April 02, 2005
Texas Democrats: A Statistical Profile
By Jim Dallas
Using the recently-released Edison/Mitofsky 2004 exit poll data (back-up | code-book | my Excel spreadsheet), a few interesting statistics about Texas Democrats can be constructed.
Exit polls are, of course, polls, so take with a grain of salt. I sure wish those urging that exit poll discrepancies prove voter fraud would take a chill-pill.
UPDATE: Some errors in the age tabs were fixed. The under 30 share of the Democratic vote is 20 percent, not 8 percent (8 percent is the 18-24 share).
The entire Texas electorate is summarized on the CNN web site, and my weighted numbers essentially match theirs (adjusting for rounding).
Numbers may not add up to 100 due to missing data, etc.
Democratic Voters (32.1% of total electorate, MoE 2.4%)
Presidential Vote:
Ideology/Philosophy (self-reported)
| Liberal | Moderate | Conservative |
| 24 | 51 | 23 |
Gender:
Race/Ethnicity
| White | Black | Latino | Asian |
| 41 | 29 | 28 | 1 |
Age
| Under 30 | 30-45 | 45-65 | 65+ |
| 20 | 21 | 42 | 15 |
Size of Place
| 500000+ | 50000 - 500000 | Surburban | Small Town | Rural |
| 31 | 23 | 31 | 9 | 5 |
N: 577, Approx Margin of Error 4.2%
I'm not going to reproduce the Kerry and Liberal voter cross-tabs here, though I will note an odd quirk - 32 percent of self-identified "liberal" voters reported voting for Bush (approximate margin of error 6.5%).
The exit poll data contains a number of other interesting variables (region, religion, income, urban/rural, etc.) but I am busy working on a paper this weekend, and don't have the time to crunch those numbers.
Posted by Jim Dallas at April 2, 2005 04:17 PM
| TrackBack
Incidentally, using the rules of joint and conditional probability:
If 41 percent of Democrats are white, and Democrats make up 32 percent of the electorate, then white Democrats make up 13 percent of the Texas electorate.
If white voters make up 66 percent of the entire electorate, then the percentage of whites who are Democrats is 13/66, or only 20 percent.
Ouch.
This is why Democrats statewide are loosing. We have to get 40 percent of the white vote and the minority vote has to be at least 30 percent.
What's remarkable, though, I think is that 2004 turnout tilted so much more conservative and Republican than the total-population polling I've posted earlier.
Changing this turnout demographic will be important thing; this was of course the logic of the "Dream Team" in 2002, which didn't really work out.
The problem, very simply, is that we have is that we have a natural constituency in South Texas which is disproportionately apolitical, particularly as you go down the SES scale.
Community organizing. I cannot beat that horse enough.
Bah... to lazy to make the edit... but the primary voters are profiled here:
http://www.cnn.com/ELECTION/2004/primaries/pages/epolls/TX/index.html
Notice the key differences between the primary and general voters, which illustrate the differences between core Democrats and "November Democrats."
