Wednesday, June 29, 2016

New Quinnipiac poll - Dead Heat

Democrat Hillary Clinton has 42 percent to Republican Donald Trump's 40 percent - too close to call - as American voters say neither candidate would be a good president and that the campaign has increased hatred and prejudice in the nation, according to a Quinnipiac University National poll released today.
This compares to results of a June 1 national poll by the independent Quinnipiac (KWIN- uh-pe-ack) University, showing Clinton edging Trump 45 - 41 percent.
This is now the second poll in the last week to come out showing the race tightening. The recent NBC/WSJ poll showed Clinton leading Trump 39-38 percent with the Libertarian and Green Party candidates included in the polling question.

Contrary to popular believe among your mainstream media folks, there is no more real evidence that Clinton is expanding her lead (Ipsos/Reuters and ABC/Washing Post) as there is evidence that the race is tightening (Quinnipiac and NBC/WSJ). In fact, of the four polls, Quinnipiac is clearly the most respected of the pollsters, as well as being the only independent (non-media commissioned) poll.


Now there has been some competing suggestions from the right regarding the polling out there. Newt Gingrich has all but accused the media of making up polling results to show Hillary with a lead, stating that if the polls show the race withing five points, that Trump will likely cruise to victory. Whereas talk radio icon Rush Limbaugh has suggested that Hillary really is up in the polls, and that people should stop writing them off and start dealing with it.

The truth is somewhere in the middle.

There is definitely some serious demographic problems with the polls putting Clinton   up big. Approximate half of the Ipsos/Reuter polling sample is made up of Democrats, which pushes her lead up to nine percent for registered voters. When they filter for "likely voters" (which traditionally benefits Republicans) the lead increases by four points? Both assumptions not only defy any empirical historical evidence, but appear to be totally without any statistical explanation. This was also the pollster that managed to show Clinton expanding her overall lead, while losing a fairly significant amount of support in every demographic breakdown. This can only be done by further manipulating the demographic breakdown to include "more" Democrats as an overall percentage of the sample. When taken as a whole, this strongly suggests that Gingrich may have a point regarding pollsters manipulating the numbers in order to champion the story of a big Clinton lead..

That being said, when I plug the cross tabs into my own spreadsheet (using historically calculated demographic percentages) I show Clinton leading, but by only around four points. The overall numbers are not unlike what we saw in 2012.  Both Obama and Clinton garnering the Democrat vote overwhelmingly, while Romney and Trump take the Republican and Independent vote.

The one demographic where Trump has slipped behind Romney is in the GOP vote. There was an overwhelming consolidation of Republicans right after Kasich and Cruz dropped out. But since then, there has been a slow loss of support among Republicans. Whether that is permanent or the result of the recent surge in the Dump Trump rumblings is not clear. But in talking to some of the "on the fence' Republicans I know, it feels like they want to vote for Trump, and are simply looking for an excuse to justify it. Trump may actually benefit from becoming official which would silence the "Dump Trump" movement once and for all. He could also benefit from staying "on point" for more than one or two speeches.

Bottom line, folks is that we may not see consistent polling for some time, if ever on this race. It seems clear that there are some substantial differences of opinion as to what the make up of the 2016 electorate will look like. If those differences are not reconciled, then we may be seeing a very wide range of polling results throughout this entire campaign.

(More on that later)

7 comments:

Honest, decent, truthful Rev. said...

https://ballotpedia.org/Ballotpedia%27s_battleground_poll,_2016

Commonsense said...

James the samples in those polls are too small and MOE is too high to make the results meaningful in any way.

And yes, I would same the same if Trump had the big lead.

C.H. Truth said...

There is a reason why neither Real Clear Politics or Pollster.com includes polls from ballotpedia.org.

Myballs said...

I think there are many voters hesitant to publicly say they will vote for trump. We've seen real violence against those who have said this. I myself had a Obama loving relative tell me that anyone who votes for Trump is a racist!

My point is, there is an unknown population that us not showing up in polls for the reasons I said.

Honest, decent, truthful Rev. said...

Balls, why shouldn't anyone call someone who votes for a racist a racist?

Honest, decent, truthful Rev. said...

CH SAYS:There is a reason why neither Real Clear Politics or Pollster.com includes polls from ballotpedia.org.
_________

Is there a reason CNN finds their polling aggregates significant?

http://www.cnn.com/2016/06/29/politics/battleground-polls-donald-trump-hillary-clinton/

Honest, decent, truthful Rev. said...

Delete the word aggregates.