Good Friday Austin:
Among the very first stories I wrote as political writer for the Statesman was a January 2013 piece under the headline, Texas House returns with largest contingent of new members in 40 years.
It began as follows:
State Rep.-elect Jonathan Stickland is 29. He left high school early and got a GED. He had never held or run for office before. His local elected officialdom was virtually unanimous in its preference for his Republican primary opponent. If he has a charisma it’s in his super-ordinariness. And he doesn’t even have the “r” in his last name that everyone assumes is supposed to be there.
And there, in brief, are the keys to Stickland’s stunning success. Every strike against him, he marvels, turned out to be an advantage in what turned out to be a crushing, 20-point primary victory. Each provided a way for people to remember and identify with him. He just had to own it, live it, be it.
Now, Stickland is one of the reasons why the new Texas House, when it convenes Tuesday for its biennial session, will be swollen with freshman – 43 in all. Together with 24 sophomores, the new and the near-new will make up close to half the 150 members of the House.
“It’s an incredible number,” said James Henson, director of the Texas Politics Project at the University of Texas.
Much of that has to do with places like Stickland’s home turf – Tarrant County – a tea party stronghold where voters gave one well-tenured Republican after another the boot.
Said Stickland, “Tarrant County lost a lot of seniority in this wave – Northeast Tarrant Tea Party. They won every single race they endorsed in.”
So, as Stickland proclaimed to huzzahs at a well-attended NE Tarrant Tea Party gathering in December, “Tarrant County just sent the most conservative group down to Austin that this state has ever seen.”
And Stickland said in an interview, “I plan on having the most conservative voting record in the entire House of Representatives.”
Well, like Babe Ruth in the 1932 World Series game at Wrigley Field, Stickland called his shot and hit it out of the park.
Here is the Texas House liberal-conservative index for Stickland’s first session (in this case just the Republican members are shown here), produced by Rice University political scientist Mark Jones and published each legislative session by the Texas Tribune.
And here is the 2017 index, in which Stickland, Rinaldi and freshman Rep. Briscoe Cain, R-Deer Park, finished in a virtual tie.
“Cain was slightly more conservative than Rinaldi who was slightly more conservative than Stickland,” said Jones. But he said the distinctions were so slight as to be meaningless, and it is hard for the human eye to contend with so many decimal places.
Notably, the 12 most conservative members of the House are the 12 members of the House Freedom Caucus, a new entity this session, which is just how they would have it.
But that does not mean that the members of the Freedom Caucus vote in lockstep.
According to an analysis by Olivia Krauth, a data and online projects intern at the Statesman, of data provided by RecordVotes.com,a legislative analytics firm that tracks every vote in the Texas Legislature, there were 259 votes in the regular session on which Cain, Stickland and Rinaldi all participated but they did not all vote the same way.
Of those votes, Cain voted differently than the other two in 66 cases, Stickland in 84 cases, and Rinaldi in 109 cases.
If you add in Rep. Tony Tinderholt, the fourth most conservative member on the scale, and Rep. Matt Schaefer, the firth most conservative, the top five conservatives did not vote in unison 483 times. Schaefer, at 161 votes, was the most likely to be in the minority among the five, followed by Stickland at 148, Cain at 111, Tinderholt at 105, and Rinaldi at 103.
“We’re independent thinkers, man,” Cain said.
Here from the Texas Tribune is the explanation by Jones that accompanies the publication of the index.
Political scientists have for decades used roll-call votes cast by members of the U.S. Congress to plot those officeholders on the Liberal-Conservative dimension along which most legislative politics now takes place. This ranking of the Texas House of Representatives does the same — by drawing on the 1,460 non-lopsided roll-call votes taken during the 2017 regular session. As with previous rankings conducted in 2015, 2013 and 2011, this one uses a Bayesian estimation procedure belonging to the family of methodological approaches that represent political science’s gold standard for roll-call vote analysis.
In the chart, Republicans are indicated by red dots and Democrats by blue ones. Based on the roll-call vote analysis, it provides a mean ideal point for each representative, referred to as the Lib-Con Score, along with the 95 percent credible interval (CI) for that estimate. If two legislators’ CIs overlap, their positions on the ideological spectrum might be statistically equivalent, even if their Lib-Con Scores are different. Also included are vertical dashed black lines to indicate the location of the respective median Democratic and Republican representatives.
The table (attached) contains each representative’s Lib-Con Score and rank-ordered position on the Liberal-Conservative dimension, ranging from 1 (most liberal) to 149 (most conservative). House Speaker Joe Straus of San Antonio, who by custom doesn’t ordinarily vote, is not included here. The table also details the ideological location of each representative in relation to his or her co-partisans. In each party, every representative’s ideological location was compared with that of his or her party caucus colleagues and then placed into one of seven mutually exclusive, albeit arbitrary, ordinal ideological categories going from left to right:
- More liberal/less conservative than 2/3
- More liberal/less conservative than 1/2
- More liberal/less conservative than 1/3
- Democratic/Republican center
- More conservative than 1/3
- More conservative than 1/2
- More conservative than 2/3
It’s important to keep in mind that Republicans can register Lib-Con Scores that are noticeably lower than those of most of their fellow Republicans while remaining conservative. It simply signifies that they have voting records that are less conservative than those of most of their fellow Republicans. For example, in the 2017 session every Republican has a Lib-Con Score that is significantly more conservative than that of every Democrat.
Jones’ index looms pretty large, especially in Republican primary politics, where being the most conservative or among the top conservative is like Olympic gold, silver or bronze. Rinaldi and Cain, like Stickland, aspire to lead the pack and they wear the distinction with pride.
“It generally tends to get used by movement conservatives in Republican primaries,” Jones said.
But during the session, Rep. Giovanni Capriglione, R-Southlake, got in touch with Jones, who shared his methodology, and in the closing days of the session, Capriglione issued his own lib-con index, using the Rice methodology with much the same result, ahead of Jones, by way of tweaking the index and its standing.
I thought this an interesting situation and have, since the regular session, talked with Capriglione, Jones, Stickland, Rinaldi and Cain about the index, its validity and what it takes to be the most conservative member of the Texas House.
Here are excerpts from those interviews.
On what cost him sole possession of first place this session. It happened, he said, when he voted with Democrats and only a handful of Republicans against HB 25 eliminating straight-ticket voting. The ultimately successful effort to end straight-ticket voting was inspired in part by the experience of some down ballot Republican judges in Harris County, including District Judge Ryan Patrick, son of Lt. Gov. Dan Patrick, who lost re-election in 2016 because of straight-ticket voting by Democrats for a ticket led by Hillary Clinton, who carried Harris County by 12 points while Patrick losing by less than 3 points.
I took I think it was like 13 votes one night on the straight-ticket voting bill. I was against that. I got scored as being liberal on that because I sided with the Democrats and voted against getting rid of straight ticket voting. For me, in Tarrant County that was the most conservative option.
I would have leapfrogged both of them (Rinaldi and Cain) substantially had it not been for that one issue.
Why was his vote the most conservative option for Tarrant County?
It’s a very red county and in my county we have a lot of minority judges at the bottom of our ticket, very diverse, lot of Hispanics, got some Asians, who are actually very conservative, but I think that may hurt them when an uneducated voter can’t go in there and pull a straight Republican ticket.
So I think that while getting rid of it may save the judges in Harris County, in Tarrant County the dynamics are it’s going to make people who are in safe Republican areas more vulnerable to Democrats.
We have been working for years in the general election to tell our voters to pull a straight Republican ticket and it has worked tremendously. So for me, the conservative movement would have been harmed by getting rid of that because of the region I’m in.
And I didn’t think that was a policy issue where liberty was at stake. That’s literally just a preference deal. No one’s liberty is going to be hurt. No one’s tax dollars are going to be spent. There were no principles at stake in that bill, it was just literally a preference. So once the Constitution wasn’t involved, and once no tax dollars were involved , I defaulted back to what’s the best thing for me district, which is vote against this.
That cost me 13 votes that night.
I will tell you that originally every single Republican representative from Tarrant County was going to vote “no.” One of the reasons was that (Rep.) Stephanie Klick, from Tarrant County, chairman of the GOP for years, and is on the Elections Committee, and she pulled all the data for our county and said this is bad for us and then I think some of the Tarrant County folks got their votes whipped by the lieutenant governor and some other folks, and I think it ended up being me, Stephanie and Gio sticking with the original plan, which was we were going to vote “no.”
That was very purposeful for me.
And to me, the whole motive from the people who were pushing it was how can we get a leg up on our competition through procedure and not based on merit, and I don’t ever play that game, ever. I don’t like changing the rules in the middle of the game.
Stickland cited another occasion when he cast a vote that was out of sync with his party.
On the sanctuary cities bill I was the only Republican, the only one, who voted with all the Democrats to allow them to have debate on sanctuary cities. They were trying to limit the number of amendments and time and I voted against that rule. Now, I was for the sanctuary cities bill and I wanted it to pass but I think it’s bull to remove the Demorats’ ability as members to fight, so I stood up for the correct rules.
It was literally by one vote. They recounted it twice. They thought I was going to break and go back with Republicans and I said, “screw y’all,” and I said, “I’m voting for what I think’s right.” My vote was the decisive vote, yes sir.
What’s wrong with the lib-con index?
It does not matter at all what the bill says, what the bill does. It doesn’t matter whether you voted for this because it’s for your district or what have you. All it says is who you are voting with and who you are voting with helps determine where you are on this index.
To me, and others, how is that a legitimate way to determine whether you are liberal or you are conservative when you are not really looking at whether the bill is liberal or conservative?
And what it sets up is, if 12 members decide they are going to vote “no” and they are the Freedom Caucus, then everybody else according to this program is more liberal.
Does the index ever affect how members vote?
I think it does. I have heard a certain member saying,” I am going up on the Mark Jones index” because of how they are voting.
It’s not too hard if you want to game this. Because it doesn’t matter what the bill is, you just know that if you and at least three other people vote no on a bill, any bill, that no one else is voting on. You are going to do better on this index and in the world of political consulting, in the world of campaign mailers, some people absolutely would prefer a higher score than a lower score
And that’s just an unfortunate byproduct of this. So if you care about this to the point where you want to do better on it, you absolutely can change your votes and change the outcome
Hypothetically, there’s a bill that would help your district but, if you see that you are going to be voting only with Democrats, even if it’s not a liberal-conservative thing, even if it’s a district thin , by doing that, you are hurting your score. It’s one thing if you vote with a mixture of Democrats and Republicans but if you’re a Republican and you vote just with the Dems and you’re one of six or seven who votes just with Democrats, your score is going to get pummeled.
The problem with this particular way of doing things is there is no consistency, there is no analysis. And so no, I’m not saying get rid of it or it shojldn’t exist, I guess all I’m saying ist that the value is almost zero.
Capriglione said with an interest group scorecard he can go to a town hall and answer a question about why he voted the way he voted and scored what he scored.
But with the Rice index, he said, there is no way he can explain or defend your ranking, because it is not based on the merits of individual pieces of legislation.
Before he we elected to the Legislature, Capriglione said, “I didn’t care too much about it because I didn’t understand it. Now that I’ve really dug through the code I think it’s worthless.”
One way to think about this, it looks at how everyone votes compared to how everyone else votes, compared to what the overall dynamics of the vote were. And it takes all that information to locate people on a dimensional plane going from the left to the right.
On Capriglione’s critique.
The thing he misses is that political science has been doing this for a long time.and all the issues he raised have been raised before. It’s part of he scientific process.
People have gone over this argument time and time again, and pretty much all of the evidence we have in political science is that this type of methodology, that effectively lets the legislators behavior define the dimension, is more accurate and more reliable than trying to do interest group scores because the difficulty you have when you try to move away from treating every bill equally, it gets incredibly subjective. What is a conservative vote? How do you weight different votes? Is this one more important or less important?
This is the established methodology that is used by everyone to study the Congress, U.S. legislatures across the states and legislatures around the world. It’s usd both in the academic setting but also people like Nate Silver and Rachel Maddow use it all the time.
This is effectively the gold standard for political science. All credible political scientists would agree that this is the methodology that is optimal for studying roll-call vote behavior and ideological dimensions within legislative bodies. It’s simply where we are. Theres no one that’s arguing otherwise.
That doesn’t’ mean that some of the critiques aren’t valid. It’s just that, like everything in science, it’s relative. Compared to the alternatives this has more merit. It’s the best we have.
And it’s pretty reliable. It has a lot of face validity. When people look at it and they look at their Legislature, that is pretty much what they see.
Jones said that by replicating his method with the same result, Capriglione is “actually showing a positive aspect of this methodology – it’s entirely transparent, there’s no cooking it.”
Can it be gamed?
There’s no way to really game it. You don’t know how any vote is going to weight compared to any other vote. You also need several other people doing it with you and any one vote in and of itself is going to really not have any effect when you have 1,600 votes.
Now I guess you could if you had a group of individuals who always voted he same way and only they were voting that way. But one thing you can’t know is who’s going to vote with you. Let’s say I’m trying to make this a conservative vote but then all of a sudden Sarah Davis, J.D. Sheffield and 20 Democrats vote the same way I do. Then a more conservative vote becomes a more moderate vote. There’s no real easy way to game it.
And one would hope that legislators are either voting on their principles or what their district wants and not based on some index that’s created every other year from a university.
It would be incredibly difficult to do and then you would have to contort your whole voting record, that would have to be your sole goal in voting and I don’t think anybody has that goal.
When Jones wrote up his findings for the 2015 session for TribTalk, he had this to say about Capriglione.
In 2013, Capriglione’s Lib-Con Score was significantly more conservative than that of more than two-thirds of the Republican caucus, with a mere four Republicans to his right and 89 to his left. In 2015, Capriglione’s Lib-Con Score located him two categories to the left in the Republican Center, with 43 representatives to his right and 53 to his left. Capriglione remained to the right of the GOP median in 2015, but he landed quite far from the rightward edge of the ideological spectrum where he was situated in 2013.
Jones elaborated for me:
When he came in his freshman year he was among the very most conservative and in his sophomore year he shifted dramatically to the center. What we find is that is very rare, that normally what we find is people’s ideological position based on their roll call voting behavior tends to be pretty stable. You tend not to see people move around too much.
It’s very rare when somebody goes dramatically from one part of the spectrum to another. That suggests they made a very conscious choice to change the way they were voting. That just doesn’t happen by happenstance. That involves a very conscious change that involves a conscious decision to change the way one votes and the way one relates to the leadership, that just doesn’t happen randomly.
I’m a computer geek. I find this stuff extremely interesting and I spent a lot of my time in my physics classes doing statistics, in my business classes doing statistics. I’m a computer guy. I guess htat’s why. For what it;s worth, which I think is zero, I’ve improved since last time, according to Mark Jones. (On the 2017 scale, there are 38 Republicans to Capriglione’s right, and 55 Republicans to his left,)
I graduated with a physics degree. I went into electrical engineering where I was software engineer. Then I designed semi-conductor chips for a telecom – high-speed mixed signal stuff. I’ve been programming since I was ten – I got my first computer a TRS-80. I still know between five and six programming languages today.
To me this I fascinating. More than it is a political issue it’s something I spend a lot of time on as a kid, coming up with my own algorithms to help me with my statistics and my electrodynamics homework. So I’m just here doing a little project that happens to be political but at the same time its really about computers.
Now I’m in private equity. I was doing electrical engineering and design for Lucent Technologies, and then I got my MBA and worked at first in investing in technology companies, and now I just do all sorts of businesses investing.
In other words, Capriglione said of his public critiquing the lib-con index, “People can say, “Oh he ‘s doing it because he’s sore.’ No, actually it’s because I’m a nerd.”
On Jones and his index:
He’s very well-known. He’s very well-respected.. I think he’s come up with a system that’s better than any other I’ve seen that takes into account thousand of vote on different topics without bringing into account human bias.
It’s even better than the interest group scorecards because it doesn’t pick and choose. The interest groups pick the votes that are most important to them. What this does is to give you a more complete picture by taking the hundreds of votes on little items that we do constantly.
What it does measure is partisanship, how often you correlate with members of your own party as opposed to the Democratic side, and if you start with the premise that Democrats are not voting for conservative policies, and then you make a correlation, you have a picture of who’s more conservative.
I think it’s generally accurate and it generally comes out the way you would expect it and it generally mirrors what the conservative interest groups do.
Like if you voted for continuing with straight ticket voting, that is something that would have been scored against you in the Jones rankings because most Democrats voted that way, but in reality that’s only a small proportion of the couple thousand votes we take so I think overall it paints a pretty good picture.
The fact that you’re taking every vote and putting it into the equation makes is so that those (individual votes) aren’t really determinative, which is why its good overall picture what people are doing.
A lot of the interest groups will take the big-ticket votes and it’s very easy to manipulate those. You know ahead of time they are a scoring it. But what this does, is we pass 800 bills and a lot of those bills incrementally increase he size of government, incrementally increase regulation, and it shows you who in vote after vote is just saying no. Even if it’s only a small issue, they hold the line on those little votes that will never make their way into a scorecard or a newspaper or anything else.
At the end of the session, I always hear about the five or six big-ticket conservative items hat we passed, and I’m thrilled that we passed those items. However, what I’m not thrilled about is the increase in regulation in bill after bill – the bills hat were worked out on tiny little issues between competing lobby groups that actually hurt consumers, and I see more of those bills come up, I see hundreds of those bills come up and it’s usually the bills, where you see ten to thirty people voting against – and those are just as important to Texas voters I think as the big-ticket items.
Case in point, said Rinaldi, was the bill Capriglione cited in the tweet above.
The Senfronia Thompson bill that required hairdresssers to get training in how to spot human trafficking, a bill that regulates hairdressers to say that you can’t practice your profession unless you take training in how to spot trafficking – well meaning, like most of these are, but in the end increases regulation.
For Gio it was a bill that was conservative but was scored negatively, I would say just the opposite of that. I don’t think turning hairdressers into sex trafficking vigilantes is really conservative. I think what’s conservative is, “I can provide a service and you have money. How about we trade your service for money?”
On surpassing Stickland last the most conservative representative in the 2015 session.
When I overtook Stickland last year, Jonathan, on some criminal justice issues is a criminal justice reformer more so than I am. I tend to be a little bit more retributive and harsher on penalties. So those votes put me over the top.
It was a mission to be at the top of the list. I didn’t know if we could make number one, but it was a goal.
Why would this be such a point of pride?
This is Texas.
How did he end up desk mates on the floor Jonathan Stickland?
There was a vacant seat back there and heck, going in, I wanted to sit next to Jonathan Stickland.
Stickland’s advice and mentorship came in understanding what was going on on the floor. Why people were at the microphone and the games and the theories that were really going on.
Not voting. I didn’t need somebody to tell me how to vote. For us (members of the Freedom Caucus) that’s really easy. There are really simple tests for us to decide whether or not they are going to support something..
Can the competition to be most conservative get up close and personal.
We just joke among ourselves, friendly banter but we’re not truly being insulting. We’re just kind of being fraternal brothers, joshing around.
Cain and his staff prepare meticulously for every vote.
I have a spread sheet on my desk – a few of us do – on who cares (about each bill), who testified in favor of it. who has a dog in the fight.
And then he applies a set of principles to each piece of legislation.
It’s based on whether or not I believe that vote is conservative or liberal. Does that vote grow government or spend money.
Everyone has their own little step paths and I don’t have mine written down but I know other members have it in their office, when they’re deciding something. Does it grow government, or is it a necessary function, is it good for a good people?
Even these things that might not seem to matter to everybody, it might be a district-specific thing. It might be a hotel tax in El Paso and I’m going to vote against it, and someone will be, “Why are you voting against it, it doesn’t effect you?” Well, it’s a tax. For me it’s just that simple so I am going to vote against it, even though it doesn’t affect my district.
He agrees with Rinaldi, it is the little votes that make the difference in becoming the most conservative member of the House.
They are completely innocuous I don’t’ know what they are. They are not going to be big-ticket items. They are things that have not been in the media. They’ll be just this thing that we ask ourselves., “Is this the proper role of government. Is the proper function of local government?”
My staff and the people I hire and the people I have helping me, we spend countless hours of preparation on every single bill.
A bill might be 90 percent good and 10 percent bad. Sometimes that’s OK. But if the 10 percent is violating one of our core liberty principles,then we vote against it.
So it just depends on what’s bad about it a lot of times.
Something can be more conservative but also violate a limited government ideal, and both may be correct. Big picture were talking conservative vs. liberal, but on these smaller details, you’re talking about limited government.
Prime example gay marriage. I personally think marriage should be between a man and woman, but I also think government should have nothing to do with marriage of anyone.
This generally happens on amendments, because on bills that are already filed we can do the research, but when amendments are flying on the floor you’ll see very often me jump up and say, “Hold on, give us time to read it,” or I’ll start to ask questions. If I get to the point where they’re forcing me to vote and I can’t slow it down and I can’t understand the implications of it, my default vote is always “no,” always, because I can always explain a “no” vote.
So it’s just being a stickler on stuff like that.
Generally speaking, except for a few partisan votes, Democrats generally vote yes on everything.
Meanwhile, Capriglione’s interest in what accounts for how members vote goes beyond his analysis of the lib-con index.
Something that’s extremely fascinating for me are what are some of the things that do affect how legislators vote. It’s of course the bill. It’s their party. It’s their committees. If you’re on the Transportation Committee and you already voted on that bill you are likely to vote for it again.
There are groups of people who talk among themselves, and they will go, “I’m rural, he’s rural we should talk about this bill. We should probably vote the same if we have the same interests and it affects out area.”
And where you’re seated, where you’re located has a huge correlation to how you ultimately vote because a lot of times you have discussions with them as a bill coms up and if you’re leaning one way and you talk to the person next to you – ‘No, it actually it does this or doesn’t do that,’ – proximity just lends itself obviously to how you vote.
Capriglione said that like-minded members tend to sit near one another, and their proximity re-enforces their affinity. He created the graphic – a dendrogram – below, that shows the connected pairs and clusters of House members who vote most alike.
If you look at the dendrogram, you’ll see myself, Greg Bonnen, Craig Goldman, Drew Springer and Ron Simmons were the five closest when it comes to voting, we were the most similar and we sit right next to each other. I decided to sit next to them because I talk to them a lot and we agree a lot, and Goldman’s from Fort Wroth and Simmons from Lewisville an Drew is my roommate
I decided to sit next to them because I guess we agree on a lot and I guess that’s why we vote the same.
(The Giovanni Capriglione Dendrogram of the Texas House showing pairs and clusters of members who vote most alike.)