The US presidential election is a interesting source of analyses for data scientists. In the following weeks, I will try to bring interesting perspectives to the election of Trump as President of the United States.
The following figure shows the number and percentage of votes each candidate had. Red arrows shift right for Trump, and blue arrows shift left for Obama. The greater the shift of the arrows, the greater the percentage difference between the candidates:
Barack Obama had 3 million more votes in 2012, compared to Trump in 2016, although you would not know this from this map. The first fact to know to understand the election is that not all counties are created equal, some are larger in land, population, some are more rural, and others more economically developed. The election of counties is also not arbitrary or meaningless. As you may know, the US president is elected by the electoral college and not by the vote count. This in turn gives an advantage to smallest counties and states in the US.
I won’t presume to understand the election of Trump, a very different candidate and the most inexperienced politician to achieve the presidency and in US modern political history. But let’s analyze the features of the election by county, the most geographically detailed data available for the US presidential election to understand why voters decided to vote for him.