2019 Symposium

7:30-8:30am

Registration and Breakfast

8:30-9:00am

Opening Session

Chair: Xiao-Li Meng, Editor-in-Chief, HDSR

Alan Garber, Provost, Harvard University

Francesca Dominici, Co-Director, Harvard Data Science Initiative

Robert Lue, Co-editor for Data Science Education, HDSR

9:00-10:30am

Differential Privacy for 2020 U.S. Census (I)

Chair: John Eltinge (U.S. Census Bureau)

After a background introduction by the Census Bureau, three research teams present their findings from comparing analyses of 1940 Census files with and without the differential privacy protection protocol planned for the 2020 Census. (Speakers are indicated by *.)

Introduction and Overview

John Eltinge

Background on Differential Privacy at the U.S. Census Bureau and 1940 Census Application

Michael B. Hawes* and Philip Leclerc* (U.S. Census Bureau)

The Effect of Differentially Private Noise Injection on Sampling Efficiency and Funding Allocations: Evidence from the 1940 Census

Quentin Brummet*, Edward Mulrow, and Kirk Wolter (NORC at the University of Chicago)

Assessing the Impact of Differential Privacy on Racial Residential Segregation

David Van Riper*, Tracy Kugler, Jonathan Schroeder, José Pacas, and Steve Ruggles (IPUMS, University of Minnesota)

Differential Privacy and Scrubbed Segregation

Brian Asquith, Brad Hershbein*, Shane Reed, and Steve Yesiltepe (W.E. Upjohn Institute for Employment Research)

11:00am-12:30pm

Differential Privacy for 2020 U.S. Census (II)

Chair and Moderator: Erica Groshen (Cornell University)

Scholars and experts from multiple disciplines will comment on the findings from Part (I), particularly focusing on implications for social science research and insights into the unavoidable trade-off between data privacy and data utility.

Panelists:

Cynthia Dwork (Computer Science, Harvard University)

Ori Heffetz (Economics, Cornell University and Hebrew University of Jerusalem)

V. Joseph Hotz (Economics, Duke University)

Salil Vadhan (Computer Science, Harvard University)

Ruobin Gong (Statistics, Rutgers University)

Mark Hansen (Journalism, Columbia University)

Paul Ohm (Law, Georgetown University)

2:00-3:45pm

Redistricting in 2020

Chair and Moderator: Moon Duchin (Tufts University)

The science of elections is very much in the news, from high-stakes redistricting lawsuits to reform initiatives creating independent redistricting commissions, to the adoption of alternative voting systems like ranked-choice voting. This session features four mathematicians (with backgrounds in applied analysis, probability, combinatorics, and geometry) working on the edge of theory and data to improve how we think about and practice redistricting.

Partisan gerrymandering versus geographic compactness
Dustin Mixon (Ohio State University)

Markov chain theorems for redistricting
Wes Pegden (Carnegie-Mellon University)

Visualizing and communicating data in alternative voting systems
Bridget Tenner (DePaul University)

Data and negotiation: Weighing complex and conflicting information in redistricting
Moon Duchin (Tufts University)

4:15-6:00pm

Data Science Education: 2020 Vision and Versions

Chair and Moderator: Dustin Tingley (Harvard University)

Leading educators from multiple universities and disciplines will discuss their visions and initiatives in providing general data science education on their campuses, followed by Q&A.

Panelists:

Michael Jordan and Ani Adhikari (Computer Science & Statistics, Berkeley)

Michael Franklin (Computer Science, University of Chicago)

Matthew Jones (History, Columbia University)

Joseph Blitzstein (Statistics, Harvard University)

Alison Gibbs (Statistics, University of Toronto)