Lesson 11: Forecasting the 2024 US Presidential Election
March 12, 2024
Review:
- Singapore non-strata housing models
- Final Exam at 14:30-16:00 on Monday, 6 May in B.5.12/B.5.13
- Exam 2 next week
Presentation:
- Election Forecasting
- Recent US Elections
- 2008 Obama defeats McCain (+6.8%)
- 2012 Obama defeats Romney (+2.8%)
- 2016 Trump defeats Clinton (-2.1%)
- 2020 Biden defeats Trump (+4.5%)
- Nate Silver and fivethirtyeight.com
- 2016 forecast and interview (PBS, 07:00)
- Recent article on economic sentiment and the 2024 election
- Demographic data from the US Census Bureau
- County Level voting histories by State (2020, 2016)
- Swing States
- State Assignments – one per student
- Recent US Elections
- Colorado data
- Demonstration with recent data
- Demonstration of downloading new US Census Bureau data
Activity:
- Begin assembling data to generate forecasts for your assigned State
- Population and demographics from Census
- Voter Turnout and % Trump vs Clinton (2016) and % Trump vs Biden (2020)
- Polling results (save this for later)
- Run correlation and/or initial regression analysis to identify and assess relationships
- Use Trump vote % as dependent variable
Assignment:
- Test various demographic variables to include in a forecast model
- Construct an initial multivariate model to estimate Trump vote % by County in your assigned state