The suspense is killing me

November 4, 2006 7:32 PM

Will the Dems take the Senate and/or the House on Tuesday?

I'm interested in the conventional wisdom, as well as your crystal ball predictions. Also, where's the best place online to get the latest polling data?

posted by Optamystic to law & government (18 comments total)

Slate's Election Scorecard is nice. So many polls are in the "statistically insignificant" margins of error that there can be no reasonable predictions about the Senate. Call your friends and make sure they vote.
posted by LarryC at 7:36 PM on November 4, 2006

Electoral-vote has all the polling data.

"Our" predictions are over in the MeFi Election Prediction contest.
posted by smackfu at 7:37 PM on November 4, 2006

My preferred site for polling info is pollster.com, which also has a really nice Senate map.

It looks like the conventional wisdom is that the Democrats will take the House. The Senate is a lot harder, since they need six seats, and not very many Republican seats are up this cycle. But Democrats' chances have improved a lot in the last week or so, such that "On NBC this morning, Tim Russert said that strategists in both parties think there’s a 50/50 chance that Democrats retake the Senate."
posted by raf at 7:49 PM on November 4, 2006

Dems take the house, but it takes a perfect storm for them to take the Senate. We're likely to see a divided Congress.
posted by chrisamiller at 8:07 PM on November 4, 2006

so, short answer, House, yes most likely barring the unthinkable. Senate? WTFK? Right now it is as tied as it can be, I'd say the odds are slightly in favor of them not taking 51 seats in the Senate. But all in all it is a pretty big shift, and could end up with 2 years of nothing happening except bickering then in '08 everyone tries to blame everyone else for not doing anything.
posted by edgeways at 8:08 PM on November 4, 2006

I'm interested in ... your crystal ball predictions.

Opinion: Dems take the House by a smaller margin than is predicted in the polls. Republicans retain the Senate by a comfortable margin. Republicans proclaim a victory.

Reason: The polling often misses a larger point, which is that the hard-core right is harder core than the hard-core left. Karl Rove knows something most people don't, and that is thirty percent of Americans will vote Republican no matter what. The same cannot be said of the Dems, who are traditionally fractured and distracted (except for that guy Clinton, who apparently owned a genie and a lamp). So, the right is really playing for only 21 percent of the vote. Play to the base, add a few wedge issues, like abortion, gays and guns, and frankly, it becomes a lot easier than it looks.
posted by frogan at 9:17 PM on November 4, 2006

the woman I am being told I was making out with last night said the dems would win hands down.

and we all know women are never wrong.
posted by krautland at 9:21 PM on November 4, 2006 [1 favorite]

I think there are so many many close races that are real squeakers, it might be a few weeks before we actually know the final answer.
posted by buggzzee23 at 9:40 PM on November 4, 2006

I think the Dems take the house by a slim margin if at all. Very little chance of taking the senate, not impossible, but damn close.
posted by !Jim at 9:57 PM on November 4, 2006

Electoral-vote has all the polling data.

I think that was the site that got my hopes up before the last presidential election, and then Bush won -- bleh -- so I don't know if I'd trust it. They seem to somehow work too much Democratic optimism into the counts. Other sources are not predicting a Democratic Senate.

My crystal ball says Americans will always vote for the best personal financial deal -- "What's in it for me?" -- which they still irrationally believe (because they were raised that way) requires a vote for the Republicans.
posted by pracowity at 12:32 AM on November 5, 2006

The Washington Post just released their 13th Biennial Outlook Crystal Ball Contest with Republican, Democratic, and independent predictions.
posted by armage at 5:25 AM on November 5, 2006

The betting markets are currently offering about 20% odds that Republicans will keep the House, and 70% odds they'll keep the Senate.
posted by mbrubeck at 5:41 AM on November 5, 2006

The polling often misses a larger point, which is that the hard-core right is harder core than the hard-core left.

I think that's not necessarily true this year, though. The Democratic base is very energized (read: pissed-off), and portions of the Republican base are turned off because they perceive the administration either as crazy or (more often) not crazy enough.

The thing to remember is that the polls don't lie. There's margin of error, yes, but usually when there's a "surprise" election result, it's because the conventional wisdom was ignoring what the polls were saying.

In this case, I think the predictions are probably about right. Dems take the House and drive the Republicans to a margin of one or two seats in the Senate. I'll be delighted if they take both houses, but I don't think it'll happen.
posted by EarBucket at 6:34 AM on November 5, 2006

Polls don't lie, but there is a human factor involved that can cause incorrect results. Each pollster uses something called a "likely voter model" to improve accuracy. They take a bunch of the answers and try to figure out whether this person is going to actually vote. If not, they basically throw out their answer. If the rules they come up with don't match the eventual reality, then their results were wrong. More info here.
posted by smackfu at 9:44 AM on November 5, 2006

Political Wire: Cook's Last Forecast
posted by gimonca at 9:47 AM on November 5, 2006

The electoral-vote.com map this morning says 51 D, 49 R in the Senate. That's based on an assumption that Democrats win the trifecta in VA, MO and MT, all three of which are too close to call at the moment.

And that Lieberman in CT caucuses as a Democrat.

The reason you heard more predictions about a possible Democratic Senate a couple of weeks ago is because Harold Ford had a tentative lead in Tennessee. That lead has evaporated after a very, very negative campaign barrage by the Republicans.

Tester in Montana (D) has had a small lead all season, that race has tightened to neck and neck.

George 'Macaca' Allen in Virginia (R) was leading Jim Webb (D) for most of the year, that race has also tightened into a very slight possible D lead. Still too close to call.

The Missouri race has been flipflopping since the beginning. Very volatile.

raf mentioned pollster.com, which has nice illustrations where you can follow the trend lines in each race: graphs of how the "last 5 poll averages" have changed over time. This helps to smooth out differences in polling practices, outliers, partisan leanings in polls, etc.
posted by gimonca at 10:14 AM on November 5, 2006

Pollster, which is run by careful professionals, seems more reliable to me than electoral-vote or political wire. But, to be honest, I obsessively check all three.

A good cure for the suspense is to volunteer for a campaign!
posted by escabeche at 11:16 AM on November 5, 2006

I like pollingreport.com. It is straight forward. Nothing fancy, but it is a quick overview of all kinds of polls.
posted by psergio at 5:41 PM on November 5, 2006