Tuesday, July 02, 2019 1:00 am
Court's discomfort with analytical math evident in gerrymandering decision
The Supreme Court finished its session last week and issued several important decisions. The one I cared most about involved gerrymandering. Unfortunately, the majority decision shows either a strong partisan bias, which I would hate to admit exists, or a dramatic discomfort with any sort of numerical analysis.
The decision came out of two cases, one from North Carolina and the other from Maryland. In North Carolina, the federal House districts were redrawn to strongly favor Republicans. In 2016, for example, the entire North Carolina House delegation was 23% Democratic, despite 47% of the popular vote being Democratic. (By the way, 2018 election results for the state are not yet complete because of election fraud.)
In Maryland, on the other hand, the House districts were redrawn to strongly favor Democrats. In 2018, for example, the entire House delegation was 13% Republican, despite the popular vote being 32% Republican.
In both cases, the people involved in doing the redistricting openly admitted that their goal was to increase their own party's representation. For example, Rep. David Lewis, chair of the North Carolina state House redistricting committee, said, “I think electing Republicans is better than electing Democrats. So I drew this map to help foster what I think is better for the country.”
The majority decision, written by Chief Justice John Roberts, is a little shocking in how divorced it is from what I thought were common American values, such as free and fair elections and equal representation. The court majority wrote that the principal reason for the decision was that there were not “judicially discoverable and manageable standards for resolving [these cases].” That is, the court majority believed that gerrymandering was too complicated for a court to become involved in.
They also seemed remarkably comfortable with these gerrymandering situations. They wrote, “Experience proves that accurately predicting electoral outcomes is not so simple, either because the plans are based on flawed assumptions about voter preferences and behavior or because demographics and priorities change over time.”
I feel like these people have never actually dealt with accurate, large-scale data before. Ten people, split 6-4 between Democrats and Republicans, will usually but not always have a majority for the Democratic candidate. However, 100,000 people, split with strong confidence 60% to 40% between Democrats and Republicans, will basically always vote in the majority for the Democratic candidate.
Gerrymandering cases in the past were fundamentally different because the split could not be done with nearly the same level of confidence. That confidence has changed because we now have very small-scale demographic information and cheap computing power to calculate different potential district maps. Reading this decision makes me feel like the majority doesn't know the difference between a physical map with lines on it and a computer map with demographic information for every block in a city. This isn't rocket science anymore.
Justice Elena Kagan wrote a dissenting opinion, which is very readable. It is clear and avoids much of arcane language typical of constitutional lawyers. I recommend it.
She refers to a study that semi-randomly created 3,000 electoral maps of North Carolina. Each map used population, physical geography (e.g., rivers), city lines and other standard criteria as potential boundaries. The key here is that the 3,000 different maps did not use any information about party affiliation. The political scientists then used recent election results to simulate previous elections but with the new district maps. All 3,000 maps produced results more balanced than the actual map. Typical results were House delegations that were between 46% and 54% Democratic. As a reminder, the popular vote in 2016 was 47% Democratic and the actual House delegation was 23% Democratic.
The truly sad aspect of this result is that we, as voters, will have a hard time changing gerrymandering. As Kagan writes, “Politicians' incentives here are very different from voters' interests.” The court is most critical in these situations, and it has stepped away from its role.
For example, the previous Supreme Court gerrymandering case came from Wisconsin's state assembly. After the 2018 election, 36% of the assembly is Democratic. The popular vote was 53% Democratic. What possible motivation would the Republican majority have to avoid gerrymandering in this situation?
In the 1960s, the court ruled that House districts had to be of roughly equal population size. That decision was based on the principle of one person, one vote. That is, if one district had dramatically fewer people, the voters in that district would have too much influence. That logic is good and the court appears comfortable with it, in part, because counting is math they are comfortable with. I wish they were also comfortable with modern, computer-based mathematical analysis.
Christer Watson, of Fort Wayne, is a professor of physics at Manchester University. Opinions expressed are his own. He wrote this column for The Journal Gazette, where his columns normally appear the first and third Tuesday of each month.