A Bush uptick?

The first arrests and disclosure of the London bomb plot were news in this country on the morning of August 10, two days after Sen. Joe Lieberman lost the Democratic primary to Ned Lamont. I speculated on Fox News that the London plot may have reminded voters that we are really threatened by vicious terrorists and that that may prompt them to be more supportive of George W. Bush and his fellow Republicans. Now we have some polls that tend to support that view.

Here are some numbers via Real Clear Politics.

Bush's job approval was 42 percent in the CNN and USAToday/Gallup polls conducted August 18-20 and in Rasmussen's three-night track August 19-21. That's higher than in all but one other poll in RCP's list going back to the beginning of May, though it should be noted that Rasmussen had him averaging 42 percent for the month of July. (Rasmussen asks the question in a form that tends to elicit slightly more positive ratings than other polls.)

Then there is the generic ballot question: Which party's House candidate would you vote for? On their August 18-20 polls, CNN had the Democrats ahead of Republicans 52 percent to 43 percent, while USAToday/Gallup had Democrats ahead by only 47 percent to 45 percent. The USAToday/Gallup was a sharp shift from their most recent previous poll, taken June 23-25, which showed a 54 percent to38 percent Democratic lead. Gallup, which does not weight results by party identification, tends to have more volatile results than other polls, but still that's a statistically significant shift, with Democrats down 6 percent and Republicans up 7 percent. Rasmussen, who does weight by party identification and whose poll results are much less volatile than Gallup's, shows the Democratic lead in the generic vote as 41 percent to 35 percent (likely voters: 46 percent to 38 percent) (members only). His June survey showed the Democrats with a much bigger lead in generic vote, 47 percent to34percent. So, assuming the June results are likely voters, he's showing only a 1 percent drop for Democrats but a 3 percent gain for Republicans: not statistically significant.

Should we care about the generic vote? As I've written before, it's a poor guide to how elections turn out; Democrats have been ahead in generic vote questions about 70 percent of the time since Republicans won their House majority in 1994, but over that period we have had five House elections and in each one Republicans won more votes and more seats. Nonetheless, I think a shift in the generic vote may tell us something. And what it may tell us is that there's a change in interest and enthusiasm, which may be relevant to turnout. Gallup's numbers are volatile partly because it screens out less interested or motivated respondents, and so the changes in its numbers suggest changes in Democratic/Republican morale and intensity. The shifts in Gallup's generic vote, together with its higher Bush job approval, present the hypothesis that the London bombing plot increased the determination of Republicans to support their president and their party and their determination to vote. Raises the hypothesis; doesn't prove it.

There will be more to say when more polls come out. So the answer to the questions, a Bush uptick? a Republican uptick? is: possible but not yet for sure.