Vol. 17(3)- Replication Data for: How Unusual Was 2016? Flipping Counties, Flipping Voters, and the Education-Party Correlation Since 1952

Many explanations of the 2016 election result, a seemingly anomalous macro-level phenomenon, have centered on two seemingly anomalous micro-level phenomena: that many counties and citizens voted for Obama in 2008 and 2012, but then flipped and voted for Trump; and that low-education whites gave more of their votes to Trump than to Clinton. In this article, I first assess the novelty of these facts by placing them in the context of past elections. Compared to past presidential elections, the number of flips in 2016 was not unusually large, even in the Midwestern states. In contrast, the partisan divide by education was the highest ever in 2016. Using a series of counterfactual analyses, I then assess whether these factors were pivotal. If flipping counties had not flipped, Clinton would have won the electoral college by three votes, and if the lowest-educated 20% of counties voted as they did in 2012, Clinton would have won the electoral college by about thirty votes.