The new week brings little data, but does bring a couple of things worth talking about as we lumber through the dog days of summer. The national tracking polls yield few clues about what the hell is going on in the presidential race, but the DCCC gets a few points for proper technique in the always intriguing internal polling game.
On to the numbers:
PRESIDENTIAL GENERAL ELECTION TRIAL HEATS:
NATIONAL (Gallup Tracking): Obama d. Romney (46-45)
NATIONAL (Rasmussen Tracking): Romney 46, Obama 43
MINNESOTA (SurveyUSA for KSTP): Obama d. Romney (46-40)
DOWNBALLOT POLLING:
CA-47 (DCCC Internal Poll): Alan Lowenthal (D) 47, Gary DeLong (R) 36
MN-SEN (SurveyUSA for KSTP): Sen. Amy Klobuchar (D) 55, Kurt Bills (R) 31
PA-SEN (Rasmussen): Sen. Bob Casey (D) 49, Tom Smith (R) 38
A few thoughts, as always, await you just past the jump ...
The two lead stories on the Wrap tonight play on recurring themes—trying to draw conclusions about the two daily presidential tracking polls (which is uncharacteristically difficult to do), and trying to draw conclusions about internal polling (which is substantially easier).
It's not just the fact that Mitt Romney has enjoyed a mini-surge in both tracking polls with no discernible reason for the movement. It is the fact that the slight fade in Obama's numbers do not appear to have much of a relationship with the president's job approval numbers, even according to the same pollster!
Consider: Gallup's job approval for President Obama released one week ago was a slight net positive (47/45). Today, the president's job approval numbers were underwater, and not by a point or two (44/49). The represents a net drop of seven points in one week.
In that same time span, though, the presidential head-to-heads have scarcely moved: Obama had a 47-45 lead over Mitt Romney a week ago, while today that lead is ... 46-45. For the math-impaired, that is a one point net drop in the same week.
Rasmussen, meanwhile, inexplicably went the other direction. Barack Obama has never had great approval numbers, according to the House of Ras, but today marked one of his better performances (49/50) this month. That was actually a few points better than Obama performed at this point last week (47/52). Again, though, the numbers barely moved in the head-to-head. However, for Obama, he is actually in (statistically insignificant) worse shape than he was last week, with today's poll matching Romney's biggest lead for the month (Romney +3).
Gallup's could be explained plausibly by one key difference in their data: Their approval numbers are off a three-day sample, while their horserace numbers are off a seven-day sample. So, if the president's horserace numbers continue to erode in Gallup's tracker this week, we might know why. And, of course, with shifts this minimal, it could all be float. But you would still expect things like presidential job approval and head-to-head performance for the incumbent to roll on parallel tracks. This week, thus far? Not so much.
Meanwhile, kudos (of a sort) to the DCCC for playing smart defense for one of their candidates. Last week, as we noted in Friday's Wrap, the campaign of GOP contender Gary DeLong (CA-47) fired off a poll showing him within three points of Democrat Alan Lowenthal, generally considered to be the betting favorite in the newly constructed district that straddles the Los Angeles and Orange County borders.
So, the D-Trip (who apparently, according to Kyle Trygstad's article, has their own polling shop. Who knew?) countered with a poll of their own, giving Lowenthal a more comfortable 11-point edge. Always good to try to smother a narrative before it takes hold, and good on them for that.
In other polling news ...
- The new Rasmussen Senate poll in the Keystone State might weaken the cred for that presidential poll last week, which had Mitt Romney within 4 points of the president. Simply put, it is pretty tough to find another pollster in the game (and multiple pollsters have tested this race) that has had Republican Tom Smith this close to Sen. Bob Casey. Could the race end up this close? Sure. Smith is still comparably little-known, and Casey's approval numbers in all polls to date have been kind of "meh." But is it 11 points now? I am skeptical.
- Meanwhile, another case study in not-very-good poll reporting comes from KSTP in Minnesota, which contracted with SurveyUSA to poll the state's presidential and senatorial contests. The KSTP write-up of their own poll notes that the lead for Barack Obama has been "cut in half" from the previous SUSA poll in the state. What the write-up fails to note is that SUSA added a likely voter screen in the interim, which study after study (including a recent piece by Mark Blumenthal confirms will goose support for Republicans. Rookie move by the client for SUSA to not clarify that this really is not an apples-to-apples comparison.
- Is Barack Obama's approval rating a harbinger of doom, as many pundits have long opined? Maybe, but maybe not. Gallup dug into their own archives, and found that Obama's approval rating from April 1 through June 30 was quite a bit worse than most presidents who were successfully reelected. That's the bad news for Obama supporters. The good news? He also was ahead of the pace of all three presidents of recent vintage who failed to be reelected (Ford, Bush 41 and Carter). Obama's 14th quarter net approval numbers (46.8) was almost exactly at the midpoint between that of Gerald Ford (46.0) and George W. Bush (47.9). That is notable, of course, because Bush narrowly won reelection, while Ford narrowly lost. Yet another data point that screams "coin flip" about this upcoming battle for the White House.