By Dan Balz
Washington Post Staff Writer
Wednesday, September 1, 2010; A2
The Gallup organization dropped a bomb on the political world this week. In shorthand, the pollsters said Monday that if the midterm elections were held now, Republicans would take control of the House - and probably by a comfortable margin.
On Tuesday, James Campbell, a professor of political science at the University of Buffalo, weighed in with a prediction based on his modeling of the political climate. He said that Republicans are poised to gain 51 or 52 House seats, at least 11 more than needed to depose the Democrats.
Election Day is still two months away, but the twin findings added to the fear among Democrats that their House majority - and possibly their Senate majority as well - is in jeopardy.
For decades, Gallup has asked voters the following question: "If the elections for Congress were being held today, which party's candidate would you vote for in your congressional district?"
This week's survey produced the largest lead for the Republicans in the history of asking that question: 51 percent to 41 percent. Ninety-six percent of Republicans said they would vote for the GOP candidate, while 88 percent of Democrats said they would support the Democrat. Independents, who helped power Democratic victories in 2006 and 2008, split 48 percent to 31 percent for Republicans.
This measurement (known as the generic ballot question) has sometimes been considered an imperfect or misleading indicator of House election results. Gallup begs to differ. Frank Newport, editor in chief of the Gallup poll, said that Gallup's final survey of likely voters before Election Day has been an accurate predictor of the two parties' share of the national vote in House elections. The national vote, in turn, he added, is an excellent predictor of seats won or lost.
Four years ago, when Democrats won control of the House, the final Gallup survey of likely voters gave Democrats an advantage of seven percentage points over Republicans. Their actual share of the national two-party vote was eight points more.
In 1994, when Republicans won the House and Senate, Gallup showed the GOP with a seven-point advantage in its final survey - exactly the margin between the two parties on Election Day.
In both those elections, the Democrats' share of the two-party vote was almost identical to their share of House seats after the election. And in those elections - and all midterms between 1994 and 2006 - the number of seats won by the Democrats was almost identical to the number predicted by Gallup's models, within three.
It's worth noting that in midterm elections before 1994, Democrats ended up with a higher share of House seats than their take of the vote.
The current Gallup survey was based on interviews with registered voters. Gallup won't start measuring attitudes among likely voters until late September or early October. But everything suggests that a likely-voter survey this year would give the Republicans a greater advantage.
There are some cautionary notes. Gallup's generic ballot measure has fluctuated between late August and early November in past midterm contests. Four years ago, Democrats were ahead by six points in August, by 19 points in October and by eight points in the final survey. In 1994, the two parties were tied in August, October and early November, but the final survey showed the GOP with a clear advantage. So things can change.
Another caveat: Barely two months ago, Democrats held a lead in Gallup's tracking. That caused a moment of hopefulness among party strategists that has quickly disappeared.
Newport said he could not identify empirically the reasons the bottom suddenly fell out for the Democrats. Could it be the worsening economic news, the controversy over the mosque in New York or something else?
"I hesitate without some kind of data to document what's happening out there," he said with admirable restraint.
Caveats aside, Democrats are in deep trouble. Their breaking point is 48 percent in the Gallup generic measurement. At that level, they would be projected to win 216 seats - and become the House minority party. A support level of 46 percent would leave them with an estimated 197 seats. No party has won more than 54 percent of the two-party vote in a midterm election since 1986.
Campbell's paper highlights other reasons he thinks Democrats are in serious danger of losing their House majority. One is a political climate that has energized Republicans far more than Democrats.
But Campbell also says Democrats are overexposed. After picking up 54 seats in 2006 and 2008, they hold territory that would be difficult to defend under any circumstances. More significant, many Democratic-held seats are in trouble, while Republicans have few in danger. That combination has led to upheavals in past midterm contests.
Campbell also dashes Democratic hopes that a spike in President Obama's approval rating will fend off big losses. He said there is little evidence to support the commonly held view that 50 percent is the line of demarcation. In midterm elections, he said it's closer to 65 percent.
"A 65 percent rating in a midterm sounds great but, politically, is approximately neutral and only a precursor to holding their status quo," he writes.
It is clear that, by almost any measure, Democrats badly lost the month of August. Is there any wonder Republicans are optimistic about November?