What accounts for the movement? Strategists offer a lot of theories. Unenthusiastic Democrats are finally paying attention and deciding to vote. Ads marking a clear contrast between candidates are sinking in, and so is the president's message about this election being about a "choice" between two parties instead of an up-or-down vote on the Democratic agenda. There is no single, compelling GOP leader, and what people know about those leaders that do exist, they don't like. In a recent National Journal/ Pew Research Center poll, 60 percent said they disapproved of GOP leaders. (Fifty-three percent had the same view of Democrats.).The GOP message—we're not Obama—has penetrated all whom it will. In sum: The Republicans peaked too early.
Now that Congress has gone home, it can't help but improve the situation for Democrats. The public does not like Congress or its leaders. There's no more chance for them to bicker on television about stalled legislation and who is to blame for it.
But now for a reality check: How much does all of this good news get the Democrats? About 10 seats, says one longtime Democratic strategist involved in the races. Which is to say, they may lose 10 fewer seats than they expected. The math still looks bad for Democrats. The economy is still terrible, and people are still extremely glum about the country's future. That bad atmosphere, plus Democrats defending seats in historically Republican territory, plus Obama's low approval rating, means more than 20 or so seats currently held by Democrats are all but certain to go to Republicans. This tracks with the lowest possible prediction of GOP gains among political scientists who have made a study of making this guess. (The highest guess in the group of political scientists was a 51 seat gain by Republicans.)
After those certain victories, Republicans need 19 seats to take control of the House. If Democrats win a few GOP-held seats, maybe that number increases to 25. There are 40 to 50 Democrat-held seats, from which Republicans could take that number. The GOP doesn't have to run the table. It just has to win about half the available seats. Nate Silver puts that probability at 67 percent, based on historical trends.
There are two pieces of data that may temper Democratic hopes. The first is that independents have been breaking away from the president and Democrats. How much stock you put in this group depends on how many actual moderates and independents you think there are. If you're a Democrat and you think a lot of so-called moderates and independents lean Democratic, then you think they're yours to get. If not, you believe that they're gone, that it's too late to convince them, and that it's going to be a bad year. Third Way, a centrist Democratic group says that turning out the Democratic base won't be enough to avoid a bad wave (PDF).
But for Democrats, the biggest bucket of cold water comes from looking at voters most likely to vote. In that group, the GOP has a big advantage that can't be dented by the Democratic Party's superior turnout model. With the election less than a month away, Gallup has now refined its methodology: Instead of looking at "registered voters," where Democrats and Republicans are roughly even, it looks at "likely voters." Among that group, and under two different turnout scenarios, Democrats are down by 12 percentage points. Gallup's final likely-voter model has been historically accurate. If it is this year again, then a huge splash of water is heading straight for the Democrats.