Monday, January 01, 2007

Poll Numbers

As the number of U.S. military deaths in Iraq passes 3,000, the Gannett-owned "Military Times" newspapers are out with their annual survey of members of the armed services, which (at first blush) shows decreased support for the war and President Bush.

According to the poll, only one-third of the nation's military personnel approve of Mr. Bush's handling of the war, and just 50% believe we'll be successful in Iraq. Those numbers represent a significant slide in support over previous surveys. According to the Times, more than 80% of survey respondents in 2004 expected success in Iraq; in that same poll, 63% of military personnel expressed support for Mr. Bush's handling of the war.

But, as we cautioned in our analysis of the 2005 poll, there are potential problems with the survey and its results. As in the past, the Military Times used a mail-in survey, directed to its subscribers. Internals for this year's survey were not available, but if the 2005 breakout is any indication, the poll had an overall response rate of 30%--not bad for that type of survey, but it still begs an obvious question: why use such a dated technique when far more effective methods are available? Military Times claims that the mail-in approach makes it easier to reach readers in combat zone, but that assertion is questionable. Afterall, opinion surveys are conducted in Iraq on a regular basis, despite the dangers. If pollsters can reach the Iraqi public, it shouldn't be that hard to measure opinion among military personnel at various bases and camps within that country.

We've also questioned the ability of the Times' mail-in survey to accurately reflect its audience. When the 2005 poll was published, an alert Michelle Malkin reader noted that 58% of those respondents had never deployed to Iraq or Afghanistan, suggesting an under-representation of combat veterans. Among this year's participants, 66% said they had deployed to one of those combat zones, an 18% increase over the 2005 poll. That begs a couple of obvious questions: first, why were past polls under-weighted in that category (suggesting that some respondents based their opinions more on media reporting than personal experience), and secondly, how did the Times manage to increase the number of combat veterans in its poll by almost 20%--in one year--using a voluntary, mail-in survey?

Finally, the Times publications manage to contradict themselves on the poll's reliability, all in the span of a few sentences. At one point, editor Robert Hodierne notes:

"The results should not be read as representative of the military as a whole; the surveyÂ’s respondents are on average older, more experienced, more likely to be officers and more career-oriented than the overall military population."

But, just two paragraphs later, Mr. Hodierne trumpets his survey

The poll has come to be viewed by some as a barometer of the professional career military. It is the only independent poll done on an annual basis. The margin of error on this yearÂ’s poll is plus or minus 3 percentage points.

So, what is it? Either you're an accurate barometer of the nation's military, or you're a flawed, mail-in survey that may or may not be an accurate reflection of opinions in the armed services. After last year's poll was published, a number of bloggers and posters suggested that readership of the Times publications had actually declined among career military personnel, who are suspicious of the paper's ownership and editorial slant. If that's accurate--and I would appreciate any historical data on readership/circulation trends among these publications--it brings us back to that originalfundamentalal issue: exactly who is responding to this survey and how accurately does it reflect thsentimentsts of those who wear the uniform? To its discredit, the Military Times papers typically require a separate, e-mail request from readers and pollsters who want to examine their data, and how it was derived.

Given the steady drumbeat of bad news from Iraq, there is little doubt that support for the war--and Mr. Bush--has declined among military personnel. But has it plunged as much as the Times' survey would indicate? Absent more detailed data on the poll's internals, we'll offer the same advice as in early 2006--take this one with a huge grain of salt.


Careful readers may recall that Mr. Hodierne is the same journalist whose reports from Vietnam were oncdescribeded as "giving aid and comfort to the enemy."


Coach Mark said...

Wasn't this group the same hacks who "called for" Rumsfeld's resignation and then had their partisanship exposed?

Mike H. said...

Trust the Leatherneck. Don't trust any thing with Times in the name.