The real point is that Kerry is doing better than Gore did among military voters

In this op-ed in today's Washington Post, Peter Feaver raises several important questions about the military and politics, but his failure to note the difference between a scientific poll and a voluntary email survey is shocking. He acknowledges that the Military Times "survey method is tilted in Bush's favor," but he claims this "cannot account for a spread of 55 points." Really? Does that mean we should seriously believe comparable media web surveys that showed numbers north of 70 percent believing Kerry won Friday's debate? Of course not.

Moreover, both Feaver and the Military Times in its own reporting overlook another factor skewing the results. As a former Air Force public affairs officer, I can testify -- and internal research supports this -- that military public affairs staffs are constantly trying to overcome the perception among military audiences that the Military Times newspapers are official, command-driven publications. Therefore, when Air Force Times asks an Airman about his or her vote, a natural “command influence” effect skews results further in favor any sitting Commander-in-Chief.

In any case, notable Democrats have not suggested that military voters will flock to Kerry over Bush. (Certainly, they think that military voters ought to support Kerry, but that is different.) Democrats have argued simply that, among military voters, Kerry will do better than past Democrats – and that is exactly what the survey shows. Among active-duty troops responding to the initial survey, 17 percent said they support Kerry, while 12 percent said they supported Gore in 2000. Among Guard and Reserve troops, 18 percent said they support Kerry, while 16 percent said they supported Gore. Kerry did better than the previous Democrat in both cases. Moreover, the follow-up survey after the first presidential debate showed Kerry gaining six more points. The broad-brush conclusion that military voters support Bush over Kerry is almost certainly correct. But Kerry is clearly doing much better among military voters than Gore did, which could easily swing an election that was decided last time by just 537 votes in Florida. (Detailed survey results are available here to subscribers.)