Our exploration of straight-ticket voting trends in competitive or potentially competitive districts turns at long last to CD23, the state’s lone competitive congressional district and one of only a handful of swing districts nationally. CD23 sprawls across 29 counties in two time zones from El Paso to San Antonio, including some of the reddest and bluest portions of the state.

CD23 has sent a different representative to Congress each of the last four election cycles, two Republicans and two Democrats. Only former U.S. Rep. Ciro Rodriguez (D-San Antonio) has won a re-election race – just one – since former U.S. Rep. Henry Bonilla (R-San Antonio) won his final term in 2004.

U.S. Rep. Will Hurd

U.S. Rep.
Will Hurd

Pete Gallego


This year is a rematch of 2014, when now-U.S. Rep. Will Hurd (R-San Antonio) defeated then-U.S. Rep. Pete Gallego (D-Alpine), 49.8%-47.7%, thanks to a 9,300-vote margin in Bexar Co. Gallego won the rest of the district by 6,900 votes, although Hurd won half of the other counties. Libertarian Ruben Corvalan, a San Antonio electrical engineer, received 3% of the vote and is on the ballot again this year.

The partisan leanings of the district depend on whether it is a presidential or a gubernatorial election year. This in part explains why the seat flipped from Republicans to Democrats and back again twice in the last decade. It was held by the conservative Bonilla for seven terms, largely because of margins built in Bexar Co. Redistricting in 2003 removed heavily Democratic Webb Co. from the district, along with Bonilla’s formidable 2002 challenger, and added some friendlier Hill Country counties. Bonilla won re-election easily. The district was adjusted by the courts in 2006, replacing those Hill Country counties with more Democrat-friendly South and West Texas counties, increasing the Hispanic/Latino share of the voting-age population, which today is nearly two thirds of the voting-age population of CD23.

Bonilla nearly won an 8-way special election outright over six Democrats and one independent, but he lost a December runoff election to Rodriguez, who was re-elected in the Democratic boom year of 2008. The seat returned to Republican hands for one term in the Democratic bust year of 2010. After Gallego’s one term, the seat is once again held by a Republican.

Recent CD23 Election History

2014 – Will Hurd (R) 49.8%, U.S. Rep. Pete Gallego (D) 47.7%

2012 – Pete Gallego (D) 50.3%, U.S. Rep. Quico Canseco (R) 45.6%

2010 – Quico Canseco (R) 49.4%, U.S. Rep. Ciro Rodriguez (D) 44.4%

2008 – U.S. Rep. Ciro Rodriguez (D) 55.8%, Lyle Larson (R) 41.9%

2006 – Ciro Rodriguez (D) 54.3%, U.S. Rep. Henry Bonilla (R) 45.7% (special runoff election)

2004 – U.S. Rep. Henry Bonilla (R) 69.3%, Joe Sullivan (D) 29.4%

2002 – U.S. Rep. Henry Bonilla (R) 51.5%, Henry Cuellar (D) 47.2%

2000 – U.S. Rep. Henry Bonilla (R) 59.3%, Isidro Garza Jr. (D) 38.8%

Just 16% of Bexar Co. is part of CD23, but it is far and away the largest geographic source of votes, and it’s growing faster than the rest of the district. Bexar Co. accounted for 47% of all votes cast in the district in 2012, up from 44% in 2008, 39% in 2004 and 34% in 2000. Republican-friendly Medina Co. provides the second-largest number of voters, just edging out the portion of Democrat-friendly El Paso Co. located in the district.

Mitt Romney narrowly won the district in 2012, 51%-48%, over President Obama, and Ted Cruz carried the district by six points. One Democratic Supreme Court nominee carried the district, but Republicans won the rest of the statewide races. Obama narrowly carried the district in 2008, 49.9%-49.3%, as did Democratic challenger Rick Noriega over U.S. Sen. John Cornyn, 50%-48%, and all of the Democratic statewide candidates. Democrats had a huge advantage in straight-ticket voting that year (~17,700 votes), but Republicans made up much of that ground in 2012. Republicans held the edge in straight-ticket voting in 2010 and 2014.

That back-and-forth shift in straight-ticket voting is a large driver of the biennial changing of the partisan guard in CD23.

Estimates of the number of straight-ticket and full-ballot votes cast in the 2004, 2008 and 2012 presidential elections in the counties and precincts currently comprising HD43 are shown below.


  • Straight Republican – 38,100 40% 40%
  • Straight Democratic – 42,900 44% 44%
  • Full Ballot – 90,700 93% 93%

Democratic Advantage: ~4,800 votes


Straight Republican


Straight Democratic


  • Straight Republican – 40,000 41% 41%
  • Straight Democratic – 57,700 59% 59%
  • Full Ballot – 97,800 100% 100%

Democratic Advantage: ~17,700 votes


Straight Republican


Straight Democratic


  • Straight Republican – 54,000 55% 55%
  • Straight Democratic – 59,300 61% 61%
  • Full Ballot – 80,700 83% 83%

Democratic Advantage: ~17,700 votes


Straight Republican


Straight Democratic

Straight-ticket voters are an increasing share of the overall electorate in the counties and precincts currently comprising CD23. They represented 47% of all votes cast in 2004, 50% in 2008 and 59% in 2012. Democrats saw their straight-ticket numbers surge in 2008, jumping 34% from 2004, then rise slightly in 2012. Republican straight-ticket voting grew slightly between 2004 and 2008, then surged 35% by 2012, cutting the Democrats’ advantage from ~17,700 votes in 2008 to around ~5,300 votes in 2012.

We expect Democrats to maintain an advantage in straight-ticket voting, but it would not surprise us if the advantage shrinks further as Bexar Co. voters gain increasing sway in the district and population growth, and voter registration, in Republican areas adds to their numbers.

Republicans’ advantage in Bexar Co. is increasing. In 2000, there were ~2,100 more Republican straight-ticket votes than Democratic ones in the Bexar Co. precincts currently included in CD23. The number of Republican straight-ticket votes increased 64% in 2004, increasing their advantage in Bexar Co. to ~7,400 votes. Strong Democratic turnout in 2008 cut the Republican edge to less than 2,000 votes, but it rebounded to 7,565 in 2012.

In 2012, there were eight precincts in Bexar Co. where more than 1K Republican straight-ticket votes were cast. Canseco received 71% of the vote in those precincts – a margin of 8,683 votes. More than 72% of registered voters cast ballots, and almost half of them cast a Republican straight-ticket vote, and they represented more than 10% of all votes cast in CD23. Voter registrations are up 13% in those precincts as of the 2016 primary. Gallego edged out Canseco in the rest of Bexar Co., so those eight precincts are crucial to Hurd.

In 2012, Canseco, an incumbent, received 12K fewer votes than Romney and 11K fewer than Cruz while Gallego slightly outperformed Obama. Hurd will need to hew closer to Donald Trump – probably outperform him – to retain his seat. Trump’s performance in Hispanic/Latino-majority counties will go a long way in determining how uphill Hurd’s climb actually is.

Population growth is faster, and voter turnout is higher, in portions of the district that are friendly to Republicans. We have already detailed the growth in Bexar Co.’s most heavily Republican (at least in terms of volume) CD23 precincts. Aside from it and El Paso Co., six other counties have a population above 15K: Frio, Maverick, Medina, Pecos, Uvalde and Val Verde Cos. Gallego won Frio, Maverick, Pecos, Uvalde and Val Verde Cos. by a combined 9,882 votes in 2012. Non-suspense voter registrations are up 3.6% in those counties since then. Canseco won Medina Co. by 4,974 votes in 2012. Its non-suspense voter registration total is up 12.5% since then. Voter turnout in 2012 was 61% in Medina Co. It was 45% in the other six counties.

Gallego should have a slight advantage from straight-ticket voting – probably less than 3% – and he has historically performed better than other Democratic candidates among full-ballot voters.

However, the district is “swing” enough to provide Hurd a path to a narrow victory, unless that path is ruined by Corvalan, the Libertarian candidate. A little over half of all the votes received by the Libertarian candidate in 2012 came from Bexar Co. Many of those 3,058 would have gone to Canseco had it been a head-to-head matchup. That would not have been enough to beat Gallego then, but it might be enough now.

We expect a narrow Gallego victory but would not be surprised if the district swings, ever so slightly, the other way. We expect neither major party candidate to break 50% despite the millions and millions of dollars pouring into the district. A Hurd victory could set the stage for another Bonilla-like run, particularly if a Democrat happened to occupy the White House.

If Gallego were to win, we would expect his re-election race to be uphill regardless of who is president. Republican candidates had about a 2,700-vote advantage in straight-ticket voting in 2014, which was just about the margin between Gallego and Hurd. A certain Castro atop the Democratic ticket could change that math, but we’re getting ahead of ourselves. As U.S. Sen. John Cornyn said today, when asked if he would endorse Cruz’s re-election bid, “One election at a time.”


Because of redistricting and shifts in precinct boundaries over time, we can only estimate straight-ticket voting prior to the last time the district was drawn (2011). We estimated straight-ticket vote totals by applying the current district boundaries backward in time to the precincts as they existed in each general election. The farther back we go, the more estimating is required. We use mapping software to identify which voting precincts lied entirely or partially in the current boundaries of CD23 in each even-year general election. We used precinct-level results from each general election. For each election prior to 2012, we allocated 100%, 75%, 50% or 25% of the votes cast in those precincts to today’s districts based on the approximate geographic area of the precinct within the district. We totaled the resulting precinct- and split precinct-level data to estimate district-wide straight-ticket votes. We rounded to the nearest 100 to avoid appearing too precise in our estimation. For the 2012 and 2014 general elections, we were able to use the current voting precincts within the current districts without the need to allocate votes across split precincts. Because we could use complete precincts, we chose to provide the accurate vote totals. We thank the Texas Legislative Council for providing the shapefile data necessary to conduct this analysis. We obtained precinct-level election results data from county election officials.

County-level Data Availability, 2000-2014

Bexar Co. – 2000-2014
Brewster Co. – none
Crane Co. – 2008, 2014
Crockett Co. – 2000-2004, 2008-2014
Culberson Co. – none

Dimmit Co. – 2000, 2002, 2006, 2010-2014
Edwards Co. – 2006, 2010-2014
El Paso Co. – 2000-2014
Frio Co. – 2000, 2002, 2008-2014
Hudspeth Co. – none

Jeff Davis Co. – 2006-2014
Kinney Co. – 2014
La Salle Co. – 2000, 2002, 2008-2014
Loving Co. – none
Maverick Co. – 2008-2008, 2012, 2014

Medina Co. – 2000-2004, 2008-2014
Pecos Co. – 2000-2014
Presidio Co. – none
Reagan Co. – 2008-2014
Reeves Co. – 2000-2014

Schleicher Co. – 2006-2014
Sutton Co. – 2008-2014
Terrell Co. – 2004, 2008-2014
Upton Co. – 2000, 2012-2014
Uvalde Co. – 2006, 2008, 2012, 2014

Val Verde Co. – 2000-2014
Ward Co. – 2000-2014
Winkler Co. – 2002-2014
Zavala Co. – 2000-2006, 2010-2014