House Race Summary for Nov. 3:
A new poll from WMUR and the University of New Hampshire shows Democratic challenger Paul Hodes running eight points ahead of Republican Rep. Charlie Bass (45 percent to 37 percent) in New Hampshire-2. Hodes leads by 2.6 points on average of three polls released this week (46.3 percent to 43.7 percent)—not quite enough to merit "lean" Democrat status but still very much a tossup race. That standing is remarkable given that these candidates faced each other in 2004, and Bass in round one won by a 20-point margin (58 percent to 38 percent).

House Race Summary for Nov. 2:
If 2006 turns out to be a "wave" election that sweeps unexpected numbers of Republicans out of office, it will be districts like Arizona-5 and Ohio-2 that tell the story. We have new SurveyUSA polls in both districts that show the Democrats with small but statistically insignificant leads (two points in Arizona-5 and three points in Ohio-2). While these results do not alter the "tossup" status of either race, the fact that either district is a dead heat this late in the campaign is remarkable. In 2004, George Bush received 54 percent of the vote in Arizona-5 and 64 percent in Ohio-2, the same year that the respective incumbents in each district won with 59 percent and 72 percent, respectively.

House Race Summary for Nov. 1:
Today we have 15 new Reuters/Zogby polls in 15 hotly contested House races. The only change in any of yesterday's assessments is in Indiana-9, where they show incumbent Republican Mike Sodrel running just two points behind Democratic challenger Baron Hill. The automated Majority Watch poll released earlier in the week had Sodrel running eight points ahead of Hill, just enough to make Sodrel's average lead in recent polls statistically meaningful. This new poll, however, returns this race to tossup status.

The Reuters/Zogby release had a few head-scratchers, such as the poll in Illinois-6, which shows Democrat Tammy Duckworth running 14 points ahead of Republican Peter Roskam. Three other polls released earlier in the week showed the race too close to call: Two had Duckworth leading by a single percentage point, and another showed her trailing by four points.

House Race Summary for Oct. 31:
More than 200 polls have been released since Oct. 1. Most of the closely watched races have been polled at least two or three times this month, and for the purpose of this scorecard, we will focus on them. Those October polls show Democrats heading toward enough seats to win a majority, although how many seats they win may depend on which polls you trust.

The math is easier given one important finding: Not a single Democratic candidate in a district now held by a Democrat is trailing, regardless of the combination of polls examined. So, we will focus below on potential Democratic pickups.

We count 10 seats currently held by a Republican where multiple surveys shows the Democrat with a statistically meaningful lead, regardless of what combination of polls we look at: Arizona-8, Colorado-7, Indiana-2, Indiana-8, North Carolina-11, New Hampshire-2, New Mexico-1, Ohio-18, Pennsylvania-7, and Pennsylvania-10.

One seat gets its own category: The one and only poll in the Texas-22 district formerly represented by Rep. Tom Delay shows Democrat Nick Lampson leading. However, a complicated ballot (Republican Shelley Sekula-Gibbs is a write-in candidate) makes this result tenuous.

Seven more Republican seats look to be in statistically meaningful jeopardy, but only when you count the automated Majority Watch surveys (either because they are the only ones done or because they tip the balance, making the Democrat's lead statistically meaningful): Indiana-9, Iowa-1, New York-24, New York-25, New York-26, New York-29, and Ohio-15.

Three more Democrats would show significant leads if we include the internal surveys released by partisan pollsters: Florida-13, Nebraska-3, and Ohio-1.

Thus, if you trust the Majority Watch surveys and assume a pickup in Texas-22, then Democrats are leading in exactly the 18 seats the need to win a majority. If you trust all polls, then they currently lead in enough districts to win 21 seats.

But even more important: Polls have been conducted in October in another 25 seats where the averages indicate a statistical tossup. Only two of these seats are currently represented by Democrats. How well the Democrats ultimately do depends on how many of these still-too-close-to-call races they ultimately win.

House Race Summary for Oct. 20
Math is hard. Polling House races is harder. That's the message you get if you look closely at the numbers in our tables below.

Since our last update, we have added six new surveys from the Majority Watch project and other polls from a variety of pollsters. We continue to focus on nonpartisan surveys conducted since late August. If we average results in districts currently held by Republicans that have been polled at least twice, we see Democrats leading beyond the margin of error in the nine districts: Arizona-8, Colorado-7, Indiana-2, Indiana-8, Indiana-9, North Carolina-11, New Mexico-1, New York-26, and Pennsylvania-10. This list has not changed since our last update.

We now show six GOP districts with the Democrat leading beyond the margin of error, but based on only a single non-partisan poll: Florida-16, New York-19, New York-24, New York-29, North Carolina-8, and Ohio-15. All six polls were conducted by the Majority Watch project.

If we average surveys in seats currently held by Democrats, we see many close contests, but not a single race shows a Republican leading by a significant margin.

Our count by no means preclude further Democratic gains (or some Democratic losses), because for every race where the polling creates a reasonably consistent picture, we see many more that are very close, or where no polls have been taken, or where the results conflict. Consider New York's 20th District, where Democrat Kirsten Gillibrand is challenging Republican Rep. John Sweeney. A Majority Watch survey conducted Sunday to Monday this week showed Gillibrand leading by 13 points. Then, Siena Research Institute released a survey conducted Monday and Tuesday nights showing Sweeney ahead by 14 points. Later, Gillibrand's own pollster released a survey, conducted Monday to Wednesday, that showed Gillibrand leading by only 2 points.

Neither sampling error nor the slight difference in timing explains that potpourri of results. The more likely explanation involves the degree of difficulty involved in polling House races, which involves as much art as science. For all the numbers, unfortunately, we still see quite a bit of statistical noise. Any one poll should be taken with a grain of salt.

House Race Summary for Oct. 17:
Since our last summary, we have changed from talking about the generic ballot to a more precise look at what's happening in contested races. Below is a list of the 63 hottest House races. Click on an individual race to jump to the most recent polling, or scroll down the page to see how all the competitive races have been shaping up.

One of the challenges of using polling data from individual races is that public polls are plentiful in a small handful of competitive districts and rare in most others. For example, the majority of the most closely watched races have been polled only two or three times since late August. So, we have nowhere near as much data for House races as we do for closely contested Senate and governor races.

We do see some clear leaders. Keeping in mind that Democrats need to gain 16 seats to win control of the House, we looked closely at the most hotly contested seats currently held by Republicans. We looked only at nonpartisan surveys conducted since late August. In districts that have been polled at least twice, we see Democrats leading beyond the margin of error in nine districts: Arizona-8, Colorado-7, Indiana-2, Indiana-8, Indiana-9, North Carolina-11, New Mexico-1, New York-26, and Pennsylvania-10.

Democrats also hold significant leads in three more districts that have been polled only once by nonpartisans since the summer: Florida-16, New York-24, and Ohio-15.

That would make a total pickup of 13 seats, but Democratic gains could be much greater if present trends continue. At least a dozen hotly contested seats currently held by Republicans look too close to call based on the available data, and no Republican leads by a significant margin in any district currently held by a Democrat.