When I was a teenager I enjoyed reading the historical fiction of authors like James Clavell. In one of his stories, Tai-Pan I believe, there was supposedly an old Chinese curse that went like this: May you live in interesting times. I have no idea if such a curse exists, but if it doesn’t it certainly should, because we are certainly cursed to live in interesting times.
Many have asked me to write a postmortem of sorts for the 2016 presidential election. As to oblige, the following is a list (in no particular order) explaining what I believe led to the outcome in the 2016 Presidential election. Many of the items on this list bleed together and intensify the impact of one another.
Word of warning, this is a long list for this kind of format, and I’m not going to put together a tl; dr version. These assessments come from a combination of expertise, the publicly available information released up to this point, and some of the semi-inside info I’ve received from several people who were in the field during the campaign. Since it is still early, some of these assessments may change as new information emerges. That being said, here’s what I think happened.
1. Social desirability bias is alive and well. Many people, especially college educated voters (both men and women), did not want to fess up to pollsters that they supported Trump. Of course, in hindsight this is not surprising, and I had concerns about this problem early on, but there was little good evidence to suggest it was a factor during the primaries (and what evidence that did exist was mixed). Moreover, the classic examples for social desirability bias in surveys occurred when white people would lie about supporting African American candidates for political office. The incredible accuracy of the overall polling environment during the 2008 and 2012 elections led many to believe that social desirability bias was no longer as major a problem as it was in the past. We were lulled into a false sense of security, and the consequences were both great and terrible.
The bad polling data led to bad polling aggregates and bad models. Or as many methodologists like to joke about these kinds of situations: garbage in, garbage out. Faulty data leads to faulty decisions. This led many within both campaigns to make faulty decisions about vote targets, potentially effective messaging, canvassing goals, targeting areas, etc. It also helped fuel a lot of confirmation bias across the board (myself included). Though to be clear on this point, damn near everyone was fooled by this. There are several reports originating from within the Trump campaign that suggest that they expected to get no more than 240 Electoral Votes on election night. They were caught flat footed by this as much as everyone else.
Similarly, most pollsters used faulty assumptions for the population weights they used in their survey results. They assumed that the election would look like 2012. This turned out to be wrong. Some pollsters released multiple findings for their polls using different weights. I think this should become the norm.
2. Sexism is alive and well. I have had people (who I thought were smart, until they said this) tell me that Clinton was nominated solely because she has a vagina. Period. The exit polls are suggestive on this matter. Damn near 80% of males overall voted for Trump. Moreover, the large number of sexual harassment stories in the days since Trump’s victory makes it clear that it may be debatable as to whether sexism was the most important factor for deciding this race, but you cannot deny that it was a factor, and likely an important one.
3. Racism and bigotry are alive and well. The exit polls and on-site interviews are suggestive on this point. People who care deeply about illegal immigration overwhelmingly went for Trump. And they were very vocal about it. We are talking about a winning candidate endorsed by both Neo-Nazis and the KKK. Moreover, many hate groups bussed ‘their’ voters to the polls in several important states. On a personal (albeit anecdotal) level, I know many of us have had to look no further than our own social media environments to see the role bigotry played in this race. I also know many people who have already had to deal with this in the real world. It is unclear at this juncture how important this was for the outcome, but it’s pretty clear that it was part of the equation.
4. White working class/white rural rage is real, and Democrats need to ditch the corporatists that moved the party away from the working class in general when it comes to economics. When it comes to economics, Trump ran as an old-school Democrat in many ways (at least rhetorically). Superficially, much of what Trump said on corporate greed, the system being rigged, etc. is similar enough to what Bernie Sanders said that many voters would not be able to tell the difference. And the Democrats let him get away with it. Instead of showing how full of shit Trump is on these issues the Democrats largely ceded the ground. This had consequences which are now plainly seen.
5. I’m sorry, but Clinton was a terrible candidate. I know I’m going to get heat for saying it, but it’s true. To be fair, it’s not entirely her fault. Her enemies have been very effective in painting her in the worst possible terms for decades, and the aforementioned problems with sexism make the impact of these attacks even worse. However, she ran a terrible campaign (especially considering who she was running against).
Many of us hoped that the Clinton camp learned from 2008, but as of right now that does not appear to be the case. It appears that they still valued loyalty over competence, and suffered from group think to the extreme. For example, I’ve gotten both inside accounts and read media reports stating that the Clinton campaign did not try to mobilize support or persuade voters in rural areas. Nor did they try to target working class whites.
Moreover, there are media accounts suggesting that this decision was made over the loud objections of Bill Clinton (who knows a few things about this subject). Both Bill Clinton and Barack Obama made plays for these constituencies. Even if you don’t win them over, you must try to counteract your opponent’s efforts in trying to win the day with these voters. The Hillary Clinton campaign ignored these constituencies almost entirely. The damage on this point is evident.
Similarly, insiders have told me, and media accounts now confirm that the Clinton campaign overly relied on simulations for determining messaging, canvassing maps, GOTV priorities, and other matters. In and of itself heavy use of data analytics isn’t necessarily bad. If you do it right I believe it can pay dividends. However, it’s what happened next that makes me want to scream. After putting their plans into effect for messaging, canvassing, etc. the people on the ground reported that it wasn’t working.
This went on for months in both critical and noncritical states where people on the ground warned the campaign brass that it wasn’t working. Messages were falling flat. They couldn’t build enthusiasm or get voter commitments. Targets were not being met. The brass in Brooklyn told them to stay the course. That is political malpractice, and whoever was in charge of making that decision should never get paid to work in politics ever again. Ever. In general, using data and running simulations to help make strategic decisions is a good thing, but once you’ve tried to implement those plans and the evidence suggests it’s not working then you need to rethink what you are doing. You need to challenge your assumptions, question the data, and listen to the people you have entrusted to put your plans into effect. Ugh. And again, you can’t just focus on minority populations, young voters, and women, and you certainly can’t just focus on the urban and suburban areas. This is a recipe for disaster.
6. I am certain I’m going to get heat for this too, but the evidence is already mounting that it’s true. Third Party support did constitute a spoiler effect for both Clinton and Trump. Exit polls suggest that roughly a third of Johnson voters in critical states, states decided by less than one percent were self-identified Democrats. Moreover, nearly all of Stein voters were self-identified Democrats. If these people had voted for Clinton in places like FL, WI, PA, or MI it could have swung the election. That’s a textbook spoiler effect. Though I of course concede that this, and all the other factors identified in this dissection do not exist in a vacuum. Regardless, that is all I’m going to say on this subject.
7. The media dropped the ball. In the attempt to adhere to the ‘balanced coverage’ approach to covering campaigns many in the media pursued a false dichotomy that led many to believe that the campaigns were comparable in many ways. This is nonsense. Yes, the media covered the ‘Grab them by the pu**y’ video, but they also dedicated a ridiculous amount of coverage to the email story, even though it was a nonstory compared to everything else going on in the campaigns. For example, even now the majority of voters are unaware that there are around 75 pending lawsuits against Trump, many involving allegations of serious crimes. Similarly, the media gave Trump so much free uncontested air time it was insane. Not to mention that issues were generally ignored. Yes, the media always focuses more on the horse race, but not to the exclusion of damn near everything else.
To be fair, television and cable are dying platforms, with fewer viewers by the day. Similarly, much of the traditional print media did a fantastic job, but fewer and fewer people read their content. Instead we get the vacuous nonsense that passes for discourse in social media. And most of what gets circulated in social media comes perilously close to rank propaganda or are downright hoaxes. I had a conversation post-Election Day with a family member and Trump voter where they listed a litany of well-known hoaxes to help justify their vote. To hammer the point home that these stories were false I read them the top Google search results for each story. In each case, the second Google result was the Snopes story explaining that the story was a hoax. I suspect many people have had similar conversations in the wake of this election. It’s has gotten to the point where Silicon Valley elites are starting to ask if Facebook has too much influence over American political discourse, especially given that most of what gets disseminated is patently false.
8. The Clinton campaign is arguing that the Comey letter is to blame for their loss. Given all the other problems I believe this is suspect as the sole cause. However, it undoubtedly played an important role. The media jumped on this story, and though they quickly dialed back the more explosive aspects because they were entirely unfounded the damage had already been done. And to add insult to injury the second statement from Comey stating that the new investigation was already over and had found nothing was nowhere near as widely circulated.
As for the impact of the FBI putting its finger on the scale, the best evidence at this point suggests that it influenced voter behavior in a number of ways. The exit polls and a lot survey data suggest that the Comey letter had a profound impact on late-deciders, with the majority of them breaking for Trump. This makes the scant coverage of Comey’s retraction all the more damning. Moreover, there is evidence that the Comey letter demoralized a significant number of Democratic leaning voters, and helped bring a lot of wavering Republicans back into the fold. Though I think Harry Reid goes too far in arguing that Comey violated the Hatch Act, his behavior came perilously close, and I believe the Justice dept. should have ordered him to not send the letter.
9. There could be many culprits as to why, but this was a low turnout election, especially compared to the two previous elections. This hurts Democrats for a number of reasons, but I will focus on what I think is the most damning reason that has the farthest reaching consequences for future races. This was the first presidential election in 50 years that was not conducted under the most important provisions of the Voting Rights Act. After testing the waters in the 2014 midterms the Republican Party pushed for measures in many Republican controlled states that depress turnout, and focused these efforts in traditionally Democratic leaning areas or targeted Democratic leaning constituencies. Unless stopped by the courts, Republican controlled states cut polling places in Democratic leaning areas, reduced the number of days for early voting, reduced the number of early voting locations, and passed voter ID laws purposely designed to make it harder for Democratic leaning constituencies to get lawful IDs.
We know that the goal was to target Democratic leaning constituencies because the GOP admitted it through the evidence presented in federal court, and both federal and state rulings were supposed to stop both NC and WI from implementing these laws. Both NC and WI ignored these rulings and did it anyway. In WI this likely threw the election to Trump given that over 300,000 voters were turned away for not having IDs (that the courts said they didn’t need to have), and as of now, Trumps vote margin is around 27,000 votes. Voter suppression (and that is what it was) likely played a critical role in the results for FL, NC, WI, and MI.
I could go into greater detail and deeper in the weeds for every single one of these points, but this is already longer than I wanted to make it, and I have other work to do.
I hope this clarifies things for many of you or at least gives you something to think about. If anyone wants them I can provide links to support every single one of these points.
I have been getting a lot of questions about what I think the Democratic Party should do or what liberals/progressives should do about all of this. I have a lot of thoughts on this matter, but I am unsure if I will put them out into the ether.
Though I will say this, when there is no hope there is no fear. For years now the left has allowed the more conservative factions to govern the Democratic Party because many believed the stakes were too high (protecting existing gains, the Supreme Court, etc.). Well, that’s over now. The right now controls all three branches of government, and they are already making plans to do what many have feared. So now I suggest that the left should fight back. Band together and fight for what you actually believe in, and don’t be afraid to lose. When the conservatives were out in the wilderness for decades they were not afraid to lose, and it helped them immensely. Democrats can’t just be the ‘not crazy’ party anymore. The party needs to actually stand for what its voters believe in. Enough is enough.