Census Research Data Centers

Census Research Data Centers
Name	Census Research Data Centers
Abbreviation	CRDC (informal)
Formation	1980s–2000s
Purpose	Secure access to restricted microdata for research
Headquarters	Various locations across the United States

Census Research Data Centers

Census Research Data Centers provide secure, on-site access to restricted microdata from the United States Census Bureau, enabling scholars, analysts, and policy researchers to conduct statistical analysis using confidential datasets. They operate through partnerships with academic institutions, Federal Statistical Research Data Centers, and state agencies to grant vetted access under strict protocols tied to legal frameworks such as the Confidential Information Protection and Statistical Efficiency Act and the Privacy Act of 1974. These centers support peer-reviewed research with data that underlie major publications in demography, United States Department of Commerce reports, and interagency studies involving the National Institutes of Health and the National Science Foundation.

Overview

Census Research Data Centers are facility-based nodes that allow approved users to work with restricted datasets including decennial census microdata, American Community Survey files, and administrative records from the Internal Revenue Service, Social Security Administration, and Centers for Medicare & Medicaid Services. Hosted at partner sites such as Columbia University, Harvard University, University of Michigan, and Stanford University, these centers follow protocols comparable to those at the Inter-university Consortium for Political and Social Research and mirror access models used by the UK Data Service and the Canadian Research Data Centre Network. They serve investigators from institutions like the Brookings Institution, Urban Institute, and state research units that collaborate with federal agencies including the Department of Health and Human Services.

History and Establishment

The genesis of on-site research centers traces to collaborations in the late 20th century among the United States Census Bureau, leading universities, and funders such as the National Science Foundation and National Institutes of Health. Early prototypes were influenced by secure data enclaves at the Bureau of Labor Statistics and the development of the Federal Statistical Research Data Centers network. Legislative milestones—like provisions in the Confidential Information Protection and Statistical Efficiency Act—and institutional partnerships with entities such as Yale University, University of California, Berkeley, and Princeton University solidified the model. Over time, expansions incorporated administrative data from the Internal Revenue Service and health records from the Centers for Disease Control and Prevention and Centers for Medicare & Medicaid Services.

Access Policies and Eligibility

Access requires affiliation with an approved institution or sponsorship by investigators at sites including University of Pennsylvania, University of Texas at Austin, or University of Wisconsin–Madison. Prospective users undergo vetting by review boards analogous to Institutional Review Board procedures and must execute confidentiality pledges under statutes such as the Paperwork Reduction Act and the Privacy Act of 1974. Projects often receive review from advisory bodies similar to those at the National Academies of Sciences, Engineering, and Medicine or funders like the Robert Wood Johnson Foundation and Alfred P. Sloan Foundation. Clearance includes background checks used by partner agencies including the Department of Defense for certain linked administrative files.

Data Holdings and Services

Data holdings encompass restricted microdata from the Decennial Census, American Community Survey, and linked administrative sources like the Internal Revenue Service returns, Social Security Administration earnings records, and health datasets from the National Center for Health Statistics. Services include secure remote job submission, statistical software support with packages such as R Project, Stata, and SAS, and methodological consultations akin to offerings by the Institute for Social Research. Centers also facilitate data linkage projects with records from the National Science Foundation-sponsored initiatives, the National Longitudinal Surveys, and longitudinal studies connected to institutions like Rutgers University.

Security and Disclosure Avoidance Practices

Security measures mirror those at major statistical organizations such as the United Kingdom Office for National Statistics and the Statistics Canada research data model, with controlled access facilities, on-site monitoring, and output review to prevent re-identification. Disclosure avoidance employs techniques parallel to practices advocated by the National Academies of Sciences, Engineering, and Medicine—including data masking, aggregation, and more recently, methods inspired by differential privacy implementations used by the United States Census Bureau in 2020. Legal controls reference statutes such as the Confidential Information Protection and Statistical Efficiency Act and procedures consistent with guidance from the Office of Management and Budget.

Research Output and Impact

Research produced in centers has informed policy analyses at organizations like the Brookings Institution, academic publications in journals associated with American Sociological Association and Population Association of America, and studies cited by agencies such as the Department of Labor and Department of Health and Human Services. Investigations range from income and taxation work tied to Internal Revenue Service data to health disparities leveraging Centers for Disease Control and Prevention records, influencing debates in commissions such as those convened by the National Academy of Sciences. Collaborations with institutions like the Russell Sage Foundation and the Alfred P. Sloan Foundation have further amplified methodological advances and replication studies.

Criticisms and Controversies

Critiques address access equity, administrative burdens, and disclosure risks raised by scholars at Harvard University, University of California, Berkeley, and advocacy groups including the ACLU. Controversies intensified around the introduction of differential privacy by the United States Census Bureau and its effect on research accuracy, prompting debate in venues like the National Academies of Sciences, Engineering, and Medicine and commentaries from researchers at Massachusetts Institute of Technology and Stanford University. Concerns also focus on reliance on partner institutions such as Columbia University and University of Michigan for infrastructure, and on legal interpretations of statutes like the Confidential Information Protection and Statistical Efficiency Act.

Category:Statistical research