Polls, if you must

Jay Cost today in The Weekly Standard should provide more than a little encouragement to conservatives who've prematurely taken to drink because of Romney's current alleged poll deficits. Cost's central point - and he is among the most prescient of poll observers -- is that most major polling organizations are oversampling Democrats, based on an assumption that the 2012 electorate will demographically match 2008's, when blacks, Hispanics and under thirty's constituted record percentages.

In discussing Ohio, for example, Cost points out that if this year's electoral demographics turn out to be a compromise between those of 2008 and 2004 -- a reasonable assumption, he thinks -- then this year's Ohio electorate would contain only a one to two point Democrat edge, which in turn would produce a razor thin outcome that would be determined by independents.  But in its current poll of Ohio showing an alleged eight point Obama lead (52-44), the Washington Post sample is based on a seven point Democratic voter turnout advantage.

Whatever one thinks of the polls in general, it is a fact that in 2008, Rasmussen's call at the end was the most accurate.  Today Rasmussen's polling organization has it 46-46 (actually, 48 Romney/46 Obama, with leaners), and the following are Rasmussen's most recently published numbers for six battleground states:

Ohio: Obama +1 (at 47%)

Florida: Obama + 1 (at 48%)

Virginia: Obama + 1 (at 49%)

Colorado: Romney +2 (Obama trailing at 45%)

Iowa: Romney + 3 (Obama trailing at 44%)

New Hampshire: Romney + 3 (Obama trailing at 45%)

For conservatives these are far from wrist-slitting numbers, and they come from the polling organization that has proven itself least partisan, most methodologically sensible and, in general, most reliable. More encouraging than the razor-thin deficits in Ohio, Florida and Virginia, and the slightly better Romney leads in Colorado, Iowa and New Hampshire, is Obama's absolute percentage in every Rasmussen poll: below the 50% threshold that an incumbent needs to reach to feel any measure of security given that the undecided or undeclared vote almost always breaks against a presidential incumbent.

If you have to watch polls, watch Rasmussen.  Better still, stay engaged, stay positive and -- especially if you live in battleground state or have friends or family who do -- keep trying to influence those who can be reached.

If you experience technical problems, please write to helpdesk@americanthinker.com