This post was mentioned on Twitter by StevenLandsburg: How to fix politics: http://www.thebigquestions.com/2010/03/09/fixing-elections/…

Divide congressional districts within a state based on birthdays. I like this better than the last name division, because it can’t be legally changed and because last names starting with X and Z are disproportionately Asian, as well as other more subtle skewing. It solves the problem of upstate New York v. NYC. It doesn’t solve the problem of Texas v. New York.

It’s no secret that the House is vastly more polarized than the Senate. This moves the House up to the level of the Senate. A small step, yes, but it happens through a realistic rule change.

It doesn’t require changing the Constitution (unlike 1), it doesn’t require cooperation among all 50 states to get started (unlike 2), and it doesn’t require massive procedural changes impacting more than special interests (unlike 3.)

While it’s not a comprehensive reform, I can see it actually happening. Even if only 10 states did it, that would be progress. There’s something to be said for wild ideas. There’s also something to be said for trying to figure out ways to make the brilliance not seem wild, even at the expense of dampening the brilliance.

These are good ideas.

The Steve’s alphabet option is intersting but has a significant downside: it nationalizes elections in every district. Every candidate would have to advertise and campaign across the country to reach every voter in her alphabet district. This would be far more expensive than local campaigns in individual districts or states, greatly increasing the power of money. All other options that ignore geography face the same problem.

There are many redistricting techniques used today to protect incumbents or benefit one party or the other, too many to mention. Overcoming these devices is straightforward if redistricting is taken away from state legislatures and given to independent commissions constrained by requirements to make districts of equal population and compact, with boundaries following natural geographic and administrative boundaries, and prohibiting gerrymandering.

Steve is correct that the goal should be to increase diversity and the competitiveness of districts.

Here are some starting points for automated congressional redistricting (some links to other pages on the bottom, but there are example maps on the page itself):

http://bolson.org/dist/ (nice pictures)
http://rangevoting.org/SplitLR.html (a version called splitline districting)

My personal preference would be having congressional districts that were “centroidal voronoi tessellations” of equal population (if you do a Google search for “centroidal voronoi” you will see what I mean). However, a practical, low-tech way to get good districts would be to allow each major party (and any group that had enough signatures on a petition) to offer a districting map, and then have people vote for the one they wanted. I think most people would instinctively choose one that is more visually compact and less gerrymandered.

I would hold congress members responsible for their votes on the order of the following:

Every law would have a stated outcome with a defined measurement of success. All who voted for the law would incur a significant financial penalty as long as the law was in force without meeting its stated goal.

I guess we don’t want to minimize the amount of land required, since that can also result in funny shaped districts, i.e. a district 100 yards wide and miles and miles long. Perhaps we could impose a rectangular shape with a given ratio of length to width? Would something like that work?

Is there some way to design districts mathematically so that each district minimizes the amount of land required for whatever the number of constituents reside in each congressional district? Shouldn’t that eliminate, or at least improve, the weird shapes that now show themselves as congressional districts?

To me, the point of the piece is pretty clear, and it’s not “here’s my radical solution to our problem”. Rather, it’s simply a way to get people to think about how the Senate works now, and positing some mental excercises that show it in contrast. I think it makes its point reasonably, and that point is to help explain what is happening.