class: left, top, title-slide # What survey data and election results can tell us about voter suppression ## Who loses votes? Can this bias election outcomes? ### <strong>G. Elliott Morri</strong>s <br />Data journalist <br /><em>The Economist</em> ### <class=‘date’>October 22, 2020 <br> Prepared for a guest lecture to Bryan Jones’s class, UT Austin</class> --- class: center, inverse, middle # What if everyone voted? --- # Guiding questions -- ## 1. How many Democrats and Republicans are there? Given data constraints, we're really asking: How many Clinton and Trump voters are there? -- ### 2. How are they distributed geographically? The answer lets us assign Electoral College votes. --- # Data -- ## 1. Cooperative Congressional Election Study (CCES): A survey of 64,000 Americans Includes demographic data and 2016 vote choice for 40,000+ validated voters -- ## 2. American Community Survey (ACS): A Census Bureau survey of 175,000 Americans Includes the same demographic data as the CCES 32,640 “cells” --- # Method -- ## 1. Train a predictive model on CCES data - Multi-level logistic regression - Predict vote choice with: age, gender, race, education, region and interactions between them -- ## 2. Use the model to predict voting habits for every eligible American Via “post-stratification” on the ACS --- # ACS Post-stratification -- ### 1. Each "cell" (row) is one "type" of person - One cell for white men ages 18-30 without college degrees who live in the Northeast - Another for white men ages 18-30 without college degrees who live in the South - Another for non-white men ages 18-30 without college degrees who live in the Northeast - etc. -- ### 2. We know how many voters in that "cell" live in each state -- ### 3. So we can say that x and y% of each "cell" vote for Clinton or Trump, then add up - For example, a Latino female age 18-30 with a college degree in Texas is 85% likely to vote for a Democrat for president (White man 65+ is 80% Republican) --- # Results <img src="figures/states_demography.png" width="3968" /> --- # Results <img src="figures/votes_bystate.png" width="4056" /> --- # Results: If everyone voted <img src="figures/everyone_votes.png" width="80%" /> --- class: center, inverse, middle # What does this tell us about voter suppression? --- # Voter suppression -- ### - We can modify the percentage of each group that turns out to vote, then re-predict the election - What if only all whites vote? - All non-whites? - Whites without degrees? Etc. -- ### - Democrats do better when non-whites turnout; Republicans have a vested interest in keeping turnout rates low - Especially in southern states with large minority populations - Or on college campuses --- # Suppression of whites votes <img src="figures/nonwhites_vote.png" width="80%" /> --- # Suppression of POC votes <img src="figures/whites_vote.png" width="80%" /> --- # Considerations -- ## What this doesn’t tell us: - That Clinton/Trump/etc would have won if certain x, y or z restrictions had been put in place - Downstream effects (AKA party positions and coalition changes) -- ## The balancing act: - There are a ton of white, non-college educated voters in the Midwest that tilt national scales if we increase turnout - Especially because increases in turnout are not uniform - And because of their geographic distribution, small relative increases in white turnout can tip the Electoral College to Republicans (see: 2016) - But on the other hand, some organizations are explicitly targeting non-whites and young voters for turnout purposes --- class: center, inverse, middle # Application to 2020 --- # Application to 2020 -- ### 1. Use YouGov data and MRP for turnout, vote choice, and vote method ### 2. Train models on NC ballot rejections to predict rejection likelihood ### 3. Calculate vote rejections for all absentee votes ### 4. Tally up lost votes for each party --- # Model NC rejection rates <img src="figures/nc_rejections.png" width="80%" /> --- # Calculate partisan impact <img src="figures/mrp_rejections.png" width="70%" /> --- class: center # Thank you! <br> <br> #### Website: [gelliottmorris.com](https://www.gelliottmorris.com) #### Email: [elliott@thecrosstab.com](mailto:elliott@thecrosstab.com) #### Twitter: [@gelliottmorris](http://www.twitter.com/gelliottmorris) <br> <br> <hr> _These slides were made with the `xaringan` package for R from Yihui Xie. They are available online at https://www.gelliottmorris.com/slides/2020-10-22-ut-austin/_