So, I'm no statistician, but I've been looking at the early voting totals coming back from the precincts in Texas' 26th Congressional District, and it is surprising how it's looking for Democrats.
What I'm looking at in these numbers is the breakdown of known partisanship among voters who have and have not yet voted. General election ballots are of course secret, and we have no way to know how an individual voted. What we can look at though is whether that person has any history of voting in any party's primary elections. In that case, we may not know who they voted for, but we know which party.
So, this morning, I pulled the numbers for district 26, and here's what we think we know about those who have voted:
Table 1: Early Voters
Hard Dem
Soft Dem
Unknown
Soft Rep
Hard Rep
TOTAL
4,392
64,808
57,102
10,299
30,929
167,530
2.6%
38.7%
34.1%
6.1%
18.5%
100%
So, with a current turnout of 41.01%, Dems have 41.3%, and Republicans have 24.6%. That looks good, but there is a 34.1% unknown turnout. these are voters who haven't affiliated themselves via primary voting.
We also have a couple of offsetting factors: The Limbaugh effect of Republican voters crossing over to vote for Hillary in the Democratic Primary, as well as Republican voters who legitimately have switched their support to Barack Obama. It's hard to say what those numbers might be, but they probably don't exceed 5%. We're also making the huge assumption that voters who vote Democratic will vote for Ken Leach - the Democrat. Some low-information voters vote on the basis of name recognition, which favors the incumbent, Michael Burgess.
But before we declare it over, we have to look at who has not yet voted. So we pulled a list of voters who have not yet voted in early voting, and here's what we got:
Table 2: Voters not yet voted
Hard Dem
Soft Dem
Unknown
Soft Rep
Hard Rep
TOTAL
980
53,054
152,415
14,897
19,622
240,968
0.4%
22.0%
63.03%
6.2%
8.1%
100%
So, even with these numbers, Democrats seem to have more votes "in the well" so to speak. Republicans thus far seem not to have been as energized about early voting.
But the large amount of unknowns still leave a lot of room for error. So what I attempted to do was model how the unknowns and the rest of the unvoted partisans would add up. You can't just add the numbers, because there is some variability in their probability of voting. Unknowns tend to be less likely to vote than partisans from either side.
Using voter probability rankings provided by the Texas Democratic Party, I weighted the remaining voters by likelihood of their turnout. I then added these numbers to the early voting totals and came up with this:
Table 3: Predicted Totals based on Weighted Turnout Plus Early Voting
Hard Dem
Soft Dem
Unknown
Soft Rep.
Hard Rep
TOTAL
5299
102588
152916
21274
48772
330849
1.60%
31.01%
46.22%
6.43%
14.74%
100%
This model leaves us with 32.61% leaning Democratic, 21.17% leaning Republican, and 46.22% unknown turnout. Mathematically, it could go either way, depending on how you think those unknowns will vote.
There are several possibilities for that: - They'll mirror the past election percentages in CD 26. - They'll mirror the national breakdown. - They'll vote in proportion to the partisan breakdown. - They'll vote somewhere between the partisan proportions and last percentages shown in CD 26.
It's hard for me to guess where that will fall. But I went ahead and figured out what the unknowns would have to be:
If the unknowns vote at least 38% Democratic, the Democrat wins. Republicans have to get at least 63% in order to get a majority. There is a very slight chance of a runoff being necessary between Burgess and Leach, if the Libertarian draws enough votes.
Given that identified Democrats in this last scenario outnumber identified Republicans by approximately 3 to 2, and that Republicans would have to more than flip that number around, it seems unlikely. Polling analysis website FiveThirtyEight.com puts Texas right now at 55.8% to 44.1% in favor of John McCain.
Further, we think a heavy turnout and national momentum for Barack Obama may benefit Texas Democrats more. How motivated will older Republicans be to turn out and vote for John McCain - a guy they only begrudgingly accepted, and who seems destined to win Texas but lose the national election.
I think there is a very real chance that sleeper candidate Ken Leach may actually win this thing. It would be a huge upset, but I think North Texas voters would reap a benefit from the choice.
Dems, you can do this. Get out and vote on November 4th, and pull it straight ticket for the Democrats. Then you can be assured that Obama will get all the help he needs to undo years of disastrous Bush policies, and send Bush's backers to the unemployment line.
On election night, I was surprised. I expected it to be much closer. With 20/20 hindsight, it's easier to see what happened.
Here's the Secretary of State's totals in the CD 26 race:
Candidate
Party
Early
Total
Michael C. Burgess - Incumbent
REP
142,257
59.05%
194,849
60.18%
Ken Leach
DEM
91,642
38.04%
117,895
36.41%
Stephanie B. Weiss
LIB
6,987
2.90%
11,002
3.39%
Early Votes: 240,886 Total Votes: 323,746
So basically, what appears to have happened is that a good portion of those who we thought were soft Democrats due to their primary voting history actually voted Republican. The Limbaugh effect was real. The other possibility that people are discussing is that as much as Obama energized the Democratic base to get out and vote, he had the opposite effect on Republicans.
Based on early voting, we all thought election day was going to be really heavy. But it turns out that possibly for the first time ever, more people turned out to early-vote than on election day.
The other thing that weighed in significantly, it appears, is that Burgess received a higher percentage of the vote than John McCain. Although it's unfathomable to me how someone could see it in their best interest to vote for Barack Obama, but then send Michael Burgess back to Congress, there are a significant number of people who vote based on name recognition and who they think is a nice guy. Burgess does get out into the community and get his name out there.
The Calendar
Our Polls
Our Videos
Our Podcasts
On The Facebook
In the Twitter
Country Radio
What went wrong?
