The U.S. Census Bureau has stated that enhancements made to the structure of the 2020 census questions and the processing of responses have contributed to a notable rise in the number of individuals identifying as multiracial in the United States. This assertion comes in response to critiques suggesting that this increase may primarily stem from a statistical anomaly.
According to the Census Bureau, which stands as the nation’s principal entity for statistical data, millions more individuals indicated multiple racial identities in the 2020 census compared to prior data from 2010. “The 2020 Census race and ethnicity results reveal the complex way people identify themselves,” the bureau highlighted in a recent statement.
This response follows a research paper published last month by two sociologists from Princeton University, who contended that the remarkable 276% surge in multiracial classifications from 2010 to 2020 chiefly resulted from altered classification techniques employed by census officials rather than reflecting substantial changes in racial or ethnic identity. For the first time, the census allowed participants to specify their family origins, significantly influencing the categorization process.
From 2010 to 2020, the percentage of individuals identifying as two or more races increased from 2.9% to 10.2% of the total U.S. population, with a particularly prominent rise among Hispanic individuals. During the same period, the proportion of the population identifying as “white alone” fell from 72.4% to 61.6%.
Some analysts have suggested that creating a new dataset using the 2010 categorization method could facilitate a clearer, “apples to apples” comparison of demographic transitions. However, the Census Bureau has countered that adopting the 2010 methodology for comparison would not yield accurate results, as the 2020 census included more write-in options and captured a broader spectrum of responses than its predecessor.
Paul Starr, one of the Princeton sociologists who authored the paper, expressed in an email that the census bureau should make its data accessible to allow researchers to assess the influence of methodological changes independently. Starr also noted that the bureau did not adequately address their concerns regarding the classification of “origins” as synonymous with racial identity, nor did they clarify the relationship between national origins and specific race categorizations.
Andrew Beveridge, a sociology professor at Queens College and the CUNY Graduate School and University Center, remarked that while Census Bureau officials have characterized the new methods as an advancement, they have yet to provide clarity on the extent to which the changes contributed to the rise in multiracial identification. He further indicated that this situation was reminiscent of “survey method malpractice,” as described by a peer in the field.
