[ENVIRONMENTAL RESEARCH 69, 15-30 (1995) Race, Class, and Environmental Health: A Review and Systematization of the Literature Puit Brown Brown University, Box 1916, Providence, Rhode Island 02912 Roceived May 24, 1994 ‘This paper analyzes and system differentials in exposure to toxic hazards and actual -alth outcomes. Research is categorized into the follow- ing: Proximity to known hazards includes (1) presence of hazardous waste sites and facilities (landfills, incinera- tors, Superfund sites), (2) exposure to air pollution, (3) exposure to various environmental hazards, e.g., toxic releases and hazards in pesticides and foods; Regulation, ‘amelioration and cleanup includes (4) record of decisions (RODs) and cleanups at NPL sites, (5) regulatory action, fas measured by assessed fines for environmental pollu- tion; Health effects includes (6) specific health outcomes which are related to environmental burden (e.g., blood lead levels). Proximity to prospective hazards includes (7) siting decisions for incinerators, hazardous waste sites, and nuclear storage sites. The overwhelming bulk of ev- idence supports the “environmental justice” belief that environmental hazards are inequitably distributed by class, and especially race. ss the race and class In recent years there has been a steadily expand- ing interest in questions of unequal exposure to toxic hazards on the part of poor and minority peo- ple. While some of this evidence (concerning air pol- lution) was known as early as the 1970s, public and scientific awareness was slow to develop. The cur- rent upsurge of interest was propelled by residents, of minority communities who felt themselves at ex- cess risk. Subsequently, this issue has received ad- ditional interest from government agencies, envi- ronmental organizations, social scientists, and the media. ‘This paper analyzes the race and class differen- tials in exposure to toxic hazards and actual health outcomes to determine if existing research confirms the presence of unequal exposure. This exposure in- cludes lead exposure, air pollution, hazardous incin- erator licensing, farmworker pesticide exposure, and Superfund cleanups. Despite the growing atten- tion to this area, there is no adequate systematiza- 18 tion of the empirical data that could confirm the ex- istence of unequal exposure. Such systematization is, necessary in order to make sense of the existing re- search, to shape future research, and to take part in, public policy discussions. INTRODUCTION—RACE AND CLASS VIEWED THROUGH THE ENVIRONMENTAL, JUSTICE FRAMEWORK Out of the activism of people and organizations in affected minority communities, the movement now widely termed “environmental justice” has become the leading approach within the toxic waste move- ment, itself the fastest growing social movement in, recent years (Szasz, 1994), The toxic waste move- ment is composed of working class and lower middle class activists, with a predominance of women. These activists typically have no prior political or environmental involvement, although minority ac-~ tivists are likely to have had some civil rights polit- ical background, Health concerns predominate among these toxic activists (Brown and Ferguson, 1995; Brown and Masterson-Allen, 1994). Environ- mental justice developed ftom a civil rights tradition and a grassroots democratic urge, rather than out of mainstream environmentalism; yet it has also af- fected the larger environmental movement (Gott- lieb, 1993). ‘The environmental justice framework probably can be best dated to the 1978 protests by residents of Warren County, North Carolina over a polychlori- nated biphenyl (PCB) dump. Through the 1980s ‘more minority antitoxics groups organized. Their ef- forts were aided by a body of research that showed excess minority environmental burden. A 1990 Uni- versity of Michigan conference on “Race and the In- cidence of Environmental Hazards” pressured the EPA to establish an Office of Environmental Equity that same year. The 1991 People of Color Summit, 0013-9951195 $6.00 Copyright © 1996 by Academic Press, Ine All rights ol reproduction in any form reserved16 PHIL BROWN held in Washington expanded the environmental justice movement further. On February 9, 1994 President Clinton an- nounced that he would sign within days an Execu- tive Order requiring all federal agencies to work to- ward ending the disproportionate exposure of minor- ity and poor people to many environmental hazards (Cushman, 1994). This was timed to coincide with the February 10-12, 1994 Environmental Justice Conference sponsored by EPA, the Centers for Dis- ease Control, the National Institute for Environ- mental Health Sciences, the National Institute for Occupational Safety and Health, and the Agency for Toxic Substances and Disease Registry. That confer- ence, attended by large numbers of minority activ- ists, was quite contentious. Following a walkout by some community activists, EPA Administrator Carol Browner turned the microphone over to lay- people, who denounced the EPA’s role (MeFarling, 1994) Tam concerned here with human-produced toxic substances because they represent a major threat which has been inadequately understood, because they stem largely from the post-World War II eco- nomic system, and because they represent a social perception of dreaded phenomena. These substances, include chemicals, industrial by-products, pes cides, insecticides, fungicides, and radioactive waste. While certain other human-made products, such as tobacco products, are intensely dangerous, they are not included here, nor are naturally occur- rring substances such as radon or solar radiation. RESEARCH ON ENVIRONMENTAL JUSTICE Most research on environmental justice demon- strates that race and class are important determi- nants of environmental exposure and environmen- tal health effects. In a majority of cases, race is more significant than class. Different areas show different minority groups (blacks, Hispanics) to be affected, though blacks are the most typically overexposed to hazards. Studies have examined a variety of geographic units (zip codes, census tracts, cities, SMSAs) and a variety of hazards (Superfund sites, toxic emissions, existing waste facilities) In the first published overview of research in this, area, Mohai and Bryant (1992) reviewed 15 studies. ‘They found that over a wide range of geographical areas (local, regional, national), race and class, es- pecially race, were associated with increased expo- sure to environmental hazards. Eight of the studies they examined looked at both race and class, and in five of those race was more powerful. More studies have been published since Mohai and Bryant's re- view, some of which deal with areas not previously studied. In addition to incorporating a larger body of research, I want to categorize the literature so that we can see how class and race are associated with specific areas of environmental quality. Another ad- vantage of this categorization is that we can focus on the measurement and design issues that are often, specific to certain categories. Research can be categorized into the following areas: Proximity to known hazards (1) Race andior class differences in presence of hazardous waste sites and facilities (land- fills, incinerators, EPA sites—CERCLIS and NPL). (2) Race andor class differences in exposure to air pollution. (3) Race and/or class differences in exposure to various environmental hazards, e.g., toxic releases measured by the EPA’s Toxic Re- lease Inventory (TRI), hazards in pesticides and foods. Regulation, amelioration, and cleanup (4) Race and/or class differences in Record of Decisions (RODs) and cleanups at NPL sites (5) Race and/or class differences in regulatory action, as measured by assessed fines for en- vironmental pollution. Health effects (6) Race and/or class differences in specific health outcomes which are related to envi- ronmental burden (e.g., blood lead levels). Proximity to prospective hazards (D Race and/or class differences in siting deci- sions for incinerators, hazardous waste sites, nuclear storage sites. Proximity to Known Hazards (1) Hazardous Waste Sites In 1983 the General Accounting Office (GAO) is- sued a report which found that three of the four haz- ardous waste landfills in the Southeast were located in predominantly poor or black areas. Despite the small number of sites, this study was notable in that it was the first report of such disparities in facility location. It was also notable because it resulted from the request of District of Columbia Rep. Walter Fauntroy, who had been arrested in 1982 with 500 others in a protest against a polychlorinated biphe-RACE, CLASS, AND nyl (PCB) landfill in Warren County, North Caro- lina. With a civil rights background, but no environ- mental movement experience, these poor black citi- zens had launched a powerful new movement. Rep. Fauntroy wanted further evidence of racial injustice in hazardous waste facility siting; hence, he re- quested the GAO to investigate (Lee, 1992). ‘Another early study of Houston found that six of the city’s eight municipal incinerators (including all five of the large ones) and all five of its landfills were in predominantly black areas. One of the other two incinerators was located in a predominantly His- panic area, Income did not protect blacks—even middle-class black neighborhoods contained facili- ties (Bullard, 1983). Both the GAO report and Bull- ard study dealt with small numbers of sites, al- though they certainly pointed to the need to address, the environmental equity question. Further study was needed to demonstrate bias across large num- bers of sites. In perhaps the most widely cited research in this area, the Commission on Racial Justice of the United Church of Christ (1987) conducted two na- tional studies of the racial and socioeconomic char- acteristics of people residing near commercial haz- ardous waste facilities and uncontrolled toxic waste sites, ‘The first study examined the relationship between race and location of all commercial hazardous waste facilities then in existence in 1986 (N = 415), in- cluding incinerators, transfer stations, and disposal sites other than landfills. The study sought to iden- tify predictors of proximity to sites from the follow- ing: minority percentage of population, mean house- hold income, mean value of owner-occupied homes, number of uncontrolled toxic waste sites per 1000 persons, and pounds of hazardous waste generated per persons (p. 10), The first hypothesis predicted that the “mean minority percentage of the popula- tion was a more significant discriminator than the other variables for differentiating communities with a larger number of commercial hazardous waste fa- cilities and the largest landfills” and the second pos- ited that the “mean percentage of minority popula- tion would be significantly greater in communities with facilities than those without” (p. 11) Location was determined by dividing zip codes into four groups. Group 1 had no hazardous facili- ties; Group 2 had one operating commercial hazard- ous facility that was not a landfill; Group 3 had one hazardous waste landfill that was not one of the five largest; Group 4 had one of the five largest landfills or more than one hazardous facility. As hypothesized, results indicated that those com- NVIRONMENTAL HEALTH 7 munities with the greatest percentages of minority residents had the most toxic waste facilities. In fact, as shown in Table 1, in the communities housing @ commercial toxic waste site, the percentage of the minority population was double the minority per- centage in communities without a waste facility (24% vs 12%) (p. 13). Similarly, in areas with two or more hazardous waste sites or one of the five largest, landfills, the mean minority population percentage was three times that of communities without waste facilities (p. 13). Percentage of minority population, proved to be the strongest predictor of communities with the greatest number of waste facilities and the largest landfills. Mean income (a disparity of $2,745) and mean house value (a disparity of $17,391) were also significant, but less so. ‘The second study determined the race of individ- uals who reside near all 18,164 uncontrolled waste sites (Comprehensive Environmental Response, Compensation and Liability Act Information Sys- tem, CERCLIS; these are the sites from which Su- perfund sites are drawn) existing in 1985 which are situated in residential zip code areas. This demon- strated that three of five black and Hispanic indi- viduals resided in a community with a CERCLIS site (p. 13). The UCC study also noted that three of the five largest commercial hazardous waste land- fills in the United States, making up 40% of the nation’s total capacity for hazardous waste landfills, were located in predominant black or Hispanic com- munities. One alone, in Emelle, Alabama, accounted for 23% of national capacity. Some scholars and officials have objected to some methodological issues of the UCC report. The UCC. study used a high P value of 0.10. The four groups of ip codes are not the most exact measure, and may TABLE 1 ‘Minority Status of Residential Zip Codes with Hazardous ‘Waste Facilities, 1986 ‘Type of ip code jority Group 1: residential zip codes with no treatment, storage, oF disposal facility 123 Group 2: residential 2ip codes with one troatment, storage, oF disposal facility that is not a Tand&ll 237 Group 3: residential zip codes with one hhazardous waste landfill that is not one of the nation’s five largest 220 Group 4: residential zip codes with more than fone treatment, storage, or disposal facility, or with one of nation’s five largest hazardous waste landfills 318 ‘Source. Commission on Racial Justice (1987)