House Race Summary for Oct. 10:
Since our update last week, we have five new national polls on the generic House vote—the new Gallup poll plus surveys from ABC News/ Washington Post, CBS/ NY Times, CNN/ORC, and Newsweek—and four of the five show some improvement in the Democratic margin. Our chart shows a sharp upward movement in the last-five-poll average of the Democratic margin from roughly 10 percentage points two weeks ago to almost 17 points today.
Then there's the new USA Today/Gallup poll which is puzzling. A month ago, the voters Gallup considered to be most likely to vote were evenly split (48 percent each) on whether they would support the Democratic or Republican candidate in their Congressional District. Their latest survey shows Democrats surging to a 23-point advantage (58 percent to 35 percent). Has the race for control of the House of Representatives really changed that much? We don't know.
One technical note: From now on we are plotting results for the likely voters rather than registered voters for surveys that report both. While the debate among pollsters remains heated over which subgroup is most reliable, we have examined the polls released since Labor Day and on average we see very little difference in the results.
To Review: Why do we use this chart?
Unfortunately, for now, public polling data on individual House races is too rare and incomplete to create a comprehensive scorecard. The chart above shows the trend in results from national polls for the so-called "generic" ballot, which asks respondents whether they would vote for "the Democratic candidate or the Republican candidate in your district." For many months, respondents have consistently favored Democrats, which is why, for now, the graph plots how much they favor a generic candidate from the minority party over one from the Republican majority. (Click here for an explanation on how to read the chart.)
While the generic House ballot has been a reasonable indicator of which party is faring better, it is a very imperfect predictor of both the total national congressional vote and, perhaps more importantly, how that vote translates into seats. We have done more commentary on this issue on our blogs (Mark here, Charles here and here).