So Gallup’s final read on the 2010 elections features a generic advantage for Republicans of 15 percent, 55-40. That’s been making people shake their heads in astonishment all day. Never before on election day have Republicans even led on the generic ballot (the question Gallup asks is whether the person polled will vote for a Republican or a Democrat). In 1994, the best midterm for Republicans in our time, the final Gallup tally had the two parties tied.
This is why people are saying something is happening here that has never happened before. The “poll of polls” at Real Clear Politics, which averages out all reputable surveys, has the Republican advantage tonight at 8.7 points. Which means even if you think Gallup is screwy, there’s still no way to avoid the conclusion that Democrats are in for a horrific day tomorrow.
But wait. There’s more. The Gallup number today is 55-40 assuming a voter turnout of 45 percent nationally. It is assumed that the higher the turnout, the better the number is for Democrats owing to the Democratic edge in the number of registered voters. Fine. 45 percent. Except that the midterm in which more voters participated than any other in the past 28 years was 1994 — and in that year, turnout was 41.1 percent. This year, a voting expert named Michael McDonald thinks the number could be a record-breaking 41.3 percent.
Think this through. An amazing number for turnout would be around 41 percent. Gallup is using a model predicting 45 percent turnout — that’s a differential of 10 percentage points. On other words, Gallup might be wildly overstating the size of tomorrow’s electorate. And what does this mean? It means that the Republican advantage of 15 points might be low. Might be very low. That the actual Republican advantage might be closer to 20 points.
The low end prediction by Gallup of the number of House seats Democrats will lose at a 45 percent turnout? 80 seats. (The best Democrats can hope for, according to Gallup, is 55.) But what if the turnout model is off significantly, as is likely? Could the Democrats actually be on track to lose 90 seats or more? Could the best they can hope for be a loss of 70? (Sean Trende, the impressive number-cruncher at Real Clear Politics, says the Gallup number translates into a Democratic loss of 98 seats.)
The problem with these percentage guesses is that the Republican advantage is not evenly distributed across the country; it might be close to 30 percent in the Southwest but only a point or two in the Northeast. Republicans can’t win many more than 90 seats because they don’t even have a sufficient number of candidates to do so.
But — and this is the big but — numbers this large, should they hold, presage doom for Democrats in the Senate. A wave this large is unlikely to tilt any close race into Democratic hands. And it might mean a shocking Republican victory in a Senate race no one has even paid attention to (Oregon? Vermont?)
Meanwhile, the story that has barely been told over the past 20 years is this: American elections have become the greatest public dramas I can think of. Clinton and Perot and Bush in 1992. The Republican Revolution of 1994. The 36 Days of Florida in 2000. The Bush-Kerry seesaw in 2004. The Democratic surge in 2006. The Year of Obama, guest-starring the surprise rookie phenom Sarah Palin, in 2008. And now this. Anybody who thinks he knows what 2012 is going to look like is living in a fantasy world. Reality is much too twisty for us to have any sense where all this will go after tomorrow night.