Early voting, Day Three: The case for pessimism

Dave Mann tells Texas Democrats to put those rose-tinted glasses away.

On Monday, the Real Clear Politics site declared that Texas is up for grabs in the presidential election. The shift comes after a series of polls showing a tight race in the state between Hillary Clinton and Donald Trump, and makes for a dramatic image on the site’s election tracking map, where Texas is no longer colored its usual red but is now the dark gray that connotes a “toss up.” For Democrats, seeing their state change color on one of the most widely read and respected campaign outlets—after decades of Republican dominance and years of unfulfilled hopes that Texas might turn blue—must be cathartic. And it might be tempting to view this sudden shift to competitiveness as the start of Democrats’ long-hoped-for return to relevance, as a turning point.

Well, they should keep the cork in the champagne, because Texas remains a Republican state.

As my colleague R.G. Ratcliffe pointed out, the Texas polls are close not because of a huge spike in Democratic voters—Clinton’s numbers are roughly in line with Obama’s totals from 2008 and 2012—but because Trump’s support has cratered. He’s drastically under-performing previous Republican presidential nominees. John McCain and Mitt Romney garnered 55 percent and 57 percent of the vote in Texas, respectively. Trump is polling 10 to 12 points below that.

[…]

While it’s true that the national GOP looks like a smoking ruin right now, the state party remains fairly strong. It still has huge advantages over Texas Democrats in money, organization, and candidate depth, and Republicans start every statewide race with at least a ten-point edge, if not more. And if you’re thinking that built-in advantage may be shrinking, keep in mind that we’re just two years removed from an across-the-board Republican blowout of nearly 20 points. In Wendy Davis, the Democrats had their best known and best funded candidate in years, and she lost to Greg Abbott by nearly a million votes.

It’s also worth remembering that most statewide offices in Texas come up for election in non-presidential years in which the electorate generally tends to be whiter and older—in other words, more Republican.

The one caveat is the potential increase in Latino voters. R.G. wrote on Monday that more than 530,000 people with Latino surnames have registered to vote since 2012, according to the Texas secretary of state’s office. It’s not hard to envision Trump’s candidacy increasing the number of Latino voters who turn out to vote in Texas, offering Democrats the opportunity to begin building a coalition that could one day make them competitive again. But capitalizing on that opportunity requires the difficult party-building, community-organizing, voter-turnout work that Democrats in this state haven’t exactly excelled at in recent years.

In other words, two years from now—without Trump at the top of the ticket—Texas Republicans will once again be heavily favored to sweep the statewide offices.

See here for my discussion of RG Ratcliffe’s article. First, let me say that I agree with Dave Mann in that it’s at least premature, if not downright silly, to call Texas a swing state right now. It’s a lot closer than we’re used to seeing it, but the numbers aren’t there for swing state status. The Real Clear Politics average for the two-way race has Trump leading by 4.6 percentage points. FiveThirtyEight has Trump’s lead at 6.2 after applying their secret sauce. Out of thirteen poll results that I’ve tracked, only that one wacky WaPo/Survey Monkey one from September had Clinton in the lead, by one point. I think to be a real swing state, your polling average has to be within, say, two or three points, with more than one result disagreeing with the others about who’s in the lead. Texas doesn’t make the cut on either of those.

That said, I think Mann is underplaying how well Clinton is doing, both in absolute terms and relative to Obama. The more recent polls have shown her increase her total more than Trump has done. I split the thirteen poll results I’ve tracked into pre-October and October results and averaged each. That works out as follows:

Pre-October: Trump 42.0, Clinton 35.7
October: Trump 46.2, Clinton 41.5

Clinton has gained 5.8 points in the average to Trump’s 4.2, cutting the margin in the average from 6.3 to 4.7. Moreover, she’s considerably ahead of where Obama was in the October polls from 2012:

October 2012: Romney 55.8, Obama 39.0

You can also use the YouGov tracker for a direct comparison. The election eve result in 2012 had Obama at 38%. As of yesterday, Clinton was at 41.4; she was up at 42.0 over the weekend. And remember, that 2012 YouGov result underestimated Obama by three and a half points. It’s possible they’ve changed their model to account for that, but it’s also possible they’re underestimating Clinton.

I don’t want to get too deep into that, because as the Devil can use scripture for his own purposes, one can read whatever they want into an individual poll. The thing is, though, we also have actual votes that have been cast, which really do tell us something. I can tell you that Democrats have done much better so far in Harris County than they did in 2012, and have won each of the first two days of early voting, after winning with mail ballots. Some of this is surely regular voters getting out there earlier than usual, and I don’t have the same data on the rest of the state, but just as surely Harris County isn’t an anomaly.

What I’m getting at is this: I think one has to strain to argue that Hillary Clinton won’t exceed Barack Obama’s vote total from 2008. I think she’s got a very good chance to exceed his vote percentage, though I’m not ready to declare that as a sure thing. We may argue afterwards if the increased vote total I expect Clinton will get represents a real bump in Democratic turnout, as 2008 for Obama did compared to 2004, or just a raise that was proportional to the overall population growth. But I don’t think we’ll be arguing over whether or not she did outperform him, in 2008 as well as in 2012.

As for 2018, I’m going to wait till this one is in the books before I get into that. It’s true that Donald Trump won’t be on the ballot, but that doesn’t mean he can’t be used as a motivating tool. It’s also true that while 2014 was a disastrous year for Texas Democrats, it wasn’t just a Texas problem. National conditions had a big effect on state elections in 2014, and in 2010 and 2006 and 2002 and so on, for that matter. What will national conditions be like in 2018? You’re a lot smarter than I am if you know the answer to that today.

Anyway. Early voting turnout was even higher on Day Two than it was on Day One. That’s actually in line with the historical pattern, as you can see from the handy early voting tracker spreadsheet that I’ve so thoughtfully included for you. Day Two was busier than Day One in all three previous Presidential years. Day Three was busier than Day Two in 2012 and 2008, too. And guess what? As you can see from the Day 3 EV report, Day Three was busier this year than Day Two was, too. It’s like there’s an established pattern or something, it’s just a matter of at what level. Another 76,098 in person votes, with 5,646 mail ballots arriving, and 287,134 total votes cast so far. The Day Three amount in 2012 was 197,987. We’re going to run out of voters eventually, but we could get an awful lot of votes cast before that happens.

Related Posts:

• Initial thoughts about the election
• Early voting, Day Six: A good first week for Democrats
• Early voting, Day Four: Just the facts