I am a data journalist at The Economist based in Washington, DC. I write mostly about American politics and elections, usually by engaging in a close study of political science, political polling and demographic data. I am responsible for many of the paper’s election forecasting models, including our 2020 US presidential election forecast.
I am writing a book for W. W. Norton on public opinion polling — its history, influence, successes, failures, and future — and why it is a crucial tool for a healthy democracy. Sign up for my newsletter to get updates as publication day approaches. No Margin For Error is due out in the fall of 2021. You can read a short blog post about the project here.
I used to write a lot over at my old blog thecrosstab.com, but nowadays I do most of my blogging via my newsletter. You can read it and sign up at thecrosstab.substack.com. One day I may transfer that old content here.
I also post a lot on Twitter: Follow @gelliottmorris
The work I am perhaps best known for is my election forecasts, which have enjoyed varied success since I began blogging about political statistics in 2015. These days, I am interested in how we can improve over the popular aggregation-based election forecasting methods by incorporating raw polling micro-data into our models. Such an approach can make crucial improvements in state-level polling, decreasing the biases in our national probabilistic estimates and producing more reliable day-to-day forecasts.
I received my undergraduate degrees in government and history from The University of Texas at Austin in 2018. As part of my coursework I also studied statistics and computer science. I used to intern at the Pew Research Center and briefly produced statistical models for the election returns startup Decision Desk HQ.
Broadly speaking, I am motivated by an interest to better understand the world using computational social science and predictive analytics. I also have a firmly-held belief in the power of political polling to improve democracy. Got any leads? Drop me a line!
Writing + code
Forecasting the (2020) US elections • The Economist
We spent the lockdown sorting American voters into 380,000 distinct groups • The Economist
Who is winning the race for Westminster? • The Economist
When to pay attention to 2020 forecasts • The Economist
If everyone had voted, Hillary Clinton would probably be president • The Economist
Should political parties really let anyone run for president? • The Economist
The failure of gerrymandering • The Economist
Two Ways of Thinking about Election Predictions and What They Tell Us About 2018 • The University of Virginia Center for Politics
How Much Can the Youth Vote Actually Help Democrats? • The New York Times Upshot
My package for doing political data analysis in the R programming language,
politicaldata. • Link
My online course on R, “Analyzing Election and Polling Data in R”. • Link
- October 18, 2019: I will be speaking at The George Washington University on data journalism, political analysis and election forecasting. Slides
- September 30, 2019: I will be speaking at The University of Texas at Austin about my work for The Economist on what would happen in Americans elections if everyone turned out to vote. Slides
- August 30, 2019: I will be presenting at the annual meeting of the American Political Science Association on best practices in forecasting elections. Slides
- January 19, 2019: I will be presenting at the annual meeting of the Southern Political Science Association on the success of my forecasting model for the 2018 mid-term elections, and on what the mid-terms tell us about politics (and 2020).
Want to get in touch? My email address is elliott (AT) thecrosstab (DOT) com.