What Happens If Democrats Win By 26%?

The latest Cook-RT Strategies poll, from 10/26-10/29, begs the question (last week's numbers in parenthesis):

Generic Ballot, 1,764 Registered Voters, Moe 2.3%
Democrats: 52% (49%)
Republicans: 39% (37%)

Generic Ballot, 807 Most Likely Voters, MoE 3.5%
Democrats: 61% (57%)
Republicans: 35% (35%)

A thirteen point difference between registered voters and likely voters is extremely difficult to swallow. The CNN poll that came out today showed a five-point gap between registered voters, which Democrats ahead 54%--38%, and likely voters, with Democrats ahead 53%--42%. The two polls show very similar results among registered voters, but utterly incompatible results among likely voters.

The likely voter numbers in the Cook poll certainly show why Charlie Cook thinks a Democratic pickup of more than 35 seats are quite possible. It also helps show why the Majority Watch polls, in which RT Strategies is a partner, look so good for Democrats. Whatever likely voter screen they are using, that screen clearly indicates that Democrats and Democratic-leaning Independents than Republicans and Republican-leaning Independents.

Democrats would win at least 45-60 seats in the House if they won the national popular vote by 26%. However, since no other poll is showing a Democratic lead of over 19%, I clearly have to consider the Cook / RT Strategies poll an outlier. Don't get me wrong--I hope it is absolutely correct, which would mean that we are heading toward a massive, pro-Democratic realignment. There just isn't much supporting evidence for this poll.

One positive to polls showing 26% Democratic leads, however, is that polls showing 11% Democratic leads look not only extremely believable, but probably quite conservative in their estimates. When Republicans have to be happy about being down by 11% instead of 26%, and keeping in mind that in 1994 Republicans only won by about 8%, the scope of our opportunity becomes clear. At this point in the House, they are fighting to lose by only 6% or so, and to prevent a generational landslide. We are not going to win by 26%, but with polls showing us ahead by that amount, it is very hard not to like our chances.



Display:


Re: What Happens If Democrats Win By 26%? (none / 0)

A question for all you poll mavens: Why is there so much attention to the generic macro polls when there are actual polls of the individual races? Ultimately isn't it a race between actual candidates?


by ab initio on Tue Oct 31, 2006 at 02:05:24 PM EST

Yeah... (none / 0)

Chris Bowers focuses almost solely on generic ballots. He never ever ever says any caveats when talking about generic ballots. I hardly ever see him mention "actual polls" of "individual races". I mean... except for that major "House Forecast" thingy he's got going, and except for the almost constant updates to that forecast... Other than that, "poll mavens" such as him never seem to utilize those "actual polls" of "individual races"... Instead focusing so much attention to the "generic macro polls".

---
That was snark.
Translation:
"WHAT IN THE HELL ARE YOU TALKING ABOUT?!?"

Bowers' two posts today focus on control of the Senate, and both posts cite state-wide polls... not "generic macro polls".
Yesterday, his last post listed 40-something congressional polls...
These are not inconsistent with the vast majority of his postings prior to that.  
So I ask you again, "What in the hell are you talking about?!?!?!?"


Invest in nature
by NCDem on Tue Oct 31, 2006 at 03:27:40 PM EST
[ Parent ]

Re: What Happens If Democrats Win By 26%? (none / 0)

Because in a close race, turn out makes all the difference.  The generic poll gives an indication of turnout among the parties.  If Dems turn out 11-26% more Democratic voters than the GOP, then many close races in the polls will probably tip our way.  


http://www.imvotingrepublican.com/ McCain Sucks!
by yitbos96bb on Tue Oct 31, 2006 at 02:28:54 PM EST

Democrats Win By 26%? (none / 0)

What Cook is saying aligns with my hopeful thesis;  that Democrats in this election cycle enjoy a big advantage in the 'likely voter' column because the D's and Indpendents are so outraged by the antics of the Bushniks that they are unusually committed to voting this year.  Most everyone I know would crawl through broken glass in the middle of a snowstorm to vote on Nov. 7--  that is not true of the GOP constituency.


by global yokel on Tue Oct 31, 2006 at 02:37:29 PM EST

Generic vs Specific poll (none / 0)

Historically, the Generic poll shows Democrats about 8-11 points higher than their actual returns. I would be shocked by a margin higher than 5%.


by niq on Tue Oct 31, 2006 at 02:50:07 PM EST

That's close to where I am (none / 0)

My target has always been 52%. Anything above that would be remarkable. It depends on turnout among independents.


by Gary Kilbride on Tue Oct 31, 2006 at 07:36:34 PM EST
[ Parent ]

RV vs. LV (none / 0)

I was looking at Senate polls today and in almost every case the democratic advantage decreased when switching from registered voters to likely voters.  Why Democrats would be less likely to turn out this year is beyond me.

But, isn't that the exact opposite of what we are seeing here?  Does it suggest generic polls are bubkus?


BlueNC - Progressive NC Politics
by Robert P on Tue Oct 31, 2006 at 02:51:34 PM EST

Re: What Happens If Democrats Win By 26%? (none / 0)

But, if Democrats win by 26%, this is what the Senate map will look like.  All that wonderful blue.

blue


BlueNC - Progressive NC Politics
by Robert P on Tue Oct 31, 2006 at 02:59:25 PM EST

Re: What Happens If Democrats Win By 26%? (3.00 / 1)

Actually, niq, polls usually give Dems 4-5 points that we don't actually receive, not 8-11.
In recent years Republicans have had better peformance -the share of votes actually cast- than Democrats as compared to party affiliation.  Repblicans have been more motivated, which is what the likely voter model attempts to predict.  This is why in places that have, for example, 39% Democratic registration and 36% Republican registration, Republicans often win.  This year, it looks like the opposite may occur.  In addition, a majority of Independents say they will vote Democratic this year, and in many races it appears that the Democrat will receive more Republican votes than the Republican will Democratic votes.
But we shouldn't count on it.
"And so in the place of the palace of privilege, we seek to build a temple out of faith and hope and charity."-FDR
by jallen on Tue Oct 31, 2006 at 03:27:41 PM EST

Exactly (none / 0)

It's a motivation issue, a get-your-base's-butt-off-the-couch-and-i nto-the-voting-booth issue.  The Republicans are, typically, better at this than we are, hence the reason they get about 4 or 5 points more than we do, comparing the generic numbers vs. the actual results.

This year is quite different.  26 points is possible if their side is really demoralized and ours is pysched up and the Indies all go our way.  Now, I think the religious conservatives will always vote (although the notion of closeted gay Republicans (such as Mark Foley) may temper them down a bit, although not as much as we would hope for).  However, fiscal conservatives and "mainstream" Republicans and Libertarian-lite folks may all either be demoralized or even be voting our way this time.

If there really is a 26 point swing, we could win a hundred more seats, a blowout much bigger than 1994 even.

This is still all very best-case-scenerio stuff, very unlikely, but it's fun to speculate.


by Geotpf on Tue Oct 31, 2006 at 06:28:50 PM EST
[ Parent ]

Re: What If Democrats Win By 26 Points? (none / 0)

Actually, the answer to Chris' question is in this paper (PDF) by three academics.  They analyze how many seats a generic ballot advantage translates into.  26 is pretty much an impossible generic ballot advantage and would translate into a ridiculously huge number of Democratic seats (a national popular vote of 63-37??).  But the authors do suggest a Democratic gain of about 32 seats (to 235) based on the generic ballot tests as of last week.


by Sandwich Repairman on Tue Oct 31, 2006 at 09:00:43 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.