It appears that the signs of a possible Democratic rebound were just a false alarm. After almost a year of unrelentingly bad poll numbers, back-to-back weeks of Gallup Polls showing Democrats with eyebrow-raising leads might have just been statistical flukes.
For the weeks of July 12-18 and July 19-25, the Gallup Organization's weekly aggregation of daily tracking polls showed Democrats ahead among all registered voters on the generic congressional ballot test question 49-43 percent and 48-44 percent, respectively. Each poll of more than 1,500 registered voters had 3-point error margins.
In the 21 weeks since Gallup began asking the generic ballot test question daily, releasing the aggregate numbers once a week, the two parties have averaged a tie.
Importantly, this tie is among all registered voters, which tends to not correspond to actual turnout numbers, and certainly not in midterms, which tend to skew Republican.
Midterms also particularly show a decline in turnout among minority and younger voters, who trend Democratic.
But still, a Democratic generic ballot test advantage among registered voters would be a big deal. It would suggest a rough tie in the national popular vote for the House, which would translate into an even-money situation for House seats for each side.
My view for over a year has been that a very bad political environment for Democrats was developing, and I have been convinced for some time that we might well see a wave reminiscent of those of 1994 and 2006, frequently pointing to data that corroborated that view.
Thus, when there were back-to-back weeks of the respected Gallup Poll pointing toward Democrats by margins a bit wider than the normal statistical flutter, it seemed fair to point to it in my National Journal magazine column.
I did so with a healthy level of caution, even skepticism, pointing out in the first paragraph that the numbers could be "a statistical anomaly that is the political equivalent of a false positive medical test" and noting that there hadn't seemed to be the kind of defining event that would cause a substantial shift in momentum.
It was also noted that President Obama's job rating hadn't budged, something that would be inconsistent with some kind of Democratic surge and that a quick canvass of pollsters on both sides had failed to find any others seeing meaningful movement.
The concluding advice was that readers should sit tight and watch for next week's Gallup release and for other reputable pollsters to weigh in.
Subsequent polling suggests that nothing has happened.
A July 16-21 CNN/Opinion Research Corp. poll of 945 registered voters with a 3-point error margin, put Republicans ahead, 50-44 percent; Democrats had been ahead in their April and May surveys. The Republican firm of Public Opinion Strategies had the two parties within the 4-point error margin in its July 18-20 survey of 800 likely voters.
Some pollsters privately suggested that there might have been some Democrats coming home, returning to the fold if you will, but there was no key change among the key groups that Democrats need to move.
Monday afternoon, Gallup released its July 26-Aug. 1 sampling of 1,562 registered voters, with a 3-point error margin. It showed Republicans up, 48-43 percent, with Obama's approval rating steady at 45 percent.
The environment still looks exceedingly bad for Democrats and last week's disappointing Commerce Department report that showed the gross domestic product grew 2.4 percent for the second quarter points to an economy so anemic as to make any meaningful improvement in unemployment extremely unlikely between now and Election Day.
But this recent fire drill serves as a useful reminder that there are 91 days between now and the Nov. 2 midterm elections and that things can always change. The closer the election gets without change makes it less possible, but things change. As this column noted last week, Republicans may have to gross as many as 43 seats to net the 39 needed to tip control, as there are as many as four Republican seats that could very plausibly turn Democratic this year, even in a terrible environment for Democrats.
Democrats also have a spending advantage that will no doubt help them and, as was seen in the special election to fill the seat of the late Rep. John Murtha, D-Pa., when Democratic Rep. Mark Critz was elected over Republican businessman Tim Burns. Democrats seem to have an edge on organizational mechanics.
The Cook Political Report last week upped its projection from a 30-40-seat net gain for Republicans to a 32-42-seat net gain, with our model showing +34 seats for the GOP. This count is more heavily weighted toward individual race dynamics, but my gut feeling -- which, in years like this one, factors in national mood, relative intensity and enthusiasm between the parties and how independents seem to be breaking -- is suggesting somewhat bigger numbers.
The wave is still the wave, and it still looks pretty strong and unabated.
DON'T MISS TODAY'S TOP STORIES