Skip links

Why 2020’s Third Party Share Should Be Lower Than 2016

Dear Readers: We wanted to let you know about two upcoming University of Virginia Center for Politics events.

The Center for Politics in partnership with the UVA Parents Fund Committee and UVA Office of Engagement will present a live, online celebration this Saturday, May 16, beginning from 11:30 a.m. to noon.

Hosted by Center for Politics Director Larry J. Sabato, the online event will feature “Thomas Jefferson” (played by historical reenactor Bill Barker) live from Jefferson’s actual study at Monticello followed by a special tribute from Sen. Mark Warner (D-VA).

The public is invited to join online at here

This celebration will end promptly at noon in advance of the official University of Virginia virtual celebration and conferral of degrees, which begins at 1 p.m. on Saturday. For more information, see here.

Then, next Thursday, May 21, at noon eastern, an expert panel will examine the “Veepstakes.” Marquette University’s Julia Azari; the Washington Post’s David Byler; vice presidential historian Joel Goldstein; and the Brennan Center’s Theodore Johnson will examine Joe Biden’s options for vice president, whether Donald Trump might consider a new running mate, the historical importance of vice presidential choices, and other factors that will go into the VP selection. Crystal Ball Managing Editor Kyle Kondik will moderate the panel.

For more information, and to register, visit our Eventbrite page here.

The Editors


— Despite — or perhaps because of — the relatively high share of the vote third party candidates received in 2016, we expect the two major parties to have a better showing in 2020.

— Voters generally feel better about their major party nominees this year than they did in 2016, leaving third party options with less of a raison d'etre.

— The field of third party candidates this year doesn’t seem especially strong, and even when prominent names have launched third party bids recently, they’ve struggled to gain traction — even in their home states.

— The public health crisis could make it harder for third party candidates to get on some state ballots.

— Rep. Justin Amash’s (L, MI-3) run for president prompts us to change the rating in his now-open House seat; we discuss that change as well as the GOP’s strong showing in the CA-25 House special election at the end of this article.

Table 1: Crystal Ball House rating change

Member/District Old Rating New Rating
MI-3 Open (Amash, L) Toss-up Leans Republican

Five reasons why the third party vote should be smaller in 2020

When sizing up the 2020 election, it is natural for campaign pros and analysts to consider the kinds of voters who might defect from their 2016 selection. Can Joe Biden claw back some small but important share of the Barack Obama voters who defected to Donald Trump? Can Trump stop his party’s slide in the suburbs?

These are important questions, but there’s a significant bloc of 2016 voters who did not express support for either Hillary Clinton or Donald Trump, and who hypothetically may be gettable for both sides: the third party voters.

About 6% of all the ballots cast in 2016 for president were for candidates other than Trump and Clinton.

It’s our belief that this third party share will be smaller in 2020 — meaning that we think there’s a decent chance that at least some voters who opted out of the major party contest in 2016 will opt-in this time.

First, we’ll review the third party vote both in 2016 and historically. Then, we’ll get into five reasons why we think the third party share will be smaller in 2020. As a note, all of the numbers cited in this piece are from Dave Leip’s Atlas of U.S. Presidential Elections, an excellent source that we (and other analysts) frequently use.

Libertarian Party nominee Gary Johnson won about 3.3% of the vote in 2016, and Green Party nominee Jill Stein won about 1.1% of the vote. Another half a percentage point went to Evan McMullin, a conservative alternative to Trump who got a third of his votes from a single state: Utah, where a significant slice of the electorate (21%) opted for McMullin. Utah is the most Mormon state in the country, and some voters there who were skeptical of Trump rallied to their Trump-critical co-religionist, McMullin.

Write-ins accounted for almost another full percent of the third-party vote; probably the most notable write-in effort came in Vermont, where home-state Sen. Bernie Sanders (I-VT) received nearly 6% of the vote as a write-in.

Votes for minor candidates amounted to the remainder of the third party mix.

In some ways, one could argue that the performance of third party candidates in 2016 was impressive. Johnson and Stein, who also were on the ballot in 2012 as their respective party standard bearers, both roughly tripled their share of the vote from their previous election. Johnson, at a little over 3%, performed way better than any Libertarian ever had before — Ed Clark, at about 1.1% of the vote in 1980, was previously the best-performing Libertarian (Johnson himself won about 1% in 2012).

On the other hand, that third party candidates only received 6% of the vote despite the fact that Clinton and Trump both had horrible personal favorability numbers shows the enduring strength of the two major parties — and the challenges for third party candidates in general.

Since the end of the Civil War, there have been 38 presidential elections. The 2016 third party share ranks as the eighth-highest in that timeframe.

The history is in Table 1, with the elections sorted from highest third party share to lowest.

Table 1: Third party presidential voting, 1868-2016

Note: *In 1968, one faithless elector in North Carolina cast his vote for George Wallace rather than Richard Nixon. In 1972, one faithless elector in Virginia cast his vote for Libertarian John Hospers rather than Nixon. The “Oth. EVs” column only includes electoral votes cast for third party candidates who competed in the election.

Source: Dave Leip’s Atlas of U.S. Presidential Elections

This history provides a good jumping off point to our five arguments as to why we think the third party vote will be smaller in 2020 than 2016.

1. The parties are largely unified

One thing that is somewhat surprising about the history of third party presidential voting in the post-Civil War era is that some of the best years for non-major parties came in years when there was an incumbent on the ballot. This throws some cold water on what we thought might be the case before we took a look at the numbers, which was that we thought it was possible that the third party vote might go up in years when the presidential race does not feature an incumbent. The reasoning would be that races featuring incumbents might represent more of a referendum on the incumbent, and therefore would sort the electorate into two distinct camps: Those who support the incumbent and want to see him reelected, and those who do not and want to see him defeated.

Yet of the seven elections with higher third party shares than 2016, six were in years where there was an incumbent on the ballot. The one exception was the explosive 1968 election, when Lyndon Johnson could have run for a second elected term but opted not to.

A better way, perhaps, to look at the third party votes in some of these elections is to see them as representing divides within the major parties.

Certainly, the top election on this list — 1912 — is the cleanest example of a divided party leading to the rise of a big third party vote. Theodore Roosevelt, upset with the performance of his Republican successor, William Howard Taft, tried to win the GOP nomination. He was rebuffed, so he created his own party and ran for president. The Republican vote splintered, and Democrat Woodrow Wilson won the presidency easily despite getting only 42% of the vote.

But we can also see this phenomenon in some of these other elections. George Wallace, the conservative, segregationist Democrat who ran third party in 1968, ran strongest in the South, the conservative region that had once formed the backbone of the Democratic Party but was in the midst of breaking away from its ancestral party over the party’s leftward evolution on civil rights and other issues. This process did not happen overnight: The presidential candidacy of Strom Thurmond two decades prior, in 1948, also represented a backlash spasm by southern conservatives against the growing liberalism of the Democrats.

While Ross Perot took some of his many votes in 1992 (and fewer votes in 1996) from both major parties, he in some ways represented a continuation of Patrick Buchanan’s 1992 primary campaign against incumbent George H.W. Bush: Both ran as critics of big institutions and were skeptical of free trade (one can see shades of Perot and Buchanan in Donald Trump, all of whom appeal to a group of voters that the late sociologist Donald Warren dubbed “Middle American Radicals”).

In 1980, moderate Republican Rep. John Anderson of Illinois ran as something of a liberal alternative to both major party candidates. This came after Ronald Reagan won the GOP nomination, representing a triumph for conservatives in the party over moderates, while Jimmy Carter fended off a challenge from the left via Sen. Ted Kennedy of Massachusetts. The exit poll in 1980 suggested that Anderson took slightly more votes from Democrats than Republicans, and he also turned in some of his better performances in counties with major college campuses.

While there are of course divides in both parties, these sorts of major fissures do not seem present in 2020. Donald Trump had only nominal opposition in the Republican primary, and he dispatched that opposition with impressive ease. After early stumbles, Joe Biden effectively knocked out his rivals over the course of a few weeks in March. While there is a portion of the left that is supportive of Bernie Sanders and highly skeptical of Biden — how seriously someone is taking Tara Reade’s sexual assault allegation against Biden is a good test of Biden skepticism on the left — Biden and Sanders themselves seem to get along well, and Biden performing better head-to-head against Sanders than Clinton did suggests more acceptance of his nomination among Democrats.

This naturally removes some of the oxygen for third party candidates, and the lack of major intraparty strife makes this election, to us, more reminiscent of 2004 and 2012, when George W. Bush and Barack Obama won second terms in competitive elections that featured very low levels of third party voting. Indeed, in 2012, Florida was the only state were neither major party candidate took a majority of the vote — by 2016, there were 14 states where both major candidates polled under 50%.

2. Opinions of the major party candidates are better

This second argument goes hand in hand with the first. The parties, and the public at large, seem to like their major party choices better this time than four years ago.

One of the key exit poll findings from 2016 was that almost 20% of voters held an unfavorable view of both Hillary Clinton and Donald Trump, an unusually high number. The president won this group 47%-30%, according to the exit poll, which was a crucial factor in his victory. The remaining voters who held unfavorable views of both candidates (23%) reported that they voted third party, according to the exit poll. Given that only 1% of voters with favorable views of just Clinton or Trump voted third party, it’s reasonable to assume that the vast, vast majority of third party voters in 2016 disliked Clinton and Trump. That certainly seems to make sense.

In the exit poll, Trump’s favorable/unfavorable split was 38%/60%, while Clinton’s was 43%/55% (the RealClearPolitics average of their favorability on Election Day was very similar to the exit poll).

It seems possible, if not likely, that there will be fewer voters who have unfavorable views of both Biden and Trump this time.

Trump’s favorability/unfavorability split in the RealClearPolitics average right now is 42%/53%. That’s not good, but it’s significantly better than when he was a candidate (his approval rating is 45% approve, 51% disapprove). Meanwhile, Biden’s favorability split is 42%/48%. Again, that is not good, but it’s also better than Clinton.

If there are fewer voters with unfavorable views of both candidates, that may deprive the third party candidates of voters given how dependent the third party candidates were on critics of the major party candidates in 2016.

(As an aside, it appears that Biden is doing much better than Trump among the voters who do hold both an unfavorable view of both major party candidates. This is a major trend to watch as we get closer to November. Politico’s David Siders wrote about this trend earlier this week.)

3. Third party candidates may be weaker

Many of the most successful third party candidates historically had some sort of previous political success or platform. Theodore Roosevelt, a former, well-regarded former president, is obviously a great example.

Sen. Robert La Follette Sr. of Wisconsin and former Gov. George Wallace of Alabama were already nationally prominent politicians when they mounted their historically noteworthy third party bids.

Ross Perot had not been a candidate before, but he was a wealthy, prominent business leader and advocate for Vietnam war POWs and their families.

Although he did not achieve the electoral success of the aforementioned candidates, Ralph Nader was a nationally-known consumer advocate and public figure before his electorally consequential Green Party presidential run in 2000.

The current roster of possible third party candidates lacks such a big name.

The most notable candidate so far appears to be Rep. Justin Amash of Michigan’s Third District, who left the Republican Party over his opposition to Donald Trump on July 4 of last year. Amash recently announced that he was running for the Libertarian nomination.

At this point, we would not consider Amash a huge name in American politics: It’s debatable whether he has more notoriety than Gary Johnson, the aforementioned 2012 and 2016 Libertarian nominee. Johnson is a former two-term governor of New Mexico who sought the GOP presidential nomination in 2012 before his Libertarian runs. That represents much more national exposure than Amash can currently boast. Amash also is not necessarily guaranteed to be the Libertarian nominee; he has several lower-profile opponents for the bid, which will be decided late next week.

The Green Party nomination appears likeliest to go to Howie Hawkins, a frequent Green Party candidate in New York state. Given that Jill Stein had already been the Green Party presidential candidate in 2012 before running again in 2016, whatever paltry name identification Stein possessed four years ago is probably greater than what Hawkins would start his campaign with. Former Minnesota Gov. Jesse Ventura, a retired professional wrestler elected to a single term as an independent in 1998, recently floated the idea of seeking the Green nomination, but he announced he would not run late last week.

Earlier this month, the Constitution Party nominated Don Blankenship as its nominee. Blankenship, a former West Virginia coal baron, was last seen seeking the Republican nomination to take on Sen. Joe Manchin (D-WV) in 2018. He’s known to hardcore political watchers — mainly as a novelty thanks to the bizarre “Cocaine Mitch” moniker he attached to Senate Majority Leader Mitch McConnell (R-KY) — but that’s probably it.

There isn’t really another third party/independent candidate of note. Former Starbucks CEO Howard Schultz floated a bid last year, but after widespread criticism from Democrats that he would simply help Trump get reelected, he decided not to run.

In other words, it appears that the 2020 third party options will be even more anonymous than the 2016 options.

Even when third parties have landed candidates with previous national experience — as the Libertarians may do if Amash is the party’s nominee — they have not recently had an especially strong effect in their home state. In 2008, former Rep. Bob Barr (R, GA-7) ran as the Libertarian nominee. In Georgia, Barr took just seven-tenths of a percentage point statewide. In counties that he previously represented, he struggled to crack 1% of the vote.

In 2012, former Rep. Virgil Goode (R, VA-5) ran as the Constitution Party’s nominee. Despite any local name recognition he brought, he ended up finishing behind what Gary Johnson, a non-native, received on the Libertarian line — both received under 1% statewide. In Franklin County — which has been his longtime political base — Goode took less than 3%.

Though Gary Johnson received relatively robust home state showings in New Mexico — 3.6% in 2012, and 9.3% in 2016 — it’s unclear that his presence helped or hurt either party. Hillary Clinton’s 48%-40% margin in the state was only down slightly from Obama’s 53%-43% four years earlier.

As both parties will certainly be ramping up their mobilization efforts in his state, Amash could end up like Barr or Goode. Amash’s district covers much of Kent County (Grand Rapids), a usually Republican county that is arguably trending Democratic. Depending on how close Michigan is, Amash could be a factor. But when it’s all said and done, Amash may struggle to gain significant support even in his own backyard.

4. The pandemic may limit ballot access

It can be challenging, at least in some states, for non-major party candidates to get on the ballot. The public health crisis might make it even harder to gather the signatures they often need to get on ballots.

>Map 1 shows where the two most significant third parties, the Libertarians and the Greens, are on the ballot so far. In 2016, the Libertarians were on the ballot in all 50 states, while the Greens were on the ballot in all but a few.

Map 1: Libertarian/Green Party progress in getting on state ballots

So far, the Libertarians are on the ballot in 36 states, while the Greens are on the ballot in 23, according to the websites of the respective parties. Some states might dial back their ballot access requirements or may be compelled to do so by lawsuits, which only seems fair given the circumstances (and third party proponents argue that such rules only exist to advantage the major parties, which is a decent point).

If you’re curious about the deadlines and requirements for specific states, and to track third party efforts going forward, see the April issue of Ballot Access News. Bill Scher also wrote a helpful overview of the challenges for third parties for Politico Magazine a few weeks ago.

We would expect the third party share in 2020 to drop even if the level of third party ballot access remained the same. That the third parties may be on fewer state ballots this time could also serve to reduce the share of the vote they collectively receive.

5. Voters may be more aware of the stakes

It is a hard thing to prove, but we think it’s possible that some number of voters cast third party votes in 2016 because they didn’t think their votes would matter, and they wanted to protest the major party candidates.

Election forecasters, betting markets, and the general public (as measured by polls asking voters who they thought would win) all favored Hillary Clinton to win in 2016. As we’ve argued in the past, these expectations may have aided Republicans down ballot, and they may also have led to some voters casting protest votes because they didn’t think the race was as close and competitive as it ended up being. This time, even though Trump generally trails nationally and in at least some of the most important swing states, he still is favored by betting markets, and he usually does better in polls asking people who they believe will win as opposed to those that ask who voters are supporting. Democrats, burned by expectations in 2016, likely will remain guarded no matter what the polls say.

Additionally, there may be a stigma attached to voting third party that didn’t exist in 2016. Democrats have noted that the number of votes cast for the Green Party’s Stein exceeded Trump’s margin of victory in Michigan, Pennsylvania, and Wisconsin, the three crucial northern states that effectively determined the election. Now, we think it’s fallacious to argue that every single one of those Stein voters would have backed Clinton had Stein not been on the ballot — some third party voters are just not gettable by the major parties, or at least not by establishment candidates like Clinton (and Biden) — but there probably also are some Stein voters who regret their choice given that Trump upset their expectations and won.

That Democrats essentially shamed Howard Schultz into not running is another indicator of the social pressure many Trump opponents may bring to bear on those considering backing someone other than the Democratic nominee.

This pressure may work the other way too: Republicans who don’t like Trump and who voted third party in 2016 might have come around to Trump because he has in many ways governed like a standard-issue Republican. If they do not, they could face social pressure from their own Trump-voting family and friends who do not want to see the White House fall into the hands of the Democrats.


Asserting that the third party vote should be smaller this time than in 2016 does not mean we don’t necessarily think it won’t be impactful. In a very close election, as 2020 may be, all sorts of factors can contribute to the outcome.

In 2008, then-GOP Rep. Ron. Paul, who was known for his libertarian brand of conservatism — indeed, he was that party's 1988 presidential nominee — was on the ballot in Montana as the Constitution Party's nominee. Paul didn't actively campaign but ended up taking just over 2% of the vote. John McCain carried the state, but with just a 49% plurality, to Barack Obama's 47%. McCain, who favored a hawkish foreign policy, was anathema to Paul's base. Still, if Paul had received a higher share of the vote of disaffected Republicans — and thus spoiled the state in Obama's favor — it would have provided Obama just three additional votes in the Electoral College, where he had already won overwhelmingly.

Another instance of when a third party candidate may have influenced a state was Minnesota in 2016. That year, the aforementioned Evan McMullin was on the ballot there on the state's Independence Party's line. In a state that Hillary Clinton carried by less than two percentage points, McMullin courted the anti-Trump bloc of the GOP. Clinton's raw vote margin in Minnesota was about 45,000 votes — less than the 53,000 votes McMullin earned. Would his support have otherwise gone to Trump? Perhaps, but looking to the fall, McMullin's supporters may be poised to make the transition to the Democratic Party. McMullin received his highest vote totals in MN-2 and MN-3. In 2018, Democrats flipped both those Twin Cities-area seats at the congressional level, and the region has lurched left in recent statewide races.

Generally speaking, we think Biden would benefit more from a straight-on head to head election with Trump, because it would force Trump disapprovers — a larger group than Trump approvers nationally — to choose the only anti-Trump alternative. If Trump disapprovers have a larger menu of anti-Trump options, some of them might choose candidates other than Trump.

There’s not a whole lot of polling that includes named third party candidates. Monmouth University, a respected national pollster, recently found Biden leading Trump 50%-41%, and 47%-40% with Amash on the ballot, receiving 5%. So, in this poll, the inclusion of Amash mildly hurt Biden more than Trump. A Morning Consult poll last month testing the race with and without Amash found Biden up four points both times, with Amash at just 1%.

Third party candidates typically poll better than they perform: Johnson hit double-digits in several national polls throughout 2016, and Stein routinely polled better than what she ultimately received.

Before his 1935 assassination, Huey Long — a fiery Democratic senator from Louisiana — was weighing a third party run against Franklin Roosevelt, in the 1936 election. With his “Share Our Wealth” slogan, Long famously joked that one party is “skinning you from the ankle up and the other from the ear down.”

More than three decades later, another southern populist would borrow from Long's rhetoric. During his 1968 presidential bid, George Wallace asserted there was not a “dime’s worth of difference” between the Democrats and the Republicans. As Ronald B. Rapoport and Walter J. Stone note in their study of Ross Perot’s candidacies, “All successful third-party candidates follow Wallace’s strategy of emphasizing the absence of a ‘real choice’ between the two major parties.”

Partisan polarization, as well as the grim realities of the 2020 national outlook and the starkly opposing political beliefs of Trump supporters and opponents, should serve to focus the minds of voters, and accentuate the very real differences between the two major party candidates. That all may act to reduce 2020’s third party vote as well.

P.S. U.S. House updates

— Amash’s Libertarian presidential run likely means he won’t be running for reelection to MI-3, a conservative-leaning district in western Michigan, although if Amash is not the Libertarian nominee, he technically still has time to get on the ballot and be part of a three-way race to retain his seat in the fall. But at this point it’s reasonable to expect a conventional race between a Republican and a Democrat here. House Minority Leader Kevin McCarthy (R, CA-23) recently endorsed veteran Peter Meijer (R), whose family founded the chain of retail stores that shares his last name. Meijer has to navigate a primary — state Rep. Lynn Afendoulis (R) appears to be the other top contender — and the GOP nominee will face attorney Hillary Scholten (D) in the general election. Trump carried MI-3 51%-42%, and it covers a battleground area of the state where Biden may need to run ahead of Hillary Clinton’s 2016 showing to win statewide. In fact, even as Gov. Gretchen Whitmer (D-MI) was elected by a robust 53%-43% vote in 2018, she came up just short in the district. This is a competitive race, but one where we think the GOP has an edge: We’re moving MI-3 from Toss-up to Leans Republican.

— Rep.-elect Mike Garcia (R) has recaptured CA-25 for Republicans less than two years after they lost it to Democrats in 2018. Garcia was leading by 12 points in the special election with an uncertain number of outstanding ballots remaining, and his opponent, state Assemblywoman Christy Smith (D), conceded Wednesday afternoon (the Associated Press called the race soon thereafter). This is a district that Hillary Clinton won by seven points, so the victory by Garcia is quite impressive even given the circumstances. Garcia clearly ran a better campaign than Smith, but he also probably was aided by an electorate that was smaller and more Republican-leaning than what we might expect in the fall. If you’re curious about more of the details from the race, see this Twitter summary that we put out late last week. We are still debating what we’ll do with the rating here after all the results come in — we left it as a Toss-up for the special election after moving it from Leans Democratic more than a month ago, and we’re keeping it there for the fall, at least for the time being.

— Rep.-elect Tom Tiffany (R, WI-7) easily held a sprawling northwest Wisconsin district in Tuesday’s other special House election: He won by about 14.5 points in a district Trump carried by about 20 points. Margins in any Wisconsin race will attract attention just because the state is so important to the presidential race; in order to hold the state, Trump probably cannot afford any erosion in a district like this, although he may have a personal vote in the district that doesn’t necessarily extend to other Republicans. Interestingly, the total turnout in the WI-7 special was almost exactly the same as in last month’s closely-watched state Supreme Court race, which the Democratic-backed candidate won comfortably statewide. But Tiffany won 57% of the vote in WI-7, while the GOP-backed court candidate took 53% in the district last month.

— All in all, Tuesday was a good night for House Republicans, although our overall outlook in the House (Democrats favored to retain their majority) remains unchanged.

Nebraska held its primary on Tuesday night, and progressive 2018 nominee Kara Eastman (D) won the right to a rematch against Rep. Don Bacon (R, NE-2). Eastman easily defeated Ann Ashford (D), the wife of former Rep. Brad Ashford (D, NE-2), who Eastman also dispatched in a primary upset two years ago (Brad Ashford lost this seat to Bacon in 2016). Bacon beat Eastman by two points in 2018 — he was one of the few swing district Republicans to survive the Blue Wave – as national Democrats largely abandoned Eastman because they thought she was too liberal to win the seat. Immediately following Eastman’s primary win, the Democratic Congressional Campaign Committee released an internal poll showing Eastman up 48%-47% on Bacon, while Biden was leading Trump in the district 52%-41%. Always take internal polls with a grain of salt, and the size of Biden’s lead (Trump won this district by two points in 2016) calls the House result into question. However, what is perhaps more notable is that establishment Democrats have come to terms with Eastman as the nominee, and she may get more support than last time. She could use it, as Bacon starts the general election with a $765,000 to $99,000 cash-on-hand edge over Eastman, per the most recent Federal Election Commission reporting.

This district awards its own electoral vote, and we would not be surprised at all if Biden won it, which could threaten Bacon. NE-2 covers a major city, Omaha, and features some diversity and higher-than-average four-year college attainment rates — all characteristics of places where Trump has seemed to lose typical Republican support. We rate the district a Toss-up for president, and Leans Republican for the House. Map 2 shows the results of the primary as compared to 2018, which was much closer. The maps are broken down by legislative district. Note, specifically, the 11th district — the one in deep Ashford blue on the 2018 map that flipped strongly to Eastman this time. It’s the only black-majority legislative district in the state, and turnout there will be crucial to Democratic chances in the district.

Map 2: NE-2 Democratic primary results