3 Questions: Melissa Nobles on the U.S. Census

April 1 marks National Census Day, the official date of this year’s U.S. Census. To help put the census in context, MIT News spoke with Associate Professor of Political Science Melissa Nobles, whose teaching and research interests span the comparative study of racial and ethnic politics, and issues of retrospective justice. Her book, Shades of Citizenship: Race and the Census in Modern Politics (Stanford University Press, 2000), examined the political origins and consequences of racial categorization in demographic censuses in the United States and Brazil.

Q. You’ve noted in your book that the initial impetus for census-taking was political, and yet the earliest censuses also included racial categories. Why are race and ethnicity included in the U.S. Census?

A. Census-taking in the U.S. is as old as the Republic. The U.S. Constitution mandates that an “actual enumeration” be conducted every 10 years to allow for representational apportionment. The initial impetus for census taking was political. Yet the earliest censuses also included racial categories. The inclusion of these categories offers important insights into the centrality of racial and ethnic identifications in American political, economic and social life. This centrality continues to this day.

So, if the initial reason for census-taking was determining apportionment for representation, why did the earliest censuses include race? Representation depended on civil status — whether a person was free or a slave — and not on racial status. There were free colored persons, after all. Yet racial identification was combined with civil status in the census because race was a salient political and social marker. Censuses from the years 1790 through 1840 asked few questions beyond those related to population. 

These censuses variously counted free white males and free white females, subdivided into age groups; slaves; and all other free (colored) persons, except Indians not taxed. The 1840 and 1850 censuses were directly intertwined with debates about slavery. Data from the embattled, and largely discredited, 1840 census purportedly disclosed higher rates of insanity among free blacks, thereby proving that freedom drove free black people crazy. The 1850 census first introduced the category “mulatto,” at the behest of a southern physician, in order to gather data about the presumed deleterious effects of “racial mixture.” Post-Civil War censuses, which continued to include the “mulatto” category, reflected the enduring preoccupation with “racial mixing.” The introduction of “Chinese” and “Japanese” in the 1870 and 1890 censuses, respectively, also reflected growing concerns about Japanese and Chinese immigration.

Q. What are the census data on race and ethnicity used for today?

A. Twentieth-century racial and ethnic census categorization remained intertwined with the century’s core political and social issues: de jure and de facto racial segregation, the eventual establishment of legal equality, and immigration. In regards to segregation, categories and instructions for the censuses from 1930 to 1950 largely mirrored the racial status quo in politics and law. Southern laws defined persons with any trace of “Negro blood” as legally “Negro” and subject to all of the political, economic and social disabilities such designation conferred. Southern law treated other “non-white” persons similarly. Census categories and definitions followed suit, essentially bringing the logic of racial segregation into national census-taking itself. The early 20th century was also a period of tremendous demographic change. Census-taking too sought to register these changes, with the addition of several new categories.

The Civil Rights movement and resulting Civil Rights legislation dramatically changed the political context and purposes of racial and ethnic categorization. Today, racial and ethnic categorization supports civil rights and minority representation. Categorization today also seeks to capture more reliably our country’s growing diversity and the self-identification of citizens. Indeed, for the first time in American census history, the 2000 census allowed respondents to check more than racial category. Moreover, prior to the introduction of self-identification in the 1960 census, enumerators determined the person’s race by visual observation, based on the definitions provided in official instructions.

The history of race and ethnicity in the census is complicated, and at times deeply discomforting, but revealing. It clearly shows that categorization cannot be separated from its larger political and social context. Fortunately, census-taking in the 21st century reflects our country’s tremendous political and social progress.

Q. This year’s census form is much shorter than previous versions. Why is that, and does this mean that the data will be less complete/useful than in previous years?

A. Yes, the 2010 census form is much shorter than in previous years. The Census Bureau has introduced the American Community Survey (ACS), which basically asks the same questions that past censuses have. The Census Bureau will send the longer ACS to a random sample of addresses and send the shorter census form to all addresses. Most importantly, the ACS is conducted every year, rather than once every 10 years, making it quite useful to researchers, government and the general public because it provides the most current information.