توجه رسانه ها و موجودی انتشار مواد سمی

This paper explores the relationship between the print media and toxic releases in the first wave of Toxics Release Inventory (TRI) filings. It first studies the degree to which neighborhood characteristics like racial composition and income status associate with the number of newspaper articles written about a TRI establishment, controlling for the volume of toxic releases, industry and observable establishment characteristics. It follows up to study whether establishments that receive media attention reduce toxics releases more than those that do not. Neither a qualitative review of the articles nor regression results show any significant correlation between race or income and the likelihood of being included in media reports. A difference-in-difference approach shows a statistically significant decrease in the toxic releases of establishments that received media attention compared to those that did not.

Since 1987, all U.S. manufacturing facilities with at least 10 employees and producing more than 500 lb of each of the 320 listed chemicals, must annually report an inventory of toxic releases to the EPA. Information about these releases is then publically disseminated through the Toxics Release Inventory (TRI). Such a requirement informs the public and allows individuals to minimize or avert exposure to toxic substances.1 The requirement also creates negative publicity, which imposes a cost on firms and provides incentives to reduce the production of or prevent the release of toxic chemicals. This paper studies the role of the print media in generating such negative publicity. Specifically, it studies the role the media played when the TRI program was first implemented. With limited preconceived notions about the polluting behavior of facilities around the early years of TRI reporting, media responses at this time provide a rare opportunity to isolate and study the behavior of the media to new pollution news and study how TRI establishments responded to a sudden wave of media attention. This paper takes two perspectives on the relationship between media attention and toxic releases. It first studies the degree to which neighborhood characteristics like racial composition and income associate with the number of newspaper articles written about an establishment, controlling for the volume of toxic releases, industry and observable establishment characteristics. The results will show little association between non-white neighborhoods and media reporting. The empirical analysis then uses a difference-in-difference approach to show that facilities that receive media attention reduce toxic releases dramatically more than facilities that do not. Furthermore, establishments in non-white neighborhoods are more likely to reduce releases. The results contribute to the research on “environmental justice,” the concept that environmental risks and hazards should be equitably distributed regardless of race, color or income. Prior studies on environmental justice don't incorporate the media and instead typically focus directly on the behavior of and location decisions of individuals and firms (Boer et al., 1997, Wolverton, 2009 and Zimmerman, 1993). There are a number of reasons to believe that neighborhood characteristics like income or racial composition can affect the media's decision to report about a particular establishment. Choices over what to report are influenced by the preferences and worldviews of reporters, editors and the newspaper owner (Bennett, 1988, Entman and Rojecki, 2000, Groseclose and Milyo, 2005 and Wilson and Gutiérrez, 1995). If reporters or editors have a liberal stand on public policy issues, they may be more likely to cover issues related to the poor and racial minorities. On the other hand, the motive of profit maximization might lead them to report less on these neighborhoods. The largest media audience in the U.S. is white and middle-class (Larson, 2006 and Shirley, 1992). Reporting about poor and minority neighborhoods may not appeal to these readers. Furthermore, the costs of reporting about toxic releases in high-income neighborhoods might be lower. If higher income neighborhoods are more vocal about their disamenities (and therefore more responsive to reporters), and lower income neighborhoods attach less weight to environmental quality, then toxic releases in higher income neighborhoods may get more attention. To the extent that media activity is associated with neighborhood characteristics, the second objective of this paper is to explore how media activity might affect toxic releases. If media attention imposes costs, facilities have incentives to change their subsequent behaviors. Prior research on the TRI has explored numerous ways that the requirement to report releases affects firms, but to our knowledge no study has focused on the relationship between TRI-related media attention and behavior of the facilities.2 The difference-in-difference approach used in this study aims to provide insight into this relationship. Although the results should not be interpreted causally, they do show that establishments which receive media attention behave differently from the ones that do not. The remaining sections of this paper present a background of the TRI program and its association with environmental justice, data, measures and the empirical strategies, and results. The section on results first identifies the association between media attention and neighborhood characteristics, and then shows results on how the toxic releases of establishments with media attention differ from the releases of those establishments without media attention.

The unconventional nature of the TRI program and the success attributed to it in reducing toxic releases over two decades have made it a popular alternative to traditional environmental regulations. Although the media has played an important role in disseminating TRI-based information, economists have generally neglected to include consideration of the media and its impact in this regulatory framework. Looking back at media attention generated at the time the program was introduced, this study investigates the role that the TRI played in generating media responses and how establishments receiving media attention behaved differently from those that did not. The role of the print media is particularly interesting when viewed through the prism of environmental justice, since a number of prior authors argue that that poor and minority neighborhoods are disproportionately exposed to toxic releases (Arora and Cason, 1999 and Gamper-Rabindran, 2006). Media attention, if it were influenced by socioeconomic factors, could impose a cost on facilities that either mitigates or exacerbates the problem of environmental (in)justice. Our estimations find no evidence for this influence of socioeconomic factors. The racial composition and income of a neighborhood matter very little in media reporting of TRI-related news. Regardless of what drives media attention, facilities that are highlighted in the press diminish their releases more than those that are not. The magnitude of that difference is quite large: over 40% in column 1 of Table 5. The magnitude is sensitive, however, to exactly how media attention is measured and to the type of estimator used. While the combination of our two findings appears to suggest that media attention, targeted towards poor or minority neighborhoods, could go a long way to remedying environmental injustice, we are reluctant to endorse this conclusion. The observable characteristics of facilities with media attention are different from those without media attention. This raises the possibility that unobserved characteristics, which are correlated with media attention, might be driving decisions about toxic releases. One such possibility is a link between regulatory enforcement, media attention, and firm responses. From the estimations, we cannot be sure that media attention causes establishments to reduce their toxic releases. However, having acknowledged the limitations of a difference-in-difference approach, it is worth highlighting that the texts of the articles in our database do often point to a causal relationship. Articles frequently include quotations by company spokespersons, who would highlight plans for reducing releases. In fact, large emitters of toxic releases, like Monsanto, Union Carbide, and Dow, announced initiatives, specifically in response to the negative publicity associated with the TRI, to drastically reduce production of toxic substances (Newsweek, 1989 and Rotman, 1989).