Forecasting Presidential Elections: A Comparative Study of Two Multivariate Models
Loading...
Authors
Migliore, Lauren
Issue Date
2008
Type
Thesis
Language
en_US
Keywords
Alternative Title
Abstract
The primary aim of this paper is to study the developing science of election
forecasting and determine the two best models created by scholars in this field and update
them with current data. After choosing models by Alan Abramowitz and Ray Fair, the
equations were updated using time-series data ranging from the early 20111 century until
the present. Both Abramowitz's and Fair's models ex post facto predicted elections on
average more accurately than the Gallup poll conducted closest to the election time and
are still considered valuable tools-although Fair's model performed more consistently
against the test of time. Using these updated regressions, I created conditional forecasts
for each model based on an array of economic and popularity rating indicators. Finally, I
created my own prediction-based on the results of the two models-that forecasts a
crushing defeat for the incumbent Republicans.
Description
39 p.
Citation
Publisher
License
U.S. copyright laws protect this material. Commercial use or distribution of this material is not permitted without prior written permission of the copyright holder.