Friday, June 15, 2007

Define Hispanic

I used to think the Nicene Creed was the most convoluted collection of words on the planet. Then I read the definition of Hispanic prepared for a study on racial profiling by police in Illinois. I believe it will give the Creed a run for its money.

Our dear Governor requires that all police traffic stops be cataloged and analyzed to see if the police are discriminating against minorities.

Below is the official definition of Hispanic from the study. Can you imagine a police officer trying to figure out this crud?

From the document: “Northwestern University Center for Public Safety
Illinois Traffic Stop Statistics Act 2004
” pg. 45-54

One of the major decisions that we had to make was with respect to the
categorization of Hispanics. The problem that we encountered here was that the Illinois law designated a category that was not immediately reconcilable with the Census. The solution to the Pacific Islander quandary was fairly simple: combine two Census categories. With regards to Hispanics, however, the issues were more complex and had the potential for much larger ramifications. Because our decision in this area is undoubtedly one of the most important ones in this analysis, some attention should be devoted to elaboration.

The U.S. Census considers Hispanic origin to be an ethnicity, not a race. There
are, therefore, two separate questions on the Census form. The first asks whether the
respondent is of Hispanic origin and requires specification of how such origin is claimed.
In answering this question, the respondent can check “No, not Spanish/Hispanic/Latino”
or one of four “yes” boxes. Three of the “yes” boxes are associated with a particular
Hispanic or Latino group(s): “Mexican, Mexican Am., Chicano,” “Puerto Rican,” and
“Cuban.” The fourth “yes” box is for all “other” Hispanic, Latino, or Spanish affiliations
and requires the respondent to print the name of the ethnic group.
The second question asks the respondent to identify a race. Here the person can claim “White,” “Black, African Am., or Negro,” “American Indian or Alaska Native,” Asian Indian,” “Chinese,” “Filipino,” Japanese,” “Korean,” “Vietnamese,” “Other Asian,” “Native Hawaiian,” “Guamanian or Chamorro,” “Samoan,” “Other Pacific Islander,” and “Some other race.”

Anyone designating himself as “Other Asian,” “Other Pacific Islander” or “Some other
race” must print the name of that other race. Thus a person can claim Hispanic origin
and be of any race. In order to conform to the Illinois law, which separates Hispanic
from the other races, we have to determine a method of extracting the Hispanic
population estimates in the Census from the other races. Not doing so will result in
double counting the Hispanic population (that is, counting them once as their race and
once as Hispanic) and thus erroneously inflating the number of Hispanic drivers “at risk”
of being stopped.

In considering this issue, we looked at other studies for guidance. There are, in
essence, three ways to approach the problem. First, anyone of Hispanic origin can be
considered “Hispanic,” meaning that a “Hispanic” individual may be of any race. This
was the approach taken by the team working on the Missouri study. Second, any
person claiming “Hispanic” origin can be classified according to the race that he
indicated on the Census questionnaire. This means that an “African American Hispanic”
person would be, for the purposes of the study, considered “African American” and not
Hispanic. This method, however, will only work where the Legislature has tailored the
law such that Hispanic is not an option for the officer making the stop. If the officer is
not able to designate the driver’s race as Hispanic, then he will necessarily have to
categorize that driver as one of the other racial groups represented in the Census. Put
differently, if the officer has the ability to select “Hispanic” as the driver’s race, then the
benchmark must have a “Hispanic” category against which to compare that stop. The
third possible solution is a “split the difference” or “hybrid” approach used by the
Institute on Race and Poverty at the University of Minnesota in the Saint Paul data
collection study. There, any Hispanic person claiming Black, Asian, or Native American
identification was counted according to that racial designation (that is, as Black, Asian or
Native American). Any white or “other” Hispanic, however, was counted as “Hispanic.”
In other words, the only “Hispanics” in the Saint Paul study were “white” or “other.”

As noted in earlier sections, the Illinois Legislature included “Hispanic” as an
option for the officer designating race. Therefore, approach two above will not work in
Illinois. As between the first and third approaches, we chose the former for a number of
reasons. Initially, it seems to be the one favored by most researchers, including Lorie
Fridell. Moreover, however, counting as “Hispanic” all who claimed such ethnicity has
an additional benefit of including many individuals in the benchmark who would be
otherwise lost. Consider, for example, that approximately 5.2% of all driving age
respondents in Illinois (or 503,021 people) described themselves as “some other race.”
However, 98.4% of those individuals also identified themselves as “Hispanic.” The
Illinois law does not allow an officer to select “some other race.” Therefore, by using
the first approach, we managed to capture 98.4% of Illinois residents (age fifteen and up)
of “some other race” that would have been lost had we classified Hispanics by their

There is, however, another, more fundamental reason for using the first approach
and designating all Hispanics, no matter their race, as “Hispanic.” The goal in data
collection studies is to determine whether an officer’s decision to stop a motorist is
influenced by the physical appearance of the driver. In addressing this issue, we (both as
a society and as researchers) classify “Hispanics” as a minority group. But 46.1% (or
487,779) of all driving age Hispanics in Illinois (1,058,323) designated themselves as
“white.” If we were to divide Hispanics by their race, rather than designate them all as
“Hispanic,” 46.1% of all Hispanics would be subsumed within the “white” category,
erroneously minimizing the number of minorities of Hispanic descent at risk of being
stopped by the police.

Finally, there is the issue of individuals claiming to be of two or more races. The
Census Bureau allows a respondent to choose any combination of races (up to six, the
total number of races on the Census). As part of our study, we chose to focus on those
individuals claiming only one race. As previously noted, 1.52% of all Illinois residents
age 15 and up (or 148,034 people) claimed to be of two or more races. However, 33.9%
of these individuals (or 50,179 people) also claimed to be Hispanic. Therefore, by
choosing to designate every person claiming Hispanic origin as Hispanic, we were able to
capture an additional 50,179 Illinois drivers by including multi-racial Hispanics in the
“Hispanic” category.

For all of these reasons, we concurred with the Missouri approach and counted as
“Hispanic” any person who had identified himself as “Hispanic.” Consequently, in our
analysis, “Hispanic” can be an individual of any race.

- - - - - - - - - -

See, I told you it would give you a headache! There was more but I spared you the charts and graphs!

No comments:

Post a Comment