Skip Navigation

A Critical History of the U.S. Census & Citizenship Questions

Summary: The Trump Administration is relying on history to defend the 2020 Census citizenship question. New Brennan Center research shows this historical narrative is wrong.

Published: April 2, 2019

On March 26, 2018, Commerce Secret­ary Wilbur Ross announced that the 2020 Census would ask about the citizen­ship status of every person in the coun­try. Since this announce­ment, the Trump Admin­is­tra­tion has relied heav­ily on broad histor­ical argu­ments to defend Secret­ary Ross’s decision. In both the courts of law and the court of public opin­ion, the Admin­is­tra­tion has repeatedly insisted that Secret­ary Ross’s “citizen­ship ques­tion” has a deep histor­ical pedi­gree stretch­ing back more than two centur­ies. This histor­ical narrat­ive, however, is mislead­ing where it is not outright false.

This Article—the first schol­arly rejoin­der to the Trump Admin­is­tra­tion’s use of history in the citizen­ship ques­tion case­s—­demon­strates that the Admin­is­tra­tion’s histor­ical account is flawed in at least two signi­fic­ant respects. First, the census has never asked for the citizen­ship status of every­one in the coun­try. Secret­ary Ross’s proposal is there­fore histor­ic­ally unpre­ced­en­ted.

Second, the Admin­is­tra­tion relies on an impov­er­ished view of census history to suggest that Secret­ary Ross can find a histor­ical warrant for his decision in citizen­ship ques­tions that were posed only to small subsets of the popu­la­tion at vari­ous points in Amer­ican history. Viewed in context, these citizen­ship ques­tions origin­ated as sporadic compon­ents of an approach to census-taking that the Census Bureau long ago rejec­ted as incom­pat­ible with its found­a­tional, consti­tu­tional goal of actual enumer­a­tion. These early citizen­ship ques­tions were part of an increas­ingly sprawl­ing census that was attempt­ing—with mount­ing diffi­culties—to pursue two object­ives at once: first, count­ing every­one; and second, collect­ing addi­tional inform­a­tion that was used for a mixture of collat­eral stat­ist­ical, polit­ical, and economic object­ives. In the wake of the 1950 Census, the Census Bureau rejec­ted this older paradigm of census prac­tice in favor of a radic­ally differ­ent model. Indeed, once social science tech­niques like sampling gran­ted the Bureau the tech­nical abil­ity to identify and remedy substan­tial prob­lems in its approach to the enumer­a­tion, the Bureau over­hauled its approach dramat­ic­ally. As part of this over­haul, the Census Bureau rebuffed citizen­ship ques­tions as viable items for any census survey designed to obtain a complete count of the popu­la­tion. Due to inter­ven­ing devel­op­ments in the Amer­ican immig­ra­tion envir­on­ment, these ques­tions have never been deemed fit to return to the complete-count form; they have been confined solely to sample surveys sent only to subsets of the popu­la­tion.

This Article will show that under these circum­stances, the Admin­is­tra­tion cannot plaus­ibly invoke census history to justify its current decision to add a new, untested citizen­ship ques­tion to the 2020 Census under either the Enumer­a­tion Clause or the Admin­is­trat­ive Proced­ure Act. History instead creates a broad presump­tion against Secret­ary Ross’s proposal, one which the Admin­is­tra­tion has not succeeded in rebut­ting.