Not over.
I was just
mentioning how hard the GOP's path to the Senate majority was, and Karl Rove's Crossroads decided to support my thesis by polling a bunch of the key Senate races. Politico's Morning Score email newsletter:
Arkansas: Democratic Sen. Mark Pryor 39, GOP Rep. Tom Cotton 39, not sure 22. Pryor approval rating: 38 percent approve, 44 percent disapprove; Cotton favorability: 31 percent favorable, 39 percent unfavorable. (Margin of error: ± 4.3 percentage points.)
Colorado: Democratic Sen. Mark Udall 45, GOP Rep. Cory Gardner 43, not sure 12. Udall approval rating: 38 percent approve, 46 percent disapprove; Gardner favorability: 30 percent favorable, 38 percent unfavorable. (Margin of error: ± 4.4 percentage points.)
Louisiana: In Nov. 4 primary: Democratic Sen. Mary Landrieu 40, GOP Rep. Bill Cassidy 35, Republican Rob Maness 4, Republican Paul Hollis 3, not sure 18 percent. In hypothetical runoff: Landrieu 43, Cassidy 47, not sure 10. Landrieu approval rating: 39 percent approve, 51 percent disapprove. (Margin of error: ± 4.2 percentage points.)
Michigan: Republican former secretary of state Terri Lynn Land 43, Democratic Rep. Gary Peters 40, not sure 18. Land favorability: 32 percent favorable, 32 percent unfavorable; Peters favorability: 25 percent favorable, 35 percent unfavorable. (Margin of error: ± 4.2 percentage points.)
Montana: Democratic Sen. John Walsh 35, Republican Rep. Steve Daines 42, not sure 23. Democratic former Lt. Gov. John Bohlinger 33, Daines 44, not sure 23. Walsh favorability: 33 percent favorable, 22 percent unfavorable; Daines favorability: 43 percent favorable, 31 percent unfavorable; Bohlinger favorability: 27 percent favorable, 23 percent unfavorable. (Margin of error: ± 4.3 percentage points.)
The polls were conducted by Harper Research, a new GOP robo firm designed to
counter PPP—so flooding the zone with polling, but without PPP's proven track record. Also note that these polls do not include cell phones, a methodological flaw that can no longer be ignored. Even PPP, the prototypical robo-poller, is now supplementing its samples with
cell phones internet surveying.
Still, it does give us some numbers to work with, and I take a closer look below the fold, including some (mostly positive) trend lines.
Right off the bat, note that Georgia and Kentucky are missing. No doubt that Crossroads polled those states and didn't like the numbers enough to release them. Or it's studiously avoiding them to avoid stepping on their narrative. But even these numbers, which are supposed to show Republicans "within striking distance" of the majority should deflate GOP hopes.
You see, at this point, the narrative is that Republicans have this thing in the bag. So why release polls showing that it's not? That Arkansas poll is now the fifth in a row without a Republican lead. And while 39 percent for an incumbent would be a guaranteed kiss of death, it's only that low because Crossroads seemed to go out of its way to not press leaners. They are absurdly high in this polling. So let's look at trend lines from the three states the organization also released in February previous polling:
Arkansas (1/20)
Pryor 39 (36)
Cotton 39 (42)
Colorado (3/9)
Udall 45 (45)
Gardner 43 (44)
Louisiana (1/20)
Landrieu 43 (44)
Cassidy 47 (45)
Michigan (1/20)
Peters 40 (37)
Land 43 (42)
Montana (1/22)
Walsh 35 (29)
Daines 42 (43)
Pryor eliminated his six-point gap. Udall held steady. Landrieu lost ground, but is still within striking range. Peters closed the gap. Walsh cut his deficit in half.
For a party coasting to November victory, those aren't the kind of numbers that should inspire confidence. Not this early in the game, and not after tens of millions of dollars of Koch dollars spent attacking these Democrats.
Udall in Colorado has been weathering his own Koch assault, and leads despite facing a top-tier Republican recruit fresh off his "entering the race" media honeymoon. Polls have undercounted Democratic performance in this state the last several cycles, so he might even be doing better than this.
That leaves Michigan, a state that is certainly too close for comfort, but it's early, in a state where its partisan lean should assert itself. If the race remains this tight late in the cycle, we might need to start sweating it, but given the low name recognition of both candidates, it's clear that the race hasn't truly begun yet. And note, also, that Peters has already faced over $2 million in Koch attack ads without much of a Democratic response.
So yeah, if inevitability is the goal, these polls come up short. On the other hand, perhaps that is the point. If conservative donors are complacent because they've bought the media "inevitability" hype, that would affect Crossroads' bottom line. Showing those donors that the race isn't over would be a good way to scare up more donations.
Either way, the bottom line is clear: Republicans are cheering their Senate majority way too early. There is still a great deal of race to run, Democratic incumbents are still in the game, Republicans have their own turf to defend, the teabaggers are still causing havoc, and it's only a matter of time before we get this cycle's Todd Akin and Richard Murdock.
There will be no greater feeling than seeing Democrats retain our majority on Election Night this November. It'll be Mitt Romney, the Sequel. And it will be delicious.
Update: There were older Harper polls for Colorado and Michigan, but for a different client. I added them to the trendlines above.