Off the Kuff Rotating Header Image

turnout

Quinnipiac: Cruz 47, O’Rourke 44

Pretty good poll result, with the ever-present proviso that it’s just one result.

Rep. Beto O’Rourke

The closely watched U.S. Senate race in Texas is too close to call, with 47 percent for Republican incumbent Sen. Ted Cruz and 44 percent for U.S. Rep. Beto O’Rourke, his Democratic challenger, according to a Quinnipiac University Poll released today.

There are wide party, gender, age and racial gaps, the independent Quinnipiac (KWIN- uh-pe-ack) University Poll finds:

  • O’Rourke gets 87 – 9 percent support from Democrats and 51 – 37 percent backing from independent voters, as Republicans go to Cruz 88 – 6 percent;
  • Men back Cruz 51 – 40 percent, while women go 47 percent for O’Rourke and 43 percent for Cruz;
  • Voters 18 to 34 years old go Democratic 50 – 34 percent, while voters over 65 years old go Republican 50 – 43 percent;
  • White voters back Cruz 59 – 34 percent, as O’Rourke leads 78 – 18 percent among black voters and 51 – 33 percent among Hispanic voters.
  • Sen. Cruz gets lackluster grades, including a 47 – 45 percent job approval rating and a 46 – 44 percent favorability rating. O’Rourke gets a 30 – 16 percent favorability rating, but 53 percent of Texas voters don’t know enough about him to form an opinion of him.
  • Texas voters “like Ted Cruz as a person” 47 – 38 percent. Voters “like Beto O’Rourke as a person” 40 – 13 percent with 47 percent undecided.

“Democrats have had a target on Sen. Ted Cruz’s back, and they may be hitting the mark. Once expected to ‘cruise’ to reelection, the incumbent is in a tight race with Democratic U.S. Rep. Beto O’Rourke,” said Peter A. Brown, assistant director of the Quinnipiac Poll.

“The key may well be independent voters. O’Rourke’s 51 – 37 percent lead among that group is key to his standing today. But Texas remains a strong GOP state so O’Rourke will need the independent strength to pull the upset.”

[…]

In the Texas governor’s race, Republican incumbent Gov. Greg Abbott tops former Dallas County Sheriff Lupe Valdez 49 – 40 percent and leads entrepreneur Andrew White 48 – 41 percent.

Voters approve 54 – 33 percent of the job Gov. Abbott is doing and give him a 51 – 33 percent favorability. His challengers are largely unknown as 65 percent don’t know enough about Valdez to form an opinion of her and 72 percent don’t know enough about White.

“Gov. Greg Abbott has a modest lead over each of the two people vying for the Democratic nomination. But what is significant is that governors with 54 percent job approval ratings rarely lose,” Brown said.

Texas voters disapprove 52 – 43 percent of the job President Donald Trump is doing. Republicans approve 85 – 13 percent. Disapproval is 90 – 8 percent among Democrats and 64 – 28 percent among independent voters.

President Trump will not be an important factor in their U.S. Senate vote, 43 percent of Texas voters say, while 26 percent say their vote will be more to express support for Trump and 27 percent say their vote will be more to express opposition.

The poll was of “1,029 Texas voters”, which I assume means registered voters. For comparison, the earlier poll results we have re:

PPP: Cruz 45, O’Rourke 37
Wilson Perkins: Cruz 52, O’Rourke 34

Not too surprisingly, this one has one of the lower approval ratings for Donald Trump, which is no doubt correlated to the overall numbers. What stands out the most to me is that all three Democratic candidates score at least forty percent even though their name ID is quite low – in the questions about favorability, the “haven’t heard enough about them” choice is 53% for Beto, 65% for Valdez, and 72% for White. I’d usually expect that to be in conjunction with a “vote for” number at best in the low 30s. The fact that it’s higher suggests to me this is another piece of evidence for the higher level of engagement.

Another thing that would suggest more engagement will be poll numbers that are consistently at least in the high thirties and forties. That may not sound like much, but look on the sidebar at the numbers from 2014 and 2016. I did a little figuring, and I found that Hillary Clinton had a 38.53% poll average across 19 polls,with a high score 46 (twice) and a low score 30. Wendy Davis in 2014 had a 36.87% poll average across 15 polls. Her high score was 42, and her low score was 32 (twice). One poll number above those totals doesn’t mean anything – remember, the first two results we saw in the Senate race had Beto and 34 and 37 – but a string of them would.

I say all that as a way of trying to put this into perspective. I’ve seen some good poll results before – again, look at that sidebar. It’s just that for each good one, there are four or five not so good ones, so we fixate on the good ones. These are good numbers, but if you read the whole poll memo, you see that Cruz beats O’Rourke in all the “who do you prefer on this issue” questions, and Abbott as noted has a shiny approval rating. Plus, you know, we Texas Democrats don’t exactly have a track record for turning out in the off years. By all means, take this as something positive, but for crying out loud don’t take it as gospel. The Observer, the DMN, RG Ratcliffe, Mother Jones and the Trib have more.

Are Texas Republicans really worried about Trump for November?

I mean, I guess they are. They’ve seen what has happened around the country, in other elections and other red places. I don’t know how worried they are, or how worried they should be.

This is a Red State where Trump remains popular among most Republicans and even some Democrats, a state that he won in 2016 by a slimmer margin that Republican Mitt Romney had four years before. And while political consultants for both parties agree he will be a drag on Texas Republicans in November, the growing question is just how much.

Fact: For the first time in a decade, Texas Republicans are having to worry about the vote drag their president could have on elections, much as Democrats suffered through eight years of Barack Obama, whom Texas Republicans loved to hate.

“Trump won’t be as much of an effect as he is in some northern states, but he will have impact here,” said Mark Jones, a Rice University political science professor who has been monitoring the “Trump effect” in Texas races for months. “In some of the down ballot races in Texas, the strategy you’re seeing is a modest embrace of Trump. … Republicans with strong brands like Abbott are not going to tarnish themselves by agreeing with him.”

Even so, Abbott confidantes privately acknowledge he likely will be re-elected by a slimmer margin than four years ago, when he beat Democratic rising star Wendy Davis by 20 points. They blame Trump.

“It’s not a question about whether there will be a Trump drag in Texas, the only question right now is how big it will be — and how many Republican incumbents will be in trouble,” said Harold Cook, a political consultant who is a former executive director of the state Democratic party.

“If you’re a member of Congress or state Senate or House incumbent who has a credible Democratic opponent, and you’re in districts that went for Trump less than 7 or 8 … points, you better be out working your ass off to get re-elected.”

I don’t disagree with any of this, but it’s hard to contextualize. The last two midterm elections in which there was a Republican President were 2002, when Dubya Bush was very popular, and 2006, when his approval rating had crashed and Rick Perry was fending off multiple challengers. The former was a great year for the GOP here and the latter was a lousy year for them (downballot, at least), but neither is a great comparison for this year. In the limited polling we have, Trump’s approval ratings are good with Republicans, terrible with Democrats, and not great with independents. How that compares with 2002 and 2006, I couldn’t say.

I think everyone expects Democratic turnout to be up this fall from previous midterms. The problem is that Democratic turnout has been so consistently crappy in previous midterms that there’s a lot of room for turnout to improve without making that much difference. We’d need to boost our baseline offyear performance by about fifty percent, to get to around 2.7 million, to really see significant gains. The good news is that basically everything would be competitive at that point, from the statewide races to a half dozen or more Congressional seats to enough legislative seats to make 2019 look like 2009. The bad news is that we’ve never come withing hailing distance of such a total, and Republicans have every reason to feel confident that it’s all just talk since we’ve never done anything like it.

So I just don’t know. I’m optimistic in general, and I really think good things will happen in Harris County and other areas that have been trending blue. I wish I knew how to quantify it, and I wish I knew how confident to feel beyond that.

The math on redistricting

It’s not just that Republicans drew themselves a favorable map. It’s that they drew a durable favorable map.

Thanks to those very effective Republican redistricting maps, Texas Democrats would have to improve their statewide election results by more than 10 percentage points to gain more than one seat in the 36-member U.S. House delegation, according to a report from the non-profit Brennan Center for Justice.

The political maps in Texas and elsewhere across the country could ultimately protect the Republican majority in the U.S. House even if it turns out to be an otherwise mediocre midterm election for the president’s political party.

Overall, Republicans have a 24-seat advantage in the U.S. House. Democrats have an advantage over Republicans in recent polling, the report says, but gerrymandering makes a party switch much less likely. To win two dozen seats, by Brennan’s figuring, Democrats would have to win the national popular vote by 11 percentage points.

“Even a strong blue wave would crash against a wall of gerrymandered maps,” the report says.

The 2016 elections put 25 Republicans and 11 Democrats in the state’s delegation to the U.S. House. Democrats got 42 percent of the state’s votes that year, according to the report’s authors. A modest improvement in the share — as little as 2 percent — could move a seat from the Republicans to the Democrats. It’s not hard to figure out that the 23rd Congressional District that runs along the border is what’s in play here; it’s the only true swing seat in the state, regularly primed to go to whichever party is having the better election year.

But here’s the house-on-stilts aspect to the maps. According to the Brennan Center’s projections, the Democrats could improve their statewide vote share by as much as 7 points — to 49 percent — and that’s still the only congressional seat they would pick up.

Listen carefully right now and you’ll hear protests from other parts of the state, like CD-32 in Dallas and CD-7 in Houston, where optimistic Democratic challengers are vying to unseat Pete Sessions and John Culberson. They might be right. The study isn’t trying to predict races. It’s trying to show how strongly the Republicans cemented their advantage in Texas, given normal conditions. Actual mileage may vary.

It would take a tsunami — a double-digit leap in Democrat’s percentage share — to gain more than a single seat in Texas. Something like that would still leave the Republicans in the majority, but it would be a 19-17 advantage instead of the 25-11 edge they have now. “For Democrats to win more than one-third of seats under the 2011 Texas map, they would need to win close to half the vote,” the report says.

The report is here and the executive summary is here. It looks at multiple states and is worth reading for its methodology and thoroughness. One way to look at this is that if Democrats can get to fifty percent of the statewide vote, then there are an awful lot of Congressional seats that would be poised to topple in their direction. Republicans drew this map on the quite reasonable so far assumption that Dems will not get to a majority of the statewide vote, but if that assumption were to fail they’d go from a trickle to a flood in a big hurry.

If it’s too daunting to think about like that, the way I look at it is that the magic number for Democrats is 2.7 million, which is to say 75% of their 2016 vote total. I’ve noodled around with the numbers before now, and that’s where things get interesting, in multiple districts. Not just Congressional districts, either – State Senate seats start to flip as well. On the one hand, that’s a huge increase over the usual off-year total. On the other hand, it’s asking people who have at least some history of voting to vote this year. Democrats gained 800,000 votes from 2004 to 2008, so a big jump can happen. What this report is saying, and I agree with it, is that this is what needs to happen. Are we up to it?

On Latino primary participation

Time for some numbers.

The predictions about Harris County Latinos becoming more engaged in the recent mid-term primary were right: The number of Latino voters who cast their ballot more than doubled compared to the previous primary of the same kind, in 2014, with an overwhelming majority voting in the Democratic election. Experts attribute the increase to factors such as the national political climate polarized by the immigration discussion and a high number of Latino candidates, among others.

According to the office of Harris County Clerk Stan Stanart, 36,184 Spanish-surnamed voters voted in the 2018 primary election compared to 13,721 in 2014.

The increase in turnout –which is the percentage of eligible voters who cast a ballot in a particular election– also doubled: 491,912 Spanish-surnamed voters were registered in the county as of February, which means the turnout was close to 7.4 percent, compared to the 370,293 Spanish-surnamed voters who were registered in the county in 2014, which means the turnout that year was 3.7 percent.

The break down by party was also significant.

In 2014, 53 percent of Latino voters participated in the Republican primary and 47 percent voted in the Democratic election, while this year 70 percent of that segment of the electorate took part in the Democratic primary and 30 percent voted in the GOP election.

Let’s break this down a little more, since this jumble of totals and percentages and whatnot muddles what it is we’re actually comparing.


Year  LatinoR  LatinoD    All R    All D  LatinoR%  LatinoD%
============================================================
2014    7,272    6,449  139,703   53,788     5.21%    11.90%
2018   10,855   25,329  156,387  167,982     6.94%    15.08%

“LatinoR” and “LatinoD” represents the number of voters with Latino surnames who voted in the respective primaries for the given year, with those numbers derived from the percentages given. The percentages are the share of Latino voters in that primary.


         Growth
===============
LatinoR   49.3%
NonLatR    9.9%
All R     11.9%

LatinoD  292.8%
NonLatD  200.7%
All D    211.2%

“Growth” is the percentage increase of the group in question for the R or D primary from 2014 to 2018. The number of Latino Republicans increased by 49.3% from 2014 to 2018, the number of all other Republicans increased by 9.9%, and so on.

I’m presenting this all just for the sake of clarity. I don’t care to draw any conclusions because I don’t think we have enough data. Especially on the Democratic side, there was so much growth from 2014 to 2018 that it’s basically a waste of time to look at subgroups, because there’s growth everywhere. (OK, “waste of time” is an overstatement. If Latino participation had grown at a smaller rate than non-Latino participation, that would have been genuinely interesting.) A big part of the reason for this is that the turnout in the 2014 primary was so low. We won’t know for years if this is a new baseline or just a blip. As I’ve said before, I wouldn’t make any guesses about November based on what happened in March. There’s value in knowing the numbers. Beyond that, be very careful about making broad statements.

The Ted Cruz brand

It takes all kinds.

Not Ted Cruz

While some Republicans may fret about Trump’s shaky approval ratings or their party’s brand among disappointed conservatives, Cruz seems to occupy his own space in the political firmament. Launching his 2018 reelection campaign in Houston on Monday, he can fall back on his own tried-and-true persona: an unreconstructed conservative born of the party’s grass-roots base.

It is an identity that also could serve in some degree as a bulwark against the anti-Trump wave that has propelled his Democratic challenger, El Paso Congressman Beto O’Rourke, who launched his long-shot campaign a year ago Saturday.

Whatever the president’s fortunes in a tumultuous and often chaotic White House, Cruz, once Trump’s fiercest GOP critic, appears ready to stand on his own.

“He’s certainly not a Trump Republican,” University of Texas government scholar Sean Theriault said. “He’s a Cruz Republican, and I mean Cruz in all caps. He definitely marches to the beat of his own drummer.”

[…]

Although Cruz has smoothed over his differences with Trump since their bitter 2016 GOP primary rivalry, it always has been clear that Cruz represents a much more orthodox strain of conservatism.

From Cruz’s perspective, Trump’s decisive advantage was his cross-over celebrity appeal, an unforeseen X-factor that may or may not still drive voters to the polls.

“The percentage that Trump has that like him, they’re not political people,” said GOP strategist Rick Tyler, who worked on Cruz’s presidential campaign. “They’re not going to show up for other Republicans, and never have. So, the idea that Trump is going to drive people to the polls in Texas is not there. But I don’t think Cruz relies on those people.”

Indeed, all the early signs point to a familiar Cruz strategy of focusing on his conservative base and it die-hard tea party and evangelical activists.

I agree with Sean Theriault that Cruz is a TED CRUZ Republican, whose thoughts and instincts are always about himself and his own interests. I don’t agree that this means that Cruz is a not-Trump Republican, because the only time Cruz has ever opposed Trump, even to the minimal and mostly performative extent of other Republicans when they clutch their pearls over some horrible thing Trump has said, is when he was still running for President. I think he doesn’t like talking about Trump because that takes away from talking about himself.

For sure, this election is about base turnout, and if I were somehow in the position of advising Ted Cruz I’d tell him to focus on making sure his people are engaged. Given Cruz’s lukewarm poll numbers – really, most people just don’t like the guy – it’s an interesting question how big that base is. The Venn diagram of “Ted Cruz Republicans” and “Donald Trump Republicans” surely has a lot of overlap, but I’d bet the not-overlapping part of that set is significant. He’s the favorite to win because he starts out with more voters, but one way or another we’ll get a better idea this year just how much support he really has. The Trib has more.

Precinct analysis: HCDE Precinct 1

After the last precinct analysis post, I got an email from Danny Norris, one of the two candidates in the runoff for HCDE Position 6, Precinct 1, asking if I intended to look at this race. My answer at the time was no, mostly because it’s not as straightforward to do this kind of analysis on non-countywide races. There’s only a subset of the other districts within the area in question, and some of them only partially intersect. Though there are some examples that work well in this framework, it’s generally not very useful. At least, I don’t think that it is.

But I thought about it, and I thought about it in the context of what I was trying to learn from the other examples, which mostly was about how the runoffs might play out, and I thought I could get something of interest from this exercise. There are three non-countywide races in which there are runoffs – CD07, HCDE6, and JP7. They all overlap to some extent. Let’s see what their cross-section looks like:


       Miller   Bryant  Norris
==============================
CD07      709      358   1,306
JP7     6,585    8,209   6,528

Danny Norris and Prince Bryant are the candidates in the HCDE6 runoff. Norris has a big advantage in the part of HCDE6 – which is to say, Commissioners Court Precinct 1 – that overlaps with CD07. Unfortunately for him, that’s a small part of the district. Bryant has a larger absolute advantage in Justice of the Peace Precinct 7, but it’s smaller as a percentage of the total vote there, and there are a lot of voters who went with Johnathan Miller. About forty percent of the vote in HCDE6 was also cast in JP7, so turnout in one will affect turnout in the other. The money is in CD07, which will drive people to the polls there, but that’s mostly a factor for the countywide races. There’s not enough of CD07 in HCDE6 to have much effect on it.

The other perspective is for the countywide races. I didn’t include HCDE6 as a district when I did the analysis of the countywide races, for no particular reason. Let me correct that oversight here, with a look at how each of those races played out in HCDE6/CC1:


District Clerk

Howard  Burgess  Jordan  Shorter
================================
 9,466   24,089   7,598   14,566

County Clerk

  West  Mitchell  Trautman
==========================
 8,151    24,945    21,809

County Treasurer

Garcia  Copeland   Osborne
==========================
15,743    16,087    21,722

HCDE Position 3 At Large

Wallenstein   Cantu  Patton
===========================
     15,006  19,271  19,558

I don’t think this tells us anything we didn’t already know, but there you have it anyway. What I did notice that I hadn’t spotted before was that HCDE6/CC1 contributed about a third of the overall vote total. Technically, HCDE6/CC1 is one fourth of Harris County, but it’s also by far the most Democratic of the four Commissioners Court precincts. I’m not sure what ratio of the vote I’d expect, but it seems like it might normally be a bit higher than one third. The fact that it isn’t is probably one part the CD02/CD07 primaries, one part the other races, and one part the overall level of engagement this year. I’ll be interested to see what the ratio looks like from the runoff.

Still discussing flood bonds

It’s complicated.

Harris County officials Tuesday said the “clock is ticking” on its call for a bond referendum for $1 billion or more in flood control projects, as requirements to provide matching funds for federal grants being disbursed in Hurricane Harvey’s wake threaten to deplete local coffers.

Commissioners Court on Tuesday stopped short of setting a date for the possible election amid questions about what projects could be included in such a bond issue and how much it would cost per year to complete them. The court directed staffers to hammer out specific proposals that would help determine how much debt the county should ask voters to approve.

Calling Harvey a game-changer, Harris County Judge Ed Emmett and other members of Commissioners Court pledged last September to call for a bond election for upward of $1 billion to pay for wide-ranging flood control projects. The bonds likely would come with an increase in property taxes.

At the heart of Tuesday’s discussion was concern over the increasingly high stakes surrounding the fate and necessity of the bond, as well as the county’s ability to take on a host of large-scale projects aimed at preventing a repeat of the flooding and devastation wrought by Harvey.

See here and here for the background. Federal grants, some of which have already been approved, require local matching funds, which constrains what the county can do right now. The county will need to figure out how to balance what it’s doing now with what it wants to do with the bonds.

Officials also wrangled over several other logistical and political issues surrounding the proposed bond referendum, which would be one of the largest ever put before county voters.

“There are a lot of dilemmas facing us here,” Emmett said. “When do you have the election? How much is it? Do you get specific? Do you leave it general?”

The level of a property tax increase accompanying the bond likely will impact the referendum’s fate.

Harris County Budget Officer Bill Jackson said that if, for example, the bond election was for $1 billion and the debt was issued over 10 years, that would result in a $5 increase in property tax bills for the average $200,000 home in the first year. That number likely would rise to about $20 in the 10th year.

Assistant County Attorney Douglas Ray said that if voters reject a bond referendum, the county cannot put the same issue on the ballot again for two years.

Commissioners Court at its next meeting in April could vote to call an election for June 16, but Precinct 1 Commissioner Rodney Ellis expressed concern over turnout during the summer months.

An election during the summer would require a plan to locate and staff polling places around the county. The governor also would have to sign off on a summer date.

“To my knowledge, no governor has ever denied a local bond election,” Emmett said. “But there haven’t been that many that have been called for a special date.”

Pushing the election to November would mean more turnout but also would raise the possibility that voters cast straight-ticket ballots for political parties and ignore the bond, Emmett said.

Ellis said he also worried about limiting the scope of the bond issue to focus on matches for federal grants, stating that he would like to see more investment in lower-income areas, and a bigger bond package to pay for it.

“After the most horrific and historic storm event we’ve had, I’ve heard members of this body say it’s our opportunity to do something big, and we may not get another bite at that apple,” he said.

I don’t think we’ve had a June election (not counting runoffs from May special elections) anytime recently. As far as the voters ignoring the bond question, Harris County hasn’t had a bond election in an even-numbered year recently. The city of Houston bonds in 2012 had undervote rates in the 20-30% range, but that still meant over 400K people voting on them. Metro’s referendum that year had a 21% dropoff but nearly 800K votes cast, while bonds for HISD (19% undervote, 315K ballots cast) and HCC (23% undervote, 352K ballots) were similar. If all those entities could have bonds in a Presidential year, I think Harris County could make do with a referendum in a non-Presidential year. (Metro is planning on one this year, remember.) Plenty of people will still weigh in on it, and if the county can’t successfully sell flood control projects post-Harvey then something is really wrong. I say put it up in November and start working on the campaign pitch now.

It’s about more than the statewide races

Two articles coming to basically the same conclusion. First, from the Observer:

People watching Texas from afar are naturally not very interested in the balance of power in the Legislature, or county government. They’re interested first in whether Texas could flip in a presidential race and, secondly, whether they can be rid of Ted Cruz. So when more Democratic ballots were cast than Republican ones in the largest counties, many read that as evidence that a Democrat could win a statewide race in November, even though the link between the two is pretty specious and at any rate Texas has open primaries. (I mostly vote in the Republican primary, and a lot of people switch at will between the two depending on what’s going on in their district.)

But on election night, the statewide results, from across all 254 counties, were quite different — because of course they were. In the end, there were still more Republican ballots, 1.54 million, than Democratic ballots, 1.04 million. Some observers, hyped on the foggy narrative that lauded the early voting turnout, decided that the results were a dud and lost interest, because the numbers no longer indicated that a statewide election could be won. One national forecaster, Harry Enten, wrote that the primary results were a disappointment because they were comparable to 2006, when the party didn’t win any statewide elections. But Texas Democrats don’t remember that year as a disappointment — they made extraordinary headway in the state House, part of an effort that almost won a majority in 2008.

The Blue Wave was real, and then it wasn’t, in the course of about a week. Stranger still, the made-up national story arc seemed to influence in-state coverage as well. Even though Democratic turnout was better than in any midterm primary since 2002, and more than than double 2014, commentators have consistently described the night as at least a mild disappointment, where the Democrats “fell short” of a goal that had been imagined for them.

The thing is, the way the state goes on the electoral college map doesn’t mean very much at all for the way Texas is governed. And while it’s possible that the party jumps back to life with the shock of winning one or two statewide elections — that there will be a proof of concept, and then everyone suddenly gets serious — it’s more likely that things change slowly, over an extended period of time, and that small gains and positive signs feed bigger gambits. What’s most important in the long run is the overall composition and strength of the Texas Democratic Party at the local and state level.

In that light, the fact that Democratic turnout doubled in urban counties while Republican turnout stayed essentially flat is significant. There are quite a few winnable legislative districts around those cities. The whole ballgame for the party is getting people to vote and to make a habit of voting. Trump is helping them do that — the trick now is to get it to stick, which it most certainly did not after the elections of 2006 and 2008.

And from the Trib:

Texas didn’t see a blue wave in its March primaries. Measured by the number of voters they attracted to their primaries earlier this month, Republicans outnumber Democrats in Texas by a 3-to-2 margin.

Dallas County did see a wave, though, and that could be important in November. The same is true, to some extent, in Harris and Bexar counties. Democrats, judged by turnout in the major party primaries, have a numerical advantage in three of the state’s biggest counties.

Another way to put this: In the three biggest counties in Texas, Democratic primary voters outnumbered Republicans in 2018 — after trailing them in 2014.

[…]

It’s not the blue wave Texas Democrats were hoping for. Texas Republican primary turnout was 1.54 million, while the Democrats attracted 1.02 million voters. But you’ll have to forgive Republicans in Dallas, Bexar and Harris counties if they start hollering for life preservers. Democrats improved their turnout numbers, in comparison with Republicans, in 18 of the state’s top 25 counties (measured by the number of registered voters) — an urban trend that’s been previously noted here and elsewhere.

What’s notable now is the electoral danger posed to incumbent Republicans. They are numerous in the three big counties, providing the Democrats with ample opportunities. They’re nervous because their party’s president is facing his first mid-term election, often a perilous time for that party’s candidates. Meanwhile, the Democrats have candidates in place to pounce as opportunities arise.

That article goes on to list the targets from Dallas County, a list with which we are familiar. The full list goes well beyond these three counties – again, we know what that’s about – but the point is simply that Democrats have a lot of ways to win this year. Obviously, becoming credibly competitive at a statewide level is the overarching goal, but as we get there a lot can happen to make the government we have better. Winning even two Senate seats would be a big step forward, not to mention a key point of leverage, thanks to the “three-fifths rule” (formerly the two-thirds rule) in the Senate, which would allow Dems to block bills they can’t abide.

There are many more lower-level targets to aim for – breaking through in Harris County, including Commissioners Court in Precinct 2, lots of State House seats, and so on – and who knows, Ken Paxton may get convicted, or Sid Miller may finally say something that alienates people who aren’t dead-enders. We’ve been over this before, you know the drill. Winning a statewide race would be huge, but it’s not the sole criteria for success in 2018. Let’s not lose sight of that.

Precinct analysis: Countywide candidates

We have four – count ’em, four – runoffs for Harris County office nominations for May. Every contested countywide non-judicial primary – that is, everything other than County Judge – is going to overtime. I’m going to look at the data from these four races with an eye towards the runoffs. As a reminder, my analysis of the Senate primary is here, and my analysis of the Governor and Lt. Governor races is here. Let’s start with the District Clerk race.


Dist   Howard  Burgess Jordan Shorter
=====================================
CD02    3,161   15,405  2,276   4,938
CD07    3,254   16,917  2,307   5,271
CD08      234      819    160     435
CD09    3,918    7,493  3,185   5,959
CD10    1,000    3,442    769   1,578
CD18    5,631   13,574  4,807   8,922
CD22      438    1,458    355     708
CD29    2,850    6,260  2,562   3,739
CD36      993    4,150    726   1,508
				
HD126     712    2,089    577   1,010
HD127     772    2,505    635   1,220
HD128     486    1,559    344     659
HD129     712    3,509    534   1,207
HD130     610    2,156    421     904
HD131   1,669    2,943  1,389   2,477
HD132     758    2,529    689   1,393
HD133     741    4,486    490   1,213
HD134   1,262   10,294    681   1,813
HD135     713    2,586    700   1,376
HD137     443    1,442    350     677
HD138     623    2,580    433   1,016
HD139   1,535    3,372  1,373   2,232
HD140     479      890    424     602
HD141   1,047    1,714  1,048   1,531
HD142   1,299    2,090  1,216   2,091
HD143     803    1,508    810   1,020
HD144     373      943    340     445
HD145     655    2,149    525     929
HD146   1,735    3,857  1,242   2,687
HD147   1,817    5,482  1,241   3,154
HD148     885    4,795    611   1,249
HD149     622    1,625    532     910
HD150     728    2,415    542   1,243

Marilyn Burgess was above the magic 50% line for most of the evening as Primary Day returns came in, but fell just short in the end, leading the pack with 49.22%. She was strong everywhere, getting at least a plurality in every district except HD142, which she missed by one vote. Stranger things have happened, but it’s hard to imagine her losing in the runoff given the data.

Next up is County Clerk:


Dist    West  Mitchell Trautman
===============================
CD02   3,368     8,412   13,817
CD07   3,824     8,739   15,009
CD08     255       729      651
CD09   3,418    10,215    6,620
CD10   1,222     2,798    2,708
CD18   5,071    15,336   12,068
CD22    418      1,283    1,222
CD29   2,777     6,286    6,160
CD36   1,051     2,687    3,599
			
HD126    783     1,881    1,683
HD12     784     2,152    2,205
HD128    488     1,296    1,257
HD129    756     2,110    3,047
HD130    674     1,713    1,678
HD131  1,340     4,511    2,506
HD132  1,037     2,304    1,972
HD133    878     1,939    4,080
HD134  1,336     2,830    9,754
HD135    956     2,342    2,028
HD137    490     1,105    1,285
HD138    720     1,693    2,214
HD139  1,405     4,216    2,756
HD140    476     1,003      884
HD141    847     3,141    1,312
HD142    954     3,951    1,741
HD143    737     1,953    1,438
HD144    406       716      934
HD145    677     1,247    2,253
HD146  1,513     4,351    3,507
HD147  1,785     4,299    5,328
HD148    922     1,935    4,655
HD149    647     1,613    1,410
HD150    793     2,184    1,927

I’ll be honest, I thought Diane Trautman would do better than she did. She’s been around for awhile, she’s run and won countywide before, and she was a very active campaigner. I wasn’t the only one who was surprised to see this race be as close as it was, with Trautman at 44.27% and Gayle Mitchell, who lost a primary for County Clerk to Ann Harris Bennett in 2014, at 40.42%. When I say that Trautman was an active campaigner, I don’t just mean on Facebook and via email. I mean I saw her at multiple events, including all of the CEC meetings from 2017. Nat West was present at CEC meetings, as he is the SDEC Chair for SD13, but as far as I know Gayle Mitchell never attended and of those or any other event that I did. Be that as it may, she finished just 5,500 votes behind Trautman, and she won or ran strongly in numerous districts. She also did better on Primary Day than she did in early voting; the same was true for Rozzy Shorter and the other non-Burgess District Clerk candidates, which probably just suggests when different types of voters were voting.

Trautman has the advantage of the runoff in CD07 going into May, as that was a big driver of overall turnout and it was her strongest turf, though she wasn’t as strong there as Burgess was. Mitchell will likely benefit from the runoffs in JP7 and HCDE Position 6, Precinct 1 – there is significant overlap between the two – though neither of those will draw people out the way CD07 will. I guess that makes Trautman a slight favorite going into May, but we all thought she was a strong favorite going into March, so who knows. If I had one piece of advice for Trautman, it would be to see if she can get some elected officials to do some outreach on her behalf. Those of us who think she’s the strongest candidate to face Stan Stanart, especially if we’re not in CD07, need to make sure we bring some friends to the polls for her.

I’m going to present the last two races together. They are Treasurer and HCDE Trustee Position 3 At Large.


Treasurer

Dist  Garcia Copeland  Osborne
==============================
CD02    8,841   4,988   11,335
CD07    9,412   5,635   11,931
CD08      685     408      533
CD09    6,404   6,742    6,729
CD10    2,826   1,763    2,060
CD18    9,634   9,856   12,141
CD22    1,226     702      989
CD29    8,533   3,170    3,816
CD36    2,835   1,493    2,910
			
HD126   1,762   1,154    1,391
HD127   2,001   1,280    1,752
HD128   1,268     733    1,005
HD129   2,185   1,166    2,512
HD130   1,679   1,024    1,324
HD131   2,478   2,999    2,711
HD132   2,289   1,508    1,472
HD133   2,209   1,222    3,260
HD134   3,581   1,897    8,060
HD135   2,251   1,485    1,537
HD137   1,193     691      996
HD138   1,849   1,047    1,689
HD139   2,390   2,746    3,051
HD140   1,333     521      573
HD141   1,569   1,964    1,589
HD142   2,038   2,353    2,061
HD143   2,146     978    1,039
HD144   1,301     332      479
HD145   2,399     576    1,295
HD146   2,645   2,898    3,568
HD147   3,264   2,888    4,983
HD148   3,066   1,034    3,373
HD149   1,469   1,029    1,150
HD150   2,031   1,232    1,574

HCDE

Dist Wallenstein   Cantu  Patton
================================
CD02       8,942   8,497   7,619
CD07      11,269   8,813   6,864
CD08         511     610     497
CD09       5,001   7,639   7,290
CD10       2,086   2,570   1,985
CD18       8,126  12,111  11,627
CD22         909   1,258     755
CD29       2,894   9,410   3,240
CD36       2,667   2,856   1,725
			
HD126      1,291   1,760   1,245
HD127      1,487   1,958   1,572
HD128        909   1,370     747
HD129      2,336   2,101   1,408
HD130      1,340   1,515   1,159
HD131      1,956   3,182   3,094
HD132      1,457   2,166   1,629
HD133      3,179   2,017   1,499
HD134      6,878   3,163   3,495
HD135      1,424   2,240   1,593
HD137        872   1,164     834
HD138      1,617   1,752   1,175
HD139      1,961   3,391   2,853
HD140        442   1,530     458
HD141      1,160   2,042   1,971
HD142      1,225   2,811   2,447
HD143        779   2,422     979
HD144        473   1,350     278
HD145        943   2,465     841
HD146      2,590   3,244   3,333
HD147      3,178   3,583   4,486
HD148      2,388   3,150   1,952
HD149      1,018   1,477   1,120
HD150      1,502   1,911   1,434

Treasurer is just a tossup. Dylan Osborne led Cosme Garcia by two thousand votes, and for the most part they were pretty close to even across the districts, with Garcia having a clear advantage in CD29. I don’t see enough of an advantage for either candidate to take a guess at who might have the edge in May. Neither outcome would surprise me.

Richard Cantu has a much more distinct advantage in HCDE, leading Josh Wallenstein by over 11,000 votes. Wallenstein came close to not making it to the runoff – he actually ran third in both phases of in-person voting, but had a big enough lead over Elvonte Patton in mail ballots to hang onto second place. Runoffs can be weird, but Cantu seems like the clear favorite for May.

That wraps it up for the Democratic primary precinct analyses. I have one more of these to present, from the other side. Hope you’ve found these to be useful.

Chron overview of CD07 runoff

I have three things to say about this.

Lizzie Fletcher

Democrats looking for a ray of hope in Houston’s Republican-leaning Seventh Congressional District have their sights locked on an apparent upset victory in a conservative Pennsylvania district that President Donald Trump won by 20 points in 2016.

But the lessons learned from Conor Lamb’s surprise 600-vote win – barring legal challenges – could mean very different things to the two Houston Democrats squaring off in the May 22 primary runoff to face nine-term Republican incumbent John Culberson.

In a race that Democrats see as one of their best pick-up opportunities in the nation, the two rivals, attorney Lizzie Pannill Fletcher and writer-activist Laura Moser, both have started fundraising off Lamb’s victory.

[…]

Rice University political scientists Mark Jones notes that although Trump lost the district to Hillary Clinton in 2016 by nearly 1.5 percent, it still remains decidedly GOP ground that routinely favors Republican candidates by wide margins.

To Jones, who once worked for former Missouri U.S. Rep. Dick Gephardt, what that means is that Culberson will likely downplay Trump in the election – if that’s possible. And for Democrats, given their recent upsets in Alabama and western Pennsylvania, it suggests a tack to the middle.

“Actually, there does seem to be a formula,” Jones said. “The formula is, give Republicans somebody they don’t feel uncomfortable voting for.”

Laura Moser

Moser, in a Chronicle interview before the primary election, said she doesn’t see it that way. “We have tried something over and over in Texas politics, which is to run to the middle and to the right, and it’s not working,” she said. “So why not stand firm for the values that we share? I’m progressive, but I don’t think that the things I stand for are out of keeping with what the majority of this district believes.”

Other Texas Democrats see merit in trying to harness the party’s new-found energy since Trump’s election. Some argue that much of that energy comes from the left with groups such as Our Revolution, a spinoff from Bernie Sanders’ presidential campaign that has endorsed Moser.

“Midterm elections are base elections,” said Ed Espinoza of Progress Texas, a group that represents the liberal wing of the state’s Democratic Party – but which remains neutral in the Moser-Fletcher runoff. “Your task is getting more of your people out than they can get of their people.”

The March 6 primary, however, offered up some sobering math for Houston Democrats. Some 38,032 voters cast ballots in the Seventh District’s GOP primary, a sleepy affair in which Culberson faced just one largely-unknown challenger. In contrast, only 33,176 people came out to vote in the Democratic primary, a seven-way contest with at least four well-financed contenders.

Though turnout on both sides can be expected to increase in the November general election, Jones argues that a base-election strategy for Democrats can only work in a truly swing district – which the Seventh is not. “Even if the base is super-enthused, you’re still going to lose by five or 10 points,” he said.

But to Espinoza, the likelihood of flipping anti-Trump Republican voters in the Seventh District seems remote. “Republican voters have either embraced the crazy, or they’ve jumped ship and they’re going to stay home,” he said. “Any voters who have left the Republican Party, they are not looking for Trump-lite. They’re looking for Trump-opposite.”

1. In the matter of the “turnout or persuasion” debate, the correct answer for this district, and likely some (though not all) others, is “both”. Mark Jones is correct that CD07 isn’t really a swing district, at least not based on 2016 results, in which the average Democratic judicial candidate received 43.5% of the vote. Trying to win here on increased turnout alone is a heavy lift, one that depends to some extent on a factor you can’t control, which is the other side’s turnout level. On the other hand, the fewer voters you need to persuade to cross over, the better. If you can boost turnout enough to make this, say, a six-point district instead of a 12-point district, you have a much better shot at getting a sufficient number of crossovers, if they are there to be had. What the proper mix is, and how to maximize them simultaneously, is the challenge.

2. I’ve already expressed my skepticism about the primary turnout/November turnout connection. For what it’s worth, of the roughly 39K total votes cast in the Republican primary in CD07, over 8,700 people voted for Scott Milder instead of Dan Patrick, and about 6,000 people voted for a Senate candidate other than Ted Cruz. Make of that what you will.

3. I hope all of the other Democratic runoffs that cover part or all of Harris County get as much press combined as CD07 is likely to get by itself.

Precinct analysis: Beto in Harris County

I now have a canvass of the primaries in Harris County, so you know what that means – time for some precinct analyses. I’ve got a few of these to do, so let’s dive right in. First up, a look at the Democratic Senate race.


Dist      Beto   Sema    Kimb
=============================
CD02    20,865  5,038   3,388
CD07    24,094  5,473   3,202
CD08     1,122    429     303
CD09     9,188  5,123   7,181
CD10     4,528  1,787   1,153
CD18    17,597  7,087  10,491
CD22     1,901    811     569
CD29     7,915  5,920   3,094
CD36     5,289  1,807   1,157
			
HD126    2,639  1,186     932
HD127    3,082  1,354   1,158
HD128    1,895    837     612
HD129    4,647  1,319     811
HD130    2,863  1,006     656
HD131    3,358  2,103   3,343
HD132    3,170  1,661     970
HD133    6,644  1,103     621
HD134   15,443  1,401     742
HD135    3,187  1,612   1,052
HD137    2,016    793     460
HD138    3,341  1,176     673
HD139    3,971  1,953   3,039
HD140    1,032    921     595
HD141    1,582  1,400   2,441
HD142    2,497  1,830   2,577
HD143    1,756  1,734     991
HD144    1,101    892     281
HD145    3,120  1,385     406
HD146    5,086  1,986   3,071
HD147    7,747  2,113   2,787
HD148    7,075  1,363     515
HD149    2,031  1,088     860
HD150    3,216  1,259     945

Rep. Beto O’Rourke

Beto O’Rourke did slightly worse in Harris County (59.08%) than he did statewide (61.79%). The Narrative – I feel like it needs to be considered a proper noun at this point – has focused on his weakness in several heavily Latino counties, where Sema Hernandez drew more votes than he did. That’s not the story here, however, as O’Rourke had at least a plurality in all of the Latino-majority districts. He could have done better, sure, but he did have majorities in HDs 145 and 148, and came close in HD144. Stace talked about why Beto didn’t do so well in South Texas, but those issues were not as prevalent for him here.

Where he was weak was in the African-American areas. Beto had pluralities at the Congressional level, but came in second to the even-less-heralded Edward Kimbrough in HDs 141 and 142, and won HD131 by a whisker. I honestly don’t know if Kimbrough did any outreach on his own, but I do have a theory as to what may have been an obstacle for Beto. My guess – and Greg Wythe can correct me if I’m wrong – is that the voters in these districts are on the whole older than voters in other parts of the county, and therefore less reachable by the social-media-driven campaigning that Beto leaned on. I’m open to other suggestions, but if I’m right then I hope this gives his campaign some useful information about where and how to improve going forward, which I hope they use.

Beto was solid everywhere else, and downright dominant in your inner-Loop and higher-income places – 87.8% in HD134 is certainly nothing to sneeze at – so I think we can say that where his campaign penetrated, it resonated. I don’t know if anyone in the pundit class noticed, but there were 18,268 Democratic primary votes cast in HD134 versus only 15,068 on the Republican side, and that was with the Greg Abbott-driven conflagration over Sarah Davis. I have to think Beto helped drive the turnout there on the Dem side, with the Governor’s race also pushing things. If he can refine his approach in the places where he needs improvement, only good things can result. I’ll look at the Governor and Lt. Governor races next. Let me know what you think.

This is our most “run everywhere” election ever

We already knew this, and have quantified it in a number of ways, but it’s still worth taking a moment to marvel at the surge of Democratic candidates this year.

Lisa Seger

Before she could talk about her campaign for the Texas House of Representatives, Lisa Seger needed to check on her goats. Seger, who lives with her husband and 30 goats on a farm 40 minutes outside of Houston, had a doe in the maternity stall that was due any minute. “Spring is kidding season,” she explained.

If elected, the 47-year-old Seger, a sustainable agriculture proponent who got into farming after reading Michael Pollan’s The Omnivore’s Dilemma, would likely be the only member of the legislature with her own brand of yogurt. But what makes her so unusual in the state’s third district isn’t her background, it’s her party—Seger is the first Democratic candidate to run for the seat since 2010, when the Republican incumbent Cecil Bell Jr. was first elected. Seger’s state senator also ran unopposed in her last election.

“I couldn’t remember the last time I was even able to vote for a Democrat in one of our elections here,” Seger says.

In West Texas, two millennial friends, Armando Gamboa, a 25-year-old from Odessa, and 24-year-old Spencer Bounds of Midland, decided to run for neighboring state house districts where Democrats have been AWOL for at least a decade. No one has run in Gamboa’s district since 2004; Bounds’ opponent is a 50-year incumbent who last faced a Democrat in 2008.

Seger, Gamboa, and Bounds are part of a trend. Call it the “Virginia Effect”: A little more than a year after the inauguration, Democrats in deep-red districts are running for office at a historic clip, determined to find and turn out progressive voters in places where no one has competed in years. It’s a sign that the enthusiasm that swept progressive activists in the first year of the Trump administration and led the party to big gains in the Old Dominion and elsewhere in 2017 is still burning heading into the midterm elections. These local races, flying mostly under the radar, could also give a party struggling for relevance in large swaths of the country a quiet boost this fall.

I should note to begin that my wife is friends with Seger, and we are regular buyers of her farm’s goat cheese. Let’s be clear that Seger, Gamboa, and Bounds are running in really tough districts – Donald Trump got 75.2% in HD03, 70.3% in HD81 (the one in Odessa), and 75.7% in HD82 (Midland, and yes that’s Tom Craddick’s district). I don’t know what set of circumstances might be needed to win one of these races, but it would not be something I would expect. That said, there are three obvious reasons why what these folks are doing is important:

– Their odds of winning may be minimal, but they are still greater than zero. You can’t beat something with nothing, and having no candidate to run is the definition of “nothing”.

– Having local candidates to vote for – remember, everyone will have a Democratic Congressional candidate on their ballot this year as well – gives people in these “can’t remember the last time I had a Democrat to vote for” places a reason to show up and vote. Beto O’Rourke is doing a great job getting out to places that seldom if ever get visited by a Democratic candidate, but it’s still the case that someone in Odessa or Midland or the nether regions of Montgomery County is more likely to have their door knocked by one of these three. If we want Beto and maybe some other statewide candidates to win, they’re going to have to do better in these places than previous Dems have done as well as better in the big cities.

– Long term, of course, things can and do change – remember, Republicans were once an extreme minority in Texas. They built up their base one election at a time, competing and eventually winning in places where they had once not existed. There’s no reason why Democrats can’t do well in the not-quite-as-big cities like they do in the big cities, but it’s not going to happen by itself.

That latter point about the medium-sized cities is one I’ve mentioned before – I mentioned it and covered a lot of this same ground in that Rural Dems post – and one I think deserves a lot more thought and effort, but I don’t want to sidetrack this post. What I do want to do to finish this up is to note that right now, Democratic legislative candidates are not doing so hot in fundraising. Some of that as I noted before is due to late entrances, some is due to the zealous focus on the Congressional races as well as Beto’s butt-kicking in that department, and some of it is because the rest of us aren’t paying much attention to State House races. Which, not to state the obvious, we need to do a lot more of, since the Lege is where the really bad stuff will happen if the Republicans have the numbers and the wingnut concentration to run amok again.

So let me put forth a modest suggestion to the big-money types that exist in Democratic politics here: Put together a pool of money to distribute to these lower-profile candidates running in unusual places, so they can at least pay for some campaign materials and maybe hire a manager or the like. I’m thinking something like $50K per candidate, which once you subtract out the incumbents and the candidates in higher-profile races who are already on track to raise plenty of their own money, would probably require $3-4 million all together. That’s actually not much at all in the grand scheme of things – I mean, Sen. John Whitmire could pay for that by himself, twice over – but it could make a real difference in the performance of these candidates as a group, which again would be a boon for Beto and probably more than a few Congressional hopefuls. If nothing else, it would be a loud signal that we’re not screwing around this year. Everyone likes to talk about the examples that Virginia and Alabama set for us in recent months. It would be nice if we did more than just talk about it.

More primary stuff

I don’t have a canvass of the primary vote from the County Clerk yet – sometimes they send out a draft canvass on their own, or they send one to someone I know who shares it with me, and sometimes I have to ask. I’ll probably ask later this week if I don’t have one soon. Primary canvasses are less interesting than November canvasses for obvious reasons, but there are a few questions I have that the data may help me with.

State data is still being compiled as well, but if you want an interesting look at the data we have from early voting – which remember is only for the top fifteen counties by voter registration, then Austin political consultant Derek Ryan has you covered. See here and here for the breakdowns. If you saw any references to who was voting during the EV period, including here, it came from his work.

What I have done as we await more data is put together this spreadsheet that compares turnout in the 2014 and 2018 gubernatorial primaries, on a county by county basis, for both parties. I’ve sat on it for a couple of days because I couldn’t think of anything to say about it that was both sufficiently interesting and not obviously BS in terms of analysis. In the end, I figured I’d just share the spreadsheet and let people do what they want with it. There are tabs for the 2014 and 2018 results by county for Dem and GOP primaries, then there are summary tabs (Dem Sum, GOP Sum) that show the change in turnout from 2014 to 2018 – positive means 2018 was higher than 2014. The Overall Sum tab shows the Democratic share of the primary vote in each county per year. What that means is that in 2014, 40.03% of the votes cast in the gubernatorial primaries in Bexar County were in the Democratic primary, while in 2018 that figure was 54.69%. This is a way of showing how the turnout changed from county to county.

Another way of doing that is on the last tab, the Per County tab, where I sorted everything by voter registration population, so those top 15 counties are at the top. The numbers in the unlabeled columns are the sums of the Growth columns to that point. What that means is this: Turnout in the gubernatorial primaries increased by 406,335 for Democrats in the top 15 counties, and by 104,357 for Republicans in those counties. It increased by 33,472 for Dems in the next fifteen counties, and by 26,759 for Republicans. Finally, it increased by 23,868 for Democrats in all other counties, and by 98,131 for Republicans in all other counties. You can see why this contributed to the surprise many people had when the results for the full state came in and they seemed to differ from the top-15-centric early vote results.

Anyway, there’s that data. I may return to this kind of analysis for other things if I can think of an angle. If you have any questions, let me know.

Does primary turnout in a district predict the November result?

Karl Rove would like you to think so.

At the House level, Democrats hope to win three districts won by Hillary Clinton and now held by Republican incumbents, as well as some of the six seats opened up by GOP retirements. Here again, the primary results are not heartening for Democrats.

In two Clinton-GOP congressional districts—the Seventh, in Houston, represented by Rep. John Culberson, and the 32nd, in Dallas, held by Rep. Pete Sessions—more Republicans voted than Democrats: 38,032 Republicans to 33,176 Democrats in the Seventh and 41,359 Republicans to 40,084 Democrats in the 32nd. Mrs. Clinton carried both districts by less than 2 percentage points in 2016.

Moreover, no Democrat won a majority in either district’s primary, forcing runoffs in May. In the Seventh, journalist Laura Moser —endorsed by the Bernie Sanders-connected “Our Revolution”—is pitted against Clinton loyalist and attorney Lizzie Pannill Fletcher. The Democratic Congressional Campaign Committee targeted Ms. Moser with an opposition-research dump arguing she was too liberal to win in the fall. The attack backfired: Ms. Moser was trailing Ms. Fletcher in early voting before the DCCC assault but won more votes among those who turned out on election day.

Democrats outvoted Republicans in a GOP-held seat that Mrs. Clinton carried by 3.4 percentage points—the massive 23rd Congressional District, which sweeps across West Texas. This year, after Democratic candidates spent a combined $1.1 million, 44,320 voted in their primary to 30,951 Republicans. Still, that is 5,000 more Republicans than voted in the 2014 primary, which launched Will Hurd into Congress. A former undercover CIA officer, Rep. Hurd is one of the GOP’s most effective campaigners. His “DQ Townhalls” at Dairy Queens across his largely Hispanic district helped him hold the district by 1.3 points in 2016 even as Mr. Trump lost by more than 3 points.

Democratic aspirations to take some of the six open Republican congressional districts also appear slim: Republicans turned out more voters in all six, with the GOP’s margins ranging from roughly 16,000 to 22,000 votes.

If we’re talking about CD23, I can tell you that the Democratic candidates have received more votes than the Republican candidates in each primary since 2012, which includes one year that Pete Gallego won and two years that Will Hurd won. As such, I’m not sure how predictive that is.

More to the point, I am always suspicious when a data point is presented in a vacuum as being indicative of something. We’ve had primary elections before. How often is it the case that the party who collects the most primry votes in a given race goes on to win that race in November? Putting it another way, if one party draws fewer votes in the primary, does that mean they can’t win in November? Let’s step into the wayback machine and visit some primaries to the past to see.


2004

CD17 - GOP            CD17 - Dem

McIntyre     10,681   Edwards      17,754
Snyder       11,568
Wohlgemuth   15,627

Total        37,876   Total        17,754

November result - Edwards 125,309  Wohlgemuth 116,049

HD134 - GOP           HD134 - Dem

Wong          4,927   Barclay         771
                      Daugherty     4,193

Total         4,927   Total         4,964

November result - Wong 36,021  Daugherty 29,806

HD137 - GOP           HD137 - Dem

Witt          1,291   Amadi           376
Zieben          970   Hochberg      1,012

Total         2,261   Total         1,388

November result - Hochberg 10,565  Witt 8,095

HD149 - GOP           HD149 - Dem

Heflin        2,526   Vo            1,800

November result - Vo 20,695  Heflin 20,662


2006

HD47 - GOP            HD47 - Dem

Welch         2,349   Bolton        1,569
Four others   3,743   Three others  2,071

Total         6,092   Total         3,640

November result - Bolton 26,975  Welch 24,447

HD50 - GOP            HD50 - Dem

Fleece        1,441   Strama        2,466
Wheeler         294
Zimmerman     1,344

Total         3,079   Total         2,466

November result - Strama 25,098  Fleece 13,681

HD107 - GOP           HD107 - Dem

Keffer        3,054   Smith           724
                      Vaught        1,169

Total         3,054   Total         1,893

November result - Vaught 16,254  Keffer 15,145

HD134 - GOP           HD134 - Dem

Wong          3,725   Cohen         2,196

November result - Cohen 25,219  Wong 20,005


2010

HD48 - GOP            HD48 - Dem

Neil          9,136   Howard        6,239

November result - Howard 25,023  Neil 25,011


2012

SD10 - GOP            SD10 - GOP

Cooper        6,709   Davis        17,230
Shelton      28,249

Total        34,958   Total        17,230

November result - Davis 147,103  Shelton 140,656

HD144 - GOP           HD144 - Dem

Pena          1,030   Perez         1,149
Pineda        1,437   Risner          462
                      Ybarra          591

Total         2,467   Total         2,022

November result - Perez 12,446  Pineda 10,885


2014

SD15 - GOP            SD15 - Dem

Hale         13,563   LaCroix       3,239
                      Whitmire      9,766

Total        13,563   Total        13,005

November result - Whitmire 74,192  Hale 48,249


2016

HD107 GOP             HD107 - Dem

Sheets       10,371   Neave         6,317

November result - Neave 27,922  Sheets 27,086

Some points to note here. One, I’m cherry-picking just as Rove had done. There were plenty of examples of one party outvoting the other in a given primary race, then winning that race in November. That’s why I don’t have an example to cite from 2008, for instance. It’s also why I concentrated on the legislative races, since outside of CD23 there haven’t been many competitive Congressional races. Two, as you can see most of the examples are from last decade. That’s largely a function of how brutally efficient the 2011 gerrymander was. Three, these are actual votes cast, not turnout, as that data doesn’t exist on the SOS page and I was not going to trawl through multiple county election sites for this. It could be in some of the closer examples that adding in the undervotes would have flipped which party led the way.

All that out of the way, as you can see there are plenty of examples of parties trailing the primary votes but winning when it mattered. In some cases, the March tallies weren’t close, like with SD10 in 2012. In some other cases, it was the November races that weren’t close, like HD50 in 2006 and SD15 in 2014. The point I would make here is simply that this doesn’t look like a reliable metric to me. If you want to make the case that these Congressional races will be tough for Democrats to win regardless of the atmosphere and the demographic trends and the relative level of enthusiasm in the two parties, I’d agree. The weight of the evidence says that despite the positive indicators for 2018, we’re still underdogs in these districts. Our odds are better than they’ve been, but that doesn’t mean they’re great. I don’t think you need to use questionable statistics to make that case.

One more thing to consider: There was an effort, mostly driven by educators, to show up in the Republican primary and vote against Dan Patrick. It didn’t work in the sense that he won easily, but some 367K people did vote against him. I’m sure some number of those people are reliable Republicans, but some of them were likely new to the primary process. This probably had an effect on overall Republican turnout. A small effect, to be sure, but if it’s a little more than half of the anti-Patrick vote then we’re talking about 200K people. Take them out of the pool and the Republicans are back down at 2014 turnout levels.

I have no idea how much this effect might be. It’s certainly small, and I doubt you could measure it without some polling. But we know it’s there, and so it’s worth keeping in mind.

Evaluating Beto

I think this is about right.

Rep. Beto O’Rourke

Beto O’Rourke, the candidate running the most high-profile statewide race, [scored] only 61 percent in his primary, against two lesser-known candidates.

[…]

A lot of weird things happen in the Democratic Primary, because the party is far from cohesive. A few years ago, a LaRouche acolyte made it into a Senate runoff, and it’s not unheard of for the party’s contender to get crushed in the first round for unclear reasons. The fact that [Sema] Hernandez and [2014 gubernatorial candidate Ray] Madrigal won in many of the same places seems to point to the benefit of running with a Hispanic last name in the Democratic Primary. It’s possible voters really took to Hernandez’s and Kimbrough’s message, of course, but it seems likely more evidence that lots of Democrats enter the primary booth with limited knowledge of who is on the ballot and select names — ask Grady Yarbrough and Jim Hogan. And it’s hard to blame them, because the “frontrunners” that usually are on the ballot aren’t exactly titans.

That said, O’Rourke’s soft spot so far has been name recognition. If you’ve seen 30 news stories a day about O’Rourke for the last six months and seen some of his packed rallies, that might seem strange, but there’s room to question whether all the hype about the “punk rock Democrat” is translating to the masses.

The Trib has a map showing the county-by-county results, and now they have a story covering the same topic. Some polls have shown that O’Rourke’s name recognition, while perfectly decent for a three-term Congressman making his first statewide run, is hardly universal. I think that’s exactly what these results show, and it’s the basic weakness of his otherwise well-lauded “visit everywhere” campaign strategy. The simple fact is that even in a low-turnout statewide election, there are way more voters than there are opportunities to meet and interact with them. If you’re not already well-known in the state, a condition that describes nearly every current Texas Democrat, you’re going to have to fortify your outreach with some old-fashioned communications. O’Rourke has raised an impressive amount of money so far, and is close to even with Ted Cruz in fundraising. It would have been a good investment to drop a few of those bucks on something other than a volunteer-powered text message outreach to voters (which annoyed a few of them of my acquaintance, by the way). This is again a reminder that one should never overestimate one’s name ID.

All that said, this is hardly a disaster. He still won handily, which is mission one. He’s getting under Ted Cruz’s skin, which ought to provide a little free advertising for him as Cruz generates news about him. I doubt he has to worry about people voting on a name in November, when party affiliation will be part of the process. But if O’Rourke wants to be someone who will push people to the polls – and Lord knows, we all want that for him – and not just someone who will be voted for by those who do show up, he’s going to need to look at these result and figure out what he could be doing better. He has time to introduce himself to a (much) wider audience, but he needs to be a bit more strategic about that. You can do this, Beto.

UPDATE: Stace has more.

Post-Primary Day thoughts

Various thoughts and observations that are better aggregated into one post…

– I believe this is the first time that all of the statewide candidates I voted for either won or advanced to the runoff in the primary. Statewide primaries are tricky, and one should never overestimate one’s name ID. Thankfully, there were no zeroes in the downballot races, and the ones that were in the top level races lost.

– As primary season began, I had expressed hope for a high level of primary turnout, to provide further evidence of our level of engagement in this election. We topped one million votes cast, and the total number of votes in the Governor’s primary (1,017,150) bested the total number from 2002 (1,003,388). There are more voters now, of course, and Republicans topped 1.5 million total, but still. It’s nearly double what we had in 2014 and it’s the second best total basically ever, after 2008. I’m happy with that.

– Of course, the fact that Republicans did cast more primary votes than Democrats is being cited as evidence that there’s no “blue wave” coming. I thought the fact that Democrats vastly outvoted Republicans in the 2008 primary was supposed to be evidence that primary turnout doesn’t really tell you anything? I’m confused. Be that as it may, Democrats had a bit less than double the turnout from 2014, while Republicans were up about fifteen percent. You can feel however you want to about that, I feel good about it.

– Looking at election night returns, a bit more than half of the Democratic primary vote was cast early, and the same was true for the Republican primary vote. It was basically the same in Harris County, where about 55% of the vote in each party was cast early. Final Harris County turnout for Dems was 167,396, and for Republicans it was 155,798.

– Which means, if primary turnout is indeed destiny, that Republicans are doomed in Harris County, right? You tell me when this matters and when it doesn’t.

– Democratic runoffs include Governor, eleven Congressional races, SBOE12, SD17, seven State Rep races, and all of the countywides plus one more HCDE and one JP races in Harris County. There are surely other county race runoffs elsewhere, but I’m not going to go looking for them at this time. Republicans have six Congressional runoffs, seven State Rep runoffs, two district Courts of Appeals, and in Harris County one District Court race and one JP race. That suggests to me there will be more media attention being paid to the Democratic runoffs, especially given the lack of a Republican statewide race for May. Of course, that may not all be good attention, but it’s another difference from 2014, and 2012 for that matter.

– I’m still digesting all the numbers, and will have more thoughts and tidbits as we go. I expect to get a canvass report from the County Clerk in the next couple of days and will of course play with that. For the most part, I’m happy with how the primaries went. People were engaged, turnout was good, no obvious clunkers got elected or into runoffs. You always want more, but overall I have no complaints. May I say the same about the runoffs in May. How do you feel about how the primaries went?

2018 primary results: Statewide

Rep. Beto O’Rourke

Statewide Dem totals
Statewide GOP totals

Harris County Dem totals
Harris County GOP totals

(Please note that all results were coming in very slowly. I expect there will still be some precincts not yet reported by the time this publishes. So, I’m going to be less specific than usual, and may have to make a correction or two by Thursday.)

No real surprises here. Lupe Valdez and Andrew White will fight it out in the runoff. They combined for about 70% of the vote. Beto O’Rourke was a bit over 60% on his way to the Senate nomination. To be honest, I thought he’d score higher than that, but whatever. Statewide primaries are hard.

Miguel Suazo was near 70% for Land Commissioner, and Roman McAllen was near 60% for Railroad Commissioner. Mike Collier was leading by about seven points for Lt. Governor. The closest race was for Comptroller, where Joi Chevalier had a tiny lead over Tim Mahoney.

On the Republican side, Greg Abbott (90%), Ted Cruz (85%), Dan Patrick (75%), and Railroad Commissioner Christi Craddick (75%), who I didn’t even realize had an opponent, all cruised. Baby Bush and Sid Miller were in the high 50’s and so also on their way to renomination. That means the only statewide runoff will be for the Democratic gubernatorial race.

One note on turnout: In 2014, there were 554,014 total votes cast in the Democratic primary for Governor. The early vote tally for the Dem gubernatorial primary was 555,002. So yeah, turnout was up. Republicans will probably have 30-40% more total turnout statewide, but I fully expect Dems to top one million at this point.

2018 primary results: Harris County

Statewide Dem totals
Statewide GOP totals

Harris County Dem totals
Harris County GOP totals

(Please note that all results were coming in very slowly. I expect there will still be some precincts not yet reported by the time this publishes. So, I’m going to be less specific than usual, and may have to make a correction or two by Thursday.)

Short and sweet, because it’s late and I’m tired:

– Marilyn Burgess fell just short of 50% for District Clerk. She will face Rozzy Shorter in May.

– Diane Trautman and Gayle Mitchell will run off for County Clerk.

– Dylan Osborne and Cosme Garcia were the top two finishers for County Treasurer.

– Richard Cantu led for HCDE Position 3 At Large, with Josh Wallenstein just ahead of Elvonte Patton. In a very tight race, Danny Norris was ahead of Prince Bryant by a nose for HCDE Position 6, Precinct 1, with John Miller farther back. There were only a few precincts out as I wrote this, but things were close enough that the standings could change.

– Adrian Garcia and Penny Shaw will be the nominees for County Commissioner in Precincts 2 and 4, respectively.

– Lucia Bates toppled Don Coffey for JP in Precinct 3. Sharon Burney and Cheryl Elliott Thornton will compete for JP in Precinct 7.

– There were only a couple of races of interest on the R side. Josh Flynn won the nomination for HCDE Trustee in Place 4, Precinct 3. Current HCDE Trustee and total chucklehead Michael Wolfe will face Jeff Williams for JP in Precinct 5. Paul Simpson held on as party chair.

– Dem turnout was 160,085 with about fifty precincts left to report. Republican turnout was 148,857 with 85 precincts still out.

Early voting in the “next” 15 counties

As you know, there’s been a lot written about primary turnout in the top 15 counties by voter registration in Texas. Much has been said about the large increase in Democratic turnout, accompanied by the much milder increase – and in some counties, decrease – in Republican turnout when compared to 2014 and 2010. This is great, but Texas has 254 counties, and there are a lot of decent-sized metro areas that are not represented in the coverage we’ve seen, Moreover, while the top 15 counties include many blue and purplish counties, the next 15 are much more tilted to the red side. Here, by my reckoning, are those counties:

Bell (Killeen/Temple/Belton)
Lubbock
Jefferson (Beaumont)
McLennan (Waco)
Smith (Tyler)
Webb (Laredo)
Hays (San Marcos)
Brazos (Bryan/College Station)
Ellis (Waxahachie)
Guadalupe (Seguin)
Comal (New Braunfels)
Johnson (Cleburne)
Parker (Weatherford)
Randall (Amarillo)
Midland

Webb is strong Democratic; Hays and Jefferson are quasi-Democratic; the rest are varying shades of red. I wanted to know how voting was going in these counties, so off to Google I went. The best story I found in my searches came from Smith County:

Early voting ticked up among Smith County voters for the March primary, and about half of the increase came from people casting ballots for Democrats.

A total of 12,926 early ballots were cast in Smith County, according to the county’s elections division. By party, there were 10,994 ballots cast for Republicans and 1,932 cast for Democrats.

Overall, the numbers represent a 9.5 percent increase in early voting overall as compared with 2014, the last time there was a primary election for local and statewide offices but no candidate for president.

By party, the 2018 early voting numbers represent a 5.6 percent increase for Republicans, who cast 10,409 early ballots in 2014, and a 38.5 percent increase for Democrats, who cast 1,395 early ballots in 2014.

Early voting lasted approximately two weeks, from Feb. 20 through Friday. Polling places were open in five locations in Tyler, Lindale, Whitehouse and Noonday. The primary election is Tuesday.

Mark Owens, a political science professor at the University of Texas at Tyler, said much of the increase in early voters in 2018 could be attributed to an increasing population in Smith County.

The number of registered voters in Smith County is 131,007 in the 2018 primary, a 5.8 percent increase over the 123,867 registered voters at the time of the 2014 primary, according to Owens.

Owens called the early voting turnout “just on par for a growing area.” However, he credited Democrats for having an impact on the increase in early voters in conservative Smith County.

In raw numbers, 1,122 more Smith County residents voted in 2018 over 2014. Republicans accounted for 585 of those ballots, and Democrats accounted for 537 of them.

“To the Democrats’ credit, the voter mobilization efforts are stronger in the fact that this isn’t a primary with as many leading elections at the top of the ticket, so they would see it, from their perspective, of people wanting to vote for (their candidates),” Owens said.

“A really big part of it is candidates coming out to East Texas to listen and encouraging people to go vote,” Owens said. “I think if you look at the numbers, that means something to people.”

I’d call that encouraging. Dems are still vastly outnumbered, but they showed up and increased their totals over 2014. Indeed, the total number of votes cast in the Democratic primary in 2014 was 2,328, so early voting turnout there came close to matching that by itself.

That’s about as good as it gets in terms of being specific. This Lubbock story is pretty representative:

Heading into the last day of early voting for the 2018 primaries, the Associated Press reported that Texas had already set a non-presidential cycle record for the number of people turning out. Before Friday, more than 583,000 Texans in the 15 largest counties had cast early ballots in person, which was already more than the then-record of nearly 510,000 who did so during early voting for 2014′s midterm election.

In Lubbock County there were 15,430 total ballots cast during the 11 days of early voting. That means about 9 percent of registered voters took advantage of the early voting period.

About 400 more votes in Lubbock county were actually cast this year than during early voting in 2014, the last midterm election. This year’s total is about 8,700 votes less than in 2010. During the last primary in 2016, more than 25,000 votes were cast in early voting.

[…]

The Lubbock County Elections Office hasn’t yet released the separate vote totals for the Republican and Democratic primaries.

Some of these places make you downright wistful for Stan Stanart. Here’s Hays County:

As of Feb. 26, 4,658 early votes have been accounted for at seven different locations spread across the county. This does not account for the nearly 2,000 votes submitted to the county by mail.

In total, around 6,600 have been counted for, shattering the numbers from previous election cycles in 2014 and 2016.

According to Hays County numbers, roughly 4,500 people voted early in the November Presidential 2016 election, while only 1,768 early votes were counted in November 2014 race.

“We’ve had a very high turnout considering the political season we are in,” said Jennifer Anderson, elections administrator for Hays County. “Democratic turnout has been good and that is to be expected considering the national swing we had with the Presidential election.”

[…]

So far, roughly 53 percent of the early voting population voted in the Republican Primary, while 46 percent of the early votes took part in the Democratic Primary.

At least that’s something to go on. In 2014, 8,521 votes were cast in the Republican primary for Governor (this isn’t the same as turnout, since people do undervote in individual races, but I can’t get to the Hays County elections page as I write this, so it will have to do), compared to 3,131 votes in the Dem primary for Governor. If the split this year is something like 53-46, then the Dem share is up by a lot. That’s very good to see.

From Comal County:

Registered voters in Comal and Guadalupe counties have their last chance to cast early ballots today for candidates competing in Tuesday’s Republican and Democratic primary elections.

Voters in both counties flocked to the polls during the 12-day early voting period, which began Feb. 20. Through last Tuesday, 5,654 Comal County residents — about 6 percent of the county’s 95,353 registered voters — had cast early ballots, running ahead of the number and percentage of registered voters who turned out in the 2014 midterm elections.

The rest is behind a paywall. Comal is deep red – think Montgomery County-deep red – so this will be worth watching. In 2014, there were 14,458 Republican primary gubernatorial votes, and 1,647 Democratic votes, so you can see what I mean. Neighboring Guadalupe County has a bit more detail:

Guadalupe County Elections Administrator Lisa Adam said area residents have slowly begun increasing their presence at the polls.

“Our numbers this week have already been higher than in the 2014 gubernatorial primary,” she said. “The first week’s numbers for this year were a little lower than they were in 2014. This week we are actually ahead than the second week of early voting in 2014; not by leaps and bounds, but we are ahead.”

[…]

“In the 2014 primary, we had 81,217 registered voters,” she said. “Right now, as of Feb. 1 we have 95,717. We’ve come a long way. We’re adding 300 to 400 registered voters a month. The growth our county is experiencing is incredible.”

In that election cycle, the county saw 14.2 percent of the voting population turn out for the Republican Primary and 2.1 percent for the Democratic Primary, Adam said.

1,688 Dem gubernatorial primary votes, 11,196 Republican. Again, there’s lots of room to grow here.

Brazos County:

Early voting before the March 6 primaries wrapped up Friday with 5,933 Brazos County voters casting ballots.

Most of those, 4,144, came from Republicans, and 1,789 Democrats voted early. The total for the two-week early voting period was helped by a push of 1,467 voters Friday. There are about 105,000 registered voters in the county.

That’s burying the lede here. In the 2014 gubernatorial primary there were 1,927 total Dem votes, and 10,665 total Rep votes. In other words, Dems are way up. Republicans, not so much.

For McLennan County, I turn to my friend Carmen Saenz:

Final numbers for 2018 early voting in McLennan County primary:
Dems: 3054 – 28% of total
GOP: 7778 – 72% of total

Relative to 2014 early voting in the McLennan County primary:
Dems 1085 – 18% of total
GOP 4940 – 82% of total

Although there is a 181.5% increase in the number of Dems voting and only a 57.5% increase in GOP, with an overall increase of 80% these numbers say a lot about the McLennan County Democratic Party.

In a lot of the counties, we’ve seen Dem numbers up a lot with Republican numbers not up much if at all. Both are up here, which makes McLennan a bit of an outlier.

The city of Amarillo is in both Randall and Potter counties. I didn’t find a good story for Randall County, but I did find this for Potter:

In Potter County, there have been 4,940 votes in-person and mail-in since Feb 27. That number is expected to increase by seven tonight, at the end of early voting.

In the 2016 Presidential Primaries, there were 5,284 early votes cast in Potter County.
Breaking down the numbers even further, 4,128 Republicans cast their vote in Potter County, during early voting.

That has surpassed the numbers from the 2016 election, which topped out at 4,031 votes. The Democrats have cast 821 votes, slightly less than 2016 early voting at 988.

That’s 2016. If you look at 2014, there were 810 total votes cast in the Democratic gubernatorial primary. So yeah, it’s up.

Last but not least, Midland County:

The elections offices in Midland and Ector County has seen a dip in voter turnout this early voting season.

As of Friday afternoon in Ector County, election officials counted 5,300 voters.

There are over 74,000 registered voters in Ector County.

[…]

As of Friday morning in Midland County, the total number ballots counted was a little over 6,200.

There are over 81,000 registered voters in Midland County.

2014 gubernatorial primaries:

Ector County – 1,320 Dems, 7,778 Republicans
Midland County – 960 Dems, 12,640 Republicans

If there’s a downward trend in these places, it’s probably not because of the Dems.

I’ll return to this later in the week. For now, this is where we stand.

The Republican poll of the Democratic CD07 primary

I would not pay too much attention to this.

Lizzie Fletcher

A poll in a pivotal Democratic congressional primary in Houston shows that activist Laura Moser could be in a position to make the run-off despite recent attacks by the Democratic Congressional Campaign Committee DCCC).

The poll, by the Republican-aligned Congressional Leadership Fund, shows Moser with 17 percent support in the seven-way primary race in the Seventh Congressional District. She trails only Houston lawyer Lizzie Pannill Fletcher, with 28 percent.

In third place on the last day of early voting in the primary is cancer researcher Jason Westin, at 14 percent.

Laura Moser

Alex Triantaphyllis, the top fundraiser in the Democratic field with more than $1.1 million in receipts, trails with 13 percent, virtually tied with Westin, who has raised half that amount.

Fletcher’s double-digit lead over three rivals in a virtual statistical tie still leaves open the likelihood of a runoff election on May 22. If nobody reaches 50 percent of the vote, the top two vote-gettters go on to a runoff.

The others in the field, James Cargas, Joshua Butler and Ivan Sanchez each polled at about 1 percent in the survey, which was conducted Thursday. Cargas, an assistant city attorney, was the Democrats’ 2016 standard-bearer. About 25 percent said they were undecided, with less than a week before next Tuesday’s primary.

[…]

The automated Interactive Voice Response and phone survey of 726 likely Democratic primary voters found that 61 percent disapproved of the DCCC attacks on Moser, while 13 percent approved and 26 percent said they weren’t sure.

The reason I am dubious is not because this is a Republican firm doing the polling but because nobody knows what a “likely voter” is in this race. Primary polling suffers from the same problem that polling in municipal races suffers, which is that the composition of the electorate can vary widely based on turnout. We already know that a significant number of people voting in the Dem primaries have little to no primary voting history. By definition, these people are not “likely Democratic primary voters”, but here they are anyway. It’s possible that this firm has guessed well as to who is likely to show up and thus arrived at an accurate result, and it’s possible they’ve produced a 2014 UT/Trib debacle. They have no track record in Dem polls to examine, so we’re left to judge this poll for ourselves. My judgment is to note it as a data point and move on. I’d advise you to do the same.

Rural Dems

They still exist, and they’re making some noise.

Trish Robinson was dropping off supplies for Hurricane Harvey relief efforts in Liberty County, about 40 minutes from Houston, when a handful of people scowled at her left-leaning political T-shirt.

David DeLuca, head of the Fayette County Democrats, said he recently introduced himself to a Republican volunteer poll worker, but the woman declined to shake hands.

And during a Tyler County town hall hosted by Senate hopeful Beto O’Rourke on a recent Friday in February, multiple people thanked the Democrat for coming to the Republican stronghold.

[…]

As a result, many of the remaining rural Democrats say they’re leery of making their political views public.

Now, as national and state party leaders talk a big game about a blue wave this November, some Democrats in rural East and Central Texas say they’re working to overcome a drag on local momentum ahead of the primary: stigma.

[…]

Something changed after Trump’s election. Through social media, Democrats in rural areas began finding each other and organizing in small but meaningful ways. County chapters have dusted off their welcome signs and other Democratic-leaning groups have emerged, both publicly and privately.

Robinson said she began the Liberty County Indivisible chapter after reading the handbook published by its national founders, which gives guidance on anti-Trump grassroots political organization.

“That sort of fed into what I believe and how I felt, and what I wanted to do,” she said. “Liberty County [Democrats] need to know there’s options for them, too. We shouldn’t always have to leave the county to feel like we belong or have a purpose or can speak up.”

She began with a Facebook post, and then held meet-ups at local restaurants. Sometimes, only one person showed up. But “if one person comes every time, I feel like that’s progress,” she said. The group has since grown to about 40.

This isn’t exactly profound, but the main thing these folks can do is believe that their votes matter, so that they actually do show up and vote in November. Dems have made big gains over the past several cycles in the big urban and suburban counties, but there are a lot more small and rural counties, and the steady degradation of the vote in those places has largely canceled those gains out. Compare the 2012 and 2012 Presidential results on a county-by-county basis sometime – it’s a thousand votes here and a thousand votes there, but in a state with 254 counties that can really add up. This is basically what the Beto strategy is all about: Narrow the margins in the unfriendly places enough so you can bridge the remaining gap in the big counties.

So having local candidates to vote for, even in uncompetitive districts, helps. Having the statewide candidates remember that these places exist helps, too. I don’t pretend to know when Dems might be able to truly contest these places, nor do I know what issues might hasten that day, but I believe the Republican Party is doing its best to marginalize itself, and the effect that has will not remain limited to the cities and the suburbs. In the meantime, let’s run candidates in races we can win – city councils and school boards and the like – and put some resources into figuring out how to make common cause and gain ground with voters in the small and medium-sized cities in these counties. I believe the opportunities are there, and we’ve taken the important first step of showing up. Let’s keep it going.

2018 primary early voting, Day 11: That’s a wrap

Records have been set, and more are likely to be set before all is said and done.

Texans have already set a record for early voting in a non-presidential primary election year, and there’s still one more day to go on Friday.

More than 602,000 voters had cast ballots in the state’s largest counties in either the Democratic or Republican primaries through Wednesday. That does not count Thursday’s totals that were not available late Thursday, or Friday’s, when polls will again be open from 7 a.m. to 7 p.m. Four years ago, fewer than 600,000 people voted in the entire early voting season.

Harris County has also seen a new record. More than 116,000 people have voted early or by mail already with two days remaining to add to that total. Four years ago, just 105,508 people in Harris voted during the entire early voting period.

Democrats represent a major reason for the records and have been out-voting Republicans since the start of early voting on Feb. 20. There have been 25,000 more Democratic ballots than Republicans have cast. That is a big change from the last two gubernatorial election cycles when Republicans dramatically outvoted Democrats in the primaries by well over 100,000 in each year.

Those numbers are partly driven by people who are new to primaries, said Austin-based political analyst Derek Ryan. In looking at voter data from about 50 counties, Ryan said he’s seeing that almost 20 percent of the Democrats voting in that primary have never voted in a Democratic Primary in Texas before. For Republicans that has been closer to 8.5 percent.

More from the DMN.

According to the Texas secretary of state’s website — which tracks only the 15 counties with the most registered voters — 161,607 people voted in the Democratic primary in 2014 during the first 10 days of early voting. This year, 310,275 people voted in the Democratic primary in the same span — a 92 percent increase. Polls closed Friday at 7 p.m., with Election Day on Tuesday.

On the GOP side, 273,293 people had voted in the Republican primary as of Thursday. That’s still an 18 percent increase from 2014, when 231,530 voted in the Republican primary during the first 10 days of early voting.

Democrats may hold a 36,982 vote lead, but that doesn’t mean all of those voters are Democrats. Since Texas has semi-open primaries, voters can choose which party’s primary to vote in. (There is a caveat to choosing: In a runoff, voters must stick with the same party.)

Political experts attribute much of Texas’ increased voter turnout as a reaction to the election of President Donald Trump in 2016, as well as the state’s eight open congressional seats.

Harold Clarke, a political science professor at the University of Texas at Dallas, said one reason Democratic primary voting numbers are up is because people think Democrats have a reasonable chance of winning.

“One of the things we know is that competition stirs turnout, and it looks now that here in Texas, perhaps especially for the first time in a long time, winning the Democratic primary is really a prize worth having because you have a real shot at the general election,” Clarke said. “That perception is fairly widespread both among potential candidates as well as Democratic voters.”

I’m interested in seeing how Dems did in the fifteen or twenty counties after the top fifteen. These are much more heavily Republican, but as we’ve seen Dem turnout has still been up in places like that. I’ll check the news over the weekend to see if I can get any previews of that.

Anyway. Let’s wrap this up:

EV 2010
EV 2014
Day 11 EV 2018 totals


Year  Party     Mail In Person    Total
=======================================
2010    Dem    6,250    33,771   40,021
2010    Rep   12,399    50,250   62,649

2014    Dem    7,359    22,749   30,108
2014    Rep   17,628    57,772   75,400

2018    Dem   17,744    70,172   87,916
2018    Rep   20,075    61,462   81,537

There were 17,728 in person Democratic votes cast on Friday, which as you can see is not much less than the entire early in person Democratic vote from 2014. That right there is this EV period in a nutshell. I’ll have more thoughts on the EV period either tomorrow or Monday. For now, know that this is going to be the best year for Democratic primary turnout outside of 2008, and any time you can make a legit comparison to 2008, it’s a good thing.

On Beto’s campaign strategy

The Trib writes about the places Beto O’Rourke has gone.

Rep. Beto O’Rourke

There was a frantic energy in the air Saturday morning amid a downpour of chilly rain as cars circled the sprawling John M. Tidwell Middle School parking lot in search of spaces. But the candidate was about to speak, so a number of cars gave up.

Crying uncle here, however, didn’t mean leaving. Instead, they haphazardly parked in fire lanes and were out of their vehicles practically before the engines stopped running.

It’s the type of scene that plays out every four years in states such as Iowa and New Hampshire: Curious voters cram into gyms and cafeterias and warehouses to see the latest buzzed-about candidate who’s come to town.

Except this wasn’t an early voting state, and this was not a candidate for president.

Instead, hundreds of Texans turned out on Saturday to hear U.S. Rep. Beto O’Rourke, a Democrat from far away El Paso, make his case for the U.S. Senate. This has been happening around the state but O’Rourke assured these Texans that they were special in his path to victory against Republican U.S. Sen. Ted Cruz.

“If there is one county that is the linchpin of the 254 [counties] that will determine the outcome of this election, it is Tarrant County,” O’Rourke said.

[…]

Consciously or not, O’Rourke is following then-U.S. Sen. Barack Obama’s Iowa strategy in 2007 of campaigning hard in conservative areas with an aim to lose less badly there and make up the difference in urban areas.

There is scant reliable, public polling on this race, so the best indicators of where it stands are anecdotal evidence like crowd sizes. If a candidate drew these kinds of crowds in smaller states, alarm bells would ring.

But Texas is massive in population and in geography, and it is an open question whether this strategy will work.

Beto’s visit-everywhere strategy has been written about multiple times, but the Trib’s summary of it is accurate. It could work, but we have no way of knowing right now. It’s different, which makes it interesting, and it’s not like the playbook everyone else had been running has much to recommend it. We just don’t have any basis for comparison except the other not-similar campaigns. And when all is said and done, we won’t know how much of Beto’s final result will be because of this strategy, and how much will be due to other factors like the national atmosphere and the loathing all decent people feel for Ted Cruz. I think the strategy he is using is a sensible one, and I think this is the perfect year to try it. It may or may not work, and even if it does work as well as it could have he may not win. We’re still a Republican state, even if the prognosticators warm to his chances. Whatever does happen, I hope we learn the right lessons from it for the future.

Early voting, Day 10: Same day service

Hey, guess what? The EV numbers for Thursday came in early enough for me to post a truly up-to-date update. So here we go:

EV 2010
EV 2014
Day 10 EV 2018 totals


Year  Party     Mail In Person    Total
=======================================
2010    Dem    5,728    23,914   29,642
2010    Rep   11,478    36,321   47,799

2014    Dem    6,802    17,092   23,894
2014    Rep   16,696    42,975   59,671

2018    Dem   16,532    52,344   68,876
2018    Rep   18,848    47,298   66,146

Thursday was slightly bigger than Wednesday, which is actually a little lighter than I might have expected. It was still another Democratic-majority day. The gap in in-person voting is starting to become wide. Dems have not yet returned half of their mail ballots, but Republicans have only returned 61.6% of theirs, which as we know is a bit light for them. If normal patterns hold, today should have about double the in-person votes as Thursday, but who knows what might happen with this unusual election. If you haven’t voted yet, what are you waiting for?

Early voting, Day 9: Around the state again

Let’s just jump right in again…

EV 2010
EV 2014
Day 8 EV 2018 totals


Year  Party     Mail In Person    Total
=======================================
2010    Dem    5,035    16,107   21,142
2010    Rep   10,327    24,759   35,086

2014    Dem    5,456    12,405   17,861
2014    Rep   14,687    29,287   43,974

2018    Dem   12,914    36,785   49,699
2018    Rep   15,512    33,140   48,652

The second Wednesday is usually where the turnout curve starts to swing upward, with the last Friday always being the busiest day. We could see a lot more voters come out before the end of the week. And notice, for all the justifiable excitement about Democratic participation, Republican turnout so far is up, too. It’s just not up as much. There are many explanations for this, some of which have elements that ring true to me. I’m going to let the numbers speak for themselves.

Now let’s look statewide. Here’s how the top 14 counties by population have been voting on the Democratic side.


County      EV 2010  EV 2014  EV 2018     2010    2014
======================================================
Harris       21,142   17,861   49,699  101,263  53,788
Dallas       11,745   20,258   35,580   56,896  68,053
Tarrant       4,696   13,663   21,507   23,704  42,209
Bexar        13,741   17,431   28,630   43,338  44,835
Travis        7,904   13,597   32,139   38,915  48,495
Collin        1,475    3,102   10,986    6,229   9,584
Denton        1,006    2,311    8,644    4,678   7,212
El Paso      11,199   11,246   19,255   36,353  32,571
Fort Bend     2,472    2,559    8,935   14,890   8,549
Hidalgo      18,647   22,479   23,645           47,350
Montgomery      893      910    3,212    4,539   2,561
Williamson    1,688    2,397    8,846    7,148   7,243
Galveston     2,402    1,802    3,686            4,680
Cameron       5,395    6,777    6,019           18,472

Totals      104,405  136,393  260,783          395,602

All numbers are for the Dem primaries; some counties did not have 2010 data available. The first three columns are for the first eight days of early voting – i.e., through the second Tuesday – while the last two are the final totals for those years. With the three busiest days of early voting plus Primary Day to go, several counties have already exceeded their final tallies from 2014, with Harris County right behind. We’re already two thirds of the way to the final overall total for these counties. I’m not going to make a projection, I’m just pointing out where we are. I think cracking one million voters and having the strongest showing after 2008 is well within reach. Keep voting, y’all.

UPDATE: Here are your Day 9 totals. Dems are now almost at 60K.

Early voting, Day 8: So how worried should Republicans be?

Worried enough to fundraise off of the Dems’ EV numbers, for what that’s worth.

Through Sunday in the 15 Texas counties with the most registered voters, 135,070 people had voted in the Republican primary and 151,236 in the Democratic. Compared to the first six days of early voting in 2014, Democratic turnout increased 69 percent, while Republicans saw a 20 percent increase.

The Democrats even surpassed their early voting totals from the 2016 primary — a presidential election year.

Sen. Ted Cruz told a group of Republican voters this month that the left would “crawl over broken glass in November to vote … We could get obliterated at the polls,” and other Republicans appear to be taking the Democratic surge seriously. Gov. Greg Abbott’s campaign sent supporters an email Monday asking for donations to help him get out the vote, warning that the early voting numbers “should shock every conservative to their core.”

“I’ll be blunt: Democrat voter turnout is surging statewide during Early Voting,” reads the email, using bold and italicized red print. The email states that the last time Democratic primary voters came out so strongly was in the 1990s, during a gubernatorial election cycle, and that Democrats are flipping seats in special elections across the country in Republican strongholds.

“We’ve seen a surge of liberal enthusiasm in deep red states like Georgia, Alabama, and Oklahoma,” the email says. “We had always hoped the liberal wave would never hit Texas, but these Early Voting returns aren’t encouraging so far.”

Brandon Rottinghaus, a political science professor at the University of Houston, said it’s interesting that Democrats are turning out at a rate more frequently seen in presidential election years. After looking at the relationship between primary and general election voters, he concluded that more votes in Democratic primaries correlate with more Democratic votes in general elections. But he said Republicans usually turn out in higher numbers to vote in the general election no matter how they voted in the primary.

“Usually Republicans tend to run up the numbers in the general and are beating their opponents by big margins, so the relationship is not positive, but it is for Democrats,” Rottinghaus said. “Because the [Democrats’] enthusiasm is so high, you’re likely to see more support for Democrats in November and that’s likely to cut into the margins that they’ll lose to Republicans.

Fearmongering isn’t the same as being fearful, and it’s not like we haven’t seen this kind of language before. Republicans used Battleground Texas to scare the yokels in 2014, after all. It’s just that this year the voting numbers back up their apocalyptic pronouncements. It doesn’t mean anything yet, but it should at least quiet the narrative that Dems don’t turn out for primaries.

And here are those Monday numbers that didn’t come in till late Tuesday morning:

EV 2010
EV 2014
Day 7 EV 2018 totals


Year  Party     Mail In Person    Total
=======================================
2010    Dem    4,571    14,018   18,589
2010    Rep    9,376    21,421   30,797

2014    Dem    4,471    10,210   14,681
2014    Rep   13,573    23,930   37,503

2018    Dem   11,207    30,664   41,871
2018    Rep   13,812    27,497   41,309

Dems outvoted Republicans in person and by returned mail ballot on Monday, and thus took the lead in overall turnout. They have already exceeded their early vote total from 2014, and ought to surpass the overall 2014 turnout on Wednesday. I feel like Dems will easily top the 101,263 ballots cast in 2010, thus making this the biggest primary outside of the insane 2008 experience. Whatever it means, the excitement is real.

UPDATE: Here are Tuesday’s numbers, which did come in on Tuesday evening. Let’s just assume I’m going to be a day behind on these, OK? Dems outperformed Republicans by another 500 votes, and are just shy of 50K votes overall.

2018 primary early voting, Day 7: Projecting final turnout

KUHF starts with the speculation.

Harris County Democrats are voting in record numbers ahead of next week’s primary. Total returns for the first six days of early voting put Democrats nearly even with Republicans.

As of Sunday night, Democrats’ combined in-person votes and mail ballots received totaled 34,555, an increase of nearly 200 percent over the 2014 congressional midterm election.

“They have an unprecedented number, the biggest they’ve ever had,” Jay Aiyer of Texas Southern University said on Houston Public Media’s Party Politics Podcast, “and it’s still counting. It’s important because about 60 to 65 percent of the total vote will come from these early votes.”

By comparison, Republican votes over the first six days totaled 35,036, up just 11 percent from the last midterm.

With all due respect, I think Jay is overestimating the share of the vote that will be cast early, and thus underestimating the amount that will be cast on Election Day. Here’s a look at past performance in Democratic primaries:


Year    Early    E-Day   Early%
===============================
2006   11,500   23,947    32.4%
2008  179,348  231,560    43.6%
2010   40,963   60,300    40.5%
2012   38,911   37,575    50.9%
2014   31,688   22,100    58.9%
2016   87,605  139,675    38.5%

There’s not much of a pattern here, but in no year has as much as 60% of the Democratic primary vote been cast early. My guess, when I put these numbers together, was that we’d be around fifty percent early (this includes mail ballots in all cases). I won’t be surprised if that’s an underestimate, but I don’t think it will be by that much. One reason for this is that it hasn’t been just the old reliables voting so far.

An analysis of the first four days of early voting in the March 6 primaries indicates that the fabled rebellion against the Republican social conservative leadership may not be materializing. On the Democratic side, it shows a surge of new voters—a fifth of the primary turnout is from people with little to no history of voting in a Democratic primary.

The new analysis of the early voting turnout comes from Derek Ryan, a Republican consultant. Ryan builds off of a Texas Secretary of State database of who voted in which elections. The database does not tell anyone how you voted, but it does reveal the names of who votes in party primaries and general elections. He then receives a daily report from the election administrators in eighteen of the top Texas counties to compare current voters to past voters with an eye toward spotting trends.

What Ryan found on the Republican side is a usual primary for a non-presidential election year. So far, more than 86 percent of the Republican primary votes have been cast by people who voted in past Republican primaries. Only about seven percent of the vote has come from people who do not vote in party primaries. Crossover voting from Democrats is almost nonexistent, with only a single percent of the GOP vote coming from 2016 Democratic primary voters.

Business and education groups have been urging members to vote in the Republican primary because of opposition to issues like bathroom bills or private school vouchers. These initial numbers indicate a weak rebellion. At the same time, social conservatives regularly make up less than 42 percent of the Republican primary vote. If enough of the Republican regulars combine with the new voters, some upsets are possible, although right now they look unlikely.

Over on the Democratic side, almost eighteen percent of the voters are people with no history of voting in a primary of either party; another three percent are people with no history of voting at all in primary or general elections; and 1.5 percent were Republican primary voters in 2016. Without polling the individual voters, Ryan told me there is no way to tell whether the surge is from motivated general election Democrats or from “purple” voters prompted to vote Democrat because of anger over the national Republican party politics.

I agree we can’t tell yet if the level of primary voting means anything for November. At this time, pending a change in the makeup of the Democratic primary electorate, I think we can say there’s still a decent reserve of regular voters who haven’t shown up yet but who almost certainly will. That to me suggests that the turnout on March 6 will be higher than one might think. I reserve the right to change my mind about this later in the week.

So what happened yesterday? Well, as of 11 PM, the daily vote report had not arrived in my mailbox. That happens when the hours change to 7 AM to 7 PM, so I’m afraid we’ll just have to wait. I may post an update later, but most likely I’ll just save this for tomorrow. Sorry.

UPDATE: Here at last are Monday’s numbers – apparently there were some technical difficulties. I’ll have full details tomorrow, but Dems outvoted Republicans in person and in returned mail ballots, and have overtaken the Rs for the lead in total votes. Boo yah!

2018 primary early voting, Day 6: The fifteenth county

Sunday is the shortest and least busy day of early voting, and it is the transition to Week 2, when all the days are 12 hours long and numbers start to go way up. Here’s what this Sunday looked like.

EV 2010
EV 2014
Day 6 EV 2018 totals


Year  Party     Mail In Person    Total
=======================================
2010    Dem    4,129    11,533   15,662
2010    Rep    8,498    17,900   26,398

2014    Dem    3,592     8,399   11,991
2014    Rep   12,288    19,649   31,937

2018    Dem    9,620    24,935   34,555
2018    Rep   12,642    22,394   35,036

Another day where more Democrats voted; Dems have almost caught up to Republicans in overall turnout. Dems have already exceeded their early vote total from 2014 (which was 31,688) and should pass 2012 (38,911) and 2010 (40,963) no later than Tuesday morning. Tomorrow I’ll look at the historical pattern in early voting turnout in Democratic primaries so we can begin to get a feel for what final turnout might be.

I’ve looked at the daily early vote returns from the Secretary of State, which tracks the numbers from the 15 biggest counties – the totals through Saturday are here. The thing about this is that the composition of the top 15 changes over time – for 2010 and 2014, Nueces County was on the list, but this year Brazoria County made the cut. As such, we can’t do the same-day comparisons for Brazoria, but we can get a bit of context by looking at the final EV totals, which you can see here: 2010 Dem, 2010 Rep, 2014 Dem, and 2014 Rep. In short:

2010 Dem = 5,828 total votes, 3.15% turnout – 2,189 votes were cast early
2010 Rep = 23,514 total votes, 14.01% turnout – 12,019 votes were cast early

2014 Dem = 2,933 total votes, 1.64% turnout – 1,542 votes were cast early
2014 Rep = 18,842 total votes, 10.56% turnout – 11,275 votes were cast early

2018 Dem = 2,133 votes so far, 1.06% turnout
2018 Rep = 7,123 votes so far, 3.54% turnout

Remember that the 2018 numbers are through Saturday, which is to say Day 5 of 11. This is more than the entire early turnout from 2014 and almost as much as 2010. I’d expect the early vote in Brazoria County to surpass final turnout from the 2014 primary on Tuesday, and will probably bypass final turnout from 2010 on Friday. So there you have it.

2018 primary early voting, Day Five: A Democratic day

The early vote hours for Saturday are 7 to 7, so the report comes out later. So, I’m just going to cut to the chase:

EV 2010
EV 2014
Day 5 EV 2018 totals


Year  Party     Mail In Person    Total
=======================================
2010    Dem    4,129    10,516   14,645
2010    Rep    8,498    16,533   25,031

2014    Dem    3,592     7,765   11,357
2014    Rep   12,288    18,176   30,464

2018    Dem    9,620    22,252   31,872
2018    Rep   12,642    20,730   33,372

The Saturdays of early voting are always strong for Democrats, and this one was no exception. Between mail and in-person, Dems led on Saturday by over 2,100 votes, thus closing the overall turnout gap to within 1,500. Note that the five-day turnout of 31,872 is more than the final EV turnout in 2014, which was 30,108. I’d guess the Dems will exceed their entire primary turnout from 2014 by Thursday. Week 2 is always busier than Week 1, so we’ll see how high the ceiling is.

2018 primary early voting, Day Four: On to the suburbs

The Chron’s Mike Snyder ventures outside Harris County for early voting numbers.

As the Chronicle’s Jeremy Wallace reported, nearly 50,000 people voted in the Democratic primary on Tuesday, the first day of early voting, in the state’s 15 most populous counties. That’s more than twice the total from the first early voting day in 2014, the last midterm election.

 Despite the Democrats’ improved turnout, however, Republican numbers were greater. Through Wednesday, 14,493 people had voted in the Republican primaries compared to 12,627 in the Democratic primary in Harris County, Wallace reported.

The Republican primaries also drew more voters in the two other Houston-area counties that rank among the 15 most populous in the state.

In Montgomery County, 8,364 early votes had been cast through Thursday in the Republican primary, compared to 1,437 in the Democratic primary.

The numbers were closer in Fort Bend County: Republicans 4,878; Democrats 3,403. (Totals in both counties include in-person and mail-in ballots.)

This is true, but it’s also not the whole story. Let’s go back to the SOS early voting page for a little context.


Party     County      2010    2014    2018
==========================================
Rep   Montgomery     3,851   6,944   8,364
Dem   Montgomery       383     393   1,437

Rep   Montgomery     1.58%   2.57%   2.61%
Dem   Montgomery     0.16%   0.15%   0.45%

Rep    Fort Bend     3,486   3,755   4,878
Dem    Fort Bend       871     921   3,403

Rep    Fort Bend     1.16%   1.07%   1.18%
Dem    Fort Bend     0.29%   0.26%   0.82%

Democratic turnout is up by a lot more in Fort Bend than Republican turnout is. Democratic turnout isn’t up as much in Montgomery County as Republican turnout is, but relatively speaking it’s up by a lot more. Another way of saying this is that as a share of registered voters, which is what those percentages represent, Republicans are up a pinch from 2014, while Dems are at triple their levels from 2014. I submit that’s a notable development.

So what about Harris County? Well, here you go:

EV 2010
EV 2014
Day 4 EV 2018 totals


Year  Party     Mail In Person    Total
=======================================
2010    Dem    4,129     8,229   12,358
2010    Rep    8,498    12,571   21,069

2014    Dem    3,592     5,949    9,541
2014    Rep   12,288    13,901   26,189

2018    Dem    8,844    16,110   24,954
2018    Rep   12,530    16,053   28,583

Dems had 5,214 people show up yesterday, which as you can see is only a bit less than their entire four-day total from 2014. Republicans had a few more than that and have nearly closed the in-person gap, but it’s the increase from previous years that is the real story. And while Dems still haven’t done as good a job returning their mail ballots, the 8,844 they have returned is more than the entire number of returned mail ballots in every primary before 2016, and should be on track to beat that year’s total. The news continues to be good. Let’s keep it going.

2018 primary early voting Day Three: A look around the state

Let’s just jump right into the numbers:

EV 2010
EV 2014
Day 3 EV 2018 totals


Year  Party     Mail In Person    Total
=======================================
2010    Dem    3,851     6,132    9,983
2010    Rep    7,929     8,803   16,732

2014    Dem    3,048     4,228    7,276
2014    Rep   11,464     9,678   21,142

2018    Dem    7,641    10,896   18,537
2018    Rep   11,558    10,781   22,339

I had the mail and in person totals for 2018 backwards in yesterday’s post, so sorry about that. Republicans had the better day yesterday, both in person and absentee – at this point, they have returned more than a third of their mail ballots, while Democrats have not yet returned one fourth of theirs. They’re only slightly ahead of their pace for 2014, however, while Dems are way ahead of theirs – their three-day total is about 60% of their entire early vote tally from 2014, and more than a third of their overall final turnout. And as we’ve been observing, this has been the pattern in the big counties around the state. Here are the two-day totals for the big counties:


Party     County      2010    2014    2018
==========================================
Rep        Harris   13,044  16,633  14,493
Dem        Harris    7,676   5,316  12,627

Rep        Dallas    4,617  10,251   6,226
Dem        Dallas    3,491   5,533   9,768

Rep       Tarrant    5,720  11,096   8,293
Dem       Tarrant    1,676   4,739   8,087

Rep         Bexar    5,107   8,484   6,329
Dem         Bexar    4,835   5,741   7,100

Rep        Travis    3,177   2,149   3,021
Dem        Travis    2,394   4,244   8,382

Rep        Collin    3,797   4,654   5,098
Dem        Collin      359     728   2,529

Rep        Denton    2,414   4,588   3,773
Dem        Denton      244     615   1,826

Rep       El Paso    1,531   1,214   1,334
Dem       El Paso    3,935   3,971   6,885

Rep     Fort Bend    2,779   2,945   3,342
Dem     Fort Bend      607     649   2,391

Rep       Hidalgo      614     879     891
Dem       Hidalgo    6,964   7,676   8,782

Rep    Montgomery    2,685   5,282   5,824
Dem    Montgomery      271     283   1,061

Rep    Williamson    2,397   2,573   3,799
Dem    Williamson      532     840   2,456

Rep     Galveston    1,004   3,040   3,385
Dem     Galveston    1,041     636   1,285

Rep       Cameron      410     528     468
Dem       Cameron    2,022   2,479   2,513

Some of these numbers are just insane. Democrats basically even with Republicans in Tarrant County? I didn’t see that coming. Even in the big red places, Dems have taken big steps forward, while Republicans have either had smaller increases or even fallen back. It’s just two days and anything can happen, but so far so good.

2018 primary early voting Day Two: When is it a trend?

I think we can say that people noticed the Day One early voting numbers.

Democrats have more than doubled their early voting in the state’s biggest counties compared to four years ago, leading some party leaders to point again to a growing wave election they think will send a dramatic message to Republicans.

But while Democrats are voting better than they did four years ago, Republicans still are near where they were four years ago, even though the lack the same star power in the primary that they had four years ago at the top of the ballot.

In the state’s largest 15 counties, nearly 50,000 people voted in the Democratic primary elections on the first day of early voting.

In 2014 — the last mid-term election cycle — only about 25,000 Democrats voted in the primary. Never have the Democrats had so many early voters in a primary in a gubernatorial election cycle going back to the mid-1990s when early voting started.

[…]

Meanwhile, Republican numbers in Texas early voting are essentially flat, with 47,000 Republicans voting on the first day of early voting — slightly lower than the 49,000 that voted four years ago.

But Republicans say those numbers don’t mean Democrats are suddenly about to overtake Republicans in both energy and at the ballot box.

We’ll talk about the rest of the state in a minute. For now, let’s update the Harris County numbers.

EV 2010
EV 2014
Day 2 EV 2018 totals


Year  Party     Mail In Person    Total
=======================================
2010    Dem    3,466     4,210    7,676
2010    Rep    7,264     5,780   13,044

2014    Dem    2,484     2,832    5,316
2014    Rep   10,514     6,119   16,633

2018    Dem    6,976     5,651   12,627
2018    Rep    6,676     7,817   14,493

Republicans had slightly more Day 2 in person voters, and more mail ballots returned, but Dems still lead in the in-person total. Of interest also is that another 2,239 mail ballots were sent to Dem voters, for 32,311 total mail ballots, while Republicans received only another 349, for 29,935 total.

Now, as Campos says, it’s one thing to request a mail ballot and another thing to return it. So let’s look at some past history of mail ballots in primaries:


Year  Party   Request  Return  Return%
======================================
2008    Dem    11,989   7,056    58.9%
2008    Rep    18,415  13,432    72.9%

2010    Dem    11,847   6,250    52.8%
2010    Rep    17,629  12,399    70.3%

2012    Dem    13,087   7,735    59.1%
2012    Rep    23,584  17,734    75.2%

2014    Dem    12,722   7,359    57.8%
2014    Rep    24,548  17,628    71.8%

2016    Dem    19,026  13,034    68.5%
2016    Rep    29,769  20,780    69.8%

One of these years is not like the others. Dems have emphasized mail ballots in the past couple of cycles, and you can see the difference in 2016. If that behavior repeats this year, Dems will reap the benefit of their larger pool of voters with mail ballots. We’ll keep an eye on that.

Finally, the DMN has a good look at voting around the state on Day One.

Of the 51,249 Texans who cast ballots Tuesday on the first day of early voting, more than half voted in the Democratic primary.

The total number of voters from 15 of the state’s largest counties is high for a midterm year. In 2016, a presidential election year, 55,931 Texans voted on the first day of early voting for the primary. But in the last midterm election in 2014, only 38,441 Texans voted on the first day.

Even more surprising is the turnout among Democrats. Since the last midterm election, the party saw a 51 percent increase in first-day early voting turnout, while Republicans saw a 16 percent increase.

You can find daily EV totals for the 15 biggest counties here, and for past elections including primaries here. I’ll return to these numbers later on, as they lag a day behind.

2018 primary early voting Day One: Let’s get this started

And we’re off, with a few concerns about aftereffects of Harvey.

Hurricane Harvey may loom large in many Houston-area residents’ minds, but the storm is expected to have a limited impact on participation in the Texas primary, which kicks off Tuesday with the start of early voting.

Nearly two weeks of early balloting precedes the Lone Star State’s March 6 primary, the first in the nation.

“On one hand, we’re going to see a decline in turnout among some individuals who are displaced. On the other hand, I think there are some people who will counterbalance that decline because they’ve become more politically active and aware as a result of Harvey,” Rice University political scientist Mark Jones said. “The net effect is likely to be pretty neutral.”

Harris County Clerk Stan Stanart, whose office administers local elections, agreed.

“If it does, it’s going to be so small you won’t be able to measure it,” Stanart said. “Your primary voters are your core voters, your most loyal of voters, so those people tend to vote no matter what’s happening. So, I don’t anticipate much disruption in their voting patterns.”

I think turnout is going to be up due to a higher level of engagement this year, but we’ll soon see. It will be interesting to track the vote by State Rep district, to see how things may have changed from previous years.

Speaking of which, of course I have those totals, from 2010 and 2014. Google Drive is an amazing thing. And now we can add the 2018 totals and have a look at them all.


Year  Party   Mail In Person    Total
=====================================
2010    Dem  2,886     2,190    5,076
2010    Rep  5,946     2,774    8,720

2014    Dem  2,080     1,276    3,356
2014    Rep  9,048     2,807   11,855

2018    Dem  4,174     3,833    8,007
2018    Rep  6,138     3,509    9,646

So more Dems voted in person, but more Republicans voted overall because of more mail ballots being returned. Note, however, that more mail ballots were sent to Democratic voters (30,072) than to Republican voters (29,566), which is a big change from 2014. It’s one day and there’s a long way to go, but this is a strong start. I’ll keep an eye on this as we go. When do you plan to vote?

Already projecting ahead to November turnout

Some in the political chattering class think the end results in Harris County this yearwon’t be all that different than what we’ve seen before.

Harris County may be awash in Democratic hopefuls for the upcoming primary elections, but don’t expect that enthusiasm to translate into another blue wave this fall.

Yes, local demographics are slowly pushing the region further left, and President Donald Trump – who dragged down the Republican ticket here two years ago – gives progressives a ready campaign talking point. Democrats also point to their nearly full primary slate as evidence of newfound strength.

It is unlikely those factors will be enough, however, to counteract Republicans’ longtime advantage in Harris County midterms, political scientists and consultants said. Not only do local conservatives turn out more consistently in non-presidential years, but Republicans also have the benefit of popular state- and countywide incumbents on the ballot, advantages made only more powerful by straight-ticket voting in November.

“There is a very slow, but steady demographic shift that will favor Democrats. I don’t know if it’s enough this year for a gubernatorial cycle,” Democratic strategist Grant Martin said.

Rice University political scientist Mark Jones agreed.

“Greg Abbott represents a red seawall here in Texas that I think will in many ways blunt the anti-Trump wave, and in doing so help hundreds of down-ballot Republican candidates across the state achieve victory,” he said.

[…]

Fewer than 54,000 Harris County voters cast ballots in the Democratic primary four years ago, compared to nearly 140,000 in the Republican primary. Come November, Republicans dominated down the ballot.

Though primary turnout certainly is not predictive of November performance, it can be, as University of Houston political scientist Brandon Rottinghaus put it, “a good pulse check.”

Rottinghaus said he anticipates Democrats will perform better locally than they did in 2014, but still come up short in most local races, in large part because of their turnout problem.

“You’re definitely going to find a narrowed margin for most of these offices,” Rottinghaus said. Still, he added, “it would be hard to unseat the natural advantage Republicans have in the midterm.”

I feel like there are a lot of numbers thrown around in the story but without much context to them. Take the primary turnout totals, for instance. It’s true that Republicans drew a lot more people to the polls in March than the Democrats, but their margin in November was considerably less than it was in 2010, when the primary tallies were 101K for Dems and 159K for the GOP. Will anyone revise their predictions for November if the March turnout figures don’t fit with this “pulse check” hypothesis? Put a pin in this for now and we’ll check back later if it’s relevant.

But let’s come back to the November numbers for 2010 and 2014 for a minute. Let’s look at them as a percentage of Presidential turnout from the previous election


   2008 Pres  2010 Lt Gov    Share
==================================
R    571,883      431,690    75.5%
D    590,982      329,129    55.7%

   2012 Pres  2014 Lt Gov    Share
==================================
R    586,073      340,808    58.2%
D    587,044      317,241    54.0%

I’m using the Lt. Governor race here because of the significant number of crossover votes Bill White – who you may recall won Harris County – received in the Governor’s race. He did so much better than all the other Dems on the ticket that using his results would skew things. Now 2010 was clearly off the charts. If the share of the Presidential year vote is a measure of intensity, the Republicans had that in spades. I’m pretty sure no one is expecting that to happen again, however, so let’s look at the more conventional year of 2014. The intensity gap was about four points in the Republicans’ favor, but that was enough for them to achieve separation and sweep the downballot races.

What does that have to do with this year? The key difference is that there were a lot more voters in 2016 (1,338,898) than there were in either 2008 (1,188,731) or 2012 (1,204,167), and that the Democratic advantage was also a lot bigger. I’m going to switch my metric here to the 2016 judicial average, since there were even more crossovers for Hillary Clinton than there were for Bill White. In 2016, the average Republican judicial candidate got 606,114 votes, and the average Democratic judicial candidate got 661,284. That’s a pretty big difference, and it has implications for the intensity measure. To wit:

If Democratic intensity in 2018 is at 55.7%, which is what it was in 2010, then Dems should expect a base vote of about 368,335.

If Democratic intensity in 2018 is at 54.0%, which is what it was in 2014, then Dems should expect a base vote of about 357,093.

Well guess what? If Republican intensity is at 58.2%, which is what it was in 2014, then the Rs should expect a base vote of about 352,758. Which, you might notice, is less than what the Democrats would expect. In order to match the Democratic base, Rs would need 60.8% to equal the former total, and 58.9% for the latter.

In other words, if intensity levels are exactly what they were in 2014, Democrats should expect to win most countywide races. Republicans will need to be more intense than they were in 2014 just to keep up. And if Democratic intensity is up, say at 60%? That’s a base of 396,770, and it would require a Republican intensity level of 65.5% to equal it.

Where did this apparent Democratic advantage come from? Very simply, from more registered voters. In 2016, there were 2,182,980 people registered in Harris County, compared to 1,942,566 in 2012 and 1,892,731 in 2008. I’ve noted this before, but it’s important to remember that while turnout was up in an absolute sense in 2016 over 2012 and 2008, it was actually down as a percentage of registered voters. It was just that there were so many more RVs, and that more than made up for it. And by the way, voter registration is higher today than it was in 2016.

Now none of this comes with any guarantees. Democratic intensity could be down from 2010 and 2014. Republicans could be more fired up than we think they will be, in particular more than they were in 2014. My point is that at least one of those conditions will need to hold true for Republicans to win Harris County this year. If you think that will happen, then you need to explain which of those numbers are the reason for it.

Oh, and that “red seawall” that Greg Abbott represents? Republicans may have swept the races in 2014, but they didn’t actually dominate. 2010, where they were winning the county by 12-16 points in most races, that was domination. Abbott got 51.41% in 2014 and won by a bit less than four and a half points. Which was enough, obviously, but isn’t exactly a big cushion. Like I said, the Republicans will have to improve on 2014 to stay ahead. Can they do that? Sure, it could happen, and I’d be an idiot to say otherwise. Will it happen? You tell me, and account for these numbers when you do.