L’EMbeDS DS^3:“Identification and Semiparametric Estimation of Conditional Means from Aggregate Data” with Cory McCartan

The DS3 organizing group, alongside the L'EMbeDS Department of Excellence of the Sant'Anna School for Advanced Studies, is launching a new online seminars series devoted to frontier research in data science, its applications and its implications across disciplines.

The L'EMbeDS Data Science Seminar Series (L'EMbeDS DS3) hosts national and international scholars to discuss cutting-edge methodology, applications in economics, the social sciences -- and beyond, societal implications and governance issues. For announcements and further information please visit our web page and join our Google Group.

The third seminar will feature Cory McCartan (Penn State University), who will present a talk entitled: “Identification and Semiparametric Estimation of Conditional Means from Aggregate Data”.

>>Join the event via the following link

ABSTRACT:

We introduce a new method for estimating the relationship between two variables at the individual level when only aggregated data are available. A common application in the US is estimating the voting preferences of different racial groups from aggregate election totals and census data. I will discuss how common approaches rely on overly strong assumptions and will present a new identifying assumption along with a more flexible, machine learning-based estimation method. Time permitting, I will discuss an accompanying sensitivity analysis that can be used to probe the key identifying assumption.