A second shortcoming in the literature is the converse of the first, namely, that studies of adverse health effects fail to differentiate the population by race, ethnicity, or socioeconomic class and examine differences in outcomes among them (Frumkin and Walker, 1997). As noted above, occupational health studies are a primary source of knowledge about the health effects of various hazards. However, only a small fraction of the existing occupational health research has provided meaningful information on either the prevalence of occupational diseases in specific racial, ethnic, or socioeconomic population groups or the relation of race, ethnicity, and socioeconomic factors to occupational health. For example, in one systematic review of 116 studies of the epidemiology of occupational cancer published in four journals, only 14 studies (12 percent) provided data on a "non-white" group. The investigators concluded that the published literature contributes little to understanding the complex relationships among occupation, cancer, and race (Kipen et al., 1991).
One rationale given for the exclusion of minority subgroups is that the number of subjects would be too small to provide an acceptable statistical power to test the primary study hypotheses with these subgroups. (Appendix A discusses these and other related issues in greater depth.) Other reasons include difficulties in long-term follow-up; inadequate measurement, classification, and reporting of data on race, ethnicity, and relevant socioeconomic variables (Feinleib, 1993; Montgomery and Carter-Pokras, 1993); and difficulties in researcher access to high-risk workplaces that employ low-wage workers.
Many of the challenges posed by an analysis of environmental justice issues are illustrated by the example of urban asthma. A doubling in the rate of asthma in the United States since 1980 and its apparent association with industrialized, urbanized areas has led some investigators to suspect increased levels of exposure to environmental factors as one possible cause (Vogel, 1997). As noted above, Hispanics and African Americans are more likely than whites to live in areas where the levels of particulates, sulfur dioxide, and ozone exceed National Ambient Air Quality Standards—circumstances that may contribute to the prevalence and severity of asthma.
The prevalence of asthma appears to be more strongly correlated with lower socioeconomic status than with race and ethnicity (Institute of Medicine, 1993). However, the use of data on numbers of hospitalizations, or emergency room visits or even the rate of morbidity due to asthma as a measure of the relative impact of environmental hazards on those with low socioeconomic status is problematic, because these indicators may be strongly influenced by other factors such as lower rates of health insurance or lower levels of access to high-quality primary health care.