NEW REPORT: Democrats Lost South Texas Voters on Values. Republicans Are Now Losing Them on the Economy.
A new study of 2,866 voters across four South Texas districts finds the people leaving each party give completely different reasons and the region is up for grabs.
May 27 2026 – A new report from the Welcome Democracy Institute finds that the voters who turned South Texas red and the voters now drifting back to Democrats are moving for opposite reasons, a pattern suggesting the region is still up for grabs.
When South Texas voters left the Democratic Party between 2016 and 2024, 28-38% said the same thing: the party walked away from their values. Immigration came second (18-26%). The economy was a distant fourth (13-19%).
When voters now leaving the Republican Party explain why, the picture flips. 29-42% point to the cost of living. Just 3-12% mention immigration.
The report, “Examining the South Texas Swing,” draws on two live focus groups in Pharr, Texas, closed-end polling of 2,190 registered voters, and 2,866 open-ended survey responses across TX-15, TX-28, TX-34, and TX-35—the four districts at the center of the region's transformation.
Key Findings
Two defections, two stories. Voters leaving Democrats tell a story about values and identity. Voters leaving Republicans tell a story about groceries and rent. Each party is losing ground on the issue it was supposed to own.
The swing was real, but the region is volatile, not realigned. The aggregate presidential margin across the four districts moved from D+27.6 in 2016 to R+8.2 in 2024, a 35-point swing.
Voters want moderation, and they're specific about it. In the open-ended responses, 36 respondents used the literal language of "move to the center" or "be more moderate" unprompted, and voters across party lines described the same candidate profile: pro-Second Amendment, fiscally disciplined, tough on the border with legal pathways, and pro-choice without absolutism.
Immigration attitudes are more nuanced than national narratives suggest. Voters overall preferred Trump's approach to immigration over Biden's by 18-24 points depending on the district. In the open-ended responses, voters rejected both "open borders" and aggressive deportation of longtime residents, consistently describing a preference for border security, orderly processing, legal pathways, and prioritization of violent offenders. The winning position, as one Democrat in Hidalgo County put it, is that there has to be a middle ground.
Blue Dogs are running ahead of the trend. Reps. Henry Cuellar (TX-28) and Vicente Gonzalez (TX-34) outperformed the 2024 presidential ticket in their districts by 12.7 and 6.7 points respectively.
"Moderates" and "Independents" are not the same voter. On the five key measures tested, the two groups sit 25-34 points apart on average. Moderates lean center-left; independents lean policy-conservative but partisan-skeptical. Campaigns that target them as a single "persuadable middle" are mis-targeting.
Affordability is the one word every group defines the same way. 877 respondents used concrete affordability language in their open-ends: groceries, gas, rent, healthcare, wages that don't keep up. It is the cleanest cross-pressure vocabulary in the dataset, and it is the issue Republicans are now bleeding voters on.
Water is the largest untapped local issue. 169 respondents raised water, drought, desalination, or aquifer depletion unprompted. Corpus Christi's stalled desalination plant draws cross-partisan blame in identical terms.
Hispanic voters do not respond as a bloc. On immigration, affordability, and reasons for switching parties, party ID predicts views far better than ethnicity does. Messages built around Hispanic identity will likely fall flat. Messages built around partisanship, working-class status, and place will likely land.
What voters actually said:
The open-ended responses are the heart of the report. The asymmetry runs through them:
"I'm 57 years old and I have been a lifelong Democrat, but the Democratic Party, they're off their rocker and my values are more in line with the Republican Party these days."
— Republican | Hidalgo County, TX-15 | Hispanic man, 57
"I was a Democrat back when Democrats were big with the NRA. They were much more conservative. They're so far left that it's ridiculous."
— Republican | Cameron County, TX-34 | White woman, 81
And on the other side:
"Affordability, for me as a teacher and a mother of three, means being able to meet my family's basic needs without constant financial stress. It looks like keeping everyday essentials within reach, like not having to spend $200 a week on regular groceries or worrying about whether I can afford to fill up the gas tank."
— Democrat | Bexar County, TX-35 | White woman, 38
"A ballroom? Really? Remodeling the White House, when here in Corpus Christi, we're still arguing about whether or not we can have a desalinization plant because we don't have enough rainfall to keep our Corpus Christi Lake in top shape."
— Ind-lean Rep | Nueces County, TX-34 | White woman, 34
Methodology
Closed-end polling was conducted across TX-15 (n=406), TX-28 (n=412), TX-34 (n=549), and TX-35 (n=823) between April 1 and April 14, 2026, weighted by race, age, gender, education, and 2024 presidential vote recall. The open-ended instrument drew 2,866 responses across the four districts in the same window. Focus groups of independent Hispanic voters were conducted by Global Strategy Group and Applecart in Pharr, Texas on March 11, 2026.
###
About Welcome Democracy Institute
The Welcome Democracy Institute works to strengthen American democracy by encouraging pragmatic political leadership, solutions-oriented governance, and independent political participation. Through research, convenings, strategic communication, and voter engagement, the Institute advances centrist approaches to local, state, and federal policy challenges. Its flagship work includes the quarterly Congressional Competitiveness Index, which tracks where parties are failing to contest winnable districts, and The Depolarizers podcast. Learn more at welcomedemocracy.org.