11

Reflections and Ideas for Moving Forward

The final session of the workshop was prompted by a discussion between Co-chair Joe Hotz and Director Robert Santos on the first day of the workshop. Hotz explained that there were conversations and sidebars during the breaks and lunches that did not get captured in the workshop presentations. Consequently, to keep with the requirements of all discussions being on the record, this session was added to the final day; Ron Jarmin (U.S. Census Bureau Deputy Director) responded in his personal capacity.

The last session was raised as an opportunity to reflect on what the committee heard over the last two days and to offer their ideas for moving forward. Co-chair Elizabeth Garner (Colorado Department of Local Affairs) specified that it was important to move beyond some of the flaws and challenges raised with some use cases and identify actionable steps for developing best practices; partnerships that can be leveraged; who is not at the table; and what resources are needed to do this work.

Although the points raised by each planning committee member are listed below in order of speaking (see Box 11-1), there were several overarching themes. Chief among these was the equity impact on rural and other small areas that will be the “losers.” Options identified for moving forward included creating special tabulations, but the clear action item from several speakers was the importance for the Census Bureau to communicate in advance that data will not be available for these smaller areas. The committee discussed the difficulty it had in locating use cases and that it would be important for the Census Bureau to not only catalog these but also begin gathering them for 2030. Finally, committee members urged the Census Bureau to make the unperturbed data available to Research Data Centers.

Ron Jarmin responded on behalf of the Census Bureau (in his personal capacity) and thanked the committee and all presenters. He characterized that the two days had been great and are part of a continuing process. He was glad to hear that there has been some positive feedback on the new techniques introduced for the 2020 decennial data to protect privacy, and to hear that the Census Bureau has made some progress especially since the first workshop in 2019. He acknowledged, “We need to do more obviously.”

The challenge for 2030 is to rethink entirely what it means to release data. He discussed the tension involved in “publishing oodles of data and let people do whatever they want with it when most of those data are probably

never used in a computation.” Privacy-loss budget is spent implicitly and now explicitly to publish these data, and yet the Census Bureau knows from American Fact Finder and Census.gov that millions of American Community Survey estimates are never downloaded.

Jarmin questioned how to optimize what the Census Bureau produces in a way to “give people more information for the things they actually need and not giving the things that they claim they needed but never got used.” He concluded by emphasizing that there is an additional priority beyond identifying what data will be produced. He stated, “There’s also the priority not to violate the oath of confidentiality. We advertised for schools, hospitals, and roads in the 2020 Census, but we also had a really cool advertisement for privacy.” He noted the tension between privacy and granularity of data.

This page intentionally left blank.