Gender, Race, and Perceived Environmental Risk: The “White Male” Effect in Cancer Alley, LA

Abstract

Research on risk perceptions are replete with race- and gender-specific hypotheses attempting to account for attitudinal variation. However, race and gender differences may mask more notable patterns across subgroups, patterns that lie at the intersection of race and gender. Recent national studies suggest that being a White male leads to lower risk perceptions and greater willingness to accept risks. This article extends this research by examining the “White male” effect in a chronically polluted context, an area where industrial pollution is palpable and well-documented. Data are drawn from a survey of a population living in “Cancer Alley,” a stretch of the Mississippi River from Baton Rouge to New Orleans. We find that women more than men and Blacks more than Whites perceive environmental risks as serious. Further, evidence suggests that these differences are mostly due to the relatively extreme perceptions of risk accepting White males and risk adverse Black females. After controlling for select variables in hierarchical multiple regression analyses, being a White male or Black female still has a statistically significant impact on risk perceptions.

Notes

¹It should be noted that disaster researchers make a distinction between natural and technological disasters, documenting that the latter”… create a far more severe and long-lasting pattern of social, economic, cultural and psychological impacts than do natural [disasters]” (Freudenburg 1997:26). As such, using data on technological disasters to test a hypothesis drawn from studies on natural disasters is questionable.

²Those living in Cancer Alley have been somewhat empowered in the 1990 s due, in part, to the environmental justice movement. Two of the four case studies presented by Roberts and Toffolon-Weiss (2001) are in Cancer Alley. In the LES case, a community group blocked the siting of a uranium-enrichment facility. In the second case, a coalition of residents and outside groups prevented the siting of a multi-million-dollar chemical plant in Convent, LA by the Shintec Corporation.

³This survey was funded by the Louisiana Environmental Resource and Education Program (LEREP) at the University of New Orleans. Dr. Raymond Burby of the College of Urban and Public Affairs at the University of New Orleans served as the principal investigator. Sampling decisions were made based on the mission of the LEREP. The simple random sample was stratified (in proportion to parish population) to provide information that would be useful in delivering technical assistance to each Parish. The survey's focus on low- and moderate-income families reflects LEREP's mission to assist this population, since they are the most at risk from the adverse health effects of living in Cancer Alley.

⁴With a relatively low response rate, an assessment of nonresponse bias is warranted. The most common method (the other two techniques—intensive postsampling and wave extrapolation—must be used during the survey or shortly after its completion) of estimating nonresponse bias is through a population comparison, where survey averages are compared with known population averages for selected variables. Since the study population (people living within one mile of the Mississippi River) does not correspond with extant political boundaries (parishes or Louisiana), I do not have data on the population. Comparing survey averages to parish averages is doable and might be interesting, but it would not allow me to address the issue of nonresponse bias.

Cell entries are the percent reporting each risk as serious. Pearson's chi-square statistic was used to test significance.

^asignificant at p<.05;^bSignificant at p<.01;^csignificant at p<.001.

⁵One caveat worth noting is that the number of cases for the “communities with plants” subgroup is considerably greater than the “communities without plants” subgroup. Thus, the lack of statistical significance found for mean differences in the latter (e.g., water pollution and hazardous waste disposal) may be due to small number of cases analyzed.

Cell entries are the percent reporting each risk as serious. Pearson's chi-square statistic was used to test significance.

^asignificant at p<.05;^bsignificant at p<.01;^csignificant at p<.001.

Cell entries are the percent reporting each risk as serious. Pearson's chi-square statistic was used to test significance.

^asignificant at p<.05;^bsignificant at p<.01;^csignificant at p<.001.

Main cell entries are percentages of each subgroup responding that the problem is “serious.”

^aF value significant at the p<.05 level.

GLM One-Way ANOVA (two level between subjects factor) significant (p<.05) differences between:^bBlack females and White males;^cBlack females and Whitefemales;^dBlack females and Black males;^eBlack males and White males;^fWhitemales and White females.

⁶Although risk perceptions vary by subgroup, the internal consistency of the eleven risk questions is quite high for each subgroup, with a Cronbach's Alpha of .91 for White males, .90 for White females, .89 for Black males, and .90 for Black females.

Main cell entries are unstandardized regression coefficients; parenthetical cell entries are standardized regression coefficients.

^asignificant at p<.05;^bsignificant at p<.01;^csignificant at p<.001.

⁷One reason for this pattern is that with few options after slavery was abolished, former slaves continued to work on the plantations as wage laborers or sharecroppers. Nearly a century later, these plantations, as large tracts of inexpensive land next to the Mississippi River, became ideal locations for petrochemical industries. In many cases, descendants of former slaves continue to live around the perimeter of the petrochemical plants (Roberts and Toffolon-Weiss 2001).