Bernie Sanders is the Most Electable Candidate
Or: Why Conventional Wisdom is Dead
[The opinions expressed here are solely my own and do not express the views or opinions of my employers, past or present.]
Introduction
The Iowa Caucus, which marks the true beginning of the 2020 Democratic Presidential Primary, is in less than two weeks. An excessive 29 candidates ran for the nomination at some point or another, but we are now mercifully down to a final 12. I’ve liked a few of the candidates who have run so far, but after roughly a year of watching the field make their cases, I keep coming back to Bernie Sanders. When I first sat down to write this article, it was going to be an explanation of why I’m choosing Sanders over my second place choice of Elizabeth Warren. But I realized that many Sanders supporters have already argued the case for why he has the best views and record in this race. Independent of that argument, there’s a related but less commonly-stated case which seems just as important to point out the truth of: Bernie Sanders is the most likely candidate to defeat Donald Trump in a general election.
Polling tells us that Sanders is well liked among Democratic primary voters, but many of those who like him are concerned that he cannot realistically win the Presidency. The urgency of removing Donald Trump from office has placed a major emphasis on the trait of “electability” in this primary, and much of the electorate has made what they view as the risk-adverse choice to support candidates that have the strongest chances of beating Donald Trump, regardless of their views. When polls pose the two options against one another, the majority of Democratic voters would prefer a candidate better able to beat Trump that doesn’t share their views than a candidate less able to beat Trump who does share their views.
This likely explains the consistency of Joe Biden’s first place lead in the primary. The Biden campaign’s primary argument is that he’s the pragmatic choice for defeating Donald Trump- after all, he is a widely-known figure with inoffensive views and has eight years of executive experience under his belt. All conventional wisdom points towards the idea that Biden is as “electable” as it gets. The only problem is that conventional wisdom is dead.
2016 was the perfect natural experiment to test whether the rulebook for American campaign politics still holds up. On the Democratic side stood a woman with lots of legislative and executive experience, a massive fundraising network, the institutional support of her party, and a set of focus-group-tested positions that fell very comfortably within regular American politics. On the Republican side stood one of the most unorthodox and inexperienced candidates in history; a wealthy elitist known for his reality TV show, shady business practices, disgusting treatment of women, and colossal ego; a man who was hated by much of the establishment of his own party; and a man who spewed a noxious combination of nationalism, racism, and right-wing populism in the most brash and vitriolic fashion possible. And when November came, the latter won.
Donald Trump’s victory shocked the American political class. Politicians, consultants, campaigners, pundits, and academics with a wide assortment of views had to face what could only be described as a direct assault on how a Presidential election was supposed to work. How one diagnosed “what went wrong” in 2016 is a major factor in what one decided needed to be done about it. Many rightly observed the role that anti-democratic institutions and practices like the Electoral College and voter suppression played in this outcome, and they turned their focus towards democratic reform. Others believed the result could be attributed directly to Russian interference, and they turned their focus towards election security. Those who viewed misogyny as the issue turned to feminist activism and education. And on, and on, and on.
I think many of these ideas hold explanatory weight to varying degrees, and the work being done in the above examples is all important and worthwhile. After all, it was only 78,000 votes in three Midwestern states which earned Donald Trump his victory, and with a margin that slim many different factors could have been enough to change the outcome.
What is worrying is that all of these explanations also seem to miss something important. Not only does each explanation ignore the deep problems with the Clinton campaign’s strategy, but even more importantly each seems to start with the assumption that Trump’s victory was in some sense a fluke, the result of a bug in a predominantly functional democratic system which can be fixed in order to return to “normal.” In my view, to hold this opinion is to have learned nothing at all from 2016. Donald Trump’s victory wasn’t a random error, but was instead enabled by long-coming shifts in the American political landscape and the tone deaf failure of the entire political establishment to recognize and adapt to these changes. It’s different this time, and even any successful attempt to go back to “normal” will still fail in the long-term, as it was the very state of pre-Trump “normalcy” which made his rise possible.
Entire chapters of the rulebook for political campaign professionals are now obsolete, and the willingness of said professionals to acknowledge this appears minimal. For this reason, I am unconvinced by the story of Joe Biden’s “electability.”
When I excitedly voted for Bernie Sanders in the 2016 primary, I did so quite sure he would lose to Hillary Clinton (even I was surprised how well he ultimately did). Regardless, I was never settled at the time on the issue of whether he had a better chance at beating Trump than she did, and despite my complaints I voted for her in the general election. In the years since then, I have realized that I was wrong about a lot of things. But the more I have read, discussed, thought, and worked, the more I have realized that the thing I’d been wrong about was Bernie Sanders. He isn’t just the best candidate for the job, a pie-in-the-sky candidate for progressives. A hard-headed look at the facts tells us that he is the single most realistic and pragmatic candidate in the 2020 Democratic primary.
When the Median Voter Theorem Is Wrong
It seems important to start with the fact that most Democratic candidates have a fair chance at beating Trump. At the time of writing Trump has a lower net approval rating than every President since Truman did at this point in their tenure, including a net disapproval in all of the Midwestern states which granted him victory in 2016. Though the margins aren’t huge, most of the leading Democratic candidates come out ahead in head-to-head polling. And despite popular belief, election polling is still relatively accurate. Democratic voters are anxious about assuming they have an advantage over Trump after the shock of an unexpected defeat last time around, but even with an economy at the peak of its cycle Trump has a lot of electoral liabilities. But for those who don’t want to leave anything up to chance, the case for Sanders is especially strong.
Before building the case for Sanders specifically, I’ll start by addressing the single most common concern about his electability: that he is “too far left” to win. In one of many examples, Obama’s 2012 campaign manager Jim Messina recently stated:
If I were a campaign manager for Donald Trump and I look at the field, I would very much want to run against Bernie Sanders. I think the contrast is the best. He can say, ‘I’m a business guy, the economy’s good and this guy’s a socialist.’ I think that contrast for Trump is likely one that he’d be excited about in a way that he wouldn’t be as excited about Biden or potentially Mayor Pete or some of the more Midwestern moderate candidates.
This argument is an article of faith among pundits. But if there is one single lesson to take from 2016, it is that the traditional campaign strategies and political analyses of America’s pundits and consultants are no longer in touch with actual political reality. In the Spring of 2016, Messina also said he “prayed every night for a year for Trump to win the R[epublican] primary” because of how easy he thought he would be to beat. Indeed, the 2016 Clinton campaign intentionally promoted Donald Trump and other far-right candidates during the Republican primary in hopes that their extremism and controversial personalities would make for an easy opponent. We can see how that worked out. For his part, Messina went on to work for the failed campaign against Brexit.
The “too far left” argument has its roots in a concept known as the median voter theorem (MVT). MVT assumes that voters are normally distributed across an ideological spectrum ranging from left to right, meaning the number of voters is highest near the center and declines as you go towards either fringe. Voters in the center can be swayed by either a left-wing or right-wing candidate, while voters on the ends have no alternative consistent with their beliefs other than abstention or voting for the candidate who falls closest to their end. In this situation, it makes the most sense for politicians to try and position themselves as closely as possible to where the average (median) voter lies on this distribution, thus winning most of the voters on their end of the spectrum and the voters in the middle who are closer to their position than their opponent’s.
This has been a long-standing principle that is quite useful in the many cases where it applies. The problem is that, like any theoretical model, MVT rests on a set of assumptions, and those assumptions aren’t true in our present circumstances. Donald Trump’s defeat of Hillary Clinton serves as a clear indication that we should call the idea of moderation winning elections into question. Something is missing.
There are two separate reasons the median voter theorem holds limited analytical power in the 2020 election, and thus that the idea of “too far left” holds much weight as an argument against Sanders’ electability.
The first criticism has always been a weak point of MVT: it presents an oversimplified model of ideology, and many voters don’t have a coherent ideology they can plot on it. Anthony Downs, the political scientist who most famously laid out the median voter theorem, explicitly stated among his assumptions that “[t]he political parties in any society can be ordered from left to right in a manner agreed upon by all voters.” Research, however, has repeatedly found that this is not the case. Major works of political science published as early as 1964 found that a “maximum” of half of the population could correctly associate the terms “liberal” and “conservative” with both their true meanings and the parties associated with them.
The trickiness of trying to apply clean and clear definitions of ideology to a population which doesn’t adhere to them calls the entirety of MVT into question. One influential paper from 2009 finds:
Although political rhetoric today is clearly organized by a single ideological dimension, we find that the belief systems of the mass public remain multidimensional, with many in the electorate holding liberal preferences on one dimension and conservative preferences on another. These crosspressured individuals tend to self-identify as moderate (or say “Don’t Know”) in response to the standard liberal-conservative scale, thereby jeopardizing the validity of this commonly used measure. Our analysis further shows that failing to account for the multidimensional nature of ideological preferences can produce inaccurate predictions about the voting behavior of the American public.
MVT’s very foundations are challenged by the fact that “moderates”- those who consistently hold views close to the center- are quite rare. When measuring both economic and social views, the study quoted above finds that at most “35 percent of self-identified moderates hold centrist positions on both policy dimensions.” In fact, an early version of another study found that “moderates” were (non-significantly) more likely to hold at least one extreme position than non-moderate voters, the only difference being that they aren’t as consistently ideological.
The majority of voters we call “moderates” are those with incoherent viewpoints, contradictory viewpoints, or ideologies that simply don’t fit neatly onto a liberal-conservative spectrum. Many “moderates” hold extreme right-wing views on some issues and extreme left-wing views on others, a combination which places them in the middle when one aggregates each viewpoint and flattens ideology out to one dimension. As you can see in the diagram to the left, a large portion of American voters are socially conservative but economically liberal. Even the most extreme of those voters are counted as “moderates” in traditional measures.
Another way to understand this is to look at it issue-by-issue. Another study finds that the most common position taken on an issue in their sample was rarely the one in the center. Of 13 issue areas measured, none perfectly resembled a normal distribution, and many come nowhere close. Based on these particular findings, MVT would actually tell politicians to take a very liberal position on taxes and a rather conservative position on immigration.
Messina argued against the idea that Sanders “can turn out tough-to-turn-out voters” because “most aren’t [progressive]… they tend to be more moderate.” In actuality, they’re neither. Tough-to-turn-out voters, including most “moderates,” hold a wide scattershot of different viewpoints not easily captured in ideological labels.
You could say that all this means median voter theorem still does technically apply, but that it functions based on what Americans think are moderate views. After all, Trump’s own ideological incoherency meant that many voters saw him as more moderate than usual for Republicans. But if that’s the case, the MVT loses much use in predicting electoral behavior, and instead just tells us that what happens depends on how voters see themselves and the candidates in any given election (duh).
But let’s say you don’t buy this critique, and that you still think MVT is a good rule of thumb for US elections. This leads to the second critique, which is that voter ideology is no longer normally distributed around the center. For a long period in American politics, it was the case that the largest group of voters were in the center, the next largest groups were on the center-left and center-right, and so on. But research tells us this is becoming less true over time.
As political polarization and party sorting grows, the median voter from each party is growing farther away from the center, leaving a growing gap where “moderates” once were. The implications of this are profound: this hole means that politicians have less and less incentive to appeal to the “middle,” as there are fewer and fewer people who are actually there.
Take a look at the diagrams to the left. Anthony Downs states that MVT holds “if voters are distributed along the scale as shown in Figure 1… But, if the distribution is like that shown in Figure 2, the two parties diverge toward the extremes rather than converge on the center. Each gains more votes by moving toward a radical position than it loses in the center.”
Looking at the distribution diagram above illustrating how the ideology of American voters has been gradually polarizing and sorting in recent decades, you can see that we’re now in a situation where neither of these predictions holds. With two separate peaks each close to but not quite at the center, the incentive structure is for each party to balance between outreach to the middle and to maintain the enthusiasm of their base closer to the end.
So what does this all actually mean? The median voter theorem, which tells us that parties should pivot towards the center to win, has always been undermined by the fact that the “center” is rarely actually moderate; more recently, it has also run into the problem that the actual number of people in that “center” is shrinking. The always-incisive Eric Levitz puts it more bluntly: “Democrats Can Abandon the Center — Because the Center Doesn’t Exist.” In the United States in 2020, ideology by itself can’t actually tell us much about a candidate’s prospects. A better predictor of how a candidate will do in a Presidential election is policy and popularity. These are places where Bernie Sanders excels.
Why Clinton Lost
Many in the Democratic Party fear that running a candidate on the left could result in a rerun of George McGovern’s disastrous 1972 presidential campaign. But historically speaking the party has had more of the opposite problem: running candidates who lose due to their blandness, unpopularity, and lack of inspiring platform (Hubert Humphrey in 1968, Walter Mondale in 1984, and more controversially, Hillary Clinton in 2016).
One of the biggest culprits for Clinton’s loss was low turnout. The number of registered voters who didn’t vote in 2016 is larger than either Clinton or Trump’s vote total. Despite the population growth that had occurred in the time since, Clinton’s popular vote victory saw less people vote for her than voted for Obama in both 2008 and 2012. In Wisconsin, Donald Trump won with a lower number of votes than Mitt Romney lost with- while he got 2,000 less votes than Romney, Clinton got 238,000 less than Obama. The difference between those two numbers was enough to turn the state from blue to red.
This is partially because Clinton and Trump were both historically unpopular. In Wisconsin specifically, much blame can also be assigned to the Clinton campaign for their failure to advertise and campaign there until the last minute. But at least as much blame can be placed on what the campaign was trying to advertise.
The Clinton campaign believed they could win by attracting moderate and center-right suburban voters who disliked Trump on a personal level. This has been a general strategy of the Democratic Party in recent years, with Senate Minority Leader Chuck Schumer remarking that “[f]or every blue-collar Democrat we lose in western Pennsylvania, we will pick up two moderate Republicans in the suburbs in Philadelphia, and you can repeat that in Ohio and Illinois and Wisconsin.” But the Clinton campaign was so confident that they could play this strategy offensively they even told supporters who were planning to campaign in Michigan that they weren’t needed there, but were instead needed in the more Republican state of Iowa “to fool Donald Trump into competing there.” After all, the Clinton campaign’s internal model showed them winning Michigan by 5 points even on the day of the election. In the end, she lost Iowa by 9.4 points and Michigan by 0.2 points. One DNC consultant remarked that the Clinton campaign “believed they were more experienced, which they were. They believed they were smarter, which they weren’t. They believed they had better information, which they didn’t.”
The Clinton campaign adhered to traditional campaign strategy when it was unwise to do so, and only strayed from tradition when it was also unwise to do so. The most important example of this is that her campaign messaging often didn’t discuss policy issues at all in order to focus on Donald Trump’s character and unfitness to serve. One of the most insightful studies published about the 2016 campaign found that:
In total, over 60 percent of ads supporting Clinton were solely about candidate characteristics, while only about 25 percent were focused on policy. This is a huge difference from Trump’s advertising, over 70 percent of which was focused on policy, and it is a huge difference from every other presidential campaign for which we have comparable data… In a typical campaign, ads that focus on candidate character have comprised less than 20 percent of total ad airings…
For all of the talk of the unusual advertising campaign that Trump ran in 2016, his message strategy was more traditionally policy-focused. Ironically, it was the Clinton campaign that deviated sharply from the conventional playbook when it came to messaging despite following conventional norms in terms of volume, placement and targeting of ads.
The results of election night 2016 show us that this decision was a disastrous one. Any center-right independents or moderate Republicans won over by the Clinton campaign’s personality-focused strategy were easily drowned out by other voters traditionally willing to support a Democratic candidate who simply weren’t presented with exciting policy arguments for doing so. It is unfortunately rather common to see Democratic pundits angrily castigate voters for failing to turnout for Clinton, or to vote for third-party candidates instead. But this gets the obligation backwards: it is the party’s job to convince voters, not the voters’ job to please parties.
Why Bernie Can Win
If the Democratic party wants to win in 2020, they will have to work to convince people to vote for them, both among their base and among undecided independents. The case that Bernie Sanders will energize the Democratic base is easy. Though there are some vocal party elites and dedicated Clinton supporters who dislike Sanders, he’s not actually very controversial among party supporters: he has a 51 point net favorability among Democratic voters (the highest of anyone in the entire field) and is in first among liberals. Combined with his nearly-fanatical organizing base, an undeniable ability to attract massive crowds, and a popular policy platform, Sanders can certainly ensure high turnout among the Democratic base. The entire “electability” debate in this primary ultimately comes down to this: who will be able to reach out to the population of non-voters who ultimately decided the 2016 election? To get an idea of this, let’s look to see who these people are.
Compared to Democratic voters, a 2019 New York Times-Sienna College poll found that Democratic-leaning nonvoters were substantially younger, lower-income, less white (slightly more black and substantially more Latino), and less educated. When looking at all non-voters, the exact same trends were identified by the Pew Research Center in 2018. Similar trends appear in a 2018 USA Today-Suffolk University poll of unlikely and unregistered voters, though that poll seems to over-represent whites and the wealthy.
The NYT-Sienna poll looked at Democratic-leaning nonvoters in six battleground states that wound up voting for Donald Trump. Though each head-to-head has a significant proportion of “unsure” voters, it found Warren 1 point down against Trump, Biden up 3 points against him, and Bernie up 4 points. As we’d expect from my discussion above, these voters as a group seem quite inconsistent in their ideology, though they appear more culturally conservative and slightly more economically progressive than Democratic voters overall. They’re less likely than Democrats to want a President with a bold progressive agenda, but slightly more likely than Democrats to want a liberal. They’re substantially less supportive of immigration and an assault weapons ban, but more supportive of single-payer healthcare. For their muddled and mixed views though, they are mostly just frustrated. 51% of these nonvoters want “a Democrat who will fundamentally change America,” nine points more than Democratic voters.
This frustration is a common trend. The USA Today-Suffolk poll that looked at unlikely and unregistered voters gave a list of answers for respondents to agree or disagree with regarding why they don’t pay attention to politics. 67.6% agreed that part of the reason was because they saw politics as corrupt, making it the single most commonly agreed-to answer. These non-voters had unfavorable views of Hillary Clinton and Donald Trump, but had a net favorability of 24.3 points for Joe Biden and a net favorability of 15.4 points for Bernie Sanders. When given a list of 12 figures that they “would be certain to go to the polls to vote for if that person would decide to run for President,” Bernie Sanders came in first.
Thus, the type of non-voters that a Democratic candidate will have to win over are disproportionately young, low-to-middle income, lacking college degrees, Latino (and to a lesser extent black), frustrated with the political process, and an ideological mixed bag. Winning the turnout of these marginal voters will likely be the difference between a Democratic victory and four more years of Donald Trump.
Who is polling in first among young people? Bernie Sanders, with more support among generation Z than the next four candidates combined and more support among millennials than the next two combined. Who is polling first among low-to-middle income people? Bernie Sanders. Voters without college degrees? Biden, with Sanders in second. Latino and black voters? Bernie is in first and second, accordingly. Voters frustrated with the political process? I don’t know of any polling on this, but it seems a safe assumption that the anti-establishment independent in this race would be doing pretty well.
Head-to-head polling consistently finds Bernie beating Trump in Michigan, Ohio, Pennsylvania, and Wisconsin (not to mention North Carolina). My case is complicated by the fact that his current margins of victory in many of these states are smaller than Joe Biden’s, but there is strong reason to believe this wouldn’t hold in the general. Biden- Sanders’ only competitor for many of the groups listed above- has a record far more susceptible to accusations of corruption than Sanders does, something important when remembering how many of these voters cite corruption as a leading frustration. While Sanders’ record is unusually clean of corruption scandals, Trump has already spent months priming his base and muddying the waters for others on Biden’s issues so that he can repeat the success of distractions like Clinton’s email servers.
Bernie is a uniquely effective messenger for his ideas. I’ve written before about how even very far-left policies gain widespread support in traditionally conservative areas through appropriate framing, and Sanders’ idiosyncratic brand of outsider populism is the perfect vehicle for doing the same with his platform. When sold to voters using the same type of progressive framing Sanders is known for, ideas like a Green New Deal, stronger labor rights, a living wage, and Medicare-for-All poll rather well. Sanders can be relied upon not to make the same mistake as Clinton of running a policy-free campaign, and is instead far more likely to run on an inspirational platform filled with ideas that would demonstrably improve the lives of the vast majority of Americans.
Conventional wisdom would say it’s a liability that Sanders’ principled progressivism has lead him to take positions before they’re popular. But it’s just as often the case that this dynamic places Sanders on the same side of issues as the American people long before most other politicians are, something true even on the social issues that Sanders spends less of his time on. He has been calling for the legalization of marijuana since the 1970’s. Today, even the majority of Republicans agree with him while most Democratic politicians are still nervous to come around to the position.
Everyone Loves an Underdog
In closing, it seems relevant to me to discuss Sanders’ early political career. Sanders worked on the reelection campaign of a liberal city alderman who supported civil rights and opposed the Chicago Democratic machine while in college. When he moved to Vermont, Sanders ran for Governor and Senator with the left-wing third party Liberty Union a total of four times, losing each race. In 1981, he won the mayoral election for Burlington, Vermont’s largest city, by 10 votes. A bipartisan majority of the town’s leadership- 11 of the 13 alderman- were against him.
When it came time for reelection in 1983, the local political system seemed dedicated to removing him from office. To win, his strategy was simple: “direct confrontation with conservative institutions and legislators, blunt talk about economic grievance and an unrelenting effort to inspire demoralized, lower-income voters with promises of true societal transformation. He believed that only a sweeping vision of a better system could summon the kind of grass-roots mobilization he needed to achieve even more modest goals.”
It worked. The people liked Bernie more than traditional local politicians, and through a tireless campaign of direct contact and outreach to voters, his first reelection saw him win a three-way race by 20 points and successfully replace three incumbent alderman who opposed him with progressive allies.
As mayor, he combined progressive policies with good government reforms while running a budget surplus, eventually even earning him the begrudging respect of some conservatives. In 1987, U.S. News and World Report named Bernie Sanders one of the nation’s 20 best mayors. The next year, the U.S. Conference of Mayors named Burlington the “most livable city” in America with less than 100,000 residents. In 1990, Sanders won a four-way election to the US House with a majority of the votes cast, easily defeating an incumbent Republican to win over a seat that had been held by the Republican party for 30 years. He helped found the Congressional Progressive Caucus, which now has more than a fifth of the House of Representatives in it. When he won his Senate seat in 2006 and took over from a Republican-turned-independent, it marked the first time that Vermont had two Senators caucusing with the Democratic Party in its entire history. He passed few bills but instead chose to work through amendments, passing more roll call amendments than anyone else in congress even when Republicans were in the majority.
Looking back on his earlier local efforts, Sanders commented that his current strategy is mostly the same, just “on a somewhat larger scale”: “…if government does respond to the needs of working people, they will come out and participate.”
It’s not just that the most common argument used against Sanders’ electability is empirically unsupported, it’s not just that he can ensure a high turnout among the Democratic base, and it’s not just that he consistently polls well with every single demographic that the Democrats need to turnout the most in 2020. What truly makes Bernie the most electable candidate in this primary is his unique understanding of what it takes to for him to win. He is the only candidate with the unique combination of an unprecedented nationwide grassroots campaign, a boldly inspirational platform, an outsider persona, and the high popularity necessary to mobilize the American people en masse to fight for a brighter future. The question for Democratic primary voters isn’t how to handle a trade-off between a progressive candidate and an electable one, simply because that trade-off does not exist. In 2020, the most progressive candidate is the most electable one.
