Deep in the Heart of Texas

What happened with the Hispanic vote in Texas in 2020? It tanked big time for the Democrats and analysts are still trying to figure out why. I present below a valiant attempt by Benjiman Lefkowitz of Decision Desk HQ to do just that. And I reproduce his entire analysis since I can't seem to find anywhere where it lives on the web-- I got the analysis in the newsletter they send out. Unfortunately the neat maps, of which there are many do not show up, so you'll have to get those elsewhere--perhaps by subscribing to their newsletter.

"Texas Democrats went into the 2020 election with high hopes. Congressman Robert “Beto” O’Rourke's 2018 narrow U.S. Senate seat loss to Republican incumbent Ted Cruz seemed to show the way forward towards future victories. O’Rourke won a majority of State House Districts on a map designed to reelect a GOP majority. The Congressional gerrymander, built by Republicans to protect Republican congressmen, strained under the weight of new suburban and Democratic voters. Texas’s booming growth in urban and suburban areas fortified O’Rourke’s 48.3% tally to position Democrats for statewide success. Turnout numbers before the 2020 general election signaled that voters across Texas recognized their state would end up competitive, and therefore acted accordingly.

The electorate’s votes failed to match pre-election expectations. Texas Republicans won the majority of competitive races and proved their well of support is deep enough to keep the state red. Most surprisingly, former President Trump netted votes in the majority-Hispanic counties along the Rio Grande, leading to two opposite yet equally important electoral stories. The Democratic margin collapsed by over 13 points from 2018 in these combined South and Southwest Texas counties, going from an O’Rourke 15-point margin over Ted Cruz (Clinton also won the region by a 15 margin in 2016), or 57.5% to 41.9%, to a Biden victory of just over 2 points, or 50.5% to 48.2%. Trump won the rest of Texas outside of these counties by 52.5% to President Biden’s 46%. This was barely an improvement over Cruz’s 4.6% margin in 2018, or 51.9% of the vote to O’Rourke’s 47.3%.

Swing In Texas Counties between 2016 and 2020, and 2018 and 2020. Do note that I use the Simple United Kingdom Political Swing formula, which is different than certain other ways of calculating swing.

These results present more questions than answers. Polarization has increased measurably every Presidential election since the historic low in the 1970s. Voters become more loyal to their parties’ tickets when the Presidency is on the ballot. Why then did Texas’s Presidential and Congressional results diverge markedly? Why then did majority Hispanic areas along the Rio Grande shift hard towards former President Trump? Why did President Biden surge and even outperform O’Rourke’s margins from the 2018 Senate race in Suburban Texas, despite 2020 being a better year for the GOP than 2018? Why did both areas have unusually high rates of split-ticket voting? In this article I will examine why South Texas Hispanics defied expectations and their traditional loyalties to vote for Trump’s GOP. In the next article I will look at why mostly White suburban areas swung hard towards Biden, but not towards the rest of the Democratic ticket.

Presidential Results

Texas’s majority-Hispanic border region was among the most consistently Democratic areas in the country. The region was the only part of Texas to vote for both Al Gore and John Kerry over the former Texan Governor George W. Bush. The region did not back President Reagan in 1980 or 1984, and the region is home to seven of the 130 counties won by George McGovern in 1972. One has to go back to the early 1900s to find a time when some of these seven counties last voted for a Republican candidate for President. The last time Zapata County, one of the seven, voted for the GOP was 100 years ago, one year after the eponymous revolutionary met his end at the hands of the post-revolutionary Mexican government's assassins. The 1950s and 60s saw the rise of LBJ's South Texas political machine. A network of local bosses and patronage ensured Democrats would keep winning the state despite the emerging strength of the GOP-inclined suburban voter. Not all of the tactics used by this machine were legal.

This is why the results from the counties along the Rio Grande shocked both observers and analysts. Pre-election polls hinted at a national shift among Hispanics in favor of former President Trump but many viewed this as the normal failure of polls to accurately model Hispanic communities. There was no history of Republican favoritism among these voters, unlike the Cubans of South Florida who shifted towards former President Trump by large margins from their 2016 results. The lurch towards Republicans did not conform to Rio Grande media markets or traditional Texas geographic regions. The data matches best to the proportion of the Texas voting population with some Hispanic ancestry. The overall result is that former President Trump performed better and Joe Biden worse than any nominee for their respective Party in the modern era. Trump received 574,433 votes, or 48.2%, to Biden’s 601,930, or 50.5%, in the analyzed area. Comparatively, Hillary Clinton received 543,387 votes in 2016 to Trump’s 396,355, and O’Rourke defeated Cruz by 496,859 votes to Cruz’s 361,863 in the analyzed counties.

[The defined regions in these articles. These regions are defined based on political results, so San Antonio, for example, is not with the rest of South Texas because it behaved like the other cities in the state.]

Congressional Results

Five Congressional Districts represent this majority Hispanic region in the south and southwest of the state. One of these is Texas’s 16th, which thanks to urban El Paso, remains an overwhelmingly safe Democratic district. South Texas and the geographic Rio Grande Valley is covered by three narrow districts, Texas’s 15th, Texas’s 28th, and Texas 34th. These districts are the product of Voting Rights Act litigation that required pairing the near-100% Hispanic-by-Census data counties with more diverse areas to their north. Republican gerrymandering ensured that even under these restrictions, these “fajita strips” (a term applied after the 2003 Redistricting, derived from the districts’ Hispanic majorities, their narrow shape, and how they paired disparate Democratic groups together) would remain Safe Democratic vote sinks. Despite the seismic voter shifts, Biden won all three districts by less-than-5% margins.

Incumbent Democratic Congressmen Vicente Gonzalez (TX-15) and Filemon Vela (TX-34) did not expect to be in serious danger of losing reelection, yet Joe Biden's weak Presidential results narrowed their once-comfortable margins of victory. Henry Cuellar (TX-28) comparatively benefited from split-ticket voting and outperformed Biden’s result by over 15 points. Cuellar’s moderate to conservative, Pro-Life Democratic brand received plenty of media coverage during his 2020 primary against a progressive challenger, which endeared him to the Trump-voting Hispanics in his district.

Texas’s 23rd is the fifth district along the Rio Grande but is different from the others in several key ways. Demographically, TX-23 is similar to a spiraling fajita seat, but geographically it remains in South and Southwest Texas. Politically, the seat is competitive but favors the GOP. Moderate incumbent Republican Will Hurd's retirement was supposed to make TX-23 an easy flip for Democrats, yet voters preferred Republican Tony Gonzales to Democrat Gina Ortiz-Jones. On the surface, this seems like an obvious result: former President Trump got a majority of the vote in TX-23, moving the seat from its 2016 left-of-the-nation result to one noticeably to the right of the popular vote. However, TX-23 remained a close contest because the district includes parts of Texas that lurched in favor of the former President and parts that moved in favor of Biden.

Swing between 2018 Senate and 2020 Presidential Elections in TX-23's portion of Bexar. Biden gained in the whiter north, and Trump did better in the more Hispanic south.

Usually, about half of TX-23’s vote comes from its piece of Bexar county and historically it favors Republicans compared to the districts Democratic-leaning rural Counties. This year though, the suburban piece of Bexar County voted for President Biden. Even though the district's western counties combined vote favored former President Trump, suburban Bexar kept the district close. Biden's margin of victory was slightly wider than O’Rourke's within the examined precincts because of Democratic improvement among the vote-rich suburbs. Inside suburban Bexar there was significant split-ticket voting in favor of the GOP, similar to Texas’s other suburbs. Gonzales won a plurality in Bexar, and this secured his victory. The reversal of traditional loyalties in the region led to the peculiar result of Texas legislative Democrats flipping the State Senate district congruent with TX-23 while losing the congressional seat. This is because a greater percentage of the State Senate district’s voters are in Bexar.

So, Why?

The Demographical Diversity Theory

The most commonly cited reason for the shift against Democrats along the Rio Grande is that Hispanic voters are not a monolithic group, and as such, react differently to the same political message. This theory was one of the first that emerged after the election, and it therefore got the most media coverage. This rationale further explains that an individual’s ethnic identity is not easily categorizable, even by census questions, and has more to do with that individual’s experiences. Democrats assumed that all Hispanic voters would be repulsed by former President Trump’s bigotry, but this message failed to resonate equally with every voter. Not all Hispanic or Latino voters in the Rio Grande immigrated to the United States at the same time over the past century, or from the same country. A few families are even descended from those who watched the border cross them in 1848. it was these groups differences in background, industry, and history within the United States that led to some groups decision to break with the Democratic party.

Swing between the 2016 and 2020 Presidential Elections in the Houston area. Note how the minority districts, TX-09, TX-18, and TX-29, are the ones that swung towards Trump.

Yet an examination of predominantly nearby Hispanic areas in Dallas and Houston - an area with different Hispanic groups in different circumstances than along the Rio Grande river - reveals that South Texas was not unique. President Biden underperformed previous Democrats within these cities Hispanic neighborhoods. The New York Times's recent compilation of national precinct data makes it clear that Hispanic communities across the country were less favorable to Biden than they were to Hillary Clinton. The size of the swing toward Trump differs depending on the state and community in question, but it is always there. Even though Florida Cubans and rural Rio Grande Hispanics were the most eager and most visible Republican-swinging Hispanic groups, their political behavior was not unique among the American Hispanic population.

The Economic Theories

Two broad other theories that seek to explain Hispanic voters behavior are economic. The first theory is holds that Trump’s four years were economically beneficial for disadvantaged Hispanics, and this is the first sign of integration into conservative-leaning White America. This theory looks to the past, and sees elections where Catholic “ethnic” Whites (a derogatory term used historically to differentiate them from other Whites) were once loyal Democratic voters. These groups are now viewed as simply Whites, due to decades of financial and cultural integration, and their voting behavior is similar to other White voters. However, predominantly Hispanic areas remain among the poorest in America, this did not change during former President Trump's term in office. The well-off suburban areas with a heterogeneous Hispanic and other minority populations swung to President Biden, similar to other suburbs across the nation. It was the heavily Latino neighborhoods in cities and the rural Southwest, the isolated communities most vulnerable to economic fluctuation, which swung in favor of the GOP across the county.

The second economic theory postulates that those groups most exposed to pandemic-related job insecurity were uniquely disposed to voting to reelect former President Trump. Disadvantaged minorities, including urban and rural Hispanic populations, struggle to make ends meet during the best of times. Coronavirus only compounded traditional insecurity and financial hardship. Some lost jobs during the spring lockdowns and were subsequently rehired when businesses reopened. Another lockdown would mean financial ruin, so these voters turned towards the party fighting to keep the economy open. There however are two things that limit the theories efficacy. First, if the coronavirus's economic effects were the sole reason for voter behavior, then the swing among minority populations would differ between states. Some states embraced reopening, others remained in lockdown or returned to it after a period of time. Hispanic voter movement however does not correlate to the variations in state policy. Hispanic swings in Los Angeles and New York City were comparable to Houston or the Rio Grande Valley. Secondly, not all were rehired after reopening. Unemployment on election day was higher than before the pandemic. Studies previous found that a worsening personal financial situation correlates to movement away from the governing party. There however is no sign of a backlash to Trump among the unemployed Hispanic voter pool that would hypothetically blunt the movement towards the GOP among employed yet insecure Hispanics.

The Straight-Ticket Theory

Another theory is that the 2020 swings along the Rio Grande were the product of Texas Republicans breaking local political networks that held voter loyalties in stasis. South Texas’s overwhelming Democratic loyalty is partially a legacy of the patron-client style politics of the 20th century, such as the LBJ machine. (this is a style of politics where politicians invest in communities only if they vote as desired, ensuring name-recognition and voter loyalty) When patron-client politics collapsed across the rural south, similar legal organizations survived in South Texas. All local South Texan politicians are Democrats, and advancement means working within the party machine.

This year Texas Republicans removed the straight-ticket checkbox from the ballot. (Straight-ticket voting is an option in some states where one checks a box for a political party, and this vote applies to all candidates under the party’s label) Some believe that by removing the Party checkbox from the ballot, voters were forced to think whether each individual politician deserved their vote, and this process flipped voters from the Democrats. South Texas previously had above average straight-ticket voting rates compared to other parts of Texas, but it did not deviate widely from the averages by county. Removing straight-ticket voting may have been a factor in 2020, but it wouldn’t "free up" enough voters on its own to explain the swings from previous Democratic results. Patron-client networks are a localized phenomenon, and the swing in urban Latino neighborhoods nationwide was comparable to voter movement along the Rio Grande, casting doubt on whether this theory explains why South Texas lurched towards the GOP.

The 2020 Theories

There are several presently unprovable hypotheses (the data is not yet available) that are particular to this election. Hispanic voters are historically harder to turn out, and Democrats opted against in-person canvassing because of the pandemic. Republicans continued their local networking without opposition. Lower-income and minority groups are unfortunately more likely to come into contact with coronavirus because of economic circumstances, so the canvasing discrepancy was more visible in minority communities. This theory was brought forward by Texas Democrats immediately after the election in response to their failure to match expectations.

Another unproven theory is the potential of a resource allocation discrepancy. 2020 was a high turnout election, and both parties worked hard to turn out low-propensity voters. Democrat-favoring early voting numbers in the Rio Grande lagged the rest of Texas, suggesting Republicans turned out more of their voters in the region than Democrats. Both parties likely turned out all their reliable Hispanic voters. If the GOP ensured that more infrequent conservative voters turned out than infrequent Democrats, then it would appear that Hispanics swung towards the GOP, even though the overall population’s views may still resemble those of previous elections. Precinct level data from the Georgia runoffs in January lend some credibility to this potential explanation. Turnout decreased by more in Georgia precincts with a significant Hispanic population than in neighboring precincts with differing demographics. These same Hispanic precincts swung towards the Democrats when compared with President Biden's November results. This data suggests that the infrequent voters removed themselves from the pool and returned to their normal ambivalent behavior, leaving only the frequent and more Democratic electorate.

Finally, there is the potential that late-election wedge issues like Amy Coney Barrett’s Catholicism and President Biden's Democratic commitment to reduce fossil fuel consumption isolated Hispanics, especially in South Texas. South and Southwest Texas Hispanics are likely to be practicing Catholics, and have a higher proportion who work in or around the Texan oil industry. These type of one-off issues are always important, but they fail to produce a lasting change in party loyalty.

The Hidden Voter Theory

The final explanation scares Democrats and excites Republicans the most, as this is potential evidence that these Republican voters were always there and just needed outreach and a motivating issue. South Texan voter turnout rates are low, and low voter registration rates make the pool of voters disproportionately small when compared to the overall population. Webb County, home of Laredo, had a 49.6% turnout in 2020, one of the highest turnout elections in recent memory (68,397 votes were cast out of a total projected population of over 276,000). Webb County gave President Biden 41,820 votes, only 487 less than the county gave to Hillary Clinton in 2016. The county swung away from his Democratic ticket because the GOP doubled their 2016 vote total. This story is similar across the region; the Biden vote total only varies slightly from the 2016 Democratic vote totals, and all the new voters on the surface appear to have gone for the GOP.

2020 was the first time in 20 years that Texas Republicans expected a fight to hold on to statewide office, because of O’Rourke’s performance in 2018, rather than their usual cakewalk. This was the first time the GOP had to assemble a coordinated statewide organization to register every potential conservative voter, convince every potential swing voter, and turn out as many Republicans as possible. These voters may have been in South Texas for years, but just like the suburbanite Democrats who moved to the city’s further north, they didn’t bother to participate in politics because the GOP previously had total control over government and could afford to ignore South Texas outreach.

Overall

There is no single factor that explains what happened among Texas Hispanics. Every theory, despite its convincing explanations, has flaws when examined individually. Yet, when combined the contributing factors in 2020 paint a grim picture for Texas Democrats. Change does not happen overnight, and ironically, Bernie Sanders’s 2020 primary victories among Southwestern Hispanics suggest that these voter movements had been occurring for several years. If the GOP convinced conservative and moderate Hispanics to change allegiances, the pool of remaining Democratic Hispanic voters would be more liberal than the party overall. It benefited Republicans that the economic fallout from the pandemic, Black Lives Matter, the Supreme Court confirmations, and other issues overshadowed Immigration this year (immigration wasn’t even an option on exit polls when they asked what issue was most important to voters this cycle). They may not have this luxury in the future."

So he doesn't really claim to solve the mystery. But there's a lot to think about here!