I saw this article recently and it definitely gave me pause.
It is now common knowledge that polarization is at some of its highest levels in recent decades. Each party has largely been scrubbed of those that do not tow the overall party line, which has led to more homogenous parties. Naturally, this has also pushed the parties more to the extremes. Primaries do not help – as all candidates have to pass through a trial by fire from a narrower political spectrum compared to the past. (For a great visual of this, see XKCD – GREAT.) There seems to a natural ebb and flow to polarization, but we are at an extreme.
I have to admit that I was surprised. I do not think I had ever really encountered an article that took a data centric view at whether gerrymandering has actually changed elections. It is oddly reassuring to have that point asserted.
Gerrymandering is one of those points that is so infuriating no matter what side of the political spectrum one falls on. Districts such as Louise Slaughters’ “earmuffs” between Buffalo and Rochester are case in point. Who would come to the conclusion that this makes sense to represent a certain population effectively? District lines are arbitrary, and when the power to make them is given to individuals, especially state legislatures, the tendency to use that power for self-benefit is overwhelming. Incumbents love to have their seats manipulated to be more safe.
Without doing enough research to know precisely what options and policies I would support in this area, I do know that I would support a more rules based approach. On what criteria? I am not sure. Some limit on the ratio of the border length to the area of the district (to prevent massively concave and ridiculous geometry) would be one potential idea. We could easily create an automated system of manipulating districts that would be more fair and less under the control of party bosses. Whether or not we could ever agree on what that system should optimize for is another question all together.