Principles for Redistricting

In my recent post on the importance of redistricting, I think I fudged an important distinction. By focusing on the difficulty of representing sprawling, politically drawn districts that take in just part of a community and then add far-flung counties, I implied that I think such values as "compactness" of the district, or following the boundaries of existing jurisdictions or communities, should be the highest priority in redistricting reform.

Several readers asked for more information about the options for redistricting reform, so I'll take up that invitation and mention some of the basic principles or tradeoffs in redistricting reform, especially as they apply to reform proposals under consideration in California, Ohio, and Florida.

All the reform proposals seek to make redistricting nonpartisan, usually by establishing an independent commission, such as in Iowa. The nonpartisan commission is an alternative to partisan districting, which is the situation in both Ohio and Florida, where Republicans controlled both houses of the legislature and the governorship during the last redistricting cycle, as well as California, where Democrats dominated, and the better-known cases of Texas and Georgia where Republicans took full control after the normal cycle and reopened the process to push through a new map. Purely partisan redistricting is less common than bi-partisan redistricting, which is normally the process when one party does not control the whole legislature and statehouse. Bipartisan redistricting, such as in New York, tends to lead to a mutual incumbent-protection scheme, whereas partisan redistricting can sometimes make a state seem more competitive, at least for a time.

The nonpartisan commissions proposed in California, Ohio and Florida must be given some guidance. In the case of California's current proposal, priority would be given to the rule that districts be compact and contiguous, and follow existing jurisdictional boundares. Competitiveness would be a secondary criterion. Florida may have a ballot initiative that creates a nonpartisan commission, and a separate initiative to require the districts thus drawn to be compact and contiguous.

Ohio reformers are taking a different approach. There, the priority is on competition, and the system will be a brilliant experiment: Anyone at all can design and submit a map. Each proposal will then be given a point score through a computerized system, with the highest priority placed on competitiveness, though points can also be awarded for compactness or subtracted for lack thereof. The commission can choose from among the three highest-scoring plans. (The addition of the commission is apparently a compromise from an older proposal that would have had a computer do the whole thing without human discretion.)

At a time when congressional reelection rates have returned to the 98% level and only a handful of races are competitive, when the same is true in most state legislatures, it would be a pity to go through the exercise of redistricting reform and not see a significant increase in competitiveness. That's entirely apart from party preference; it's simply the fact that much of what's wrong with politics -- from personal corruption to policies that ignore public preferences -- thrive in an environment of lack of competition. Putting the priority on compactness of the district, or following existing lines, might make some districts more competitive, but in many other cases might make them much less competitive.

Further, I used two examples of split-up communities in my previous post: Austin, Texas and Jacksonville, Florida. Both of those are mid-size cities that would fit almost completely in a single congressional district, and should. But they are rare cases. Any rural district is likely to sprawl across a wide area that has no natural sense of community, just as any district within a city like New York is likely to take in a collection of neighborhoods that may have little in common or little sense of community. That's particularly true if all districts are required to be the same size.

That would tend to argue for a higher priority on competitiveness as a goal. But simply constructing districts that have a relatively comparable number of Democrats and Republicans doesn't necessarily improve the quality of representation. If the Ohio reform passes, probably many districts will be constructed in Ohio that link heavily Democratic urban areas with distant Republican suburbs. Will the winner be the person who can most faithfully represent the broad majority of voters in that district? Maybe. Or will the outcome will hinge on mobilization, on whether the Democratic candidate is more successful at boosting turnout in the urban areas, or the Republican more successful at mobilizing white suburbans: a replay of the Ohio presidential campaign strategy of 2004? If so, assuming the winner achieves a reasonable majority, he or she is not likely to feel any more obligation to represent the views of, or win over, the 47% minority than Tom DeLay feels to a duty to represent the smaller number of people in his district who would never vote for him. And after that first victory, the other advantages of incumbency kick in, principally the ability to raise enough money to scare away any challenger, so the district may not be very competive for long.

There are two kinds of "competitiveness." One is a kind of "natural" competitiveness that one finds when people are genuinely up for grabs as congressional voters. Usually this kind of competitiveness is found at moments of transition. When I first started following congressional races closely, in 1988 or so, I remember that a good number of the competitve seats were in the Carolinas. But these were districts that had been voting for Reagan, and yet still sent Democrats to Washington. They were becoming Republican districts, and their apparent "competitiveness" reflected a deadlock between the incumbency advantages and moderation of their Democratic rep, and the ideological claims of the Republican challenger. It sometimes took a few tries before the Republican broke through, and the Republican takeover in 1994 was really the sound of the last dominoes falling. Today and in the years ahead, the competitive districts will be those historically Republican Northeast and Midwest districts that have been voting for Democrats for President. For now, Republicans like Chris Shays, Nancy Johnson and Mark Kirk hold on, either they will step aside or they will eventually lose, and at that point, the districts will cease to seem "competitive" but will become the Democratic districts they are meant to be, and the "moderate Republican" will become extinct.

It seems to me that some of the "competitiveness" that these reform efforts are trying to restore is trying to artificially recreate circumstances that existed only at a time when party and ideology were not fully aligned. In the 1950s, James McGregor Burns wrote of four parties -- each party having a presidential and congressional wing, with the Democrats' presidential wing liberal and its congressional wing more conservative; the Republicans more or less the opposite. That four-way dynamic, while it had many ugly outcomes such as the denial for decades of civil rights, nonetheless created some possibilities for movement and uncertainty that are less likely in an era when there are only two parties.

In the current climate, competitiveness in many districts might be found in primaries, and I don't really object to that. In some cases, first-past-the-post primaries may lead to polarization, as more extreme candidates have an advantage, but I doubt that's always the case. The bigger problem with relying on primaries for competition is that it is very rare and difficult to wage a primary challenge against an incumbent.

All these tradeoffs could be avoided by adopting the most compelling of the reforms in the proportional representation/instant-runoff voting universe: Multi-member districts. (See www.fairvote.org for background on PR, IRV, choice voting, etc.) A single district of, say, 2 million residents might include all of Houston, with three representatives elected at large, through a system of choice or ranked voting. There would likely be competition for all three seats, perhaps two Democrats and one Republican elected, and almost everyone in the city would be in some sense spoken for.

Unfortunately, these reforms, at least at the congressional level, are not yet within the realm of the possible. I'm not one of those who believes that anything short of eliminating winner-take-all elections is basically worthless. The redistricting reforms in California, Ohio and Florida, if they pass, will be great experiments and could bring real improvement in the quality of elections and representation.

Comments

[Electoral Math]

Aren't first-past-the-post primaries, aka "blanket" primaries, unconstitutional? Or are you talking about something else?

I'm not a big IRV fan ... it appeared to have really wacky effects on local elections in San Francisco. Plus in the end it encourages fringe candidates, discourages coalition building, and doesn't eliminate tactical voting.

Non-partisan redistricting with a bias towards competitiveness + CFR seems to get you most of the way.

[C.J.Colucci]

I think the question of "naturalness," for lack of a better word, gets short shrift in discussions of compactness and competitiveness. I live in a racially-mixed but plurality white middle-class neighborhood in the Bronx. Nothing but the most outlandish gerrymander could make it other than solidly Democratic. But the State Senate Republicans got p.o.'d at my then-Senator, whose district then included along with mine affluent Riverdale and a large chunk of Upper West Manhattan (both the stereotypical Upper West Side and Hispanic Washington Heights). Knowing that a Democrat would win in the district and any surrounding district, the Republicans gerrymandered him by cutting off different white enclaves and giving him a more Hispanic district in hopes that some Hispanic Democrat would beat him in a primary. (It didn't work.)
I was quite happy with the Senator I had. His concerns and interests very much coincided with those in my neighborhood, and we were part of his "natural" constituency. As part of the redistricting plan, however, my neighborhood became a tiny appendage to a largely poor, Hispanic district represented by a Senator whose concerns, understandably and rightly, were those of his main constituents, who have little in common with us. He spends next to no time cultivating us, and I don't criticize him for that. He is, and ought to be, predominantly an advocate for the far more needy constiuents who are his natural base. But the fact remains that his concerns are not ours and we have no political leverage over him.
Can computers and what-not factor this sort of thing in?

[Ellen1910]

An aesthetic argument might tempt voters.

Anyone know of a site out there dealing with topological issues of design of voting districts?

[Marvyt]

For true reform we need to eliminate winner-take-all races. Multi-member districts would go a long way toward giving everyone a feeling of being represented. Even if a district is 65-35, the 35percent minority deserves to be represented. I would even be in favor of proportional statewide voting. The independent redistricting committees are better than what we have now, but incumbents will still overly dominate the races.

[Mark]

As long as massive political reform is on the table, is there any good reason that every state has a Senate and House? Both are based on population, but Senate districts are bigger. What's the point? What's the problem with unicameral legislatures?

[yoyo]

Nebraska has a unicarmaral legislature.

But we have to have to bodies, since american law must follow English law, and England has to legislatures. so we do too.

[Tim O'Keefe]

Re: unicameral state leges.

What I like is the idea of having a house and senate, leaving the house as is--fairly small single-member districts--whereas the senate would be like many European parliaments. Have party slates, and then voters state-wide would vote for a party, and the number of seats for each party would be proportional to the % of votes received. That way you'd get both representation of regional interests and a voice for third-party type ideas (e.g., Greens, libertarians) that are currently excluded.

[Joe]

The independent redistricting committees are better than what we have now, but incumbents will still overly dominate the races.

Marvyt: I think that you focus too much on incumbents winning as a problem. As long as the official is responsive to the district he/she represents and CAN be held politically accountable, incumbency is actually a good thing. The last thing we need is to have a different set of 435 Reps every 2 years. Incumbents have the experience, expertise, and understanding of government and politics necessary to craft effective legislation. If incumbency advantages were taken away, lobbyists would be the only power in Washington. Independent redistricting is a powerful force to prevent safe districts (I live in Gingrich's old district, and even with an open seat race in 2004, no Dem ran for Congress, that is ridiculous) while still maintaining an effective Congress

[murky]

What is the point of county-size districts in an age of hour-and-a-half commutes by car, cross-country career moves,and instantaneous electronic communication with the other side of the globe? Thanks to zoning and real-estate prices, we could use geography to make political representation more class-based, and I suppose Marx if not other political scientists would say this would make politics more natural. But economic districting doesn't seem to be the rule. It looks like we game the system so as to foster a competition between ideas whose appeal doesn't have any obvious relation a person's wealth. Is that natural or good? I don't know, but it reflects some theoretical assumption I doubt I've ever had explained to me.

[murky]

I'm worried my point about gaming the system is opaque: I'm thinking in a Darwinian way about what kind of political ideologies or party platforms would evolve in two kinds of environments. In an environment of class-based districts, platforms that appeal to class would have a selective advantage and so could be expected to evolve. In an environment where districts are roughly even class mixtures, there's little advantage to tailoring platforms to class, and so we can expect political ideas will compete for votes by appealing to something other than class. I suppose race and religion are sufficiently common denominators to win some districts, but to win nationally I imagine you need to tap something deeper--human nature or "national cultural values." That seems a little crazy to me--like voting on a religion, and it seems to encourage the emergence of political platforms that are more like religions, having a wholly abstract appeal. "Special interests" has become a bad word, but what is geographic political representation but a harnessing or tapping of special interest? The problem is that the interests become less special to the extent populations are dynamic and districts are badly drawn.