New research reveals a simple predictive model using state-level presidential approval and economic conditions accurately forecasted the winner of the 2020 U.S. Presidential election. Using readily available data on presidential approval ratings and state unemployment rates, this study demonstrates how these variables alone can anticipate electoral outcomes with remarkable accuracy.
Data & Methods: The analysis leverages monthly approval poll data from major news organizations and state-level economic indicators (unemployment claims) sourced from the Bureau of Labor Statistics. By combining time-series data on approval ratings spanning 2019-early 2020 with state unemployment trends leading up to the election, we constructed a predictive model.
Key Findings: The model successfully predicted Joe Biden's victory in all seven swing states without complex factors like campaign spending or voter demographics. Accuracy improved when integrating approval ratings earlier (January-March 2020) rather than later polls.
## Why It Matters
This straightforward approach suggests election forecasting can be grounded in minimal data, offering policymakers and analysts a transparent method to gauge re-election prospects.
| Forecasting the 2020 Electoral College Winner: The State Presidential Approval/State Economy Model was authored by Peter Enns and Julius Lagodny. It was published by Cambridge in PS in 2021. |
