PUFFING UP DEMOCRAT VOTER SAMPLES

Headlines! PPP declares to all who will listen (not RCP, btw), Clinton Leads in Key Battlegrounds; Seen as Big Debate Winner. Oooh, big news? Trump stalwart Laura Ingraham, strangely, announces the PPP battleground polls just after they come out and for some reason says, "well, Trump isn't destroyed, anyway."

Okay, so what's all the fuss about? PPP shows Hillary leading in five key battlegrounds; Pennsylvania, Florida, North Carolina, Colorado and Virginia. Most strikingly, she's up by two points in Florida and North Carolina, two must win states for Trump. Well never fret Trump supporters, cause it just aint so.

It looks like the Colorado poll could be legit. The party split may actually favor the GOP in this poll, so that's a plus for the poll's legitimacy. What about the sample? The sample size may have been big enough, but we're not sure because the report on the sample is sketchy. Here it is, "Public Policy Polling interviewed likely voters on September 27th and 28th on behalf of VoteVets.org Action Fund. There were 694 respondents in Colorado, 826 in Florida, 861 in North Carolina, 886 in Pennsylvania, and 811 in Virginia."

So it looks like every respondent was a likely voter, what amazing luck! In reality, it was probably people who had been predetermined to be likely voters, or I guess it would have to be, or there's no possible way that everyone they randomly called is a likely voter. Or they may have small likely voter samples, which they didn't report.

PPP is using ridiculous voter turnout assumptions to pad the Democrat vote

So if PPP predetermined who it would talk to, they should have done a better job of getting a more representative partisan distribution of respondents. In the two states where Clinton ekes out a +2 lead, Florida and North Carolina, PPP assumes that Hillary not only mobilizes the African-American/millennial powerhouse turnout that propelled him to reelection in 2012, but adds about 500,000 additional Dems to that Obama turnout in Florida and 134,000 to the Obama coalition in North Carolina.

This . . . assumption . . . is . . . ridiculous.

It will not happen. Black women voted in higher numbers than any other demographic group in 2012. That will not happen in 2016. PPP's polls should show Trump up in Florida and North Carolina, but they don't, because they were partisan hacks on this one.

Now for Pennsylvania and Virginia. These two polls assumed Hillary would do about as well as Obama at getting out the Democratic vote. They figure that Hillary will draw about 110,000 less voters than Obama in PA and 80,000 less in VA. I just find this assumption hard to defend because Obama had organization and passion on his side. Hillary lacks the passion which is really necessary for good organization. Perhaps it will happen, but I do think it's unlikely.

PPP is assuming that 2016 will be a repeat of 2012 massive Democratic excitement, plus some more Democrats on top of that. There's no good reason to assume this. For that reason, the PPP battleground polls are not predictive. IGNORE

Follow my new account on Twitter for daily updates on the election!