Incinerators, landfills, and other "locally undesirable land uses" (LULUs) are not popular neighbors, however essential they may be to the community at large. The fact that many of them are located in poor and minority communities may look at first glance like a clear case of discrimination in siting. A closer look reveals, however, that the problem may not be solely with the siting of LULUs, but also with the housing-market dynamics that come into play after a LULU has been established. If so, attempts to achieve a more equitable distribution of LULUs would have to extend beyond changes in the siting process.
Studies show that communities hosting waste management facilities and other locally undesirable land uses (LULUs) have, on average, higher percentages of racial minorities and the poor than other communities. Advocates of environmental justice contend that this is unfair, arguing that environmental risks should be distributed more equitably among races and socioeconomic classes. They assert that the disproportionate burden LULUs impose on poor and minority communities is the result of racism and classism in the siting process. As evidence, they point to studies that reveal a correlation between the racial and class characteristics of communities and the presence of LULUs in those communities.
If the siting process does discriminate against the poor and racial minorities, it should be reformed. However, there may be other reasons why poor and minority communities host a disproportionate number of LULUs; one may be that housing-market dynamics lead the areas surrounding LULUs to become disproportionately poor or minority after LULUs have been sited.
If this is the case, we must look beyond the siting process if we are to remedy inequities in the distribution of LULUs. Indeed, if the free market is in part the cause of these inequities, even a siting system that ensured a perfectly fair initial distribution of LULUs would not result in any long-term benefit to the poor or people of color.
The GAO and Bullard studies
More than a dozen studies document the fact that poor and minority communities now host a disproportionate number of the nation's LULUs. Two of the most notable studies are "Siting of Hazardous Waste Landfills and Their Correlation with Racial and Economic Status of Surrounding Communities," which was conducted by the General Accounting Office (GAO) in 1983, and "Solid Waste Sites and the Black Houston Community," which was published by Dr. Robert Bullard in the same year.
The GAO study examined the racial and class characteristics of communities surrounding four hazardous waste landfills located in three southeastern states. GAO found that, in 1980, African Americans made up between 52 and 90 percent of the population of three of the four communities where the landfills were sited, but only between 22 and 30 percent of the host states' populations. The study also found that between 26 and 42 percent of the population of the host communities was living below the poverty level, but that the host states' poverty levels ranged only from 14 to 19 percent.
Bullard's study sought to determine whether the siting of waste facilities in Houston, Texas, discriminated against African Americans. It found that, in 1980, six of Houston's eight incinerators and mini-incinerators, as well as fifteen of its seventeen landfills, were located in predominantly African American neighborhoods. At that time, African Americans made up only 28 percent of the Houston population.
Both the Bullard and GAO studies are cited as proof that the current distribution of LULUs is the result of discrimination in the siting process. However, neither establishes that the siting process caused the disproportionate distribution. Each study considered only the current demographics of host and non-host communities, ignoring the demographics of communities at the time siting decisions were made. This failure begs an obvious question—namely, whether host communities were poor and minority communities at the time they were selected as LULU sites or only became so in subsequent years.
If neighborhoods were minority neighborhoods at the time they were selected to host LULUs, the choice of sites may have been racially discriminatory. If so, then the siting decisions would have been unfair. But if the neighborhoods were not minority neighborhoods when they became LULU hosts, some factor other than discrimination must account for the fact that they now are disproportionately populated by minorities. That factor may be the dynamics of the housing market.
The role of housing-market dynamics
Each year, between 17 and 20 percent of the U.S. population moves to a new home—often to a different neighborhood in the same city or to a different city. The decision to move is based in part upon individuals' dissatisfaction with the quality of their current neighborhoods. A new neighborhood is selected in part because of its characteristics and cost of housing. These two factors are interrelated because the quality of the neighborhood affects the price of housing.
Accordingly, the siting of a LULU can influence the characteristics of a neighborhood in two ways. First, an undesirable land use may cause those who can afford to move from the neighborhood to do so. Second, it may decrease property values in the neighborhood, making housing available to low-income households and unattractive to high-income households. As a result of both influences, the neighborhood is likely to become poorer than it was before it hosted the LULU.
The neighborhood also is likely to become home to an increasing number of people of color, whenever racial discrimination in the sale and rental of housing relegates them to less desirable neighborhoods than are available to whites. Once a neighborhood becomes a community of color, racial discrimination in the promulgation and enforcement of zoning and environmental protection laws, the provision of municipal services, and the lending practices of banks may cause neighborhood quality to decline further. That further decline will induce those who can leave the neighborhood—namely, the least poor and those least subject to discrimination—to do so.
The presence of a LULU in a neighborhood can lower the neighborhood's quality and thus its property values, making housing there more available to low-income families.
The dynamics of the housing market, therefore, are likely to force the poor and people of color to move to or remain in the neighborhoods in which LULUs are located, regardless of the demographics of the communities when the LULUs were first sited. Indeed, as long as the market depends upon existing wealth to allocate goods and services, it would be surprising if, over the long run, LULUs did not impose an undue burden upon the poor. And as long as the market discriminates on the basis of race, it also would be remarkable if LULUs did not impose an undue burden upon people of color.
Extending the studies
To determine whether the current distribution of LULUs is the result, at least in part, of market dynamics, I extended the GAO and Bullard studies. While those studies documented only the then-current demographics of the communities in question, I documented demographics roughly concurrent with the years in which siting decisions were made. I then traced subsequent demographic changes through 1990.
The GAO study examined the racial and class characteristics of communities surrounding four large hazardous waste landfills in the southeast. Sites for three of these landfills were probably chosen in the early or mid-1970s; the site for the fourth landfill was chosen in the late 1970s. Therefore I examined the 1970 demographic data for the first three sites and the 1980 demographic data for the remaining site.
My analysis of these data reveals that all four host communities studied by the GAO were predominantly African American at the time they were selected as LULU sites. The percentage of African Americans in the host communities' populations at the time the LULUs were sited ranged from 1.6 to 3.3 times that of the host states' populations. In the GAO's analysis, however, only three of the communities were predominantly African American in 1980.
Accordingly, demographic data from the time of the sitings, rather than from the 1980 census, strengthen the inference that siting choices had a disproportionate impact upon African Americans. This does not necessarily mean, however, that the process was discriminatory. Siting decisions may be based upon land prices, proximity to sources of waste, transportation networks, or other factors unrelated to race or poverty that nevertheless have an incidental, disproportionate effect upon people of color or the poor.
At the same time, the data provide no support for the theory that market dynamics cause host neighborhoods to become increasingly populated by African Americans. In all the communities the GAO studied, the landfill sitings were followed by decreases in the percentage of African Americans populating the communities. Between 1970 and 1990, the decreases in two host communities were 32.3 and 35.8 percent, even though the decrease in African Americans making up the total population of South Carolina, where the communities are located, was only 2.3 percent.
Demographic data for the time when the landfills were sited also provide no support for the theory that market dynamics cause host neighborhoods to become increasingly populated by the poor. If this theory were correct, the data should show decreases in relative median family income and relative median housing values, as well as increases in relative poverty subsequent to the sitings. According to my analysis, the relative poverty and relative median family income of the host counties changed only marginally between 1970 and 1990. During the same period, the relative median housing value also changed only slightly, and in two of the four host communities it actually increased.
My extension of the Bullard study offers somewhat different results. As noted above, Bullard examined demographic data for Houston communities hosting solid waste management facilities. In redoing his study, I eliminated the community demographics for communities where facilities had ceased to operate by the 1970s, since these sites were selected long ago and meaningful demographic data were not available. Consequently, my analysis was confined to the Houston communities that host the three mini-incinerators and seven landfills cited in the original study. While the original study used "neighborhoods" as its unit of analysis, I examined census tract data. I changed the unit of analysis because I had no information about how Bullard defined neighborhoods and therefore could not replicate his analysis.
Waste incinerators tend to be located in poor and minority communities. The inequitable distribution of these and other locally undesirable land uses has been blamed on discrimination in the siting process, but may result from housing-market dynamics
In my extension of Bullard's study, then, I examined the 1970 census data for seven communities, because all three of the mini-incinerators and four of the landfills were sited in the early 1970s. For the community hosting the two landfills sited in the early and mid-1950s, I examined both 1950 and 1960 census data. (Because the 1950 census tract containing the landfills was so large, the 1950 data are not particularly meaningful.) The remaining landfill was sited in 1978, so I examined the 1980 census data for its host community.
My analysis of the census data reveals that three of the seven landfills and two of the three mini-incinerators in question were located in areas where the percentage of African Americans was significantly greater than that of Houston as a whole at the time the facilities were sited. Even though about 25 percent of Houston's population was African American, five of the ten facilities were sited in areas where African Americans made up 60 percent of the population. This indicates that the siting process had a disproportionate effect upon African Americans.
Analysis of the Bullard sites shows that the siting process had a disproportionate effect upon African Americans, but it also lends force to the argument that LULUs change a community's demographics by driving down property values.
Yet analysis of the host communities' demographics in the decades after the LULUs were sited reveals that the siting process was not the sole cause of the undue burden Houston's African Americans now bear. Between 1970 and 1980, the percentage of African Americans in the neighborhoods surrounding the landfills increased by as much as 223 percent, while the percentage of African Americans citywide increased by only 7 percent. And while the number of African Americans as a percentage of the total Houston population changed little in the following decade, the number of African Americans as a percentage of host communities continued to increase in all but one of the communities. By the 1990 census, all of the communities hosting landfills had become home to a disproportionate percentage of African Americans.
Analysis of the host neighborhoods' economic characteristics reveals a similar pattern. Only three of the ten communities studied had poverty rates significantly higher than Harris County, where the communities are located, at the time the facilities were sited. Between 1970 and 1980, the poverty rate of all but two of the host communities (as measured by the percentage of the communities' population with incomes under the poverty level) increased, while that of Harris County dropped. Between 1980 and 1990, most of the communities hosting landfills experienced significantly higher increases in their poverty rates than did Harris County. By the 1990 census, five of these communities and two of the three communities hosting mini-incinerators had become significantly poorer than the county.
Similarly, median family incomes in all but one of the communities hosting landfills decreased relative to those of Harris County between 1970 and 1990. In addition, all but one of the communities in which landfills were sited suffered marked declines in their housing values relative to Harris County in the decades following the sitings.
According to my analysis of data from the census closest to the date of the siting decisions in question, the siting process had a disproportionate effect upon African Americans. But it also provides considerable support for the theory that market dynamics contribute to the burden LULUs impose upon people of color and the poor. The data I examined lend force to the argument that LULUs change a community's demographics by driving down property values. True to that argument's prediction, the homes surrounding the landfill sites in most of the host communities became less valuable properties relative to homes in other areas of Harris County after the landfills were sited. The host communities then became increasingly populated by African Americans and the poor.
Implications
Using demographic data from the census nearest in time to siting decisions (rather than data from the most recent census) and then tracing changes in demographics significantly changes the implications of the GAO and Bullard studies. My analysis of the sites in the GAO study indicates a correlation between neighborhood demographics and siting decisions, but suggests no evidence that market dynamics are forcing the poor or people of color to "come to the nuisance." My analysis of the sites in Bullard's study, on the other hand, indicates that market dynamics may play a significant role in the distribution of the burdens LULUs impose. This finding suggests that even if siting processes can be improved, market forces would be likely to create a pattern in which LULUs become surrounded by people of color or the poor.
My research shows that we can make no easy generalizations about the cause or causes of the current inequity in the distribution of LULUs. More data and analysis are needed to prove either that discrimination in the siting process is the sole cause of this inequity or that both siting decisions and housing-market dynamics—demographic changes caused by a LULU's effect on property values and by discrimination in the sale and rental of housing—are to blame.
If further study of LULU sitings confirms the findings of my analysis of the sites in Bullard's study, however, the debate about the fairness of the distribution of environmental risks would have to shift gears. It would become a debate, not just about the process of siting LULUs, but also about the free market and poverty and racial segregation in residential areas. Moreover, discussions about remedies would have to extend beyond the siting process, because changes in this process would be unlikely to achieve real, long-term improvement.
Vicki Been is an associate professor of law at New York University. This article is based on a seminar she gave at Resources for the Future on October 7, 1993. For a complete description of the above-described research, see "Locally Undesirable Land Uses in Minority Neighborhoods: Disproportionate Siting or Market Dynamics?" in The Yale Law Journal, vol. 103 (1994), p. 1383.
A version of this article appeared in print in the May 1994 issue of Resources magazine.
A version of this article appeared in print in the May 1994 issue of Resources magazine.
