Republicans need to win six Senate seats to take control of the upper chamber, and most scenarios for victory include the Southern seats up for grabs. A poll out today from the New York Times and the Kaiser Family Foundation suggests that may be tougher than first thought. Mark Pryor, considered to be one of the most vulnerable incumbents in the midterms, has a ten-point lead over his Republican challenger, Rep. Tom Cotton:

The survey underscores a favorable political environment over all for Republicans in Kentucky, North Carolina, Louisiana and Arkansas — states President Obama lost in 2012 and where his disapproval rating runs as high as 60 percent. But it also shows how circumstances in each state are keeping them in play for the Democrats a little more than six months before the midterm elections.

Senator Mark Pryor of Arkansas, a two-term incumbent who has been considered perhaps the most imperiled Democratic senator in the country, holds a 10-point lead over his Republican opponent, Representative Tom Cotton. Mr. Pryor, the son of a former senator, has an approval rating of 47 percent, with 38 percent of Arkansas voters disapproving of him.

Senator Kay Hagan, Democrat of North Carolina, appears more endangered as she seeks a second term. She has the support of 42 percent of voters, and Thom Tillis, the Republican state House speaker and front-runner for his party’s nomination, is at 40 percent. Unlike Mr. Pryor, however, Ms. Hagan’s approval rating, 44 percent, is the same as her disapproval number. In Kentucky, Mitch McConnell, the Senate Republican leader, is also effectively tied with his Democratic rival, Alison Lundergan Grimes, a race that may be close because Mr. McConnell, first elected to the Senate in 1984, has the approval of only 40 percent of voters, while 52 percent disapprove. But Ms. Grimes must overcome Mr. Obama’s deep unpopularity in the state, where only 32 percent of voters approve of his performance.

With 42 percent support, Senator Mary Landrieu, Democrat of Louisiana, has an early lead in a race that is not fully formed against a large field of Republicans. Representative Bill Cassidy, the Republican front-runner, was the choice of 18 percent, and 20 percent had no opinion. There are two other Republicans in the race, but Louisiana has no primary. So all candidates of both parties will be on the ballot in November and, absent one of them taking 50 percent, there will be a runoff in December.

The outcome in Arkansas, at least, seems a little odd in contrast to other numbers inside and outside of this poll. For instance, Barack Obama’s approval rating in Arkansas within the NYT/KFF poll is 32/59, and even when his national numbers were much better in 2012, Obama lost Arkansas 37/60 to Mitt Romney. No Democratic candidate for the House even got to 40% in that election.

Furthermore, a quick check of the RCP poll average for Arkansas shows this as an outlier. That’s more true of Cotton’s standing, though, than Pryor’s. His support has been consistent at 46%, although not a great number for an incumbent, especially one with Pryor’s prominent family name. Cotton, though, also polled in the mid-40s even up to earlier this month, when he got 43% in the Talk Business poll, and going all the way back to last summer. Why would he drop to the mid-30s at this point?

I’m inclined to chalk this up as an outlier, but it still should serve as a warning to the GOP, too. It won’t be a cakewalk to beat Pryor, or for that matter to win any of these races. Republicans can run against Obama, whose approval numbers are atrocious across the board, but they’d better be clear as to what they’re for as well. Names like Pryor and Landrieu still carry weight in states Republicans should win this year.

Update: Dr. Pradheep Shanker makes an interesting catch:

Usually, it’s presidential cycles that bring out significant numbers of new voters.

Update: Bill Kristol finds another reason to distrust the numbers:

In other words, the Times and Kaiser have produced a sample in Arkansas that reports they voted in 2012 for Romney over Obama–by one point. But Romney carried Arkansas in 2012 by 24 points. Similarly, the Kentucky sample is +3 Romney when reality was +23. The Louisiana sample is +3 Obama in a state Obama lost by 17, and the North Carolina sample is +7 Obama in a state he lost by 3.

The whole point of question 12 is to provide a reality test for the sample. That’s why they ask that question–we know what happened in 2012, so the only thing to be learned by asking the 2012 question of the sample is to ensure that it’s a reasonably accurate snapshot of voters in the state. Of course there’ll always be some variance between reality and the sample’s report of its vote a year and a half ago–but not a 23 point variance.

Yikes. Interestingly, though, the difference didn’t actually boost Pryor, as I already noted above.