How Liberal Media Use Polls as Weapons

To hear the major media tell it, this year’s election is already over, with Barack Obama the landslide winner.  

Given the economic meltdown unfolding since September, the unpopularity of the Bush White House and Obama’s huge financial edge over John McCain, some such result might be expected.  Even more to the point, it’s been relentlessly foretold by poll surveys appearing in the Washington Post, New York Times, and other media outlets with the same agenda.  

Typical headlines in recent days include "Polls Point to Struggle for McCain" (Post) and "Polls Show Obama Gaining Among Bush Voters" (Times) — both items running on October 24.  Similar stories appear on virtually a daily basis, showing eight to ten point national margins for Obama and similar Obama numbers in crucial states for John McCain, such as Ohio and Virginia.  

Frequently included with these surveys are comments about the hopelessness of McCain’s position, dim prospects for the congressional GOP, and/or musings about a "transformational" Democratic landslide like the election of 1932.  And such polls, of course, are potent factors in themselves, spurring on Obama backers and demoralizing some conservative GOPers (see David Frum’s throw-in-the-towel column last weekend in the Post).  

Against that backdrop, it may come as a shock to many to learn that there are polls out there quite different from those showcased by the Post and Times.  Numerous survey groups of good repute track elections on a regular basis, and a dozen or so of these show McCain running dead even on a national basis and/or leading Obama in battleground states Ohio, Florida, North Carolina and Virginia — something nobody could ever guess from reading the Washington Post or New York Times.  

Among the more notable of these surveys was a national wrap-up from the Associated Press on October 22, headed "AP Presidential Poll: All Even in the Homestretch." Considering the hype we’ve been getting about a landslide for Obama this must have come as a jolt in certain newsrooms, and would seem to be a significant news item in its own right.  But apparently not.  Diligent searching reveals no sign of this survey in the New York Times, Washington Post and kindred liberal press outlets.  (Though stories following up — and debunking — the AP poll did appear in Newsday and the Atlanta Constitution.)  

Similar shrift has been given surveys from Gallup — not exactly an unknown in these circles — which twice in recent days (October 16 and October 28) published national polls showing a two point Obama advantage among likely voters measured by "traditional" standards — well within the margin of error. Likewise for a battleground poll (Lance Tarrance and Celinda Lake) showing a national one-point (48-47) Obama lead on October 20. Again so far as careful checking shows, none of these neck-and-neck poll data appeared in the Washington Post or New York Times.

(On October 29, two weeks after the initial Gallup story, the Post finally ran a CYA reference to these numbers, albeit somewhat indirectly in paraphrase of a McCain complaint about the landslide polling.)  

In state-level surveys, the Post this week (October 27) ran a top-of-page-one story about the campaign in Virginia, headed "Poll Gives Obama 8-Pt. Lead in Va." Unmentioned by the Post was an October 24 Mason-Dixon poll, a respected source on southern elections, that showed McCain ahead by three points in Virginia (47-44). Evidently no poll can make it as a Post news story unless it shows Obama winning by a hefty margin.  

Likewise ignored by the Post and Times are polls that show McCain winning by two points in Florida (Mason-Dixon, October 23; Strategic Vision, October 22), leading by two points in Ohio (Strategic Vision, October 22), ahead by a percentage in Missouri (Suffolk University, October 19), winning by two points in North Carolina (Rasmussen, October 24), and so on. These results, also within the margin of error, suggest a close election far different from the landslide being predicted by the Post, the Times and some of their media brethren.  

Granted the different assumptions and sampling methods used by various pollsters, there is no sure way of telling which of these surveys is right or wrong, or whether they are a bit of each.  Maybe the landslide polls are right, and maybe the photo-finish ones are more on target. But all of them ought to be reported.  In the pages of the Washington Post and New York Times there’s been space over the past month or so only for the landslide version.  

Look at enough of these survey data, and it’s apparent that, for the establishment media, the polls are a weapon selectively used to create a bandwagon psychology for Obama and spread defeatism among his opponents.  There is a term for this:  It’s called a "self-fulfilling prophecy."  To judge by the dispirited comments of many conservative pundits, apparently it’s working.  Whether this will translate into GOP voter apathy on election day there is no way of telling, but if not, it won’t be for lack of effort by the liberal press corps.