Mellman: The 2016 and 2020 races: What's different?

Mellman: The 2016 and 2020 races: What's different?
© Getty Images

We’re just past the traditional Labor Day start of a presidential campaign that feels like it’s been engaged for four years, so it’s finally reasonable to ask how it’s looking.

The answer’s simple: good for Joe BidenJoe BidenBiden adds to vote margin over Trump after Milwaukee County recount Krebs says allegations of foreign interference in 2020 election 'farcical'  New DOJ rule could allow executions by electrocution, firing squad MORE.

He’s ahead in the national polls and in states worth more than 315 electoral votes, with 270 needed to win.


Modelers at FiveThirtyEight give Biden a 71 percent chance of winning, while The Economist’s statisticians set the chances at 84 percent.

But not so fast.

One could travel back to September 2016, substitute Hillary ClintonHillary Diane Rodham ClintonValadao unseats Cox in election rematch Trump says he'll leave White House if Biden declared winner of Electoral College Federal workers stuck it out with Trump — now, we're ready to get back to work MORE for Biden and The New York Times for The Economist, and written the same two sentences.

No one read the evidence quite correctly then, a failure that should elicit some modesty in pronouncing upon this year’s election.

But modesty in predictions shouldn’t mean ignoring the evidence.

This strength of this year’s evidence rests on important quantitative and qualitative differences between this election and the last.


Biden is polling meaningfully better than Clinton did. The Sunday prior to Labor Day, FiveThirtyEight’s polling average gave Clinton a lead of 3.4 percentage points — overstating her final national margin by a small 1.3 percentage points.

This past weekend, Biden was ahead by 7.5 points.

Pundits love saying the popular vote doesn’t count, which is true constitutionally and theoretically, but ridiculous in practice.

It’s relatively easy for a Democrat to win the popular vote by 2.1 points, as Clinton did, while losing the Electoral College, but it’s virtually impossible to win the national vote by 7.5 points (or even 6 points) and lose the White House.

It’s also worth noting that at this point in 2016, some 19 percent of voters were opting for third party candidates or telling pollsters they were undecided, creating significant room for volatility.

Today, that number is less than half its 2016 level, suggesting less room for movement.

Moreover, those who voted for third party candidates in 2016, or didn’t vote at all, are listing in Biden’s direction this time.

Differences between 2016 and 2020 go beyond the horse race data.

As I pointed out in 2016, that election was decided by the 18 percent who disliked both candidates and gave Trump a whopping 17-point vote margin nationally, and similar, or greater margins in swing states.

2020 presents two differences. Because Biden’s unfavorables appear to be about a dozen points lower than Clinton’s, the double-haters are likely to be a smaller segment this time around.

Moreover, folks in that group are leaning toward Biden in 2020 by even bigger margins than they supported Trump in 2016.

That may reflect a vitally important qualitative distinction between the two elections: in 2016 some voters said to themselves, “Trump will change, he’ll grow into the office. He’ll be a better man as president.”

If you believed that, you had an excuse to vote Trump. That excuse has disappeared.

Regardless of what you think about Trump’s performance, no one now expects him to change or be different in a second term. He is who he is, and he will be what he is.

Who he is, and how he’s done, are central questions this year. The natural state of presidential elections featuring incumbents is to be a referendum on the president.

That’s why approval ratings and votes are so tightly correlated. Since Harry Truman, only one sitting president has seen his percentage of the vote exceed his approval rating by more than 3 points. Jimmy CarterJimmy CarterCan Biden vanquish Democrats' old, debilitating ghosts? CNN acquires Joe Biden documentary 'President in Waiting' French radio station mistakenly publishes obituaries of celebrities MORE’s vote grew 4 points above his 37 percent September approval rating.

With Trump currently clocking in at 43 percent, matching Carter’s boost would still leave him losing by 6 points.

Despite this catalogue of advantages for Biden over Clinton, one important discordant note rings out.


In only one swing state (Arizona) is Biden leading by a margin greater than Clinton was at this point in 2016.

Modesty in predictions should remain our guidepost.

Mellman is president of The Mellman Group and has helped elect 30 U.S. senators, 12 governors and dozens of House members. Mellman served as pollster to Senate Democratic leaders for more than 20 years, as president of the American Association of Political Consultants, and is president of Democratic Majority for Israel.