The obviously electable candidate

It is obvious to many of us why Sanders lost to Biden, just as it was obvious why Clinton lost to Trump. But we’re still arguing about the lessons of 2016 and I see no end to arguments about the lessons of 2020. Was the intervention of the party establishment the decisive factor? Or was it voter preference for Biden? Was Sanders too “left,” but voters too “moderate”? Could a better campaign have made much difference?  If strategy matters at all, movement activists will need answers to these questions.

When I began my survey of the already massive literature about the campaign, I thought that the Democratic Party establishment had so overpowered the campaign that a different strategy could not have changed the outcome. But this only pushed the question further back: How do you explain why the voters went along with the party’s choice, when polls showed that both Sanders and his signature policies were popular?  Voters wanted Medicare for All, free public college, and a wealth tax, all of them with majority support among all voters, even more among Democrats.

The answer may seem obvious, since everyone seems to agree that the voters chose Biden because he was more “electable.” People talk about “electability” as if it is an inherent attribute of a candidate. Yet in 2016 Clinton was electable and Trump unelectable … until they weren’t. And then there were these two paradoxical developments this year: First, the “unelectable” socialist Sanders became the frontrunner, but then Biden, exposed in the debates as the feeblest of the “moderates,” defeated Sanders in a landslide in one primary after another. Uneasiness about
Biden was reflected in the betting markets — they predicted Biden would be the winner five times, but each time losing to other favorites soon after, according to British political scientist David Runciman.

Only a few days in March

The “electability” of Biden was something created by other people in only a few days in March, just before and just after he won the South Carolina primary 48% to 20%. And within two weeks, his polling tripled. A recap will show how quick — and odd — the reversal was.

Sanders had been the clear front-runner — apparently the most “electable” after winning the popular vote in Iowa, New Hampshire, and Nevada — the only person ever to have won the popular vote in contested races in these first three primaries. In Nevada he won in a landslide,  despite a fierce attack on him and his Medicare for All plan by powerful union leaders. He showed he could win with Medicare for All, and win with Latinx voters; if he could win Black votes in the South, that would clinch the nomination. This was not so unlikely — one Reuters/Ipsos poll showed him ahead of Biden by double digits among registered Democrats and independents, and three points ahead of Biden among Black voters, and Morning Consult had him beating Trump by the same percentage as Biden even among Black voters over 65. The next primary, in South Carolina, would be an important test.

Earlier, the conventional wisdom was that Biden would sew up the Black vote in South Carolina, where Blacks were 61% of Democratic voters in the 2016 primary, and where Clinton had trounced Sanders 86% to 14%. But polls just before the primary surprised observers — Biden had only a “narrow lead” over Sanders.

Then Rep. James E. Clyburn’s endorsement “changed everything for Biden’s campaign.” Clyburn, majority whip and the third ranking member in the House, was a civil rights movement veteran, popular statewide and trusted by Black Democrats. (He also is one of the biggest recipients of pharma money, over $1 million.) Three days after his endorsement, Biden won in a landslide. Over 60% of Black voters said Clyburn’s endorsement was “an important factor” in their decision.

The Umpire Strikes Back!

It looked like the Democratic establishment had finally thrown its support to Biden. What happened, according to observers across the spectrum, is that Sanders had “sent … the Democratic establishment into panic mode” after he won the Nevada caucuses (47% to Biden’s 20%). Moderates “hastened” Clyburn to endorse Biden and “pressured” centrist candidates to leave the race. Then, when Biden won decisively, “the party establishment kicked into full gear.”

Clyburn’s support was a “signal to rally around Biden.”  It brought seven more endorsements before the South Carolina vote, including Virginia’s Sen. Tim Kaine and former governor and DNC chair Terry McAuliffe, although Biden had only a “modest lead” in their state. By the close of voting on Super Tuesday, Biden scored 31 more endorsements  — former Senate Majority Leader Harry Reid, two former DNC chairs and more office holders from the Senate on down.

Party leaders made their approaches, advising the other candidates to withdraw. Pres. Carter had breakfast and a photo opp with Buttigieg in Plains, “top Democratic donors” urged him to drop out and help unite the party, and Pres. Obama counseled the mayor that this was the moment he had “considerable leverage … and should think about how best to use it.” Meanwhile, Harry Reid was “working behind the scenes” to induce Klobuchar and other  “former candidates” to endorse Biden. And so Biden received the high profile withdrawals and endorsements of Buttigieg, Klobuchar, and Beto. Soon to follow was Bloomberg, who was “experiencing intense pressure from multiple sources from inside the Democratic Party.”

Biden had won “the endorsement primary,”one of the key signs of intervention by the party establishment.

“The Party Decides”? The system is rigged? 

After months of panic and indecision, Biden’s win in South Carolina gave party elites the rationale they needed to coalesce around him. Was this a rerun of “the system is rigged,” with the establishment working behind the scenes to throw the election to a centrist? The narrative supports a favored view among political scientists that the party establishment decides on the nominee before the convention. Political scientists Marty Cohen, David Karol, Hans Noel and John Zaller developed this idea in “The Party Decides” (2008). They put aside conventional focus on the candidates as the authors of their own success or failure, looking instead at the role of the party elites in uniting the various party factions around a trusted, electable candidate — and filtering out the untrustworthy and unelectable.

There had been democratizing reforms of the primaries, with the most recent one reducing the power of the unelected “super-delegates” in 2016. But Cohen et al. argued that the establishment had been able to adapt, through exploiting elite networks and resources to privilege a favorite candidate — or to meet threats from a dangerous outsider like Sanders. During the “invisible primary” before the convention, party elites promote candidates through fundraising and endorsements, while exploiting media and influence networks. Biden should have been the frontrunner, given his early dominance in the “endorsement primary,” beginning with Diane Feinstein and Andrew Cuomo as early as January 2019. But donors had doubts about his viability, no doubt for the same reasons debate watchers cringed at his performances, and so he did poorly in the “money primary.” (His fundraising is still worryingly low.)

This time the party united not around a candidate, but around fear of a candidate — fear which rose to “panic” after Sanders’ crushing victory in Nevada. Journalists reported on a Stop Sanders movement active behind the scenes, and we saw a steady stream of stories about worried insiders, warnings and prophecies by “experts,” freakouts by columnists and talking heads, rumors about phone calls and meetings, pressure and promises. The opposition to Sanders was so extreme that influential insiders felt they could even talk to reporters about damaging the party’s chances rather than allowing him the nomination.

“Objective reporting,” bias, and demonization

Much of this coverage is in the traditional “objective” style of reporting — third-person articles stating the “facts,” with the writer avoiding editorializing  but quoting the negative opinions of others. Campaign strategists, party insiders, big donors, academic experts, voters, various opinion leaders can be quoted “factually,” while the articles reenforce negative stereotypes and promote the perception that Sanders is unelectable. One study showed Sanders and Warren receiving far fewer mentions, and most of them negative, in the cable network with the largest progressive audience, MSNBC. Meanwhile, Biden received far more positive media coverage, with  “nearly $72 million in almost completely positive earned media in the days between South Carolina and Super Tuesday.”

It is too soon to see other systematic research, but the anecdotal evidence gives ample support to claims of corporate media bias. So, for example, the New York Times assigned Sydney Ember to cover the Sanders campaign. What more qualified reporter than a former analyst for BlackRock and wife of a senior consultant at Bain Capital, whose father was Bain’s CEO? Critics called her out for quoting political opponents of Sanders as if they were experts or neutral observers, without mentioning their political or corporate ties (Fairness & Accuracy in Reporting’s list is here ).

The Times did better with a column by David Leonhardt, which pointed out the centrist bias of political journalism. He took as example the wealth tax, which both Sanders and Warren make central to their messaging. Leonhardt cites negative slanting, centrist worries, while quoting expert supporters far less than opponents. “For that matter,” he writes, “the complaints of obscure billionaires have gotten more attention than the arguments of sympathetic experts.”

The opinion pieces and cable rants of talking heads go far beyond this more professional, “just-the-facts” form of bias.  Some likened Sanders to Trump — negative and angry authoritarians — with false equivalence, equating egalitarian populism of the left and white supremacist populism of the right. Chris Matthews famously compared Sanders’ Nevada victory to the Nazi invasion of France; he dismissed Sanders’ “democratic, Swedish-style socialism,” imagining executions in Central Park if “the Reds had won the Cold War,” saying “certain persons” would be cheering. James Carville compared Sanders to the UK’s socialist Labour Party leader Jeremy Corbyn and said his nomination would be “the end of days.” “To nominate Sanders would be insane,” wrote Jonathan Chait. A David Brooks op ed had the headline, “No, Not Sanders, Not Ever”; Goldman Sachs CEO Lloyd Blankfein said he might find it harder to vote for Sanders than Trump, a warning heard elsewhere on Wall Street. All of this was in a context made unsurprising by the years of stereotyping.

When a candidate’s ideas are popular, attack his character. Sanders is angry, humorless, an ideologue, intolerant and disrespectful of people who disagree with him, an extremist, a totalitarian, the mean Bernie Bros are a cult and he’s responsible for their behavior, he’s sexist, not a good Jew and even anti-Semitic — and he never accomplished anything in 20 years in Congress. None of these characterizations were given scrutiny. Some have no basis in fact to begin with; others pick out an incident or a comment which is given arguable interpretation and is in itself a poor basis for a grand summation of a person’s character or career.

The power of demonization is that you don’t need evidence to make your case. All you have to do is prime your audience with the “us” and “them” frame and demonstrate that the target is one of “them.” Character attacks that would be questioned about another person need no corroboration. The media plays an important role in political demonization campaigns, simply by covering the news and by giving privileged voice to the accusers. Repetition gives the smears legitimacy; quoting respected sources endow them with credibility, either explicitly endorsing them or “objectively” reporting what others are saying. Years of this since Sanders’ 2016 campaign and yet polls still showed he was popular. But for many these attacks must have strengthened the perception he was not electable.

President Bernie Sanders addressing the Bernie Bros, as seen by the Democratic establishment. (And by Gustave Doré, illustrating the moment in Milton’s Paradise Lost. Engraving, 1886.)

Why the freakout?

The conventional view is that party elites were afraid Sanders’ radical program would turn off voters and drag the party down with him. But the best evidence we have, polling, doesn’t support this fear. Sanders had serious problems in broadening his coalition to include older Blacks and suburban women, and youth were not turning out as hoped. But then Biden faced an enormous challenge also, in gaining the support of the progressive wing of the party, youth, and Latinx voters. It was no wonder that polls showed Sanders doing as well as Biden against Trump, showing his policies more popular and his personal favorability as high or higher.

Voters may identify as moderate or conservative, but many of these same voters like what Sanders stands for. A Times article summed this up in a paradoxical headline: “Progressive Ideas Remain Popular. Progressive Presidential Candidates Are Losing. Why?” The reporter points out “in state after state” voters say they support the signature progressive policies of Sanders and Warren — “single-payer health care, canceling student debt and the Green New Deal,” yet vote for Biden, “who has proposed largely moderate ideas.” “The ideas that Sanders has popularized,” wrote John Nichols in The Nation, “were running better than Sanders himself.”

Immoderate moderates

If anything was more surprising than Sanders’ wins in the first primaries, it was the votes of the moderates in the primaries he lost. Journalists reported that his program was widely popular, but a look at the exit polls shows this popularity in a curious light: Moderate and conservative voters favored Sanders’ “extreme” agenda.

Take Florida, for example, a state which Trump won, with a Republican legislature and governor. Biden crushed Sanders (62%/23%), and the exit polls show 45% of voters identifying as moderate or conservative. Yet a majority favored single payer health insurance (“a government health plan that would replace private insurance,”  55%/33%). Another signature position of progressives: 46% thought the economic system needed “a complete overhaul” — and half of those voted for Biden, only a third for Sanders.

Or Texas.  Biden won 35%/30%. 42% identified as moderate or conservative, but over half the voters had a favorable opinion of socialism,  supported single-payer (63%/34%), and free tuition at public colleges  (76%/20%). The voters for Biden and other “moderate candidates” made up close to half of the voters who wanted single payer (45%) and  free tuition (48%) and who had a favorable opinion of socialism (43%).

If you go through the exit and entrance polls for the other states, you’ll find similar patterns. Just one more: Mississippi. Biden 81%, Sanders 15%; 46% liberal, 54% moderate/conservative; but 60% want single payer, and 3/4 of those voted for Biden.

Is the U.S. really “center-right ideologically”?

Who are those “moderates” and what do they believe? This recent Gallup poll concludes, “The U.S. remained center-right, ideologically, in 2019.”  The poll, based on 29,000 phone interviews, concludes that 37% of Americans are conservative, 35% are moderate — and only 24% are liberal. The poll, like many others about “ideology,” sorted people as liberals, moderates and conservatives, but didn’t ask them about their preference for specific policies. The exit polls do better here: They ask whether voters identify themselves “ideologically,”  but also ask their policy preferences. You get a different sense of the electorate when you ask the specific and not just the abstract question about their politics, and then you find large numbers of “moderates” have some “extreme” preferences.

Polls show how popular Sanders’ policies were, especially among Democrats. FiveThirtyEight reported these surprising results as late as mid-March, when the tide had already turned against Sanders. Medicare for All “enjoyed 63% support among Democrats” (42% nationwide). Free public college and university? 58% nationwide support it, only 39% oppose it. Tax on extreme wealth? 68% nationwide, 83% of Democrats. Most surprising to me, in red-state Florida more than 70% of all the state’s voters “favored replacing private health insurance with a ‘Medicare for all’-type system” (AP VoteCastsurvey).

Our political culture is so polarized that it may seem quite natural to classify voters as liberal, moderate and conservative — politicians describe themselves that way all the time. But it turns out that most people don’t have consistent or coherent ideologies; maybe elites do, but ordinary Americans don’t. Political scientists Donald R. Kinder and Nathan P. Kalmoe marshaled the evidence in their recent book, “Neither Liberal nor Conservative: Ideological Innocence in the American Public.” Only a small slice of the electorate are accurately described by these or other ideological categories, and, as you expect, those voters are people like you, well-educated people who follow politics closely. When most people self-report as liberal or conservative, they may have quite different positions (or none firmly held) on immigration, affirmative action, gun control, health insurance, tax reform and warmaking. Ezra Klein’s summary: “Most voters aren’t ideologues, and even accounting for that, most ideologues aren’t particularly ideological.”

Misunderstanding polarization

How does this square with how polarized our politics and culture have become? Americans are deeply divided, but it’s more accurate to say they are divided into opposing parties, not opposing ideological camps.

Most voters have strong group party identities and hostility to the other party. They choose their party because it’s the party of their family members, their peer group, community, social group, church. If most people around you are Republican or Democrat, maybe you are too. They choose the party first, then adopt the party’s ideological label, without necessarily supporting the party’s signature policies.

How else can you explain how Donald Trump could became the Republican nominee in 2016? He campaigned against the Republican party’s fundamental ideological principles; he came out strongly against free trade, anti-Russian foreign policy, personal morality, small government. He promised health insurance for everyone, he promised to protect Medicare, Medicaid and Social Security, and he promised to tax the rich. He won the nomination with only 44.9% of the vote, and then Republican voters came home to the party in the election and are loyal to the party he remade in his interest. And then his broken  campaign promises didn’t matter.

America is not “moderate” or “centrist” — it is — or the voters are — Democrat or Republican, and on our ballots we don’t have a Liberal Party or a Conservative Party or a Labour Party.

Super Tuesday’s lesson in the power of party loyalty

One consequence of the power of party identity is the influence of endorsements. When the party establishment coalesces behind a candidate, the party’s voters take the cue. When the primary season  began, there were 28 Democratic candidates; but when only a few were left, the party establishment moved fast. We saw how the flood of endorsements by party leaders created momentum for Biden before the South Carolina and Super Tuesday primaries. His “electability” was created then by influential party leaders, together with the media … and then Democratic voters. Such is the “aura of inevitability” a candidate needs to secure the nomination.

Sanders’ supporters are right to celebrate as well as grieve — his campaign, the culmination of years of mass movements, mounted a serious threat to party elites, moving the party to the left and forcing progressive political ideas into the mainstream. But they — we — have to come to terms with Sanders’ failure to win over the people who like his program but voted for Biden. Bernie was right when he said that the corporate establishment, the media establishment, the Democratic establishment were all working against him; he was right to say that his policies actually were not “extreme” and were popular with voters; and he was right to say his campaign won the ideological battle yet failed to convince voters of his electability. But the polls tell us that progressive policies and an enthusiastic movement are not enough.  The party establishment’s real power is the loyalty of its voters.

—Paul Elitzik

Notes:
This article focuses on the role of the party establishment in the campaign, only one part of the explanation of Biden’s success. My next article will survey campaign postmortems, looking at mainstream explanations, but also how Sanders veterans and the left understand the lessons for movement strategy. In particular, I’ll look at explanations of the failure to win Black voters and mobilize youth.

The feature image is from Bernie Sanders’ Facebook page. I was behind the huge “E” at the back of the crowd. Disclosure: I like Bernie and supported his campaign.

The thesis of Kinder and Kalmoe’s “Neither Liberal nor Conservative” was argued in an influential paper by Philip E. Converse in “The nature of belief systems in mass publics” (1964). Converse believed that “ideological innocence” is common in all societies with mass democratic parties. He has an interesting section on Hitler’s rise to power in this paper and in later work studied French parties. There are a number of articles about Kinder and Kalmoe’s book in Vox.com, this article in the NY Times, some pieces by Lee Drutman in various publications.

Ezra Klein has an article on David Broockman’s research on “moderates,” which criticizes the idea that America is “moderate” from another angle. Broockman criticizes polling’s use of the “moderate” category for people with both liberal and conservative views, view that often are not at all moderate, even “extreme.” 

If you don’t want to spend your money on “Neither Liberal nor Conservative,” you can find some papers by Kinder online about this area of their research:  here, here, and here.