agencies, such as the National Institutes of Health, which are required to comply with the U.S. Office of Management and Budget's guidelines for data collection with regard to U.S. population groups. The U.S. Office of Management and Budget's Directive 15 requires federal agencies to report population data on the basis of five "racial and ethnic" groups (U.S. Office of Management and Budget, 1978). The U.S. Office of Management and Budget notes that such classifications do not carry scientific or anthropological validity; rather, they are based on social and historical considerations. The American Anthropological Association and the IOM Committee on Cancer Research Among Minorities and the Medically Underserved note that this reporting requirement may handicap health researchers, who are often unable to draw meaningful inferences regarding the source of group differences because "racial" groups do not vary systematically with regard to biological or genetic makeup, socioeconomic status, culture, or other relevant variables.
The use of "race" in health research will be further complicated in the future because the U.S. census will allow respondents to list more than one "racial" category to describe themselves. Although it is expected that only a small fraction of the U.S. public will seek to describe themselves as belonging to more than one "racial'' group (U.S. Bureau of the Census, 1997), health researchers will have to develop means of accounting for such populations.
Notwithstanding these concerns, the committee believes that the collection of race- and class-specific data is crucial if environmental justice is to be achieved. Data that lack specificity or that gloss over demographic and material realities will not support adequate analysis, regulatory intervention, or remediation of environmental health risks. In many instances, race is a critical variable with respect to environmental justice because (1) the racial segregation of neighborhoods remains a common feature of the United States, and (2) lower-income, predominantly minority neighborhoods may be especially vulnerable to environmental degradation and abuse primarily because of relatively lower levels of political and economic clout of the populations in those neighborhoods to gain redress or restitution from environmental polluters. "Race" therefore remains an important term because of its social and political implications, but it should not be assumed to have scientific validity in the absence of evidence.
To better understand the consequences of exposure to environmental hazards, public health officials and researchers should pay close attention to the experiences of individuals in local communities and should systematically collect and validate data on those experiences. Adverse health effects from environmental hazards are often suspected first by the people who experience them rather than by the health care or scientific community. In other cases, a toxicant's effects may be known to the health or science professionals but particular routes of exposure may have yet to be discovered for a particular community. In