About Those PPP Polls

First let me say this about PPP polls:

1. They should not be summarily dismissed with a wave of the hand
2. They should not be relied on as heavily as some people do.

Two polls that have made some noise are post convention polls from Ohio and North Carolina.

Doug Mataconis of Outside The Beltway, a libertarian who spends far more time whining about and bashing Republicans than he ever does Democrats, of course, highlights these two polls and joins in the current narrative that has formed that basically says Romney’s goose is cooked. About Ohio:

Republicans tend to dismiss PPP as a “Democratic” polling firm, however it has tended to be fairly accurate in recent years, more so that Rasmussen has in any case. In this specific poll, it’s worth noting that the D/R/I breakdown in the poll seems to be fairly reasonable, with only a slight +4 Democratic advantage in that area, which is actually less than the Democratic advantage in the actual 2008 Exit Polls. So, it strikes me as hard for the poll to be dismissed as mere “bias.” Of course, that doesn’t mean that the poll itself is an accurate measure of the race, we’ll have to wait for other in-state polling to see if that’s the case.

What the hell is he saying here? He earlier said Romney should be “concerned.” Then he says PPP is more accurate than Rasmussen without providing any evidence to suggest that. Then he says he doesn’t even know if the poll is accurate. Well then why write about it then?

Anyway, here are his comments about North Carolina:

The fact that state is so close despite all of this suggests several things to me. First, the previous assumptions that North Carolina would naturally drift back in to the GOP camp appear to be incorrect, and it’s rather irrelevant at this point whether that’s for demographic reasons, or because of something to do with Governor Romney. Second, the fact that the President is keeping the race close here despite not really campaigning there recently suggests that he has indeed held on to many of the independent voters who put him over the top there four years ago.

This is ridiculous and I’ll explain why shortly.

This is not so much about Doug as it is the narrative that has quickly formed following the Democratic National Convention. Obama has gotten a bounce, Romney is in deep doo-doo and everybody should start packing it in. If you want to look at this race from the point of view of a pro, I’d suggest you take a look at what Rick Wilson had to say about it.

With regard to PPP, their track record lately has not been that good, particularly as it relates to state based polling. Remember, PPP does automated telephone polling instead of traditional polling where there are two humans talking. Again, this does not mean their polls should be dismissed outright but it does invite valid skepticism.

The first example would be the Texas primary. Just days prior to the vote (not the runoff), PPP showed David Dewhurst with a 17 point lead over Ted Cruz and was suggesting Dewhurst could get the 50% needed to avoid a runoff with Cruz:

David Dewhurst leads Ted Cruz 46-29 in the Texas Republican Senate primary and the main question on Tuesday will be whether he can sneak over 50% and avoid a runoff.

They even went on to suggest that if there was a runoff, Dewhurst would win easily:

Even if Dewhurst does get stopped short of 50%, his prospects in a runoff election look pretty promising. He would lead Cruz 59-34 in a head to head, mostly because Leppert voters prefer Dewhurst over Cruz by a 77-13 margin.

It turns out, Dewhurst only won by 10 points, forcing that runoff and Cruz won quite easily.

So PPP was off.

The Wisconsin recall election is another example wherein PPP was not in line with other polls. Their final poll for the recall election between Governor Scott Walker and Mayor Tom Barrett showed Walker squeaking by with a lead of 50% to 47%.

The actual result (which closely represented other polls) was 53% to 46%.

Again PPP was off.

Nate Silver, who writes for the NY Times about polling has suggested PPP has a built in bias towards Democrats of anywhere from 3-4 percent, which is pretty big.

If you take into account the bias of the PPP polls, then the race in Ohio is just as close as it has been for some time, with the gap narrowing in Romney’s favor as we get into the home stretch.

Ladies and gentlemen, the important thing to keep in mind are the trends, not any one or two polls during a 7-10 day stretch. Please note that both of the graphs I am going to show here include these latest PPP polls. Here is the chart for North Carolina:

Here is the chart for Ohio:

So please, dispense with the pearl clutching and fainting couch acts, ok? There is zero reason to panic or to be “concerned” at this point, contrary to what the narrative is at this moment.