Enter the Libertarian, again

Four years ago, the Libertarian Party in the United States was hopeful that, with the two major parties on the way to nominating seriously unpopular candidates, it could be their year for a big showing. At their convention, held at the end of May, they nominated (for the second time) a credible candidate, former New Mexico governor Gary Johnson.

And by Libertarian standards, Johnson did do well in the presidential election: he took 3.3% of the vote, easily the best of the minor party candidates and three times the previous best for a libertarian. But it was still a long way short of the ambitions that had been entertained mid-year.

It was also a long way short of breaking into the electoral college. Even in Johnson’s best state, New Mexico, he had only 9.3% of the vote, more than 30 points behind Trump and almost 40 behind Clinton, who carried the state with 48.3%.

Last week when we looked at the US pendulum, I explained that when calculating margins I factored out all the votes for minor candidates. Since there are no preferences in the American system, votes for the also-rans are conceptually the same as those who vote informal or just stay home.

Unless, of course, a minor candidate is in a position to carry a state and therefore win seats in the electoral college. But no-one has done so since white supremacist candidate George Wallace, of the American Independence Party, who carried five states in the deep south back in 1968.

In the twelve elections since, only one candidate has ever come close to beating the major parties in a state: independent Ross Perot in 1992, who won 18.9% of the vote overall and more than a quarter of the vote in eight states. His highest vote was 30.4% in Maine, but even there he was more than eight points behind the winner, Bill Clinton.

Apart from Perot, the best third-party result in an individual state in that time was in 2016. But it wasn’t Johnson; it was conservative independent Evan McMullin, who won only 0.5% of the total vote but managed 21.3% in Utah (and 6.7% in neighboring Idaho). But that still wasn’t enough to put Utah in play – Donald Trump won it with 45.5%.

So we can probably put aside the idea of the Libertarians winning seats in the electoral college.* Realistically, their hope is to win enough votes to be seen to be able to make the difference between the major candidates, thus raising their profile and giving them some potential bargaining power.

Last time around, there were eleven states – the five most marginal on each side of the pendulum, plus New Mexico – where Johnson’s vote was larger than the gap between Trump and Hillary Clinton. If his voters had all supported Clinton instead, she would have won the election. If they had all voted for Trump, he would have extended his margin and won the popular vote as well.

Neither of those things, however, was at all likely. Libertarian voters come from across the spectrum; while the majority appear to be former Republicans (seven of Johnson’s ten best results were in Republican states), it doesn’t follow that, in the absence of a Libertarian, they would therefore vote Republican.

Consider New Mexico again. In the House of Representatives election held on the same day, with no Libertarians in the field, Democrat candidates in aggregate improved by 7.7% on Clinton’s total, winning 56.0%. Republicans, with 44.0%, were running only 4.0% ahead of Trump. That doesn’t mean the majority of Johnson’s vote necessarily went Democrat – there might be other reasons why people vote differently for the House and presidency – but it certainly doesn’t look as if it went strongly to the Republicans.

It’s reminiscent of the discussion last year on the death of Perot about his effect on the outcome of the 1992 election. The short answer is that no-one knows for sure what Perot’s voters would have done without him (or where they would have given their preferences, if they could), but that they were sufficiently diverse that it seems most unlikely they would have concentrated enough in one place to make a big difference.

This year the Libertarians again have a relatively high profile candidate in the pipeline, and again he’s a former Republican. Justin Amash, member of Congress for the third district of Michigan, has formed an exploratory committee to seek the nomination and is regarded as the frontrunner to get the nod when the party’s (virtual) convention meets in a fortnight’s time.

Amash left the Republican Party last year over his opposition to Trump, and it’s expected he will have some appeal to voters who dislike the president but are otherwise conservative. Democrats, understandably enough, are worried that the people he will attract would otherwise be willing to take the plunge and vote for Joe Biden.

It’s particularly an issue because Michigan, Amash’s home turf, is a crucial swing state. Trump won it last time by just 10,700 votes. Johnson collected ten times that many, and Amash may do better still. As Geoffrey Skelley and Julia Azari at FiveThirtyEight explain:

Amash, unlike Johnson, hasn’t held a statewide office, but he probably would still have an advantage in the western part of Michigan, which he’s represented in the House for nearly 10 years. That could get in the way of Democrats’ hopes of making gains in traditionally Republican Kent County — Amash’s base and home to Grand Rapids — because of its sizable share of college-educated voters. While Biden has led most early Michigan polls, Amash could complicate the race there.

But let’s not get ahead of ourselves too far. If Amash does actually get the nomination, we’ll have another look at where his appeal might lie – and what that says about the dynamics of party politics in the US.

* I ignore the possibility of electors who win seats on a major party ticket then voting for someone else in the college, as occasionally happens (so-called “faithless electors”). Libertarian John Hospers received a vote this way in 1972.