How the Gerrymander Wastes Votes

People concerned about growing partisanship in the electorate, as I am, find solace in responses to two questions on the most recent Marquette University Law Wisconsin poll. The first question asked how state court justices should be chosen:

Some states appoint state supreme court justices, and other states elect them. Do you think it is better to have state supreme court justices appointed or to have justices elected?

As the next graph shows, Republicans, Independents, and Democrats all overwhelmingly opted for the election of justices.

As the next graph shows, the respondents also heavily favored that judicial candidates discuss their views on issues likely to come before them:

Should judicial candidates discuss during campaigns issues likely to come before them if elected so voters know what the candidates stand for or should they avoid talking about such issues in order not to seem to be pre-judging the issues?

The desire to hear from candidates as to their views on issues likely to come before would seem to contradict the demand from some Republican politicians that Janet Protasiewicz recuse herself from considering challenges to Wisconsin’s partisan gerrymander that assures Republicans control the Legislature even if they fail to win a majority of the state vote.

Certainly, having open-minded judges is a good thing. In practice, however, the demand that judicial candidates not say anything of substance has served as a tool for hypocrisy. Maine senator Susan Collins voted to confirm both Neil Gorsuch and Brett Kavanaugh, two of President Donald Trump’s nominees to the Supreme Court under the impression that neither would vote to overturn Roe v. Wade. She said that both men assured her that Roe was “settled as precedent.”

The responses to the two questions are consistent with each other. Voters from both parties want a say in who becomes a judge, and they want information that will help them make an informed choice.

However, that bi-partisan consensus only goes so far. Consider the response to the following question:

A case currently before the state Supreme Court could require maps of the legislative districts for the state Senate and Assembly to be redrawn for upcoming elections. Do you favor redrawing the district maps or should the maps created prior to the 2022 elections remain in place until the next scheduled redistricting in 2031?

The next graph shows the response by Republicans, Independents, and Democrats. Bipartisan agreement now disappears. The same group of voters who agreed on Supreme Court elections and discussion by candidates for judge now widely disagreed depending on their political preference. Republicans opted to keep the map favoring Republican control of the state Legislature, while Democrats wanted it redrawn.

What are the actual and proposed maps? A website named Dave’s Redistricting allows maps to be uploaded and analyzed. Among its reports is one on how well a map performs on four measures:

• Proportionality, how well the map reflects election results. For example if 60% of the voters support a particular party, a proportional map will result in 60% of the districts being won by candidates of that party (also called “fairness”).
• Competitive, based on the number of seats that could be won by either party.
• Minority based on seats where the minority voters make up a majority.
• Compactness.
• Splitting. The more counties are split between districts, the lower the score.

I asked the program to calculate the measure using results from six elections: President in 2016 and 2020, Senator in 2020 and 2022, Governor in 2022, and Attorney General 2022. The results appear in the graph below.

The bars show the ratings for four maps of Assembly districts (bigger is better). The blue column shows ratings for the current map, that passed the Legislature in SB621. This map was passed by the Legislature, vetoed by the governor, and then revived by the former Supreme Court.

The second, rust-colored, columns reflect a map developed by Matt Petering, a member of the engineering faculty at UW-Milwaukee.

The third column results from one of the maps that came out from the People’s Map Commission, established by Governor Tony Evers.

The final, gray column rates a map called  “WI state house fair.” As with Petering’s map, Dave’s Redistricting gave it the highest possible proportionality score. (This analysis was done before Petering published his map.)

Not surprisingly, the current SB621 map has much the poorest proportionality score of the three maps. This is not surprising since the aim of the map adopted following the 2010 census was to assure that Republicans never lost control of the Wisconsin Assembly and Senate, and the current map was updated to reflect population changes while making the “least change” to the previous map.

The next graph shows the ratings for the Senate districts of these same maps. In this case Petering’s map is substantially superior to the current map. It is notable too that that map outscores the others on compactness. One might expect that a higher score on proportionality would come because of odd-shaped districts, but apparently that is not the case.

The final graph shows the “efficiency gap” for these four district maps (including both Assembly and Senate districts). This is based on the concept of “wasted votes.” Votes are considered wasted two ways: First, if a party wins a district, any votes above what is needed to win the district is wasted. Second all votes to losing candidates are wasted. The aim of gerrymandering, then, is to maximize the other party’s wasted votes. This is accomplished by “packing” the other party’s voters into a few districts. This allows other districts to be “cracked” by adding enough voters from one’s party to switch the district to one’s own party.

The next graph shows the efficiency gap for each of the four maps. By custom, efficiency gaps favoring Republicans are given positive numbers; those favoring Democrats are negative. The very large efficiency gap for the present maps confirms that it is virtually impossible for Democrats to win control of either the Senate or the Assembly.

Democracy is based on the concept of majority rule. By entrenching one party to control the Legislature gerrymanders threaten democracy.

Protasiewicz won 55.5% of the vote for Supreme Court, but that was not enough to win most of the assembly’s districts. She carried only 47 of Wisconsin’s 99 assembly districts. Of the 33 senate seats she carried 16, almost—but not quite—a majority.

Applying this same rule to races for state-wide office would require that a Democratic candidate for governor would need to be supported by at least 55.5% of the votes to be elected.

How will the Supreme Court rule on this issue? Such speculation is dangerous, but here goes. It seems to me that the case against the so-called “islands”—the parts of districts completely surrounded by other districts is overwhelming. The claim that the islands should be grandfathered in because they were included in previous district maps strikes me as farfetched. It is especially violates the concept of “originalism” — the concept of interpreting the meaning of the constitution based on the meaning at the time it was written. If the authors of the constitution meant to say that “contiguous” meant close, they could have said so.

If the court rules against the part of the present map establishing the islands, would we still be stuck with the gerrymander that assures continued Republican dominance of the Legislature? At the hearing, supporters of the gerrymander argued that the gerrymander part of the districting plan should be left untouched. That argument strikes me as similar to arguing that if you bring your car to a mechanic because of one problem and other major problems are discovered they shouldn’t be fixed.