<<Back to the Questions
There is still a lot to be sorted out in the aftermath of the 2016 presidential election, but one thing is quite clear: Donald Trump significantly outperformed the predictions of both polls and pundits. Politicos and data scientists will no doubt devote years attempting to answer why Trump’s support surged on election day, but one potential key to this question can be found in the abundance of social data in the leadup to the election. Could a closer analysis of the social conversation about the election have provided a more accurate projection of election results?
When we analyzed the emotions expressed in social posts surrounding both major party candidates, we found one emotion much more present than in the 2012 election: fear.
The Atlantic insightfully pointed out the potential implications of this increase in fear prior to election day:
“[Fear] is a unique political force. Its ebbs and flows through American political history have pulled on elections, reordering and destabilizing the electoral landscape… If Trump still stands a chance to win in November, fear could be the key.”
Looking at the difference in emotion between the last two presidential elections, we also see a rise in “joy” from 2012 to 2016. Despite that rise in the happy emotion, “joy” is heavily outweighed by the dominance of “fear,” and the closely related “disgust.” Those two emotions accounted for more than half of the overall conversation about the candidates.
While Hillary’s campaign focused on the fear and disgust of Trump, the Trump campaign capitalized on many of the larger fears and disgusts in the country including terrorism, crime, illegal immigration, and economic instability. Had political analysts paid more attention to the dominant feelings of fear and disgust on social media, perhaps election results would have been less surprising on November 9th.