Back to MyDD Weblog

Survey USA has some sort of fault with their Hispanic methodology. You can see it in their NC Senate results, where they show Dole getting 69% to Bowles 19% of the total number of Hispanic votes. Granted, it's small in the NC sample (23 likely Hispanic voters), but the fault is across the board, and just seems magnified with their Texas results (could be accent-related, since it's automated).

In the Texas polling, they use only a 5% likely voter composition of Spanish speaking voters. Based on the EV turnout in S. Texas, that's likely to be higher. I've been following the Travis County EV results, which ended yesterday at 19% of the registered voters, with the average being 8-9%, and the results down further in S. Texas are even higher:

The border counties combined; El Paso, Webb, Hidalgo, Cameron, Nueces, and Yellow Dog Jefferson, have outvoted the Republican strongholds in Collin, Denton, Williamson, Montgomery, Midland, Lubbock, and Randall counties. It's hard to predict Dallas, Harris and Bexar.

In terms of the statewide average, as of Friday, Early voting in the largest 15 counties was up 71% over the last midterm election. Texas is going to have close to 1,000,000 early votes in the 15 largest counties, which should project to 1.5 million across the entire state.

Given the straight-ticket voting going on with Hispanics in S. Texas, if Kirk is getting what SUSA says he is of the white vote, 31%, he's probably going to win.

Back to MyDD Weblog