Hoover Institution visiting fellow Daron R. Shaw is University Distinguished Teaching Professor and Frank C. Erwin Jr. Chair of State Politics at the University of Texas at Austin. He specializes in American government, campaigns, and elections; political parties; public opinion and voting behavior; and applied survey research. He spoke to Chris Herhalt about some of his recent work tracking the impact of interstate migration on the Texas electorate and why it sometimes pays for a politician not to respond to a media request.
Chris Herhalt: Let’s start with your Texas migration research. You point out that lately the average interstate migrant to Texas is in fact slightly more conservative than the Texas population as a whole, and that the strongest predictor of being a Texas migrant is itself a strong personal interest in politics. What does that mean for elections in Texas?
Daron R. Shaw: The conventional wisdom, which has been around ever since I’ve come to Texas, is that even though this state has a conservative bent and a solid Republican majority, the demography of the state is going to push it in, at the very least, a more moderate but probably ultimately a more Democratic or liberal direction. And the underlying demography of that intuition is fairly obvious to anybody who’s looked at Texas census data or Texas population figures. The ascendant population is younger, it’s more Latino, it’s less white. It’s comprised of these groups that have historically tended to lean Democratic.
The algebra is pretty straightforward, but I’m a political scientist. And one of the things that drove me crazy and that continues to drive me crazy about a lot of these analyses is that all of that is true, if you hold politics constant. The idea that political issues and candidates and coalitions will remain unchanged in perpetuity always struck me as farcical. It just makes no sense. And when it comes to Texas, there’s also been, at least until perhaps the past ten years, a tendency to underrate the extent to which migration from other states and countries might impact things. In the mid-2010s, you began to see interesting analyses from Jim Gimpel (Wendy Cho has done some too) assessing migration particularly from the Pacific states, and to a lesser extent the upper Midwest states, to Texas and how that might change the politics.
Gimpel, Cho, and other researchers noted that the census showed a significant migration into Texas from California and other states, but they could only look at very broad measures of politically relevant variables. For instance, in Texas we don’t register by party. You just register, and then you can vote in either Republican or Democratic primaries. Somebody moving from California might be a registered Democrat, but after they come to Texas the best we can say is “well, they vote in the Democratic primary.” So, the analysis tended to treat that as a Democratic addition to the Texas electorate. But that’s imperfect. Registration and primary participation doesn’t necessarily reflect attitudes perfectly. So, our measurement at the back end—this is what these people think when they get to Texas and how they behave—was very difficult to assess from voter files.
At the back end of our surveys of Texas, we ask, “Are you from Texas or did you move here from somewhere else?” But what we’ve also been able to do, since it’s a survey, is ask, “Who’d you vote for in this election? Are you a Republican or Democrat? Liberal or conservative?” We also ask them if they have moved from another state, and which one. And as you pointed out, the main distinguishing variable is the attitudinal structure of people moving here: they tend to be more engaged and more politically interested, but they also tend to exhibit higher levels of political independence, which very much surprised us. So, the other key element to the story is that there’s a complexity to migration that I think you and I would probably understand if we thought about it.
People coming from Louisiana and Florida are much more likely to be African-American. They tend to be more Democratic, more liberal. A lot of these people, we postulate, came here because of Hurricane Katrina or some of the hurricanes in Florida, and they don’t move randomly throughout the state. Most go to Houston. People coming from the upper Midwest, and especially some of the border states like Oklahoma and Arkansas, are moving into the [Dallas–Fort Worth–Arlington] metroplex, and they tend to be a little more conservative.
But there’s also migration into Dallas and Fort Worth from California, which is a little more liberal on the whole. And there are two hypotheses associated with that migration. One would be that these are Silicon Valley tech workers coming to live in the tech corridor between Dallas and San Antonio and up through Austin, and therefore they’re going to bring their liberal sensibilities with them. Another hypothesis is that these are disaffected conservatives who are getting out of California because they’re mad and they don’t want to pay high taxes. Both are true, so the net tends to be not terribly distinct. And it doesn’t seem to be pushing politics clearly one way or the other.
So, what we’re finding is that in the metropolitan areas, you tend to see people moving in who are more liberal than the people already there—not hugely, but enough to move the dial a point or two in the rest of the state. But outside Dallas, Houston, Austin, San Antonio, the people moving in are slightly more conservative. The rural and small-town places are getting redder, and the metropoles are getting a little bluer. The result is that we’re seeing greater polarization, even though the net effect is almost zero.
Chris Herhalt: It’s not fulfilling the myth of the past several decades that the state is going purple or it’s going blue?
Daron R. Shaw: “They’re turning Texas blue”—it’s been a Democratic project for a while. There’s another metaphor that drives me a little crazy, personally: it’s the “sleeping giant” reference to the Hispanic population. It’s not that I don’t think Latinos are going to have a huge impact on politics—that Latinos are becoming a larger portion of the population is undeniable—but the political impact of that has yet to be determined. You’re aware there’s been a huge movement in the Rio Grande Valley and other places, I wouldn’t say necessarily towards Republicans, but away from the Democrats. In other words, towards a sort of place in the middle where you have to prove it to me (as a Latino) to win my allegiance.
Chris Herhalt: I’ve heard Senator Ruben Gallego [D-Arizona] talk about something similar: you approach the Latino in the Southern United States with your Los Angeles liberal sensibility, and that only puts them off.
Daron R. Shaw: I’m from San Diego, and San Diego has yet to have a Hispanic mayor. San Antonio has had half a dozen. It’s interesting that there’s this notion that immigration attitudes are much more progressive in California, and the Latino population itself is different. But another fact is that the divide within Texas among Latinos is becoming more acute. I was listening to Sergio Garcia-Rios and Ross Hunt. Sergio’s here at the University of Texas and Ross is a Republican consultant. They both agreed that what you’re seeing now is that the movement away from the Democrats in the Rio Grande Valley has not abated. It continues, but north of San Antonio, Latinos look like they’re kind of boomeranging back to the Democratic Party in the first year or so of the [second] Trump administration. I guess that’s a long-winded way of speaking to the complexity and diversity of these geographic politics within Texas. In California, you see much the same thing.
Chris Herhalt: In your research about whether elected officials should respond to the media, you find something that really puzzles me as a former political journalist. You found that “a strategic silence,” or situation where a public officeholder just simply refuses to comment, not even to say “no comment,” when asked about something damaging can minimize the electoral damage of a negative story. Now, in all my experience, doing that just lets the other stakeholders in that story, the opposition or members of the public, crowd the officeholder out of the story and gain the attention and oxygen that would otherwise go to the officeholder. I don’t dispute your findings but I’m wondering, why do you think this approach is effective?
Daron R. Shaw: There’s a theory of political science called issue ownership. It’s associated with John Petrocik, who was at UCLA and later chair at Missouri. Petrocik argued that the parties have differential credibility on different issues, and this is born of the coalitions that serve the political parties and the dynamic between party constituencies and issues. For instance, I could ask you or anybody in an audience, straight up, do you think the Democrats or Republicans would do a better job of handling issue X? If I said health care, most people would say the Democrats. If I said crime, most people would say the Republicans. We could go down the list: for the environment, civil rights, taxes, national security, voters tend to see one or the other party as better able to handle the issue. Petrocik’s argument is that electorally, the consequence of this is that campaigns are about agenda setting. If the election in 2026 is going to be about border security, taxes, and national security, I can tell you right now that the Republicans are going to win. If the election, on the other hand, is about health care and climate change—and I was going to say protecting the poor, but that’s kind of up in the air in recent years—and Social Security, the Democrats are going to win.
So, the controversial part of Petrocik’s notion is that engaging on an issue where you don’t have an advantage isn’t going to be very useful.
This has informed our expectations about how public officials and candidates handle media inquiries. We developed a study that looked at Governor Greg Abbott in Texas and all the media inquiries that came to him, and we kept a record of who he responded to and how his staff was doing so. Was it a text message, an e-mail, a phone call? What was the issue? Was it a substantive response? And what was the request about: was it for an interview or just for a comment? And then we measured public opinion in response to these inquiries and responses.
This period, May and June of 2023, at the very end of the Texas legislative session, was a very hot political time. A lot was going on. The central question asked by Abbott’s staff and by his campaign team, who spent enormous effort and resources responding to media inquiries: Was it worth it? Should we even talk to the media anymore? I think a lot of your audience would think that’s an interesting question in an age where you can simply put out your own version of the story on social media. Do you need to talk to the Dallas Morning News or the local Fox News affiliate in San Antonio? Do they matter anymore?
And the short answer is they do matter. But engaging on unfriendly issues—perhaps that’s not the phrase, but issues on which you’re on the defense—is probably a mistake.
I think their assumption was that for a lot of these issues, the media had a particular frame they were pursuing, and they did not want to be involved in what they saw as the media’s effort to reframe it. It wasn’t done reporter by reporter. It didn’t seem to matter where it came from. They were a little more likely to respond to local broadcast outlets than to newspapers, because they clearly think local broadcast still reaches people they care about, and I don’t think they believe that of newspapers.
Chris Herhalt: In in the same study, you talk about the concept of helping a candidate “prime” a voter rather than “persuade” a voter. Could you explain that?
Daron R. Shaw: It also draws on this theory of issue ownership. I don’t actually think the Abbott people had a consistent approach to deciding who to respond to and who not to. For instance, we saw—and anybody who’s worked in a campaign can verify this—their staff is very young, and the staff was more likely to respond to a text message than an e-mail. I think that absolutely has to do with young people and their preferred mode of communication, not some strategic decision that text responses were going to be better than e-mail responses.
What they did seem to want to do was not to get “their side” of the issue out there. I think that’s where you’re coming from: the traditional approach to the news media. The reason you do “earned media” is to get your side of whatever the argument is, whatever the debate is, into the story. And I think their approach was that they were not interested in trying to persuade voters on certain issues because they didn’t seem to think the media afforded them an opportunity to do that. I believe they thought, if you’re doing a story on this (an issue on which we have limited credibility with the public and the news media), that’s just not going to be good for us. Conversely, they were interested in lending their credibility, lending their name, to stories they thought were inherently good for them, where it didn’t matter what the media said.
Chris Herhalt: Is it still true that the pool of independents, true independent voters, across successive elections, keeps shrinking? Are people locking in at an ever-greater rate?
Daron R. Shaw: Well, it depends. The way we typically measure party is to ask an initial question: “Do you generally consider yourself a Republican or Democrat, or don’t you think of yourself as either?” We call it the directional component of party ID: R versus D. About 30 to 40 percent of Americans say, “I’m neither.” Then we ask a secondary question: “Well, if you had to say, which way would you lean: closer to the Democrats, or closer to the Republicans?” So, when people look for a percentage of independents in a poll story, sometimes they’ll see 40 percent and sometimes they’ll see 12 percent, depending on whether the pollster or the writer treats independent leaners as partisans or as true independents. The 12 percent is closer to my view—I view leaners as partisans. When someone says, “I lean Republican,” that means there’s a 90 percent chance they’re voting Republican.
So, to answer your question, there’s been a significant uptick in the percentage of independents if you count leaners as independents. What you’re getting is a lot more people who say they’re independent, but in fact, they’re essentially partisans. And that speaks to your point. The portion of people who truly are movable is one in eight or one in ten these days, and it has decreased over time. You do see important increases, though, especially with younger people.
This interview was edited for length and clarity.
