Revisiting an Idea On Redistricting Reform

[Reposted; hopefully this time the tables are legible.]

Long-time readers of this blog are aware that I periodically enjoy pondering how the electoral system in the U.S. could be reformed to address certain shortcomings in the current arrangement.

Specifically, I have two pet peeves with how our elections operate:

• I dislike gerrymandering — the playing of political games in drawing representative districts so as to increase the odds of a particular party having a stable number of seats in a given state; and
• I dislike the idea of “winner take all” elections, which tends to polarize politics, enshrines the power of the two dominant political parties in the country, and seemingly stifles alternative ideas

I’m not alone in those thoughts, and various folks have proposed a wide array of ideas on how to address those shortcomings in our current political system, including suggesting algorithmic ways to draw legislative boundaries, as well as offering up alternative methods of vote-casting or vote-tallying to shake the tree of power a little bit.

However, I haven’t heard too many folks advocating my preferred idea…so I’ll share here.

For discussion’s sake, let’s assume that I’m talking about reforming how members of the House of Representatives are elected.   The logic could be similar for electing representatives to state or municipal governing bodies, or even to allocating votes in the electoral college.   With that aside, my idea is:

1. I’d like to see regulation/legislation that requires representative districts to be drawn such that the length of the boundaries of districts be minimized.   Doing so would inherently create “compact” or “square” districts that should tend to keep communities together, rather than these odd, long-shaped districts drawn to favor one political party or another.
    
2. I’d like to see an end, or at least a relaxing of the idea that each legislative district can be represented by one-and-only-one candidate, determined by a winner-take-all vote.  In the case of the House of Representatives, rather than electing 435 representatives from 435 districts, why not draw the 435 representatives from, say, 233 districts (minimum one district per state, but half-as-many “excess” districts, with the “excess” districts permitted to cross state lines)? 
    
3. Shifting to having potentially multiple representatives from a district doesn’t alleviate the polarization and the enshrinement of duopoly unless you limit individuals to voting for only one individual, and unless you permit candidates to run in multiple districts (presumably with a relaxation of any residency requirements).
    
   In such an environment, the highest vote-getter in each district would be declared elected.  Then, from the list of not-yet-elected candidates, take the top vote-getters to fill the remaining seats.   If a mid-term vacancy occurs, take the next willing candidate from the list (avoiding the shenanigans of gubernatorial appointments or special elections).

This scheme might be easier to understand with an example.

Let’s assume that we live in a jurisdiction with a 10-seat legislature, 10,000 voters, two major parties (”D” and “R”) and two minor parties (”G” and “L”).  Today, that legislature would likely take its members from 10 districts, where election results could look something like this:

(In the candidate abbreviations, the first letter represents the party, while the second letter represents a particular candidate within the party.)

Candidates R-A, R-B, D-C, R-D, R-E, D-F, D-G, R-H, D-I, and D-J are elected, giving us a legislature made up of 5 R’s and 5 D’s.   In aggregate, the legislature would look sort-of-kind-of like the political leanings of the entire jurisdiction (if the L’s are assumed to be like R’s, and the G’s are assumed to be like D’s), but it does have the unfortunate side effect of many voters potentially being unserved by the representative from their district.   Minor parties are completely unrepresented in the legislature.

Under my scheme, we could see that same jurisdiction divided up into five districts.  D’s and R’s would still run one candidate per district, but the G’s and L’s would each pick one candidate to run across all five districts.   The results would then look something like this:

With the votes cast, we have ten representatives elected:

We have 4 D’s, 4 R’s, and both the G’s and the L’s were able to gather enough votes to elect a candidate each by running one candidate each across all five districts, rather than splitting their votes.  Thus the legislature is more representative of the political makeup of the jurisdiction than under a traditional one-representative-per-district scheme.

Also, you have more potentially satisfied voters, since more folks had “their” candidate elected, thus reducing the problem of citizens feeling disenfranchised because of the “don’t blame me; I didn’t vote for him” syndrome.

One criticism that you might have with this example is that fewer candidates ended up running for office, and almost everybody was elected, making this not much of a contest.  However, I think this is more of an artifact of presenting a simplified example than an honest flaw.

It’s true that the major parties would likely get into some heavy-duty statistical analysis to develop strategies of who should run in how many districts to optimize their results.  If the parties were able to maintain their duopoly of power, then this would make it difficult to have a legislature that was appropriately responsive to the people.

However, the ability for a particular candidate to gather votes from multiple districts opens the doors for third parties to play a role, as well as special-interest offshoots of the major parties.

For example, for some voters race or ethnicity matters.  It’s easy for me to imagine an “African-American Democrats” caucus or a “Latino Republican” faction running candidates across multiple districts, to obtain votes from interested voters.  This has the happy side-effect of potentially alleviating systemic racial discrimination, the combating of which has been the cause for some mighty funkily-shaped districts under the current system.

Perhaps a Conservative Christian group would splinter off from the GOP to offer up its own slate of candidates, as could a LBGT special interest group.  Even Cindy Sheehan could work to get herself on the ballot countrywide to get a seat in Congress, rather than opting to pick a fight with Nancy Pelosi in an ultra-liberal district.

Because of the potential for rather specific groups to get candidates elected, and because of the reduced influence of geography, I can see porkbarreling being slightly less attractive, as well as perhaps a bit of an erosion in the tendency of the legislative sessions degenerating into a sequence of party-line votes, regardless of the range of opinions held by the folks being governed.

And, because voters could conceivable elect officials from anywhere, you’d eliminate another gripe I have with the way Congress currently operates.  Today, if there’s a bill being considered in committee that I’d like to comment on, as an ordinary citizen, I have to hope that my local representative is on the committee.  If he/she isn’t, then I may be out of luck since the members of the committee have no real incentive to listen to me.   However, under this system, I should be able to attract the attention of at least a few committee members, since they could conceivably be interested in my support next election, regardless of where I live.

I realize that supporters of other reform ideas have their own thoughts on how their systems would break the stagnation of duopolistic power.  However those systems often require significant changes to how voters vote — e.g. ranking candidates by preference, or scoring candidates.  My system doesn’t require significant reeducation, and only minimal adjustments to the processing of votes.  Everyone still casts a vote for their preferred candidate, and those votes are still counted in the normal way.  The only change is in how the tallied votes are aggregated to identify who’s been elected.

I don’t hold my breath for a system like this, or any real reform, from ever taking place.  The folks in power have too much invested in the status quo to consider alternatives.  However, I hope you’ll forgive me if I happen to ponder from time to time how things might be made better.