« Previous | Next »

Why Third Parties Can’t Work
Posted by tgirsch

Duverger’s Law. Via FactCheck.org

January 10th, 2008 Politics, Weekend Flame Bait | 20 comments

20 Comments »

  1. Peter Hughes writes:

    True. But here in the UK, where we have a similar winner takes all system to elect MPs, with the administration and official opposition based on the number of MPs, we still have a reasonable sized third party, with at least an outside chance of getting in power one day. The fact that the US has nothing like that does make me wonder if there is something else going on to explain the difference. Maybe its something to do with size? Or maybe something in the American psyche keeps pushing towards dichotomy, evidenced by a stout refusal to move beyond a liberal / conservative divide (a gradient at most)? Any thoughts?

    Comment 1/10/2008

  2. tgirsch writes:

    Peter:

    My guess, without knowing, is that your third party is strong in certain regions and weak everywhere else. Otherwise, I’m not sure of the logistics of how it can work. They have to be able to at least win SOME seats, yes? Also, is your voting system plurality-based, where a candidate can win without a true majority? I’d be interested in learning more. (PS: What’s an “MP?”)

    In the US there are two problems, as I see it. The first is that the third-parties are nationally rather than regionally based, which dilutes their support. The second is that there simply isn’t enough to distinguish them from the two major parties to siphon off many votes. Further complicating the latter is that the third parties tend to exist on the edges — they’re either VERY liberal or VERY conservative, which naturally limits your voter base. There aren’t any “moderate” third parties, at least not with any name recognition.

    Comment 1/10/2008

  3. Stormy Dragon writes:

    >There aren’t any “moderate” third parties, at least not
    >with any name recognition.

    There’s two types of moderates:
    1. People who are rightly seen as non-moderate, but in a way that doesn’t map well to our single-axis way of looking at politics.
    2. People who are moderate purely for the sake of being moderate.

    This first group is too small to form a competetive party. The second is big enough to form a competetive party, but is generally unable to campaign effectively because they have no specific ideas to run on.

    Comment 1/10/2008

  4. Dan M. writes:

    TG
    MP := Member of Parliament.
    Analogous to a congressman.

    Comment 1/10/2008

  5. Nomen Nescio writes:

    Duverger’s law rests on the premise that districts elect only one single representative each, the winner being whoever gets more votes than any other single candidate in that district.

    i would submit that this demonstrates a bug in that method of electing representatives, and we should change to a different method.

    Comment 1/11/2008

  6. digglahhh writes:

    The link provided doesn’t explain why third party a can’t work.

    It simply posits a when A is the case, B likely develops relationship.

    Such an observation neither explains why B occurs, nor expresses an opinion on whether the development of B is a good or bad thing.

    Maybe Duverger himself addressed these questions in his own work (I don’t know if I’m sufficiently motivated to research this), but they aren’t addressed in the link.

    Comment 1/11/2008

  7. Peter Hughes writes:

    Yup, Dan M got what an MP is!

    The third party is indeed strong in some regions rather than others, but still spread across the country (so all three are nationally organised). My question is really why this could not happen in the US. There are enough political configurations: At the very least, the Republicans could split into a big business / libertarian faction and religious nutjob faction. Obviously there are reasons why that coalition is desperately being held together, but it does show that there’s no reason why only two worldviews should dominate.

    Comment 1/11/2008

  8. tgirsch writes:

    Nomen:

    I agree. I’m in favor of instant runoff voting, myself (sometimes referred to as “single transferable vote” or STV voting.

    Digg:

    Duverger’s Law:

    The most obvious inhibiting feature unique to the SMDP voting system is purely statistical. A small third party cannot gain legislative power if it is based in a populous area. Similarly, a statistically significant third party can be too geographically scattered to muster enough votes to win seats, although technically its numbers would be sufficient to overtake a major party in an urban zone. Gerrymandering is sometimes used to counteract such geographic difficulties in local politics, but is impractical and controversial on a large scale. These numerical disadvantages can create an artificial limit on the level at which a third party can engage in the political process.

    The second unique problem is both statistical and tactical. Duverger suggested an election in which 100,000 moderate voters and 80,000 radical voters are voting for a single official. If two moderate candidates and one radical candidate were to run, the radical candidate would win unless one of the moderate candidates gathered less than 20,000 votes. Observing this, moderate voters would be more likely to vote for the candidate most likely to gain more votes, with the goal of defeating the radical candidate. Either the two parties must merge, or one moderate party must fail, as the voters gravitate to the two strong parties, a trend Duverger called polarization[1].

    And, as Peter has already pointed out, there are at least some counterexamples. But the American experience has borne out the theory time and again. And as I’ve mentioned before, in those infrequent occasions when third-parties do gain power, they generally supplant one of the previous two parties, such that you’re still left with a two-party system. As has happened here in the US:

    A third party can only enter the arena if it can exploit the mistakes of a pre-existing major party, ultimately at that party’s expense. For example, the political chaos in the United States immediately preceding the Civil War allowed the Republican Party to replace the Whig Party as the progressive half of the American political landscape. Loosely united on a platform of country-wide economic reform and federally funded industrialization, the decentralized Whig leadership failed to take a decisive stance on the slavery issue, effectively splitting the party along the Mason-Dixon Line. Southern rural planters, initially lured by the prospect of federal infrastructure and schools, quickly aligned themselves with the pro-slavery Democrats, while urban laborers and professionals in the northern states, threatened by the sudden shift in political and economic power and losing faith in the failing Whig candidates, flocked to the increasingly vocal anti-slave Republican Party.

    Peter:

    At the very least, the Republicans could split into a big business / libertarian faction and religious nutjob faction.

    But neither of those factions, by itself, has enough support to win any seats. And without winning seats, they have no power.

    it does show that there’s no reason why only two worldviews should dominate.

    Actually, they don’t. Two parties dominate, not two worldviews. Important distinction. A Democrat from Louisiana or Mississippi is a much different animal than a Democrat from New York or California.

    The parties don’t represent worldviews, but rather coalitions of worldviews, that compromise amongst themselves on the thinking that some influence is better than no influence.

    Comment 1/11/2008

  9. digglahhh writes:

    I’m not sure I understand the first paragraph. So, a third party, by nature presumably, must be either small but concentrated, or large but scattered? Why?

    Anyway, the reason given in the second paragraph is his acknowledgment of the same thing I was rambling about (unprovoked) in The Onion piece thread.

    People won’t support the third party because they care more about choosing the lesser of two evils than exercising their political will to support a candidate that most accurately represents their views. To me, that’s more of an issue of a voter limiting him/herself by neutering his/her vote, than it is some endemic feature of the system.

    The problem with supporting the candidate who has a shot to win as opposed to who actually represents your views is that it further marginalizes your own views. It continues to keep the “market” for real alternatives hidden, deterring others from supporting it, and ensuring major party candidates don’t have to re-align themselves to gain support.

    It’s like choosing to sit with the popular kids in the cafeteria even though you hate them. You’re a geek at heart, but don’t want to be socially ostracized.

    Perhaps if people actually voted for who they support, we’d see the true extent of the displeasure out there, and fringe voters would be encouraged by the showing. I don’t know…

    Also, a true third party would not only re-distribute existing voters’ support, it would summon and engage support from those who are silent in the political sphere.

    Comment 1/11/2008

  10. tgirsch writes:

    So, a third party, by nature presumably, must be either small but concentrated, or large but scattered? Why?

    I don’t think the “large but scattered” holds. That’s detrimental, not beneficial. As to why, it’s because unless a party can actually win seats, it has no meaningful power or influence.

    People won’t support the third party because they care more about choosing the lesser of two evils than exercising their political will to support a candidate that most accurately represents their views.

    I disagree with this assessment. People are not absolutists, and therefore they will not bet the farm on a hopeless cause. People would vote for a third-party candidate if they believed that candidate had a chance at winning. But in a winner-takes-all system, a vote given to a non-viable candidate is just as good as a vote given to the viable candidate you like the least.

    In other words, people are willing to compromise, and I contend that more often than not this is a good thing. If everybody took the stance of saying “give me everything I want or nothing at all,” the most frequent outcome by far will be getting nothing. Instead, they take what they can get. You might disparage that as “settling,” but it doesn’t make sense to refuse to settle unless the act of doing so somehow makes it more likely that you can get what you want at some later time. But most often in our system, precisely the opposite occurs.

    The problem with supporting the candidate who has a shot to win as opposed to who actually represents your views is that it further marginalizes your own views.

    That depends. Suppose viable candidate A matches 40% of your views, and viable candidate B matches 5% of your views, and you vote for non-viable candidate C, who matches 85% of your views. Suppose candidate B wins in a close election. How does this advance your views, in a way that voting for candidate A (and helping her win) wouldn’t?

    Your hypothetical only makes sense if your views are between those of candidate A or B, not if they’re at one end or the other. (Or, if the positions of A and B on the issues you care about most are so close as to be indistinguishable.) Because the inevitable response of the losing major party will be a move to the middle, to attempt to capture that median voter. People emulate winners, it’s a natural tendency. So if a conservative loses to a liberal, he’s not likely to respond by becoming even more conservative. That doesn’t make sense.

    Perhaps if people actually voted for who they support, we’d see the true extent of the displeasure out there, and fringe voters would be encouraged by the showing.

    But that can only ever work if everyone does it. What you’ve had for the past couple of decades is that one side (liberals) has done it a lot more than the other (conservatives), and so the conservatives win. Otherwise, it’s like a football team who has decided that the forward pass is unfair and detrimental to football. Sure, they could refrain from passing and just run on every down, but unless they get the other teams to agree to do the same thing, they’re going to lose a whole lot of football games.

    Also, a true third party would not only re-distribute existing voters’ support, it would summon and engage support from those who are silent in the political sphere.

    Previous caveats aside, that would depend on what that third party actually stood for. It would have to be viewed by a significant portion of the populace as being better on most issues than either of the two main parties. And that’s a tough trick to pull off.

    Honestly, I understand your objection in principle, I really do. And I’d actually like to see a system in which more parties could compete and have influence. But that is simply extremely unlikely to happen in our system. So unless we change the system, strategic voting continues to make sense. As long as you have a preference (positive or negative) between the two viable parties/candidates, it behooves you to vote for the one you prefer, or against the one you like less.

    I again return to the Bush/Gore/Nader problem. Were the differences between Nader and Gore really more pronounced than the differences between Gore and Bush? Not even remotely close. Nader was farther to the left than Gore, such that if you liked Nader but didn’t like Gore, you were going to absolutely despise Bush. But Bush is arguably what your Nader vote actually got you. [Insert angry response from LarryE here. :) ]

    Now, in my perfect world, as I’ve said above, you’d have IRV (instant runoff voting), and you’d be able to list first choice Nader, second choice Gore. In that scenario, when Nader is eliminated, his votes shift to the second choice. (Ditto for, say, Buchanan/Bush.) That way, choosing to cast your vote for a third-party candidate no longer is a high risk proposition in a close race, because if your candidate doesn’t get the support you hope for, your vote falls to your second choice.

    Comment 1/11/2008

  11. Lean Left writes:

    […] Let’s try this another way. […]

    Pingback 1/11/2008

  12. Stormy Dragon writes:

    >I agree. I’m in favor of instant runoff voting, myself
    >(sometimes referred to as “single transferable vote” or
    >STV voting.

    Every voting method has flaws (see Arrow’s Impossibility Theorem). IRV in particular is non-montonic (it’s possible to lose because you have too much support), it’s suceptible to cloning (having two or more candidates in the race that the voters perceive as essentially identical), and favors candidates with narrow but enthusiastic bases over those with broad but milkwarm support.

    Comment 1/11/2008

  13. Stormy Dragon writes:

    IRV doesn’t help third parties like you think either. Your argument is that people can afford to vote for third parties without fear they end up spoiling the election for the major party the like best.

    This is only the case with non-competive third parties. To use your example, sure, you can vote for Nader without costing Gore the election only if Nader’s support is 5%. If, however, Nader’s support is something like 25%, then he can suddenly start acting as the spoiler again.

    The other point is that eliminating a non-competetive third party’s spoiler effect isn’t necessarily a good thing either, since it means that the major parties can ignore them entirely. Using the threat of costing them an election in order to get a major party to implement some of your platform is often the only way supporters of an non-competetive party have anyway of affecting politics.

    Comment 1/11/2008

  14. tgirsch writes:

    Stormy:

    I don’t claim IRV is perfect; it’s just a lot better than what we have today.

    And, of course, the “competitive third party” problem is purely hypothetical, since our system all but guarantees that no third party can ever garner such support, except in the very rare case that one of the two major parties is in such serious decline, that another party must rise up to replace it. And in that case, you’re still dealing with a two-party system.

    As an aside, I have to say that I find it thoroughly amusing that someone who just accused me of being contrarian spends the vast majority of his time on this blog pooh-poohing everything. :)

    The other point is that eliminating a non-competetive third party’s spoiler effect isn’t necessarily a good thing either, since it means that the major parties can ignore them entirely.

    And they don’t ignore them entirely right now?

    Using the threat of costing them an election in order to get a major party to implement some of your platform is often the only way supporters of an non-competetive party have anyway of affecting politics.

    I don’t suppose you’ve got some discrete examples of this ever actually working, do you? I suppose if you wanted to use the Dixiecrats as your shining example, but that’s not a true third party; it was the result of a rift in one of the two major parties, and quickly re-consolidated into the two-party system.

    Comment 1/11/2008

  15. Stormy Dragon writes:

    Well, as an example, I’d say that when the Libertarian portions of the vote was increasing in the early 80’s, the Republicans started focusing a lot more on limiting the size and power of the government in an attempt to coopt those voters. Now they’re much more concerned with the Constitutional party peeling off voters, so they’re abandoning that and pushing for more social regulation.

    I’d expect you could find similar reactions by the Democrats after 2000 in order to capture Green voters, but I don’t follow DNC intraparty politics closely enough to give specifics.

    Comment 1/11/2008

  16. tgirsch writes:

    I’d say that when the Libertarian portions of the vote was increasing in the early 80’s, the Republicans started focusing a lot more on limiting the size and power of the government in an attempt to coopt those voters

    Two problems with this. First, Reagan was campaigning on small government in 1980, so your “early 80’s” timeframe has to be a little bit off. Second, unless you can point to libertarian candidates actually costing Republicans elections in ‘76 and ‘78, the third-party apologist’s theory doesn’t hold water. (It’s possible that the Republicans noticed and taken advantage of a growing libertarian trend in the electorate, but that doesn’t necessarily have anything at all to do with third-party politics.)

    I’d expect you could find similar reactions by the Democrats after 2000 in order to capture Green voters

    Not really. I think you’re missing a key part of the dynamic here. Any change in a party’s policies/stances is going to attract some new voters, but will also deter some of your existing voters. Unless the number of new voters attracted significantly outweighs the number of existing voters you’re likely to deter, the change in platform doesn’t make sense.

    In any case, because the third parties in this country tend to be closer to the poles than the major parties, rather than in between the two parties, it makes more sense for the major parties to continue to ignore the third parties, and instead try find a way to attract the middle-of-the-road voter who may have stayed home in past elections.

    Comment 1/12/2008

  17. Dan M. writes:

    SD,

    I think you need to re-read that Wikipedia article on Arrow’s Impossibility.

    It makes perfectly clear that “every voting system has flaws” is only one possible interpretation of Arrow’s proof, and a fairly unsophisticated one at that. Very bluntly “If we assume that X, Y, and Z are reasonable, then domain D has no solutions.”, suggests that at least one of X, Y, or Z is not reasonable, not that D has no solutions.

    Even assuming Arrow’s right (which he certainly seems to be), the conclusion is that his ideas of “fairness” are wrong, not that there are no fair voting systems.

    In particular, the formalization of “non-dictatorship” is muddled. It amounts to “Nobody is always right.” rather than “Everybody matters.” A much clearer articulation would be that forall R in L, for all permutations R’ of R, F(R) == F(R’). Most likely, that plus Pareto efficiency imply something that would meet the vernacular demand of non-dictatorship.

    Comment 1/12/2008

  18. digglahhh writes:

    TG,

    I understand your position too.

    One of my main problems with your position is that it treats “support” as some sort of mythical property bestowed on a party by a force beyond our control.

    I don’t assign as much blame to the system as you do. When you make statements along the lines of, I’d like to see a system where that is possible, you downplay individual agency in the whole process. “The system” prevents a lot of things - it did not prevent anybody from checking of Nader’s name! Only people prevented themselves from doing that. You can say that the reason you didn’t vote for Nader didn’t have to do with Nader himself, but nothing was stopping you from flicking his lever.

    Support has to be built, people potentially have to sacrifice in order to help build that support. They have to come to terms with the idea that their vote for the third party may be rendered meaningless, or even detrimental in the short term. It’s kinda like rebuilding a team in sports. Playing the rookie with potential over the aging veteran with the hefty contract is not going to help your team win more games right now - but your goal is to build a strong team for years to come, and rid yourself of the dependence on this washed-up, overpaid, vets.

    Finally, I will address a couple of things about the Nader/Bush/Gore example you raised.

    1. Gore won the election - so the whole point is moot. Bush didn’t win the presidency because people voted for Nader. He won the election because the Court stepped in and gave it to him. Perhaps, Nader made it close enough that something like that could happen - but the Court stealing an election for Bush is not something that you can blame on Nader. Sorry.

    2. I live in NYC, so I can vote for whoever I want, and the party of lesser evil will not miss my vote. :)

    3. I also think it is entirely legitimate to argue that the difference between Nader and Gore is greater than the difference between Gore and Bush.

    Comment 1/14/2008

  19. tgirsch writes:

    digg:

    I understand your point about the system versus the individual. But the system is corrupt in a lot of ways, and people (in the aggregate) are very easy to manipulate. The current system makes it far too easy for the unscrupulous to take advantage of that fact.

    I also understand the idea of a “rebuilding year,” but it’s a little different when you’re talking about a rebuilding four years or a rebuilding eight years. The sports analogy isn’t great, anyway, because every year you get to start over at zero, while in real life, the losses of the previous year carry over and have a cumulative effect.

    I will say this, though. I won’t go so far as to say it never makes sense to vote for a third party. Just that it only makes sense in very limited circumstances, which don’t occur very often. In my mind, the question is “What is that third-party vote likely to accomplish?” If the answer is “nothing,” or, worse, “something counterproductive,” then I think that vote is a bad idea. And I think that that’s usually the answer. Honestly, I don’t see why that’s such a controversial proposition.

    As to your three points:

    1. Gore won the popular vote, but that’s not the same thing as winning the election. Maybe there’s one out there, but I have yet to see a count of Florida that has Gore winning that state.

    2. And in that circumstance, I have no problem with you doing that.

    3. I’d be interested in hearing that argument. Because on the issues I care the most about, the differences between Bush and Gore are cavernous. Gore and Nader, not so much.

    4. (I know you didn’t have a 4, but I’m adding one anyway). This has nothing to do with anything, but Nader’s kind of a dick. Don’t get me wrong, he did great things for consumer protection, especially in the 70’s, but you can do great things and still be a dick. :)

    Comment 1/14/2008

  20. Morris writes:

    “the differences between Bush and Gore are cavernous.”

    You got that right.

    Comment 1/14/2008

Leave a comment