IL-06: Inside the Polls (or Why Duckworth Isn't Winning)

Following a comment at SoapBlox/Chicago, I took a closer look at the internals on the IL-06 polls from Constituent Dynamics. (IL-06 Cross-tabs for August and October.) What I found offers a suggestion of some of the internal dynamics at work in the district and may explain why Duckworth can't seem to pull away.
In August, Constituent Dynamics had IL-06 at 47% Duckworth, 46% Roskam. In October it had barely changed; 47% Duckworth, 47% Roskam.

Here are some other metrics on IL-6 between the August 27-29th survey and now (October 8-10th):
Party AffiliationAugustOctoberRepublican46%46%Independent27%28%Democratic28%26%
Bush ApprovalAugustOctoberApprove34%41%Undecided8%10%Disapprove58%49%
Head-to-headAugustOctoberStrong Roskam35%39%Weak Roskam11%8%Undecided7%6%Weak Duckworth10%10%Strong Duckworth38%37%
Here's where it get's interesting...

Head-to-headRepublicans onlyAugustOctoberStrong Roskam69%66%Weak Roskam11%11%Undecided3%5%Weak Duckworth5%6%Strong Duckworth11%11%
Independents onlyAugustOctoberStrong Roskam17%23%Weak Roskam17%6%Undecided7%11%Weak Duckworth17%13%Strong Duckworth42%48%
Democrats onlyAugustOctoberStrong Roskam0%9%Weak Roskam4%5%Undecided6%3%Weak Duckworth8%13%Strong Duckworth83%70%
Bush ApprovalRepublicans onlyAugustOctoberApprove60%64%Undecided10%14%Disapprove30%22%
Independents onlyAugustOctoberApprove18%29%Undecided16%8%Disapprove66%63%
Democrats onlyAugustOctoberApprove1%13%Undecided0%4%Disapprove99%83%
So the Republicans and Independents are pretty much the same as last time, but the Democrats have changed significantly.

The gender and age breakdowns are nearly identical between the two time periods, but there is a difference in the racial breakdown (but it changed in the Independents too and there was little to no movement there in the head-to-head matchup).

And the money tables...

WhiteAugustOctoberStrong Roskam37%39%Weak Roskam11%8%Undecided4%6%Weak Duckworth10%10%Strong Duckworth38%38%
Non-WhiteAugustOctoberStrong Roskam14%41%Weak Roskam5%12%Undecided42%13%Weak Duckworth5%8%Strong Duckworth34%27%

If this data is to be believed then where Duckworth will win or lose this race is with non-white (predominantly Latino) Democrats. Right now she's losing it.

Tags: IL-06, Peter Roskam, Tammy Duckworth (all tags)



2000 words on why we need to defeat Roskam">
by ltsply2 2006-10-13 12:37PM | 0 recs

I looked at the sample size (and the subsamples) on one of the reports. Note that these are weighted samples - this injects additional error.

The 95% confidence level (margin of error) for the 256 Democrats in the sample at 50%/50% is 6.125% [1.96 X square root ((50 X 50)/256))].

The 95 % confidence level for the 104 non-white repondents is even larger at 9.609%

The overall picture is pretty decent. I wouldn't get too "het up" about the subsamples supposedly say. Not with those relatively small subsamples, anyway.

by Michael Bersin 2006-10-13 01:07PM | 0 recs
Re: IL-06: Inside the Polls (or Why Duckworth Isn'

So, um, that immigration flyer didn't sit too well with some folks, then?

by palamedes 2006-10-13 03:34PM | 0 recs
Constituent Dynamics

There are many more Hispanics and Asian/other in the October polling than in August. Which poll better reflects the district?

I've also seen another recent poll that has Duckworth clearly ahead.

by EvanstonDem 2006-10-13 05:53PM | 0 recs
Re: Constituent Dynamics

First, the very pro Duckworth polling that recently came out was a DCCC internal (I believe). It is certainly good news, but partisan polls are always somewhat suspect.

Second, the actual demographic makeup of the district is somewhere in between the two survey samples. According to the 2000 census it is:

89% White
5%  Hispanic
3%  Black
3%  Other

by ltsply2 2006-10-16 08:56AM | 0 recs
Re: IL-06: Inside the Polls (or Why Duckworth Isn'

Judging by the debate last night we're clearly winning or Roskam can't get his supporters out.

by markg8 2006-10-13 06:01PM | 0 recs
An objection

Not to the idea that non-White voters will make the difference in the race but rather that Duckworth is loosing the race.  You seem to assign Duckworth part of the blame when you say that she is loosing.  I personally would blame it on the sample or George Bush.  For example Mr. Bush's approval rating goes up in both NAV and D samples but in the D sample it goes up 15% while Duckworths support only sinks 6%.   As you yourself noted the number of hispanic democrats went up in the sample as did Bush approval.  I think it is clear that what happened is the poll included some Hispanics who weakly self-identify as Democrats who liked Bush's recent set of speeches.  The difference between the two polls lies not in Duckworths actions but in th sampling, we'll see who shows up in November.

by descartes 2006-10-14 08:37AM | 0 recs
Re: An objection

Yes, this is exactly what I was trying to get across. If this sample is to be believed, then there is a significant group of Hispanics in the district who self-identify as Democrats, but seem to show belief patterns more similar to Republicans. That they exist surely isn't Tammy's fault, but it does seem like a natural place to look for votes would be in people who self-identify as Democrats.

by ltsply2 2006-10-16 08:46AM | 0 recs


Advertise Blogads