Becoming (Non)-White by U.S. Standards: Recategorization in the 2020 Census Race Question
摘要
The present study explores categorical shifts in the White Latino population in the U.S. Census following recent changes to the 2020 race question. We use ancestry data from the American Community Survey to simulate differences in how the U.S. Census Bureau may have coded race write-in responses both before and after changes to the 2020 design. We find that White Latinos who solely wrote Hispanic or Latino ancestries in the 2020 design have higher odds of multiracial classification than those who reported two racially distinct (multiracial) ancestries. In the pre-2020 question design, this association was reversed, indicating that entry into White categorical membership has become more rigid for White-identifying Latinos across question designs. Second, we observe a systematic shift among various Latino ethnic groups from White categorical membership to multiracial between designs. This shift is particularly notable for Cubans, who prior to the new race question, typically reported singularly as White in the Census. We refer to the reclassification of White Latinos from White categorical membership to multiracial membership as statistical bronzing. Finally, we find that first-generation, Spanish-speaking immigrants were more likely to be classified as multiracial in the 2020 design despite having a higher likelihood for White identification in the previous design. Native-born Latinos with limited Spanish-fluency and higher educational attainment levels appear more likely to achieve White classification in the new race question, supporting Bonilla-Silva’s ‘Honorary White’ thesis that situates assimilated White Latinos alongside traditional White ethnic groups in an emerging “tri-racial” U.S. racial classification system.