Optimized spatial information for 1990, 2000, and 2010 U.S. census microdata.

Area of Science:

• Demography
• Geographic Information Science
• Statistical Data Management

Background:

• Historical U.S. decennial census data lack precise spatial accuracy.
• Prior to recent improvements, spatial information in census responses was limited.
• Analyzing small geographic areas over time presents significant data challenges.

Purpose of the Study:

• To upgrade the positional accuracy of U.S. decennial census responses from 1990, 2000, and 2010.
• To create a linked dataset (OSCILAT) for improved spatial and longitudinal analysis.
• To enable exact tabulations of census data for arbitrary spatial units.

Main Methods:

• Developed and applied methods to enhance the spatial accuracy of historical census response data.
• Integrated 1990, 2000, and 2010 census data into a unified, spatially improved dataset.
• Stored the Optimized Spatial Census Information Linked Across Time (OSCILAT) in a restricted Federal Statistical Research Data Center (FSRDC) data warehouse.

Main Results:

• Successfully upgraded the positional accuracy for all responses in the 1990, 2000, and 2010 U.S. decennial censuses.
• The OSCILAT dataset significantly improves spatial accuracy and completeness for older census data.
• Enabled precise spatial and longitudinal analysis, including tabulations within 2020 block boundaries.

Conclusions:

• The OSCILAT dataset provides a foundation for more accurate historical demographic research.
• Enhanced spatial data facilitates precise measurement of change over time in small geographic areas.
• Access to OSCILAT, via the FSRDC system, supports detailed analysis of U.S. population dynamics.