A theory about third parties

Before I get to that theory, have you ever wondered about the people who vote straight ticket Libertarian or Green in Harris County? I got to wondering about them, because that’s the sort of thing that I think about at times like this. Here are the total numbers of such people, grouped by Presidential and non-Presidential years, going back to 2000:

Year  Total votes  SP Lib  SP Green   Lib%  Green%
==================================================
2000      995,631   1,935     4,503  0.19%   0.45%
2004    1,088,793   3,343            0.31%
2008    1,188,731   4,017            0.34%
2012    1,204,167   4,777     1,759  0.40%   0.15%
2016    1,336,985   8,781     4,577  0.66%   0.34%

2002      656,682   1,159     1,399  0.18%   0.21%
2006      601,186   3,052            0.51%
2010      798,995   2,506     1,110  0.31%   0.14%
2014      688,018   2,922     1,180  0.42%   0.17%

“SP Lib” is the total number of straight party Libertarian votes, and “SP Green” is the same for the Greens. “Lib%” and “Green%” are the share of these straight party votes to all votes cast in the county. If you look at the election result pages on the HarrisVotes.com website, you will see that my percentages are lower than the ones shown there. That’s because they calculate the percentage of these votes as a share of all straight-party votes cast, not a share of all votes. I did it this way to see what if any trend there was for Libertarian and Green voting overall. For comparison purposes, 30.01% of all votes in Harris county this year were straight ticket Republican, with 35.35% of all votes being straight-ticket Democratic.

As you can see, in the Presidential years the Libertarians had been slowly ticking upwards, with a bit of a jump this year, though the trend is more erratic in the off years. The spike in 2006 is odd, because the Libertarian candidate for Governor received only 0.61% of the vote that year. If you wanted to vote outside the two-party box for Governor in 2006, you had plenty of choices. The Greens weren’t officially on the ballot in 2004, 2006, or 2008, so there’s less of a trend to spot. I’d say they do better in or right after a year where they have a Presidential candidate who gets some attention. Whether any of this will hold next year is not something I’m going to speculate about at this time. My mantra for the next twelve to eighteen months is “conditions in 2018 will be different than they were in 2014 and 2010”, and leave it at that.

That brings me to my theory, which applies to low profile races – not President, not Senate, not Governor, sometimes not other races. I’m limiting myself to statewide contests here, since that’s where you get most of the third party candidates that an individual voter sees. In my case, there was a Green candidate for CD18, a Libertarian for SBOE, and nothing else below the state level. I believe that in these races, which this year would be the Railroad Commission and the two state courts, voters for third party candidates can be broadly sorted into one of three groups. The first group is the party faithful, which as we have just seen is a relatively small cohort. There are probably a few more people who vote L or G as a first choice but don’t vote straight ticket, but that’s still a small group even in the context of just third party voters. Most of the people voting third party in these races aren’t voting third party as a matter of course.

So who are they? Group Two I believe is people who normally vote for Rs or Ds but who refuse to vote for their candidate in this particular instance. That may be because the candidate of their party is too/not sufficiently liberal/conservative for them, because that candidate supports or opposes a specific thing that is of great importance to them, because the candidate has ethical baggage, or because they just don’t like that candidate for some reason. In these cases, they don’t want to vote for the candidate of the other party, so a third party it is. Gary Johnson obviously got a lot of these votes in the Presidential race, but the downballot exemplar for this one was the Railroad Commissioner race, where Libertarian Mark Miller got a bunch of newspaper endorsements for being the most qualified candidate running.

The thing is, I don’t think there are that many races like that. I think in a lot of these races, people just don’t know anything about any of the candidates. So if you’re someone who (say) generally votes Democratic but aren’t that committed to it and you’re looking at a race for the Court of Criminal Appeals, you may say to yourself “well, I know I don’t want to vote for the Republican, but I don’t know who any of these other people are, so I’ll just pick one and move on”. These people are my Group Three.

What that says to me first of all is that both Republicans and Democrats are leaving some votes on the table in these downballot races by not doing a better job of getting their candidates’ names out there. That’s not much of a concern for the Republicans, who continue to win by double-digit margins, but it could eventually matter. I see this as an extension of a problem that Democrats are increasingly having in their primaries, where candidates like RRC nominee Grady Yarbrough have won races by a combination of pseudo-name recognition and random chance because no one knows who the hell these people are. I have many wishes for Texas Democrats going forward, and high on my list is for the party and the donor class to take these downballot primaries seriously.

One possible exception to this may be for Latino candidates. Look at the top votegetters for each party: Supreme Court candidates Eva Guzman and Dori Contreras Garza. My hypothesis is that Latino voters in a Group Three situation will choose a Latino candidate, even possibly one from their non-preferred party, instead of just randomly picking someone. Again, this is in races where none of the candidates are known to the voters, and thus there could be a different outcome if people had more knowledge. If we ever get to that point, maybe we’ll see that difference.

Finally, I believe my theory is consistent with the Libertarian candidate almost always doing better than the Green candidate does in these situations, for the simple reason that the Libertarian candidate appears on the ballot above the Green candidate. If it’s true that some people just pick a name after having moved past the first two candidates, then it makes sense that the first candidate listed after those two would get a larger share.

Anyway, that’s my theory. I could be wrong, and I doubt anyone other than me had given this much thought. I’ll get back to the precinct analyses tomorrow. Let me know what you think about this.

Related Posts:

• Endorsement watch: A Libertarian moment
• Initial thoughts about the election
• State and county election result relationships, part 1: How Harris compares