Example: The ACS 2020 Non-Release Crisis. The American Community Survey (ACS) 1-year 2020 dataset was not released due to pandemic disruptions and nonresponse bias. The Census Bureau determined that response rates were too biased for a valid national 1-year estimate. Programs depending on ACS like housing, transit equity, poverty estimates were suddenly without their core inputs.
Play: Non-profit Organizations
Identify alternate datasets that answer the same question, even if they come from non-federal sources. Use KFF, PRB, CIDRAP, KidsData, and similar sources. https://www.kff.org/
Play: Modelling
Use modelling (MRP, small-area estimation, crosswalks) to rebuild the indicator from partial data.
These tools allow you to estimate outcomes for small populations or geographies. https://cran.r-project.org/web/packages/lme4/lme4.pdf
Play: Administrative Databases
Shift to administrative datasets that still exist and are consistent across years.
Consider USPS, NHTSA, hospital discharge, school enrollment, and utility arrears. https://cdan.dot.gov/
Play: Reconsider
Re-ask the underlying question with your community or funder to determine whether the indicator is still the right one. Use facilitated reconsideration methods. https://belonging.berkeley.edu/
Play: Replace
Replace national survey data with regional collaboratives or multi-state pooled data. These can answer similar questions with different inputs. https://www.nyc.gov/site/doh/about/press/pr2025/announce-northeast-public-health-collaborative.page
Play: Build Back Better
Build a community-owned data cooperative to re-collect the needed indicators from the ground up. Community Data Cooperatives (Civic Data Cooperative). https://civicdatacooperative.com/project/community-data-conversations/
Play: Synthesize
Social scientists used synthetic microdata derived from earlier stable years.
Play: Public-Private Alliance
The National Labor Exchange (NLx) is a public-private partnership between DirectEmployers and the National Association of State Workforce Agencies (NASWA). predicts BLS job openings a month in advance, leveraging cooperative data https://usnlx.com/
Anything else we should add?