Uncategorized

Predicting the Presidential Election

The winning presidential candidate has enormous powers to shape the nation’s domestic and international policies. They will also appoint the nation’s cabinet members and judges to the United States Supreme Court and lower federal courts, all of which require Senate confirmation. They will also have a role in the budgetary process.

As supporters of Mr Gore and Mrs Clinton will no doubt remember, a lead in the popular vote does not necessarily guarantee election to the White House. America’s unique electoral college system ensures that the winner of each state’s electoral votes, not just those from the national popular vote, will win the presidency. To predict the outcome of this race, our model uses engineering principles based on probability and statistics to analyze election polling data from several well-known sources, including ActiVote, Bloomberg/Morning Consult, Dartmouth College, New York Times/Siena College, Emerson College, SurveyUSA, UC Berkeley’s Institute of Governmental Studies and The Washington Post.

A national election requires 270 electoral votes to be declared a winner. To determine the outcome, our model predicts the results of each state’s electoral votes, and then combines these results to assign a national winner.

Our predictions are based on our own analysis of polling data, but also incorporate the fact that different kinds of errors can affect individual states. For example, our model takes into account whether a state’s registered voter polls are subject to biases in one direction or another. This produces a set of 10,001 simulations, each of which represents a different path that the election could take.