Public information as a deterrent to environmental infractions

Abstract

Past studies of the impact of publicly announced violations of environmental laws on the market value of the offending firms have yielded equivocal results. A particularly troubling aspect of this research is whether the market imposes a reputation penalty for such violations. This article uses the event-study methodology to estimate the effect of announced environmental infractions on the market returns of the violators. Stockholders experienced significant negative abnormal returns in the event window. Cross-sectional analysis showed these cumulative losses to be a function of firm size, type of legal action and specific toxin. The reputation component of abnormal returns was influenced by these factors as well.

Notes

¹ Cohen (1992, p. 1066) refers to such violations as conditionally deterred offences; i.e. activities society wants to discourage but not entirely prohibit.

² For instance, Karpoff, Lott and Rankine (1998, p. 8) note that the penalties for ‘environmental violations depend to an unusually large extent on the discretion of prosecutors, courts and juries’.

³ Russo and Fouts (1997) see environmental policy as a source of significant market advantage. They note that firms from Atlantic Richfield Company (ARCO) to Procter & Gamble attribute part of their profitability to a reputation for pro-environment corporate behaviour. Furthermore, environmental reputation must support a reputation for product quality. Damage to a firm's environmental reputation can impact its overall reputation, leading to reduced expected future cash flows. Arora (2000) concludes that firms falling short of prevention expectations realize a decline in market value while companies that meet or exceed their pollution prevention goals do not earn positive abnormal returns during the event window.

⁴ For example, a firm may be disqualified from bidding on government contacts.

⁵ For example, Helland (1998) found that avoidance of compliance costs added 2.8% to the probability of an environmental violation among a sample of pulp and paper manufacturers.

⁶ The probability of being discovered also varies with the type of waste produced by the company. Some emissions, like smoke, are easy to detect, while other chemicals are more difficult to find.

⁷ Announcements of environmental infractions provide information about a firm's attitude toward abatement and disposal of potentially hazardous chemicals. The ethical investor or socially responsible investing hypothesis would foresee negative abnormal returns resulting from unanticipated announcements of environmental infringements. Unexpected information about abatement activities would elicit positive returns (Wall, 1995).

⁸ The deep pockets effect was measured as the ratio of the estimated remediation cost to the number of potentially responsible firms that were publicly traded.

⁹ The sample of firms studied was relatively small. The authors began with 77 companies (out of an original 283) that had announced penalties but eliminated 53 that were judged potentially biased, so their conclusions are based on a sample of only 24 companies.

¹⁰ Both the pre-event windows −30 to −1 and post-event windows +1 to +30 are reported as well.

¹¹ Concurrent events include major announcements of plant closures, suspensions of merger talks, introduction of new products, stock splits and similar announcements that might cause a firm's market price to vary irrespective of the announced environmental infraction.

¹² The chemical categories listed in Table 3 have recognized or suspected health effects such as cancer, developmental toxicity, reproductive toxicity, neurotoxicity, cardiovascular disease and skin disorders.

¹³ Quoted in Karpoff et al. (1998, p. 7).

¹⁴ Breusch–Pagan tests (Gujarati, 1995, pp. 377–8) rejected the null hypotheses that the models’ regression residuals were homoskedastic at a p-value of 0.0001. The t-values and SEs are based on White's heteroskedasticity-consistent variance matrix.

¹⁵ They limited their sample to electric utilities and oil companies.

¹⁶ Uncertainty about the varying hazards associated with different toxins may help explain Hamilton's (1995) findings: the larger the number of chemicals a firm reported on the TRI, the greater is its market penalty.

¹⁷ A joint test of the subset of toxin dummies and interaction terms yielded an F-statistic of 2.048, which is significant at the 0.01 level (F _critical = 1.774). Thus, this subset of variables adds to the overall explanation of abnormal returns. See Pindyck and Rubinfeld (1998, pp. 128–130).

¹⁸ Actually, the companies had to meet two criteria. First, the monetary sanctions had to be announced with the environmental infraction. Second, the announcement of the environmental infraction had to elicit a negative market response.

¹⁹ In this estimation, the predicted is a function of the actual monetary sanction announced in the legal action and the expected reputation effect estimated from Model 3. Theoretically, the abnormal return should reflect the reduction in expected cash flow due to both the monetary sanctions and the loss of reputation among stakeholders and third parties. The predicted values from Model 3 are proxies for the unobserved reputation component of abnormal returns estimated from a set of structural variables (Table 3) that influence decisions about a firm's reputation in cases involving environmental infraction.

²⁰ Karpoff et al. (1998) and Jones and Rubin (2001) found no reputation effect for the average violator.