The Big Data Paradox and the Limits of Modern Polling

Why do modern polls fail to predict elections despite our increased ability to collect complex, high-volume data? Is big data always better data?

In the context of the most significant discrepancy between election forecasts and eventual outcomes in 40 years, Dr. Michael Bailey will explain the hidden biases in big data and engage in a discussion with the audience on the reliability of future polls.

Dr. Bailey is the Colonel William J. Walsh Professor of American Government in the Department of Government and McCourt School of Public Policy at Georgetown University. He is well known for his work on public opinion polling; statistical analysis of phenomena at the intersection of political science, policy, law and and economics; welfare policy; and the US supreme court.

Bailey is the author of two statistics books, Real Stats and Real Econometrics both focusing on statistical methods and the the concept of endogeneity (correlation does not imply causation). He was also coauthor of The Constrained Court: Law, Politics and the Decisions Justices Make from Princeton University Press.

Currently, Dr. Bailey’s work is focused on the concept of text as data, an emerging field of research in the computational social sciences. His new book, still unpublished, focuses on the future of polling.

Moderation by Dominique Cardon, Scientific Advisor, Digital, New Technology and Public Policy stream at Sciences Po's School of Public Affairs. 

This Master class is organized by the Digital, New Technology and Public policy stream of the School of Public affairs. The event is reserved for Sciences Po communities (students, teachers, employees). A Sciences Po card will be required to access the building.