Gerrymandering, Math and You
I had known peripherally about gerrymandering since a great high school government class. I’ve been scared by the increasing schism between the two American political parties, and I have wanted a way to help return the country to moderation. A note about the first Metric Geometry and Gerrymandering Group workshop in Boston came across in one of my mailing lists. I thought I’d finally found a good way to apply my nerd skills to help restore some moderation to the country — perhaps some kind of tool to automatically draw district lines. I learned much more than I was hoping at the MGGG workshop in Boston.
I learned about the legal framework of redistricting, in terms of the constitutional amendments and historical Supreme Court decisions. Many people have spent a great deal of time arguing that gerrymandering is unfair, but it is mostly ultimately legal. The main exception, thanks to the Voting Rights Act of 1965, is when it can be proven to hurt a racial minority. Unfortunately, that burden of proof is high, and there’s no legal barrier to putting bad rules in place in the first place. With section 5 of the Voting Rights Act, there were barriers to any changes that affected voting in areas with histories of discriminatory practices, but that section was gutted in the supreme court’s Vieth v. Jubelirer ruling. Since then, harm has to happen before anything can be legally done about it. Preventing harm requires carefully monitoring and staying involved in any redistricting or legislative efforts, and protesting changes before they go into effect.
I learned that automatically drawing districts is not a good way to do things. There are too many competing concerns, and there will always be the question of who programmed the computer with what conditions to optimize. Instead, computers can be used for analyzing drawn districts and enabling people to draw maps themselves. Each of these are powerful tools for challenging legislators on their schemes. Computers are also critically important for analyzing geographical data to make cases for specific populations being hurt by proposed legislative changes.
And finally, I learned fascinating things about how the responsiveness of an election and the representativeness of the population can compete with one another, and we must choose an appropriate compromise. These can be seen on vote-seat curves, where the slope of the curve is higher for more competitive elections.
In more competitive elections, because a small change in the electorate can swing results more strongly, minorities can have more impact. However, when the vote-seat curve has a slope other than 1, by definition, the seats are not made up exactly according to votes cast. That compromise is the crux of many of the metrics being debated before the supreme court right now. This was wonderfully unintuitive to me, and fun to learn.
I’ve been excited to share what I learned at MGGG in Boston, but I’m even more excited for the coming workshop in Austin. This will be held February 1–4, 2018, at UT. Registration opens Dec 1.
A note from Open Austin: Like this post?
This is a follow up blog post from our Open Gov & Civic Tech Meetup on Gerrymandering & redistricting where Michael Sarahan was the speaker.
Join us on Slack to continue the conversation.
