An In-Depth Look At Key Congressional Races — NPR Survey

Back in June, the survey that Stan Greenberg and I conducted for NPR was huge news. Conducted in 60 Democrat-held and 10 Republican-held seats, it showed Republicans ahead by a 48%-39% count in the Tier 1 seats (30 most vulnerable) and up 47%-45% in the Tier 2 seats (next 30 seats).  It was the first indicator to the national media that a wave was building.

Once again working with Greenberg, Quinlan, Rosner Research (GQR) we just did a similar survey for NPR. However, we expanded to all the seats on the Cook Political Report’s list at the time – 86 Democratic held seats (53 in Tier 1) and 10 GOP seats.

As always, any analysis on this blog is solely mine and does not reflect that attitudes and opinions of either NPR or GQR.

The ballot test is NOT the same as a generic ballot. Rather than simply ask if they were voting for the Republican or the Democratic candidate, we tested actual candidate names by district. While the sample size per district is too small to look at any single campaign, it does give clear direction on the situation in the House Battleground seats.

In the Tier 1 seats, the Republican candidates lead 48%-44%. The GOP candidates are tied with the Democratic candidates at 45% each in Tier 2, and in the GOP seats (which are actually heavily Democratic), the GOPer leads 49%-42%. Combined across the 86 Democrat-held seats, the GOPer leads 46%-44%.

(For reference, averaged across the Tier 1 Dem seats, McCain actually won those districts in 2008 by a 50.1% to 48.6% margin. In the Tier 2 Dem seats, McCain lost 48%-50.9%. The GOP seats voted just 40.6% McCain and 58.1% Obama (remember, the GOP seats wouldn’t be on this list if they were safe seats!).

Not every one of those Dem seats are going to go Republican – clearly much hangs in the balance. However, incumbents who are polling 45% or below when there is a wave against them lose.

Underscoring the importance of the enthusiasm gap, Republicans lead 50%-41% in the 86 Democrat seats among high interest voters – those who rate their interest as an 8-10 on a scale of 1-10. That is reinforced by the finding that low interest voters prefer the Democrat by a 32% GOP/55% Dem margin. So, the group that Democrats are doing best with don’t care.

Stan Greenberg makes the point that these are heavily screened voters, so even the less interested ones are likely to vote. I’m a little more skeptical, and it underscores the hard road the Democrats have on motivating their voters. If they succeed, they’ll hold their losses down below 50, and if they fail, the losses will cross over the 50 level and could be huge.

There were 58 seats that we tested in both June and October. In June, the GOPer led 49%-41% in those seats. Now, the GOP lead is smaller – 47%-44%. But again, these are mostly incumbent seats and, aggregated, the Dem is both losing AND below 45%.

Incumbent and open seat polling is not created equal. Undecideds tend to break against the incumbent, while open seat undecideds tend to split more evenly. Undecided voters in the 86 Democrat seats aren’t too fond of the President – he has just a 35% approve/55% disapprove rating. They are also very pessimistic about the direction of the country – 15% right direction/77% wrong track. The pessimistic voters – wrong track – break 66% GOP/23% Dem on the ballot. These undecideds will break more to the GOP than to the Dems.

While each race is an individual story, the broader trends are helpful to look for clues as to what is driving the electorate. Key findings among subgroups include:

  • Despite Republicans dominating among Independents in the year old Virginia, New Jersey and Massachusetts elections (I know, Massachusetts was in January), the ballot is tied among Independents in these 86 Democratic held CDs.
  • There is an eighteen point gender gap – GOPers lead 51%-40% with men, and trail 41%-48% with women.
  • The African American vote is maxed out for Dems – they break 6% GOP/87% Dem, so there is no easy gains for the Dems among that base group.
  • Conservatives are united for the GOP – 71%-18%, and mod/lib GOPers are also strong (80%-14%). Moderates break heavily against the GOP (30%-63%). However, mod/lib GOPers back the Republican 80%-14%.
  • Despite (or maybe because of) the gender gap, Republicans would do well to remember that there is a marriage gap. GOPers lose married women just 45%-46%, while trailing 33%-53% with single women. Conversely, GOPers win single men by a modest 45%-41%, but carry married men 53%-39%.
  • GOP candidates need not fear the union label. Republicans are getting 36% among union households, while Dems take just 52%.
  • In the Dem-held districts where Obama got under 50% of the vote in 2008 (that’s 46% of these districts), the GOP candidates leads 49%-42%. In the 28% of the districts where Obama got 50-54% of the vote, GOPers lead 46%-44%. The 26% of the districts where Obama earned 55% or more of the vote, the GOP candidates trail 41%-49%.
  • Another cool crosstab is that Dems who were elected in 2008 trail 50%-41%, while those who were elected in 2006 are in similar trouble (49% GOP/40% Dem). The 47% of the seats who are longer term Dems are safer – the GOP candidates trail 42%-48%. Thus, some longer term members will get knocked off, but the bulk of the GOP wins will be among districts won in the recent GOP bloodbaths – a return to the old normal is likely.

 

Methodology

The survey was conducted October 7-10, 2010 among 1,200 likely midterm election voters. The survey is best considered in three parts – the 53 Tier 1 Democratic seats, where we conducted 450 interviews, the 33 Tier 2 Democratic seats, where we conducted 450 interviews, and the 10 GOP seats, where we conducted 300 interviews.