The Odds of a 2016 Redux

The anxiety levels of more than a few Democrats increased exponentially last Friday with the almost simultaneous publication of two articles by really bright, independent political analysts who reached the same conclusion: Despite his very low job-approval ratings, President Trump could win another split-decision election, prevailing in the Electoral College while losing the national popular vote.

In 2016, Hillary Clinton won the national popular vote by 2.9 million votes out of 137 million cast, representing 2.1 percentage points, but as we were reminded that Election Night, the popular vote and $5 will get you a cup of coffee at Starbucks. By a scant 78,000 total votes in Michigan, Pennsylvania, and Wisconsin, Trump won 30 states that have a combined 306 electoral votes, to Hillary Clinton’s 20 states plus the District of Columbia and their combined 227 electoral votes.

The New York Times’s Nate Cohn and my Cook Political Report colleague David Wasserman make persuasive cases as to how it could happen again. In fact, some of the factors that contributed to the split decision last year could be even more pronounced in 2020. Cohn’s piece, “Trump’s Electoral College Edge Could Grow in 2020,” argued that the same states that were so close last time will be again, focusing mostly on Wisconsin and Florida.

Wasserman’s piece for NBC News, “How Trump could lose by 5 million votes and still win in 2020,” takes a slightly different angle but is equally persuasive. Wasserman points out, for example, that a Democrat might well win California by an even bigger margin in 2020 than Clinton did in 2016, in effect “wasting” more votes by running up the score there. Meanwhile, Trump could win Texas this time around by a much smaller margin, meaning fewer “wasted” GOP votes.

It goes to the core of a problem for Democrats—they tend to have their votes highly concentrated, while Republican votes are more efficiently allocated around the country. Both analyses also point to the importance of white, working-class voters, a group that was key in Michigan, Pennsylvania, and Wisconsin, the three states that tipped the election to Trump, and the fact that he has seen little erosion among that group.

A partisan Democrat reading the two pieces would, at best, get a bad case of indigestion. At worst they might be looking for a ledge to climb out on. Both writers back up their arguments with solid numbers and analysis and should not be dismissed. But these aren’t the only two ways of looking at the 2020 election math. Other, equally solid cases with good data point to a strong chance for a different outcome.

For example, Ronald Brownstein’s piece for The Atlantic, “Trump’s Base Isn’t Enough,” looks at “conflicted voters,” swing voters who approve of Trump’s management of the economy but still disapprove of his overall performance as president. Brownstein points to the continued support of those same working-class white voters who keep Trump in the hunt in key swing states that have disproportionately large numbers of them. But the pull of the economy (assuming it remains strong—no guarantee there) may be offset by concerns about his character. How many times can he say “God-d----d” in speeches (as he did in Greenville, North Carolina last week) before it troubles at least a few in his flock, enough to tip a tight race?

While it is almost universally believed that we will have a huge voter turnout next year (perhaps even setting a modern record as the 2018 midterm elections did), it isn’t clear which side will benefit. I would argue that the bigger the turnout, the more the electorate looks like all registered voters and the more likely that polls will be accurate. For all of the problems in survey research today (and there are plenty), polling has the toughest time when voter turnout is low or even average. But when it is huge, the disparities between various groups are mitigated. High-propensity voter groups go up less because they are already high. Low- to moderate-propensity voter turnout increases more, because they have a lot of room for growth.

The fact is that the country is not only narrowly divided but hyper-partisan as well. That means nobody is going to win big. As we learned on Election Night 2016, small things can make big differences. These three pieces are well worth reading; just keep in mind that there are a lot of ways to look at this election and it is important to not engage in confirmation bias—seeing what you want to see and ignoring any new information that might lead to a contrary conclusion.

This story was originally published on nationaljournal.com on July 13, 2019