

One of the lessons learned from the past election is about the over-reliance on polling data, and a reconsideration of polling methods. On election day the voting results deviated from the polling data, especially in states like Wisconsin, Michigan, and Pennsylvania, states previously believed to be part of Hillary Clinton's impenetrable "blue wall". Throughout the campaign, polls which showed Donald Trump performing better, like the LA Times, were assumed to be flawed at least or biased at worst. Many simply assumed that since the numbers didn't match those of blue chip pollsters like Nate Silver, who consistently assured us that Clinton had odds of winning of better than two-thirds. But no real explanation was ever provided as to why the polling methodology of Silver and many others was imperfect or inexact. As far as I've been able to determine, no one has offered any solid explanation since the election, other than to remind us of how, like weather prediction, polling is still an inexact science.
Pre-election speculation offered reasons suggesting that Trump's polling numbers were inaccurate, both high and low. Those who argued that Trump's numbers were too high suggested that voters were telling pollsters that then intended to vote Trump, but really wouldn't be able to bring themselves to do it when they walked into the voting both. They might be angry at the establishment, but when it came time to mark their X, they wouldn't really give the nuclear codes to someone as bombastic as the Donald. Those who though his numbers were too low said that many believed that they would be criticized or attacked if they honestly admitted that they intended to vote Trump because of the high levels of anti-Trump sentiment in the media. They planned on voting Trump, but didn't want to be attacked for doing so, so they lied to pollsters.
It's true that polling will never be an exact science. Polls may purport to answer which candidate a person supports at a given time. The more important question is: how solid is that support? In elections where voters "settle for" a candidate, rather than support one or believe in one, how they answer a poll today may be completely unreliable on election day. A strong political wind can easily sway poll numbers back and forth under such circumstances. And yet if one looks at polling trends, that didn't seem to happen.

Trump was strongly criticized in the national media, especially after the Access Hollywood tape was released. His surrogates were mocked on most panels and any Hollywood celebrity of note was quick to disparage him, many believe for good reason. But to voters in regions that had once been industrial hotbeds, this did not matter. To many women in these regions, allegations of Trump's misogyny, sexism and perceived immaturity were not reasons to vote against him. In many regions with high Latino demographics such as Florida and Arizona, his past remarks considered disparaging by many did not prevent them from voting for him.
From the results of Tuesday's vote, the majorities in the states that went for Trump seemed to want to teach the following lessons:
1. There is vast economic disparity. On one side of that disparity, many people are afraid for their economic future. On the other side are people who are not afraid for their economic future, and are oblivious to the concerns of those who are. For those in the second group, it is easy to ignore or write off the concerns of the first group. But the concerns of the first group are very real.
2. Voters no longer trust the promises of traditional politicians, if they ever did. Something more drastic is called for.
3. In the face of the problems presented by this economic disparity, political correctness is not a priority. This does not mean that people are generally racist or sexist. But they will forgive or overlook political incorrectness in favor or expressing their displeasure and unhappiness about their economic future.
4. Sometimes voters' strongest impetus is to vote against, rather than for. Throughout all the polling, the strongest poll numbers were those that suggested that voters were unhappy with the direction of the country and by large margins. When voters are strongly unhappy with the status quo, they will pick what they perceive to be the best alternative choice, even if the alternative would not be their first choice.
Voters' emotional reaction was as much about trust as it was about their fears. It did not matter which candidate was promising tax cut or to what segments of the population. Nor did it it matter which candidate was promising to increase the minimum wage. It did not matter how the candidates might deliver on the promises of great things to come that they were making. Their analysis was much more visceral.
So why didn't the polls reflect this? Guardian writer Mona Chalibi writes:
"I spent almost two years working for Nate Silver’s website FiveThirtyEight, where I hoped to learn the secrets of political forecasting. I walked away totally disillusioned. It sometimes seemed as though their interpretation of the math wasn’t free from subjective bias. There was also a certain arrogance that comes from being part of an elite that 'gets the numbers', and an entrenched hierarchy meant that predictions weren’t properly scrutinized."
Recent polling failed to predict the Brexit vote, as well as Justin Trudeau's majority victory in Canada and the Colombian Peace Agreement referendum. The LA Times said that one reason they believe their polling was more accurate was because they didn't underestimate Trump's support, and their online system captured people's will better than phone polls. Their pollsters found that Trump voters were noticeably less comfortable about telling a telephone pollster about their vote. This was especially the case for women who said they backed Trump.
Knee-jerk reactions are rarely helpful in assessing social trends, especially when they go against the current of what is actually happening. "There's no way Trump will win" was a comment frequently heard, one motivated by the candidate's unconventional political style and his "shoot from the hip" rhetoric. This in turn invited an unwillingness to examine what was really behind the large number of voters who supported him in most of the Republican primary and caucus states. As William Paley (and not Herbert Spencer, who is often credited) once wrote: "There is a principle which is a bar against all information, which is proof against all arguments, and which cannot fail to keep a man in everlasting ignorance. That principle is contempt prior to investigation."

The question that now must be examined and respectfully debated is not: "why did Trump win", or "should Trump have won". Rather, it should be: What can be done to address the fear of the many who worry about their economic future and security? This is not a red or blue issue, nor a conservative or liberal issue. It is one that everyone should be concerned about.