This series of posts will attempt to answer a question I’ve been seeing a lot of my facebook feed recently: Could Bernie sanders have won the election? Before we get into answering this question or whether or not Bernie Sander could have won the election, we must admit this is a question about blame. People want someone to blame for the madman now in the White House. This question is an implicit attack on the Hillary Clinton’s candidacy. Blame can be a loaded concept accompanied by intense anger. To get at the truth of the situation, if there is any one truth, we will need two basic assumptions: (1) Anger must eventually give way to a hard look at the data without (as much as humanly possible) prejudice and (2) We must understand that the data will, inevitably, send us in contrary directions because data must always be interpreted. We must take those varied interpretations and present a meta-analysis of the data with commentary if we want to understand how the loss to Trump will inform future directions.
I’m in favor of a meta-analysis because there is a danger in finding a “silver bullet” that someone believes explains the election. There is no silver bullet. As Jordan Greenhall notes in his post on medium
The cultural s-curve associated with the rise of “black music” (jazz, blues, rock and roll, hip hip hop) played a role in significantly changing many aspects of American (and global) culture. Identity, family, language, fashion, how we move through space, what we value. There were multiple different s-curves (disruption waves) passing through each of those aspects of culture. Even as rock and roll was moving through culture, we saw economic disruptions relating to the triumph of the union movement, the migration of labor in the post-War environment. We saw s-curves in family dynamics associated with the expansion out of dense multi-family urban environment into nuclear (and relatively mixed) suburbs. We saw s-curves in gender relations. These waves are passing through many, many different aspects all the time and they are all interacting with each-other all the time.
To give an analogy of this somewhat heady metaphor, imagine a traffic jam on a freeway. There are any number of reasons for traffic such as collisions, fender benders, road work etc. If there is a collision and they drivers did not pull over to the side of the road, every person behind that car is affected. They are further affected by whether or not the parties call the police, how long they stay on the side of the road, etc. There are thousands of people affected, but this does not in turn account for the thousands of people who are subsequently affected by those who decide whether or not to slow down and look at the crash, which lane they choose to merge into, or a myriad of other driver decisions. Those thousands of decisions then trigger thousands of more decisions (a kind of “butterfly effect”).
Continuing with this traffic analogy, we can make a guess as to why the traffic is occurring, but nine times out of ten, we never actually get to the source of the traffic because it clears up before we can find it. While traffic is an immensely complicated endeavor for people to understand, politics is infinitely more complex with many more millions of individual causes. The problem with politics, however, is that everyone thinks they can understand it. In the same way that people often talk about fixing education and schools because they have experienced it, people often think their pat answers to political problems are commonsense solutions that make sense to everyone. The reality, however, is that most answers we give in answer to politics are local or regional solutions that might work for our city, county, or state, but would make little sense in a different state with a different set of circumstances. We also tend to become overconfident that the electorate looks like or behaves in a way similar to the people we are around on a day to day basis. In fact, Nate Silver of fivethirtyeight makes the argument that many pundit’s overestimation on an “emerging democratic majority” can explain many of the problems of coverage in 2016:
The “emerging Democratic majority” also hadn’t held up that well empirically. Since the book was written, Democrats had good election cycles in 2006, 2008 and 2012 but bad ones in 2004, 2010 and 2014 — and had decent results in federal elections but had fared miserably in elections for governor and state legislatures.
Why then, did the idea have such currency? One reason may be that it was seductive to liberal cosmopolitans, a category that includes most journalists at the Times and at other news outlets (including FiveThirtyEight). If you live in a big city and work in an industry dominated by college-educated professionals, you might intuitively overestimate the education level of the electorate and how rapidly it was diversifying.
I believe that any discussion about Bernie Sanders needs to be come within the understanding that literally everything affects everything. While I’m not a mathematician, another way to put it is that one calculation will affect all other calculations which will effect all other calculations infinitum. Talking about Bernie is really to talk about a progressive movement which is really to talk about a diverse coalition which is really to talk about politics which is really to talk about America. All of these small “bits” affect the other “bits.” This is especially important when discussing theoretical scenarios like “the Bernie effect” would have been on an election like 2016.
We must also admit from the outset, that this is, by nature, a gendered discussion. This entire post actually started after my wife commented on a Bernie supporter’s Facebook wall with support for the candidacy of Hillary Clinton. There was a lot of back and forth, but their main disagreement was their understanding of how Hillary was perceived by the public. Perception, in my opinion, is the key to unlocking politics and must be central in our understanding of why Bernie is perceived positively by a subset of Americans and also why Hillary is often perceived negatively — to varying degrees — by this same subset. Some of these perceptions are influenced by gender, race, income level, college attainment, geography, and sexual orientation (this is obviously not a complete list) and each of these perceptions, to use Greenhall’s s-curve model, cut across one another. No one perception can be viewed in vacuum (which is also why doing such a theoretical analysis is fraught with numerous problems).
False Belief in “American Narrative”
People like me who grew up voting in the Obama era were presented with a kind of false narrative. We saw a president twice elected who leaned to the left. We saw barriers broken down by the Supreme Court mostly upholding the important parts of Affordable Care Act and the rights of LGBTQ people to marry. We saw increased environmental regulations and a call for justice in wage labor. We had very few real political fights (besides the partisan bickering) that we felt called to us. But now we must stare in the face the fact that Mr. Trump has been elected president of the United States. For those like me who turned 30 in 2016, it was a real wakeup call that we need to take a closer look at our country.
We are not walking in a straight line toward justice. There is a belief in many liberal circles that Americans have picked up the mantle of MLK and are making this country more progressive everyday. Because, as I pointed out in the introduction, everything affects everything, trying, as some do, to write a narrative of American history as a march toward justice is not just an oversimplification, buf actually walks in the face of much of our history.
Consider for example, the emancipation of slaves. Life was supposed to moving toward justice and yet Jim Crow and Black Codes pushed black Americans into economic wage slavery and often into literal debtors prisons. The civil rights movement was supposed to finally bring this issue out of the closet, but education, wage, and socioeconomic gaps persist between people of color and white Americans. Wealth has become more concentrated in the hands of fewer people and the standard of living has gone down for many in America.
The Basic Argument and Possible Refutation
According to Sanders supporters, there is a fairly basic argument made as to how Bernie Sander might have won the election. According to David Horsey in an LA Times OpEd
Trump won the election by prevailing in the Rust Belt states of Michigan, Wisconsin and Pennsylvania that, together, gave him 46 electoral votes. In Michigan, he edged Hillary Clinton by just three-tenths of a percent. In Wisconsin, the margin was eight-tenths. In Pennsylvania there was a slightly larger gap of 1.2%.
All three of those states usually lean toward the Democratic candidate. This time around, most working-class white voters — many of whom voted for Barack Obama in the last two elections — saw Clinton as the incarnation of a political establishment that was indifferent to their struggles. They were won over by Trump’s boasts that he would protect American jobs and challenge the influence of Wall Street. Who else in the 2016 campaign made similar promises, with far more conviction? Bernie Sanders, of course.
The reason behind this shift, according to Bernie supporters, is encapsulated by Fredrik deBoer in his post Hillary Clinton lost. Bernie Sanders could have won:
Indeed, turnout overall was a major problem for the Clinton campaign; though not all votes are yet counted, it’s clear that Clinton received millions fewer votes than Obama in several states, while Trump frequently received more than Mitt Romney did in 2012. Nor did Clinton enjoy the benefits of party crossovers. There was much talk of “Clinton Republicans” who would, in the spirit of the Reagan Democrats, cross party lines to oppose Trump. But according to the exit polling of the New York Times, more Democrats crossed over and voted for Trump than Republicans crossed over and voted for Clinton. Sanders, notably, never had trouble drawing crowds, and in the Democratic primary campaign, turnout rebounded from 2012 lows. Whether that rebound was a result of voters’ enthusiasm for Sanders or the opposite is hard to say; what’s clear is that Clinton wasn’t able to get out the vote herself and that she lost both Democrats and independents to Trump, while Sanders had notorious luck with independent voters.
The most common refutation of these arguments that I have read are from those like Kurt Eichenwald in Newsweek who argued:
It is impossible to say what would have happened under a fictional scenario, but Sanders supporters often dangle polls from early summer showing he would have performed better than Clinton against Trump. They ignored the fact that Sanders had not yet faced a real campaign against him. Clinton was in the delicate position of dealing with a large portion of voters who treated Sanders more like the Messiah than just another candidate. She was playing the long game—attacking Sanders strongly enough to win, but gently enough to avoid alienating his supporters. Given her overwhelming support from communities of color—for example, about 70 percent of African-American voters cast their ballot for her—Clinton had a firewall that would be difficult for Sanders to breach.
Accounts on both sides like the ones above pepper the internet in droves and read almost like carbon copies of one another. Such a debate on these terms, as I’ve found on social media and other places, is about as fruitless as the debate on abortion because the debate is, for the most part, binary. One side says Bernie could have gotten the working class and tipped the election toward the democrats and the other side says he would be hit by such strong negative ads that he would have lost to a candidate that was famous for pulling no punches. I could see either account playing out especially in a world where every decision affects every other decisions. Clouded, as Yoda would say, this vision is. To find the data that will help us make the case one way or another will be in the detailed analysis of the electorate and a more nuanced look at sociological, psychological, and economic factors and how those currents cross each other.
In the next section of the series, we will look at three types of voters: Bernie, Hillary, and the Donald.