It’s election season in America, and once again lots of people who are dissatisfied with their choices are calling for the creation of a third party. But I’m sorry to say that will never happen – at least not without major changes to the Constitution.
America has never had a third party that had any major impact at the presidential level that has lasted for more than one election cycle. There have been times when one of the two parties came undone and another rose to replace it (when the Federalists died, to eventually be replaced by Whigs, or when the Whigs died and the Republicans were born). But when a third party did well at the presidential level (never winning, but having an impact – 1912 and the Bull Moose Party, 1968 and the Dixiecrats, 1992 and whatever Ross Perot called himself), the party did not survive beyond an election cycle or two.
I thought about that the other day. There seems to be something about our system that keeps a third party from getting big. And I think I’ve figured it out.
(Note: I feel confident that political scientists have already figured this out. I don’t claim a unique genius. But it made for some interesting thoughts, so I figure I’ll write up my explanation. And next time online I have to tell someone that we won’t have a third party, I can just link to this.)
In the American system, for the most part, our elections are winner-take-all. Further, they are first-past-the-post – whichever candidate gets the most votes, even if it’s less than a majority, wins the election. There are no run-offs, and coming in second doesn’t gain you anything.
The implication is that if you have a large number of people who can work together within one party, they benefit from doing so. Let’s look at an example.
Suppose we have two parties: the Cat Party and the Dog Party. The Cat Party controls 60% of the electorate while the Dog Party has the other 40%. So the Cat Party will win the election. Would it make sense for the Cat Party to break into two other parties – say, the Siamese Party and the Persian Party?
Suppose the Siamese and Persian wings each control half of the Cat Party. So they could break away and form their own parties, with each ending up with 30% of the electorate. But in that case, both would lose the election to the Dog Party and it’s 40% support.
So there’s a strong incentive for the Siamese and Persian wings to stay together. After all, while each would prefer its own candidate to win, each would also prefer any Cat candidate to a Dog.
If there were a way for Siamese and Persians to form a post-election coalition to put the coalition candidate in office, as happens in parliamentary systems, it would make sense for them to stay distinct parties. But as long as the person who gets the most votes in the election gets the office, there’s a strong disincentive to have multiple parties. And making such changes would require changes to the Constitution.
(I’ll note that our system does have this in common with parliamentary systems: both encourage coalitions. It’s just that parliamentary coalitions tend to be between parties and form after the election, while in our system the coalitions forms within the parties and unite to contest the election.)
But it goes further than that. In American history, the two parties tend to be of roughly equal strength. And I think that’s also built into the system.
Look again at the Cat and Dog parties. In the real world, parties are not just cats and dogs. They are coalitions of groups, each of which has its own set of priorities.
Suppose, for example, that a piece of the Cat Party cares most about having milk with dinner. Suppose the Milk Coalition consists of 8% of the total electorate.
Non-milk Cats might say hey – we can win without the Milk Coalition. So let’s not bother supporting milk policies – they just use up resources that we’d rather apply to other things. We don’t need them, so why bother?
But the Dogs need more voters to have a chance at winning. And they don’t really care all that much either way about milk. So why not start supporting milk in order to get all those Milk Coalition voters into the Dog Party?
We’ve seen over the past fifty years a number of such shifts. As one example, the GOP has become the party of strong national security, so those Democrats who care most about strong military and foreign intervention have shifted over to the GOP as the Neo-Conservatives. In this and other cases when one party came to dominate, things eventually balanced out and the parties ended up at rough parity.
So to those who want a third party in this country, I’m afraid you’re out of luck. Unless, of course, you find a way to change the way our elections work. And I don’t see that happening.
And to those who want their party to become a permanent majority, you’re equally out of luck. By the nature of things, while you might get short-term dominance, things will balance out over time.
Note: I’ve just discovered that this phenomenon has a name. It’s called Duverger’s Law, and it even has its own Wikipedia page.