When Third Parties Do Well

The news today about the third party joke “Americans Elect” is that, well, it’s still a joke — they have lots of ballot spots, but still no candidate.

John Avlon thinks that’s too bad, but moreover thinks that Americans Elect is having trouble because they picked the wrong cycle:

Perhaps the biggest obstacle was the basic fact of this particular election cycle—when a president is running for reelection, it tends to be a referendum. Third-party candidacies do best when there is not an incumbent on the ballot or after an extended period of one-party rule with weak opposition.

Now, first of all, that’s an (unmarked) change from the original column, which made the factual error of claiming that Ross Perot wasn’t running against an incumbent in 1992 — it was immediately caught by lots of people over the twitter machine this morning, with Steve Kornacki being the first one I saw. I don’t have any problem with someone for making factual mistakes…I certainly make my share, and as long as you’re willing to correct it when called on it, that’s not a problem in my book.

However, the underlying point is wrong. The truth is that weak incumbent presidents create opportunities for third party candidates.

The strongest third party runs in the 20th century were Ross Perot in 1996, Perot in 1992, John Anderson in 1980, George Wallace in 1968, Robert La Follette in 1924 and Teddy Roosevelt in 1912. Five of six were against incumbents. If you go back to 1892, and the Populist campaign, make it six of seven. This makes the conclusion that they do best “when there is not an incumbent on the ballot” somewhat odd.

How about “after an extended period of one-party rule with weak opposition”? I’m not sure exactly what that means, but let’s see. If we limit it to control of the presidency, the incumbent party had been in for one term in 1996, three in 1992, one in 1980, two in 1968, one in 1924, and four in 1912 (plus, just one in 1892). I’m not really seeing a pattern there, but where there is doesn’t support Avlon.

It’s more complicated if we include control of Congress as part of “one-party rule,” but basically there’s nothing there.

Or, to put it another way…which elections have not had an incumbent president or an extended period of one-party rule? If we consider the latter any time in which the incumbent party has won three or more terms in the White House, then we won’t have much. I believe the only ones that fit are 1952 and 1908, neither of which produced a notable third party run. If we dial it back to two terms (and no incumbent on the ballot), we get 2008, 2000, 1988, 1968, 1960, 1952, 1928, 1920, and 1908 — and that picks up Wallace in 1968, but one out of nine isn’t a very good record (compared with five in the other 19 elections, if I’m counting correctly, from 1900 through 2004).

Again: third party opportunities are mostly created by unpopular presidents.

[Cross-posted at A plain blog about politics]

Jonathan Bernstein

Jonathan Bernstein is a political scientist who writes about American politics, especially the presidency, Congress, parties, and elections.