The Emergency Planning and Community Right to Know Act (1986) has mandated Toxic Release Inventory (TRI) disclosures in the United States. This Act requires all manufacturing companies (SIC code 20-39) who employ more than 10 people to provide an annual report about the release of more than 300 specified toxic chemicals. Similar legislation exists in other countries as well. How is this information used by investors and corporations? We develop and test a regression model to answer this question. We also perform a few robustness tests. Our sample comes from TRI disclosures for "top 100 " corporate polluters based on COMPUSTAT data. Descriptive statistics and correlation measures are also provided. The higher the return on assets the higher is Tobin's q (a proxy for firm value or shareholder wealth). The waste disposal variable (toxic air release) is a statistically significant predictor of Tobin's q. As expected, the sign of the regression coefficient for waste disposal is negative. In addition, firm size has a significant impact on Tobin's q. A firm's beta, P/E ratio, and the corporate governance variable are all statistically insignificant.
The disastrous Union Carbide accident that occurred in India in 1984 and other smaller chemical accidents have caused anxiety in the public's mind about the release of chemicals from factories. The Emergency Planning and Community Right to Know Act (1986) has mandated Toxic Release Inventory TRI disclosures. This Act requires all manufacturing companies (SIC code 20-39) in the United States who employ more than 10 people to provide an annual report about release of more than 300 specified toxic chemicals. The TRI program offers environmental performance information to the public and is administered by the Environmental Protection Agency (EPA). How is this information used by investors and corporations?
EPA's Environmental Economics Research Strategy (EPA, 2004) identifies measuring the benefits of environmental information disclosures as one of its high priority research areas. Some interesting research results have already been published. For example, Konar and Cohen (1997) report negative stock price reactions to TRI disclosures in 1989. These negative stock returns forced companies to change tiieir behavior. Those firms with the largest negative stock market returns to TRI announcements in 1989 subsequently reduced their emissions more than other firms in their industry. The purpose this research project is to examine the association between the TRI disclosures and firm value as measured by Tobin's q. The goal of examining the association between the TRI thsclosures and firm value will be accomplished through the development and testing of a regression model. A few robustness tests are also conducted. Tobin's q is a widely used proxy for firm value in the finance literature (Gompers, Ishii and Metrick, 2003) and is used in this study as the dependent variable.
Several researchers have conducted event stuthes and documented negative stock price reactions to TRI announcements (Hamilton 1995 and Khanna et al. 1998). Event stuthes examine the stock price reactions on one or two days when the environmental information is thsclosed. Klassen and McLaughlin (1996) also reported significant negative stock price reactions to bad environmental news such as oil spills. These event stuthes do not analyze longer-term stock price trends. These stuthes have generally used smaller samples. Moreover, they have used data from 1989 which are eighteen years old. To overcome these thfficulties, a new regression model is developed which uses more recent data from 2000 TRI thsclosures. The TRI thsclosure data is compiled from raw data reported to the EPA on a facility-by-facility basis and not on a company-bycompany basis. The thfficulty of aggregating to company-level data makes the 2000 TRI thsclosures the most recent data currently available.
2. Prior Research
Karpoff and Lott (1993) report that when corporate illegal activities and other fraudulent financial schemes are revealed, stock price declines have been the result. In order to estimate the value of intangible assets, we propose to include environmental performance information among the explanatory variables (see Konar and Cohen 2001). Good environmental performance can translate into a good reputation for the firm as an ecology-friendly company and this can increase investor trust (Ragothaman and Lau, 2000). Similarly, bad environmental performance can lead to stock price declines.
This research builds on prior research and expands knowledge in several thfferent and new ways. 1) Data used in this study are more recent (than 1989) and come from TRI thsclosures for the year 2000; 2) Tobin's q is measured in accordance with suggestions from finance scholars; 3) The regression model includes some new variables; and a crosssectional regression model is used. Descriptive statistics and correlation measures are also provided. New insights are gained about the impact of environmental thsclosure programs on stockholders' wealth. Our formulation for Tobin's q follows that of Chung and Pruitt (1994) and Hirschey and Connolly (2005), where q is measured as the market value of common shares outstanthng plus the book value of total assets minus common equity, all thvided by the book value of total assets. Tobin's q is viewed as a marketbased measure of firm valuation. In this paper, the effect of environmental thsclosures on the market valuation of firms, proxied by Tobin's q, is examined.
Beta is a measure of the risk associated with owning shares in a firm and is commonly used to measure market risk. Konar and Cohen (1997) utilize beta to control for the systematic risk in security returns. Beta is included in this study as a control variable. Various measures of firm size appear in the literature. Dowell, Hart, and Young (2000) use the logarithm of total assets with mixed results in examining whether corporate global standards create or destroy market value. Hamilton (1995) uses the number of employees as a proxy for firm size in examining the relationship between Toxic Release Inventory data and media and stock market reactions. The logarithm of the number of employees (LEMP) is used as a proxy for firm size, and is included in the model as another control variable.
Waste (toxic air release) is measured as waste disposal in pounds per revenue-dollar. Waste should be negatively related to Tobin' s q, as it measures the extent to which firms are "dirty." Konar and Cohen (1997) use toxic chemical releases and the number of lawsuits to proxy waste. Hamilton (1995) uses the number of superfund sites to proxy waste. Return on assets (ROA), defined as net income divided by total assets, is used as a measure of firm-level performance. It is a proxy for profitability. ROA should be positively related to Tobin' s q, since better performing firms should be more highly valued in the marketplace, ceteris paribus. Hirschey and Connolly (2005) use profit margin to measure profitability.
Another control variable used in this study is the price-to-earnings ratio. The price-toearnings (PE) ratio is measured as the market price of a firm's common stock divided by the firm's income-per-share of common stock. The PE ratio is included in the model as a control variable to pick up the effect of firm-level growth. Firms that are growing rapidly should have a higher market valuation, as measured by Tobin's q. Yet another control variable used in this paper is "audit opinion" which is a proxy for the corporate governance mechanism. Li et al. (2005) found that firms with higher stock market return tended to receive more clean (unqualified) audit opinions. In other words, audit opinion is negatively related to market value of the firm. Hodge et al. (2004) conducted an experimental research project and concluded that investors reacted to audit qualifications as if it signaled that management was strategically understating financial results. It could be posited that management was concerned about future performance and consequently understated the current performance. According to Choi and Jeter (1992), audit qualifications indicate that uncertainties associated with future cash flows have increased and consequently, the future market value of the firm can be adversely affected.
3. Methodology and data sources
Researchers at the Political Economy Research Institute (PERI) at the University of Massachusetts released, in 2004, the list of the top 100 corporate air polluters based on TRI data disclosed by companies in the year 2000. The toxic (air release) waste data are reported in pounds per revenue dollar. Data from COMPUSTAT were used to compute several operating and financial ratios for these 100 firms. The following independent variables were obtained from the COMPUSTAT database: market beta, return on assets, logarithm of number of employees, P/E ratio and audit opinion. Following Hirschey (2005), the following formula is used to estimate Tobin's q: Tobin's q = [Total assets + Total market value of equity - Book value of equity] / Total assets. Tobin's q was also computed from the data obtained from the COMPUSTAT data base. Due to missing variables in the COMPUSTAT database, 9 companies were dropped. One more firm was deleted because of an extreme outlier. The final sample used in this study contains data from 90 companies.
The multiple regression model used in this study is:
Tobin 's q =/ market beta (risk), logarithm of number of employees, waste discharge per revenue dollar, return on assets, P/E ratio and audit opinion}
The research questions are transformed into null hypotheses as given below:
Hl: Beta has no significant effect on Tobin' s q.
H2: Size as measured by number of employees has no significant effect on Tobin's q.
H3: Waste discharge has no significant effect on Tobin's q.
H4: Return on assets has no significant effect on Tobin's q.
H5: Growth as measured by the P/E ratio has no significant effect on Tobin's q.
H6: Corporate governance as measured by audit opinion has no significant effect on Tobin's q.
4. Results and discussion
The descriptive statistics are reported in Table 1 . The average Tobin's q for the sample firms is 2. 176. The average amount of toxic air release (waste discharge) is 0.0009 pounds per revenue dollar. The mean for return of assets is 4.648 percent. The average beta (risk measure) is 1.121.
A correlation analysis of these six explanatory variables with the Tobin' s q and other independent variables was performed. The correlation results are reported in Table 2.
The correlation analysis results indicate that Tobin's q is strongly related to return on assets. The higher the return on assets, the higher is Tobin's q. Beta, firm size and waste discharge are all negatively related to Tobin's q. Beta and return on assets have strong negative correlation. Firm size and waste discharge are negatively correlated.
Multicollinearity among independent variables may be present in the data and can potentially lead to unstable regression coefficients. A rule of thumb is suggested by Judge et al. (1985) to assess the impact of multicollinearity. They argue that a serious multicollinearity problem arises only when correlations among the explanatory variables are higher than 0.8. In our dataset, the highest correlation is between return on assets and beta at -0.41 1. Hence, the degree of collineari ty present appears to be too small to invalidate estimation results.
An ordinary least-squares regression model was developed to investigate the relationship between Tobin's q and toxic air release, beta, return on assets, growth and other independent variables. Regression methodology permits the testing of six null hypotheses simultaneously. Tobin's q was the dependent variable and the six explanatory variables mentioned earlier were the independent variables. The regression coefficients, t-statistics (in parentheses), and significance levels are reported in Table 3, column I. The multiple regression model has a respectable adjusted R-squared of 31.3 percent.
The regression results indicate that return on assets is the independent variable that has the highest t-statistic (t = 4.343) and is significant at the 0.000 level. Thus, the null hypothesis H4 can be rejected indicating that return on assets has significant impact on Tobin's q. As expected, the coefficient is positive for return on assets. The higher the return on assets the higher is the Tobin's q. This result is consistent with prior literature on Tobin's q. Beta has a t-statistic of -0. 186 and is not statistically significant. Thus, the null hypothesis Hl cannot be rejected. However, beta has a negative coefficient consistent with prior literature.
The waste disposal variable has a significant t-statistic (t = -2.244) at a 0.028 level of significance. Thus, the null hypothesis H3 can be rejected indicating that waste disposal (toxic air release) is a statistically significant predictor of Tobin's q. As expected, the sign of the regression coefficient for waste disposal is negative and is consistent with event study results by Hamilton (1995) and Khanna et al. (1998). It is possible that some managers of firms in the private sector may want to maximize short term profits and hence, may not invest in pollution control programs. When the EPA and other regulators mandate TRI disclosures to the investing public, they are creating market mechanisms to enhance pollution reduction. Evidence from this hypothesis indicates that large polluters are punished by the capital markets as indicated by the significant negative coefficient for the waste disposal variable.
The independent variable "logarithm of employees" has a significant t-statistic (t = 3.319) at the 0.001 level. Thus, the null hypothesis H2 (size has no significant impact on Tobin's q) can be rejected. The negative coefficient for the size measure is in the expected direction and this result is consistent with prior research. The independent variable, P/E ratio, has t-statistic of 1.574 and is not significant at conventional levels. The positive coefficient for P/E ratio is in the expected direction. Hence the null hypothesis H5 (growth has no significant impact on Tobin's q) is not rejected. The final independent variable, audit opinion, does not have a significant t-statistic at conventional levels and the null hypothesis H6 is not rejected. We performed three robustness tests by including additional control variables such as R & D intensity, capital intensity and another measure of size (logarithm of property, plant and equipment) in me regression. The results are reported in Table 3 - columns II, III and IV. The results of robustness tests were essentially the same and the same three variables were significant either at 5 percent or at 10 percent level.
The purpose of this research project is to examine the association between the TRI disclosures and firm value as measured by Tobin's q. Research into the relationship between environmental disclosures and firm value has shown that the market responds efficiently to disclosure information. Results in the event-study literature use disclosure data from 1989 (because TRI disclosures were mandated for the first time in 1989), and are now dated. This paper has advanced our understanding of the literature by using new disclosure information from 2000, and by using a measure of Tobin's q that is consistent with that found in the finance literature. Results from OLS estimations show that return on assets has a significant positive effect on firm value, while the actual reported waste discharge significantly reduces firm value. Evidence from this research indicates that large polluters are punished by the capital markets as indicated by the significant negative coefficient for the waste variable. Investors may consider large polluters to be riskier investments. Consequently, investors impose a higher required rate of return on these objects and the present values of such firms will be lower. Firm size is negatively related to Tobin's q and this result is consistent with prior research. A firm's beta, P/E ratio, and the corporate governance variable were statistically insignificant.
Future work regarding environmental disclosures needs to address limitations in the literature and in this paper. Updating the results using data from 2000 is helpful, still newer data needs to be integrated into the literature. Besides providing more current estimations, additional data should provide more observations, which is a limitation of both the prior and current findings in this literature. Larger and updated datasets will increase the explanatory power of our estimations, and potentially will allow for comparative studies of the effects of environmental disclosures on an industry- byindustry basis. There are currently not enough firms included in the sample to allow for this kind of analysis. Additional data will also allow the effect of disclosures on individual companies to be studied over time.
Choi, S., and D. Jeter. (1992), "The Effects of Qualified Audit Opinions on Earnings Response Coefficients," Journal of Accounting and Economics (June-September): 229-247.
Chung, K. and Pruitt, S. (1994), "A Simple Approximation of Tobin's q," Financial Management, Vol. 23, pp. 70-74.
Gompers, P., J. Ishii, and A. Metrick. (2003), "Corporate Governance and Equity Prices," Quarterly Journal of Economics, February: 107-155.
Hamilton, J. (1995), "Pollution as News: Media and Stock Market Reactions to the Toxics Release Inventory Data," Journal of Environmental Economic and Management, Vol. 28, pp. 98-1 13.
Hirschey, M. and R. A. Connolly (2003), "Firm Size and the Effect of R&D on Tobin's q," R&D Management, 35(2), 217-223.
Hodge, F., R. Martin, and J. Pratt. (2004), "The Effect of Audit Qualifications on Investor Assessments of Financial Performance when Management Reports Conservatively in Periods of Strong Financial Performance," Working paper, Indiana University.
Judge, G, W.E. Griffiths, R.C. Hill, H. Lutkepohl, and T. Lee, (1985) The Theory and Practice of Econometrics (New York, NY: Wiley).
Karpoff J. M. & J. Lott (1993), "The Reputational Penalty Firms Bear from Committing Criminal Fraud," Journal of Law and Economics, 36: 757-802.
Khanna, M., W.H. Quimo, and D. Bojilova. (1998), "Toxics Release Information: A Policy Tool for Environmental Promotion," Journal of Environmental Economics and Management, 36, 3: 243-266.
Klassen, R. and McLaughlin (1996), "The Impact of Environmental Management on Firm Performance," Management Science, 42: 1 199-1214.
Konar, S. and Cohen, M. (1997), "Information as Regulation: The Effect of Community Right to Know Laws on Toxic Emissions," Journal of Environmental Economics and Management, Vol. 32, pp. 109-124.
Konar, S. and Cohen, M. (2001), "Does the Market Value Environmental Performance?" Review of Economics and Statistics, 83(2): 281-289.
Li, C, F. Song, and S. Wong. (2005), "Audit Firm Size Effects in China's Emerging Audit Market," Working paper, University of Hong Kong.
Ragothaman, S. and R. Lau, (2000), "The Strategic Orientations of Firm's Environmental Management Programmes: An Analysis of Survey Data," International Journal of Management, Vol. 17(4), December, 468-475.
U.S. Environmental Protection Agency. (2004), "Environmental Economics Research Strategy" EPA/600/R-04/195.
University of South Dakota
University of South Dakota
Contact email addresses: Srini.Ragothaman@usd.edu David.Carr@usd.edu
School of Business
University of South Dakota
SD 57069 USA…