MEDIA CONTACTS:
Jason Shevrin: [email protected], 202-994-5631
Timothy Pierce: [email protected], 202-994-5647
WASHINGTON (March 3, 2020)—Sen. Bernie Sanders is expected to earn the largest share of votes across 14 primaries and caucuses taking place today, according to the latest projection from the George Washington University Graduate School of Political Management (GSPM). Sanders’ closest competitors are former Vice President Joe Biden and former New York City Mayor Michael Bloomberg.
The GW research team based its projections off of updated data gathered Monday afternoon, after candidates Pete Buttigieg and Amy Klobuchar dropped out of the race, but before either of them endorsed Mr. Biden. The results vary significantly depending on which other details the team incorporated into the model.
A “basic” model, which considers Twitter mentions, cash on hand and endorsements, projects Mr. Sanders will garner 34.4% of the vote. Mr. Bloomberg would pull into a distant second place with 21.9% of the vote, Mr. Biden would be expected to take 15.2% and Sen. Elizabeth Warren is projected to earn a 14.1% vote share.
The researchers’ “momentum” model considers the same factors as the basic model but also takes into account the results from the South Carolina primary. The momentum model projects a tight three-candidate race, with Mr. Sanders in first place (29.5%) and Mr. Biden (25.8%) and Mr. Bloomberg (21.8%) close behind.
Both models project Mr. Sanders to gain the highest percentage of the vote, with a relatively small range of forecasted outcomes. There is more uncertainty when it comes to the projected performance of Mr. Biden and Mr. Bloomberg. The basic model is confident Mr. Biden will earn around 15% of the vote, with a small range of projected outcomes. However, the momentum model (which considers the most recent primary) not only projects a much higher vote share for Mr. Biden, but also demonstrates a much wider range of forecasted outcomes. Mr. Bloomberg’s possible vote share range also is quite large in both models.
“Large error estimates occur for Biden in the momentum model and Bloomberg in both models due to recent electoral results and financial reporting, respectively,” Meagan O’Neill, GW’s lead research scientist for the project, said.
The central premise of this election prediction project is to determine if social media activity can accurately predict electoral results. This project is the first of its kind that takes into account the number of times a candidate’s name is mentioned on Twitter. Through the first few state primaries and caucuses heading into Super Tuesday, the basic model has performed somewhat better than the momentum model. A complete explanation of the project’s methodology and performance is available online.
The 2020 election prediction model project is an initiative of GSPM’s Public Echoes of Rhetoric in America (PEORIA) Project, which strives to quantify how voters react to campaign messages. Model projections will be published periodically throughout the 2020 campaign.
-GW-