WPA's Election Update: 3 Weeks Left

Some weeks in politics not a lot really happens.

Last week was that kind of week in the Presidential race as the Vice Presidential candidates met in a debate that wound up producing more stories about Joe Biden’s veneers and Martha Raddatz’s activist moderating than it did about anything that would actually move votes.

This week we’ll see the actual Presidential candidates meet in a town hall that could move numbers more than the Vice Presidential candidates trading smirks and Irishisms did.

With not a lot happening in the Presidential race, I’ll spend time this week focusing on the race (if you can call it that) for control of the House.

The Presidential Race

Still tied, more to come.

An ABC News/Washington Post poll from this morning that shows how the big TV News/Newspaper polls continue to be outliers in the pro-Democrat direction notwithstanding, the race is basically the same today as it was a week ago.

Of the six polls completed last week, two show a one point Romney lead, two show a two point Romney lead, and two show a one point Obama lead.  Averaging across those polls shows a very close race with Romney leading by around two-thirds of a point.

Even if we include ABC/WaPo’s seeming outlier of Obama +3, we still see the average from last week as a very slight edge for Romney.

That’s about where things stood in the first post-debate polls.  It appears that the ground Romney gained in the debate has held up, but nothing that happened last week improved his position beyond that.

Earned media, and especially earned media surrounding the debates, has taken an outsized role in this campaign beginning with the Republican primary.  I anticipate that, unless the town hall debate produces no moments of interest at all, another meaningful shift in the data will come after tomorrow night’s debate.

The Senate Race

Full of polls and numbers, signifying nothing.

Not a lot seems to have changed in the Senate races.  Or at least not that we can determine looking at the polls.  But, man, those polls.

Last week there were 41 polls released in Senate races around the country.

  • Many of these were IVR polls that exclude cell phones and historically have been more accurate in the closing week of an election than earlier (read: this far out).
  • A number were from organizations with a history of releasing outlier numbers that benefit one side or the other, seemingly more for the purpose of creating an earned media “buzz” about a race than offering a scientific assessment.
  • Some others were media or university polls whose biases due to choices about sampling and weighting I detailed two weeks ago.

What this all yields is an extraordinarily wide range of estimates in any given race:

  • Florida:
    • On Sunday, PPP released Florida data showing Bill Nelson up by eight.
    • This morning Rasmussen Reports released data showing Nelson up by only a single point.
  • Virginia, all on the same day:
    • On Thursday we saw numbers from NBC/Wall Street Journal/Marist University showing Tim Kaine with a one point lead in Virginia.
    • But WeAskAmerica released numbers that same day showing George Allen up by five.
    • And CBS News/New York Times/Quinnipiac University showed Kaine with a seven point lead.

I could go on.  The point is that, for right now, there is little consistency in the Senate polling to suggest anything other than what I described last week—there are a lot of close races, but Republicans would need to come close to running to table to tie or win the Chamber.  It’s still an uphill race for GOP control.

The House Race

Unless something big happens, Republicans will hold the House. 

I don’t like to talk in absolutes because, despite what I often tell perspective clients, I can’t actually see the future.  It’s entirely possible that some major scandal could come to light in the next few days that would totally reshape the battle for the House.

But, assuming that doesn’t happen, Republicans should feel pretty safe.

Assessing something like the House is a bit tough because there is simply too little public polling on a seat-by-seat basis to make good judgments.  Some prognosticators will use fundraising and story-based judgments about momentum to judge individual races.  We have found that looking at a few environmental variables makes more sense.

So let’s look at two indicators of how things might go:

First, the generic Congressional ballot (that’s a ballot with no names associated with the candidates, just their partisan labels).

Here is some historic data from the last few election cycles.


Generic Ballot

Seat Gain


Republicans +2%

Republicans +7


Democrats +8%

Democrats +37


Democrats +10%

Democrats +29


Republicans +7%

Republicans +69

Today, according to the Real Clear Politics polling average, Democrats lead on the generic ballot by less than one percent (1%).  Historically, small margins like that are associated with a relatively static House make-up.  With a big lead in the House to start, this is reassuring for Republican chances of holding the Chamber.

Our second key indicator is the margin of the Presidential popular vote.

Let’s look at a handful of key years, focusing on Presidential election years.


Presidential Popular Vote Margin

Seat Gain


Republican +10%

Republican +46


Republican +18%

Republican +12


Democrat +9%

Republican +10


Democrat +8%

Republican +11


Republican +3%

Republican +7


Democrat +7%

Democrats +29

The trends here are somewhat obscured by the slow erosion of Blue Dog Democrats throughout the 1990s, but the bottom line is that in all but the biggest landslide Presidential years the changes in the House are relatively modest.  And even a landslide year isn’t guaranteed to produce a big shift in the House.

Combined with what is, at this point, a very close Presidential race, this data suggests that we won’t see the kind of “wave” in the House that would flip the Chamber.

Follow Ryan on Twitter