Aside Posted on
New from Weldon Cooper Center for Public Service at University of Virginia
Memo on 2020 Census Data Distortion
Differential privacy is a new mathematical procedure devised by the Census Bureau that will be applied to the 2020 Census data before it is released to enhance data privacy protection. Our analysis indicates that data accuracy at the sub-state level will be sacrificed as a result of this new approach to data release. This inaccuracy may lead to misallocation of funds, poor capacity for planning, substandard service provision, and a competitive disadvantage in economic and workforce development.
Find it here: https://sdcclearinghouse.files.wordpress.com/2020/01/censusdistortionprogram_vagovernor_2020-01-23.pdf
Aside Posted on Updated on
New from Minnesota Population Center IPUMS
Differentially Private Census Data
To help data users assess the impacts of differential privacy, the Census Bureau has released 2010 demonstration data products, which supply differentially private versions of 2010 data. To facilitate comparisons between the original and differentially private 2010 data, we provide both versions joined together in a simplified format through the links below.
Find it here: https://www.nhgis.org/differentially-private-2010-census-data
Aside Posted on Updated on
New from California Department of Finance
Review of U.S. Census Bureau’s 2010 Demonstration Data Products
Simple online app comparing population characteristics with and without Differential Privacy processing.
Find it here: https://shiny.demog.berkeley.edu/eddieh/DP2010DemoData_CA/
The steering committees of the Census Information Centers (CIC), Federal-State Cooperative for Population Estimates (FSCPE), and State Data Centers (SDC) transmitted a letter, on Nov. 27, 2019, to Census Director Dillingham. This letter outlines our questions and concerns regarding the Census Bureau’s move to a new Differential Privacy Disclosure Avoidance System (DAS).
When: Wed., December 4, 2:00 Eastern (11:00 Pacific). The webinar will be 90 minutes.
Presentation slidedecks are now available:
- Census implementing new disclosure avoidance methods with 2020 Census public data products (Beth Jarosz)
- Overview of Differential Privacy, effects on data accuracy, examples based on Census’s decennial demonstration data tables (David Van Riper)
- Proposed data products, what tables will be available (Jan Vink)
- How you can review the demonstration data tables now online (Beth and Jan)
- Suggested next steps, and opportunities for feedback (Todd Graham)
Presenters for this webinar are:
- Jan Vink, research specialist at Cornell Program on Applied Demographics
- Dave Van Riper, spatial analysis director at Univ of Minnesota Population Center
- Beth Jarsoz, demographer at Population Reference Bureau
- Todd Graham, principal demographer at Metropolitan Council of the Twin Cities.
Thanks to Jenn Shultz, Pennsylvania State Data Center for hosting this webinar meeting.
From US Census Bureau, Customer Liaison & Marketing Services Office:
In a new Research Matters blog, Balancing Privacy and Accuracy: New Opportunity for Disclosure Avoidance Analysis, we discuss the just-posted new 2010 Demonstration Data Products designed to help data users better understand the 2020 Census disclosure avoidance system (DAS) and its impact on data quality and privacy protection.
You can access the data via the new 2020 Census Data Products page on census.gov. You can also get there by searching “2020 data products” from the census.gov home page.
Please note that the data are only released via the FTP and in summary file format. Data will not be released on data.census.gov. As a data intermediary, you may be asked to provide or extract data in a different format. So we encourage you to take a look at the summary file format to become familiar with the data.
Data user feedback will help the Census Bureau’s senior leadership make informed decisions about the final structure of 2020 data products and the fine-tuning of the disclosure avoidance system. Because decennial census data are widely used in ways that go beyond the Census Bureau’s needs, we want to ensure that they are fit for as many data users’ needs as possible. Feedback on the demonstration files should be sent to firstname.lastname@example.org
Please note that there will be a live Webcast of a National Academy of Sciences Center for National Statistics (CNSTAT) workshop on this topic on December 11-12, 2019. We’ll post more information on our Web site closer to the date.