California’s Redistricting Experiment

This is a guest post from previous guest contributor Eric McGhee.


The latest legal decision is in, and things are looking even better for California’s nascent Citizens Redistricting Commission (CRC).  Last Friday, the California Supreme Court ruled strongly in favor of the CRC, declaring unanimously that the commission’s state Senate map should be used this year even if a pending referendum to overturn the map qualifies for the ballot.  The decision, Vandermost v. Bowen, can be found here .  This follows on a decision last fall, where the Court unanimously (and without any written opinion) dismissed a more substantive legal case against the state Senate and congressional maps.

The outcome was consistent with an evaluation of the CRC plans that Vlad Kogan at UCSD and I conducted for the California Journal of Politics and Policy (in fact, the Court cited the paper in its decision).  Not only are the new maps an improvement over the old ones on a variety of criteria the CRC was supposed to consider, they are also a little more competitive and show no signs of clear partisan favoritism.  That doesn’t mean that one party won’t benefit—Democrats can generally expect to do a little better—but these gains have as much to do with particular outcomes under the old maps as anything with the new ones.

My sense is that most political scientists and other savvy political observers would not have expected the commission to be such a success.  There were plenty of good reasons to be skeptical:

  1. It was composed of ordinary citizens with no formal connections to the Sacramento political establishment and little experience with redistricting;

  2. It had 5 Democrats, 5 Republicans, and 4 independents, and a majority of each group had to approve the plan;

  3. The criteria the commission was required to consider (compactness, minority representation, etc.) often contradicted each other (though the commission improved their own draft maps on virtually all these dimensions);

  4. The maps benefit the Democrats, yet a Supreme Court composed almost entirely of Republican appointees (6 out of 7) has repeatedly and unanimously ruled in the commission’s favor.

It makes me wonder whether the commission model is actually the right one for the post-Baker v. Carr era.  By requiring a new redistricting every 10 years, that US Supreme Court decision forced a contentious political fight that many states used to studiously avoid.  The result has been endless litigation, with the courts trying and often failing to stay out of the fray.

A commission like the one in California both avoids conflicts of interest and offers maximum transparency.   Indeed, the CRC held dozens of public hearings, all those hearings were filmed on line, all the data from the draft and final maps were publicly available, and the commission released a final report (mandated by law) detailing how its maps met a variety of legal requirements.  In essence, the commission drew the lines the way a Court would draw them—with documentation about why the lines were drawn and which legal requirements they satisfy—and while the commissioners might have had strong political points of view they did not benefit personally from the lines.  Considered this way, the commission becomes a quasi-judicial body, and its products correspondingly more valid in the eyes of a Court.  That doesn’t mean that a commission’s maps are necessarily “better” (can someone offer a definition of “better” that everyone would agree on?) except in the sense that they are a more appropriate fit for what has become a quasi-judicial process.

I’d be interested to hear reactions to this idea, especially from those who support the legislature-drawn model.  Why not a commission?  If a commission is such a bad idea, why has California’s experience been a good one (so far)?  Is it just dumb luck?  Or is there something more meaningful going on here?

8 Responses to California’s Redistricting Experiment

  1. Andrew Gelman January 31, 2012 at 8:14 am #


    t is my understanding that they have similar commissions in England and, although they’re not perfect, they work just fine. I’ve long thought such commissions would be better than the current system in U.S. states. I didn’t know anyone was arguing against them (as would seem to be implied by your question, “Why not a commission? If a commission is such a bad idea…”). Rather, it seems natural that, given the legislature is typically in charge of districting, that they wouldn’t want to give up that power.

    • Eric McGhee January 31, 2012 at 2:14 pm #

      Andrew & Quinn:

      Yes, it’s my understanding that most single-member district countries use commissions to redraw their lines. I don’t know enough about them to offer any insights, but it does turn the question around: rather than wonder why some states use commissions, we should wonder why most do not. There may be a perfectly good reason, but it would seem a question in need of an answer. Do you know of any comparative research on this subject?


  2. Quinn Albaugh January 31, 2012 at 8:45 am #

    Nonpartisan electoral commissions have existed since 1955 in Canada, starting in Manitoba for provincial elections, then moving to other provinces and the federal level. Canada developed the idea based on an earlier model from Australia. Often the Canadian commissions are smaller than 14 and full of experts. The criteria generally considered important in Canada and in the provinces are also at times contradictory. Despite that, the system has worked fairly well since then–certainly it’s not at the same level of litigation and post-commission disputes. Now, changing the number of seats in the House of Commons, that’s where you can find a dispute.

    Further reading:
    Courtney, John C. 2001. Commissioned Ridings: Designing Canada’s Electoral Districts. Montreal-Kingston: McGill-Queen’s University Press.

  3. Ben Bishin January 31, 2012 at 10:32 am #

    Nice piece. I am generally skeptical of such “post-partisan” style solutions, although in this case, the previous gerrymander was such an extreme one favoring incumbents that I thought it would be hard for an independent body not to do better. Simply put, they have reduced the incumbency gerrymander to better reflect the distribution of voters’ (partisan) preferences. And to their credit, the GOP appointed Court has not stepped in to stop them (another source of my concern). But I also think there are some interesting unanswered questions here. Do these commissions work as well in states that employ partisan rather than incumbency gerrymanders? And, what are the implications for minority representation which seem to slightly reduce the number of seats the Democrats are likely to win (at least in the Assembly). I wonder if future courts are likely to view commissions as being more race blind than the legislatures were? Ive also long wondered how such processes should account for unaffiliated voters who in California make up around 25% of the population (last I checked). If you take the view that these are closet partisans, maybe its not such a problem, but if you view them as people who have rejected the parties, or even as weak identifiers, they are left with little influence–perhaps no change to their previous situation! Finally, let me point out that Florida may provide an interesting comparison case in which voters passed rules guiding redistricting by the legislature (but not an independent commission). The results aren’t yet known there, but all of those plans seem to be moving in a fairer direction. However, while Dems look set to pick up at least 3 seats, despite roughly equal party registration, the GOP will hold on to a large majority there (I’m guessing something like 17-10; the previous one was 19-6).

    • Eric McGhee January 31, 2012 at 2:25 pm #


      Yes, the answer may just be that California’s old districts pissed off so many people that anything would have been better. The Florida solution is interesting, but it seems to be missing the independence that I assumed was a key part of the CRC’s legal success. In fact, it seems like there are two important dimensions of difference among these redistricting reforms (including the dozen or so states that use some sort of commission): the independence of the commission members from the political establishment (no mean feat, and rarely achieved), and the specificity of the legal constraints on the kind of maps the commission can draw. The CRC was at one end of both dimensions. FL, on the other hand, might be high on specificity, but it’s clearly low on independence.

      Someone ought to write a paper about this (including all the other questions you raise), if they haven’t already.


  4. Anonymous Coward January 31, 2012 at 12:16 pm #

    I generally favor legislature-drawn maps, or at least vaguely oppose commissions and the like, for a variety of reasons.

    First and foremost, most of the criteria they’re called on to use are pointless to me, because I try to keep my eyes on the endgame of who gets elected and what policies are enacted. Whether a district is a perfectly smooth polygon or has squiggly borders doesn’t impact that. All I care about is whether the people who are elected as a result generate good policies (ie, ones that I like). Irrespective of the aesthetic qualities of the districts, California’s commission won’t be a success until the California Legislature generates detectably better outcomes than it had before. If, in the end, it doesn’t create a legislature that generates detectably better outcomes, why should I say it was a good thing? I suppose the argument is that a more transparent or seems-fair-to-voters system will help create legitimacy or support for the legislatures. But Hibbing and Theiss-Morse have more or less convinced me that voters will never like or support any legislature that appears in the news, so that just seems like tilting at windmills to me.

    Second, if you care about things other than policy outputs, competitiveness is a terrible goal to pursue (as opposed to just irrelevant like compactness). It guarantees that districts will actually maximize the number of electorally losing voters who quite rightly feel unrepresented in the legislature. Secondarily, the process of trying to draw competitive districts while leaving some inevitably uncompetitive is hard to distinguish from a hard partisan gerrymander. You want a map that generates many more or less competitive districts in terms of party registration? Look at the New York Senate’s 2000 map.

    Third, any set of districts is going to encourage the election of some people and discourage the election of others, and there is no possible set of districts that is fair, neutral, or unbiased. Further, it is difficult to imagine that anyone working within the political sphere does not have political preferences, and if they don’t there’s probably something wrong with them. From my perspective, you would have to be daft to think that people drawing the lines aren’t trying to achieve their own policy or electoral ends — for example, the boundary commission in the UK seems to like favor Tories in England, to the extent that it would (almost certainly) be possible to draw a more pro-Labor map there but they choose not to. So I can’t see why getting a map that reflects a commission’s preferences or some set of civil servants’ preferences is really any better than one that reflects legislators’ preferences.

    • Redistricter January 31, 2012 at 1:07 pm #

      I work a salaried position at a certain political party’s headquarters that exists solely to give my side a leg up in the redistricting process, and even I think commissions are a good idea. It is my sincere hope that my job does not exist in 2021.

      Nothing about a legislative redistricting process ensures better policy outcomes. The two goals of any maps drawn for a partisan redistricting process are (1) incumbent protection and (2) partisan advantage. I hesitate to criticize any particular plans from this round for fear of exposing my partisan employer of choice, but the Democrats’ maps for the Illinois Congressional redistricting cynically sacrificed minority voting power for incumbency protection in the Chicago area. This is one reason why compactness matters: A district shaped like a spider having a seizure is probably drawn that way to encompass an incumbent’s house. Before you cut a map, you devise a layer that is just the lat/long of incumbents’ homes, cut around that, then remove the layer and ship it off to your state affiliate. Republicans’ Kentucky state senate proposal had some particularly pernicious pro-incumbency effects: They renumbered all of the districts the week before the filing deadline, so most Democratic incumbents woke up one day to find out they’d filed to run in districts they don’t live in; and skirted the very border of the state Constitution just to dig a home for a Republican incumbent in Warren County. Each of these actions were perfectly legal, but befuddling efforts for anyone to run for office who wasn’t in the room when the map was drawn would never occur to anyone engaging in a non-partisan process whose primary concerns were population changes and compactness.

      Immunizing political incumenbents from challenges does not better public policy make. Entrenching not just parties but specific incumbents empowers the interests that funded their campaigns in the first place. Two 75/25 districts disenfranchise just as many people as two 50/50 districts, but in a fairer district there is a higher probability the more motivated, energized side will win. A 75/25 incumbent can jettison 75% of voters and still tie up an election, while a 50/50 incumbent must respond to at least 50% of the electorate as well as, probably, swing over a few fence-sitters. Fair districts actually minimize the number of disenfranchised voters, or at least maximizes the number of voters an incumbent has to be responsive to.

    • Eric McGhee January 31, 2012 at 2:46 pm #

      My thoughts were less about the CRC’s product, and more about the process it followed. Does that process help inoculate the product against legal challenges? If the courts are going to be involved anyway, maybe it’s best to design a quasi-judicial process expressly for drawing the lines, rather than leave it to the discretion of judges who were not chosen specifically for that purpose and clearly feel uncomfortable in that role.

      Maybe the legal challenges are not worth worrying about, and the real issue is the CRC’s product. But if the growing role of the courts is a cause for concern (as I feel it is), then it seems like California’s commission may be a good bureaucratic solution.