Scott Rasmussen now predicts that the Republicans will pick up fifty-five seats in the House. Larry Sabato at the University of Virginia still has the pick-up at forty-seven but says that, if forced to tweak the numbers right now, he would increase his estimate of Republican gains by single digits – which is to say, he agrees with Rasmussen.

There are pollsters out there who are playing games, as a glance at the polls for the Senate race in West Virginia should make clear – and, of course, it is easy to play games. If one wants to encourage the Left and discourage potential Republican voters and donors, all that one has to do is to base one’s poll on the presumption that the percentage of self-described Democrats within the voting public in 2010 will be equal to the percentage in 2008.

Sabato and his associates and Rasmussen are not, however, among the gamesters. Both are aiming at accuracy. Sabato and company have a reputation to uphold (and, in the academic world, that is all-important), and Rasmussen is a nonpartisan pollster who attracts clients by way of demonstrated precision. Neither outfit can afford to make a fool of itself.

I nonetheless think that both are greatly underestimating the size of the Republican surge. Both have reason to be cautious. For understandable reasons, neither is going to climb out on a limb; and both are basing their estimates on recent electoral history. If something is in the offing that exceeds the range of political oscillation in recent decades (including, notably, 1994), if we in American live in something other than normal times, they will miss the size of the surge.

It is good to remember that not a single Sovietologist predicted the collapse and dismemberment of the Soviet regime. History has a way of lulling us into sleep. What has been in recent times we tend to think will be in the foreseeable future. Then, every once in a while, suddenly, out of nowhere, a political earthquake arrives – and only in the aftermath do the experts notice that there were ample warning signs.

(more…)

From Rasmussen Reports:

A new Rasmussen Reports national telephone survey finds that 48% of Likely Voters would vote for their district’s Republican congressional candidate, while 36% would opt for his or her Democratic opponent. The survey data was collected on the seven days ending Sunday, September 5, 2010.

This matches the largest advantage ever measured for the Republicans. Three weeks ago, the GOP also held a 12-point lead.

Still, while the margin has varied somewhat from week-to-week, Republicans have been consistently ahead in the Generic Ballot for over a year. During 2010, the GOP edge has never fallen below five points. When Barack Obama first took office as president of the United States, the Democrats enjoyed a seven-point lead on the Generic Ballot.

(more…)