A majority of Harris County residents lean Democratic for the first time ever, propelled by plummeting support for Republicans among Latinos, according to a survey released Monday by Rice University’s Kinder Institute for Urban Research.
The finding, in the midst of a particularly divisive presidential campaign, could signal an important shift in arguably the nation’s largest swing county, which narrowly went to President Barack Obama in 2012 by only about 970 votes. It might also portend that the long-sleeping giant of Latino voters will, finally perhaps, be roused from slumber in an election that has featured decidedly anti-Latino and anti-immigrant rhetoric, particularly from billionaire Republican contender Donald Trump.
“Frankly I’m not all that surprised,” said Jim McGrath, a Republican political consultant in Houston and spokesman for former President George H. W. Bush. “These are the fears realized by those on the Republican side who are worried about the irresponsible rhetoric surrounding the illegal immigration issue.”
According to the annual survey, which was conducted between January and March, 52 percent of Harris County residents said they identified more with the Democratic Party compared to 46 percent in 2012. Only 30 percent of residents leaned Republican this spring, about the same as in 2012, meaning that it is the share of undecided and new potential voters whom have swung largely Democratic.
Support for the GOP has stayed steady among white and African-American residents for the past decade, with 54 percent of the county’s white population swinging Republican and 39 percent Democrat, though there was a slight increase in Democrat support among Anglo voters in the county over the past two years. Similarly 82 percent of African-American residents lean Democratic and 8 percent Republican.
Among Latinos, however, there has been a sea change.
From about 2000 to 2008, some 40 percent of the county’s Hispanic residents identified as Democratic compared to fewer than 30 percent who felt Republican, Klineberg said. That began to change around 2009 when their support for Democrats increased to nearly 50 percent and the share of those leaning Republican dropped to 25 percent.
The gap widened once more around the 2012 presidential election when Republican Mitt Romney received the lowest share of the Hispanic vote — 27 percent — than GOP nominees had tallied in the previous three election cycles in a campaign during which immigration was particularly divisive.
This spring, Harris County’s Hispanic residents registered the lowest amount of support ever for Republicans — only 18 percent — compared to 68 percent of Latinos who said they lean Democrat.
“It’s a powerful message to the Republican party, reach out to these Latino voters, don’t push them away,” Klineberg said. “And for the Democrats, get out the vote.”
The survey is conducted by land line and cell phone calls among a statistically representative sample of 808 residents, not eligible voters, in Harris County. Among 604 Harris County residents who can vote, 46 percent leaned Democrat and 41 percent Republican.
See the Urban Edge blog for more details on the poll. There’s quite a bit more to the 2016 Houston Area Survey than this, but for now we’ll just focus on this particular data point, for obvious reasons. This is not a poll in the standard sense – it doesn’t ask which candidate you will support, nor does it try to determine who is a “likely” voter – but it is consistent with what we are seeing in national data as well as swing states. Latinos were slightly more likely to vote Republican in Texas in 2012 than they were elsewhere, though that was partly a turnout function, as polling data at the time showed that lower-propensity voters were more strongly Democratic. If – the big if – Latino voters are more strongly motivated to turn out this year, it is consistent for them to be more Democratic even without taking the Trump factor into account.
What could this mean in practical terms?
Some advocacy groups, such as the William C. Velásquez Institute, a national Latino public policy research group in San Antonio, predict Hispanics in Texas this year will account for more than 3 million registered voters and cast more than 2 million votes, both of which would be records. Overall, the state has about 14.2 million registered voters.
Their expectations are largely predicated on population growth. Since 2012, Texas gained 600,000 eligible Hispanic voters, expanding to 4.8 million – second only to California, according to the Pew Research Center, a think tank in Washington, D.C. The Latino share of Texas’ eligible voters increased 2 percentage points in that period, to 28 percent.
Bearing in mind all of the usual disclaimers, let’s do a little back-of-the-envelope math for the fun of it. Here are three statewide scenarios for this year:
Total votes Latino Not Latino Pct
4,650,000 480,000 4,170,000 58.75%
3,350,000 1,120,000 2,230,000 41.25%
4,570,000 400,000 4,170,000 54.40%
3,830,000 1,600,000 2,230,000 45.60%
4,670,000 500,000 4,170,000 53.00%
4,230,000 2,000,000 2,230,000 47.00%
Scenario 1 is basically what happened in 2012. No change in Latino turnout, which based on 2012 polling is 20% of the total, or Latino propensity for voting Democratic, which was about 70% that year. Scenario 2 is the “two million Latino voters” possibility that the Velasquez Institute mentioned. For that, I’m assuming 80% Democratic support, which is consistent with the polling data we have so far for matchups against Donald Trump, and with the data noted above that lower-propensity Latino voters are more heavily Democratic than Latinos overall. Sure, this may be a bit optimistic, but I’m playing a what-if game here, so stay with me. Scenario 3 is the bluer sky version of #2, where Latino turnout is 2.5 million at the same 80% Democratic rate. Note that in all cases, non-Latino turnout and propensity is the same. This is mostly to make the calculations simple; basically, I’m isolating the Latino voting variable. One could play around with the hypothesis that a Trump candidacy might also depress base Republican turnout, but I’ll leave those calculations to you. In scenario 2, Latinos make up about 24% of the voter universe, while in #3 they are 28% of total turnout, which as noted is about their share of total eligible voters.
I’m not arguing any of this is likely, or even realistic. I am showing that the ground is shifting, and even a relatively modest change could have a sizable effect. It’s not enough to turn Texas blue, but the state would be a lot less red. As noted before, that effect would surely be felt downballot, with Harris County likely being an epicenter. The bigger question would then be if any of that might carry over into a non-Presidential year, or if the same patterns we have observed in recent elections would persist. That’s beyond my scope here, and depending on how things end up may be irrelevant. But clearly something is happening. Even if it’s not enough to change the state, it’s more than enough to tilt Harris County, whether there is a concerted turnout effort (which I hope there is!) or not. Campos has more.