Thursday brought with it a veritable poll-a-palooza of November trial heats, with NBC/Marist dropping a six-pack of new polls at the presidential and Senate levels, and acting as a counterweight to all the breathless coverage of yesterday's Quinnipiac poll (I don't know that I've seen a single state poll get so much coverage this early in a campaign).
Today, however, let's focus on Wisconsin, whose recall election is coming up in less than two weeks. Two new polls come out of the Badger State today. On the surface, they are decidedly more pessimistic for the Democrats than a couple of other polls (the GQR poll and the St. Norbert College/WPRI) released earlier in the week. That said, there are still data points within that point to the possibility of this race tightening between now and June 5th.
But, first, as always, the numbers. And there are quite a lot of them to consume today:
PRESIDENTIAL GENERAL ELECTION TRIAL HEATS:
NATIONAL (Gallup Tracking): Obama d. Romney (47-46)
NATIONAL (Rasmussen Tracking): Obama d. Romney (45-44)
NATIONAL (YouGov): Obama d. Romney (46-42)
CALIFORNIA (PPIC): Obama d. Romney (50-39)
MASSACHUSETTS (Suffolk University): Obama d. Romney (59-34)
NORTH CAROLINA (National Research for the right-leaning Civitas Institute): Romney d. Obama (47-45)
OHIO (NBC News/Marist): Obama d. Romney (48-42)
VIRGINIA (NBC News/Marist): Obama d. Romney (48-44)
WISCONSIN (Reason-Rupe): Obama d. Romney and Gary Johnson (46-36-6)
DOWNBALLOT POLLING:
CA-PROP 29-TOBACCO TAX (PPIC): Favor 53, Oppose 42
FL-SEN (NBC News/Marist): Sen. Bill Nelson (D) 46, Connie Mack (R) 42
FL-SEN (Quinnipiac): Connie Mack (R) 42, Sen. Bill Nelson (D) 41; Nelson 43, George LeMieux (R) 36
FL-SEN--R (Quinnipiac): Connie Mack 40, Mike McCalister 8, George LeMieux 7
HI-SEN (PPP for the League of Conservation Voters): Mazie Hirono (D) 50, Linda Lingle (R) 41
MD-MARRIAGE EQUALITY LAW (PPP): Favor 57, Oppose 37
MA-SEN (Suffolk University): Sen. Scott Brown (R) 48, Elizabeth Warren (D) 47
OH-SEN (NBC News/Marist): Sen. Sherrod Brown (D) 51, Josh Mandel (R) 37
PA-SEN (Rasmussen): Sen. Bob Casey (D) 48, Tom Smith (R) 41
TX-SEN--R (PPP): David Dewhurst 46, Ted Cruz 29, Tom Leppert 15, Craig James 3
VA-SEN (NBC News/Marist): Tim Kaine (D) 49, George Allen (R) 43
WI-GOV (Garin-Hart-Yang for Barrett): Gov. Scott Walker (R) 50, Tom Barrett (D) 48
WI-GOV (Reason-Rupe): Gov. Scott Walker (R) 50, Tom Barrett (D) 42
WI-GOV (We Ask America): Gov. Scott Walker (R) 54, Tom Barrett (D) 42
A few thoughts, as always, await you just past the jump ...
First on the agenda: Wisconsin. Where, as it happens, we get a trio of results today with a stunning 10-point spread between them. On one end of the spectrum, we have the We Ask America poll, giving Scott Walker a double-digit edge. Besides the genesis of that particular polling outfit (for an explanation, David Nir touched on it this morning), there is a demographic data point that simply defies common sense. Indeed, it is so out there that W.A.A. addresses it in their own writeup of the race. They claim that Tom Barrett leads public union households by only a 50-47 margin. That would be somewhat akin to chickens being on the fence about Colonel Sanders, as it were. I am not saying that renders the entire poll suspect, but it definitely raises a pretty substantial caution flag.
On the other end of the spectrum is a Garin-Hart-Yang poll for Democrat Tom Barrett showing a two-point race. Besides the normal caveats that apply with a campaign releasing its own data, a poll with only two percent undecided seems awfully low. However, in this case, it's a definite possibility. More than one pollster, and more than one analyst, has speculated that there are few "persuadables" out there in this particular recall, and that the battle now solely revolves around the ability to energize base voters and get them to show up a week from Tuesday.
Splitting the difference is a new poll by Reason-Rupe. No one would suggest that Reason doesn't have a rooting interest here, but one does have to give them credit on a few fronts: Like Daily Kos, they chose to contract with a pollster with a solid reputation (ORC also does polling for CNN), and they put all of the issue-based questions that could spoil a sample after the horserace questions, as well.
And if you dig into those numbers, you get the same turnout-based message.
While Barack Obama only leads Mitt Romney rather narrowly (four points) with the likely voter screen in place, he leads by 10 points among the larger universe of respondents. A similar dynamic kicks in for the gubernatorial election: What was an eight-point lead for Walker with the likely voter screen in place drops to four, with a considerably higher number of undecideds.
Even though their results were wildly different, there is a common thread to be found in all three: Victory or defeat is not based on a quivering mass of indecisive "persuadables," as may often be the case in elections. This one is all about composition of the electorate. If, as Reason-Rupe and virtually every other pollster suggests, the electorate skews GOP, Scott Walker will not be recalled. Tom Barrett actually does pretty well with indies, but not enough to overcome what looks like a GOP turnout edge. If that edge narrows, or disappears entirely, a legitimate upset could be in the offing.
In other polling news ...
- Those NBC/Marist polls are getting a ton of attention today, and they paint a pretty decent picture not just for Barack Obama, but Democrats in general. That Ohio Senate poll caught my attention the most. GOP nominee Josh Mandel, the state treasurer, has taken some serious incoming as of late. One has to wonder if his crappy media week is starting to take a toll on his numbers. It has been a while since Democratic incumbent Sherrod Brown has run this far ahead of the president in Ohio.
- A word of caution on that California poll, showing Mitt Romney down just 11 points to the president in the Golden State. That is based on a "likely voter" screen for the primary election, which is coming up next month. If you apply the partisan ID of the sample (D 44, R 31, I 25) to the presidential breakdown by party in the poll, you get a pretty substantial shift (53-36 Obama). The big news out of that PPIC poll, however, was the meteoric shift in support for California's Proposition 29, a $1 per-pack tobacco tax increase. Once up 67-30, a gigantic media campaign funded by big tobacco (to the tune of nearly $40 million) has brought the vote much closer to parity. Definitely one to watch on election night in two weeks.
- When the DSCC dropped a poll a couple of days ago showing the Massachusetts Senate race tied, it was hard not to be a little worried. After all, it is easy to think, "Wow, if the Democrats' own polling has it tied, what must the real state of the race be?" Well ... it is only one poll, but Suffolk (which had Brown up by a respectable margin in the high single digits last time around) has it as a one-point race. Their result essentially confirms the DSCC's numbers. It also confirms that being the former governor is paying Mitt Romney precisely zero dividends in Massachusetts—the Suffolk poll has him getting smoked there by Barack Obama.
- Finally, a word to my friends in punditocracy: When you say over and over that Mitt Romney is a lock in North Carolina, and yet even a right-wing pollster, for a right-leaning organization, have it as a two-point race, you start looking more than a little ridiculous. Could Mitt Romney reclaim the state's 15 electoral votes for the GOP? Sure. But to call that state anything other than a tossup is simply absurd. Outside of one Rasmussen poll, it is pretty tough to find a poll outside of a very narrow band within a point or two either way. Which, for what it's worth, seems to me to be the textbook definition of a tossup.