President Obama opened a wide lead over Mitt Romney in the last few weeks, but three polls show Romney now beating Obama in the 2012 election. But if you write about that on Facebook, it could cost you friends. Here's our guide to today's polls and why they matter.
Pollster: ABC News/ The Washington Post
Methodology: Telephone interviews with 1,003 adults, some of them cell-phone-only users, nationwide from March 7 to March 10.
Why it matters: Romney was briefly beating Obama last fall, but Obama has polled ahead of him in hypothetical match ups polled during January and February, suggesting that the long primary had hurt Romney. His poll numbers among independents show he's seen more unfavorably than favorably. But in recent days either voters are starting to like him better or they're starting to like Obama worse. Rasmussen and Gallup both show Romney beating Obama, but Rasmussen leans right and Gallup has tended to understate Obama's support this year compared to other polls. But Gallup finds Romney winning 50 percent to Obama's 46 percent -- still a bigger gap than ABC, but a similar result.
Caveat: Polls showing Romney beating Obama six months out from the election are nearly as useless as polls showing Obama beating Romney seven months out from the election.
Pollster: Public Policy Polling
Methodology: Robo-calls to 600 likely Republican voters in Alabama March 10 and March 11; robo-calls to 656 likely Republican voters in Mississippi on the same dates.
Why it matters: These states are "would be really good to win" for Romney, given his lead in delegates; "probably needs to win" for Santorum, who wants to be the one true conservative alternative to Romney; and "really definitely needs to win" for Gingrich, who's turning himself into the southern candidate.
Caveat: There haven't been that many polls in these states, so it's hard to say how reliable these are. "It appears Romney’s over-performance in both public and private polls in Tennessee convinced the campaign that it had a shot in the Volunteer State only to lose it by a bigger margin than any late poll had predicted. Just something to keep in mind," NBC News' First Read writes. That could happen in Alabama too, which would mean Romney wasted energy he should have directed toward Illinois.
Pollster: ABC News/ The Washington Post; CBS News/ The New York Times
Methodology: Telephone interviews with 1,003 adults, some of them cell-phone-only users, nationwide from March 7 to March 10; telephone interviews with 1,009 adults nationwide from March 7 to March 11.
Why it matters: Americans are resigned to accepting two guys they're not completely thrilled about.
Caveat: Expectations can change easily. You never know who will get caught kicking a puppy on YouTube.
Pollster: Pew Internet & American Life Project
Methodology: Telephone interviews with 2,253 adults from January 20 to February 19.
Why it matters: If you post this story on your Facebook wall, your high school friend might unfriend you. (Go ahead and share it, you never liked her anyway.)
Caveat: Okay, old folks: let's not confuse social media "friends" with IRL friends. As The Hill points out, blockers usually unfriend acquaintances, not their bffs, the study finds.