The Disutility of SEIR Model Forecasts During the COVID-19 Pandemic

What are students in the Policy Research Laboratory working on?


PRL Students

Forecasting models developed by academics are rarely used in high-stakes decision making. When COVID-19 hit, epidemiologists built curve-fitting models (SEIR models) to forecast viral trends, and these predictions were used to justify punishing lockdowns and school closures. These models generated *ridiculously wrong* point forecasts and uncertainty intervals. Inexplicably, these researchers are receiving $27.5 million from the Centers for Disease Control and Prevention for “disease modeling and forecasting to support more effective responses during public health emergencies.” To read the full paper written by group of students at UT Austin, click here.

Research With Us

The competitive Salem Center Fellowship provides fellows the unique experience of assisting and conducting data-driven and policy relevant research with the center’s Senior Fellows.