The difficulty in blogging today is that (a) the only thing worth blogging about is tomorrow's election but (b) the only clear fact that emerges from a thorough review of all of the polls and pundits is that nobody actually knows anything. Most pundit predictions are based (at least) as much on desire as they are on data. The only course of action that makes sense is to assume nothing and do everything possible to ensure the right outcome tomorrow.
Having said that, there are a few facts worth noting:
(1) If there is to be a deep Democratic sweep, the 1994 mid-term election which swept incumbent Democrats out of office and subjected the country to a Newt Gingrich-led Congress is the likely model. National Journal's Hotline dug through its archives and posted the "generic ballot" polling results for the weekend immediately prior to the 1994 election:
ABC: --- 47-46 in favor of the Dems (a 6-point swing in the last week toward the Dems)
Gallup: --- 51-44 for the GOP (a 4 point swing in the last week toward the Dems)
NBC: --- 46-35 for the GOP (a 2 point swing in the two weeks toward the Dems)
Times Mirror: --- 48-43 for the GOP (a 7 point swing in the last month toward the Dems)
Two points to note about that: (a) the party that was swept away was the beneficiary of late swings in each of the polls and (b) the polling advantage which Democrats have over Republicans for this election is substantially greater than the one Republicans had over Democrats at the same point in 1994. But one countervailing fact should also be noted -- the incumbent-protection provisions installed by Republicans over the last ten years likely means that Democrats, in order to achieve a 1994-type sweep, will require a larger voting advantage than the Republicans had in 1994.
(2) Although it is conventional wisdom that the Democratic defeat of 1994 was due to the intense unpopularity of President Clinton, his approval ratings were substantially better, even in 1994, than Bush's are now. As Fox pointed out with regard to its newly released poll today: "According to 1994 midterm election exit poll results, at that time 44 percent of voters said they approved of the job President Clinton was doing and 52 percent disapproved." By contrast, Fox's new poll shows that Bush is a deeply unpopular President, by a 38-54 margin (CNN's new poll shows Bush's approval rating even lower, at a humiliating 35%).
(3) To defend their Leader, Republicans are claiming that it is normal and natural for a President to suffer low approval ratings after being in office for six years. But in President Clinton's sixth year in office -- 1998 -- he was a highly popular President, even though he was besieged by the right-wing and national media's filth-mongering impeachment proceedings. Put another way, when one compares Bush and Clinton at the same points in their presidency, they are on opposite ends of the popularity spectrum. From Fox: "Six years into Clinton’s presidency, the same point as Bush is today, 55 percent of voters approved and 43 percent disapproved."
(4) While a couple of polls released yesterday reflect what has become the media cliche of the day -- "tightening" -- the trend lines for the individuals races, with few exceptions, reflect no such thing. The polling data for the key Senate races have long shown -- and continue to show -- huge Democratic leads in Ohio and Pennsylvania; marginal though consistent leads in Montana, Rhode Island, Maryland and New Jersey; and tiny, precarious leads in Virginia and Missouri. Not much has changed in that regard for some time. If those leads hold and there are no major upsets either way, Democrats (even if they lose Tennessee) will take over the Senate (a result that is far, far from certain, or even likely, but it is definitely possible, and more to the point, nothing that has occurred over the past few days has changed that fact either way).
(5) Markos' predictions are here, for what they're worth (and he wisely emphasizes: "I've been way off on my predictions the past two cycles. So there's no particular wisdom or inside knowledge or track record to give these predictions any more weight than anyone else's. They're just for fun"). He predicts gains of 24-36 seats for Democrats in the House (15 are needed for a majority) and a gain of 6 seats in the Senate (exactly the number needed for a majority). Consistent with that, National Review's Editor, Rich Lowry, says that Republicans consider 12-13 GOP House seats lost for certain, and that they'd "be quite pleased if the GOP holds losses to 22 [House] seats tomorrow night."
(6) Nobody actually knows anything about the results tomorrow.