With just ten days to go, it’s an open question whether Mitt Romney can make a comeback in Ohio. As Nate Silver noted today, the polls are getting accurate at this late stage, and the odds strongly favor Obama maintaining a lead in Ohio polls heading into Election Day. That might be especially true this year, since the polls have been remarkably consistent, Obama has already banked a lead among early voters, and it's hard to imagine advertisements making a late difference after months of airing beyond saturation levels. If Obama still leads by 2 or 3 points in Ohio in ten days, the risk of a Romney victory would hinge on the possibility of a systemic error in the polls. It happens, but not especially often.
Still, while the odds are against a late Romney turnaround, it's not impossible. It’s worth recalling the Ohio polls from this time in 2004, which showed Kerry tied or slightly ahead. Starting with the polls conducted over these very dates eight years ago, Bush took a slight and consistent lead in Ohio and never relented. From this point onward, Kerry only led in one poll of Ohio (Gallup, of course), where the RealClearPolitics average showed Bush holding a 2.1 point lead on Election Day. A similar lurch toward the right would bring about a tied race in the Buckeye State in 2
Whether the movement in the polls was attributable to a genuine shift in the race or something else is impossible to say. But 2004 reminds us that late movement in the numbers is still possible, even if it's generally unlikely and even in the most pivotal state.
Still, it's unlikely that Romney will go into Election Day as a favorite in Ohio. If Ohio lurches two points toward Romney, that would only indicate a tied race and Obama would still have a 50-50 shot.
As a result, Obama is a clear favorite. If the polls stay where they are, which is the likeliest scenario, Obama would be a heavy favorite on Election Day, with Romney's odds reduced to the risk of systemic polling failure. Harry Reid can tell you all about the risk of systemic polling failure, but it's still an unlikely scenario. There's also the chance that the race tightens, but, even then, Obama would still have a 50-50 shot. Viewed collectively, Obama's odds look pretty good--but there's still ten days to go and '04 reminds us that those ten days can occassionally make a difference.