Megan McArdle responds to my previous post about third parties, suggesting that Barbara Ehrenreich (and I) have “about as tenuous a connection to reality as the folks who brought us Pepsi Clear.” Her counter-argument:
- That ‘first-past-the-post’ voting tends to produce two party systems.
- That presidential systems are much more prone to two-partyism than parliamentary ones.
- That the reason why Ehrenreich’s (or indeed McArdle’s ideas) don’t become policy isn’t because they’re blocked by the system, but because most Americans disagree with them.
- Therefore: third-partyism is an exercise in futility.
These arguments are exactly the sort of thing that we political scientists like to claim that we know something about (I note in passing that Megan’s confident assertion of these empirical relationships sits somewhat awkwardly with her belief that political science doesn’t have much to do with falsifiable predictions). On the first of these claims, there’s evidence from the literature to suggest that McArdle is sort of right (but not in a way that really helps her overall argument). On the second, there’s evidence to suggest that she’s fundamentally wrong. On the third, she seems to be on thin ice (if she’s making a limited claim) or falling through into the river beneath (if she’s making a strong general argument).
McArdle’s first claim is that ‘first past the post’ voting leads to two-party systems. This is a version of what political scientists call Duverger’s Law, which states in its original formulation that “the simple-majority single-ballot system favors the two-party system.” And indeed, there’s a fair amount of evidence to suggest that this law has real explanatory value, although there are many counter-examples.
This brings us on to McArdle’s next point. If there are important counter-examples to this general tendency, then McArdle’s big claim about American politics could be in trouble – the US could very easily be one such counter-example. Thus, McArdle makes a second claim – that the pesky counter-examples of countries with first past the post voting and more than two parties, are parliamentary democracies, not presidential ones like the US.
in those systems [parliamentary democracies], there is always the possibility that the winning party will not win quite enough votes to form a government, and thus your little party will get to be the coalition partner reaping political rewards out of its weight class—as the Greens in New Zealand, and the NDP in Canada. So in some ways, the parliamentary system actively rewards third party formation.
In the United States, on the other hand, being the guy in the little third party won’t get you the presidency, and will be a positive handicap in the legislature, because our legislature, for various reasons, doesn’t have anything like the party discipline of British-style parliamentary systems. Thus, our opposition often gets to help shape policy. This means that there is a large benefit to voting for the dominant opposition party.
This claim, while not implausible on its face, is unfortunately disconfirmed by the available evidence. Let’s take the general claim – that presidential systems are more likely to have competition between two (and just two) parties than parliamentary ones. In a forthcoming article in Comparative Political Studies, Terry Clark and Jill Witrock find strong evidence that precisely the opposite is true. On the basis of a regression analysis of electoral results in post-Communist Europe, they conclude that Duverger’s law is considerably weaker in countries with strong presidencies: these countries are more likely than countries with parliamentary systems to have several parties. As they note, there is a secondary literature strongly suggesting that this effect has more general application than Eastern Europe. Thus, it is in countries with strong presidential systems where we may expect Duverger’s law to apply weakly, if at all.
Indeed, when we look at the specific case of the US, this expectation is partly confirmed. Duverger’s law does not lead to a dearth of third parties. As the eminent political scientist, William Riker remarked in a classic article on Duverger’s Law, the US political system provides strong incentives for third parties to throw their hats in the ring, explaining what Riker describes as the “almost constant supply of third parties” (p.765). Hiroshi Okayama has a very interesting recent paper that speaks further to this, drawing on evidence from post-Civil War politics. While these parties may not last, they may have very substantial political effects.
McArdle’s final assertion is that the reason that Barbara Ehrenreich’s ideas (or indeed McArdle’s own) don’t have popular traction isn’t because of the system – it’s because a majority of the population disagrees with them. This claim is somewhat naive if it’s supposed to be a generalizable argument. If McArdle is arguing that political structures can’t restrain popular sentiment, there’s ample evidence to contradict her (cf, for example the lack of a death penalty in many countries where it would receive overwhelming popular support). If she’s arguing instead that the specific issues that Ehrenreich favours are incompatible with US public opinion, she needs to provide much stronger evidence than her (currently unsupported) assertions to make that claim stick. She may be right about some of Ehrenreich’s policy goals – but she’s probably wrong about others – and she needs to say what basis she’s arguing on. Contra McArdle , a majority of US citizens could probably be persuaded of the benefits of universal health care. The reasons that this isn’t on the agenda have more to do with the weight of special interests in US politics than public opinion. More generally, Cass Sunstein’s idea of constitutive commitments is a good way to get a handle on the question of what policies are feasible, and what aren’t in US politics. The general point still stands – some proposals that Ehrenreich (or indeed McArdle) would like to see on the public agenda are probably quite compatible with these commitments and would be on that agenda under different political structures.
Finally, it’s hard to see how any of this has bearing on the question of whether leftists can “use a third party to pull the Democrats back to where we want them.” Harry makes the counter-argument in comments, and does so, I think, convincingly.
Defensive dismissal of the likes of Ehrenriech is calculated to suppress debate within the DP both about whether it should be the place for the left, and what direction it should take. Given the rules of the game the most likely outcome of any serious and moderately successful left organising effort outside the DP is that it will get absorbed into the DP, and change the DP’s internal politics (a bit) as a result. Taking as your staring point the idea that the DP is your home (if you are on the left) is a sure-fire way of diminishing your influence on it.
BE knows all this. So, I suspect, does Brad De Long. BE is a left-of-center social democrat in European terms. Brad is a moderate conservative/Christian Democrat in European terms. The DP is very much in the hands of people with Brad’s basic outlook, which is one reason that the Nader issue still hasn’t gone away, despite the all the whining of the DPers who wish they could take the left as much for granted as they do, eg, African Americans. If the left isn’t willing to consider alternative to the DP they can forget about influencing it.
Update: See Daniel Geffen for further argument.