Carnegie-Mellon University

Investigators:
Stephen Fienberg, Principal Investigator
William Eddy, Co-Principal Investigator

The Carnegie-Mellon node will conduct research on three basic issues of interest related to the conduct of censuses: privacy, costs, and response rates. First, the researchers will address the practical problems of insuring confidentiality and privacy while still producing useful data for public and private purposes. Secondly, the researchers will investigate the use of administrative records to create a basic census frame, saving the duplicated effort of gathering that same information repeatedly. They will also look at other possible uses of administrative records as part of the census. Thirdly, they will investigate the use of online data collection as a substitute for the traditional mail-out/mail-back census-taking. With respect to response rates, the researchers will conduct experiments that implement new ways of encouraging participation in an effort to reduce the decline in (or perhaps even increase) response rates.

This research will explore the potential for a significant reduction in the costs of conducting the 2020 census by demonstrating how information already collected by the government can serve as a starting point for the census in lieu of having the Census Bureau collect that information as part of the decennial census process. By learning to effectively use the Internet for censal data collection, the research should lead to a higher initial response rate, which should lower census costs overall, and create a more accurate count. Better methods for confidentiality protection and privacy notification not only will instill greater public confidence in the Census Bureau, but they also will contribute to better response rates and greater census accuracy. All of the technical statistical tools developed by the project will have other uses, both public and commercial. The project's educational and training initiatives aim to (1) prepare an educated citizenry on census and related matters, (2) use research issues under study at CMU and elsewhere as components in courses, and (3) train a new generation of students to enable them to work in agencies such as the Census Bureau in a diverse set of capacities, including the most technically demanding ones.