Citizenship Question on the Census will Undercount Immigrant Communities


Yesterday, Commerce Secretary Wilbur Ross directed the U.S. Census Bureau to add a citizenship question to the 2020 Census—a move that may have serious consequences for Asian American and Pacific Islander (AAPI) and immigrant communities across the nation.

Every ten years, the Census counts everyone who lives in the United States—regardless of citizenship status—and plays a consequential role in American public life. The Census count determines how congressional seats are apportioned to states, dividing the 435 seats in the House of Representatives to states based on total population. The Census count informs how federal dollars are allocated—to a tune of more than $675 billion, for programs including Medicare and Pell grants. Beyond the federal government, Census data informs business investments, community decision-making, and other services provided at the state and local levels. And Census data drives much of the work here at AAPI Data.

Thus an accurate Census is critical to the lifeblood of American democracy, ensuring that we understand what our communities look like and what the needs of facing them are. Even before the inclusion of a citizenship question, AAPI communities were already at risk of being undercounted—from the introduction of a new online response option to existing language barriers. About 1 in 5 Asian Americans and 1 in 3 NHPIs live in hard-to-count Census tracts, which require more extensive follow-up to collect a response. More than 1 in 3 Asian Americans are limited English proficient, and approximately 3 in 4 Asian Americans speak another language at home other than English.

By adding a citizenship question to the 2020 Census, the Department of Commerce will risk losing an accurate count of all residents in the United States, including AAPIs. As Karthick Ramakrishnan, Director of AAPI Data, points out on Twitter, the Commerce Department is turning the entire Census into an experiment with the inclusion of this question—instead of using a test on a smaller sample, as most new questions must do before being included in the decennial count.

Turning the 2020 Census into an experiment will run “the risk of having the Census fail in its only Constitutional mandate, to accurately count the number of people in the United States. People. Not Citizens,” Ramakrishnan writes.

Furthermore, in today’s xenophobic political landscape—largely driven by policies and rhetoric from the Trump Administration that Secretary Ross serves in—trust in the government from communities of color and immigrant communities has dropped. There are more immigrants from the Asia-Pacific region coming to the United States today than anywhere else in the world, and 1 out of every 7 Asian immigrants are undocumented—the fear of deportation continues to plague Asian immigrant communities. As raids continue within immigrant neighborhoods and the looming threat of the end of the Deferred Action for Childhood Arrivals (DACA) program persists, the expectation of these communities to readily hand detailed, personal information to the federal government in 2020 already poses a threat of undercounting.

By adding a citizenship question, the threat of undercounting is fundamentally exacerbated. As the Census is expected to provide an accurate count of all residents, and not just citizens, a decrease in response rates from immigrant communities stemming from fear of the citizenship question will, as Ramakrishnan puts it, risk having the Census fail in its only constitutional mandate.

Every organization—from corporations and businesses looking to invest, to direct-service non-profits helping their community—relies on data to drive their work. Bad data would harm everyone—and as many of these organizations depend in some way on Census data, the citizenship question would threaten the accuracy of Census data in 2020 and have consequences across sectors.

Preparations for the 2020 Census have been underway for years—and while underfunding of the Census Bureau has forced it to cancel field tests that are critical to ensuring an accurate count ahead of 2020, this misguided inclusion of the citizenship question may result in a serious undercount that will threaten immigrant communities in the decade ahead.