Alternative Strategies for Grouping People into Resource Units: Measuring Poverty in the American Community Survey

Trent Alexander, U.S. Census Bureau
Misty L. Heggeness, U.S. Census Bureau
Sharon Stern, U.S. Census Bureau

Poverty experts and researchers are interested in developing Supplemental Poverty Measure (SPM) local area estimates using the American Community Survey (ACS). One challenge is that detailed information does not exist about the interpersonal relationships of individuals not related to the householder. The Census Bureau assumes those individuals are their own resource unit. Taking advantage of family interrelationship variables from the 2010 ACS Integrated Public Use Microdata Series (IPUMS), this paper investigates how poverty estimates change when additional interpersonal relationship information is available. It also analyzes what happens to poverty estimates when cohabiting partners, foster children, and other children unrelated to the householder are grouped with the primary resource unit. We find that having additional information about subfamily interrelationships of unrelated individuals does not change overall poverty estimates. However, including the cohabiting partner of the householder, foster children, and other unrelated children in the primary resource unit does influence poverty estimates.