Skip directly to search Skip directly to A to Z list Skip directly to navigation Skip directly to page options Skip directly to site content

Historical Context


The National Health Interview Survey (NHIS) is one of the major data collection systems of the National Center for Health Statistics (NCHS). In order to provide the broadest possible perspective on the health data obtained in the survey, the NHIS collects information on the sociodemographic characteristics of respondents, such as age, race, Hispanic origin, sex, education, income, employment, family size and relationships, and geographic region of residence. Demographic characteristics are collected primarily because they provide a context for the health data collected in the survey, and because they help to explain interrelated trends in the survey data. Inequalities in health status and access to care - as well as the unequal burden of morbidity and mortality - for some racial and ethnic groups in the United States have made race and Hispanic origin among the most important demographic characteristics of interest to users of the NHIS.

For almost 20 years, the NHIS collected information on the race and Hispanic origin of its respondents, following guidelines set forth by the Office of Management and Budget (OMB) in a policy known as OMB Directive 15 (1). In response to the changing demographics of the U.S. population and the growing need of persons with multiple racial heritages to indicate this on the Decennial Census and federal surveys, the OMB began a review of these guidelines following the 1990 Census. After an extensive period of research and public commentary, the OMB issued new standards for race and Hispanic origin data collection in the federal statistical system in October 1997 (2). The new standards included provisions for the creation of two new race categories – Asian, and Native Hawaiian and other Pacific Islander – and for the placement of Hispanic origin questions before race questions when the questions are asked separately. It also included a provision permitting respondents to the Census and federal surveys to select more than one group when answering questions on race. The changes were fully implemented in the 1999 NHIS.


Historical Context

From its inception in 1957 through 1975, the NHIS did not explicitly ask race of its survey respondents. Race was recorded solely on the basis of the interviewer's observation, and no information was recorded about Hispanic origin. Interviewers were instructed to record the race of the respondent and family members as "white", "black, or "other", probing verbally only when the race could not be determined. Although it was recognized that the racial and ethnic makeup of the U.S. population was considerably more complex than these categories would suggest, and that the race of many persons would be difficult to identify on the basis of observation alone, it was believed that these broad racial categories would provide an adequate and socially acceptable basis for tabulating statistical data for demographic and social research purposes (see the observed race question [PDF - 155 KB]).

Both the need for better data on minority groups in the U.S., and an acknowledgment of the inadequacy of race data classified solely by observation, led to changes in the way the NHIS collected race and Hispanic origin data. In 1976, a single "national origin" question was added to the Demographic Background section of the survey. This format was used to obtain race and Hispanic origin data for the 1976 and 1977 surveys (see the 1976-77 question [PDF - 153 KB] and 1976-1977 flashcards [PDF - 270 KB]).

In 1978, race and Hispanic origin were asked for the first time in a two-question format. The Hispanic origin items were separated from the race items, and race was asked first. These questionnaire items also allowed respondents to identify themselves as having more than one race or Hispanic origin. Those persons who indicated that they had more than one racial or Hispanic origin group were given a follow-up question that asked them to select from the responses the single group they felt best described them. In 1979, the race category "American Indian or Alaskan Native" was change to "Aleut, Eskimo, or American Indian" as defined by OMB Directive 15 (which was established in 1977). However, from 1976-1981, these questions were asked only of persons in the household aged 17 or older. In 1980-1981, the race of children in the household was imputed based on the race reported for their parent(s). The NHIS race and ethnicity questions used in the survey from 1978-1981 [PDF - 300 KB] can be found here.

As part of a major survey redesign in 1982 [PDF - 210 KB], the NHIS began asking race and Hispanic origin questions about all household members. Under OMB Directive 15, data were collected for four race groups (white, black, Aleut/Eskimo/American Indian, and Asian/Pacific Islander). Persons reporting more than one race were asked the same follow-up question first added to the survey in 1978. Information on Hispanic origin was also obtained for eight Hispanic origin subgroups (Puerto Rican, Cuban, Mexican, Mexicano, Mexican-American, Chicano, Other Latin American, and Other Spanish). The race and Hispanic origin questions and response categories, as well as their placement in the questionnaire, remained consistent from 1982-1991. In 1985, the NHIS began oversampling black households in order to increase the precision of estimates for this group.

In 1992, the NHIS started collecting information on the following Asian/Pacific Islander population subgroups: Chinese, Japanese, Vietnamese, Filipino, Asian Indian, and Korean. An "Other Asian" category was also used to capture responses that did not fall into any of the categories listed. That year also marked the first time the NHIS fielded a Cancer Epidemiology supplement for Hispanic respondents, so the Hispanic origin question [PDF - 520 KB] was moved from the Demographic Background section of the questionnaire to the Household Composition section to facilitate screening for Hispanic respondents. The race question [PDF - 362 KB] race question remained in the Demographic Background section.

In 1995, for the first time, the NHIS sample redesign included oversampling of both black and Hispanic households in order to increase the precision of estimates for both groups. To facilitate screening of households for black and Hispanic members, the race and Hispanic origin questions [PDF - 533 KB] (see also the 1992 and 1995 flashcard [PDF - 272 KB]) were moved to the Household Composition section of the survey, where they have remained.

In 1997, the NHIS underwent another major survey redesign. The placement and order of the race and Hispanic origin questions were unaffected by this new design, but other changes were implemented. Respondents were allowed to give up to five Hispanic origin groups (identified as HISPAN_1 – HISPAN_5 in the questionnaire) and up to five race groups (identified as RACE_1 – RACE_5 in the questionnaire). All questions were also assigned a new numbering system, with the prefix indicating the questionnaire section. Since 1997, the Hispanic origin and race questions [PDF - 618 KB] have a prefix of HHC, indicating their placement in the household composition section. The 1997 flashcards [PDF - 356 KB] are also included for reference.

The Office of Management and Budget revised its standards for the collection of race and ethnicity data in federal data systems in 1997. Two major changes resulted from this: the collection and reporting of data for multiple race persons and the establishment of two new categories – Asian, and Native Hawaiian or Other Pacific Islander – from the old Asian and Pacific Islander race category. The NHIS fully implemented these changes in the 1999 race and Hispanic origin questions [PDF - 1.7 MB] (see also the 1999 race and Hispanic origin flashcards [PDF - 357 KB]). The race and Hispanic origin question wording, placement, and order have remained consistent since 1999.

In the 2006 survey year, the NHIS implemented oversampling of Asian households for the first time. As has been done with black and Hispanic households in the NHIS, oversampling Asian households will help to improve the precision of estimates for Asian respondents. The draft 2006 race and Hispanic origin questions [PDF - 1.6 MB] and the 2006 flashcards [PDF - 335 KB] are provided here to give users easy access to the most up-to-date information.

The following tables summarize the history of race and Hispanic origin data collection in the NHIS:

  • Table 1 [PDF - 17 KB] - provides an overview of the major changes that took place in NHIS race and Hispanic origin data collection from 1975-2006.
  • Table 2 [PDF - 20 KB] - provides detailed information on the period from 1982 (when then NHIS began collecting race for all household members) through 1996, the last year before the survey redesign.
  • Table 3 [PDF - 20 KB] - provides detailed information on the period from 1997-2006.