Sol works at the intersection of social science and machine learning. He has led technical teams since 2006 with regular executive communication/public outreach. His work has been featured in the New York Times, NPR, CBS, Wired, FastCompany, TechCrunch, and other outlets and podcasts.
During the 2020 election cycle, Sol was Chief Scientist at ACRONYM, where his team drove high-level strategic investment by correctly modeling the electoral importance of Georgia, and leveraged hundreds of field experiments to build ML systems to produce real-time estimates of persuasive messaging impact–an amazing collaboration with Minali Aggarwal, Sylvan Zheng, Dan Fankowski and James Barnes.
At Facebook, Sol built prototype and production ML models whose offspring power ads, page-recommendations, groups, site integrity, and social science/ethical AI efforts. His team built a differentially private data platform and released a data set reflecting more than an exabyte of media data. He also worked on approaches that blend ML and experimentation (heterogenous effect estimation and contextual bandits).
He also founded the Data Lab at Pew Research Center, where his team used deep learning to produce algorithmic audits of Google Image searches and news photos on Facebook, used ML to study inauthentic and automated behavior on Twitter, and used NLP to understand the role of ideology and power structures embedded in how members of congress use social media to communicate.
Sol’s doctoral work at Stanford University consisted of research on social influence and social structure in digital environments and probably the largest study of algorithmic bias in social media, which was later published in Science and has thousands of citations. He also did work to uncover bias in visual portrayals of African American candidates in political attack ads.
His work on election forecasts sparked public debates about the ethics and consequences of models projecting the results of elections, and prompted FiveThirtyEight to change their user interface to address concerns raised by the work.
Sol is published in Science, the American Political Science Review, the Journal of Politics, Public Opinion Quarterly, Communication Research, Radiology and Pediatric Radiology. His book, The Impression of Influence with Justin Grimmer and Sean Westwood was published by Princeton University Press.
PhD Political Communication
Intl MBA (Policy emphasis)
UC San Diego (GPS)
BA in Political Science
UC Santa Barbara
Largest ever social science data set, released under differential privacy
Should We Stop Paying Attention To Election Forecasts? NPR’s Science Friday, hosted by Ira Flatow & Elah Feder
Data Science across Academia, Industry, and Progressive Campaigns, with Dr. Solomon Messing. Social Media and Politics Podcast, hosted by Michael Bossetta
Technology in Political Campaigns and Activism 2020 CODE@MIT - Fireside Panel, hosted by Dean Eckles
Using Trumps Tactics Against Him with The Leaders of Barometer. The Great Battlefield Podcast, hosted by Nathaniel G. Pearlman
A Look Beyond Traditional Audiences + Models with Solomon Messing FWIW 2020 Debrief, hosted by ACRONYM
What color is Obama? These researchers examined reactions when his skin looks darker. Washington Post’s Monkey Cage
How election forecasts confuse Americans — and may lead them not to vote at all. Washington Post’s Monkey Cage
Did Trump win in 2016 because people are bad at probability? Washington Post
Study Finds Election Forecasts Lower Voter Turnout. Political Wire
Obama’s skin looks a little different in these GOP campaign ads. Washington Post
Study: Campaign Ads with Dark-Skinned Black People Appeal to White Racism. Atlanta Black Star
Does Facebook drive political polarization? Data science and research. Harvard Shorenstein Center
FiveThirtyEight Politics Podcast. FiveThirtyEight
Clinton’s Achilles’ heel in 2016 may have been overconfidence. Washington Post
Q&A with Solomon Messing of Pew Research Center’s Data Labs. Pew Research Center
Facebook Use Polarizing? Site Begs to Differ. New York Times
Don’t (just) blame Facebook: We build our own bubbles. Ars Technica
It’s The Frequency, Not The Size: Compromise & Credit Claiming in Congress. Mischiefs of Faction (Vox)
Today we published an Op Ed in @USATODAY calling on journalists, policymakers & public intellectuals not to play to clickbait horse-race commentary in the 2020 election. That game led some to stay home in 2016 and there's too much at stake this time. https://t.co/R0J1hg2vf2— Sol Messing (@SolomonMg) October 1, 2020
🚨🚨IT'S OUT🚨🚨On January 17, we launched one of the largest social science data sets ever constructed—a compact summary of nearly an exobyte of data. It’s meant to facilitate research on misinformation from across the web, shared on FB. pic.twitter.com/pZzsKSXAb7— Sol Messing (@SolomonMg) February 13, 2020
New post + thread: why Trump's chances are better than they now look https://t.co/j2lhem3CXw.— Sol Messing (@SolomonMg) June 28, 2020
Recent polling makes Trump's chances of winning in Nov look bad. Failing to account for COVID-19 + other issues, the polling itself could lower Dem turnout.
Social Media Researchers, you'll want to check this out -- perhaps the biggest data set describing sharing news and other content on social media ever released: https://t.co/V9ojjoJas6— Sol Messing (@SolomonMg) July 11, 2018
The media failed to question Cambridge Analytica’s marketing hype, which turned out to be mostly snake oil. Political scientists have been saying so for a long time. The UK ICO recently released an investigation that found... the same thing https://t.co/6AQt1gOCrZ pic.twitter.com/qsdBO9RGrS— Sol Messing (@SolomonMg) October 30, 2020
Cambridge Analytica/Facebook Thread: I find it surprising in the wake of breathless debate about this story that there's little talk of research ethics and the potential consequences thereof.— Sol Messing (@SolomonMg) March 18, 2018
When you only train your ML in the Bay Area... https://t.co/UVm9kBEVnE— Sol Messing (@SolomonMg) March 13, 2017
Does @FiveThirtyEight *move markets?* After their real-time forecast had GOP's odds of taking the House spiking at 60% at ~8:15PM, @PredictIt's odds on the GOP rose above 50-50, & **U.S. government bond prices saw brief spike of 2-4 basis points.** pic.twitter.com/MFZ7vb1yxq— Sol Messing (@SolomonMg) November 11, 2018
The one thing that’s clear about polling in 2020 is this: the tech bros saying we need to use deep learning to forecast elections understand neither deep learning nor elections.— Sol Messing (@SolomonMg) November 14, 2020