What are your turnout scenarios?

I keep thinking about this:

County Clerk Stan Stanart predicts up to a million Harris County residents could be casting ballots in a string of hotly-contested races.

As you’ve heard me say many times, the Democrats’ main issue in off year elections in Texas has been that the base vote has not really increased at all since 2002. With the exception of the occasional Bill White or John Sharp, it generally tops out at about 1.8 million, which is what Wendy Davis collected in 2014. This year, there are multiple factors that strongly suggest Dems will blow past that number. The national environment, the plethora of candidates, as well as their terrific success at fundraising, the tremendous level of engagement, and on and on. But right up in there is the increase in voter registration, at the state level as well as here in Harris County. What do the numbers from the past suggest to us about the numbers for this year?

Let’s start with some basics:

Year      Harris      State   Ratio
===================================
2002     656,682  4,553,979  14.42%
2006     601,186  4,399,068  13.67%
2010     798,995  4,979,870  16.04%
2014     688,018  4,727,208  14.55%

Year      Harris   Register      TO
===================================
2002     656,682  1,875,777  35.01%
2006     601,186  1,902,822  31.59%
2010     798,995  1,917,534  41.67%
2014     688,018  2,044,361  33.65%

The first numbers are the turnout figures in Harris County and statewide in each of the last four off year elections. I wanted to see how big the share of the Harris County vote was. YThe second numbers are more familiar, turnout and registered voter totals for Harris County. Let’s use these to get a sense of the range of outcomes for this year. We know that we have about 2,316,000 registered voters in Harris County, based on the news reports we’ve seen. (The exact figure has not been released.)

2,316,000 at 31.59% = 731,624
2,316,000 at 33.65% = 779,334
2,316,000 at 35.01% = 810,831
2,316,000 at 41.67% = 965,077

You can see where Stanart came up with that “up to a million” figure. It’s hardly implausible, based on past performance. Even the fairly modest 35% turnout projection would give us a new record for an off year. Now what might this translate to at the state level?

731,624 at 16.04% = 4,566,941
731,624 at 13.37% = 5,352,040
965,077 at 16.04% = 6,016,689
965,777 at 13.67% = 7,034,967

Six million may well be the over/under total. The Upshot is predicting a range of 6.3 million to 7.2 million, based on the polling data they’ve seen.

Which leads to the next question. If six million is accurate, and Beto O’Rourke is headed to a 45% performance, that’s about 2.7 million votes. Remember when I said that Wendy Davis got 1.8 million in 2014? That’s a 50% improvement over her. Even if you buy into the idea that Lupe Valdez is heading for a 20-point loss, she’d still collect 2.4 million votes out of 6 million. The flip side of this is that Ted Cruz would collect 3.3 million votes, and Greg Abbott would get 3.6 million. That’s a ten percent improvement over the 2010 baseline for Cruz and 20% for Abbott, and it’s about an 18% improvement over 2014 for Cruz and 36% over 2014 for Abbott.

Frankly, all of those numbers seem outrageous to me. Not unrealistic, certainly not impossible, just amazing. A more modest scenario might be the 810K in Harris County, and Harris being about 14.5% of the state total. That gives an estimate of 5.6 million overall, with Beto’s being a bit more than 2.5 million and Lupe Valdez’s 40% translating to 2.24 million. Still a big boost over 2014, no matter how you slice it. You have to contort things to an unrealistic place to not reach historic numbers.

Personally, I do believe Democratic base turnout will be up, quite possibly a lot, over 2010 and 2014. It almost has to be for Beto to be within ten points. Given that Beto is clearly outpolling Lupe Valdez, his vote total will be even higher. You could assume that he’ll still be in the Bill White zone of 2.1 million or so votes, with Valdez doing a Wendy Davis-like 1.8 million. That would imply about 2.5 to 2.6 million votes for Cruz and 2.8 to 2.9 million votes for Abbott. Do you believe that overall turnout will be static from 2014? This scenario leads to a turnout rate of 29.5%, roughly 4.67 million voters out of 15.8 million registered. That seems far more unrealistic to me than the various vote-increasing totals.

I don’t have any conclusions to draw. I’m putting this out here because this is what the numbers we have are saying. What I want to know is, what are the experts saying? What turnout situation do the pollsters expect? The political scientists? The campaigns themselves? I’ll be happy to see a range of possibilities from them as well. It’s easy to say, oh, Quinnipiac has Beto down by 9, it’s all over, but what do you think that means the final score will be? How did you arrive at that? These are the things I think about when I see new polls.

Related Posts:

• Some opening thoughts on the 2022 election
• UT/Trib: Cruz 41, O’Rourke 36, part 2
• The Beto-Abbott voters