Census 2020 Undercount, Federal Programs, and Baltimore City Featured Image
December 7, 2018

Research Team: Mary Elizabeth Hughes, Associate Scientist, Bloomberg School of Public Health and Stuart Schrader, Lecturer and Assistant Research Scientist, Department of Sociology, Krieger School of Arts and Sciences

By definition, the Decennial Census must count every person living in the US on April 1st of a Census year. The Census Bureau thus mobilizes a vast effort to ensure no one is missed or counted more than once. Despite these efforts, the Census typically has a degree of miscount; the 1990 Census under-counted the US population by 1.6%, the 2000 Census over-counted the population by .49%, and the 2010 Census over-counted by .01%. Of much greater concern than miscount is differential undercount: when people with certain characteristics are less likely to be counted than persons without those characteristics. Differential undercount (usually referred to as “undercount”) is well-documented in the US; people most likely to be missed include race/ethnic minorities, young children, and persons who rent, rather than own, their homes.

Population counts derived from the Census are the basis for the distribution of billions of Federal program dollars to local areas. Area-specific program allocations usually depend on the population of an area or on the number of people who meet broadly-defined eligibility criteria (for example, the number of children). Any undercount has potential to shortchange areas of funds intended to improve residents’ lives. This project will conduct an analysis for Baltimore City, developing scenarios to show the effect of Census undercount on the funds allocated to Baltimore for specific Federal programs. These scenarios can then be used to inform Baltimore City’s Complete Count Committee and encourage Census participation.