PHOENIX -- The U.S. Census Bureau is questioning whether there are better ways to ask Americans about their race and ethnicity. Researchers from the Census Bureau have now released the results of experimental race questions they tested in 2010 in the hopes of getting more accurate responses.
Back in 2010, the census form most Americans received asked them to choose as their race either White, Black or African-American, American Indian or Native Alaskan, Asian, or Hawaiian or Other Pacific Islander.
A separate question asked about Hispanic origin.
That's because the federal government thinks about Latino identity as an ethnicity, not a race. But that format has problems.
"Many Hispanics skip the question on race entirely," said Nicholas Jones of the Census Bureau during a panel discussion at George Washington University on Wednesday. "Or [they] respond by responding in the same manner they answered in the question previously on Hispanic origin: That they are Hispanic, Mexican, Salvadoran, etc., expressing responses which are ultimately classified as 'Some other race' in the census."
That meant that 37 percent of Latinos wound up classified as 'Some other race,' making that racial category the third largest in the country.
But a small portion of households got experimental forms that combined race and Hispanic origin into one question. Jones announced that experiment resulted in a drop in the 'Some other race' category, suggesting the accuracy of the responses were higher.
New York University Sociologist Ann Morning, who also participated on the panel, suggested there would be benefits to the census adopting the combined race and ethnicity question.
"We are moving towards using the label Hispanic or Latino as a de facto racial category, making the combined question one that tracks better with people's sense of how we would list major groups," Morning said.
Census researchers recommend more research on that question. They also recommend eliminating the term "Negro" from census forms.