When Polls Become Idols
Donald J. Trump's victory was described as an upset in many news outlets on November 9, 2016. It was an "upset" only because of the almost uniformly failed prognostications about the results from "scientific pollsters," from Nate Silver to Quinnipiac to CNN and numerous others.
What went wrong?
The news networks and the pollsters will try to answer this question by looking at the demographics of polling, and how there were so many unanticipated "last-minute" shifts among Jews, blacks, women, educated, uneducated, Hispanics, etc. This type of analysis tends to shift the blame from the pollsters to the "unpredictable voter." The voter is thus portrayed as behaving in a surprising, even erratic way that caused the pundits to err.
Is this model of explanation valid?
In order to answer our question, it is not necessary to review the details of polling methodology. Rather, we can answer our question by going back to the sixteenth-century savant Sir Francis Bacon. Bacon perceived that advances in knowledge were being held back by our biases. He called these biases "idols" and asserted that they distorted our perceptions of reality, and even our experimental approach in science.
His choice of the word "idols" was significant in a time when Christian religion was not considered retrograde, as it is in many learned circles today. Idol-worship was considered a wrong worship when compared to worship of the one true and living God of the Old and New Testaments. The Ten Commandments includes the statement: "You shall have no other God before me." If one did have another God, then he or she would be an idol-worshiper. Thus, by designating certain biases as idols, Bacon was to some degree loading his terminology to suggest a kind of ungodly behavior. And since Christians believed that Jesus Christ was the "way, the truth, and the life," the distortion of our biases was a departure from godliness and hence from truth.
Truth or the search for truth was a search to free one's mind from biases. Nevertheless, in our time, this writer has been told good-naturedly by certain academics that it is impossible to rid ourselves of our biases, so we should extol our biases, teach our biases (as though they were truth [realizing as we do that they are nonetheless relative]), and go happily on our deconstructivist way.
However, Bacon listed four idols that this writer will briefly outline and suggest that they applied to the vast majority of pollsters in this presidential election. Their minds were distorted by mis-assumptions, inclinations, and various agendas that kept them enslaved to their biases and thus, in the last analysis, wholly incorrect in their predictions.
First, Bacon refers to "Idols of the tribe." This includes the human tendency to "accept, believe, and even prove what we would prefer to be true." Also, it includes our tendency to "rush to conclusions and make premature judgments." Thus, President-Elect Trump was right on target when he described the election as rigged. The pundits and pollsters had politically correct assumptions that attitudes of denigration of the Trump movement and the candidate contributed to their uniformly incorrect assessments. For example, how often did I hear Megyn Kelly say to guests she was interviewing that the vast and enthusiastic attendance at Trump rallies was relatively insignificant? What really counts, she often insisted, is the voting turnout and not the rallies. Yet this observation cuts directly against common sense. Trump himself, with uncanny insight, remarked frequently at his rallies that the media showed only the people behind him when he speaks and failed to pan the enthusiastic, vast crowd in front of him. He reprimanded the media, but his criticism was ignored.
A second idol for Bacon was the "Idol of the cave." This includes "high esteem for a few select authorities," as well as "a tendency to reduce or confine phenomena within the terms of our own narrow training or discipline." On election night, some of the TV commentators admitted their polling experts had had long arguments with Trump's people who talked about expected large turnouts in certain "red" counties in North Carolina. But the pollsters pooh-poohed this because they were attached to their predictability models based on a succession of past elections. They statistically evaluated the variability of turnouts in 6, 8, or 10 previous presidential elections and concluded that turnout would have to be within a certain range. In fact, the turnout in various "swing states" was greater in the pro-Trump counties and outside the range of turnout of the pollster models. They had discounted the wholly correct feedback they had received from the Trump camp.
A third idol for Bacon was the "Idol of the market place." This type of idol is "from the 'intercourse and association of men with each other.'" Here the main culprit is language. Thus, the pollsters and the commentators who are working hand-in-glove tend to live in a politically correct world, a world where definitions are distorted to mean whatever the P.C. crowd wants them to mean. There is a mental discounting of certain Trump expressions as "mere demagoguery" when the people listening are hearing words that are "very inspiring." Thus, the commenters and the pollsters who support them have an edge of contempt built into their assessments.
Hillary and her supporters on PBS, CNN, NBC, ABC, and CBS are always using the word "inclusivity" as expressing her vision for America, thus they suppose that Trump's interest in putting more controls on illegal immigration or refugee admission is exclusionary, whereas he is merely talking about some practical restraints, not the elimination of our inclusive history. Thus, polling pundits failed to factor out or offset their own P.C. bias in interpreting the feedback from voters and potential voters. Pollsters must ask questions, and the questions as well as the hearing of the answers require an interpretive cognitive framework that governs the creation of the questions and the understanding of the responses. Thus, the polling questions may be distorted and colored by bias from beginning to end.
Bacon's fourth idol was the "Idol of the theater." This is the idol of the marketplace operating at a more sophisticated or philosophical level. Thus, the pollsters are not merely carrying around the baggage of left-wing politically correct assumptions and beliefs (both educated and uneducated may have these). They are reinforced in their anti-Trump prejudices by sophisticated economic theories like those of Robert Reich, Paul Krugman, and even David Ricardo (author of the principle of "comparative advantage," which is so basic in our formulation of so-called free trade deals that Trump wants to correct). The sophisticated anti-Trump pollsters and commentators are apostles of John Stuart Mill and his ethical principle of the greatest good for the greatest number. This principle is extended to apply to a global model (not the nation-state) as the "greatest number." To this end, they are willing to sacrifice other ethical philosophies embodied in the Ten Commandments and the Sermon on the Mount that ratify the sanctity of property ownership and of individual moral liberty, not imposition of fiats based on collectivist neo-Marxist assumptions.
In essence, it is idol-worship that caused a misunderstanding of the dynamics of this year's election by the media. But truth broke through and showed itself to an amazed population.