to convert it from a win to a loss and to spend those votes for the Democratic candidate in a district the Republican is expected to win, making sure that the shifted votes are not enough for the Dem to win there. Instead, the Republican candidate still wins, albeit by a smaller margin, "wasting" fewer Republican votes and more Democratic votes.
First, a simple definition of proportionality. Proportionality in the design of district boundaries yields a party a 50-50 chance of the same proportion of seats in the Assembly and State Senate as that party's proportion in statewide elections. Under this definition of fairness, a party with, say, 30 percent of the state-wide vote should have a fifty-fifty chance at 30 percent of the seats. Or, as in the present case, since the Democrats typically earn around 51 -54 percent of the state-wide vote, they should have a fifty-fifty chance of that proportion of the legislative seats.
Second, algorithms can construct maps on the basis of the usual requirements plus proportional representation. Its maps project a fair outcome with Democrats having a 50-50 chance of earning a majority in either chamber, or both, in proportion to the statewide vote. Moreover, the algorithm can rate other maps on the basis of the proportionality standard, as well as any of the other necessary requirements. For instance, the District Solutions algorithm analyzed the map proposed by the Wisconsin Senate Democrats in November 2021 and projects that map would provide the Democrats with just a 1.7% chance of winning a majority in the Assembly and an 18.9% chance of a majority in the state Senate.
Third, two learned scholars of the problem of fair maps, Judge Lynn Adelman and Fred Kessler, have opined that any construction of fair maps must include the use of data on where voters live. (see Further Readings below). Why? Democratic voters are more clustered in cities than the more widely dispersed Republican voters. Consequently, if the algorithm is not provided residency data, it will operate as if voters had random locations and would perform its task of overlaying district boundaries only according to the other requirements of fairness. District Solutions LLC provides quantitative demonstration of the political and legal conclusion of Kessler and Adelman: Without residency data, readily available from recent state-wide elections, the algorithm cannot be calibrated to construct maps that produce proportional representation for voters of both parties.
To highlight the importance of using residency data, District Solutions has used its algorithm to estimate results obtainable without using the data. They show that, despite winning roughly 53% of the statewide vote, the most likely outcome for Democratic voters would be to win about 45% of the seats with very low likelihood of a majority in either chamber. While that is better than the current allocation of 36% of seats, it falls well short of "fair."
In other words, the use of voter data is not partisan; it is the failure to use voter data that results in partisan bias. Proceeding without such data would yield maps biased in favor of Republicans and would preserve most of the imbalance now built into the current array of maps, as we see in neighboring Iowa (see Thompson in the "Further Readings" below).
"Political Fairness in Redistricting: What Wisconsin’s Experience Teaches," The University of Memphis Law Review, Vol 49, 2019, Judge Lynn Adelman
"How to Draw Fair Maps in Wisconsin," Capitol Times, Nov 14, 2020. Fred Kessler, former legislator and re-districting expert.
"Why That 'Iowa Redistricting Plan Was Unfair" by Bruce Thompson, Urban Milwaukee, Oct 4, 2023.)
District Solutions LLC www.districtsolutions.net, Amicus Brief of Matthew Petering, Clerk of Wisconsin Supreme Court 11-08-23