Understanding the value of bloc voting, using the Congressional Progressive Caucus as an example:

Daniel Stock writes:

I’m a public policy PhD student, interested in economic policy and a bit of political science.

I recently saw that the Congressional Progressive Caucus instituted bloc voting rules a few months ago: if at least two thirds of them agree on a bill or amendment, then all CPC members are bound to vote for it.

This got me thinking: how well would this bloc work, given the observed voting records / preferences of the members? There must be a sweet spot: it is irrelevant if (i) none of the members ever agree, or (ii) the members always agree. It might even be possible to calculate which member is expected to benefit the most (AOC?), in terms of winning the most extra votes to her side (and being forced to flip on the fewest votes).

I googled “optimal voting bloc” and your 2003 paper came up. Do you know if this is still an active / promising area of study, especially on the empirics side? It seems like most papers are about detecting blocs, rather than evaluating them.

This is a great research topic! I haven’t thought about coalitions in awhile, but my guess is that this is a good area for further study. One thing I hadn’t considered at all in my earlier work was anything like this two-thirds rule. There seems to be room for both theoretical and empirical work here.

One lesson from my 2003 paper is that you can’t think about the effects of coalitions in isolation. The left-wing congressmembers have formed this caucus, but there are also conservative and moderate caucuses. I don’t know if these other groups have this sort of formal bloc-voting rule, but they must do some coordination.

Anyway, if any of you have suggestions on where to go with this research project, just let us know in the comments.

12 thoughts on “Understanding the value of bloc voting, using the Congressional Progressive Caucus as an example:

  1. So generally in voting systems, coalitions and bloc-voting formal coordination tend to have an advantage in steering vote outcomes.

    What’s the deep PoliSci insight here ?

    (seems a routine and unremarkable observation)

    • Orest:

      It depends a lot on the details, is the answer! Read the above-linked article, also there’s more stuff here. Just for example, a quick question would be how well a 2/3 rule will perform, compared to a 50%+1 rule or a 75% rule. Another question is how large should your coalition be, another question is who to include in the coalition and who benefits from being in it.

      Lots of things are unremarkable but can still benefit from careful study. For example, it’s unremarkable that if you raise the price of a product, you’ll (typically) sell less of it; it also can be useful to estimate elasticities, quantifying this effect and comparing it across items or over time. It’s unremarkable that if you add salt to a saltless dish, most people will enjoy it more, but if you add too much salt, people won’t like it. It can also be worth studying how much salt is optimal, in various senses.

  2. There’s an interesting aspect of duration to this. I assume the CPC rule applies to just a single vote or perhaps a small set of votes linked to a single item. Following the vote all members are free to revert to their previous positions, espousing and organizing for them.

    The opposite is the Leninist principle of democratic centralism, which is of unlimited duration. Once the party (or its leadership) has decided on a particular policy choice, everyone must fall into line. Expressing dissent or forming a dissident caucus is forbidden indefinitely. This is not just dustbin-of-history stuff, since some organizations today continue to operate along these lines — certain unions, for instance.

    The principle of time limitation is a very important one for democracy. I can imagine temporary unanimity rules that extend beyond a single vote. A union, for instance, might institute such a rule for the duration of a job action, or a political organization might do something like this for the duration of an election campaign. In any case, I think it’s extremely important to commit at the outset to a fixed duration after which dissidents will be free to organize to alter policy.

    Empirically, it would be interesting to know what sorts of organizations have adopted unanimity rules, how they’re conditioned, and how well they’ve worked.

    I realize my comment goes beyond the legislative voting context of the OP, including it as a special case.

    • Ironically, the Democratic party seems to have less capacity for central decision making, even if it is a “Leninist principle”–I’ll have to defer to others’ knowledge re: unions. Conservatives, however, have had remarkable success of late forging a unified platform, persuading within-party opponents to change sides (see Sen. Lindsey Graham) or primarying them out of power (see most anti-Trump Republicans). They’ve even managed to flip the platform 180 degrees virtually on a dime–with only slight, and temporary, pushback–re: immigration and free trade. I attribute the GOP’s centralization to their having essentially outsourced the party apparatus to conservative media like Fox News. MSNBC may be liberal, but they seem to have only limited ability to promote–much less define or enforce–a particular set of ideas as “the” liberal agenda.

    • +1 It would also be fascinating to learn how long a coalition using a two-thirds rule can last and under what conditions. If a subcoalition forms that is always having to vote for the proposals of the other two-thirds, that will eventually not be attractive. However, knowing that their coalitions is getting something “closer” to what they want, they may stay put. In a way our whole political landscape is dominated by something like a two-third rule where we agree to go along with the Democratic or Republican agenda even though large fractions of each side are unsatisfied with their own side. There should be spliter groups all the time that the other side can pick off, but it actually seems like the coalitions are remarkably stable. It would be fascinating to know why both mathmatically and empirically.

      • I’d argue it’s more like a 1/3 rule. Vast majorities of the population–liberal, independent, and conservative–support gun control, raising taxes on the wealthy, ending the war on drugs, scaling back military spending, reducing the debt, cheap healthcare, public safety nets of various kinds, and (until recently) greater access to voting. I would hypothesize the reason is that you only need 50.1% of those who actually vote to an election, and you generally only need a plurality to win the primary that gets you into the election. But yes, some empirical evidence one way or the other would be good. :)

  3. I know very little about the science of coalitions. However, I’ve thought a lot about them lately, because coalition building and certain perverse incentives arising from them are, I believe, what got US politics to the scary place we’re in right now. Prior to the 90’s, the GOP, as a pretty homogeneous party in a pretty homogeneous country, was able to use a couple of core sets of conservative positions (military-economic-religious) to forge a large and energized coalition. As society has become more diverse, Republicans have only doubled down on homogeneity, which has shrunk their potential base. Consequently, they’ve had to offset falling party size with rising enthusiasm and turnout, which they have achieved largely by taking increasingly divisive stances on issues. These stances have not arisen organically–they’ve been market-tested by rightwing media as the most provocative, compelling issues for their audience. This is how the party went from pro-Romneycare to anti-Obamacare, pro-immigration to anti-immigration, pro-free trade to anti-free trade, and most recently pro-corporation to anti-corporation.

    We now have a situation, at least on the political right, that reverses the assumptions of Andrew’s 2003 paper. Andrew assumed that coalitions are built primarily to facilitate mutual support for sets of issues. Now, however, conservative coalitions often are not formed in order to gain support for political positions or political actions. Political positions and actions are, rather, driven by the perceived need to hold together a shrinking coalition. The driving force behind coalition building, then, is not the desire to enact policies favored by coalition members. It is the desire to maintain control of political institutions and to “beat the other team” at the ballot box. The cart is pulling the horse.

    • Michael:

      I don’t necessarily agree with all of your premises, but I will say this: A very good trial lawyer told me that good plaintiff’s lawyers don’t try to get juries to sympathize with their clients. They try to get jurors angry at defendants because anger is a better motivator than sympathy. I say this because the Republican consulting class definitely subscribes to this view. They have believed for a long time that turn out of their base was more important to win elections than finding new votes. And, anger and fear are the best motivators. However, what gets people to the voting booth may be different than why they are part of the Republican coalition. Republicans in Congress still passed regular Republican legislation, big tax cuts, trying to reduce social spending, big defense spending. It was McConnel’s policies that got passed. There was no big infastructure bill under Trump. No real change in free trade policies. So maybe we are mistaking what motivates the base to give money or vote with the policies that the Republican coalition stands for, which I don’t think has changed that much.

      • I think you were right before 2010. Republican leaders were saying one thing and doing another. But then Cruise, Paul, and a boatload of far-right Reps were swept in, all of whom seemed to believe what they were telling voters and acted accordingly. They were a huge thorn in the sides of both Paul Ryan and McConnell–their rhetoric was party-approved, but their actions, or inactions, had practical consequences. Plus, McConnell’s a terrible example of a pre-90’s Republican: his all-out war on Obama, from calling even previously-GOP-supported positions un-American to tolerating open racism to sidelining Garland, were qualitatively different from an earlier era. Even Boehner didn’t go that far. To say that the GOP under Trump hasn’t substantively departed from normal conservatism is to accept their redefinition of normality. It’s a biased metric.

        I think we also disagree about what constitutes real change. Tariffs on China? Blocking everyone from migrant farm workers to highly-educated tech workers? Firing scientists and watchdogs? Now, perhaps you’re focusing on legislative actions only (although I don’t see why–a law is a law, whether by legislative action or consent). But congress does more than pass laws. Yes, all the laws passed were consistent with McConnell’s priorities, but very few laws were passed at all, so that’s a red herring. In the meantime, we saw substantive erosions in congressional oversight. Republican chairmen of congressional committees publicly released documents that the CIA insisted would compromise sources. Members of the FBI were subpoenaed and accused of conspiracy theories, tarnishing and ultimately weakening federal law enforcement. Presidential actions that a GOP congress would never have tolerated from GW Bush–and certainly didn’t tolerate from Nixon–were forgiven, excused, or encouraged from Trump. And don’t forget laws that weren’t passed but might’ve been with GOP centrists. The last time the Voting Rights Act came up, it passed near-unanimously across both parties. Will it next time?

        Like they say, the greatest trick the devil ever pulled was convincing moderates he was just a fiscal conservative.

Leave a Reply to dl Cancel reply

Your email address will not be published. Required fields are marked *