Conducting federal surveys and providing statistical products have become increasingly challenging, primarily due to declining response rates and rapidly increasing costs. These challenges are further amplified by the increasing demand for timelier and more geographically detailed data.
The U.S. Census Bureau wants to understand how to leverage external data sources with traditional survey data. Furthermore, decision-makers want to know the effects on statistical data quality and standards of use resulting from incorporating external data. The Census Bureau tasked the Social and Decision Analytics Laboratory with addressing this question: "How can we know if external data are useful for federal statistical needs?"
The principal goal of this study was the development of an initial data framework that encompasses the theory and methods capable of capturing, repurposing and integrating sources of data. Two specific case studies were chosen to ground the data framework development.
The United States Census Bureau commissioned SDAL to analyze the long-term viability of untapped data sources.
The first case study considered measuring housing information based on locally available data. The second case study explored the use of state longitudinal education and workforce data. For both case studies, analyses were conducted to determine statistical properties, quality, accuracy, availability, and timeliness for each data source.
SDAL’s research demonstrated that external sources can provide valuable information to support other federal data-gathering efforts such as the American Community Survey. Our report—Leveraging External Data Sources to Enhance Official Statistics and Products—developed alternative estimates of housing and education information and provided recommendations for viable external data sources. This study is a vital first step in assessing the value of these previously untapped sources for use in federal statistics.