Previously we examined delegate allocation and the caucus process in Texas. The short version is that:

1. Texas has 193 pledged delegates, as well as 35 unpledged delegates
2. The Statewide Primary vote, by itself, determines no delegates and is merely a “beauty contest,” except insofar as it translates into delegates elected in State Senate Districts, or through caucuses (in order to participate in caucuses, one must also have voted in the primary). It is very much possible that whoever wins the sum of the Senatorial District Primaries will not win the majority of Texas’ delegates.
3. 126 pledged delegates will be allocated proportionally in 31 separate primary elections conducted in each of Texas’ State Senate Districts.
4. 67 pledged delegates will be allocated through Texas’ caucus system, the first round of which will be held in precincts across Texas at 7:15 PM on March 4th. 42 of these will be rank and file “at large” delegates, and will ultimately come from ordinary people who participate at their precinct caucuses, while 25 will be party leaders and elected officials (PLEOs).

This post will analyze how Texas’ delegates are most likely to be split between Senator Obama and Senator Clinton.

This analysis assumes demographic voting patterns roughly similar to those seen in Super Tuesday and in the elections that have already occurred. It is somewhat immune to a shift in the popular vote, but if one candidate blows the other one out of the water, that will obviously shift the delegates in favor of the winner. But provided that both Obama and Hillary remain generally competitive in Texas, these projections shouldn’t be off by more than 5-10 delegates in either direction.

My analysis also does not include racial classifications other than African American, White, and Hispanic. This may very slightly skew things in a few places. Most notably, Asian Americans make up 3.4% of the population of Texas, and are concentrated mainly in Houston, DFW, and Austin. Asian American voters favored Clinton by about 3 to 1 in California, so that might give her a slight edge in a few State Senate Districts. Of course, Asian American is a complex demographic, and Asian American voters in Houston may have different voting patterns than Asian Americans in San Francisco, for example. Additionally, I am generally assuming that Hillary Clinton will do as well among Hispanics in Texas as she has done in other states. On the other hand I am assuming that while Hispanic turnout will be higher than turnout in the 2004 primary, it will not increase at a much faster rate than other demographics.

Texas State Senate Districts:

Map 1, by region – hat tip to Lone Star Project.

Map 2, showing the same thing with county lines visible.

Close Up of Dallas/Fort Worth

Close Up of Houston

Close Up of Austin/San Antonio

District 1

This East Texas GOP district includes the cities of Texarkana, Longview, Part of Tyler, and surrounding rural areas. All the demographics are unfavorable here to Obama, with the one exception of the African American percentage. However, at 17% of the population, AAs should make up about 30% of primary voters. The rural, white, over 65, non-college educated demographics all point strongly to Hillary winning the popular vote, but the 4 delegates here should split 50/50. It’s worth noting here that the data on urban/rural population seems to generously define “urban” to include suburban areas and smaller towns.

District 2

This GOP district includes the Dallas suburb of Rockwall, some East Dallas exurbs, and rural areas. The above average AA population, the above average > 50K vote (concentrated in the DFW suburbs), and the urban vote should keep Obama close enough to split the delegates evenly. Hillary should carry the popular vote, but that will have absolutely no effect on the delegate count at the end of the day.

District 3

This is yet another East Texas GOP district with 4 delegates. It is composed of towns like Nacodoches and Lufkin, along with more than its fair share of rural counties. This is a very tough district for Obama. It is (often by a wide margin) whiter, older, more rural, poorer, and less well educated than Texas as a whole. The only saving grace for Obama is that the AA population is very slightly higher than the Hispanic population. If Obama’s campaign puts resources here, it might, just might be able to get the ~37.5% of the vote it needs to split the delegates. Otherwise Obama is down 2 here.

District 4

The final East Texas GOP district, District 4 includes Beaumont, some Houston exurbs, and rural areas. This is district is a bit more urban and has more >50K voters, and somewhat resembles District 2. The 4 delegates should split evenly here, and Obama may even win the popular vote if he has a good day.

District 5

District 5 is a GOP district that includes Bryan/College Station (Texas A&M University), North-Austin suburbs around Round Rock, Huntsville, and some parts of rural Central Texas. Although Texas A&M is conservative for a University, it is still a University, and that favors Obama. There are slightly more > 50K voters and slightly more AA voters than the state average. It is highly unlikely that the delegates here split anything but 2-2, and this district should be competitive in the popular vote, perhaps with a slight Obama edge.

District 6

District 6 is a Democratic, Hispanic majority district in Houston. The fact that 72% of the district is Hispanic simply overwhelms all other demographic factors, and makes it very likely that Hillary wins 2 delegates and Obama wins 1 delegate. Obama won about 1/3 of the Hispanic vote in even in California, and did better in Arizona and New Mexico. Since this district has 3 delegates, all Hillary has to do to win 2 is to come in ahead of Obama.

District 7

District 7 is a GOP district located in the suburbs and exurbs to the north-west of Houston. With the exception of the fact that African American voters are about 4% below the average for TX, the demographics are favorable to Obama. Hillary could potentially make it competitive, but the 60% > 50k and the 41.6% with at least a bachelors degree will likely favor Obama, giving him 2 delgates to Hillary’s 1.

District 8

District 8 is a GOP district in the northern area of Dallas, centered on Plano. There are plenty of “wine track” voters who are educated and have a household income of more than 50K, a key part of Obama’s base. However, it is unlikely that he will win by a big enough landslide to secure 3 delegates, so they should split 2-2.

District 9

District 9 is another suburban GOP district, somewhat similar to district 8, which covers the area between Dallas and Forth Worth, including Grand Prairie and parts of Arlington and Irving. Obama benefits demographically from relatively young, educated, upper middle class voters and an above average percentage of African American voters, while Hillary will probably keep it close with the help of Hispanic voters, and might even win. Still, on balance I think Obama should be favored to win the extra delegate, 2-1.

District 10

District 10 consists of Fort Worth and much of its southern suburbs. It is slightly more Democratic than Texas as a whole, but nonetheless leans GOP. African American voters will help Obama and Hispanics will help Hillary, but the AA vote can be expected to more than counterbalance the Hispanic vote. Obama should also do well among white voters here, since they skew towards being > 50K and college graduate voters. All of that adds up to Obama most likely receiving 3 delegates and Clinton most likely receiving 2 delegates. It is also possible that Hillary could win 3-2, or that Obama could win 4-1, though this is less likely.

District 11

District 11 sits on the southern suburbs of Houston, and reaches down to Galveston via the Johnson Space Center. That’s right, this is one of the only areas of the country where candidates’ Space Plans will be a top consideration for a significant number of voters. I have no idea whose space plan will be more popular, but Obama generally does well among highly educated voters. The popular vote in District 11 could tilt slightly to Hillary or slightly to Obama, but it doesn’t matter which way it tilts. Both candidates will walk away with 2 delegates.

District 12

District 12 is yet another GOP district (we have already went through way too many of those) includes northern Tarrant County (Fort Worth) and the southern part of Denton County. District 12 is heavily white, and more than 50% of households have a > 50K income. The AA vote for Obama and the Hispanic vote for Clinton should roughly cancel out. This district could vote either way, but again, it does not matter. Both Hillary and Obama should get 2 delegates.

District 13

District 13 is a Democratic district which is half African American. However, African Americans should be overrepresented among Democratic primary voters, and will probably make up about 80% of those who vote in District 13. Combine that with a slightly greater than average percentage of college students and voters with at least a bachelors degree, and you have a recipe for an Obama goldmine. Clinton must rely on Hispanic support, some white voters, her ~15-20% of the AA vote, and a handful of votes from Asian Americans to secure more than a single delegate. Anything more than 30% of the vote here would be a victory for Hillary. She should take 2 delegates to Obama’s 5, but if Obama does very well he could take 6 or even all 7 delegates (because of the 15% cutoff).

District 14

District 14 consists of most of Travis County (Austin), and includes the University of Texas. There is a heavy “creative class” contingent to be found, with tech jobs, > 50K voters, college educated voters, etc. Hillary probably takes 35-40% of the popular vote at most, and thus 3 delegates, while Obama takes the other 5 delegates. A 6 to 2 split is also possible, but difficult for Obama to achieve, mainly because of the fairly large Hispanic population. In the unlikely event that Hillary manages to tie the delegates at 4-4, that is a major victory for her.

District 15

District 15 is a Democratic leaning swing district centered on the North Houston suburbs. African Americans should make up a bit under half of Democratic primary voters, but even so, Obama and Hillary will probably split the delegates 2-2. If Obama does very well he could conceivably take 3 delegates, but this is unlikely unless he substantially improves his margins among Hispanic voters.

District 16

District 16 is a GOP district located in North Dallas, which includes the affluent Highland Park area. The African American vote should marginally outweigh the Hispanic vote, while voters will be favorable towards Obama in the > 50k and college graduate demographics. The only question is whether Obama takes 3 delegates or whether the candidates split them 2-2. On the assumption that the Hispanic vote gives Hillary a sufficiently large buffer, I predict that the delegates are split.

District 17

District 17 is a meandering GOP district that stretches all the way from West Houston and Sugarland down to Galveston, and then over to Port Arthur. Favorable numbers of African American voters, > 50K voters, and College educated voters give Obama the edge and 3 of District 17’s 5 delegates. There are not enough Hispanic voters to counter the African American vote.

District 18

District 18 is a large, mostly rural GOP district including Victoria and most of the area between Houston, Austin, and Corpus Christi. All the demographics are favorable to Hillary. Simply put, most every Democrat here that is not an African American is part of Hillary’s base. This district will likely send 3 delegates to Hillary and 1 to Obama, though Obama has an off chance of keeping Hillary’s win small enough to split the delegates. If Hillary cannot win by a big enough margin to win 3 delegates here, she will not do well statewide.

District 19

District 19 covers much the same territory as Ciro Rodriguez’s US Congressional District 23. It stretches all the way from San Antonio to the edge of El Paso along the Rio Grande. About half the population of the district lives in San Antonio, and 2/3 are Hispanic. Hillary should have no trouble at all winning 3 of the 4 delegates. Obama would have to do much better among Hispanics to tie the delegates. In the highly unlikely event that this happens, it would be an enormous victory for Obama.

District 20

District 20 is another heavily Hispanic district, which includes parts of Corpus Christi and McAllen, and a sliver of land connecting the two. It should yield similar results to District 19 – 3 delegates for Hillary, 1 for Obama. It is not so much a question of whether Obama loses in a landslide, but of how big the landslide is.

District 21

District 21 includes Laredo, a small part of suburban/exurban San Antonio, and a vast rural area in between. Perhaps it is beginning to sound like a broken record, but like 19 and 20, 21 is 2/3 Hispanic and Hillary will blow Obama away. 3 delegates for Hillary, 1 to Obama.

District 22

District 22 is a GOP district that stretches from the DFW suburb cities of Burleson and Waxahachie down Interstate 35 to Waco. With the exception of the average sized African American population, the demographics are simply unfavorable for Obama, and Hillary should have no difficulty beating him and winning 2 of the 3 delegates at stake.

District 23

District 23 is a heavily Democratic district that includes much of urban Dallas. It is about 40% African American and about 40% Hispanic, but African Americans should not only make up a larger percentage of the Democratic primary electorate, but vote more strongly for Obama than Hispanics vote for Clinton. The most likely outcome is 4 delegates for Obama and 2 for Clinton, but it could also split 5-1 if Obama really runs up the score among African Americans and limits the damage among Hispanics. 3-3 is conceivable, but only has even a faint chance of happening if Clinton not only does very well generally and among Hispanics but manages to increase Hispanic turnout to well over normal Texas levels.

District 24

District 24 is a strongly GOP district which includes Abilene and Temple/Killeen, as well as much of the rural areas in between. The Demographics here are quite favorable to Clinton. While Obama may do fine in the Abilene and the Killeen area, next to Fort Hood where the Iraq War will be a large issue, Hillary should do very quite well in the rural and small town areas. On the other hand, there are fewer Democrats in those areas. Still, if Hillary can’t win 2 of the 3 delegates here, she is in serious trouble.

District 25

The best thing about District 25 is that it is home to the world famous Schlitterbahn water park in New Braunfels. But it also includes San Marcos (home to Texas State University), northern San Antonio, and parts of southern Austin. Obama’s biggest problem here is the near total absence of African American voters. On the other hand, he could form a coalition of Texas State University voters, > 50k household income voters, and college educated voters. I predict a 3-3 split, but it is also possible that Obama will take a 4-2 lead, or that Hillary will get the better end of a 4-2 split (less likely, but still possible).

District 26

District 26 is an urban, Democratic, Hispanic majority district located in San Antonio. With a 2/3 Hispanic population, this district is all about the Hispanic vote. Unless Obama can do better among Hispanics, this district will split 3-1 in favor of Hillary. With good Hispanic outreach, it is possible that Obama holds Hillary’s margin low enough that an even split will result. Poblano has suggested that “it is one’s immigration experience, and not one’s race, that appears to account for Hillary’s stronger support with Hispanic and Asian voters.” If this is true, Obama may well be able to hold Hillary’s margins among Hispanics in San Antonio and parts of South Texas below her California levels, and split this district 2-2. But without much solid data to confirm this, I have to predict 3 delegates for Hillary and 1 for Obama.

District 27

District 27 is a Democratic district in the Rio Grande Valley that includes Brownsville. It has an extraordinarily huge Hispanic population, and the Hispanic vote is virtually synonymous with THE vote in this district. The only good thing about this district for Obama is that it only has 3 delegates. They should split 2-1 in Hillary’s favor, but she could conceivably sweep it 3-0 if she does very very well among Valley Hispanics.

District 28

District 28 is a massive West Texas GOP district. It includes Lubbock, San Angelo, and utterly vast tracts of very sparsely populated West Texas space. With the exception of college students (because of Texas Tech in Lubbock), all the demographic categories are more favorable to Hillary than the State average. West Texas, however, is very different from East Texas, and the further west one moves, the more the voting patterns should start to resemble those seen in New Mexico. Hillary should easily win the delegates 2-1.

District 29

District 29 is essentially the city of El Paso in the far west corner of Texas, just south of the New Mexico border and a short drive from Las Cruces, New Mexico. Hispanics make up 78% of the population, Hispanics will vote for Hillary, and the delegates should split 2-1 for Hillary. But expect her margins among Hispanics to be more comparable to those in New Mexico than those in California.

District 30

District 30 is a North Texas GOP district which includes Wichita falls, some DFW suburbs and exurbs, and surrounding rural areas. It is one of the most heavily white districts in Texas, and it should vote for Hillary by a substantial margin. Edwards may take a fair share of the vote as well just like he did in neighboring Oklahoma, but will take votes from Hillary rather than Obama. Barring something completely unexpected, this district goes 2-1 for Hillary as well.

District 31

District 31 is our last district, and also the one with the fewest delegates – 2. It is a GOP district that reaches up from Midland/Odessa in the south all the way along the New Mexico border to Amarillo and the Oklahoma panhandle. Hillary should win the popular vote easily, but Obama only needs a little bit more than 25% to split the delegates. This actually could be a problem for him. But 25% is a very low hurdle, and given that Obama passed that hurdle even in Hillary’s best state (Arkansas), one would imagine that he would be able to pass that hurdle in District 31 on the strength of professionals in Midland/Odessa and Amarillo.

Statewide Popular Vote

My projections for the Statewide popular vote are based on the 2004 Texas Primary Exit Polls and the 2008 Missouri Primary Exit Polls, and more than a little guesswork – it is an educated guess, but a guess nonetheless. But as long as the Statewide totals are not wildly off, the actual delegates from the State Senate District primaries should not swing too far, because it takes a large swing in the popular vote to swing many proportionally allocated delegates when there are only a small number of delegates being awarded per election (2-8). This is especially true for State Senate Districts with smaller numbers of delegates.

I assumed that the racial demographic makeup of 2008 Democratic Primary voters in Texas would be the same as the demographics of 2004 Democratic Primary voters in Texas. I am assuming that African American, Hispanic, and Asian voters are monolithic blocks, which they are not, and assume that AAs will vote 80/20 for Obama and Hispanics will vote 2-1 for Hillary, simply because it is too complicated to do otherwise. I also give Asian Americans 2-1 to Hillary and split “other” 50-50, because I don’t know what else to do. 2004 Texas Primary also voters were probably more educated and wealthier than 2008 Texas Primary voters will be. A full 50% of 2004 Texas Democratic Primary voters had household incomes of greater than 50K. But only 46% of 2008 Democratic Primary voters in Missouri had household incomes of greater than 50K. Likewise, 42% of 2004 Texas Democratic Primary voters were College Graduates. But only 33% of 2008 Democratic Primary voters in Missouri were college graduates. In both cases, one would expect fewer >50K and fewer college graduate voters in Missouri because Missouri has slightly fewer of both demographics than Texas, but on the other hand turnout in 2008 will be higher than in 2004, and wealthier and better educated voters are likelier voters. I assume that all of this will roughly cancel out, and 2008 white voters in Texas should vote similarly to 2008 white voters in Missouri.

A 51.5-46.5 margin for Hillary also agrees fairly well with the only poll we currently have of Texas, conducted by IVR on January 30th/31st, just after Edwards dropped out of the race. The IVR poll had the race at 48-38 in favor of Hillary Clinton. A 51.5-46.5 result has most of the undecideds going to Obama, which seemed to generally happen from poll averages to the Super Tuesday results, and also fits with the idea that Hillary Clinton is the pseudo-incumbent.

Texas Caucuses

As mentioned previously, 67 of Texas’ delegates will be determined by caucuses. Senator Obama has been doing very well in caucuses with the exception of the Nevada caucus,, in which he won the State delegates but lost the precinct delegates. Nevada has at least two relevant similarities to Texas – it has a high Hispanic population, and both campaigns will have a major presence, have a decent amount of time to set up their organizations, and will be strongly competing. In many of the caucuses that Obama has won by 2-1 margins, Hillary had basically ceded (Idaho, Nebraska, Kansas, etc), so Obama cannot be expected to do that well. But Texas has not had real contested caucuses since 1988, so nobody really knows how they will turn out. So for the caucuses I predict:

Obama: 55% – 25 At Large delegates – 14 PLEO delegates

Clinton: 45% – 17 At Large delegates – 11 PLEO delegates

Total Projected Delegates

Obama: 99 UPDATED – 98 Delegates

Clinton: 94 UPDATED – 95 Delegates

As Barack Obama said about hope at the Virginia Jefferson-Jackson Dinner, on February 9th, “the notion is that somehow, if you are realistic, then you set your sights lower – That if you talk about hope, then somehow you have to be passive, and have your head in the clouds, and just wait for things to happen to you. I have to remind people, that’s not what hope is.”

We can’t just be passive, have our heads in the clouds, and just wait for things to happen to us. One way or another, Texas is going to be both close and crucial. So if you enjoyed this post and are an Obama supporter (or even if you are not! ), please consider sending Obama a donation to help Turn Texas Bluebama! If you live in Texas, sign up at Texansforobama. If you don’t live in Texas, you can help by phonebanking (not only in Texas, but in other crucial states as well!).

Update – I made at least 3 errors in my delegate counts. In district 3, I gave Hillary an extra delegate by mistake. In district 14, Clinton should have an extra delegate and it should be 5-3 Obama. In district 25, I mistakenly took 1 delegate from Hillary and gave it to Obama. In sum, this gives Hillary 1 more delegate, and makes the final total Obama 98, Clinton 95. Unless, of course, I made some other mistakes . Thanks to eparrot at Daily Kos and Tantris at MyDD for their heads up attention to detail.