Date of Thesis

Spring 2018


The 2016 presidential election results varied significantly from many forecasts. The media proposes that the radically atypical candidacy of Donald Trump motivated pockets of the electorate to support the Republican Party more so than they had in past elections. This paper examines the following questions: Which traditional predictors of the election failed to foresee a Republican victory? If the traditional predictors were unsuccessful, can the results be explained using county level economic and demographic data? Is there evidence to support the media’s proposed explanations of the results? By utilizing the Census Bureau’s American Community Survey data, this paper examines some major economic and demographic determinants of the 2012 and 2016 election outcomes with a particular emphasis on the role of globalization and immigration. Further, this paper examines how changes in these county level features affected differences in county level support of the two Republican nominees. Thus, this paper finds that sociotropic economic perceptions remain as a significant determinant of voter behavior, even at the county level. Trump’s campaign may have failed to muster more support in counties with a high concentration of secondary sector occupations and higher inflows of immigrants. However, the results support the media’s conventional narratives that white Americans more heavily supported Trump, while educated Americans increasingly disapproved of the Republican nominee.


presidential election, immigration, county

Access Type

Honors Thesis

Degree Type

Bachelor of Science


Mathematical Economics

First Advisor

Christopher Magee

Second Advisor

Janet Knoedler

Included in

Econometrics Commons