Statistical Tests to End the Curse of Gerrymandering

Link to Sam Wang's December 5, 2015 New York Times article "Let Math Save Our Democracy"

Link to Sam Wang's December 5, 2015 New York Times article "Let Math Save Our Democracy"

Sam Wang begins his December 5, 2015 New York Times op-ed "Let Math Save Our Democracy" (flagged by a current article about the upcoming 2020 census) with the words 

Partisan gerrymandering is an offense to democracy. It creates districts that are skewed and uncompetitive, denying voters the ability to elect representatives who fairly reflect their views.

I expressed a similar view in my early post Persuasion

Many people may not realize the extent to which political polarization in the House of Representatives arises from partisan and pro-incumbent redistricting. When electoral districts are designed to be either safe Republican or safe Democratic districts, then the main fear for a politician seeking reelection is losing in the primary. That typically pulls members of the House of Representatives toward the extremes. Nonpartisan redistricting is a way to have more districts be competitive in the general election and so make those running for Congress worry more about the general election relative to how much they worry about the primary. I believe this would pull politicians toward to center and toward a greater willingness to work with those in the other party. Getting change to happen in this area will be hard, but there are groups already working on this. I believe the long-run value to our Republic of nonpartisan redistricting would be substantial.

In my more recent post "Nonpartisan Redistricting," I added:

In other words since most (all but about 40 of 435) Congressional districts are designed to be safe for one party or the other, those in Congress often take actions to please their bases rather than the center. That in turn tends to push Congress toward being more of an arena of posturing rather and less of an arena for deliberating about helpful legislation.  

So I am pleased that the Supreme Court upheld the constitutionality of redistricting commissions. (I am not sure the text of the Constitution fully supports this decision, so ideally it would be good to have a constitutional amendment declaring them legal–and going beyond that, requiring them for all states.)

I hope with the issue of constitutionality settled that more and more states adopt redistricting commissions. Though this may involve short-run sacrifices of reelection probabilities, I think this is actually even in the long-run interest of a party in control of a given state, since parties that get used to appealing to the center to a greater degree are likely to grow in influence.

Sam Wang gives another ray of hope that things can improve even in states that continue to have legislative redistricting:

Simple criteria for identifying gerrymanders would be of great use. The Supreme Court has never rejected a voting district for giving a political party an advantage. ...

A majority of the court, however, supports the idea of finding a test that measures partisan asymmetry. Justice Anthony M. Kennedy, who in the past has provided the deciding vote on this question, has stated his desire to find a workable means of identifying partisan gerrymanders. Either the average-median difference or the lopsided-margin test could serve such a function. Justice Kennedy can have them both.

The trouble with Sam Wang's tests is that they focus on the effect of gerrymandering on overall partisan balance, rather than the number of competitive districts. To me, the number of competitive districts is every bit as important as the overall partisan balance. So a third statistical test is needed: one the counts the number of districts with with close to a 50% vote share for each party. This would force candidates in those districts to try to appeal to independent voters and even voters of the other party, not just voters of their own party.