Fair Redistricting

Aleks said I should link to this article on redistricting by Tim O’Reilly. O’Reilly suggests that computers be used to redistrict automatically. O’Reilly links to this page by Brian Olson. I think this method would be fine. Actually, many different methods would be fine. I think that the U.K. had some sort of nonpartisan boundary commissioners and that worked ok also. Of course the biggest imbalance is the U.S. Senate.

4 thoughts on “Fair Redistricting

  1. This is not the first Computer Scientist who has observed that (A) redistricting is a complete mess and (B) this is the sort of thing that is homework for CompSci undergrads. But, alas, it ain't that easy.

    I had a brief exchange with somebody who is now at Google who had observations (A) and (B) above and an assistant prof at a PoliSci department. The PoliScientist points out two problems: first, parties have to both credibly commit to using a single `objective' redistricting algorithm, and second, that redistricting algorithm can not disenfranchise any groups—and we can't define that term in English, let alone in algorithmic terms. She pointed out that "[t]he parties can't credibly commit to using the districts the computer spit out, because what if the 'perfect' lines end up disenfranchising some particular racial / ethnic group and the state gets sued for it (which happens all the time)?"

    I'd love to see the current system go away as much as everybody else, but there's no reason to think that any redistricting scheme is objective just because it was autogenerated via a transparent algorithm—or that such an algorithm will be objectively administered by humans.

  2. Moreover, the Olson admits that the method relies on some random elements, and spits out different answers. Thus, his choice as to what is best (as far as I can tell) is what is most visually pleasing to him. There are two problems: 1) That is hardly a good standard to choose districts, and 2) the parties will understand that there are multiple "right" answers and will still fight over which is "rightest."

    I suppose it is possible that is still an improvement over hand-drawn districts, but with some many possible right answers, how do the parties agree on seed value for the search process knowing that the possibility exists that they will either significantly advantaged or disadvantaged. Looking at the how visually appealing districts are without thinking about the consequences of those districts is hardly reform, and does not strike me as a responsible way to advocate reform. In some ways the worst of all possible worlds is the ability to create "biased" districting plans while saying the process is unbiased.

  3. "Ah, the genius that results from ignoring a dense, 60 year literature across political science, mathematics, law, economics, and geography. I'm sure if we add Ajax support it will revolutionize our political process."

    Thanks for that Chris, now I have to clean the coffee off my monitor.

Comments are closed.