STORY OF THE WEEK: Are projected Republican gains overstated?

For months, we and others have noted the difficult political environment facing the Democrats this cycle. To be sure, there are a number of obvious warning signs that suggest Republicans have a good shot of taking the House, and an outside chance of taking the Senate. While we do not disagree with this overall assessment, we believe projections of Democratic losses may be somewhat overstated because they rest, in part, on several less-than-stable assumptions about the electorate this cycle.

For example, tucked into a recent, highly-publicized Gallup survey showing Republicans with a 13-point advantage on the generic Congressional ballot were two questionable data points. The first shows a majority (54%) of survey respondents describing themselves as "conservative," up from 40% in 1994 and 42% in 2002 and 2006. This is a monumental shift when you consider the relative stability of this metric over the past 20 years (this picture is worth a thousand words).

Additionally, exit polling shows that since 1984 conservatives have not comprised more than 37% of the electorate, or a larger slice of the vote than moderates. Needless to say, a survey showing so many conservatives is going to overstate the Republican margin. And Gallup is not alone here; as Kristin Soltis points out, a handful of prominent public polls are showing more conservatives---and thus larger GOP margins---than we might expect on Election Day.

The second data point that caught our eye concerns the gender dynamic. Gallup's poll has 48% of the vote coming from women, which would be a record low dating back to 1994. Women have comprised a majority of the electorate in each of the past four elections (between 51% and 54%), and they are much more likely to vote Democratic (there has been a clear gender gap in every election since 1992, and there is a gender gap in every recent public poll). Short a survey of women, as Gallup has done, and it will overstate the GOP advantage.

Margie Omero points out another data point on the gender that is worth watching in the coming weeks. Since 1994, the Democrats have won or held a majority in the House when they have met one of the following conditions: 1) win the male vote by any margin; or 2) win the female vote by double-digits. While the Democrats are not currently satisfying either, they have a natural advantage with women and still have the opportunity to narrow their losses by energizing this pivotal slice of the electorate. Omero offers a number of helpful suggestions to this end.

As we noted last week, there is also compelling evidence that polls may be overstating GOP margins if they exclude cell phones in their samples. Last week, Pew Research released data showing a 4-6 point GOP advantage in landline-only surveys, up from about 2 points in 2008. However, it is important to note that many of the top national polls, such as ABC/Washington Post and NBC/Wall Street Journal, currently include cell phones, and these polls still show Republicans with modest leads on the generic ballot. But it does raise questions about the litany of statewide and congressional district polls that typically do not include cell-phones.

The "incumbent rule" has also fueled projections of a GOP landslide in recent weeks, but again, new data suggests commentators may be overstating the degree of anti-incumbent sentiment among voters. The "rule" suggests that incumbents who are below 50% are in deep trouble because undecided voters typically break for the challenger. Nate Silver at FiveThirtyEight.com finds, however, that incumbents who were leading their opponents with less than 50% in October over the past few cycles had a winning percentage of 80% (66-17). This should come as a silver lining to those incumbents, many of whom are Democrats in tough races, who are sitting just below the 50% threshold.

Finally, we have written often about how grand predictions of a GOP landslide may overlook district-specific peculiarities, and more importantly, the effect of Democrats using their sizeable war chests in the final weeks of campaigning to define their opponents and frame their races as a choice, instead of a referendum. Anecdotally, we are seeing evidence that these efforts are paying off. The Fix recently highlighted 10 high-profile races where Democrats have started pressing contrasts and creating positive separation between the two candidates (Anzalone Liszt is proud to have a hand in several of these contests). Democrats will continue to do so through Election Day.

OTHER NEWS FROM THE POLLING AND POLITICAL WORLD

Senate: GOP insiders see a Senate takeover increasingly out of reach, in a new National Journal Political Insiders poll.

Battleground districts: A new NPR poll shows the Democrats are closing the gap with Republicans in 58 competitive House districts. They are down just 3 points, from 8 points in June.

Crooked ads: Greg Sargent details the many misleading Chamber and Rove ads this cycle.

Health care: Scott Murphy goes up with an excellent pro-health reform spot.

Time "40 under 40": LA House candidate and ALR client Cedric Richmond is profiled in Time's prestigious article.

Early voting and registration: Michael McDonald examines increases in early voting and changes in party registration this cycle, and the implications (here and here).

Sexism on the trail: Fascinating article about the effect of sexist remarks on support for female candidates.

POTUS: A new Daily Kos/PPP survey finds that Obama is more popular than Sarah Palin in Alaska.

Cultural diversion: Esquire released a fun poll comparing attitudes of 20-year old and 50-year old men.

Enthusiasm gap: Nate Silver at FiveThirtyEight.com makes the case that the Democratic base is actually relatively enthusiastic about this election; Republicans are just much more enthusiastic than usual.

Voter Confidence Index: NBC's current VCI index stands at -40, down one point from September. That puts Obama and the Dems in a better position than Bush in 2006, but worse than Clinton in 1994.

PUBLIC POLLING

Follow us on Twitter: AnzaloneLiszt   Like us on Facebook: Anzalone Liszt Research