The will of the voters in first-past-the-post systems

Much has been made of how Donald Trump received fewer votes than Hillary Clinton in the 2016 presidential election, with some calling into question the legitimacy of his government because he didn’t receive the majority of support of voters. Specifically, he got 45.93%.

That’s an interesting argument, and especially so where I live, as the federal government in Canada has a majority of seats but received less, at 39.47% of the vote. So, according to this argument, Trudeau’s government has less legitimacy than Trump’s.

I think these sorts of post hoc arguments aren’t very persuasive. The Trudeau government campaigned in a system with specific rules which allow for parties with less than a majority vote (and sometimes not even the most votes) to form a majority government. Similarly, Trump’s strategy was based on the specific rules of the U.S.’s electoral college system. If there’s a problem, it’s at a higher level, in terms of electoral reform.

Having said that, one thing that’s interesting about the Trump-Clinton 2016 election is that Clinton didn’t get a majority, either. Rather, she received 48.02% of the vote. A majority of voters didn’t vote for her, either.

So, which candidate would the majority of voters have preferred, if they could transfer their vote with ranked choice? If you add up Trump, Johnson (Libertarians tend to be closer to Republicans than Democrats), and McMullin (closer to the Republican party than Democratic), you get 49.73%. If you add up Clinton and Stein (Greens tend to be closer to Democrats than Republicans), you get 49.08%. The remainder is various write-in ballots.

So, the right-wing bloc was larger than the left-wing bloc. If the U.S. had direct popular vote with ranked choice and instant run-off, which candidate would have won?

The answer is: we don’t know. Some of the people who voted for one of the smaller percentage candidates might not have ranked another candidate. For example, many people who voted Green might have done so because they excluded all the other options. Similarly, although McMullin was a protest candidate against Trump, it’s not clear how many people who voted for him would have supported Trump next. Similarly with Libertarians.

8 thoughts on “The will of the voters in first-past-the-post systems

  1. This is a far from scientific response. I also don’t think you can compare a parliamentary government with a non-parliamentary government. Although, I can see how it might come off as strange for someone used to governments with less than a majority support. In the US, we tend to expect the candidate that leads in the popular vote to get the presidency. The legitimacy has always been questioned otherwise (see JQ Adams, Hayes, Bush II, Trump, and others).

  2. I think you can Anthony,And the more diffrent ways of voting the better (I hope one day for campaigns we have an option for all ways of voteing)
    And I’m gonna bug you about this again,Primaries popular vote soon please!

  3. I honestly find this all humorous as the democrats repeatedly mocked Romney’s path to 270 electoral votes in 2012 even if he won the popular vote because of the blue wall. The reality is the same system that elected the previous 44 presidents elected Trump and while it is not perfect system everyone campaigned according to that reality. We really have no clue what would have happened had it been strictly a popular election as nonvoters in states that overwhelmingly go to one party may have voted seeing as their vote wouldn’t have been “wasted”. I’m not a fan of Trump but I am very happy that people voted and the person that won was sworn in as President in a peaceful transfer of power.

  4. @ModerateGuy,

    Ya, counterfactuals can get complex very quickly. Change one thing, and a bunch of other things might change, so I think you’re right that “we have no clue what would have happened had it been a strictly popular election.”

  5. @TheMiddlePolitical re popular vote for primaries,

    It’s on the to-do list – a workaround is to use PR for each region in the primaries with numbers of seats for each region equivalent to population.

Leave a Comment