UNO Political Scientist Predicted Trump as the Republican Nominee
In the lead-up to the Presidential primary season, UNO Political Scientist and Presidential nomination scholar, Randy Adkins, dove into the statistics.
- search keywords:
- Political Science
- Randy Atkins
- elections
During the Republican presidential primary season, political observers – pundits, journalists, and political scientists alike – watched in disbelief as Donald Trump cruised to the Republican nomination. During the primary season, many of these observers predicted that Trump would come to fail. For political scientists, a book called The Party Decides explained why. The book represented the best thinking on party nominations, and its conclusion is that the two most important factors for predicting who the party nominee will be are elite endorsements and fundraising. In 2015, the vast majority of the Republican elites who had endorsed candidates chose Jeb Bush, and he had also by far raised the most money. The clear prediction was that Jeb Bush would be the nominee.
Nate Silver, the editor of the FiveThirtyEight blog, and probably the most famous data-driven election forecaster in the country, came to the same conclusion (and for the same reason). After Trump secured the nomination, he wrote a post on the morning of May 18 titled, "How I Acted Like a Pundit and Screwed Up on Donald Trump." How could even he have made such a mistake? Silver wrote,
“We didn’t just get unlucky: We made a big mistake…"
"We didn't just get unlucky: We made a big mistake...[It was] a curious one for a website that focuses on statistics... our early estimates of Trump's chances weren't based on a statistical model. Instead, they were what we call 'subjective odds' which is to say, educated guesses. And we succumbed to some of the same biases that pundits often suffer... Without a [statistical] model as a fortification, we found ourselves rambling around the countryside like all the other pundit-barbarians, randomly setting fire to things."
In the lead-up to the Presidential primary season, UNO Political Scientist and Presidential nomination scholar Randy Adkins wondered what such a statistical model would predict. There had been previous scientific studies of this question - if these models were applied to this election, something which Silver neglected to do, what would their predictions be? Working with his co-authors at the University of Arkansas, Andrew Dowdle, Karen Sebold and Jarred Cuellar, he analyzed some of these past factors that predicted who the major party nominees ended up becoming. By using common sense variables such as who wins the Iowa Caucus, and New Hampshire primary, and how much support the candidates have in public opinion polls at the end of the year before the nomination process, his model predicted that Donald Trump would emerge as the Republican nominee (and it predicted Clinton for the Democrats, too). The model has a highly accurate record in predicting the past nominees as well. The results are forthcoming in the journal PS: Political Science & Politics, so that people like Nate Silver, who stake their reputations on predicting election outcomes, will be able to use this model to provide insights into the how and why of these kinds of presidential election outcomes. The article represents an important step forward in the scientific understanding of vote choice.