Princeton Election Consortium

Innovations in democracy since 2004

Oct 21: Biden 357 EV (D+5.4% from toss-up), Senate 53 D, 47 R (D+4.8%), House control D+5.0%
Moneyball states: President NV AZ PA, Senate AK MT IA, Legislatures KS TX NC

Constitutional rot: like an eclipse, or a mistuned clockwork?

September 25th, 2020, 5:28pm by Sam Wang


In my second contribution second contribution to the Symposium over at Balkinization, I lay out a theory for reform. Any government is composed of pre-existing and artificial features, working together to make a complex system. Within that system are positive and negative feedbacks that can make the government more responsive or less responsive to its citizens, and entrench or destabilize various groups.

In a democracy, the trick is how to tune those feedbacks. In the face of the long-term demographic shifts we face, can we manage the coming decade so that our democracy comes out better? Here at the Electoral Innovation Lab we will be interested in understanding this question – and maybe help with devising the answers.

Tags: Politics · U.S. Institutions

14 Comments so far ↓

  • UserFriendly

    “(Another difference is that there have been innumerable total eclipses, while there’s only been one Gilded Age. Really, he’s extrapolating from only one cycle. Still, you gotta have hope!)”

    There have been more:
    https://slatestarcodex.com/2019/09/02/book-review-ages-of-discord/

  • Ebenezer Scrooge

    Engineered systems are usually nicely decomposable, and thus susceptible to analysis. Evolved systems–eh. A lot of ecological feedback loops are easy to model. But social systems–what did Damon Runyon say?

  • James McDonald

    One fix would be a simple law to increase the size of the House to the maximum number consistent with the constitutional restriction of at least 30,000 voters per district.

    Also add a compactness requirement that would automatically favor (after the requirement for equal population per district) any districting map for which the total district boundary lengths was minimized.

    The House would then have about 11,000 members, so the House rules would need to be revised to accommodate that, probably including a robust process for proxy voting, but also remote voting provisions.

    The EC advantage of small states nearly vanishes.
    Gerrymandering becomes impossible.
    The average voter would be on an almost first name basis with their representative, making the money advantage for advertising dramatically less significant.
    Geographically diffuse groups with common interests would have a chance of assembling voting coalitions in the House.
    Special elections would be vastly simpler.
    In the event of a catastrophic loss of government buildings and persons in DC, the House could reconstitute itself within hours.

    And it could be passed on day one of a Biden administration.

    • Sam Wang

      No, gerrymandering becomes easier than ever with small districts.

      The compactness idea does not help because density and partisanship vary to still allow packing/cracking.

      Try this by drawing some actual maps in Dave’s Redistricting App or some other program.

    • Sam Wang

      The best example of smaller-population districts allowing gerrymanders is state legislatures. Wisconsin and Michigan have extreme partisan gerrymanders in their lower chambers despite having 99 and 110 districts, respectively.

      https://www.michigan.gov/documents/cgi/house10statewide_371473_7.pdf

      In any event, shape is not the only criterion for defining a good district. See the redistricting criteria in California, Michigan, Colorado, and Ohio reform legislation. And in any case, humans must be involved – no algorithm to date captures the real-world tradeoffs necessary to create a fair map.

      For many, many materials, see gerrymander.princeton.edu. For precinct-level voting data, see openprecincts.org. For free redistricting software see davesredistricting.org.

    • Jarrett Grindlay

      I used to be a big fan of increasing the number of districts, but if we are going to bust out our magic want, what about just increasing the number of representatives so that each district had 3-5 representatives?
      This helps combat the issue of data analytics fine-tuning micro-districts to still pack/crack, while also doing a really radical thing: giving almost every voter a representative that represents their views. If a district has 3 or 5 representatives along with a Ranked Choice voting system, it would require candidates to fight for the center. Right now something like 90% of house districts are functionally non-competitive – if you win the primary, you win the general almost by default. But under this scheme, maximizing seats can only be achieved by maximizing your broad appeal. Also, I just really like that even in most safe districts there’s a good 30-40% of people who vote for the other party and get nothing. this way they would likely at least get at least 1/3 of the representatives.

    • Sam Wang

      There’s a nonlinearity problem. This came up during the 2016 GOP Presidential nomination process, which used a rule kind of like what you are suggesting (but without ranked-choice voting). For example, what do you do with a 55-45 vote in a 3-representative district? Presumably the 55% get two representatives, i.e. 67% of them. Unless the system is crafted really carefully, it can backfire.

      Historically, multi-member districts have been used to hold down black people. For example, some southern states would have an at-large delegation elected by winner-take-all.

      Getting back to your suggestion, I think the key part might not be the multimember aspect, but the voting rule. Voting rules still have to be crafted to fit a local political landscape. That is a whole dissertation, literally.

  • Paul Marino

    Anything that promotes societal integration I.e., the breaking down of make-believe differences among humans will, produce a more peaceful and prosperous society. This has been the social evolution of our species. Democracy I.e., equal voice in how society functions is the inevitable outcome of this process. The problem in the US is that democracy is flawed, different people, depending on where they live happen to have a larger voice in how things are run than do others. Fix that problem and improve education and the nation will function better. Democracy does depend upon a well educated/informed public.

  • Bela Lubkin

    > Gerrymandering becomes impossible.

    I think the opposite. Detailed demographic and voting information is available, down to precinct level at least. To gerrymander at that level you simply have to create a lot of border districts between ‘us’ and ‘them’ in which the ‘us’ population is somewhat higher than ‘them’ (‘cracking’). Meanwhile, especially if you are the low-population-density conservative side, the interior of these areas you are edging around ‘pack’ the rest of ‘them’ into conveniently disposable high density packets.

    I would guess that 30K districts would give a Republican gerrymander a natural advantage of somewhat over 2:1 in most states (consistently win around 67-70% of seats). A Democratic gerrymander in most states would be handicapped and difficult to get more than 60% of seats. Both directions would be absolutely possible, if the legal and societal overseeing environments allowed them.

    There is probably actual research out there, better than my too-of-head spewing….

    • James McDonald

      First, the compactness requirement would trump any such attempts to move the boundaries, so that’s a moot point.

      So the only question is whether compactness itself would amount to a de-facto gerrymander. I’d need to see the numbers to be sure, but I suspect not. Most likely there would be a lot of 80%+ D districts and a lot of 80%+ R districts, and not much in between. 30K is a fairly small number — most districts would be just a few miles across, with a few extremely large rural districts making up the remainder.

      It’s perhaps worth noting that the compactness requirement could lead to some districts being entirely geographically within the inner perimeter of a “donut” district around them — wedge shaped districts (e.g. the ones used to carve up Austin, Texas) would have longer borders and would be excluded.

  • Bela Lubkin

    growf. ‘top-of-head’, you silly swipe kbd!

  • Bela Lubkin

    I should note that I’m talking about expected seat results starting with a population divided 50/50 between the two current parties. Obviously in a state with 70% one or the other, the range of results is going to be quite skewed in that direction.

    The idea that gerrymandering becomes impossible seems to me an encoding of the idea ‘gerrymandering at this level would be a tremendous amount of fiddly little work which wouldn’t happen’. This breaks with reality in two important ways:

    (1) it is amazing what sort of detailed fiddly work humans will get up to. Build a schooner out of toothpicks in a bottle? Reconstruct the Parthenon brick by brick in Minecraft? Hand-draw detailed maps of your entire home state?

    (2) this is exactly the sort of work computers are good at — and they do not get bored or frustrated.

    • James McDonald

      Again, that’s why I added a compactness requirement.

      No matter how much you fiddle with the maps, if I present another map with shorter total borders, my map wins, period. (Assuming all the maps first obey the equal population rule.)

      In most cases there will be a globally minimal drawing that would prevail over any contenders, or at most a handful of identical drawings that move one or two houses one way or the other. In theory anyone could find those, but in practice it’s plausible that someone would come up with a “near enough” solution that no one else can beat. And it is extremely unlikely that anyone who did find a more compact solution would materially affect anything.

  • ArcticStones

    One of the first key steps must be to replace the present “reality bubbles” with a shared set of facts. I believe the best way to do that is to revive the Fairness Doctrine. Introduced by the FCC in 1949, it was sadly killed off in 1987.

    Only when we have shared facts and shared news can we have the meaningful public discourse that is fundamental to a healthy representative democracy.

    The second thing is to rejuvenate what we might call The Social Contract.

    The 2016 Presidential Election has been described as a National Referendum on Decency – and decency lost. In this 2020 referendum, we have to make sure Decency wins by a landslide.

    Without core decency at the top, it is hard to imagine there can be accountability.

    I think these core ideas are far more important than structural changes.

Leave a Comment