Deep breathing over Gallup

By Tim Grieve
Published October 18, 2004 2:15PM (EDT)

The headlines are alarming for Kerry supporters: The new CNN/USAToday/Gallup Poll has Bush up by eight points among likely voters, erasing every inch of the gain Kerry made in the three debates. There's no doubt that the Kerry folks would rather have Bush's Gallup numbers than their own, and not just because polling drives headlines and headlines can shape momentum. But before anybody has a heart attack, here are a few things to remember.

Item one: At just about this time four years ago, Gallup said Bush had a 13-point lead over Al Gore. Say what you will about Florida and the flaws in the Electoral College, Al Gore won the popular vote. If Gallup's numbers were right then, Gore somehow erased a 13-point deficit in two weeks with a closing campaign that few folks thought was impressive. If Gallup's numbers are right now, Kerry has only an eight-point gap to overcome.

Item two: Gallup's numbers may well be wrong. As USAToday acknowledged last month, "Gallup's recent polls have consistently shown Bush further ahead than he is in other surveys." We haven't seen the internals on Gallup's poll yet, but we're betting they'll show an over-sampling of Republican voters. Gallup's projections typically assume that more Republicans than Democrats will turn out to vote. That's the opposite of what has happened in recent elections in what we like to call the "reality-based" world.

Item three: The race isn't won based on national results. Just ask President Gore. Most polls suggest that Kerry has the edge in the battleground states at the moment. Florida is tied, but Pennsylvania is pretty firmly in the Kerry column, and Ohio -- where the loss of jobs isn't a "myth" -- is looking a little blue these days. If Kerry can take both Pennsylvania and Ohio, Bush's national numbers may ultimately mean nothing.

Item four: While the national polls, taken together, certainly suggest that Kerry is behind, Kerry doesn't need to lead going into election day. As the Los Angeles Times explains today, an incumbent's numbers don't improve on election day; a challenger's numbers frequently do. If undecided voters swing toward Kerry at the end, Kerry could trail Bush by a few points in the final pre-election polls and still win the national vote on Nov. 2. Thus, the number to watch isn't necessarily the spread between the candidates. It's Bush's number. If Bush is polling at under 50 percent, it's going to be hard for him to win. If Bush is polling over 50 on the weekend before the election -- as he is now in Gallup's likely voter poll, but pretty much nowhere else -- it's going to be hard for Kerry to beat him.

Tim Grieve

Tim Grieve is a senior writer and the author of Salon's War Room blog.

MORE FROM Tim Grieve

Related Topics ------------------------------------------

War Room