|
|
|
Joseph J. Salvo [1]
NYC Department of City Planning
May 2001
While multi-racial persons existed prior to the 2000 Census, only in the 1990s did the federal government acknowledge this small, but fast-growing racial category. The number of multi-racial individuals has increased substantially and is likely to continue rising with each passing decade, given the increasing racial and ethnic diversity of the nation and the concomitant rise in interracial marriage.
For the first time in the 2000 Decennial Census, the question on race permitted persons to "check one or more" from the following six categories: white; black, African American or Negro; American Indian or Alaska native; Asian; Native Hawaiian and Other Pacific Islander; and Some other race.[2] The various combinations yield a total of 63 race categories. Respondents were also asked if they were of Hispanic origin. Thus in 2000, there were 64 mutually exclusive race/Hispanic categories available for tabulation: 63 race categories tabulated only for non-Hispanics, and a 64th category for all Hispanics. This is a big change from the 1990 Census, where just five categories existed for race, with no option for recording multi-race responses.
In the 2000 Census, 6,826,228 persons checked more than one race, representing a large number of combinations. This poses a significant challenge to keeping 2000 Census tabulations manageable. It also complicates attempts to evaluate change in the racial composition of areas in the 1990-2000 period - particularly for evaluating changes among children, who are more likely than adults to be multi-racial. Four percent of persons under 18 were categorized as multi-racial in the 2000 Census, compared to 2.4 percent of those 18 and over.
This article offers a few basic guidelines regarding the use of these data and some resources to help users work through the conceptual and methodological issues affecting the tabulation of race data.
First and foremost, the data user must understand how the data are being applied. What question is the researcher trying to answer? There are at least three generic types of questions or application areas: civil rights monitoring and enforcement; documenting the emerging multi-racial population; and program planning, targeting and implementation. Depending on the application, the tabulation requirements will differ. It is important to acknowledge that there is no single way to present these data, but multiple paths to understanding the data that depend on the questions addressed and the methodologies applied.
Use of the race data for civil rights monitoring and enforcement concerns compliance with laws on voting rights, equal employment opportunity, affirmative action and other laws prohibiting discrimination by race. In these cases, it is important to track changes in the numbers of persons in groups that are the focus of anti-discrimination efforts. Yet a precise analysis of these changes is hampered by the lack of strict comparability, given the absence of the multi-race response option in 1990.
In order to deal with these tabulation challenges, the federal Office of Management and Budget has worked closely with agencies on guidelines for reporting race and ethnicity. One result is the creation of specific program requirements for the tabulation of race and Hispanic data; rules have been developed, for example, to combine single response and multiple response categories for the purpose of assessing shifts in groups considered to be subject to discrimination. Moreover, an overall effort is underway to better understand how comparisons can be made over time. It is important to pay attention to the "bridging" and related methods recommended by agencies charged with evaluating compliance. More information is available at www.whitehouse.gov/omb/bulletins/b00-02.html. Detailed reports on the implementation of standards and on trend analysis can be found under Race and Ethnicity at: www.whitehouse.gov/omb/inforeg/index.html.
Documenting the emerging multi-racial population requires tabulations with details on the race combinations reported by multi-race respondents in the Census. Given that there are up to 63 race combinations available on the PL 94-171 file, presenting data on multi-race groups in a comprehensible fashion can be a real challenge. One option is to present data for only two race combinations, which accounts for more than 93 percent of all multi-race responses nationally. For examples of this approach and others see www.nyc.gov/html/dcp/html/poptable.html or www.usc.edu/sppd/census2000.
In areas with large Hispanic populations, the nature of these tabulations can change radically, depending on whether Hispanics are included or excluded from the multi-race category. Of the more than 6.8 million multi-race respondents in the 2000 Census, some 2.2 million were Hispanics, about one-third of the total. More than three-quarters of the Hispanics who checked more than one race category checked one of the five major race groups and "other," with the write-in response likely indicating their Hispanic ethnicity (e.g., Dominican, Puerto Rican, etc.). In some parts of the nation, a majority of persons checking more than one race were Hispanic. Thus, removing the Hispanic population from the "multi-race" category can have a significant impact on the size of the multi-race group. New York State's multi-race total is 590,000 (3.1% of the total population) if its large Hispanic population is included; if Hispanics are excluded from the multi-race total, the number is 366,000 (1.9% of the total population). Similarly, California's multi-race population totals change from 1.607 million if Hispanics are included to 903,000 if they are excluded - a change from 4.7 percent to 2.7 percent of the total population.
Uses of race categories for program planning, targeting, and implementation focus on addressing the needs of groups for specific programs. Data on race and Hispanic origin provide valuable context for program planners who are attempting to deliver services; data on the characteristics of these groups, which will be available in the Census Summary Files, are also important. Demographic and socioeconomic data help identify populations that are targets for programs such as day care for working mothers, job training, and English language proficiency. Decisions on categorizing the multi-race population need to be study-specific for maximum flexibility in identifying groups. For example, a language program for Asian neighborhoods may best be served by a maximum definition of the group; all respondents that listed an Asian group, either alone or in combination with another race, need to be tabulated. More information on the approaches used in Census Bureau products can be obtained at www.census.gov/Press-Release/www/2001/raceqandas.html.
A good source of practical information on the use of the new race data, especially as it pertains to children, can be found in "Using the New Racial Categories in the 2000 Census" by Sharon M. Lee (KIDS COUNT/PRB Report, March 2001). This paper is available in pdf format at: www.kidscount.org or www.ameristat.org. Finally, visit www.usc.edu/sppd/census2000 for a comprehensive look at the tabulation options for 2000 Census race data.
Educating users about the conceptual basis for data tabulations is a major task for data disseminators and nowhere is this more important than with the tabulations involving race. We should provide users with data that are flexible enough to permit the creation of alternate tabulations to fit the application at hand. In addition, those who serve as intermediaries have a responsibility to provide guidance on the correct application of these data, a role that has never been more crucial. Most people now enjoy unprecedented access to these data, and as such, are in urgent need of guidance on how to use it.
APDU welcomes feedback on this article via e-mail. Please include your name, organization, and e-mail address. The editors reserve the right to publish feedback.
2 In this presentation, the discussion refers to "Asian" and "Native Hawaiian and Other Pacific Islander" as two major race groups. The actual questionnaire contains seven subgroups that can be checked for the former and four subgroups for the latter.
Date last modified: 28 April 2001
Editorial contact: Joseph J. Salvo
and Julie Linden
Disclaimer: This site does not necessarily reflect the views of the Yale
University Library.