People are asking this question—or flat out claiming third parties did sink her—because they are worried about how such parties might affect the Democrats' chances of defeating Trump in 2020. As one example, Josh Marshall recently stated:
[I]t's really the unusually high 5.7% of the vote going to three third party candidates—Gary Johnson, Jill Stein and Evan McMullin—that made it possible for Trump to win as a minority candidate.
Thinking about the 2020 election, there are certainly scenarios where third parties, depending on their type and the distribution of their vote, could hurt the Democrats.
But to set the record straight, 2016 does not appear to have been one of those times. In a “States of Change” report, we performed the exercise of re-allocating the “extra” third-party vote to see how the election outcome might have been affected if those third-party voters had voted for the Democrats or Republicans. As we explained in the report:
One of the unique features of the 2016 election was the relatively high third-party vote. Nationally, third-party candidates in 2016 collectively garnered about 4 points more than they did in 2012—5.7 percent versus 1.7 percent. While it is possible that similarly high levels of support will appear in future elections, the historical trend would suggest that a decline is more likely after a spike. Given that trend, the authors developed a separate 2016 baseline where third-party vote share is returned to its lower 2012 levels and the rest of the third-party vote share is reallocated based on underlying partisan preferences.
The result: Trump still wins the electoral vote, only by a larger margin, 309–229. This is because he still carries the key states of Michigan, Pennsylvania, and Wisconsin, while also carrying New Hampshire by a narrow margin. This makes sense when one considers the actual distribution of the third-party vote in these states: Michigan, Johnson 3.6 percent/Stein 1.1 percent; New Hampshire, Johnson 4.9/Stein .9; Pennsylvania, Johnson 2.4/Stein .8; Wisconsin, Johnson 3.6/Stein 1.
So the third-party effect is not necessarily anti-Democratic. And Hillary Clinton did not lose the 2016 election because of it. As for 2020, we should wait until we have more information before we make a judgment on who it will help and who it will hurt.
