On September 18, Quinnipiac Polls released a poll for the Virginia gubernatorial race, showing Libertarian Rob Sarvis at 7%. Democratic nominee Terry McAuliffe is at 44% and Republican nominee Ken Cuccinelli is at 41%. Undecided is at 8%.
When Quinnipiac last did a Virginia gubernatorial poll, on July 16, it did not ask respondents about Sarvis. The July 16 poll showed that “someone else” (which had to be volunteered by the respondent) was only 1%. Thanks to PoliticalWire for the link.
It is not true that polls necessarily overstate actual support for minor party and independent candidates. In November 2012, the actual results for three Libertarian U.S. Senate candidates were approximately as high as the polls had suggested. Andrew Horning in Indiana got 5.67%, Jonathan Dine in Missouri got 6.07%, and Dan Cox in Montana got 6.56%.
Scott Rupert, an independent for U.S. Senate in Ohio, got about 6% without proper polling.
I think the main thing is polls often overstate actual support of a third party candidate when they don’t include all ballot-qualified candidates. Gary Johnson polled much higher than he finished in some polls in part because they didn’t include any other candidates voters would see, other than Mitt Romney and Barack Obama. When included with Jill Stein and sometimes Virgil Goode, a more appropriate 1-3% was usually the result. Well-publicized debates also tend to drag down a 3rd party candidate closer to election day if he or she is not included.
In this case, Robert Sarvis is one of only three candidates. Also the 1-2 debates so far (one was a “forum”/quasi-debate) that didn’t include Sarvis were not really followed or publicized too much; the only people who saw them were folks who really sought them out. People are turned off from the major party candidates in this race, so the silver lining of Sarvis’ exclusion from debates is that it doesn’t give his opponents a major edge vs. no debates at all. It looks like he may actually qualify for an October debate in the Roanoke area anyway: http://www.timesdispatch.com/opinion/our-opinion/columnists-blogs/bart-hinkle/hinkle-one-candidate-is-focused-on-issues/article_2d6dc94b-683a-5175-ac8e-cf9955eaced8.html
Oh also it should be noted Sarvis is polling at 10% according to the less-well-known Harper Polling: http://www.conservativeintel.com/917-conservative-intel-poll-of-va-mcauliffe-has-the-upper-hand/#
Rather then saying that polls overstate the support for “Third” party candidates, it would be more accurate to say that support falls off on election day. In effect, the regime parties are “stealing” votes from the opposition parties.
I agree. Support for the weakest of three candidates dropping off just before the election is pretty common, even when third parties are not involved. In Maryland’s 2006 Democratic US Senate primary, the top three candidates were Cardin, Mfume, and Lichtman, in that order. Mfume and Lichtman were the most progressive, so when it became clear about a month before the election that Lichtman was going to come in third, many of his supporters defected to Mfume, and Lichtman’s poll numbers quickly dropped by half. The polls just before the election accurately predicted the results, so it was pretty clear what had happened here.