The quality of the race and Spanish origin information reported in the 1980 census.

This paper examines the quality of the race and Spanish origin information collected in the 1980 US census using data from the Content Reinterview Survery (CRS). Specifically this paper compares responses to the race and Spanish origin questions by identical persons. Race data in both surveys represents self-classification by persons according to the race they identify with. About 98% of the persons who classified themselves as white in the census also classified themselves as white in the CRS. The remaining 2% put themselves in the "other" category which includes write-in entries such as Wesort cosmopolitan interracial and mixed. Only 58% of the 88 respondents who reported themselves as American Indian in the census also did so in the CRS. 96% of the Asian and Pacific Islander population reclassified themselves as such in the CRS; the remaining 4% classified themselves as whites. Only 36% of those who classified themselves as "other" in the census reclassified themselves that way; 46% of that group reclassified themselves as white. About 89% of those who classified themselves as Spanish origin reclassified themselves that way. About 60% of those reporting as Spanish origin in the census say that at least 1 parent was from a Spanish-speaking country; an additional 26% report grandparents from a Spanish-speaking country. Of those reporting as Mexican origin in the census 91% also reported as Mexican in the CRS 4% reported as "other Spanish" and 6% reported as nonSpanish. Puerto Ricans showed a similar high reporting consistency as did Cubans 91% and 86% respectively. The "other Spanish" category was highly inconsistent; only 55% of those who gave that answer in the census also gave it in the CRS. In general the CRS showed that response to the 1980 Spanish origin question was quite satisfactory although there is some evidence that the total number of persons reporting Spanish origin in the reinterview was slightly greater than that in the census.