Off the Kuff Rotating Header Image

Jill Stein

Precinct analysis: Ogg v Anderson

Kim Ogg had the second highest vote total in Harris County this year. Let’s see how that looked at a more granular level.


Dist  Anderson      Ogg  Anderson%    Ogg%
==========================================
CD02   156,027  117,810     56.98%  43.02%
CD07   135,065  118,837     53.20%  46.80%
CD09    26,881  106,334     20.18%  79.82%
CD10    78,602   38,896     66.90%  33.10%
CD18    47,408  154,503     23.48%  76.52%
CD29    36,581   93,437     28.14%  71.86%
				
SBOE6  328,802  277,271     54.25%  45.75%
				
HD126   34,499   26,495     56.56%  43.44%
HD127   46,819   26,260     64.07%  35.93%
HD128   39,995   18,730     68.11%  31.89%
HD129   40,707   27,844     59.38%  40.62%
HD130   57,073   23,239     71.06%  28.94%
HD131    7,301   38,651     15.89%  84.11%
HD132   36,674   31,478     53.81%  46.19%
HD133   46,242   29,195     61.30%  38.70%
HD134   43,962   45,142     49.34%  50.66%
HD135   31,190   28,312     52.42%  47.58%
HD137    8,728   18,040     32.61%  67.39%
HD138   26,576   24,189     52.35%  47.65%
HD139   12,379   39,537     23.84%  76.16%
HD140    6,613   20,621     24.28%  75.72%
HD141    5,305   32,677     13.97%  86.03%
HD142   10,428   34,242     23.34%  76.66%
HD143    9,100   23,434     27.97%  72.03%
HD144   10,758   16,100     40.06%  59.94%
HD145   11,145   22,949     32.69%  67.31%
HD146   10,090   38,147     20.92%  79.08%
HD147   12,156   45,221     21.19%  78.81%
HD148   17,538   29,848     37.01%  62.99%
HD149   15,352   27,535     35.80%  64.20%
HD150   47,268   28,160     62.67%  37.33%
				
CC1     73,521  240,194     23.44%  76.56%
CC2    123,178  126,996     49.24%  50.76%
CC3    187,095  164,487     53.22%  46.78%
CC4    204,103  164,355     55.39%  44.61%
Kim Ogg

Kim Ogg

Ogg received 696,955 votes, which is about 11K fewer than Hillary Clinton, while Anderson drew 588,464 votes, or 42.5K more than Donald Trump. I believe the differences can be accounted for as Ogg not getting as many crossovers as Clinton, while Anderson picked up most of the Gary Johnson supporters. Compare the results from the Presidential race and the judicial races to get a feel for this. In particular, compare the Presidential numbers in HD134 to the same numbers above. Ogg got 4,765 fewer votes than Clinton in the district. Add to that the 4,044 Johnson votes for a total of 8,809, and then observe that Anderson did 8,131 votes better than Trump did. Not exact, but pretty close. There are some fudge factors as well – some of those Johnson voters were straight party Libertarian, Ogg may have received some Jill Stein votes, etc. It’s good enough for a back-of-the-envelope approximation, is what I’m saying.

Outside of HD134, Ogg consistently did about two points better across the county, with slightly bigger gains in more Republican districts. Basically, Ogg is to 2016 what Adrian Garcia was to 2008. Garcia maintained his status as Democratic pacesetter in 2012, and I think Ogg will have the chance to do that in 2020 if she does a good job and accomplishes the goals she has laid out. We have seen plenty of examples of county officials and candidates for county office drawing bipartisan support, on both sides. We’ve also seen examples of failed incumbents getting turned out in emphatic fashion. Good performance is good politics in these elections.

I’ll look at the other countywide races in the coming days. Are there any particular questions you’d like me to explore with this data? Let me know.

The state of the polls

Hillary Clinton

I’m just trying to get a handle on the numbers, with the idea of establishing some kind of guide for what to expect in the Presidential race in Texas. Bear with me.

The RCP average for the two-way Trump/Clinton race is 44.0 for Trump and 38.3 for Clinton. The FiveThirtyEight polling averages, which includes some other sources, come in at Trump 45.6, Clinton 37.6. However, once you apply the 538 secret sauce, you wind up with projected totals of 49.7% for Trump and 43.2% for Clinton.

RCP does not do this kind of modeling/forecasting – it’s a straight up polling average. As such, it can underestimate final totals, since it doesn’t try to guess what undecided voters may do. The 2012 RCP average for Texas had President Obama at 39.0 and Mitt Romney at 55.7; they finished at 41.4 and 57.2, respectively. Similarly, in 2008, Obama was averaging 40.5 and John McCain was at 53.5; the final numbers were 43.7 and 55.5. In other words, RCP underestimated Obama by three points in 2008 and by 2.5 points in 2012.

(I couldn’t find 538’s data for Texas in past years, so we’ll just skip that part of the analysis.)

There are so many variables in play here that I’ve been very reluctant to even begin to guess at what the final numbers might look like. Here are some of the things that factor in:

1. Overall turnout – Voter registration is at an all-time high, but that correlates weakly at best to turnout. However, the overall voting age population is way up, and even in a modest turnout-to-VAP scenario like we had in 2012, we’re easily looking at a half million or more extra voters than we’ve ever had, and that number could be quite a bit higher without setting a record for turnout as a share of the adult population. Nine million votes is not out of the question. I have to believe that beyond a certain point, extra voters will break Democratic. Where that point is, how blue they are, and how likely that is to happen, I have no idea.

2. Undecided voters – In 2008, the Obama/McCain share of the vote in the averages was 94.0%; in 2012, the Obama/Romney share was 94.7%. This year, it’s 82.3% for Trump and Clinton. Even adding in Johnson and Stein only gets you to 91.6%. That’s a lot more undecided voters. Do they show up? Which way do they lean? There’s a lot of room for candidates to gain ground here.

3. The third-party candidates – Just as a reminder, Gary Johnson and Jill Stein combined for 1.42% of the vote in Texas in 2012. Their RCP combined average is 9.3% right now. Poll numbers for third-party candidates are almost always overstated, often by quite a bit, but we don’t have any useful data for comparison from 2012. I’m sure there are some Republicans who will vote for Johnson over Trump, but nearly the entire state GOP establishment is in Trump’s corner, so it’s not like there’s an organized #NeverTrump movement. As with the undecided voters, there’s a lot of room for the Trump and Clinton numbers to change here if as has been the norm historically the L and G numbers are exaggerated. But if there was ever a year where maybe they’re not, you’d think this would be it.

4. The other polls – There are national polls showing Hillary Clinton with a double-digit lead. That’s a landslide by any measure, and if it’s what we get, it’s entirely possible that the polls we have for Texas are underestimating her by a considerable amount, as state polling tends to lag the national trends. The fact that the one most recent poll we have is also the closest one we’ve seen since that weird Washington Post poll suggests that possibility as well. We also know that there’s a lot of polling data that is not made public but from which we can make inferences based on the actions taken by the campaigns and other actors who have that data. Here, we have multiple suggestions of Republicans being worried about their turnout in Texas, plus Hillary Clinton actually running a week’s worth of ads in Texas, online and on TV. Draw your own conclusions about that.

5. Latino voters – This is baked into some of the other factors, but I keep being struck by the differences between what national polls say about Latino support for Donald Trump – in short, he may be lucky to get 20% of the Latino vote nationally, well below what Mitt Romney got – and what the state polls have said. The latter have generally had his support in the 30s, with Clinton in the 50s or low 60s. This may be a function of small sample sizes combined with excessive weighting to compensate, or it may simply indicate that Texas Latinos are different than Latinos elsewhere. Bear in mind that we have some data to indicate that lower-propensity Latino voters tend to be more Democratic than high-propensity Latino voters, which is a fancy way of saying that higher Latino turnout correlates with better Democratic performance among Latinos.

6. Crossover voters – Mark Bluenthal wrote yesterday that the key to Hillary Clinton’s increased national lead is that she has consolidated the Democratic vote better than Donald Trump has done with the Republican vote. Another way to put that is there are more Republicans who are voting for other candidates, including Clinton, than there are Democrats who are voting for other candidates. We see that in Texas as well, specifically in that UH poll, which showed ten percent of Rs voting for Clinton or Johnson, but only five percent of Ds voting for other candidates. Hillary Clinton’s better performance in Texas is two parts turnout – there are more Democrats and fewer Republicans voting than usual – and one part crossover voting. If that latter group is bigger than we think, that will affect the outcome.

In the end, I’m less interested in the margin between Trump and Clinton – given what we do know so far, barring anything unexpected that margin is going to be smaller than the McCain-Obama margin – as I am in the absolute totals. How many people actually vote for Hillary Clinton? The high-water mark is 3,528,633, set by Obama in 2008. Just on the increase in population alone, she could top that while receiving a lower percentage of the vote (for example, 3.6 million votes for Clinton out of 8.4 million total = 42.9%; Obama got 43.7%), but I would consider that a huge disappointment. Can she get to 3.8 million, or (be still my heart) 4 million? Can she reach 44 or even 45 percent, a level not reached since Jimmy Carter in 1976? I hope to have some small amount of clarity on this before voting concludes, but I doubt I’ll get much.

I think that about covers it. What it all means, I still don’t know. But when it’s all over and we’re doing the autopsy, these are the things I’ll want to look back on.

UH Hobby School: Trump 41, Clinton 38

Damn!

Hillary Clinton

Republican presidential nominee Donald Trump has one of his slimmest leads yet over Democratic rival Hillary Clinton in Texas, 41 percent to 38 percent, according to a new poll among registered voters. Trump’s support falls within the survey’s margin of error, which is plus- or minus 3 percent, meaning the race is a statistical dead heat.

Released Tuesday by the University of Houston’s Hobby School of Public Affairs, the poll also found that 16 percent of respondents were undecided or refused to answer. Four percent chose Libertarian Party candidate Gary Johnson and 1 percent selected Green Party nominee Jill Stein.

“The national gains Hillary Clinton has made in the last two weeks are evident in Texas, where she has closed within three points of Donald Trump,” said Richard Murray, political science professor and director of the Hobby School’s Survey Research Institute. “With such a close margin, the key question will be which candidate can actually get their supporters to the polls over the next three weeks.”

Trump’s lead jumps one point – to 4 percent – when the poll considered voters who said they were certain to vote on or before Election Day. Among independent voters in Texas, Clinton dominates Trump, 30 percent to 14 percent. The GOP candidate, however, won the support of a plurality of male respondents, 44 percent to Clinton’s 35 percent, while women support Clinton by a four-point margin, 42 percent to 38 percent.

There’s also another WaPo/Survey Monkey poll that shows Trump up 2, 48-46. That same poll had Clinton up 46-45 in early September. I’m not putting too much weight into this because its methodology is weird, but for those of you that saw news of this poll, I’m letting you know that I saw it as well. Here’s the info for the UH poll. I’ll quote from their intro:

The Hobby School of Public Affairs at the University of Houston completed live telephone interviews with 1,000 registered voters in Texas who reported they were certain (77 percent) or very likely to vote (23 percent) on or before election day on November 8, 2016.
Interviews were conducted by Consumer Research International between October 7 and October 15, 2016. Interviews were conducted on landline (54 percent) and cell phones (46percent).

The margin of error for the survey is +/- 3 percent (at the 95 percent confidence level). The survey was conducted under the supervision of co-directors Richard Murray, director of the Hobby School’s Survey Research Institute, and Robert Stein, research associate at the Hobby School.

The sample was weighted to reflect the racial and ethnic composition of the electorate based on the 2008 and 2012 presidential elections.

This is consistent with their earlier poll of Harris County that showed Clinton with a lead. As I said then, you can’t have Trump leading the state by less than half of Mitt Romney’s margin – hell, less than half of John McCain’s margin – and not see that reflected at the local level as well. One could argue that the composition of the Texas electorate this year will be more favorable to Democrats this year than 2012 and possibly 2008 were, but we’ll leave that discussion for after the election. In any event, a few quick points to make here:

– I can’t overstate how shocking it is to see a Republican candidate in Texas in a top-of-the-ticket race score only 41% in a poll in October. Forget the three-point margin for a minute, how is it that Trump so consistently can’t even come close to fifty percent?

– Even worse from Trump’s perspective is there’s not that much room for him to grow. He and Clinton have about the same share of their own voters – 80% of Dems say they support Clinton, 78% of GOPers are with Trump. More to the point, here aren’t a lot of undecided Republicans out there – twelve percent fall under None, Don’t Know, or Refused, while 14% of Dems are in one of those buckets. Trump does lose more of his own voters than Clinton does – ten percent of Republicans are voting for someone else (5% Johnson, 5% Clinton) while only five percent of Democrats are defecting (2% Trump, 2% Johnson, 1% Stein). Maybe some of them will come home for him.

– There’s a large share of undecided independents (29%), but 1) Clinton leads 30-14 among indies who do have a preference, 2) we don’t know how big a slice of the sample indies are, and 3) these are probably your least likely voters in the sample.

– Unfortunately, the provided poll data does not include breakdowns by age or by race. I’d bet that Clinton leads among voters under 50, as has been the case in other polls, but I can’t confirm that based on what we have.

FiveThirtyEight has this poll incorporated into their data set for Texas, but as of this writing Real Clear Politics had not noticed it. You should also read this 538 post about the poll and why Clinton is doing as well as she is in red states overall and Texas in particular.

I’ll have some more thoughts on the state of the polls tomorrow.

UH Hobby School (Harris County only): Clinton 43, Trump 34

I have many thoughts about this.

Hillary Clinton

Between September 1 and September 20, 2016, the University of Houston Hobby School of Public Affairs conducted a telephone survey of 550 Harris County registered voters as part of a larger Hobby School study on voter participation and engagement under the direction of Mark P. Jones, Renée Cross, and Jim Granato with Ching-Hsing Wang and Wyman Wan. The survey was based on a stratified probability design, including both landlines and cell phones. The survey was available in both English and Spanish using bilingual operators, and lasted an average of 13 minutes. The final data set was weighted by ethnicity, age, and gender to be representative of Harris County registered voters. The margin of error for the survey results is plus or minus 4 percent (at the 95 percent confidence level).

Harris County is the third most populous county in the nation and is home to approximately one fifth of Texas voters. The 2008 and 2012 presidential contests were very close in Harris County, with Barack Obama defeating his 2008 Republican rival, John McCain, 50.45 percent to 48.82 percent, and his 2012 Republican rival, Mitt Romney, 49.39 percent to 49.31 percent.

The Presidential Election: Vote Choice

Approximately two months prior to the November 8 presidential election, the survey finds Hillary Clinton with a lead over Donald Trump in Harris County, with the size of her lead varying depending on assumptions related to voter participation.


Candidate    ELV    LV    RV
============================
Clinton      43%   43%   42%
Trump        39%   34%   32%
Johnson       7%    9%    9%
Stein         1%    1%    2%
Unsure/NA    11%   13%   15%

We divided the survey respondents into three groups of decreasing size: all registered voters (Registered Voters), voters who indicated that it is very likely or extremely likely that they will vote this fall (Likely Voters), and voters who indicated that it is extremely likely they will vote this fall (Extremely Likely Voters).

[…]

The Harris County District Attorney and Sheriff Elections: Vote Choice

The survey also queried respondents regarding their vote preference in the key races for Harris County District Attorney and Harris County Sheriff. In both contests, even among the likely and extremely likely voters, there existed a substantial proportion of respondents who were unsure about their preference in these lower visibility contests (compared to the presidential race).

Among those likely voters who did have a preference, in the District Attorney race 29 percent favored Democratic challenger Kim Ogg over Republican incumbent Devon Anderson with 27 percent support. In the Sheriff contest, 33 percent of likely voters supported Republican incumbent Ron Hickman while 32 percent backed Democratic challenger Ed Gonzalez.

Just as was the case in the presidential race, the Republican candidates fared better when the population was limited to those most likely to cast a ballot this fall: the extremely likely voters. Among this population Anderson narrowly bested Ogg, 30 percent to 29 percent, while Hickman increased his lead over Gonzalez, 36 percent to 30 percent.

Poll data is here. My thoughts:

– I find the distinction between Likely Voters and Extremely Likely Voters to be silly. I mean, why stop at Extremely Likely? Why not ask if someone is Super Duper Double Dog No Backsies likely to vote? There’s a reason why every other poll under the sun uses Registered Voters and Likely Voters, and nothing else.

– The press release for this poll claims that “neither party can depend on presidential coattails” for the donwballot races. I don’t see how they can draw that conclusion from the given data set. For one thing, there’s no breakdown of the vote by partisan affiliation, nor is there any generic “which party’s candidates are you more likely to support in other races?” question. We have several cycles’ worth of actual results to suggest that Democrats have done a pretty good job of voting all the way down the ballot, while Republicans have only really done that in 2012, and even then not quite as faithfully as the Dems. Don’t make suppositions about what the topline numbers mean. Ask the questions that could tell you the answers!

– I picked the “Likely Voter” numbers for my post title, and if one went by that without knowing anything else, one would feel really good about Democratic downballot chances. But to be secure in those feelings, one would have to know those partisan crosstabs. If one were to choose an archetype for the Republican Who Will Not Vote For Donald Trump, it’s probably a voter in HD134. How many self-identified Republicans say they are voting for a candidate who is not Donald Trump? How many are just in the “Unsure/No Answer” column? That would tell me a whole lot more about the downballot races than the actual Ogg/Anderson and Gonzalez/Hickman results do.

– How many new voters are in this sample, and how many of them are deemed “Likely” or “Extremely Likely”? We know voter registration is up in Harris County – indeed, it’s way up around the state – so who are these voters and how were they accounted for in the poll?

– I would love to see a similar poll – with my suggested modifications, of course – for other area counties as well. Fort Bend would be my first choice for this, as they are close to being the kind of swing county that Harris has been. Is the year that Fort Bend Democrats break through, or does the Fort Bend GOP maintain its hold? A sneak preview of that answer would have been nice.

– If the polls of Texas that show Clinton trailing by significantly less than the margins by which President Obama lost the state are accurate, then it stands to reason that Clinton would be doing better in Harris County than Obama (who won the county by small margins) did. Similarly, if this poll of Harris County is accurate, then it stands to reason Clinton would be running more strongly statewide than Obama did. It’s like the relationship between national polls and state polls – they may not be tightly correlated, but one is unlikely to make a big move in a given direction without the other following suit.

ECPS: Trump 42, Clinton 36

Another state poll result to add to the list.

A new Emerson College poll in Texas, which has long been a Republican stronghold, shows Donald Trump leading Hillary Clinton by only 6 points (42% to 36%), with 10% of respondents voting for Libertarian Gary Johnson, 6% for the Green Party’s Jill Stein and 6% undecided. Trump’s lead is within the poll’s 3.6% margin of error. The poll was completed before Clinton announced she had been diagnosed with pneumonia.

As a point of comparison, in the 2012 presidential race, Mitt Romney soundly defeated Barack Obama by 16 points in the Lone Star State (57% to 41%). Texas has not voted for a Democratic presidential candidate since 1976, when Jimmy Carter beat GOP incumbent Gerald Ford.

Trump is having mixed success winning over voters who supported Texas Senator Ted Cruz in the state’s GOP primary last March. In the Emerson poll, 76% of Cruz voters plan to vote for Trump in November, while 10% support Johnson, 6% are unsure, and the remaining 7% are split between Stein and Clinton. Cruz won the Texas GOP primary, beating Trump 44% to 27%. Clinton is struggling to win over voters who supported Bernie Sanders in the Texas Democratic primary. Only 49% of Sanders supporters plan to vote for Clinton in November; 21% say they will vote for Stein.

Trump is winning the Independent vote 36% to 19% over Clinton, with Stein taking 16% and Johnson 15%. However, it appears Johnson is hurting Trump more than Clinton, taking 12% of registered Republican voters and 15% of Independents while only winning 3% of registered Democrats.

The text above is taken from the ECPS homepage, where you can also download an XLS with poll data. The top link is to a press release with more details than what I quoted above, but it was formatted in a funky way that made copying nearly impossible. The ECPS did poll Texas in 2014, though theirs was one of the least accurate results. It was in March, for what that’s worth, but they were way off. The Libertarian/Green numbers are quite a bit higher than they were in the PPP poll, but that’s the only point of comparison we have. The one thing we can say is that this is consistent with other results showing a less-than-usual-size lead for Trump. The Texas Lyceum has poll numbers coming out today, so we’ll see where they fit in. Link via The Hill.

PPP: Trump 44, Clinton 38

We knew this poll was coming, and it confirms what we have been seeing elsewhere.

PPP’s new Texas poll finds a relatively tight race, at least on the curve of recent Presidential election results in the state. Donald Trump leads with 44% to 38% for Hillary Clinton, 6% for Gary Johnson, 2% for Jill Stein, and less than half a percent (0) for Evan McMullin. In a head to head contest Trump leads Clinton 50-44 in the state, which Mitt Romney won by 16 points in 2012.

A Democratic victory in Texas this year remains a stretch but within the numbers there are signs of Democrats being positioned to become seriously competitive there in the years ahead. Trump’s lead is based entirely on his holding a 63-33 advantage among seniors. With voters under 65, Clinton leads him 49-45. And when you look just specifically at voters under 45, Clinton leads Trump 60-35. Older voters are overwhelmingly responsible for the Republican advantage in Texas, and generational change is likely to help Democrats become more competitive.

A big piece of that generational change is the increasing racial diversity of the electorate in Texas. Trump has a 69/25 lead with white voters but the reason the state’s so competitive overall is that among non-white voters Clinton has a 73-21 lead, including a 68-27 edge with the state’s booming Hispanic population.

Clinton’s unpopular in Texas, as you would expect, with a 36/59 favorability rating. But Trump’s not a whole lot better off with only 40% of voters seeing him favorably to 53% with a negative opinion. The tax return issue continues to plague Trump with 64% of voters thinking he needs to release his returns to only 25% who don’t think it’s necessary for him to. Even Trump’s supporters, by a 43/41 spread, think he should release them. Another issue that has the potential to cause Trump problems down the road is if he refuses to participate in the debates as scheduled. 77% of voters think he needs to do that to only 14% who don’t think he needs to and among his own supporters there’s an even stronger sentiment- 82/12- that Trump needs to participate. If Trump is stubborn about that it could cause the bottom to fall out on his support even further.

The full polling memo is here. A few thoughts:

– If President Obama were running against Trump, he would be trailing by only two points, 48-46. Let that sink in for a moment. Obama’s approval/disapproval numbers are 42/54, which is a tiny bit better than Trump’s favorable/unfavorable numbers of 40/53.

– This is Clinton’s high-water mark in Texas so far, which puts her in Obama’s polling range from 2012, while Trump continues to lag way behind Romney’s poll numbers. All this is of course consistent with the race being closer now than it was four years ago, but it’s not yet suggestive of Clinton doing better than Obama did. PPP did poll Trump/Clinton straight up, and the result there was 50/44, which is more in line with her exceeding the 2012 level, but it’s not a two-candidate race, so all that shows is that she has the potential to grow.

– Trump’s numbers among white voters are closer to what Romney got, but still a few points behind it. The 69-25 figure cited about is from the Trump/Clinton two-person choice; with all four candidates listed he leads by a more modest 64-21 among whites. He does have the potential to grow as well, as Gary Johnson gets 5% and 8% are undecided. It’s also well within reason that this just portends a decrease in Republican turnout. It’s still too early to say.

– This is the first poll of Texas I’ve seen that includes all four candidates. Johnson’s 6% and Stein’s 2% would significantly exceed their numbers from 2012 if they hold up, but as we know from national polling data, third party numbers tend to be exaggerated in polls compared to what ultimately happens at the ballot box. This is a weird year, of course, so one wants to tread carefully in making any broad claims. Unfortunately, there’s no basis for comparison in the 2012 numbers, as none of the polls from September onward included Johnson or Stein, who represented the L and G parties that year as well. The one data point we have is in the UT/Trib poll from October 12, 2012, in the Senate race (see page 3), where Lib candidate John Jay Myers polled at 3% and Green David Collins was at 2%. In actual voting, Myers wound up with 2.06% and Collins with 0.86%, higher than their Presidential counterparts but lower than their poll totals. Make of that what you will.

– The age split is encouraging from a Democratic perspective, but old people vote, and a 20-year timeline as suggested in the polling memo is forever. The Democrats’ base problem remains the same – base turnout hasn’t grown, at least in non-Presidential years, since 2002. If Democratic turnout increases this year, then perhaps there is some hope to get an increase in 2018. Of course, one could have made the same claim after 2008, and we know how that went. Solving the base turnout issue is the Hilbert problem for Texas Dems.

PDiddie has more.

Three polling-related observations

This story is about the tough spot that Donald Trump is putting professional Latino Republicans in. I have no sympathy for any of them, of course, but what caught my eye in this article was this little nugget:

“I’m not on the anti-Trump movement like some of my colleagues who I talk to every day, but I’m far from an endorsement,” said Leslie Sanchez, a Republican commentator for CBS News, echoing the views of many of these notable Latino conservatives who are skeptical of Trump.

None of these top figures said they will support presumptive Democratic presidential nominee Hillary Clinton. But they are highly critical of Trump, both because they are offended by his “abhorrent racist rhetoric” (as Sanchez put it) and demoralized by his campaign’s poor engagement with broad swaths of the electorate — including, but not limited to, Latinos.

“Trump is having a maddening effect of turning solid red states purple. Areas that should not have gone in that direction for the next 30 years, he’s managed to do in about four months,” Sanchez said, citing survey work she did in Texas in April.

“Some of the Republican pollsters I talked to there said this should not be happening for at least a generation by natural demographic growth,” she added. “Trump has accelerated that by underperforming in traditionally conservative Latino households.”

In other words, the polls that aren’t being publicly released are in general agreement with the polls that have been publicly released, which is to say they agree that Donald Trump has a much more modest lead over Hillary Clinton in Texas than we are used to seeing. This also suggests that the data we haven’t yet seen in the public polls, about how the vote breaks down along demographic lines, that Trump really has been galvanizing Latino voters, in a way that could very well shake things up at least a little here in Texas. You have no idea how much I’d love to see the data that Leslie Sanchez is talking about.

Then there’s this, in a story about how even with “tighter than expected” polls, the Clinton campaign has no current plans to make a push in Texas.

Texas Republicans, of all groups, are perhaps the most enthused over the idea that the state could be in play in the fall.

Republicans say they would love to see Democrats drawn into what they view as a hopeless money pit. But also, within a state GOP torn over its own nominee, a Clinton offensive could be just what it takes to rally an otherwise morose group.

“The quickest way to activate disenfranchised GOP donors who won’t give to Trump would be an aggressive effort by Democrats to win the state,” said Brian Haley, a Texan who was a top fundraiser in two previous GOP presidential campaigns.

Abbott is one of multiple Republicans who have already sent fundraising emails on the notion.

“She has already made it known that winning Texas will be a focus of her campaign,” Abbott campaign director John Jackson wrote in a recent missive, referring to Clinton. “It’s clear that Hillary will not only continue Obama’s liberal leadership—she will be even worse!”

Hey, Trump may be a racist con man who took days to even put out a tweet about the SCOTUS HB2 decision, doesn’t really care where anyone goes to the bathroom, and scares the bejeezus out of our corporate overlords, but at least he’s not Hillary, am I right? That’s a remarkable admission of weakness, one that lends credence to the idea that Republican turnout could be lower than we have seen in recent Presidential elections. That’s got to be a scary prospect for various downballot Republican candidates, including an especially all the countywide Republicans in Harris County.

Finally, here’s the initial FiveThirtyEight view of Texas, which has Trump up by five (!) points, 48.5% to 43.5%, on Hillary Clinton. Here’s how that might break down:


Candidate     Total Votes    Pct
================================
Trump           3,880,000  48.5%
Clinton         3,480,000  43.5%
Johnson           528,000   6.6%
Stein             112,000   1.4%

I’m assuming turnout of 8 million, as was the case in 2008 and 2012. Jill Stein’s numbers are not included on the 538 page, as none of the recent polls included her by name, so I just assigned her the remaining percentage. Under this scenario, Clinton exceeds Obama’s 2012 vote total by about 170,000 votes though she falls short of 2008 by about 50,000. Trump falls well short of Mitt Romney, who drew 700,000 more votes than this, while both Johnson and Stein far exceed their 2012 numbers – I mean, the total of Johnson plus Stein in 2012 was a hair over 113,000, or just barely more than what Stein is projected to have by herself here. That’s quite a significant change, if it holds.

Now, I think Clinton will do better than 3.5 million votes, and I have a hard time imagining Trump getting fewer than 4 million. I figure the difference will come in part from total turnout being higher than 8 million, and in part from the totals for Johnson and Stein being less than what is projected here. I could be wrong about either or both of those – this is for sure a weird year – but for now at least, all the data we have points to this being a closer, possibly much closer, Presidential race in Texas than we have seen recently. Now we need to wait and see what the trendlines look like.

Precinct analysis: Third parties revisited

Politico has a question.

Is Austin’s Travis County the nation’s Libertarian Party stronghold?

The co-founders of a Libertarian political action committee based there make that case, arguing that the Texas locale is the “most Libertarian large county in America.”

Wes Benedict and Arthur DiBianca of Libertarian Booster PAC note that 31 Libertarian candidates were on the Travis County ballot this year, more than any other county in America. Among the other stats they cite:

  • Libertarian presidential nominee Gary Johnson won 2.7% there, his highest percentage nationwide for large counties outside his home state of New Mexico.
  • Four Libertarians got over 40% of the vote for the portion of their district within Travis County
  • The current chairman of the national party, Geoffrey Neale, lives in Travis County, and 2004 Libertarian presidential nominee Michael Badnarik had previously run for office as a Libertarian in Travis County, and his presidential campaign headquarters were located in Travis County.

Their argument makes some sense – certainly there’s a strong libertarian bent in tech-heavy communities like Travis County.

We reviewed this before, and indeed Travis was the best county in the state for Johnson. It was also the second-Greenest county in the state, so I think it’s more a matter of iconoclasm than Libertarianism. For the record, those four Libertarians with over 40% of the vote were the candidate in CD17, plus three statewide judicial candidates. All were the sole opponents to Republicans, and I say that’s more about anti-Republicanism than pro-Libertarianism. Besides, as we’ve also seen, Libertarian Railroad Commissioner candidate Jaime Perez did better than that in several Latino-heavy counties, winning a majority of the vote in Maverick and Webb even though he also had a Green opponent. The simple fact is that in areas that are dominated by one party, Ls and Gs serve as the default option in races where that party isn’t represented. This doesn’t detract from the claim that Travis County has a large number of people willing to push the L button – relatively speaking, anyway – just that one needs to be aware of the qualifications.

Reading that story made me decide to go back to the Harris County precinct data to see where the Lib and Green friendly areas were. I broke this down into three sets of races, since obviously not every race featured an L and/or a G. The first set is the top of the ticket, the Presidential race and the Senate race. Here’s how the L and G candidates did in those races in each State Rep district:

Dist Johnson Stein J Pct S Pct Myers Collins M Pct C Pct ================================================================ 126 531 160 0.92% 0.28% 903 424 1.58% 0.74% 127 587 208 0.88% 0.31% 977 491 1.48% 0.74% 128 450 129 0.80% 0.23% 791 325 1.43% 0.59% 129 781 284 1.21% 0.44% 1,216 582 1.90% 0.91% 130 693 199 0.99% 0.29% 1,143 480 1.65% 0.69% 131 196 93 0.45% 0.21% 384 297 0.89% 0.69% 132 549 151 1.03% 0.28% 908 405 1.72% 0.77% 133 815 241 1.12% 0.33% 1,187 561 1.65% 0.78% 134 1,114 436 1.34% 0.53% 1,697 898 2.08% 1.10% 135 548 162 1.01% 0.30% 881 447 1.63% 0.83% 137 289 113 1.17% 0.46% 486 322 2.01% 1.33% 138 540 161 1.17% 0.35% 795 403 1.73% 0.88% 139 260 132 0.51% 0.26% 513 392 1.01% 0.77% 140 152 66 0.64% 0.28% 223 215 0.96% 0.92% 141 142 56 0.37% 0.15% 282 169 0.74% 0.45% 142 166 93 0.40% 0.22% 352 271 0.85% 0.66% 143 189 104 0.62% 0.34% 336 304 1.11% 1.01% 144 238 90 0.98% 0.37% 371 241 1.55% 1.01% 145 273 161 0.92% 0.54% 481 342 1.65% 1.17% 146 376 190 0.74% 0.38% 624 438 1.25% 0.88% 147 583 304 1.06% 0.56% 944 685 1.75% 1.27% 148 640 282 1.62% 0.71% 947 553 2.43% 1.42% 149 347 131 0.80% 0.30% 594 358 1.40% 0.84% 150 598 157 0.92% 0.24% 976 478 1.51% 0.74%

The percentages here are calculated from the four-candidate totals. For comparison purposes, Libertarian Gary Johnson had 0.93% overall in Harris County, and Green Jill Stein had 0.35%; in the Senate races, John Jay Myers had 1.54% and David Collins had 0.86%. Everyone who had HD148 as the most third-party-friendly district in Harris County, come forward and collect your winnings. You would have guessed HD134, am I right? That district isn’t as Montrose-y as it used to be, which I suspect is the reason for its runnerup status. At the other end of the scale, note how third-party-resistant the African-American districts were – all but HD147 were well below the countywide levels of L and G support. Republican districts in general were also third-party-averse, with only HDs 134 and 129 overperforming for them. This is what you should expect for Presidential and Senate races – as the highest-profile races, and the ones that tend to have the fewest undervotes, people are going to stick with their home teams unless they’re crossing over for a specific reason. Once we get past these races, however, it’s a different story. There were two other statewide races that had an R, a D, an L, and a G – the Railroad Commissioner race that featured Christi Craddick, Dale Henry, Vivekananda (Vik) Wall, and Chris Kennedy; and the Supreme Court race between Nathan Hecht, Michele Petty, Mark Ash, and Jim Chisholm. Here’s how that played out for the L and G candidates.

Dist Wall Kennedy W Pct K Pct Ash Chisholm A Pct C Pct ================================================================ 126 951 758 1.69% 1.35% 1,240 530 2.22% 0.95% 127 1,060 922 1.63% 1.42% 1,438 620 2.22% 0.96% 128 785 757 1.44% 1.39% 1,117 512 2.05% 0.94% 129 1,387 1,174 2.21% 1.87% 1,677 727 2.69% 1.17% 130 1,183 861 1.74% 1.26% 1,668 607 2.46% 0.89% 131 354 550 0.83% 1.28% 452 298 1.06% 0.70% 132 906 751 1.73% 1.44% 1,207 495 2.32% 0.95% 133 1,307 1,036 1.85% 1.47% 1,674 676 2.40% 0.97% 134 1,937 1,784 2.46% 2.27% 2,373 973 3.04% 1.24% 135 964 724 1.81% 1.36% 1,187 473 2.25% 0.90% 137 494 525 2.07% 2.20% 578 317 2.44% 1.34% 138 884 748 1.96% 1.66% 1,082 490 2.42% 1.09% 139 518 744 1.03% 1.47% 676 527 1.34% 1.05% 140 213 447 0.92% 1.94% 318 307 1.38% 1.34% 141 250 362 0.66% 0.96% 332 253 0.88% 0.67% 142 347 405 0.85% 0.99% 442 297 1.08% 0.73% 143 287 611 0.96% 2.05% 448 419 1.51% 1.42% 144 361 556 1.53% 2.35% 502 345 2.13% 1.46% 145 501 795 1.74% 2.77% 690 515 2.41% 1.80% 146 626 810 1.27% 1.65% 748 433 1.53% 0.88% 147 1,022 1,197 1.92% 2.25% 1,229 719 2.32% 1.36% 148 941 1,319 2.47% 3.47% 1,319 798 3.49% 2.11% 149 607 637 1.44% 1.51% 725 353 1.74% 0.85% 150 1,093 904 1.71% 1.42% 1,475 613 2.32% 0.97%

These results just fascinate me. The total number of L and G votes in each race was nearly the same – 38,476 in the RRC race, 36,993 in the Supreme Court race – but the distribution was completely different. Wall (19,036 for 1.65%) and Kennedy (19,440 for 1.68%) basically tied, while Ash (24,665 for 2.14%) doubled up Chisholm (12.328 for 1.07%). Look in each district, and you can basically see some number of people who voted for Kennedy in one race voting for Ash in the other? You may wonder why this is. It’s possible that Christi Craddick was more acceptable, and Dale Henry less so, to the “swing” third-party voters that otherwise vote R and D, with the reverse being true for Nathan Hecht and Michele Petty. There is something to that – Henry is on the verge of morphing into Gene Kelly, while Nathan Hecht has ethical baggage and nearly foisted Harriet Miers onto an unsuspecting US Supreme Court. The total number of voters involved here is tiny enough to include the possibility that they’re sophisticated enough to make such judgments. Personally, I think it’s more likely that we’re looking at roughly the same voters in each race, and that people picked Chris Kennedy over Vik Wall as their “none of the above” choice because Wall had a funny-sounding name. What do you think?

At the county level there were no four-way races, but there was a Green candidate running for Sheriff (Remington Alessi) and a Libertarian candidate running for Tax Assessor (Jesse Hopson). Here’s how they did in their respective races.

Dist Alessi A Pct Hopson H Pct =================================== 126 866 1.54% 1,291 2.30% 127 1,180 1.82% 1,632 2.51% 128 851 1.55% 1,156 2.12% 129 1,428 2.27% 1,866 2.98% 130 1,027 1.50% 1,695 2.50% 131 603 1.41% 534 1.25% 132 903 1.73% 1,294 2.49% 133 1,317 1.88% 1,804 2.58% 134 1,952 2.49% 2,458 3.15% 135 894 1.68% 1,279 2.42% 137 622 2.61% 695 2.93% 138 868 1.92% 1,225 2.73% 139 801 1.58% 844 1.68% 140 300 1.28% 357 1.55% 141 373 0.99% 366 0.97% 142 478 1.16% 497 1.21% 143 450 1.49% 488 1.64% 144 435 1.83% 524 2.22% 145 697 2.40% 777 2.71% 146 927 1.89% 895 1.83% 147 1,383 2.60% 1,369 2.58% 148 1,226 3.19% 1,437 3.79% 149 671 1.60% 834 1.99% 150 1,070 1.68% 1,547 2.44%

These are two different races, so Alessi and Hopson’s numbers aren’t directly comparable, but it’s still interesting to see them side by side. I take this as a data point in favor of the hypothesis that Libertarian candidates tend to draw support from Republicans; based on these numbers, they do so in somewhat greater quantity than Greens do from Dems. I wouldn’t draw too broad a conclusion from this sample – there was a lot of money in the Sheriff’s race, and that tends to minimize third party support. Then again, Alessi did actually campaign – if Hopson did, it was invisible to me – and there was some criticism of Sheriff Garcia from the left, so one might expect him to do better than a generic “none of the above” candidate. Make of it what you will.

I think that about runs me out of ideas for precinct analyses. One never knows where inspiration may strike, though, so don’t quote me on that. And there’s always next year, which is to say this year now. Until then, or until I come up with another angle at which to examine the data, we’ll call it a wrap on 2012.

The third parties

While I work my way through the precinct data in Harris County, we can keep looking at the county data for Texas from last week’s election. Here are the top and bottom ten counties by percentage of the vote for Libertarian Presidential candidate Gary Johnson:

County Johnson % County Johnson % ============================================== Travis 2.72% Throckmorton 0.00% Hays 2.46% Brooks 0.25% Brewster 2.35% Kimble 0.32% Williamson 2.22% Lipscomb 0.34% Jeff Davis 2.02% Parmer 0.35% Bastrop 1.90% Refugio 0.37% Brazos 1.87% Bailey 0.39% Caldwell 1.84% Zapata 0.40% Terrell 1.80% Dimmit 0.41% Blanco 1.71% Deaf Smith 0.42%

Travis County is a hotbed for third-party voting, and apparently that fever has spread to some of its neighbors. My guess is that more people there consider their Presidential vote to be meaningless, so they feel freer to use it for personal expression. I will add that the #12 county on the “most Libertarian” list is Loving County, where Johnson collected 1.56% of the vote. Of course, there were only 64 total votes cast in Loving County (2010 population: 82 residents), so that 1.56% represents exactly one voter. How would you like to say that you’re the only voter of your kind in your entire county? For what it’s worth, Travis was the only blue county in the top ten, while Brooks, Zapata, and Dimmit are all deep-blue Rio Grande counties. Only Blanco County was more than 70% red, while five of the top ten counties were between 50% and 60% Republican; of the bottom ten counties, all but Refugio among the Republican counties were at least 70% so.

By the way, Johnson did something that no other Libertarian Presidential candidate had ever done in Texas: He got more than 1% of the vote, 1.10% to be exact.

Here are the same lists for Green Party candidate Jill Stein:

County Stein % County Stein % ============================================== Brewster 0.91% Loving 0.00% Travis 0.87% Hudspeth 0.00% Borden 0.83% Hemphill 0.00% Foard 0.81% McMullen 0.00% Presidio 0.66% Oldham 0.00% Dallam 0.65% Sherman 0.00% Kinney 0.63% King 0.00% Delta 0.59% Kenedy 0.00% Jeff Davis 0.59% Floyd 0.00% Blanco 0.58% Martin 0.00%

Note: that’s “Dallam” County in Stein’s top ten list, not “Dallas”. There is Travis again, giving Stein not just a relatively high percentage but also a huge share of her total vote: The 3,360 Greenies in Travis County represented nearly one-seventh of Stein’s final total of 24,450 votes. Only three other counties appeared on both Stein and Johnson’s lists, and outside of Travis they’re all small to tiny; besides Brewster (35 votes for Stein) and Blanco (29 votes), none provided more than 12 Green votes. Serendipitously, there were exactly ten counties that pitched a Green shutout. Hays (0.57%, #11 on the list) and Jefferson (0.13%) were the high and low Green scorers among counties with at least 100,000 registered voters, while El Paso (0.37%) and Fort Bend (0.21%) were at the top and bottom of counties where at least 100,000 votes were cast.

And finally, the same lists for John Jay Myers and David Collins, the Libertarian and Green candidates for Senate, respectively.

County Myers % County Myers % ============================================== Cottle 4.67% Glasscock 0.55% Brewster 4.62% Brooks 0.64% Travis 4.30% Sutton 0.70% Hays 4.21% Martin 0.71% Williamson 4.09% Jim Hogg 0.81% Hudspeth 3.96% King 0.82% Terrell 3.75% Dickens 0.83% Bastrop 3.53% Wheeler 0.83% Culberson 3.42% Rusk 0.85% Kenedy 3.29% Jefferson 0.96% County Collins % County Collins % ============================================== Maverick 2.34% Glasscock 0.00% Johnson 2.27% King 0.00% Presidio 2.09% Floyd 0.24% Jeff Davis 1.95% Borden 0.29% Brewster 1.87% Hartley 0.32% Culberson 1.85% Madison 0.32% Webb 1.84% Garza 0.34% Willacy 1.71% Hemphill 0.34% Loving 1.67% Lamb 0.35% Zapata 1.65% Camp 0.37%

There’s a lot of overlap between Johnson and Myers’ top lists – Hudspeth was #11 for Johnson, and Culberson was #26. Cottle and Kenedy are both tiny counties, and the differences are small but pronounced given the minimal number of voters. 31 people in Cottle votes Myers, but only 5 for Johnson, while in Kenedy it was 5 for Myers and 1 for Johnson. As for Collins, just as there was one Libertarian in Loving County, so is there one Green there. I wonder if they know each other.