Numbers can tell a story, but they can also be misleading. And as we now know, pre-election polls in the 2016 Presidential Election predicted the wrong outcome with Donald Trump beating Hillary Clinton in the Electoral vote.
Wall Street Journal Data Reporter Paul Overberg shared his insights about the role data played in the historic election in his presentation, “By the Numbers: Data, Journalism and the 2016 Election,” at the College’s Media Innovation Center on November 28.
Overberg addressed problems in polling, such as outdated data-gathering methods and the “shy voter” phenomenon, in which voters aren’t always truthful about their choices. He also discussed the possible impact of both candidates’ historically high unfavorability ratings leading up to the election.
“There was a lot of trying to figure out this year who was going to show up to vote—no one really knew,” said Overberg. “There has never been an election where both candidates were so underwater in terms of favorability.”
Overberg suggested that the large amount of money spent on polling could be better spent on funding data reporting before and after elections. And he demonstrated how demographic and economic data can be used to better explain what happens in elections and why. For example, through data analysis, he was able to show the connection between immigration patterns and voting in specific regions of the country.
Overberg said data journalism is only going to grow in importance over time, and he encouraged students to take more math classes, learn coding skills and use easy tools that enable journalists to analyze and “interrogate” data, such as Structure Query Language.
“All journalists will have to know how to use data,” he said “If you learn a little bit of data while you’re in school, you’ll be ahead of many professional journalists.