Huffington Post has this commentary by Roger Stone. It points out that in nationwide polls, it appears that a majority of voters who intend to vote for Gary Johnson would be more likely to vote for Mitt Romney if Johnson weren’t running. But it also points out that in certain crucial swing states, there is data showing that Gary Johnson voters would be more likely to vote for President Obama if Johnson weren’t running.
Much commentary about the “spoiler effect” of minor party candidates is woefully unsophisticated. The Stone piece is a good reminder of that. There is other empirical evidence that the effect is complicated. In 2004, the nation’s leading polls determined that voters who intended to vote for Ralph Nader, by a slight margin, would have been more likely to vote for President Bush than for Senator Kerry. Also, in 1950, pollster Sam Lubell published his findings in “The Future of American Politics” that Progressive Party nominee Henry Wallace caused President Truman to defeat Thomas Dewey in 1948. And, a recent book called “Predictably Irrational” presented social science evidence to rebut the simplistic interpretation of “spoiling.”