Costly effort pays for Tony Sanchez - Becomes first Hispanic Democratic nominee for Texas governor

The Houston Chronicle ^ | March 12, 2002 | R.G. RATCLIFFE

[KQQL]

so you are saying hispanics are 13% of the reg voter, blacks 11% , whites 70% and others 5%

[KQQL]

If the Hispanic turnout is at 50% of the 2.2 million reg voters and my black numbers are correct, and whites sleep in nov,2002 and Rats get 33%-39% of the white votes....I gues s RATS will win in texas

[Torie]

In 2000, it appears that Blacks voted about 52% of VAP, and Anglos about 73% of VAP. Hispanics were about 36% of registered, and that is a bit harder number. In any event, your analysis assumes rather substantial shifts in turnout percentages. There is a fairly substantial error factor here, which I think is largely reflected in a too high Anglo VAP percentage. If you take that error factor entirely out of the Anglo number, than the Anglo percentage is 58% of VAP.

[Torie]

Others I suspect is under 2%. But that would be part of the Anglo percentage mostly.

[KQQL]

remember in mid-term election turnout will be lower for whites.....and higher for hispanic and blacks,

[Torie]

Maybe. Generally it is the opposite, but maybe not in Texas this time. I can see about a 70% Anglo turnout, and 15% each for Hispanics and Blacks. That strikes me as a realistic Dem hopeful scenario. If the Dems get a third of the Anglo vote, 3/4 of the Hispanic vote, and 95% of the Black vote, that equals 48.6%. I don't think it will work out that well for the Dems, but as you can see, the Pubbie has a slight edge, assuming the Anglo split is accurate.

[Torie]

I suspect in the end nothing much will change in Penn. The litigation is over in Georgia, so legality is moot.

[Tex_GOP_Cruz]

Thank you once again, Torie - hope you don't mind me sharing the credit for the model; seems like you've done all the hard work. Anyhow, using the corrected (latter) results of the least squares analysis (Am I correct in understanding that it is based on historical turnout using racial percentages by county to determine racial turnout through regression? These would then be the numbers to use directly, without adjusting for turnout percentage again as KQQL has done in his version.):

Tex_GOP_Cruz & Torie's Revised Perry Equation II
0.1112(0.05) + .1341(X) + .7547(Y)
Where X is % of Hispanic vote for Perry and Y is % of Anglo (plus other) vote for Perry and where a 5% black vote for Perry is assumed. If the expression is greater than 0.50 (50%), then Rick Perry wins.
Previous Version: 0.1466(0.05) + 0.1216(X) + .7317(Y)

For X = 0.45 (45% of Hispanics vote for Perry)
Y > 57.52% Perry wins

For X = 0.40
Y > 58.41% Perry wins

For X = 0.35
Y > 59.30% Perry wins

For X = 0.30
Y > 60.18% Perry wins

For X = 0.25
Y > 61.07% Perry wins

For X = 0.20
Y > 61.96% Perry wins

For X = 0.15
Y > 62.85% Perry wins

These numbers aren't quite so grim. But the result is certainly sensitive to relative turnout figures.

[KQQL]

it the end it probably will be a 55% 45 split in GOPS favor,,,,,but the Lt.Gove race will very close and could go either way

[KQQL]

it the end it probably will be a 55% 44% Rat 1 % other ,split in GOPS favor,,,,,but the Lt.Gove race will very close and could go either way

[Tex_GOP_Cruz]

Question for you - did you run the regression on 1998 gubernatorial election numbers? If so, I wonder if 1994 numbers (adjusted for racial shifts in Texas) would yield comparable results or if turnout is that variable in Texas.

[Torie]

Am I correct in understanding that it is based on historical turnout using racial percentages by county to determine racial turnout through regression

I did a regression analysis on a precinct by precinct basis comparing Hispanic percentage turnout of registered voters to the precinct percentage turnout of registered voters, and for black and Anglos, their percentage of VAP in the precinct to percentage of registered voters turnout. I threw out small precincts with odd percentages, some in which the turnout was higher than Vap, or whatever. Those precincts either were altered, or grew a lot, or whatever. In general, that was less than 200 precincts out of 8300.

[Torie]

No just for the 2000 election. We have a lot of data there, because the legislatures demanded it be generated for redistricing/gerrymandering. This is the first time we got such detailed data, and we will only get it once every ten years henceforth.

[crasher]

roll call reports that a federal judge has scheduled a trial, and delayed the filing deadline in Michigan over the new Congressional lines. This could be a total disaster.

[Torie]

The article is so poorly written, that it is not clear what is before the federal court. The federal court interpreting state law the the state supreme court has already disposed of, strikes me as odd. What is the federal issue here? Regarding state law, the Dems are arguing that the map that splits the fewest county and municipal lines is the one that must be adopted. That strikes me as silly. That means that anyone who can come up with a map that splits fewer automatically trumps the rest.

[crasher]

lol... quite right about the article. Looked at the Detroit News, couldn't find anything there. Aside from what you said, I'd like to know WHO this judge is - Democrat, Republican? But the article was clear about one thing, he set back the filing deadline. And that scares me. That means he thinks there is a chance he will order new lines.

[KQQL]

!

[KQQL]

An Interesting take on the TX election by RAT Consultant:--

So much has been written about whom will vote this fall that it is now my time to forecast what might happen.

In the last U.S. Senate race in Texas during a Presidential election year (2000) we had just over 6 million voters voting in that race. In Phil Gramm's last election, in another Presidential election year (1996) 5.4 million voters voted in that race. In the last two Governor's election in non-presidential years we had a total of 3.7 million voters in 1998 and 4.3 million in 1994. Yes, we have more voters today so the question is just how many will vote this year. With about 12,500,000 registered voters on the voting lists in Texas we could see about 5.7 million voters actually go to the polls in November. That would be about 50% of the voters (11,500,000) who are actually eligible to vote in that they have not died or move out of the state. This could increase if either party does a massive voter registration campaign.

With Tony Sanchez and Ron Kirk on the Democratic ticket we should have more then a typical midterm election but less then a Presidential election turnout.

So to win in November you would need about 2.85 million votes (50% of the 5.7 million voters). How does Tony Sanchez or Ron Kirk or John Sharp get there? Here is what two well-known professors' say:

Dr. Richard Murray of the University of Houston is quoted in the Washington Post: "African Americans, who make up 12 percent of the Texas population, often account for less than 10 percent of the statewide voter turnout, Murray said. With Kirk in the race, Murray said, black participation is sure to rise significantly. Meanwhile, Sanchez's candidacy is certain to drive up Latino voting. Although Hispanics make up nearly one-third of the Texas population, they typically account for only about 12 percent of the turnout".

"Kirk can bank on overwhelming support from blacks and a large majority of the Latino vote, Murray noted. To win, though, "he'll have to get, minimally, about 35 percent of the white vote, and that's going to be a real hurdle"

But another University of Houston professor, Dr. Adolph Santos has another theory. He says in the Houston Chronicle "Sanchez is going to rely heavily on Hispanic support, and given his ethnicity, he can expect a large turnout in the Hispanic community. Currently accounting for a third of state's population, Hispanics can be expected to deliver as many as 1.2 million votes to the Laredo businessman running for governor. This assumes that 29 percent of the 4.2 million voting-age Hispanics turns out to vote".

Santos also says "And Senate candidate Ron Kirk can expect a huge turnout in the African-American community -- reaching as many as 950,000 African-American voters. This figure assumes that 58 percent of voting-age African-Americans will turn out to vote. Even this estimate may be conservative, given that the African-American community will be mobilized to elect the state's first African-American senator and the nation's only African-American senator.

"The combination of these two groups will give the Democratic candidates in statewide races 2.15 million votes. In a state with nearly 15 million voting-age residents, this may not seem like much. But, given that recent midterm elections in Texas have averaged a turnout rate of 3.9 million voters, 2.15 million voters will be more than enough to win the races" says Dr. Santos.

Dr. Murray is talking about registered voters and Dr. Santos is talking about voting age residents, not necessary either eligible to vote or even registered. Some are not citizens and others are not eligible due to criminal history or that they are currently in prison. And some folks just do not want to vote period.

And Dr. Santos assumes the two candidates would get every minority voter who showed up at the polls. In real life that does not happen. Even African-American candidates do not get 100% of the African-American vote. And you must also get Anglo voters as Dr. Murray correctly points out. You need many Anglo voters witch Dr. Santos does not even count on in his analysis.

So how do you get to 50% if you are either Ron Kirk or Tony Sanchez or John Sharp?

Let's assume that the African-American vote does increase from Dr. Murray's projection of 10% (570,000) to 15% of the total vote (855,000). That would be a 50% increase in turnout. Kirk should get 95% of that vote or about 812,000 votes. And if the Hispanic turnout increase by a similar percentage of 50% over that in the past they would go from Dr. Murray's prediction of about 12% to 18% or 684,000 votes to 1,026,000. Since neither Ron Kirk or Tony Sanchez will not get 100% of the Hispanic vote lets give them what Democrats can expect in good times about 70% of that vote or 718,000 votes.

If they each get the same percentage that leaves them with 1,530,000 votes and they will need an additional 1,320,000 Anglo votes to win. Of course a number of Anglo voters are Democrats already and can be expected to vote for their "Dream Ticket". Just looking at the Democratic Primary this year close to 300,000 Anglo voters participated and those are the real activists. So you could say that close to 500,000 Anglo voters are already committed to vote for Sanchez and Kirk. The key is to get them to the polls and then to convince Anglo ticket splitters, those who consider themselves to be independents to also go vote and vote for the Democrats.

If I was advising them I would recommend that they target Anglo women and use health care, insurance abuse, a woman's right to choose and sex education in the public schools as issues. In my scenario the number of Anglo votes left after subtracting the total votes of African-American voters and Hispanic voters is 3,819,000. To get 1,320,000 additional votes to win you would have to get 34.5 % of the Anglo vote, about what Dr. Murray predicted.

Of course this can change many ways. Change the percentage turnout among the three groups and the percentage changes on how you win. If Sanchez and Kirk can boast the turnout among traditional Anglo Democrats, Women and minority voters even more they pick up votes and do not have to rely on getting a third of the Anglo independent voters. If the GOP does a better job of motivating its voters they win.

So, watch the campaigns and see how they position their issues to motivate minority voters and attract women and independents.

George gstrong@political.com

_____________________________________________________________________ GEORGE STRONG and ASSOCIATES ** See Us At ** http://www.political.com

Providing Government Relations-Political Consulting-and Political Buttons!

*** he'S RIGHT ABOUT THE BLACK SPLIT , However I think RATS will get 75% not 70% of the hispanic votes////

[Tex_GOP_Cruz]

My initial take on Strong's analysis is that the percentages look similar to the ones we arrived at a couple of weeks ago. However, his turnout numbers for black and especially Hispanic voters seem a little arbitrary. (For example, I don't think the Orlando Sanchez race increased Hispanic turnout quite this much; then again a mayoral race with a Cuban Republican is not a gubernatorial race with a Mexican Dem.) Thus, I feel that Sanchez and Kirk need perhaps closer to 40% than 35% of the Anglo vote. It is probably easier to predict (and, as Nov approaches, to poll) percentage support in each group than turnout. Thanks for posting this, KQQL.