How I Got the Election Wrong
I was one of those pundits who confidently predicted a Hillary Clinton win. I did this numerous times on television, in interviews with reporters, in formal and informal discussions with foreign diplomats and leaders, and more times than I can count with friends and family. I predicted Clinton would get 318 or 320 electoral votes and win the popular vote by five or six points. While Donald Trump’s victory leaves me deeply concerned about the future of our democracy, the physical safety of many of our citizens and global economic and political stability, it also forces me to both publicly acknowledge that I got it wrong and to ask myself how I made such a big mistake. As for the first point-I got it very wrong. I am saddened, sorry and a bit embarrassed by that. Understanding the several reasons why got it wrong may help me in the future.
Like many other pundits and other citizens, I am exposed to an increasingly narrow range of the ideological spectrum. My social media feeds are tilted heavily towards people who share my political views. I live in a neighborhood that votes overwhelmingly Democratic and mostly work alone or with people who share my political views. Most of my friends and family are left of center. I have one very close Republican friend who has, in the past, acted as a check on this problem, but this year he was not at all enthusiastic about his party’s nominee. Thus, our email exchanges and phone calls over the last few months have focused mostly on baseball and family rather than exchanging ideas and opinions about the election. This led me to make the same mistake President Obama did when he said “I continue to believe Mr. Trump will not be president, and the reason is because I have a lot of faith in the American people.”
I compounded this by cherry picking the polls. For the last many months, going back to the primaries, I checked the polling section on Real Clear Politics several times a day. In the last five or so days before the election, a clear trend was emerging. Trump was closing the gap both nationally and in many key states. I simply refused to believe or process that. I chose, instead, to only believe the increasingly few polls that showed Clinton having a solid lead and rationalizing away the others. The evidence was there but my hopes, partisanship and fear of a Trump presidency blinded me to it. I also looked at changes in the demographic data without focusing on the now obvious finding that most of the growth in Latino population has occurred in safe states.
A related point is that I have spent enough time in graduate school statistics courses, crafting polls and talking to pollsters to know that polling is complicated. It is part art and part science. The art includes things like determining how much to weigh different respondents and what kind of turnout model to use. Despite that, I never got my hands dirty with the numbers and the modeling. Instead, I looked at the top lines, read a few analyses of the methodologies and trusted others to do the hard work with the numbers. As a result, I was susceptible to the growing media consensus that Clinton was winning and polls that showed otherwise were prima facie methodologically flawed.
I also ignored the warning signs. These included the high turnout for Trump in the Republican primaries, the failure of any of the Republicans to get any traction in their efforts to draw attention to Trump’s countless shortcomings and failures in the primaries and the inability of Hillary Clinton to connect with working class white voters. I also looked at economic data that suggested the economy was recovering while discounting the effect of years, maybe decades, of right wing media telling white working class voters how bad their lives were. This made me miss signals such as Clinton’s very light campaign footprint in those communities.
I have never liked the Clintons. This was the third time I voted for one for President, but I have always found them too conservative and, while not criminal, a little too unethical and defensive for my tastes. Nonetheless when Hillary Clinton won the nomination I was happy because I thought the one thing I can trust the Clintons to do is play campaign hardball and win elections. I trusted that instinct and believed the stories about her superior field operation. However, when you lose an election because a few close states swing against you, and turnout is low, the field operation must be blamed. The Clinton campaign was good at projecting calm and even made some deft strategic moves, but the field operation was very disappointing.
The last reason I got this wrong is the most upsetting. My 16 year old self, the one with hair halfway down his back, running around San Francisco spouting radical political slogans would have gotten this right, but I didn’t. The things I believed then, that a big majority of white America was racist, that a clownish authoritarian could get elected president and would surround himself with white supremacists, made many see me as some kind of left wing nut back then. Turns out that version of me was right.
Photo: cc/Christi Nielsen