Well, that would have to be the alternative explanation to what’s apparently happening, which is that the governor’s race is tightening in Virginia again. Cuccinelli/McAuliffe went from 42/48 in August’s Q-Poll to 41/44 now.

The poll indicates McAuliffe, the former chairman of the Democratic National Committee, at 44% among likely commonwealth voters, with Cuccinelli at 41%. The three point margin for McAuliffe, who lost a bid for the 2009 Democratic gubernatorial nomination, is within the survey’s sampling error.


The new poll suggests that Robert Sarvis, the Libertarian candidate who stands at 7%, could hold a key to the November election.

And that last factoid reminds me of the 2009 governor’s race, actually. Only, the one in NJ.

You might not remember this one: Christie/Corzine was one of those that superficially looked a lot more competitive than it actually was. And the reason? Overstating third-party performance:


Turns out that only about half the people who were supposedly voting for Daggett actually did: the rest mostly broke for Christie, thus giving him the election by a rather higher margin than RCP’s rule-of-thumb assessment would have suggested. Which is not surprising, given the way that third parties work in NJ:

Year Top 3rd 2PV
2009 6% 93%
2005 1% 96%
2001 1% 98%
1997 5% 93%

As the chart shows*, New Jersey elections are not particularly friendly to third parties: the six or so percent that Daggett received was easily in line with what New Jersey typically gives its top third party candidate in any given election. I would like to note, by the way, that I’ve been saying this about NJ and third parties in general for some time (both before [see also here] and after the 2009 election); this is not something that I’ve cooked up for this discussion.

I mention that because now let’s look at the VA third party vote situation.

Year All 3rd 2PV
2009 0% 100%
2005 2% 98%
2001 1% 99%
1997 2% 98%
1993 1% 99%
1989 0% 100%
1985 0% 100%
1981 0% 100%

Highlighted in red are elections where a third party candidate managed to get on the Virginia ballot; as you can see, the reward for such perseverance is typically somewhere between 1 and 2% of the vote. This is relevant because you have to drill down on the polls showing the current RCP average for the Virginia race:


The breakdown is as follows:

Now, it may very well be that this time the pollsters are detecting a genuine, sustainable Libertarian party surge in a state that has never been known for that sort of thing. Certainly the Libertarians would argue that! Or it could mean that Purple Strategies is correct, and that one-fifth of the electorate has simply not made up its mind yet, six weeks out. That honestly sounds more likely to me – and yes, I’m aware that it’s also the ‘better’ answer for the Republican party, thus making it automatically suspect.

Still, a tightening poll situation isn’t good news for the Democrats. Because the goal for them is to open up a lead – not least because Terry McAuliffe is apparently really bad at campaigning…

Moe Lane (crosspost)

PS: It hardly needs to be said, but: Ken Cuccinelli for Governor.

*Numbers for those two charts from Wikipedia, on the principle that surely they can get election results correct.