UCLA

In this study, we set out to better understand the voter behavior of di↵erent demographic partitions of Los Angeles County (LAC). We developed a methodology to enrich and analyze LAC voter data by joining it with US census data and modelling vote by mail rates and turnout rates using general additive models. In doing so, we were able to forecast where in LAC we expect to see the highest rates of turnout by vote by mail and turnout in person for the 2020 presidential election. These findings are timely: In 2020, LAC will roll out a new and novel voting system called Voting Solutions for All People. The findings of the current study provide guidance on how to best allocate resources to high-turnout populations and target outreach to low-turnout populations in preparation for the first election under the new system. This paper presents a combination of visualizations, predictive models, and summary statistics that can inform LAC of where to focus outreach to expand the electorate and build a more representative electorate while also improving preparation in areas where turnout has been historically high.