NBC/WSJ poll: Obama leads Romney by four nationally, by three in battleground states

Ah, the NBC/WSJ poll, the dreariest blog post of the month for the big A. Thus far this year they’ve measured registered voters, not likelies (although that’ll change soon), which means the data’s been iffy to begin with. Worse, the partisan split in the sample has bounced around enough each month to make you wonder how useful the trendlines are over time. Remember the sample they dropped on us in July? That was a big ol’ D+11, which meant NBC’s sample imagined an electorate considerably more Democratic than the electorate that handed Obama a landslide in 2008. (The exit poll on election night was D+7.) This month’s sample is D+6, which is more realistic and yet still imagines a bluer electorate than we’re likely to see, I think. The point is, when you’re looking for trends by comparing July data to August in the crosstabs, you need to remember that July was based on a strong Democratic tilt and August a more mild one. Case in point: Although O’s lead over Romney technically declined from six points last month to four points now, that’s probably more of a sample artifact than any real movement. In fact, because a lot of the numbers this month are similar to what they were in July, this poll is good news overall for The One insofar as it means his lead is holding up even with a less forgiving sample.


In a smaller sample of voters living in 12 key battleground states – Colorado, Florida, Iowa, Michigan, Nevada, New Hampshire, New Mexico, North Carolina, Ohio, Pennsylvania, Virginia and Wisconsin – Obama leads Romney by three points, 49 percent to 46 percent.

That’s a narrower edge in these battlegrounds than the eight-point lead the president enjoyed in the June and July NBC/WSJ polls.

Looking inside the numbers, Obama continues to lead Romney among key parts of his political base, including African Americans (94 percent to 0 percent), Latinos (by a 2-to-1 margin), voters under 35-years-old (52 percent to 41 percent) and women (51 percent to 41 percent).

Romney is ahead with whites (53 percent to 40 percent), rural voters (47 percent to 38 percent) and seniors (49 percent to 41 percent).

Obama leads by 35 points when voters are asked which candidate is more likable; Romney’s net favorables are -6, which is lower than any other modern GOP nominee. The news isn’t all bad, though. Here’s a good trend that isn’t explained by the July sample. Obama is the first column, Romney the second:

If Romney does well in the debates, that remaining gap will close as more voters decide he’s presidential material. Meanwhile, Obama’s still lagging on the crucial “are you better off now than you were four years ago?” question:

As for the Ryan pick, how to explain his numbers being noticeably flat compared to other recent VPs?

Only Cheney’s numbers are more or less even; everyone else was greeted warmly by voters. The reason, I assume, is that Ryan’s been an ideological boogeyman to the “professional left” for several years now because of his budget. Palin and Edwards were essentially unknown when they were nominated, and Biden and Lieberman were reasonably well regarded senators who inspired no strong reactions when they were picked. It’s a testament to Ryan’s political influence, even at the age of 42, that his selection would inspire stronger feelings in voters than an old Republican power-broker like Cheney.

The numbers on Medicare “messaging,” meanwhile, are not encouraging:

The big catch to that result is that the pollster insisted on pegging each position to a candidate instead of offering the statements in isolation. Given Obama’s likability advantage, the fact that voters seem to trust him more than Romney may be leading them on this question to embrace his position partly because it’s his. The point is, would these numbers change if a more naturally likable pol like Ryan was tasked with making a sustained Medicare pitch? We’ll find out!