Asked and Answered: Race/Ethnic Coding
To: Arthur Kennickell:
Why are the race categories of American Indian/Eskimo/Aleut and Asian and
Other combined in the public use tape? With respect to financial behavior,
it would be informative to be able to look at Asians and make comparisons to
5909 Y4. Are you American Indian, Asian, Hispanic, black, white, or other
race? (SHOW CARD 16)
For public use file, American Indian/Eskimo/Aleut (1) and Asian (2) and
OTHER (-7) are combined in -7.
Would it be possible in future SCF to have separate variables for race
(white, black, asian, american indian/eskimo/aleut)
and ethnicity (hispanic,
to enable more careful controls for the demographic
By having separate variables, it would be possible to look
at nonhispanic whites, etc.
Thanks for your response.
Subject: Re: Survey of Consumer Finances Working Group, The Ohio State
Date: Fri, 19 Apr 1996
From: Arthur Kennickell
There are two reasons have aggregated the “minority” categories.
First, the number of cases in some cells (e.g., Native American) is
too small to be analyzed separately. Normally, I would be happy to
let analysts make this decision on their own.
The second reason has
to do with privacy and the restrictions that we have to live with as a
condition of being allowed access to SOI data for sampling.
restriction binds us strongly to do a range of things to make it
virtually impossible for someone to identify respondents in the
survey. We do a number of things that we can talk about (such as
collapsing the race/ethnicity categories), and some that we can’t talk
In the latest release of the 1992 SCF, we responded to users’
requests to expand the occupation and industry data and the geographic
coding. To do so, we had to review the data intensely to make sure
that we could compensate for the potentially identifying information
that was released by blurring the data in other ways. I am very
comfortable with what we were able to do in that case.
It is possible
that we could do something similar in the future for race/ethnicity.
My feeling is that once the federal standard for race/ethnicity is
determined (you may know of a large OMB effort in this regard that
appears to be far from conclusion at this point), we should reexamine
Return to Asked and Answered