For Republicans, poll numbers have never looked better prior to a midterm election than they do today. “You’ve got to pinch yourself every time you look at the data,” pollster Neil Newhouse said. A Republican victory “could be bigger than anyone thinks.”
But, how predictive are the numbers? Are they reliable enough, assuming they don’t change significantly before the Nov. 2 election, that we should expect a massive Republican sweep?
The answer may be found by comparing today’s numbers on the two poll questions that matter most — presidential job approval and the generic ballot — with those in the landslides of 1982, 1994 and 2006.
For the comparisons, I’ve used only Gallup Poll numbers.
As of Wednesday, President Barack Obama’s job performance was approved by 41 percent and disapproved by 52 percent, the worst showing of his presidency. Republicans had a 50 to 43 percent advantage on the generic ballot, which asks respondents whether they intend to vote for a Republican or Democrat for the House.
The general rule has been that when a president’s rating dips into the mid- to low 40s (or lower), his party is likely to lose a substantial number of House seats in a midterm election and some in the Senate. And, this trend is reinforced when the opposition party is ahead on the generic ballot.
In 2006, President Bush’s approval was 42 percent in mid-August and fell to 37 percent in the final poll before Election Day. In August, Democrats led Republicans, 47 to 45 percent, on which party’s candidate was preferred. The final pre-election poll gave Democrats an advantage of 51 to 44. The election result: Democrats won 31 House and six Senate seats, capturing control of both bodies.
In 1994, President Bill Clinton’s approval rose from 40 percent in August to 46 percent in early November. But, the Republican lead on the generic question also rose, from 47 to 44 percent in August to 53.5 to 46.5 percent in November. The result: Republicans gained 54 seats in the House and eight in the Senate, winning a majority in both chambers.
In 1982, President Ronald Reagan’s performance in office was approved by 41 percent in August and 42 percent in late October. Democrats had a whopping 54 to 36 percent advantage in August and a 55 to 45 percent lead in October. The result: Democrats won 27 House seats, but lost one in the Senate.
Why didn’t Democrats win more seats in 1982, given the strength reflected in poll numbers? With 242 House seats before the election, they were near their high-water mark, post-Depression, in the size of their majority. By boosting it to 269 seats, Democrats erased most of their losses in 1980.
For the most part, the poll numbers in 2006, 1994 and 1982 changed little between August and November. That’s the usual pattern. The numbers aren’t likely to change in 2010 either. The public’s assessment of the economy — 2-to-1 negative in Gallup — is usually settled this close to an election.
Pollster David Winston said Republicans have better prospects in 2010 than 1994. But, he said Republicans “have the challenge of what people remember from 2006,” when they were voted out of office in Congress. And, they must answer what Winston calls the “why us?” question. Republicans need to tell voters “what you’re going to get if we have a majority.”
Fred Barnes is executive editor of The Weekly Standard, where this article appeared.