Wednesday, July 22, 2015

Quinnipiac is Back and Worse Than Ever

From the pollsters that brought you the miraculous 20-point Rand Paul Pennsylvania swing of Spring 2015...

From the writers of the press release gleefully stating that Hillary Clinton was "wilting" in Virginia while their own poll simultaneously showed her surging into a solid lead...

From the guys who are somehow making Gravis Marketing look amazingly reliable right now...

...comes the first real Republican swing state blood bath poll of the 2016 campaign. Make sure you're sitting down, or ready to laugh (like I did) and check it out.

There are so many strange things that could be pointed out here, such as:
-Hillary has led Jeb in every single (non-Fox News) national poll for months, often by double digits. If all those results are accurate, this poll must have been taken in an alternate universe.
-The party IDs in the samples are 2 to 8% more Republican than we saw in the last presidential election.
-Hillary's favorability in Iowa (33%) is so ridiculously dismal that even Rick Perry is doing better (36%).
-In the span of three months - for no apparent reason - Hillary's support among women in Virginia dropped by a whopping 16 points.
These are just a few of the headscratchers. For more, check out this excellent post by Steve Singiser at Daily Kos.

In the six months that Hillary HQ has been in existence, I've noticed something: Polls are coming out constantly and it can be hard to keep track of all of them, much less analyze them. To be perfectly honest, it can be exhausting. The good news is that the vast majority of these polls have ranged from decent to spectacular for our candidate, and even if there is an occasionally lousy poll, you usually only need to wait a day or two before there's another one that cheers you up.

So I'm going to let that happen and not lose any sleep over this one.

Well. Almost.

Take a look at one last thing from this new poll: The crosstabs of the Clinton vs Walker matchup in Virginia, where Walker leads by 3. You'll notice that Hillary is doing a little better with her party's faithful than Walker is doing with his, and the Indies are almost exactly split. Considering that the party ID is +1% Democrat (28% D - 27% R - 35% I), how in the world is Walker up by 3 points here? Even with the most generous of decimal rounding, it does not seem possible to this blogger who got a B in college business math.

So...calling all math nerds out there: If you can somehow get these Quinnipiac crosstabs to add up to a 3-point Walker win, please lay it out in the comments below. The Bush & Rubio numbers from Virginia also make about as much sense, if you'd like to see more.

Rest assured that I am more than willing to concede that I'm out of my league on this one if presented with enough evidence.


  1. Don't waste your time trying to make sense of their bonkers numbers.
    Conservative analyst Sean Trende says to ignore.
    Andrew Baumann noticed 2012 CO electorate was D+5, compared to R+3 in the poll
    Chris Hughes noticed their use of random digit dialing skews the independent (lean R) number in Colorado.
    Mark noticed only 28% of voters in VA were D, compared to 39% in 2008/2012, and 35% in 2014.

  2. Bloomberg explains it pretty well.

    " we see that the electorate in the Quinnipiac polls in all three states looks more Republican and more like the 2014 mid-term electorate than the 2016 mid-term electorate. "

    "While the general election electorate in 2016 may not be as Democratic as it was in 2012 (can the eventual Democratic nominee replicate the message and mobilizing muscle of the 2012 Obama campaign?) it is unlikely to look so similar to the Republican leaning 2014 mi-term electorate."

    1. That's just a nicer way of saying that Quinnipiac is most likely wrong! :)