Third party spoilers

I have seen an interesting political meme a few times of late. It says that third party candidates are always spoilers, and cites the Republican dominance of the Presidency between 1896 and 1932 as partial evidence. The argument is that Eugene Debs pulled votes away from the Democrats, causing them to lose. It also cites other examples, but it’s the turn of the century example I’m going to address here. (Because it’s the one which is wrong. The others are pretty much accurate.)

I’m making this post because I spent a while putting together the data and I don’t want to lose it; also because I believe in spreading accurate information, where such is available. My data source is David Leip’s Atlas of U.S. Presidential Elections.

Content moved into the extended entry because it was messing up the front page.

Table One: Popular Vote

This table shows the popular vote for every Presidential election between 1896 and 1928. All third parties are lumped together, because I want to provide the most favorable scenario for the Democrats.

Year	Democrat	3rd Party	Republican
1896	46.7%	2.3%	51%
1900	45.5%	2.8%	51.7%
1904	37.6%	6%	56.4%
1908	43%	5.3%	51.7%
1912	41.8%	35% (27.4% Bull Moose)	23.2%
1916	49.2%	4.7%	46.1%
1920	34.1%	5.6%	60.3%
1924	28.5%	17.5%	54%
1928	40.8%	0.9%	58.3%

In every year in which the Republicans won the election, they had a simple majority of the popular vote. There are no cases in which the third party vote plus the Democratic Party vote add up to more than the Republican Party vote, except in 1912 when the Republican Party was split by Theodore Roosevelt — the one example of a successful third party spoiler from that era damaged the Republicans.

Also, that split was anomalous, as it was the result of corrupt politics on the part of the Republican Party. More information is available here.

Note that the extremely large third party vote in 1924 was also due to a Republican Party split. Robert LaFollete was the standard bearer for the Progressive Party. Don’t be misled by the name; LaFollete, though quite radical, was a Republican earlier in his political career and close examination of the electoral votes of 1924 shows that LaFollete took Republican votes in states that had voted Republican in the elections immediately previous and subsequent. The Progressive Party did not cost the Republican Party the election, despite one of the strongest third party showings of the 1900s.

Table Two: Electoral Vote

For each year between 1896 and 1932 in which the Republicans won, I list the actual Democratic and Republican electoral votes, the states that would have swung Democrat if all the third party voters went Democrat, and the hypothetical Democratic and Republican electoral votes.

Year	Actual Democrat	Actual Republican	Swing States	Hypothetical Democrat	Hypothetical Republican
1896	176	271	CA (8) KY (12)	196	251
1900	155	292	no changes	155	292
1904	140	336	MD (1) MO (18)	159	317
1908	162	321	IN (15) MO (18) MN (3)	198	285
1920	127	404	no changes	127	404
1924	136	382	AZ (3) ID (4) KY (13) MD (8) MO (18) MN (4) NE (8) NV (3) NM (3) ND (5) SD (5) UT (4) WI (13) WV (8)	235	283
1928	87	444	NY (45)	133	399

The 1928 electoral votes don’t add up because the Progressive Party drew 13 electoral votes that year. For the sake of argument, I’ve swung all the Progressive Party votes to the Democratic Party that year; the Republican Party would still have won by a slim margin.

Corrections on math errors and so forth are welcome. So are comments on the politics of the era. Note that this is a purely mathematical analysis of the question, and does not account for the possibility that the existence of a third party drove some voters to vote Republican.