So, barring an act of God, it is now certain that the Democratic nominee for president will be either Joe Biden or Bernie Sanders.  Biden is the new front runner, but the race is far from over.  Which of these two men should we support?  To answer that question we need to ask two more: which of these would make a better POTUS, and which of these stands a better chance of beating Donald Trump in November.  The answers to these two questions could well be different.  The answer to the first question depends heavily on one’s values and ideology.  The answer to the second question should be sought by setting aside one’s values and ideology—i.e., without reference to the answer to the first.

I mainly want to deal with the second question today, but I will briefly note that if I could magically pick the next POTUS, I would certainly choose Bernie over Joe.  Bernie has serious shortcomings, but he represents the kind of change that I think the country needs in both domestic and foreign policy.  Biden represents little better than the status quo ante—which of course is far better than the Trumpian status quo, but still sorely deficient.  (The status quo ante, after all, brought us to Trump.)

The tougher question for me—and, in fact the more critical question—is which of these two men has the better chance of beating Trump. It’s the more critical because beating Trump won’t be a pushover, and because four more years of his presidency would be a terrible disaster for American politics and society.

If we simply look at the latest polls, there seems to be little difference in electability between Sanders and Biden.  The latest Real Clear Politics average of polls shows Biden beating Trump by 5.1% in the national popular vote and Sanders winning by 4.5%.  Both men beat Trump by similar margins in the crucial swing states of Michigan and Pennsylvania, and both lose to him rather similarly in Wisconsin.  And so on.

But polls at this point in time are unreliable predictors of the outcome in November.  The fear among many mainstream Democrats is that in a predictably vicious election campaign Republicans would be able to exploit Sanders’s many vulnerabilities and scare away moderate, swing voters that the Democrats need to win.  Those purported vulnerabilities include unpopular policy positions (abolition of private health insurance and decriminalization of illegal border crossings), his “socialist” self-label, and past statements suggesting insufficient disapproval of leftist regimes hostile to the US. The numerous expressions of fear by Congressional Democrats, especially House freshmen, that a Sanders POTUS candidacy would drag down the rest of the ticket need to be taken seriously.  Nicholas Kristoff reports on a dinner meeting with a dozen House Democrats who worried that a Sanders candidacy could hand the House back to the GOP.  House members—especially new ones–generally have a good feel for the sentiments of their constituents: their political survival depends on it. Their concerns, moreover, are consistent with the considerable body of political science research, mostly on legislative elections, indicating that more ideologically extreme candidates do less well electorally than relative centrists.

The Sanders campaign tends to downplay concerns about losing moderate voters and instead bases its claim for electability on Sanders’s expectation that his “political revolution” will so energize young people, working people and minorities that a great influx of enthusiastic new voters will bring victory at the polls.

So, it is said that a Sanders candidacy will scare off moderate voters.  It is also said that it will bring in new voters.  How real are these two possibilities, and how do they net out?

I don’t think there can be any serious doubt that a Sanders candidacy would turn off considerable numbers of voters who would otherwise be inclined to reject Trump. If you don’t know anybody who dislikes Trump but is also very skittish about Sanders then I would tend to assume that either you don’t talk much about politics or you are living in some kind of bubble.  Consider that nearly 3 million Hillary Clinton voters in 2016 had voted for George Romney four years earlier. (Coincidentally, that number is roughly equal to her popular vote margin over Trump).  It’s reasonable to assume that those Romney/Clinton voters were mostly moderate to conservative people.  Many of them must have held their noses voting for the centrist liberal Clinton only because they disliked Trump even more.  How many of them would hold their noses and vote for Sanders?

In a massive recent study of some 40,000 adults, political scientists Broockman & Kalla found that a Sanders candidacy would indeed lose voters who would otherwise vote for a more moderate Democrat.  To make up for that loss, Sanders would need to achieve an unprecedented increase in turnout—a greater increase in youth turnout than the increase in black turnout that Barack Obama’s historic candidacy produced in 2008. Broockman & Kalla suggest that this prospect seems unlikely.  Its unlikelihood is in fact confirmed by the results of the 2020 primary campaign thus far. The great increase in youth turnout for Sanders simply hasn’t materialized, as Sanders’s campaign has acknowledged with disappointment.  Where turnout has increased, it has benefitted Sanders’s opponents, especially Biden on super-Tuesday.

A stronger argument for Bernie’s electability would cite the third party vote. In 2016, the leftist Green Party candidate, Jill Stein, received 1.5 million votes. Her vote totals in Michigan,  Wisconsin and Pennsylvania exceeded Trump’s margins of victory in those states.  It is reasonable to assume that most Stein voters would go for Bernie, going a long way to closing the gaps in those crucial swing states. On the other hand, surely some of those Stein voters would also vote for Biden, having faced the ugly reality of Trump for four years.

Still another argument for Bernie’s electability comes out of the work of political scientist Rachel Bitcofer.  Bitcofer gained an overnight reputation as a forecasting whiz when she was practically alone in correctly predicting a 40+ gain for the Democrats in the 2018 elections for the House. Her model for 2020 predicts a win for the Democratic presidential nominee no matter who he is. It assumes that in this polarized era, swing voters are practically non-existent and there are more Democratic and Dem-leaning voters than on the other side, motivated by antipathy to Trump.  Needless to say, I hope she’s right.  (Her model was less successful in predicting results in Senate and gubernatorial races.)

In sum, I don’t think the evidence is clear enough to justify the more extreme cries of suicidal doom if Sanders became the Democratic nominee.  I do think Sanders would have a reasonable chance of beating Trump.  But as of now I think Biden’s chances would be better. Some Sanders supporters will undoubtedly react with incredulity, if not scorn, to this conclusion. Just look at the wildly enthusiastic, youthful crowds that greet Bernie everywhere! The millions and millions of small donors! How can you not feel the Bern? But I think Bernie’s movement—his revolution—is still a relatively small, intense subset of the population.  It’s a deceptive guide to his potential support among the broader electorate.  (I hate to compare Trump and Bernie in any way, but I’m sure that there are millions of Trump fans who cannot imagine that their idol is actually highly unpopular.)

This is not an endorsement of Biden.  There is still time for Bernie to demonstrate that he can ride to victory on waves of new voters.  I’ll be looking in particular to the primaries in Michigan, Wisconsin and Pennsylvania. The results thus far don’t augur well for the revolution, but who knows—there could be more surprises for us in this surprising primary season.