"Race" and such...

Bosley_J (BosleyJ@ORE.PSB.BLS.GOV)
Mon, 4 Dec 1995 15:22:00 EST

Here at BLS we have recently completed the preliminary analysis of a survey
that was taken as a supplement to the US Census' Current Population Survey
(CPS) last May, and the results may be of interest to members of this list.

The Office of Management and Budget asked BLS and Census to help them decide
on possible changes in the way race and ethnicity self-identifications are
made for use in administering federal programs where these identifications
are important--for example, enforcement of the Voting Rights Act. Certain
population groups have over the years registered some concerns about the
categories that have been used for quite some time now; for example, some
people would like to self-identify as "multiracial." I will post some
highlights below that seem most relevant to recent discussions on this list.
Interested people can obtain copies of the preliminary report by calling
202-606-7375; ask for USDL 95-428, "A CPS Supplement for Testing Methods of
Collecting Racial and Ethnic Information."

Focal issues the survey was intended to address were:

(1) What is the effect of having a "multiracial" category among the
(pre-existing) list of races?

(2) What is the effect of adding "Hispanic" to the list of racial
categories; and

(3) What are people's preferences for alternative names for racial and
ethnic categories?

Four "panels" (a 2 x 2 design) allowed examination of the first 2 issues;
data related to issue 3 came from all panels.

--In two panels, respondents were given the opportunity to identify
themselves as "Hispanic" independently of selecting a racial identification;
in the other 2 panels, "Hispanic" was included as a racial category. Less
than 10% of the sample selected this category in all panels combined, but
the category was less likely to be selected when it was folded in with other
"races." On the other hand, of those who self-identified as "Hispanic" in
some fashion, a majority thought the category should be "racial."

--The frequency of self-identification as "white" was affected by whether
"Hispanic" or "multiracial" response alternatives were available for use.
When Hispanic is presented as a separate but *non-racial* category, about 4%
more respondents identify themselves as "white." When "Hispanic" *is*
offered as a racial category, about 8% of respondents use that category,
drawing from "white" or "something else" categories where it is not
available.

--Availability of "multiracial" as a category decreases the frequency of
identification as American Indian, Eskimo, or Aleut. Neither Hispanic nor
multiracial labels for "race" affected overall proportions of Black or Asian
respondents.

--Terminology: Of those classified as "Hispanic," 58% preferred to be
*called* "Hispanic;" Latino" or "of Spanish origin" came in second at about
12% each (no preference got 10%)

"White"--62% preferred to be called "white." "Caucasian" and "no preference"
were tied for second at 17%.

"Black"--44% preferred "Black;" 28% preferred "African-American;" 12%
preferred "Afro-American;" 3% preferred "Negro;" 1% preferred "Colored;" 2%
wanted "some other term" and 9% stated "no preference."

"American Indian"--50% preferred "Native American" while 37% preferred
"Native American." (But there were very few people in this group--only about
1% of the total sample)

"Multiracial"--(About 1.5% of the entire sample)--28% preferred to be called
"multiracial;" 16% preferred "mixed race;" 14% preferred "some other term."
And 28% had no preference.

Interesting data I think on how Americans characterize *themselves.*

John Bosley
Bosley_J@bls.gov