What is it about?

Some socially marginalized populations are more likely than others to be left out of the census count. This gives rise to disparities in federal funding and political representation for these groups and the communities they live in. A survey of Latino Immigrants in California about their willingness to self-respond to the census, respond to enumerator visits, respond to requests for information about neighbors, together with analysis of availability of administrative records, shows that adding a citizenship question to Census 2020 will result in serious undercount of this large population. The undercount will include, in addition to non-citizens, many citizens in their households and social networks.

Featured Image

Why is it important?

The analysis is important because it shows that efforts to compensate for pronounced and uneven levels of response to the decennial census occasioned by adding a sensitive question about citizenship cannot be "cured" by the standard procedures proposed for enumerating households that do not self-respond: reminder postcards, reference to administrative records, proxy interviews with neighbors, and imputation of household size and characteristics using the information from households that do respond. The analysis explains how inclusion of this highly sensitive question initiates a cascade of errors which gives rise not only to differential census undercount but, also undermines the reliability of data on demographic characteristics such as age and, ultimately, makes the sought-for information on citizenship unreliable as well.


I am pleased that the research methodology used in this study provides real-world insights on the dynamics of census enumeration. This was made possibly by reliance on a team of interviewers who are immigrants themselves conducting in-person interviews in friendly locales and an ethnosurvey approach which yielded data for quantitative analysis while, also, eliciting qualitative data from respondents about their perspectives. This provides a way to cross-check observations and yields useful insights for efforts to improve survey response.

Edward Kissam
WKF Fund

Read the Original

This page is a summary of: How low response among Latino immigrants will lead to differential undercount if the United States’ 2020 census includes a question on sensitive citizenship, Statistical Journal of the IAOS, June 2019, IOS Press,
DOI: 10.3233/sji-190505.
You can read the full text:




The following have contributed to this page