WP Poll: Public Evenly Split on Afghan Escalation?

"U.S. Deeply Split on Troop Increase for Afghan War" is the headline on the Washington Post's October 21 report about its latest polling on Afghanistan. The paper reports that "Americans are evenly and deeply divided" over sending 40,000 extra troops: "47 percent of those polled favor the buildup, while 49 percent oppose it."

If you've followed polling on this question, these results are striking–most recent surveys show the public is deeply troubled by the war and opposed to sending more troops. The most recent CNN survey (10/16-18/09), to take one example, found 39 percent support for sending more troops, and 59 opposed to that idea.

So who did the Post get those results? They've been asking questions about troop buildup in their other polls, but for this one they changed the wording of the question to this:

U.S. military commanders have requested approximately 40,000 more U.S. troops for Afghanistan. Do you think Obama should or should not order these additional forces to Afghanistan?

It's very likely that including references to "military commanders" and Obama skew theresponses to the question–as has been noted, Obama tends to poll better than his policies do.One of the Post'srecent polls (8/13-17/09)on Afghanistan was more neutrally worded:

Do you think the number of U.S. military forces in Afghanistan should be increased, decreased or kept about the same?

The result then: 24 percent favored an increase, 45 percent favored a decrease, 27 percent supported keeping troop levels the same. This led the Post to report the results of that poll under the headline, "Public Opinion in U.S. Turns Against Afghan War."

So did the Post change the wording of the poll to get a different outcome? Or did public opinion just dramatically reverse course in two months? The latter seems implausible.