Russell Freedman sits outside a polling station Tuesday in Lanesborough, Mass. (Stephanie Zollshan/Berkshire Eagle/AP)

Sen. Bernie Sanders has a relatively simple explanation for why he’s the best candidate to challenge President Trump in November.

“The way we are going to beat Trump, which is what everybody up here wants,” he said during a debate in South Carolina last week, “is we need a campaign of energy and excitement.”

“We need to have the largest voter turnout in the history of the United States,” continued Sanders (I-Vt.). “We need to bring working people back into the Democratic Party. We need to get young people voting in a way they have never done before. That is what our campaign is about.”

This has been Sanders’s sales pitch since 2016: His candidacy is the one that will prompt a surge of turnout, an electoral revolution that will sweep across the country, carrying him to the White House and progressive candidates to Capitol Hill. Sanders said it in that debate: Spurring turnout is what his campaign is about.

That rhetoric is potent, and it’s clearly the case that the Democratic Party’s nominee will need to do better at turnout than Hillary Clinton did in 2016. (More than 4 million voters who backed President Barack Obama in 2012 stayed home four years ago.) In the four states that have voted in 2020, though, Sanders is not exactly demonstrating that he can spur a surge in turnout.

In all four states, he has earned a smaller share of the vote than he did four years ago. There are three states in which we can compare actual vote totals with 2016. In two, he received fewer votes than he did then. In the other, he saw a small increase in votes — but the vote total in the state grew four times as large.

Walking through those states quickly:

  • In Iowa, Sanders earned about 26,000 fewer votes than he did in 2016, though overall turnout in the state increased over that year. (The size of that increase isn’t clear. The state party says that 171,000 people caucused in 2016, meaning a 3 percent increase this year, though actual tallies have the number of votes substantially lower. State turnout increased, but it’s not clear how much.)
  • In New Hampshire, Sanders got nearly 76,000 fewer votes than in 2016.
  • There aren’t good numbers for the raw vote in Nevada from 2016. (It, like Iowa, holds caucuses, not a primary.) Sanders’s support in the state dropped 13 percentage points between the two elections.
  • In South Carolina, Sanders added more than 10,000 votes from 2016 to 2020. Across the state, turnout increased by 169,000 votes.

We can dismiss the drops in overall support as a function of the larger field of candidates. The drop in raw vote totals, though? That’s harder to wave away. Clearly, much of Sanders’s success in 2016 was a function of opposition to Clinton’s candidacy, but that Sanders hasn’t expanded his base of support in early states since 2016 suggests that one might take his assertions about being able to spur turnout with a grain of salt. If his campaign is about turning out hardcore Sanders voters, why would he have won 152,000 votes in New Hampshire in 2016 but only 76,000 this year? It suggests that the other 76,000 voters who didn’t back Sanders or didn’t vote are not motivated by Sanders’s promised revolution. (Many, polling has shown, went to Sen. Elizabeth Warren of Massachusetts.)

The campaign and its supporters might justifiably point to Sanders’s rallies as demonstrations of enthusiasm for his campaign. It’s clearly the case that Sanders does have a large base of enthusiastic support, enough to yield tens of thousands of backers at rallies. This was the argument that Trump’s campaign used in 2016: His large rallies proved that his base was enthusiastic. And, sure enough, Trump’s support edged past Clinton’s in just enough places to win him the presidency.

It’s not clear, though, how much enthusiasm exists beyond the tens of thousands at the rallies. Polling of enthusiasm for Democratic candidates has often shown that Sanders doesn’t have a huge advantage over his competitors. A CNN-SSRS poll taken in late January, for example, suggested that there wasn’t a significant difference between Sanders and Joe Biden on enthusiasm.

Part of the problem for Sanders has consistently been that his core base of support is younger people who tend to vote much less heavily than older voters do. There are a lot of reasons for this, including that owning a home and having a higher income tend to correlate to heavier voting and aren’t as common among younger Americans. But it means that Sanders’s translation of enthusiasm into votes is probably going to be trickier than Trump’s was.

As voting unfolds on Super Tuesday, it will be interesting to watch how Sanders matches up against his 2016 numbers in the 14 states casting votes. We compiled his numbers from four years ago, if you’re interested in following along at home.

State 2016 support 2016 votes (k) Statewide votes
Alabama 19.2% 76 397
Arkansas 41.4 193 466
California 46 2,382 5,173
Colorado 59 73 123
Maine 64.3 2 3
Massachusetts 48.5 590 1,216
Minnesota 61.7 126 205
North Carolina 40.9 467 1,143
Oklahoma 51.9 174 336
Tennessee 32.5 121 372
Texas 33.2 466 1,436
Utah 79.2 63 80
Vermont 86 116 135
Virginia 35.2 276 785

In those 14 states, he earned more than 5.1 million votes. It seems safe to argue that, if his advantage in the Democratic primary field is his ability to turn out voters, he should at least match that total.

In Iowa, New Hampshire and South Carolina, he has earned 91,000 fewer votes in 2020 than he did four years ago.