Pre-Election House Polls In 2006 and 2004

I will be spending the afternoon canvassing, but since I had a couple hours in the morning, I thought I would compare the district-level polling situation in 2004 to the district level polling situation in 2006. Right now, looking only at the most recent, independent, district-level polling, Electoral-Vote.com shows Democrats making 37 pickups in the House, and Republicans making none:
Projected New House*: 240 Democrats, 195 Republicans
* Where no independent polls exist, the 2004 election results have been used. See complete House polls.

Dem pickups (37): AZ-05, AZ-08, CA-11, CO-07, CT-02, CT-04, CT-05, FL-13, FL-16, FL-22, IA-01, ID-01, IL-06, IL-10, IN-02, IN-08, KY-03, KY-04, NC-08, NC-11, NH-02, NM-01, NY-19, NY-20, NY-24, NY-25, NY-29, OH-01, OH-02, OH-15, OH-18, PA-06, PA-07, PA-08, PA-10, TX-22, WI-08

GOP pickups (0): none
For good measure, let's add in the districts where the incumbent party still leads, but the challenger is within ten points in the most recent, independent poll:
Dem close (23): AZ-01, CA-04, CA-50, CO-04, CO-05, FL-24, IN-09, IA-02, MN-01, MN-06, NH-01, NJ-07, NV-02, NV-03, NY-03, NY-26, OH-12, PA-04, VA-02, VA-10, WA-05, WA-08, WY-AL

GOP close (5): GA-12, IL-08, IN-07, VT-AL, WV-01
So, according to the most recent, independent poll from individual House districts, Democrats are ahead in 37 Republican-held districts, and close in 23 Republican-held districts. By contrast, Republicans are ahead in zero Democratic-held districts, and close in five Democratic-held districts. There are a few races, such as GA-08, KS-02, and KY-02, where there are no independent polls.

Now, let's compare this situation to districts level polls in 2004. Further, since there have been far more polls in 2006 than there were in 2004, when looking at 2004 let's use both independent and internal polls. According to Polling Report, here are the races where the challenging party was ahead of the incumbent party in the final, pre-election, district-level poll:
Dems (6): CO-03, KY-06 (special election), LA-03, NY-27, NC-11, SD-AL (special election)

Reps (3):: LA-05, TX-19, TX-32
Again, according to Polling Report, here are the districts that were close:
Dems (11): CO-04, CO-07, CT-02, CT-04, IL-08, IL-11, MN-06 (tied), NE-01, NV-03, NM-01, NY-13, WA-08

Reps (6):: LA-07, MO-05, SD-AL (general election), TX-01, TX-17, UT-02
Note: when both internal and independent polls were available for the final weeks of a campaign, I used the independent poll instead.

So, here is a useful comparison chart of the pre-election polling situation in the two cycles:
  • Dems ahead, 2006: 37
  • Dems ahead, 2004: 6

  • Dems close, 2006: 23
  • Dems close, 2004: 11

  • Reps ahead, 2006: 0
  • Reps ahead, 2004: 3

  • Reps close, 2006: 5
  • Reps close, 2004: 6
The interesting thing is that eight of the nine polls showing a party switch in 2004 were correct: CO-03, KY-06 (special election), LA-03, NY-27 and SD-AL (special election) did in fact flip to Democrats, while LA-05, TX-19, and TX-32 did in fact flip to Republicans. Only NC-11, where an internal poll put Patsy Keever up by one point two weeks out, was incorrect. Further, many of the other pickups for one party or the other can also be found in the "close" districts: IL-08, LA-07, and TX-01. In fact, the only pickups that were not listed here came in GA-12, IN-09, KY-04, and TX-02. The polls for those campaigns weren't wrong, there just weren't any polls for those races released to the public. In TX-10 and TX-11, the other two Republican pickups in the 2003-2004 cycle, no polls were needed because Republicans faced minimal opposition in those seats, and the pickups were guaranteed from the start.

I am about to head out the door, but here is my point: the district level polling in 2004 was pretty good. In fact, nearly every single independent poll accurately predicted the winner, and nearly every independent poll did the same. With Democrats ahead in thirty-seven races, and close in twenty others, we should expect Democrats to win the House. There are at least 65 close races and / or near-certain pickups right now, and 60 of them are held by Republicans. Since Democrats only need a net of 15 seats to win the House, our chances to become the majority are excellent.

More importantly, at the same time that we should expect to win, we should also expect a lot of close races. Right now, polling averages show 54 races within seven points, which means we have a slew of opportunities in many districts where we might not receive another good chance to win for four, six, or many more years. In the races with a separation of larger than seven points, Democrats currently lead 204-170. While this means we only need 26% of the close races in order to take the House by one seat, it also means we can build a majority larger than anything Republicans have received since 1946 with 54% of the close districts going our way. Considering this, we have a rare opportunity to build a long-term House majority in this election, not just to take the majority. If we take that majority, you will see a wave of Republican retirements, new maps, and Republicans changing sides, which will only strengthen that majority further. Now, I will be happy if we end up with any majority at all, but it will still be a big disappointment if we don't capitalize on this opportunity and tuck the House away for a long time to come. That will stop the agenda of the conservative movement in its tracks for years.

By the way, the odds of getting 14 or more of 54 coin-flip races to go your way are at least 99.97%. Just thought I would point that out.



Display:


A Rare Opportunity, But... (3.00 / 2)

Don't say it won't come again soon.  One thing that a solid House majority can do is give us a platform for setting out an agenda that can help turn still more seats.  (As I never tire of saying, this is just what happens when a true presidential realigning election occurs.)

Yes, I know there's been a lot of gerrymandering that will make this more difficult than it was in the past.  But, OTOH, the GOP's "majority" is the most artificial and contradiction-riddled of any we've ever seen.  A standard 30% seat shift in two elections--the hallmark of past Presidential realignments--may be beyond reach, but it shouldn't be needed, either, since we start from much closer to parity.

So one good reason to push hard in the last few days is not just to take advantage of this rare opportunity, but to help increase the odds for next time as well.  The CW about how all those seats will be difficult to defend is just as questionable as the CW on everything else.


by Paul Rosenberg on Sat Nov 04, 2006 at 02:00:07 PM EST

Re: A Rare Opportunity, But... (none / 0)

[quote]The CW about how all those seats will be difficult to defend is just as questionable as the CW on everything else.[/quote]

I thought that bore repeating.


by InigoMontoya on Sat Nov 04, 2006 at 02:13:42 PM EST
[ Parent ]

Re: A Rare Opportunity, But... (3.00 / 1)

I don't even under the CW about difficulty of defending a seat. Its far more difficult to obtain a seat than defend it. That's why incumbency is such a powerful tool in the box of things that one used to gain power in Congress. I don't understand how people can ignore the incumbency rule when we have seen first hand that the impact of incumbency both with Democrats when they were in charge, and now with Republicans. Frankly, I feel the real point that most making this argument are really wanting to make is that somehow the country is really "conservative" but-for all the things that went wrong for the GOP this cycle. There are way too many people on the left (and as one expect the right) who believe the GOP hype that it is the natural governing majority of this country. That's just patently false- one need look no further than one stat to show its a falsehood (and that the GOP has moved far right through slim victories), ie, in 2004, as I remember, the number of votes cast, for example, for Democrats for Senate was higher than the numbers cast for Republicans. It's simply the way our system is structured, issues of incumbency, us coming off of a bad period for Democrats nationally, etc that explains GOP dominance. But it's silly one either policy or, for that matter, capacity to win over the long term electorally to assume that the GOP can win. Their coalition indeed is much less stable than ours in a lot of ways- but thats a subject of another post.


by bruh21 on Sat Nov 04, 2006 at 02:22:18 PM EST
[ Parent ]

Narrative vs. Facts and Values (none / 0)

The idea that the country is naturally conservative, or that there's a conservative majority, etc., is simply an example of a conservative narrative that the media has adopted, just as they've adopted the narrative that they themselves are liberal.  When it comes to a battle between the narrative and facts, the facts lose.  When it comes to a battle between the narrative and values, the values lose.   The narrative is everything.

And yet, the situation in the country is so bad, that the narrative is heading directly for a thumping.

Of course, the narrative is very much a product of who's paying for it.  And, as the recent leaked memo regarding not advertising on Air America shows, corporate America is who's doing the paying, and they aren't paying for folks to ask questions about them.


by Paul Rosenberg on Sat Nov 04, 2006 at 05:49:45 PM EST
[ Parent ]

WI-08: Here's one reason KAGEN will win (none / 0)

Bartlet posted a lovely photo of John Gard (R), the Republican running for WI-08 over at Kos.
by donnatella on Sat Nov 04, 2006 at 02:38:35 PM EST
[ Parent ]

Re: A Rare Opportunity, But... (none / 0)

I agree. Just by nature of this somewhat surprisingly vast playing field this cycle, IMHO there will be many extremely strong candidates in our current Tiers 2 and 3 races who'll walk away only just losing, by under 5%, after having been outspent 2-1 or more by the R. They will all be in excellent positions in 2 years, particularly with regard to R incumbents after yet another 2 years of Bush.


by jamfan on Sat Nov 04, 2006 at 03:21:03 PM EST
[ Parent ]

Not To Mention That (none / 0)

some of those incumbents may very well retire, much like the wave of Democratic retirements after they lost power in 1994.  In which case, the Democratic "challengers" will be the most familiar names on the ballots.


by Paul Rosenberg on Sat Nov 04, 2006 at 05:52:34 PM EST
[ Parent ]

Re: Pre-Election House Polls In 2006 and 2004 (none / 0)

So...you're saying we might win?


by Lucas O'Connor on Sat Nov 04, 2006 at 02:01:23 PM EST

Re: Pre-Election House Polls In 2006 and 2004 (none / 0)

Well, Lucas, I'd say it's beginning to look like a bare possibility.  Fingers crossed.


by InigoMontoya on Sat Nov 04, 2006 at 04:13:37 PM EST
[ Parent ]

Re: Pre-Election House Polls In 2006 and 2004 (none / 0)

If the events happen as your stats suggested, then this will go a long way to addressing my chief concern, which has not been whether the Democrats would win on Tues, but whether we would achieve a governing majority. It's one thing to have a slim lead that leads to stagnation, and it's another to win by such overwhelming numbers that you can govern. My friend who works on the Hill made that comment to me a few months ago, and it stuck with me. He said ecking out a win would be better than nothing, but obtaining a governing majority is idle because you have to assume that on each vote you will have defectors etc, and to really have a cushion you need a higher number than bare majority. This is really good to hear, and it willl make my phonebanking tommorrow (at least personally although not in what I say) more enjoyable.


by bruh21 on Sat Nov 04, 2006 at 02:15:34 PM EST

Re: Pre-Election House Polls In 2006 and 2004 (none / 0)

I'm not so concerned about a narrow majority.  Republicans had 230 after the '94 election and lost seats in '96, '98 and '00 elections, but still had no trouble being effective.


by Lucas O'Connor on Sat Nov 04, 2006 at 02:28:48 PM EST
[ Parent ]

Re: Pre-Election House Polls In 2006 and 2004 (none / 0)

they have had discipline that the Dems still lack. I believe Bowers or stoller a few months back put up a nifty comparison of defections on our side on key issues versus our side. That's where the problem lies.


by bruh21 on Sat Nov 04, 2006 at 02:42:48 PM EST
[ Parent ]

Re: Pre-Election House Polls In 2006 and 2004 (none / 0)

that might be true, but we also haven't had a chance to see what being in the majority does to loyalty.  If Democratic positions are seen as more popular nationally, it might be easier to keep people in line.  We'll see, it's a process that won't produce results for a while yet, but I'm not gonna write these folks off before they even have a chance.

Immediately after they have a chance sure.


by Lucas O'Connor on Sat Nov 04, 2006 at 06:37:19 PM EST
[ Parent ]

Re: Pre-Election House Polls In 2006 and 2004 (none / 0)

I am not writing off the new congress people- its the old entrenched ones that concern me


by bruh21 on Sun Nov 05, 2006 at 11:42:43 AM EST
[ Parent ]

Re: Pre-Election House Polls In 2006 and 2004 (none / 0)

I haven't done a study, but it's been 12 years since we've had a majority...how many members have come in since then and what chance have they had to demonstrate their chops?  Granted, these aren't the folks who will be running committees, but that doesn't mean they can't make some noise.


by Lucas O'Connor on Sun Nov 05, 2006 at 12:46:27 PM EST
[ Parent ]

Re: PA-18 (none / 0)

Looks like Tim Murphy, R-I, in the 18th is self-destructing. First he's caught on camera snatching away evidence from a KDKA reporter, now he's fired the staffer who went on record saying how Murphy broke the law by having his congressional staffers work on his reelection campaign on the taxpayers dime.


by phillydem on Sat Nov 04, 2006 at 02:31:00 PM EST

Re: PA-18 (none / 0)

Too bad Chad Kluko seems to have no money at the moment.


by Lucas O'Connor on Sat Nov 04, 2006 at 02:33:07 PM EST
[ Parent ]

Faulty assumptions (none / 0)

This election is not 2004, 2002, or even 1994.  The dynamics this election cycle are dramatically different than previous ones.  Using the predictive value of opinion polls from previous elections to predict this election cannot be correct, especially the 2004 election.

The big issue polls and all the national indicators, including the huge number of Republican-held races in play are what is important this year.  They point to only one result for Tuesday, a Democratic Party takeover of both the House and the Senate.

The Dems are going to take the bulk of the independents and undecideds this election.  This is going to turn the election hugely to the Dems.  There are going to be many surprises, and many embarrassed prognosticators on Wed morning, especially those who were saying that this election is in anyway close.

Dems take 7-8 Senate seats.
Dems take a very large number of House seats.  People calling for 40, or even 50 may be correct.  Or, it may even be more.
Dems will lose zero seats in either House.

Nothing is going Republican's way.


by Longship on Sat Nov 04, 2006 at 02:54:12 PM EST

Re: Faulty assumptions (none / 0)

I'm with you, and predicting a 42 seat pickup.  It's as good a number as any, probably better.  The dem majority was sealed weeks ago with the Foley news cycle.  

Democrats are doing a great job not getting caught with dead girls, live boys, meth or gay hookers, from what I hear.  Nothing much else they could do wrong, given they haven't been in power and can't abuse power they don't have.  

The only thing you could possibly worry about at this point is massive vote fraud (the Diebold boogeyman!  boo!), but even if 75% of votes on Diebold DRE machines go to republicans, I'd think we'd still get a congressional majority.


end the occupation of Iraq
by aip on Sat Nov 04, 2006 at 05:25:18 PM EST
[ Parent ]

Re: Pre-Election House Polls In 2006 and 2004 (none / 0)

This prediction matches the Newsweek poll that just came out today. Generic poll of 53-32 overall and 54-38 by district (which seems more predictive). There's good reason to hope, folks.


by cmpnwtr on Sat Nov 04, 2006 at 03:37:04 PM EST

Dems Will Win All Three Close Races (none / 0)

I know it doesn't usually work that way, but it will this year. Don't forget TN and AZ will be fairly close, probably within 5, so they can be counted as our close ones on the losing end.


by Davidsfr on Sat Nov 04, 2006 at 04:02:18 PM EST

Re: Pre-Election House Polls In 2006 and 2004 (none / 0)


by Davidsfr on Sat Nov 04, 2006 at 04:02:27 PM EST

Re: Pre-Election House Polls In 2006 and 2004 (none / 0)

can you explain your result on 14 of 54 coin flips? it doens't look right to me.


by along on Sat Nov 04, 2006 at 04:29:23 PM EST

Re: Pre-Election House Polls In 2006 and 2004 (none / 0)

I checked it, it's right, if you don't want to use the web site use a calculator:

(54!/ (40!)(14!)) / 2^54 +
(54!/ (41!)(13!)) / 2^54 +
[...]
(54!/ (53!)(1!)) / 2^54

Use teh google for more about coin flips.


end the occupation of Iraq
by aip on Sat Nov 04, 2006 at 04:59:08 PM EST
[ Parent ]

Re: Pre-Election House Polls In 2006 and 2004 (none / 0)

Why do you have us behind in NY-26? Is it because of that Survey USA poll?

Because the scuttlebutt around here is that Reynold's own internals show him tied and Davis' have him up by 7%.

I admit that Davis hasn't run the most exciting campaign in the world, but I seriously doubt that Reynolds is ahead by 5% (according to Survey USA).

It is possible that it was an outlier (since only 440 people were actually surveyed). Also, don't forget there were THREE different polls (including a local Buffalo News poll) that showed Davis ahead by 15% or more less than a month ago. At the same time, Survey USA had Davis up by only 7% in that same period.

Davis has also had a VERY good week with Hillary and Spitzer campaigning with him and a big televised interview (with NBC affliate) where he refuted many of the lies Reynolds is spreading. Plus, a NY Daily News story which suggested that Reynolds had lied when he underplayed his involvement in the Foley scandal. The story was picked up by local papers and media.


by JackBourassa on Sat Nov 04, 2006 at 06:33:18 PM EST

Re: Pre-Election House Polls In 2006 and 2004 (none / 0)

Other districts that might be close.  Some are my own guesses, but other districts have polls that show them close but aren't on the list.

My own guesses
1.Texas 14 (Shane Sklar vs Ron Paul)
2.Texas 23 (redistricting of Henry Bonilla)
3.West Virginia 2 (Mike Callaghan vs. Shelley Moore Capito)
4.Iowa 4 (Selden Spencer vs. Tom Latham)
5.New Jersey 5 (Paul Ahronson vs. Scott Garrett)

Districts with polls showing close races
1.Kentucky 2
2.Nebraska 1
3.Nebraska 3
4.California 19
5.Alaska-at large
6.Kansas 2

(not sure why those 6 districts weren't included)

That totals up to 71 districts.


by Adam T on Sat Nov 04, 2006 at 10:02:12 PM EST

Re: Pre-Election House Polls In 2006 and 2004 (none / 0)

I think Nick Lampson is poised to take TX-22 away from Sekula-Gibbs, Bell will boot Perry, and Radnofsky will put Hutchison out of commission.

I know the Republicans will do anything to hold onto that seat, as well as the state of TX.

Anyway, these predictions come on the heels of me getting a Perry mailer that says that Chris Bell's vote against Proposition 2 (banning gay marriage) is against the Mainstream. Chris Bell says he wants a New Mainstream, and that's what TX will get (and hopefully we'll take out that gay marriage ban).


by gameshowguy2000 on Sun Nov 05, 2006 at 06:57:10 PM EST


You are not logged in.

In order to post a comment, you must be logged in. If you have a member account, please log in to comment.

If not, you can make an account right here. It's quick and free.