What if the USA changed to proportional representation?

[I later followed this up with “Democrats, socialism, PR, and Bernie Sanders.”]

Among the small, but beneficial, ripples from the allegedly “inconclusive” result of the German election on September 18 has been some discussion of how multiparty democracy with proportional representation (as seen in Germany and most other democracies) compares to the strict two-party system (seen almost uniquely in the USA).

For example, Chris Lawrence suggests:

If the incentives for a two party system melted away, more likely than not our existing Republican and Democratic parties would melt away with them (or at least be transformed beyond recognition). And if you think our parties are bad now, wait until you see the parties led by Maxine Waters and Pat Robertson (or their acolytes) and comprised solely of their true believers.

I disagree that Dems and Reps would melt away or be transformed beyond recognition; more on that later. First, I want to stick to this “true believers” analogy, because Stephen Karlson also uses it:

Successful political parties in parliamentary republics are able to appeal to their true believers — who do not have to live in contiguous districts such as Berkeley, or Emporia — to obtain seats in proportion to the true believers’ share in the vote.

Now, let us assume that Stephen meant parliamentary systems with proportional representation (as implied by his own subject line) and not parliamentary systems more generally, given that some parliamentary systems (notably Britain and Canada, which are not “republics,” by the way) use the plurality (first-past-the-post) electoral system just as the USA does. Let us further focus only on the dynamics of legislative party positioning, and not government formation, since for the latter process the make-up of the legislature is almost irrelevant in a presidential form of government like the USA has.

Do parties in proportional representation (PR) systems appeal solely to their true believers? And would hypothetical coalitions between Democrats or Republicans and smaller parties render our two big parties more polarized than they are today (as implied at Betsy’s page)? No, and no. In fact, no one who has ever watched an election campaign in a European PR democracy or New Zealand since it adopted PR could possibly make such claims. Caricaturing the process in this way represents a fundamental—but, in America, widely held—misunderstanding of how multi-party democracy works.

In a nutshell, the point is that, with very few exceptions, parties in PR systems cannot afford to appeal solely to true believers if they seek any actual policy-making influence. Why? Because inter-election volatility (the movement of voters from one party to another) is much higher in multiparty PR systems. It is higher precisely because each voter has more choices—that is, more than one party that may be appealing to some aspect of his or her policy preferences. If parties are competing in an environment with this heightened level of competition, the ones that stick to their true believers quickly become rumps and find themselves marginalized.

In the quote above, Chris asks rhetorically, “if you think our parties are bad now…” Yes, I do think our parties are bad now, and it is largely because one of them is pulled too much one way by the likes of Maxine Waters and the other is pulled too much the other way by the likes of Pat Robertson (to use Chris’s examples). As Chris notes, our electoral system and two-party system generate broad parties that are internal coalitions of interests. And he is also right that all parties—even the small ones in some PR systems—contain internal electoral coalitions of interests. Get the Waters and Robertson acolytes out of the internal coalition of the Democratic and Republican parties and I, for one, will like both parties a whole lot better than I do now. And I suspect most voters would, too.

In broad “catch all” parties with loose internal organization like ours, political forces represented by the likes of Waters or Robertson are constantly digging in and attempting to keep their respective party from drifting too far to the center in the quest for the mythical “median voter.” A member of congress like Waters (or Frank or Conyers, etc.) can dig in her or his heels and, because our single-seat-district electoral system gives such members safe seats, these members never have to worry about losing influence by doing so. It is close to frictionless for them to stake out extremist positions in advance of bargaining over specific pieces of legislation. There are simply no electoral costs for them from doing so, and lots of potential benefits if they can keep the policy debate within their party skewed even a little bit in their direction, while at the same time appealing to their own relatively cohesive and ideologically extreme electoral districts. (The same analogy holds within either party, though we would need to use a congress member who was close to Robertson’s views, rather than Robertson himself, because he is not a member of congress, as is Waters.)

In a PR system, let us suppose Waters (and her allies, as well as her counterparts on the opposite side of the spectrum) split off and form their own parties. For the sake of argument, I am going to assume that the US has adopted MMP, like in Germany and New Zealand, though this thought experiment would not be radically different under most other forms of PR.

Now, in our hypothetical MMP system, Waters’s clout in congress (i.e. the share of seats her new party obtains) depends on her success at garnering votes from outside her own safe electoral district. She is now subject to competition with the Democratic party—which surely would survive, as would the Republican, albeit in smaller and more moderate form. The party led by Waters, Conyers, et al., is now tugged towards the center, because that is where it can gain the most in additional votes. That is where the inter-election volatility will take place, not at the fringes.

In conclusion, I do not deny that there are some parties and some party systems where PR and multipartism contribute to the kind of narrow ideological appeals that Chris and Stephen have in mind. For example, Israel, with its extremely low threshold and several tiny religiously oriented parties, or Italy from the 1950s to 1980s (but not today) with its very large and ideologically marginalized Communist Party. But these are not typical of PR systems more generally, and are not even remotely relevant comparative referents for a hypothetical PR system in the USA—for all its diversity, this country lacks the kind of rigid social divisions that give rise to parties like Shas or the former Italian Communists. Combine that with the probable high threshold our PR system would have (on the order of Germany and New Zealand) and our presidential form of government, in which the parties that can realistically elect presidents are sure to remain the most important players in the system, and it is clear that PR in the USA would hardly be as radical as Chris and Stephen fear. PR would moderate, not further polarize, our partisan competition.

22 thoughts on “What if the USA changed to proportional representation?

  1. I don’t know if proportional representation would give the Democrats the shock treatment they need, or if we’d get an even more fractured set of parties and interest groups to the right of Rush Limbaugh. I think that we’re even less likely to get any experiments in PR, now that the Republicans dominate national and state politics to the degree they do.

    But who knows? With PR, maybe even the Republican Party would fracture, with perhaps some ultra-religious and “revisionist South” parties forming on one wing, and a “traditional Republican” faction breaking off on the other. That’s exactly why, I’d guess, the Republican leadership would be strongly against any such experiments, which would more likely result in fractures at the very level–city, county, and state–where some of these experiments might take place.

  2. Well, I would expect the leadership of both big parties to be against PR. It is hard to think of a case of a transition from plurality to PR that did not take place only after one or both big parties found itself losing seats due to the emergence of a new competitor.

    That is, the break-up of the party system is a precondition for the adoption of PR.

  3. Current ANTI-Democracy math — half the votes in half the gerrymander districts is about 25 percent minority rule

    Democracy = Majority Rule

    Party Seats = Party Votes x Total Seats / Total Votes

    How soon before Civil WAR II happens if the gerrymander regimes continue ???

  4. Pingback: Fruits and Votes

  5. Hey, what if we created a mixed system? I mean have a system where 70% of the members are elected in mixed SMDs/MMDs using IRV/ STV and 30% elected in Topup areas using Open state lists. Each state should have at least 3 members and 2 Topup members, giving a total of 5 members. We’d use 2 ballots: 1 for the district members and 1 for the party vote. In states electing it’s members at-large using STV, the Topup members don’t represent a certain area but represent the state.

  6. The plurality voting system in the US has the following weaknesses:

    a) people cannot and often will not vote their conscience for a small party because they fear the vote is not only lost but will help the opposition win (e.g. green party spoiled the gore presidency in 2000).
    b) during election campaigns, only contentious voting districts and/or states are paid attention to by the respective parties.
    c) large fractions of the population feel unrepresented and therefore don’t vote.
    d) not enough competition for the 2 large parties. They never have to fear
    to go out of business or be endangered by smaller parties
    e) plurality voting systems almost always tend to produce 2 party systems.
    f) the absence of coalition building (as needed in proportional systems)
    lessens the need for compromising and abets strongarm tactics.
    g) too much stress on the individual. A proportional system leads to people
    look at the party as a whole,
    which moves the spotlight away from how sexy a candidate may be
    to what the party and its principles stand for.
    h) In proportional systems at least some members are selected from party lists.
    It may therefore not be as difficult to get in and therefore require less money
    to run for office. In a winner take it all system, like the US system, you have
    to win an entire district to get into the house. that takes a lot of effort and

    I like the German system which gracefully combines the idea of local representation and proportional vote. That system gives smaller parties a voice and even power sharing, as was
    proven with the green party-Social democrat coalition under Chancellor Schroeder.
    It can also produce a coalition of the 2 major parties (Social democrats and Christian Democrats) like is currently the case. It is just more interesting and participatory to have more than just the same 2 parties over and over again. Big parties rest on their laurels and need some fresh wind and some competition.

    In the US, it happens perhaps once per century that you have a new party arise
    that can field a good number of house/senate representatives.
    and for that, public discontent would have to be dangerously and extremely high.

    [MSS note: Footnotes to Wikipedia pages deleted. We don’t need Wikipedia here; the referenced themes are pretty much all covered somewhere here at F&V. It is also worth noting that embedding more than two URLs in a comment causes it to get intercepted by my spam filter. In any event, I recovered this seed from the compost pile, so welcome aboard, Christian!]

  7. The first job for any US electoral system is to rid itself of gerrymandering.
    The second is to guarantee freeness and fairness – stopping the sort of rorts that went on FL & OH during the last two presidential elections for example.
    Thirdly, Federal elections should be under Federal control with a neutral Electoral Commission – then you can sensibly look at voting systems.

  8. Dr, I think your formulation is sensible, except that I don’t think it is impossible to have all these reform discussions at once. And any PR system automatically fixes the gerrymandering problem, and makes the stakes of small vote differences much lower, thereby rendering moot the administration problems you note.

    But that the USA, alone among all FPTP systems that I know of, still tolerates legislators drawing district lines, and that it also, alone among all minimally established democracies, tolerates partisan officials administering elections, only shows how far we have to go before we can talk about the USA as a genuine modern democratic republic.

    (To be clear, my references above to FPTP systems and minimally established democracies means the USA stands out among a set of electorally based regimes that includes much of the developing world, as well as among its advanced industrialized peers.)

    (And also thanks for your comment to this thread. It is always nice when someone propagates an older planting; facilitating such propagation is the point of having recent comments displayed on the right sidebar.)

  9. I suspect gerrymandering is way overblown. Gerrymandering could disappear tomorrow, and the independent redistricting crowd would not be any happier with Congress or the various state legislatures. The country has gotten more polarized, polarization has a geographic flavor, and that makes “competitive” districts harder to draw.

    I can’t tell what the independent redistricting crowd fundamentally wants. Is it in fact competitiveness? If so, what does that mean? The academic and advocacy literature surrounding prop 77 contained numerous different conceptions of “competition.” One event went so far as to suggest competition would naturally flow from districting that minimized county splits.

  10. I believe that the question of polarization vs. gerrymandering has been addressed in the political science literature on district competitiveness, and that the weight of the evidence indeed suggests that gerrymandering is a much smaller part of the story than the c.w. implies. Of course, as is the case distressingly often, I can’t supply the relevant citations.

  11. STV is the way to go because it reduces gerrymandering. If 3, 4, 5 member districts are used, then more states are at large than not. Some states like Wyoming would only have one seat and the other thinly populated states with one seat now may only have two seats (If the Wyoming rule is used), but all the rest of the states elected 3, 4, 5 members.

    The question is how would a state like California carve its districts. Would communities of interest districting arise? Would rural areas get three seater and urban areas 5 seaters? Would districts be redrawn or just seats be readjusted like a 4 seater becomes a 3 seater, and a 4 seater becomes a 5 seater?

  12. MSS: gerrymandering is a much smaller part of the story than the c.w. implies. Of course, as is the case distressingly often, I can’t supply the relevant citations.

    Two that I know of are:

    Andrew Gelman and Gary King, “Enhancing Democracy Through Legislative Redistricting”, American Political Science Review, Vol 88, No. 3 (September 1994)

    John N. Friedman and Richard T. Holden, “The Rising Incumbent Advantage: What’s Gerrymandering Got to Do With It?”, Social Science Research Network (2005).

    Jack: I can’t tell what the independent redistricting crowd fundamentally wants.

    My theory is that they are driven by the belief that problems of government are the result of selfish behavior rather than institutional design. Forbid or severely constrain the bad behavior, and government will be fine. No need to consider real institutional change (e.g. PR), of which they are fundamentally afraid.

    “Non-partisan” redistricting is a pretty good example of this approach.

  13. I think that since primaries are very open in the US, closed lists (with primaries dictating the order of their candidates), would be the way to go. Every state with more than say, 7 seats, would. need to be split into districts no smaller than 4 and no bigger than 7 seats each. Seats would then be apportioned to each party according to the D’hondt or St-Lague formulas.

    Somewhat unrelated, but I think that in order to spread power a bit from the 100 ‘old men’, each state should have 3 senators. At each general election, ALL states would the have Senate races (preferabl under instant-runoff voting).

  14. In response to JD, primaries in the United States are not really that open. If nothing else, being able to raise lots of money is a huge factor.

    As for the Senate, I suppose that having a 150 member chamber, with a third of the members selected at each election from all the irregular electoral districts is a slight improvement over a 100 member chamber, with a third of the members selected at each election from most of the irregular electoral districts. But there is so much wrong with the U.S. Senate, and it is so impervious to reform, that if you could get something like that through there are many other changes I would try instead. Really this proposal is like changing the drapes in the Senate chamber. And even at that the Senate rules give so much power to individual Senators to gum up the works that the expansion to 150 Senators might be a bad idea after all; the chamber barely functions at its present size.

  15. A Senate of 150 or 200, with all 3 or 4 elected by PR within each state would be my preference. It does not run afoul of the constitutional principle of equal representation,* but makes it much less likely that a party could win a plurality of all votes cast for a given Senate, and yet fail to have a plurality of seats. (This anomaly has been the case every time there has been a Republican Senate majority; in each of these Senates, Democrats had more votes nationwide.)

    I’d probably argue for STV, given the small number elected, and the value of maintaining individual Senatorial stature. But any PR is fine by me.

    * Actually, my more sincere preference would be to dump that rule. But it is almost inconceivable. PR by party in the Senate is hard enough to imagine, but the case for PR could be made within the current constitutional confines.

  16. As I might have noted, my reading of the US Const is that changing the Senate to “4 per State by PR, 4-year non-rotating terms, VP to have casting vote” could be accomplished by normal constitutional amendment, ie 38+ States. Since no State would have its Senate voting power reduced, no State would have an individual liberum veto over the amendment.

  17. I would abolish the states altogether for other reasons, which would effectively reform the Senate as well, but that is really a comment for another thread.

    Otherwise, the current number of 100 Senators is perfect for a nationwide open party list proportional representation system. And as for the strange constitutional requirement of not reducing a states’ representation in the Senate, with this scheme each state would retain the same number of Senators relative to other states, namely zero.

    Alternately, keep the representation as it is, but greatly reduce the Senate’s powers, which was done with unrepresentative upper chambers in other countries. The selection method of the Canadian Senate is even more unrepresentative than that for the U.S. Senate, and in fact even more ridiculous, but that has not been a problem. The 1911 Parliament Act in the U.K is also a good model. If the U.S. Senate could only block legislation that altered the (admittedly ill-defined) powers of state governments, and couldn’t block budget bills at all, we could even return to the days when Senator were selected by State legislators, instead of being a second version of the House, albeit with stranger internal procedures and elected from even more irregular electoral districts.

  18. After the ambassadorial v senatorial iupper house thread I think I would advocate the governor plus 5 senators elected by STV, and the vote being by state on certain issues and by individual member on others. There may also be a case for assigning the senators by population with a state minimum of 1 or 2 although that would draw the liberum veto.

    Australia also has an equal representation + liberum vetorule but it applies only to Original States. The Northern Territory is likely to become a state at some point and will almost certainly not get equal representation with the original states. The most likely formula is to give the new state proportionate representation so that the quotas for Tasmanian and Northern Territory* senators are the same.

    *Most NT citizens seem to want to be called the State of the Northern Territory.

  19. Pingback: Democrats, socialism, PR, and Bernie Sanders | Fruits and Votes

  20. Pingback: Searching for the center: A citizens assembly for California? | Fruits and Votes

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.