The Unfavorable Favorite

Notably, the case that Clinton has the best chance to win in November does not seem to depend much on Clinton herself. This is no coincidence: by a number of measures, she profiles as a comparatively weak general election candidate.

According to national polls, nearly 53 percent of Americans have an unfavorable impression of Clinton, which would make her the most disliked presidential nominee in modern history. Even if incumbents are included, the only candidate with worse numbers was Jimmy Carter in 1980.

Public perception of Clinton has been shaped by the intense sexist and right-wing attacks that she has endured since the 1990s. These polarizing assaults, along with Clinton’s own partisan record, have helped make her very popular with loyal Democrats, but unpopular with everyone else.

The problem is that the loyal Democratic vote is simply not enough to win a general election. In 2012, Democrats made up only 38 percent of the general electorate, while registered independents accounted for 29 percent. On his way to defeating Mitt Romney, Barack Obama won almost half of them.

Clinton’s appeal among these non-Democratic voters is extremely limited. Just 29 percent of independents hold a favorable view of her, according to an average of three YouGov surveys taken since January; over 61 percent view her unfavorably. In the most recent poll, Clinton’s count was 24 to 67, with 50 percent saying they hold a “very unfavorable” opinion. These are numbers that should make even Supreme Court-first liberals feel skittish.

It’s too soon to conclude that Clinton’s historic unfavorability will spell defeat in November. Yet as Nate Silver noted with regard to Mitt Romney’s (less pronounced) unpopularity in April 2012, we should not dismiss these early numbers either. At the very least, they make it plain that Clinton faces an image deficit greater than any challenger in recent memory, including landslide losers like Walter Mondale, Michael Dukakis, Bob Dole, and John McCain.

McGoverns of the Mind

Generally, however, the “electability” argument skips past Clinton and concentrates on Sanders. And here the case against Sanders divides into three general paths — one, guided by historical analogy; another, driven by pundit fears and fantasies; and a third, oriented around voter ideology and demographics. None are persuasive.

The most common way to dismiss Sanders is to lump him in with previous progressives battered by conservatives in general elections — usually Mondale in 1984 or George McGovern in 1972. “The early enthusiasm for Sanders reminds me of the McGovern and Mondale races, where two good men were only able to win one state each in their presidential campaigns,” former Louisiana senator John Breaux told the New York Times in January.

The logic of this analogy turns on the idea that McGovern and Mondale both lost for the simple reason that they were too liberal for American voters. The first rebuttal is almost too obvious to spell out: the 2016 electorate looks nothing like the 1972 or 1984 electorate — quite literally, it is a different set of people.

A healthy majority of voters this year were not eligible to vote in 1984; almost half of them weren’t even alive in 1972. People old enough to have cast ballots against McGovern will probably make up no more than 20 percent of the electorate in 2016. These are very old historical parallels.

Very old, and very lazy. As Daniel Denvir has written, the combination of factors that produced the McGovern disaster bears almost no resemblance to the political situation today. In 1972 the Democratic Party was in a state of flux. McGovern captured the nomination with about 25 percent of the primary vote; over 23 percent went to the Alabama white supremacist George Wallace. Major party leaders like AFL-CIO boss George Meany, meanwhile, refused to support McGovern in the general election against Nixon.

Today both major parties are far more ideologically unified and more polarized. Although the Democratic Party elite has so far shunned Sanders, he is almost as popular as Clinton among the party’s rank and file. If Sanders wins a clean majority of the primary vote, it’s hard to imagine any significant chunk of the Democratic coalition abandoning him in a general election against the Republicans.

But the historical analogies miss the mark for an even more fundamental reason. McGovern and Mondale did not lose because they were too liberal, but above all because they faced Richard Nixon and Ronald Reagan, popular incumbents presiding over economic booms.

The 1972 and 1984 blowout losses conform closely to electoral models that measure vote totals based on underlying economic conditions, without taking any account of candidate identity or ideology. The Democrats were doomed no matter who they nominated.

If the primary race continues to tighten, Clinton supporters will no doubt continue to spook Democrats with the fatal visions of 1972 and 1984. These are but McGoverns of the mind, false creations conjured by elite pundits and party officials. They offer no actual evidence that can be applied to the 2016 general election.

Pundit Horsemen of the Apocalypse

Across the primary season, Sanders himself has rebuffed “electability” arguments by pointing to poll results. In hypothetical matchups against the three leading Republicans (Donald Trump, Ted Cruz, and Marco Rubio) he beats them all soundly, and polls better than Clinton in every case.

We may be skeptical about the predictive power of these findings, nine months before Election Day. But it’s wrong to call them “absolutely worthless,” as one political scientist told Vox last week.

In a comprehensive analysis of elections between 1952 and 2008, Robert Erikson and Christopher Wleizen found that matchup polls as early as April have generally produced results close to the outcome in November.

Even much earlier “trial heats” seem to be far from meaningless. As partisan polarization has increased over the last three decades, there’s some evidence that early polling has become more predictive than ever. In all five elections since 1996, February matchup polls yielded average results within two points of the final outcome.

In other words, if Sanders is the nominee, his nine-point lead over Cruz is probably safe. Even his four-point lead over Rubio likely reflects a meaningful advantage.

Of course, mainstream pundits are quick to dismiss these numbers. Once the Republican attack brigade moves into action, they argue, talk of socialism and taxes and the Iranian Revolution will fill the air like fire and brimstone; poor Bernie will quiver and flop before the vast counterrevolutionary horde.

Occasionally, these liberal concerns seem to resemble pure reactionary giddiness. But that’s not even the real problem. The pundit horsemen of the apocalypse have constructed an argument that has no relationship to actual evidence, either in the current polling or recent history.

Sanders is far from an obscure or unknown figure. Across sixteen national surveys since New Year’s Day, an average of 85 percent of Americans knew enough about Sanders to form an opinion of him.

This is not the profile of a candidate flying under the radar. John Kasich, who fits that description, elicited an opinion from just 53 percent of respondents in the eleven surveys that asked about him. More Americans have a decided view of Sanders than either Rubio (77 percent) or Cruz (81 percent).

That view is strongly positive. Sanders’s favorability ratio of 51 percent positive to 38 percent negative is the best of any candidate in the race, by far. His favorability with independent voters is also much higher than any of his rivals, including Clinton, Trump or Rubio.

There is simply no historical precedent for a major party nominee as popular and well-known as Sanders collapsing in a general election.

Some gleefully apocalyptic liberals have likened Sanders to Michael Dukakis, who held an early polling lead over George H. W. Bush before ultimately losing by a large margin in 1988. Yet the comparison falls apart before it begins.

A stiff technocrat, Dukakis won the Democratic primary not by packing arenas with passionate supporters, but chiefly by having more impulse control than Gary Hart and being whiter than Jesse Jackson. And his early polling strength was clearly a mirage, as contemporaries noted: only 52 percent of voters even had an opinion of him in May 1988. Dukakis was John Kasich, not Bernie Sanders.

None of this means we should expect to see the end of the “wait until the Republicans get him” argument any time soon. It’s a staple of the Clinton primary arsenal.

In February 2008, then-Clinton chief adviser Mark Penn discounted early polls that showed Barack Obama performing well in a general election: “Sen. Obama has never faced a credible Republican opponent or the Republican attack machine, so voters are taking a chance that his current poll numbers will hold up after the Republicans get going.”

Penn was writing about an African-American candidate whose middle name was Hussein, and who had spent much of his childhood in a Muslim country. A month later, a video surfaced showing Obama’s longtime pastor Jeremiah Wright saying, “God damn America!”

Did the vaunted “Republican attack machine” fail to take advantage of what Vox’s David Roberts might have called Obama’s “glaring vulnerabilities”? Of course not: an independent super PAC spent $2.5 million on Wright-themed advertising in swing states. None of it had much impact on Obama’s poll numbers against John McCain, which rose considerably between February and Election Day.

The attacks aimed at Obama may have reinforced his unpopularity with right-wingers, but they did little to dent his appeal among Democrats and independent voters. Why does anybody believe that red-baiting will succeed where racist innuendo failed? When the Berlin Wall came down twenty-seven years ago, today’s median voter was not old enough to drink alcohol.

In addition, this year’s polls show little sign of an electorate ready to abandon Sanders at first exposure to right-wing talking points.

Only 35 percent of Virginia independents said they would be less likely to vote for a “Democratic-Socialist” candidate. And when a conservative push-poll asked Nevada Democrats and independents how they felt about Sanders’s plans to spend “$15 trillion dollars” for “a government run health care program,” 53 percent replied that it made them more likely to support him.

Another and even less persuasive claim is that Clinton, unlike Sanders, has already withstood every right-wing attack she can possibly face. Mark Penn also made this point in 2008 — and today Clinton’s unfavorability is even higher than it was then.

Sanders has famously refused to discuss Clinton’s emails. He has denounced her Wall Street speaking fees, but has largely refrained from discussing the much larger hoard of cash connected to the Clinton Foundation — an area that Republicans seem eager to exploit. That same Nevada push-poll showed that 64 percent of Democrats and independents were less likely to back Clinton after learning about “foreign donations” given to the foundation while she was secretary of state.

Republican candidates have already made various scattered attacks on these subjects, but we’ve seen nothing like the tornado of Clinton scandal-mania likely to follow if Hillary is nominated.

Moderates for Socialism

A final line of argument, exemplified by Ruth Marcus in the Washington Post, insists that Sanders’s platform is simply too left-wing for a “moderate” American electorate. Usually this is trotted out amid broad national surveys that find the country divided between ostensibly coherent blocs of “liberal,” “moderate,” and “conservative” voters.

But as political scientists Shawn Treier and D. Sunshine Hillygus have argued, two-dimensional surveys of voter ideology do not provide a useful guide to the American electorate. To the great disappointment of the Post editorial board, many self-identified “moderates” are not sober Beltway centrists but in fact “cross-pressured” by a mix of strong liberal and conservative beliefs.

The unstable and multidimensional identity of the “moderate” voter helps explain why Sanders’s own polling numbers have regularly confounded the prejudices of pundits. In New Hampshire, for instance, where experts repeatedly stressed his strength with “liberals,” Sanders actually did even better with “moderate/conservative” voters.

It might also help explain why Sanders polls well in places like Nevada and Alaska — states not known as liberal bastions, but home to a large number of independent voters.

Who are these independents and “moderates” voting for Sanders? It seems reasonable to believe that they are not confused centrists, but “cross-pressured” voters with a wide range of views, all drawn to Sanders’s left-wing economic message. In fact, Sanders has a long record of winning over these kind of populist “moderates.”

Consider Caledonia County in Vermont’s rugged Northeast Kingdom. Contrary to media cliché, not all of Vermont was a liberal paradise in the 1980s. Caledonia County twice voted heavily for Reagan; in 1988, Bush crushed Dukakis there, 61 to 38 percent.

Yet two years later, when Sanders won his first statewide election for Congress, he defeated the incumbent Republican in Caledonia County by eleven points. Over the next decade, Sanders ran well ahead of the centrist New Democrats Bill Clinton and Al Gore — in 2000, the same year George W. Bush carried the county by seven points, Sanders won it with 66 percent of the vote.

You can chalk some of this up to the quirkiness of rural Vermont. But as the primary campaign has unfolded, Sanders has shown an undeniable ability to connect with the same kind of lower-income and less-well-educated white voters all over the country, from Iowa to West Virginia to Oklahoma.

Democrats have been slowly losing these voters to Republicans since the 1970s; in the last decade, they have almost abandoned them entirely. But non-college-educated whites still represent over 40 percent of the electorate in key swing states like Ohio, Wisconsin, and Indiana.

Many of these poor and struggling voters — however “moderate” according to Gallup — seem very receptive to Sanders’s call for universal health care and a living wage. A Sanders campaign that made deep inroads with working-class whites across the Midwest would be well-prepared to defeat a Republican in November.

It’s difficult to find an equivalent category of voters where Clinton might outperform Sanders in a general election. Women? Clinton’s most recent favorability ratio with all women voters is strongly negative: 41 to 54 percent. Sanders’s mark stands at 44 to 41 percent. In a general election, those numbers might shift — but would it be enough to give Clinton a significant advantage?

Clinton’s strongest support in the primary campaign seems to come from the most loyal Democrats, including African-Americans. But in a bitter campaign against an ethnic nationalist like Trump or a right-wing Republican like Rubio, would loyal party voters refuse to turn out for the Democrats, just because Sanders rather than Clinton was the nominee? It doesn’t seem likely.

None of this is to suggest that Sanders should take loyal non-white Democratic votes for granted. That is exactly what Clinton-style New Democrats did when they pivoted to the center in the 1980s. In a general election campaign, Sanders would have to do the opposite, and build a populist coalition that depended on solidarity between black, Latino, Asian, and white working-class voters.

Unquestionably, it would be difficult work. But the opposition of an ever-more-reactionary Republican Party would surely help. And a successful left-of-center coalition would be well positioned — in both ideological and electoral terms — to mount the much larger, long-term struggle necessary to achieve even Sanders’s social-democratic goals.

Of course, it’s impossible to predict the particular contours of a general election campaign featuring either Sanders or Clinton. Much depends on the Republican nominee, and also, perhaps, on the exact proportion of narcissism and pragmatism in the mind of a certain Manhattan billionaire.

But there’s no question that Bernie Sanders can win in November — and there is good reason to believe he would actually be a stronger Democratic candidate than Hillary Clinton.

Last April in front of the Capitol, when the skeptical reporter asked Sanders if he really intended to contend for the nomination, he replied indignantly: “We’re in this race to win!”

Sanders then continued, insisting that it’s impossible to separate the question of “electability” from the question of democracy. “If you try to put together a movement which says, we have got to stand together as a people,  and say that . . . our country belongs to all of us, and not the billionaire class  — that’s not raising an issue, that is winning elections. That’s where the American people are.”

Far more than the elite media imagined, that’s where the American people have been, all campaign long. They’ll still be there in November.