From left, Libertarian presidential candidate Gary Johnson, Green Party presidential nominee Jill Stein and independent presidential candidate Evan McMullin. (STF/AFP/Getty Images/file)

Hillary Clinton was not the only presidential candidate to lose to Donald Trump on Tuesday. Joining her were minor party and independent candidates, including Libertarian Party nominee Gary Johnson and Green Party nominee Jill Stein. Johnson and Stein were expected to perform exceptionally well for minor-party candidates — and perhaps even to break through and establish their parties as viable alternatives in future elections. That didn’t happen.

Here are answers to five key questions about the performance of third parties in 2016.

Did Johnson and Stein cost Hillary Clinton the election?

We cannot know how Johnson and Stein backers would have voted if forced to choose between Clinton, Trump and staying home. It may be some time before we have the data we need to know for sure. But let’s work with the data that we have.

First, in five states Trump won by a margin smaller than the combined Johnson/Stein vote: in Arizona (by 0.6 percentage points), Florida (1.6), Michigan (4.4), Pennsylvania (2.0) and Wisconsin (3.7).

What if we add the entire Stein vote to Clinton’s total in these states? This would have flipped the outcome in Michigan (16 electoral votes) and Wisconsin (10), and left Clinton just 19,234 votes short (as of this writing) in Pennsylvania (20) — a margin she might have overcome without Johnson in the race. Adding the first two states to Clinton’s column would give her 258 electoral votes, while Trump would have 280. Adding Pennsylvania would have put Clinton over the top, 278 to 260.

But this projection rests on the unrealistic assumption that all Stein voters would have voted for Clinton. Probably some would have stayed home, skipped the presidential race or voted for another candidate.

Second, the 2016 exit polls asked voters how they would have voted if forced to choose between the two major-party candidates. Based on those answers, in the five states identified above, Trump would have kept his lead in Arizona (48 percent to 44 percent) and Florida (48 to 46 percent). Clinton would have flipped Michigan (48 to 44 percent).

In Pennsylvania and Wisconsin, Clinton and Trump would have tied at 48 percent apiece.

In each case, the other respondents said they would have abstained from voting.

Clinton might have won, based upon these data, but only by winning both Pennsylvania and Wisconsin. If Trump held onto even one, he would have kept an electoral college majority.

But we can’t be certain that this year’s exit polls were accurate. They led some pollsters to forecast a Clinton victory. If the exit polls were skewed in Clinton’s direction by even one percentage point in Pennsylvania or Wisconsin, she would still lose in the electoral college.

How did Johnson and Stein do in 2016?

Gary Johnson won 3.3 percent of the national popular vote, and Jill Stein 1.0 percent (as of this writing). This is a dramatic improvement over 2012, when the same two won 0.99 percent and 0.36 percent, respectively. Johnson’s showing is the best in Libertarian Party history. Stein’s falls well short of Green nominee Ralph Nader’s 2.7 percent in 2000.

On the one hand, the Libertarians and Greens tripled their vote share from the previous presidential election and further established themselves as the most viable organized alternatives to the two major parties.

On the other hand, Johnson and Stein never were serious threats to Clinton and Trump nationally or in any individual states. And they were far from the most successful minor party candidates in recent presidential history: George Wallace won 13.5 percent of the vote as the American Independent Party nominee in 1968, and Ross Perot won 8.4 percent as the Reform Party’s nominee in 1996. (Perot also won 18.9 percent in 1992 as an independent candidate, and John Anderson won 6.6 percent in 1980 as an independent).

Weren’t Johnson and Stein supposed to do even better this year?

Many political observers thought this might be the year that a minor-party or independent candidate finally became a viable alternative in a presidential race.

After all, Clinton and Trump had historically high unfavorable ratings for presidential candidates. Both were scandal-prone, from Clinton’s FBI email investigation and embarrassing WikiLeaks revelations to Trump’s tax records and treatment of women.

Add to that voters’ apparent disdain for the two-party system and “politics as usual,” and conditions would seem to have been ripe for a serious minor-party threat in 2016.

Further, the Libertarian Party ticket of Johnson and Bill Weld, as we noted here earlier this year, appeared to be the most experienced minor-party presidential ticket in at least a generation. Both are former two-term governors who served as Republicans in reliably Democratic states — Johnson in New Mexico, Weld in Massachusetts. Unlike many minor-party candidates, Johnson and Weld were on the ballot in all 50 states. Their message of fiscal conservatism and social liberalism seemed to appeal to disaffected voters in both parties.

Early polling even suggested that Johnson-Weld might earn the 15 percent support necessary to qualify for the fall’s presidential debates. In the end, they didn’t come close.

Why didn’t they do better?

Let’s take this question in two parts: first, focusing on the structural disadvantages that Johnson and Stein faced; second, focusing on the weaknesses of their particular campaigns.

First, minor-party candidates are at a massive disadvantage in U.S. elections. Our electoral institutions are designed to ensure, although not mandate, a two-party system. Amanda Skuldt explained how in an excellent Monkey Cage piece.

The short version is this: Most American elections award a single office to whichever candidate wins the most votes. This creates powerful incentives for parties to build the broadest possible coalitions to win an electoral plurality — and for voters to choose among these coalitions in order to avoid “wasting” a vote on a candidate who will not win. Add a host of state laws limiting ballot access for minor-party candidates.

Second, Johnson’s and Stein’s prospects considerably weakened over the campaign. Johnson was winning from 7 to 9 percent of the national electorate’s support between mid-July and mid-September. That started to fall steadily in late September, to less than 8 percent just before the first presidential debate on Sept. 26 and below 7 percent just before the second presidential debate on Oct. 9. By Nov. 8, he was projected to win just 5 percent of the vote. But even that was too high: Johnson won only 3.3 percent of the national popular vote, suggesting that the polls had overstated his support throughout the campaign.

Similarly, Stein was polling at 2 percent, on average, at the end of the campaign, but won only half that in actual voting. She had been polling up to 5 percent in mid-July, before falling to 3 percent in late September and 2 percent in late October, according to the RealClearPolitics average.

Why such a decline? It is tempting to blame the candidates’ high-profile gaffes, such as Johnson’s infamous “Aleppo Moment” and Stein’s air travel embarrassments. Actually, the polling declines did not coincide with these or other gaffes, but with the presidential debates. Presumably, the binary logic of the two-party system kicked in and many voters chose what they considered the “lesser of two evils.” U.S. minor-party candidates usually lose ground as campaigns end, probably for that same reason.

What are the prospects for minor parties going forward?

The Libertarian and Green parties had some success in 2016; in addition to tripling their vote share from 2012, both gained significant media attention via CNN town halls and television advertisements. Yet they faced challenges that go beyond those noted above, including a lack of federal matching funds, state ballot-access restrictions, and voters’ willingness to vote for the “lesser of two evils.” There may be different presidential candidates in 2020, but probably not different rules. If so, minor parties shouldn’t expect the results to change, either.

Christopher J. Devine is assistant professor of political science at the University of Dayton. Find him on Twitter @ProfDevine. 

Kyle C. Kopko is assistant dean for academic achievement and engagement, and associate professor of political science at Elizabethtown College. Find him on Twitter @KyleKopko.

They are the authors of “The VP Advantage: How Running Mates Influence Home State Voting in Presidential Elections.”