I’ve been reading a just-released book titled “Lost In a Gallup: Polling Failure in U.S. Presidential Elections” by W. Joseph Campbell. I’ll offer up a more comprehensive review in the coming days but it’s a great read for those interested in elections, presidential polling, and American political history. Campbell delves into the controversy of polling and obviously the 2016 presidential race.
The author has an entire chapter on “experts” and the media debacle in predicting the most recent presidential outcome. While Clinton was ahead in all but one of a dozen final polls before the election (Investors Business Daily was the outlier at Trump 45-Clinton 43), the national polling wasn’t wildly off from the final outcome. In part because Clinton ran up her vote totals in California. The polling broke down in the Midwestern states and in North Carolina too. There are several reasons and two main ones were those very late decider voters in those states broke overwhelmingly for Trump and the non-college-educated were underrepresented in so many of the individual state polls.
The media interpretation of the polls were just as problematic, if not more so than the polls themselves. Newsweek ran their now infamous “Madam President” cover amongst so many examples. Some self-proclaimed experts predicted Clinton’s chances at winning to be over 98 percent. One forecaster even vowed to eat bugs on TV if Trump won, which he did. Campbell has a lot of great quotes from journalists completely dismissing any chance for Trump to win. Of course, this is all well-remembered now, what is interesting to me is if we will see some of these same patterns in 2020.
Lots of Trump supporters and some pundits believe there is shy support for Trump. Sometimes it’s referred to as the “silent majority” in politics. Basically, folks who will support Trump but won’t tell a pollster or even bother responding to a poll. I’m not sure I’m sold on this theory but admit it can’t be dismissed either. Trump has been in office for nearly four years and virtually everybody has formed an opinion on him and there may not be as many late deciders this go around. However, it is interesting that the three times the polls missed at predicting the results in the last 40 years, Republicans won all three of those presidential races. Pollsters missed the late surge by Ronald Reagan in 1980 in his first impressive national landslide victory.
What is clear is that so much of the media has still not made any attempt to understand Trump supporters or work towards impartiality in the upcoming election. They are doubling down on an aggressive campaign to drive the president from office. In part, this election feels like a referendum of the media itself and its credibility. Right or wrong to support Trump, those that do are having their worldview and concerns continually dismissed and ridiculued by so much of the media.
One of the conclusions highlighted in the book is that more “highly-quality polls at the state level, conducted close to the election” are needed. This is where the polling work of Civitas is important. We take our polling credibility and analysis of those polls very seriously. Keep up with our polls because I believe this presidential race will tighten up more in the last few months and obviously, North Carolina will be one of but a handful of states that will determine the outcome.