With one day to go until Election Day, you can add a new USA Today/Gallup poll to those showing a rather dramatic tightening in voters' preferences between generic Democratic and Republican candidates. A month ago, Democrats held a 23-point advantage in the poll; today, it's down to seven points.
As Walter Shapiro notes today, the Democrats' advantage dropped sharply in two polls released over the weekend: It's down from 14 points to six points in the latest Washington Post/ABC News poll and from 11 points to four points in a new poll from the Pew Research Center for the People and the Press.
Is this a cause for concern? Of course it is. DCCC chairman Rahm Emanuel tells the New York Times that some tightening in the polls is inevitable, but he admits: "This is making me nervous."
Even if these polls didn't exist, it would be hard not to be nervous just now. Democrats have seen what seemed like victory slip through -- or be taken from -- their fingers before. Al Gore won the popular vote in 2000 but never became president. John Kerry's team was pretty certain that he'd won in 2004, at least until the actual election results were announced. Could it be happening all over again?
It could be, and maybe it already is. Now, it's important to remember that we don't hold national elections in the United States, even when we're electing a president and especially when we're electing senators and members of Congress. But that fact cuts both ways when it comes to thinking about these generic ballot matchup polls.
The positive spin for Democrats: Changes in national generic poll results tell us very little about who will win in any particular congressional race; Rick Santorum isn't going to win in Pennsylvania no matter what voters around the country are saying about hypothetical candidates they may or may not prefer on Election Day.
The more frightening reality for Democrats: A big lead in the generic ballot polling -- or even in the national votes cast -- doesn't guarantee a win in any particular state or district. Indeed, Democrats need a lead in the generic polling even to stay close with Republicans on Election Day. Case in point: If you add up all the votes cast for the people serving in the Senate and the people who ran against them last time around, you'd come up with a number that favors Democrats slightly, yet the Republicans hold a 55-45 seat advantage in the body today. Blame the Constitution's small-state bias. California's 34 million residents get two senators; so do the 2.7 million residents of Kansas. Concentrations of Democratic voters in urban areas lead to similar problems for Democrats in the House; it really doesn't matter if the residents of New York City really, really, really don't like George W. Bush as long as a slim majority of folks in Idaho and Wyoming do.
All that being said, we should hang on to some sense of perspective here. As Charles Franklin notes at Pollster.com, the Pew, Post/ABC and USA Today/Gallup polls are just three of the six that have measured generic ballot preferences over the last few days. While those three all see a pretty dramatic tightening, the other three do not. Newsweek puts the Democrats' advantage at 16 points. The latest Time poll has the Democrats' advantage holding steady at 15 points. And the latest CBS/New York Times poll, taken just a bit earlier than the others, put the Democrats' advantage at 19 points, an increase over the poll's previous numbers. So what does that tell us? Factoring in all of the polls, Franklin moves his "trend line" down from a 15-point Democratic advantage to an 11-point Democratic advantage. That's still a bigger advantage than Democrats have enjoyed this close to an election since at least at least 1992.