Maintaining Optimism in the Face of Reality. Occasional observations on the state of the world, society, business and politics. Usually anchored by facts, always augmented by opinion.

Bush's Hidden Support in The One Real Polling Gap  | e-mail post

A lot of Dems have tried to suggest that the surge of newly registered voters or those individuals with only cellular phones means that polls undersample Kerry's support. These people simply don't understand math and/or polling. Random-digit dialing insures that new registrants are picked up as they register, and I've already completely taken down "The Cell Phone Only Polling Bias and the '50% rule'" in a post back on Oct 16th (basically, the cell-phone only thing could be an issue if and only if we assume cell-phone-only users have significantly different political leanings than their peer group, which is a poor assumption).

However, there is one group that is undersampled by random-digit dialing polls, because they are outside the country, and they do have a significantly different political bias than the population at large: the actively deployed military.

While Michael Moore and others may be convinced that the active duty military must oppose Bush, the Annenberg Center for Public Policy polled active duty military and their families and released the results in mid-October. [PDF results] [Washington Post]

It's funny, Annenberg shows its political bias in this poll. They don't actually ask who the military would vote for, I think solely because they don't want to report the heavy Bush support, and they title their press release: "Service Members, Families Say Pentagon Sent Too Few Troops To Iraq, Stressed National Guard and Reserves, Should Allow Photos of Coffins at Dover, Annenberg Data Show." From the headline, you'd think the military must hate the administration. But, in fact, if you read the release:
"The military clearly likes President George W. Bush better than Senator John Kerry, and strongly believes in its mission in Iraq and his handling of it," said Adam Clymer, political director of the survey."Sixty-three percent of the military sample, compared to only 41 percent of the public generally, approves of Bush’s handling of Iraq."
Even more, though, what he neglects to note, is that the survey says that if you throw out the families of the military, and just focus on the troops, 68% favor Bush's handling. That's almost double the support of the general population, and these are the men and women whose lives are on the line.

In addition, 79% of the troop sample says that we should keep troops in Iraq until there is a stable government in place. (Iraqi citizen polling suggests they feel pretty good about things as well, for what it's worth [Yahoo on IRI poll].)

Oh, and the question Annenberg asked about the flag-draped coffins: they asked if showing the pictures would increase respect for the military, not whether or not they should be allowed to be shown, it's an entirely different question.

As an aside: this election cycle has taken down a lot of my respect for the Annenberg Center for their partisan leanings. I have already taken their FactCheck.org to task for their handling of the issue of tort reform ("Fact-Checking FactCheck.org on Tort Reform") and this survey and most notably the press release they crafted for it strongly indicates their partisan bias. BTW, I actually sent them my post on the tort reform issue, and have never received any response to it, which I found doubly-disappointing. At least I still feel like I can trust Pew.

There is one other survey to consider, conducted by Army Times Publishing (which is owned by USA Today and publishes a variety of military service newspapers). The survey is not a scientific poll, as it relied on e-mail responses. However, 73% of respondents said they would vote for Bush, while only 18% said they would vote for Kerry. Almost 70% of the respondents said Kerry's post-Vietnam activities make them less likely to vote for him. [Military.com]

The Kerry campaign of course dismisses the poll as unscientific. Funny, they never seemed so interesting in good statistical methodology about all those horribly-slanted internet instant polls that were used during the debates.

The net is that we should assume the military absentee voters will break at least 60-40 for Bush. This is potential slight positive for states like Florida, and there are also a lot of National Guard deployments out of swing states like MN and IA as well. For instance, the head of Iowa's Republican Party (until he left, anyway), a state senator and college friend of mine, is on active deployment. I'm just guessing, but I suspect he marked his ballot for Bush.

Now, I'm certainly not going to rave that all kinds of polls have undersampled Bush's support, but there is at least a credible and logical claim that there is some level, albeit small, of underreported Bush support.

e-mail post | Link Cosmos | [Permalink]  |  | Monday, November 01, 2004
Comments: Post a Comment

This page is powered by Blogger. Isn't yours?