Democrats are surely focusing on the headline numbers in the latest Washington Post/ABCNews poll, showing the president 8 points ahead of Mitt Romney, his likely Republican opponent this fall and with approval at 50%.
Now, to be sure, those numbers do look pretty good for the incumbent, but looking more deeply into the poll, particularly at its internals, a number of bloggers have teased out details which should undermine the recent overconfidence of the Obama campaign. Over at the Washington Examiner, Charlie Spiering has found three numbers which should cause heartburn at the White House, particularly given numbers which his colleague Conn Carroll and HotAIr’s Ed Morrissey highlight. Spiering notes that Romney leads Obama on the economy, deficit and energy by 47-43, 51-38 and 47-42 respectively.
Putting the Post poll in context, Carroll reports that it oversamples Democrats, “34 percent of those polled identified themselves as Democrats, 23 percent identified themselves as Republicans, and 34 percent identified themselves as Independents.” He goes on to compare that number to the “the turnout realities of the past four elections” and finds that even in the banner Democratic year of 2008, Democrats only enjoyed a 7-point advantage.
Image what Romney’s lead on those key issues which would like in a less skewed sample.
Surveyed the skewed numbers and comparing it to the Post’s previous polls where Democrats only enjoyed a 4-point advantage (as opposed to 11 in the current survey), Ed Morrissey observes:
. . . one should be seeing huge leads for Obama in the head-to-head matchups. Instead, Obama lead Romney by only eight among general-population adults, 51/43, barely getting into majority territory, and Santorum by ten, 51/41. Among registered voters, Obama leads Romney by seven, 51/44 — in both cases, smaller than the artificial sample advantage of the poll. In fact, adding seven more points to the Democratic advantage only resulted in Obama’s support growing by five points since March, and Romney’s dropped only four in that period. Once again, the numbers are almost exactly like February’s poll, with its matching D+11 sample.
But hey — the Post got to write its headline, right?
And if we adjusted that number to reflect the past less decidedly Democratic sample, the president would have fallen back a bit and Romney edged up a notch.