Public Policy Polling (pdf) (3/5-8, likely voters, no trend lines):
John Hickenlooper (D): 50
Scott McInnis (R): 39
Undecided: 11
(MoE: ±4.1%)
Well, these numbers are a little different from Rasmussen‘s, aren’t they? PPP’s first look at the Colorado governor’s race since the tradeoff from retiring Democratic incumbent Bill Ritter (currently saddled with a 38/50 approval) to Denver mayor John Hickenlooper (51/27 favorable) finds that the upgrade looks poised to pay dividends. Hickenlooper has an 11-point lead on Republican ex-Rep. Scott McInnis, whose favorables are 28/27.
PPP’s sample composition is 39 D/36 R/24 I, which is a little optimistic compared with the actual registration numbers of 34 D/35 R/31 unaffiliated (or 33 D/33 R/34 UAF if including inactive registrations). Still, the same sample broke 49/49 for Obama and McCain, so it’s still not as Dem-favorable a sample as the actual voters who showed up in 2008.
RaceTracker Wiki: CO-Gov
Excellent news, hopefully he helps Bennet.
but nothing but noise in the middle.
I know it’s impossible to predict the actual make up of voters in Nov., but I’m not sure if trust PPP or Rass to get it close this far out.
I thought Hickenlooper was in the Gov race.
Republican – 35%
Independent – 35%
Democrat – 30%
Hickenlooper – 10/53/95 = 52%
McInnis – 90/47/5 = 48%
Toss-up.
They’ve been doing great this cycle. Absolutely knocked MA-Sen out of the park, got NJ-Gov and ME-Init within the MoE, and did pretty well in Virginia. Obviously their biggest failure was in NY-23, but that’s not a bad reflection on their polling so much as an error in judgment to continue with the poll despite the funky circumstances of the race.
Nice to see that Rasmussen isn’t the only one polling this race.
…that tweaking the turnout model in a poll to rebalance on age, or party ID, or whatnot, doesn’t change the toplines very much once you sit down and actually do the math.
And they’re right, because I discovered the same thing in my own math when doing that tweaking of polls for fun the past couple cycles.
So even if the PPP turnout model is more Hickenlooper-friendly than November’s reality, an 11-point edge wouldn’t be erased and instead would decline only a few points to stay in the high single-digits.
but Markos points out several governor’s races where Rasmussen’s polling is not even in the same ballpark as other pollsters who’ve polled the race. And that’s not just PPP, but also Quinnipiac (OH-Gov) and Research 2000 (IL-Gov). We’re talking MORE than a 10-point swing in each case.
The other way. But thank goodness for PPP upping their output this cycle else Rasmussen would be getting away with his nonsense to an even greater degree.