Rachel Bitecofer is Assistant Director of the Wason Center for Public Policy at Christopher Newport University, USA, and author of The Unprecedented 2016 Presidential Election.

Read a free chapter, "What (Really) Happened " until March 16.

Political science is an interesting field because it is the only field where non-experts regularly claim expertise. Nowhere is this more true than among political media, most of whom have no exposure to political science beyond the undergraduate level. At the undergraduate level, political science education teaches students about concepts such the incumbency advantage or the midterm effect but almost always, the methodological processes used to arrive at these concepts is kept hidden from undergraduates. Only at the graduate level does the “science” of political science get introduced.

As such it should be no surprise that most accounts of political phenomena “explained” by political media rely heavily on anecdotal evidence and case study observation. Certainly, these accounts contribute to the public’s understanding of political events, but such explanations are limited in scope and in the case of election outcomes, are quite often wrong. Take for instance the various arguments put forth by political media and political players to “explain” the results of the 2016 presidential election. These explanations range from pinning Clinton’s loss on a campaign mired by infighting and bogged down by mismanagement to the campaign’s inability to win over white working-class voters, particularly in the Midwest. But what if I was to tell you that the Clinton campaign from top to bottom was one of the best designed and resourced campaigns in history in terms of talent and treasure? Further, the Clinton team found themselves competing in the most unbalanced election in the modern era. Presidential elections usually produce two highly capable candidates running two well-resourced campaigns, deploying similar tactics, and largely offsetting each other in terms of resources and infrastructure. But the Trump campaign was inferior in every metric. Given that information, one might be incredulous at an explanation that grounded Clinton’s loss as a product of a poorly run campaign.

In the case of what I call the white, working-class hypothesis (henceforth referred to as the WWW hypothesis) Clinton’s loss was due to her failure to articulate an effective economic message to counter Trump’s populist, often nationalist, economic message which naturally resonates with a segment of the population most displaced economically by globalization and technology. This is an attractive explanation as well because it fits a narrative touted by the media throughout the campaign, especially coming out of the Democratic primary where Clinton held off a robust challenge from Bernie Sanders who ran on an unabashed platform of progressive populism.

The problem the WWW hypothesis runs into, of course, is data. I often tell my students in my political analysis class that “the data don’t lie” and the data have a tale to tell about the voting behavior of WWW voters. Analysis of voter preferences in elections as far back as 1980 shows that once white working-class voters became “Reagan Democrats” they went on to become permanent Republicans with little interruption. In fact, the last time that Democrats made gains among this voting group was in 1992, and to a lesser extent, the 1996 elections, when the party offered two southern white men on their ticket. Other than that the vote share for Democratic Party nominees among white working-class voters is remarkably stable, especially in the Midwest. Hillary Clinton did no better or worse than Barack Obama among white working-class voters yet Clinton lost the Electoral College and Obama went on to win both elections. Knowing this, people might also view an explanation of the election that pins Clinton’s loss of WWW voters skeptically.

So what does political science tell us about Clinton’s loss? In my book, The Unprecedented 2016 Presidential Election, I show that Clinton’s loss was due primarily to a decision she made months before Election Day. Hoping to capitalize on the “Never Trump” movement the Clinton team embraced a persuasion-based campaign strategy aimed at winning over Independents and disaffected Republicans. The centerpiece of the strategy was the selection of Tim Kaine for vice president. A moderate and affable well-liked senator from the key swing state of Virginia, the Kaine pick was designed to make Independents comfortable with voting for Hillary Clinton, whose image was badly damaged by a series of Republican-led investigations into Benghazi.

The problem with this strategy is that it relied on an outdated notion of the American electorate. In the age of polarization there are few persuadable voters out there and in their pursuit of Independent voters the Clinton team’s persuasion strategy further isolated the progressive wing of the Democratic Party. On Election Day, a historic number of progressive voters defected from the Democratic ticket. In Wisconsin, a state that was decided by less than 1%, the defection rate was 5 times higher than normal with 6.32% of voters casting third party or write-in ballots. Meanwhile, despite gearing her entire electoral strategy towards them, Clinton did no better than Obama among Independents. Clinton’s loss was an ideological loss. Her team conducted the nearly perfect execution of the wrong electoral strategy and in so doing, cost her the only chance she had to become America’s first female president.

The lesson drawn from political science about the 2016 presidential election is that in the age of polarization, elections are all about that base.