House and Senate Forecast Updates



I have now made new updates to both the Senate Forecast and the House Forecast. I am making these updates today for multiple reasons. First, looking at the whole situation at once helps calm me down. Second, so much information is coming in, it is forcing me to update several races every day. Third, I have been concerned in a lack of consistency in my forecast method, and I wanted to make sure it was all ironed out. Fortunately, by now I think I have enough improvements to be able to leave the forecasts alone until Monday evening, when I intend to make my final forecasts.

In addition to the changes I announced last night, here are the new changes to the House forecast:
  • CT-05 upgraded from "toss-up" to "lean Democratic"
  • IN-09 upgraded from "toss-up" to "lean Democratic"
  • TX-22 downgraded from "lean Democratic" to "toss-up"
  • KS-02 upgraded from "likely Republican" to "lean Republican"
  • WY-AL upgraded from "likely Republican" to "lean Republican"
The House forecast now shows a Democratic gain of 23-29 seats, giving me a wider range on the exact 26-seat pickup I went with last night. If nothing else, at least right now I feel those rankings are consistent with each other. They may seem conservative, but they are my best guess.

Now, onto the changes in the Senate forecast:
  • Michigan downgraded from "non-competitive" to "likely Democratic"
  • Washington downgraded from "non-competitive" to "likely Democratic"
  • Montana downgraded from "lean Democratic" to "toss-up: projected Democratic"
  • Missouri upgraded from "toss-up: projected Republican" to "toss-up: projected Democratic"
  • Tennessee downgraded from "toss-up: projected Republican" to "lean Republican"
  • Arizona upgraded from "likely Republican" to "lean Republican."
  • Arizona moves ahead of Tennessee in the overall rankings to our seventh best target
  • New Jersey moves ahead of Maryland to become the second best Republican target.
Once again, I feel these rankings are now consistent with each other. The overall Senate projection, however, is notably inconsistent with the seat-by-seat projection. While I forecast Democratic pickups in Pennsylvania, Ohio and Rhode Island, and while I give Democrats the edge in Montana, Virginia and Missouri, I find the latter three races so close that I do not project we will win all of them. In other words, looking at each seat in a vacuum, I see six pickups for Democrats. However, looking from a wider lens, I see the Democratic edges in Montana, Virginia and Missouri all being so slight, that it seems more likely to me than not that we will lose one of those races. However, I just can't pick which of the three we will lose. This is why I see a five-seat gain for Democrats in the Senate, despite seeing Democrats with an edge in six pickup opportunities.

Is that a cop-out? Maybe. Am I too gutless to ever pick Democrats to win the Senate this cycle? Maybe. However, I would much rather be wrong and have Democrats take the Senate, than be wrong and have Republicans keep the Senate. The margin between Democrats taking the Senate and Republicans keeping the Senate seems so small right now, that even if I am wrong, I shouldn't be wrong by all that much. Right now, this si the best I can do when it comes to guessing. Now, it is time to spend the weekend getting active.

Update: So much information coming in at once. Look at the new Survey USA polls. I clearly jumped the gun on IN-09, but I was clearly right about CO-04. Hard to keep track of everything. OH--and the poll showing a 47-47 tie in Maryland? It has the green party canddiate getting 3%. Yeah, that will happen. Still, I also feel good about bumping that race ahead of New Jersey. Menedez will win. Cardin--I'm not so sure yet.

Tags: election forecasts, House 2006, Senate 2006 (all tags)

Comments

40 Comments

Re: House and Senate Forecast Updates
Why has everyone seemed to have moved TX-22 back into the toss up category? Do the Republicans really have a good shot at holding this one?
by blueryan 2006-11-03 11:02AM | 0 recs
We'll get it.

Enough voters will do what I did when I was 18. Vote for everyone unopposed. No way they get enough write-ins.

by Cleveland John 2006-11-03 11:09AM | 0 recs
Hilarious DCCC mailer: TX-22

Doctor Biobrain describes an hysterical DCCC mailout in TX-22.  Basically it tells voters that Don Richardson (not shelly whatshername) is "too conservative for Texas" and voters should not vote for him.

They play it straightfaced, because they're banking on idiot conservatives saying "Damn the Democrats, I will vote write-in for Don Richardson because they told me not to!"

It's tricky but completely honest.  The DCCC isn't saying that Don Richardson is the GOP's write in candidate, but just that in their opinion he's too conservative for TX-22.  

I will really laugh hard if even 100 people vote write in for "Don Richardson" after that.

Kudos to the DCCC for a brilliant manoevre there anyway.

by scientician 2006-11-03 12:41PM | 0 recs
It's a fair tactic

Don Richardson is running as a write-in Republican candidate in TX-22.  And he definately is too conservative for Texas.  And his name is so much easier to remember...

I don't see how Republicans hold this one.  Most people will get their ballots, mark Shelley Gibbs on the special election, than see the general election and vote for either Lampson or the Libertarian candidate.

Prediction: Lampson 45%, (Libertarian) 15%, write-in Gibbs 35%, other write-ins 5%.

The real test will be holding this seat in 2008, but if anyone can hold it, it's Lampson.

by Skaje 2006-11-03 01:07PM | 0 recs
Re: Hilarious DCCC mailer: TX-22

That's awesome. I love a political tactic that smacks of folks having fun, and I'm sure the DCCC folks had a great time putting that together.

by BriVT 2006-11-03 02:16PM | 0 recs
Weak Chris, you called it .. a cop-out

Yes Santa Claus, we have Virginia.

Latest from Rasmussen ... had Tester up by 4.

Michael J Fox gave us Missouri.

Stow some guts :)  "Today" the Senate is ours.

by Cleveland John 2006-11-03 11:07AM | 0 recs
Re: House and Senate Forecast Updates

Yes I am curious why TX-22 went into the toss up category.  I know its a deeply Republican District, but the Democrat is Nick Lampson who has a ton of money, and the Republican is a write in with a hyphenated last name.

I guess my only concern is that determining the intent of the voter is allowed in Texas, so mispelled attempts or vague references to the Republican (can't say a write in candidates name!) could be interpreted by partisan vote counters as a vote for Republican.

by agpc 2006-11-03 11:07AM | 0 recs
Re: House and Senate Forecast Updates

"Is that a cop-out?"

I don't think so.  Your reasoning is consistent with standard probability theory.  For example, assume there are three races, in each of which you think the Democrats have a %70 chance of winning.  The probability that they will all win is:
.7*.7*.7 = .343
The probability that exactly two of them will win is:
3 * (.7*.7*.3) = .441.
So, in this case, it is more likely that only two will win.

by JamesRoy 2006-11-03 11:08AM | 0 recs
Re: House and Senate Forecast Updates

I've been wondering if how accessable AZ really is. Last I looked Kyl had an 8 point lead. Isn't that a lot to over come in three or so days? I know Kos and you are ranking it about TN.

by mrdavedog 2006-11-03 11:11AM | 0 recs
Re: House and Senate Forecast Updates

Great news in AZ. I think the DSCC ads start tomorrow ....that can help pull this over the top.  

by misd1925 2006-11-03 11:14AM | 0 recs
Re: House and Senate Forecast Updates

What I like about it, is that even the DSCC gets to expand the playing field this late in the election, keeping the NRSC guessing.  

That we're still on offence at this point is simply awesome.  After Foley, I really thought we'd be defending a slowly declining lead right upto election day.  

The only fear I have left is Saddam's trial on the 5th.  That the news will break on a sunday might help though.  Especially if it's sunny and people stay out doors.

by scientician 2006-11-03 12:43PM | 0 recs
Re: House and Senate Forecast Updates
The verdict, and we all know it will be "guilty",
is only going to serve to remind Americans that the
cost of overthrowing Saddam wasn't worth the cost in lives, money and prestige lost at all.
by phillydem 2006-11-04 12:09AM | 0 recs
Re: House and Senate Forecast Updates

DSCC ads were on TV tonight (ties Kyl to Bush) and there was an ad for Anne Simon (v. Renzi) in AZ-01  that highlighted all his corruption.

by Mimikatz 2006-11-03 05:39PM | 0 recs
Couple hiccups.

A new Survey USA polls has Cardin/Steele at 47% a piece.

And, apparently, the GOP internal tracking polls show Chafee and Whitehouse tied. They also, supposedly, claim Talent is gaining a bit of momentum and that Jim Webb is starting to pull away in Virginia.

The Survey  USA poll, however, is based upon a likely voter model whcih inludes a hgher African American voter turnout, and that 1/3 of them voting for Steele. So, Bill Clinton and Barack Obama need o be out campaigning with Cardin very visibly to make sure this one doesn't sneak up on us.  In fact, Obama is in Maryland today.

by Hesiod Theogeny 2006-11-03 11:16AM | 0 recs
Re: Couple hiccups.
Survey USA has ben a consistent outlier in the Mayrland race. And when it comes to two day, itneral GOP polls leaked to the National Review well, um, whatever.
B<R> They key is to always look at the overall polling picture. I think my rankings caputre those. I have both MAyrland and Rhose Isalnd as "lean Dem." I think that is where they belong.
by Chris Bowers 2006-11-03 11:44AM | 0 recs
Re: Couple hiccups.

I understand. But I think Survey USA's methodology for their likely voter screen is not implausible. It should defintely be something to take seriously.

by Hesiod Theogeny 2006-11-03 11:49AM | 0 recs
Re: Couple hiccups.
I don't think it is implausible. If I did, the race would be in "likely Dem." I just don't think it is likely.

And in Maryland, I would say that even a tied race wold probbaly go to Cardin by three or four.
by Chris Bowers 2006-11-03 11:58AM | 0 recs
Re: Couple hiccups.

I think Survey USA is usually pretty accurate, but I think they're wrong on Maryland. every other poll (Rasmussen, Baltimore Sun, Washington Post, Zogby/Reuters) shows Cardin up by 5-7 points. I think they're wildly overestimating the amount of black support Steele will get.

by johnny longtorso 2006-11-03 01:46PM | 0 recs
Survey USA

Worth noting:

"Though most other pollsters have found Cardin with a slight but shrinking lead, this is the third straight month SurveyUSA has found the race deadlocked." mydd

by Cleveland John 2006-11-03 04:38PM | 0 recs
Re: House and Senate Forecast Updates

On wednesday we may have to re-evaluate how we handicap races but I agree that the smell of the past, especially the GOPs ability to pull it out when it counts is hard to overcome.

by aiko 2006-11-03 11:17AM | 0 recs
Re: House and Senate Forecast Updates

What is going on in Montana?

by MNPundit 2006-11-03 11:33AM | 0 recs
Re: House and Senate Forecast Updates

Last poll out of AZ has Kyl up by only 4%.

It's moving our way.  If this is a wave, he's falling.  I might say this one is more likely than Missouri, and I think definitely more likely than Tennesee now.  
I agree that 2/3 is more likely, with Tester and Webb winning, but I'm not seeing Missouri.  I think McCaskill has a good chance, but I think she'll narrowly lose.
If you're in Missouri, Virginia, Montana, Arizona, Tennesee or near Rhode Island, spend the weekend working for the campaigns.  Get us a Democratic Senate.

by jallen 2006-11-03 11:35AM | 0 recs
Re: House and Senate Forecast Updates

Imagine taking both MO and AZ.  Kyl got listed as one of the "best" senators by Time not so long ago.  Evidently he is far more influential than most of us noticed.

by scientician 2006-11-03 12:46PM | 0 recs
Re: House and Senate Forecast Updates

I get the personal osychology part - you don't want to set yourself, or all of your Loyal Readers up for a disappointment. But this is politics, man!

I know picking the DEMs to only win one or maybe two of the really close races seems to make sense and is the seemingly cautious way to go, but I think iit's worng.

The trend is my friend and the DEMs will most likely take all three or none of the hot Senate seats.

by dougwood57 2006-11-03 11:43AM | 0 recs
Re: Senate races don't break evenly

IIRC, it's been shown that in wave or momentum elections, Senate seats don't break evenly, but fall disproportionately to one party or the other. If this holds, true, then the competitive seats will almost all break for the Dems.

by phillydem 2006-11-03 12:03PM | 0 recs
I also believe

that in recent history, the Senate has always changed power when the House does.  The thing to remember is that in a wave election, each race should shift a few percent one way.  All we have to do is keep it around even in the polls, and the wave can push is over.  So Ford is clearly not finished in Tennessee.  Tester, Webb, and McCaskill should be considered favorites at this point.  And Pederson has gone from long-shot to serious contender.

As for Connecticut?  If Democratic turnout dwarfs Republican turnout, Lieberman is toast.  The pollsters are using models that may not even apply in a year like this.

by Skaje 2006-11-03 01:13PM | 0 recs
Re: House and Senate Forecast Updates

"The trend is my friend and the DEMs will most likely take all three or none of the hot Senate seats."

If the probabilities of winning each race are independent, then its ok to just multiply the probabilities of winning each race to get the overall probability.  In this way, you can get high probabilities for splitting the results (ie, Dems winning 2 out of 3 close races).

However, if there is an unknown bias for or against Democrats, then this comment is correct, and you are more likely to either win or lose all the close races.

by JamesRoy 2006-11-03 12:04PM | 0 recs
Nebraska races.

   Maxine Moul and Scott Kleeb have realsitic shots at victory.  I'm surprised by all these chances in deep red states.  We seem to be highly succesful in all states, except for Nevada.  We should be doing better in Nevada.    

by cilerder86 2006-11-03 12:21PM | 0 recs
Re: Nebraska races.

Carter's health problems maybe?

Or is there residual disdain for his father still hurting him?

by scientician 2006-11-03 12:47PM | 0 recs
Not the Senate race so much.

   The House races and governor race.  Two polls came out today showing Hafen and Derby down in the high single digits.  Titus still trails despite the Gibbons scandal.  I'm not sure if Titus will win - the last polls showed her negatives at 40!  We should be doing better in the Senate race, but I never expected to win it.  Reid has just not lifted a finger to help the Nevada candidates.

by cilerder86 2006-11-03 12:56PM | 0 recs
FORD'S take

I watched MSNBC's Coverage with Russert interviewing  Ford. Ford claimed that Dems are turning out to vote at a higher percent(14%) in early voting.  He also claimed that the DNC's internalls have them up a few in the state.  What do we make of this?

by nzubechukwu 2006-11-03 12:36PM | 0 recs
Chart Prints Great in &quot;Landscape&quot; Format

This may be obvious, but the chart looks great printed in landscape format and makes a great scorecard.

by tonymoco 2006-11-03 12:42PM | 0 recs
Sorry, Chris.

   You copped out on the Senate forecast.  You're predictions race by race show the Democrats taking the Senate 51-49.  But you left the overall prediction at 50-50.  Come on!  Stop showing your conservative bias.

by cilerder86 2006-11-03 01:04PM | 0 recs
Re: House and Senate Forecast Updates

Polling has us up 36-28. THat's pretty good IMO.
7% was going to other write ins.

Oh well, I'll enjoy my 1 term democratic representation in the House while it lasts. Whoever has the seat in 2008 certainly won't be as bad a Tommy DeLay.

by Trowaman 2006-11-03 01:41PM | 0 recs
Re: House and Senate Forecast Updates

Can we have a printable copy of the forecast Monday so that we can print it out and carry it with us?

I want mine Tuesday so that after a long day of GOTV I can relax at the local Democratic victory party and check off the dominoes as they fall across the country.

by ChetEdModerate 2006-11-03 02:10PM | 0 recs
Not A Copout

Three 52% probabilities make an expected average total of just over 1 1/2.  Your prediction of 3 in individuals but 2 in aggregate is exactly the right thing to do.

That was the key insight in Sam Wang's election site at Princeton in 2004.

by Professor Foland 2006-11-03 05:18PM | 0 recs
Re: NJ

Menendez has a great spot running now pretty often on TV where he walks along a boardwalk and talks about how bad Bush and the GOP have been for the country. Menendez consistently opposed Bush and the war from the beginning and it's helping him big now. I think that ad and the one from DSCC tying Kean Jr to Bush and the war is the reason Menendez is pulling ahead.

by phillydem 2006-11-04 12:14AM | 0 recs
Senate in 2006 was always a longshot

It's truly amazing that we've got any sort of chance.

Look at the arithmetic: we started the year with the Senate 55-45 GOP.  The 33 seats up for election this year are 18 D, 15 R. And in 6 of those 15 (ME, IN, MS, TX, UT, WY), the Dems are essentially not contesting the seats.  That gives the Dems only 9 Senate seats that are even remotely potential pickups: RI, PA, OH, VA, TN, MO, MT, AZ, and NV.

So to take the Senate, the Dems have to win 6 of 9 contested GOP-held seats, without losing any of their own seats.  Jack Carter never quite got Nevada into play, so that increases the challenge to winning 6 of 8.

Pulling that off would be quite extraordinary, no matter how big the wave.  For me, this year's always been about taking the House, and getting the Senate close enough so that it's put-up-or-shut-up time for GOP moderates - no more catch-and-release, thanks - before taking the Senate for real in 2008.

If we get lucky and win the Senate this year, that would be great.  And it's been worth working for and contributing money towards.  (I don't regret a dollar that I've contributed to Lamont, Tester, Webb, or McCaskill.)  But the odds have always been against us in 2006.  In all likelihood we'll set it up this year, and break through in a big way in 2008.

by RT 2006-11-04 02:04AM | 0 recs
Re: Senate in 2006 was always a longshot

In fact, it might help our chances for the presidency in 2008 if we don't get the senate now.  Iraq is a mess.  The housing bubble may burst.  More scandals to come.  And with a dem house and rep senate, nothing much will happen.  I'm nervous about the courts, of course, and would rather see a dem senate sooner than later, obviously, but there may well be a strategic advantage of not winning it this time.  

by kekavala 2006-11-04 07:44AM | 0 recs
Re: House and Senate Forecast Updates

SUSA in MD is an outlier.  The last two SUSA polls have shown tie and Steele +1, so this one actually shows no movement.  We shall see if the IVR is the magic crystal ball in this state, but it definitely disagrees with the other polls...

by NJIndependent 2006-11-04 05:03AM | 0 recs

Diaries

Advertise Blogads