National Spatiotemporal Population Research Infrastructure

Project Summary/Abstract With 494 billion data points and 62 million map polygons, IPUMS National Historical Geographic Information System (NHGIS) is the world’s largest publicly accessible population database and is an essential component of the shared data infrastructure for population and health research. NHGIS gathers area-level U.S. census data from diverse sources, formats them consistently, develops comprehensive standardized machine-processable metadata, and creates high-precision GIS boundary files describing the spatial units. To eliminate major obstacles for studies of small-area population change, NHGIS distributes geographic time series that link comparable data across multiple census years. The data are broadly accessible to health researchers through powerful dissemination tools that make it easy to navigate the intricacies of the U.S. statistical system, enabling rigorous and reproducible population health research. NHGIS reduces costs for population and health researchers by minimizing redundant effort, simplifying data access, and improving data reliability. NHGIS currently disseminates 12 terabytes of data per year to 81,900 investigators, with a new citation of NHGIS appearing in Google Scholar once every 14 hours. To meet the demands of a rapidly increasing user base and ever-expanding requests for more and improved data, this project has five major goals: 1. Add to core census datasets. The proposed project will incorporate new releases of American Community Survey (ACS) summary data with corresponding boundary files and broaden the coverage of existing NHGIS datasets with 1980 block boundary data for all metropolitan areas, Puerto Rico data from the 1980–2000 censuses, and ACS tables for detailed race/ethnicity/tribe/ancestry groups. 2. Add resources to support analysis of noisy 2020 census data. To protect privacy, the Census Bureau is adding noise to 2020 census data, causing some data to be unreliable in ways that are difficult to predict. This project will add three key resources to NHGIS to help researchers understand and account for the added noise: confidence intervals, Noisy Measurement Files, and data for optimized block groups. 3. Extend integrated spatiotemporal datasets. To broaden research opportunities, this project will extend NHGIS geographic crosswalks to cover more levels and years, extend time series tables to cover more years and subjects, and extend annual estimates of tract-level populations through 2027. 4. Improve data access. The project will give data users new options for code-based access to NHGIS data and metadata via our Application Programming Interface (API) and a Python package. 5. Expand and support the research community. The project will invest in extensive user support, training, and outreach, leveraging two new data user networks built with recent NSF and NIA funding to reach new, underrepresented, and early-career scholars. The project team will also develop new guides to key concepts in U.S. census data.