Political vs economic competition, or why a two-party system can be OK

Max Kaehn, an occasional MR commentator, expressed a common sentiment when he wrote:

You think a voting system that sticks us with a two-party cartel instead of a diverse market in political representatives isn’t a major problem? Are you sure you’re an economist?

Here are a few reasons why political competition isn't the same as economic competition:

1. Economic competition lowers costs.  For the average worker, it cost a month's wages to buy a book in eighteenth century England and today it might cost well under an hour's wage.  The competitive incentive to use and introduce new technologies drove that change.  Political competition may support cost-reduction enterprises in an indirect manner, by providing good policy and spurring the private sector, but the mere ability to supply candidates and parties at lower cost is no great gain.

2. Having lots of parties means you get coalition government.  This works fine in many countries but again it is not to be confused with economic competition.  Coalition government means that say 39 percent of the electorate gets its way on many issues, while 13 percent of the electorate — as represented by the minor partner in the coalition — gets its way on a small number of issues.  Whatever benefits that arrangement may have, they do not especially resemble the virtues of economic competition.  

3. Many people think that greater inter-party competition, and/or more political parties, will help their favored proposals.  Usually they are wrong and they would do better to realize that their ideas simply aren't very popular or persuasive.

4. Often the U.S. system is best understood as a "no-party" system, albeit not at the current moment, not yet at least.  The bigger a party gets, often the less disciplined it will be.

5. Stronger electoral competition, in many cases, brings outcomes closer to "the median voter or whatever else is your theory of political equilibrium."  That's better than autocracy, but again there are limits on how beneficial that process can be.  It's not like economic competition where you get ongoing cost reductions in a manner which saves lives, brings fun, and enriches millions.

The bottom line: Political competition is better than autocracy, but its benefits are not well understood by a comparison with economic competition.


Many people think that greater inter-party competition, and/or more political parties, will help their favored proposals. Usually they are wrong and they would do better to realize that their ideas simply aren't very popular or persuasive.

I strongly disagree. Being a vital piece of a coalition is usually the *only* way that very unpopular (if not critically important) positions can be made into policy.

I see few true multi-party governments that work (e.g. Italy). The three largest economies for the past few decades have used a two-party systems, a de facto two-party system and a de facto one-party system.

Well, now you're changing the subject. Earlier you said "Range voting [ie, a multi-party system] is a solution in search of a problem." Now you're merely saying that political competition is different that economic competition, which is true.

Range voting, and the resulting multi-party system, would be about the best improvement you could make to American democracy. Since we'll never have it, I'd settle for congressional term limits, or quadrupling the size of the House. Of course, we'll never get those either.

@Bill: See virtually anything written by Anthony Downs.

Of course, parties don't converge on the median voter because politicians seek to maximise policy satisfaction, not just vote share. That is, they care about some things. File under ``shocking but true''.

@Ted Craig: You're selecting on the dependent variable. (And also it might be mistaken to lump China in here; we should perhaps restrict the analysis to countries in which a modicum of political competition exists). What about the two party systems with crappy growth?

Coalition government means that say 39 percent of the electorate gets its way on many issues, while 13 percent of the electorate -- as represented by the minor partner in the coalition -- gets its way on a small number of issues.

We get this anyway, for whatever that's worth. 51% of a population elects a Republican and the next thing you know they're handing out favors to the <5% hyper-religious-nutjob demographic. (I'm sure there's an equivalent interest group on the D side I just can't think of one right now.)

I don't know that parties that get less than 10 percent of the vote are significant. And does the CSU really count? That's like saying Obama presides over a coalition of Democrats and Farmer-Laborers.

Politicial systems are often a result of the make up of the population they govern. A highly fractionalized society tends toward coalition governments and multiple parties. When divisions are minor or marginalized by apathy, two party systems are more common. Political systems are organic and probably could not be constructed for optimal performance.

I think the main advantage of new voting systems is that they'd increase the number of competitive seats. I'm a Democrat that lives in Maryland, but I do think that we do have some bad features of a dominant party government despite the fact that it's my party. This isn't to say that I want the Republicans in power, but having most of the competition in the primaries doesn't seems inferior to other achievable options.

That said, many more competitive elections and two year House of Representatives terms might make a bad combination. Staggered four year terms but elections that are actually competed might involve a similar amount of time spent campaign but more effective accountability.

Tyler, the dominant strategy in a democracy is COLLUSION, which makes #4 a real stretch. The point is not whether any party acts as one, but whether the political representatives are able and willing to do their job. A lot of things "CAN be ok", but the question you should be asking is whether things CAN BE BETTER. 1990's search engines were ok, but google is way better. The web is increasingly becoming a more social animal and thus a political tool. We don't need more parties necessarily, what we need is:

1. an ability to weight sophisticated expert opinion from plumber joe
2. to reduce incentives for collusion

The algorithms that will be created to aggregate crowd WISDOM will make the simple majority rule seem barbaric. The Bayesian Truth Serum is a great start and very difficult to manipulate.

Two ideas:

1. The US system does not necessarily need to be a two-party system. That is the equilibrium for any specific election, but only one election in this country is national. There is no institutional prevention of the formation of regional parties, with two competing in any area and coalitions forming at the national level. I actually wonder what holds this back, since any individual state Republican or Democratic party would be better able to compete if it weren't bogged down by a national party's policy priorities and other baggage. Money doesn't seem to be enough, although simple inertia could be.

2. I also prefer range voting, generally, but I have also been fooling around with a random draw system, with the likelihood of winning based on the percentage take of a typical one-vote election (with a low cutoff to avoid the fringe). This would result in inferior choices for any single election, but the long-term results of a series would be much more in line with the priorities of the electorate. In a century's worth of elections, why should a district populated 60-40 for party A elect party A 100% of the time?

This is a complete non-sequitur: you make an argument that competition in politics is not the same as in economics, and then go on to claim that therefore the two party system is OK. You may be correct in the first claim, but that the two party system is OK in no way follows.

You also ignore in the first claim the effects that competition qua competition have on other market actors. In the face of competition market actors have an incentive to run more lean and efficient operations that better fulfill their customers wishes. You hint at this, but ignore the political corollary: more political competition gives political actors incentive to be more efficient: to better represent their constituents (or for parties to better represent their members). Political competition therefore produces not only more candidates, but, like economic competition, it also produced better candidates.

Competition is good as such for a variety of reasons, this is a lesson we should all extrapolate from economics to politics.

Strongly disagree as well. A two-parties cartel brings the cost of corruption down. Indeed, it is fairly easy to come up with an agreement between the parties to share the cake, whereas if there was a third or fourth party, the coallition would be unstable, and the risk of an opportunistic wistle-blower much higher.

The example that comes to mind is post-WWII Italy where the communist party and the christian democracy agreed to rule together which totally canceled the check-and-balance system. No wonder that the Sicilian mafia prospered in this environment.

Posted by: David Heigham
"A cartel of incumbents (think California legislature) becomes a bar to the flow of innovation. "

Last I heard, the California legislature has been subject to some of the more stringent term limits in the country.

Is MR aware of Buendo de Mesquita's 'Selectorate vs Winning Coalition' work?

Anyway, A lot of what would be considered inter-party bargaining in a multiparty system in fact takes place within our parties - as anyone who has had to deal with Blue Dogs would attest to of late. Also, the LDP party in Japan, which dominated Japanese politics since the end of the war until recently, has also not been a monolithic bloc but, in reality, a plethora of smaller, competing factions. So, the real issue is not the number of parties per se but competing factions working both within and outside the two mainstream US parties.

The real problem with the US system is the blatantly undemocratic spanner states throw into the work. California or Texas each have many millions of more people than states like Montana and Idaho, but they get as many Senators as these states... not to mention the additional weight they have in presidential politics due to the electoral college. This wouldn't be so bad if, like the rest of the developed world, our upper house had very little power... but it does, and so it hugely distorts our political outcomes to favor vested interests in the smaller states over ostensibly numerically far larger populations. What it really does is give too much influence to the least populated, most backwards parts of the country at the expense of the more populated, wealth-generating centers of the country. The situation is not altogether different from the problem posed by the so-called rotten boroughs in the British parliamentary system before the reform acts of the 19th century.

"Labour polled seventy thousand fewer votes in England than the Conservatives, yet won ninety-two more seats." - from Wikipedia. If America ever develops a meaningful third party, as most plurality voting countries do, expect this kind of undemocratic nonsense.

However the political scientist Lijphardt simply found that these "multi party system" states worked better in virtually all regards: http://www.amazon.com/Patterns-Democracy-Government-Performance-Thirty-Six/dp/0300078935/ref=ntt_at_ep_dpi_1

There are methodological issues here, but the discussion is somewhat more informed than the one on MR!

No mention of Duverger's law? Man, this really is an econoblog.

Agree strongly with Alex and Ben. The issue is the cost of collusion (gerrymandering) and the barriers to entry. Both would be eased by a switch to PR, but that's not going to happen anytime soon.

Between 1992 and 2004, neither major party received a majority of the popular vote in Congress. The peak for non-duopoly voting (essentially "throwing their vote away) reached as high as 6.1% in 1994. That would be 27 votes under PR to a third-party coalition, certainly enough to throw some weight around and, more importantly, to provide an example that there is an alternative way of doing these things. Furthermore, it is evidence of a dissatisfaction amongst the American voter.

I don't believe a third-party coalition is that difficult to imagine. Anti-status quo pols do (should realize) their commonalities outweigh their differences.

Also, John has it precisely backwards with regards to the interaction between electoral systems and the politicalization of society.

Comments for this post are closed