CONTEXT EFFECTS FOR CENSUS MEASURES OF RACE AND HISPANIC ORIGIN

This paper reports on the results of a split-ballot experimentconducted in 1987 to test alternative versions of the decennialcensus long form. Two forms were randomly assigned and self-administeredin group sessions involving a total of 515 respondents. Theorder of race and Hispanic origin items was experimentally manipulated.The standard long form asks race, then Hispanic origin. Theexperimental form reversed the order of the items in order toreduce perceived redundancy, and to create a more restrictedframe of reference for the race item. The objectives of thecontext manipulation were (1) to reduce item nonresponse forthe Hispanic origin item, and (2) to reduce reporting of "Otherrace" by Hispanics in the race item. Objective (1) was met.Objective (2) was met for Hispanics born in a U.S. state, butnot for immigrants. The results are interpreted as reflectinga process of acculturation that affects how Hispanic respondentsapply U.S. racial categories "white" and "black" in the census.