Presidential Nomination Forecasts: Iterative Updating for Error Reduction

• In Event: Forecasting Political Outcomes with Different Methods

Abstract

A number of papers have forecast the results of the presidential primaries with a high level of success (Mayer 1996; Adkins and Dowdle 2001; Steger 2008; Dowdle et al. 2016). Besides early popularity polls, most studies typically concluded that elite endorsements and financial resources played a significant role in predicting nomination outcomes (Dowdle et al. 2009).

However, forecasters have had less success in predicting Democratic races with crowded fields (Steger 2015). We plan to utilize a pooled OLS model that will create a forecasting model using the 1980-2016 presidential nomination contests and apply it to the 2020 Democratic presidential nomination contests. This model will include both measures of pre-primary resources (e.g. a candidate’s standing in the preprimary Gallup polls, percentage of party endorsements and fundraising success) and early campaign momentum (e.g. finishes in the Iowa Caucus and New Hampshire Primary).

Authors

• Andrew J. Dowdle, University of Arkansas, Fayetteville

• Karen Denice Sebold, University of Arkansas, Fayetteville

• Randall E. Adkins, University of Nebraska at Omaha

• Wayne P. Steger, DePaul University