Environmental disclosure in Spain: Corporate characteristics and media exposure

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=refc20 Spanish Journal of Finance and Accounting / Revista Española de Financiación y Contabilidad ISSN: 0210-2412 (Print) 2332-0753 (Online) Journal homepage: https://www.tandfonline.com/loi/refc20 Environmental Disclosure in Spain: Corporate Characteristics and Media Exposure Manuel García-Ayuso & Carlos Larrinaga To cite this article: Manuel García-Ayuso & Carlos Larrinaga (2003) Environmental Disclosure in Spain: Corporate Characteristics and Media Exposure, Spanish Journal of Finance and Accounting / Revista Española de Financiación y Contabilidad, 32:115, 184-214, DOI: 10.1080/02102412.2003.10779479 To link to this article: https://doi.org/10.1080/02102412.2003.10779479 Published online: 15 Jan 2014. Submit your article to this journal Article views: 118 Citing articles: 16 View citing articles articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 184 Spanish Journal of Finance and Accounting nº 115, April 2003 pp. 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE. By Manuel García-Ayuso Universidad de Sevilla Carlos Larrinaga Universidad de Burgos Fourth draft: February 17, 2003. Address correspondence to: Manuel García-Ayuso Departamento de Contabilidad Universidad de Sevilla Avenida Ramón y Cajal, 1 41018 Sevilla SPAIN E-mail: [email protected] Manuel Garcia-Ayuso is Carlos Cubillo professor Accounting and Auditing at the Department of Accounting of the University of Sevilla. Carlos Larrinaga is associate professor in accounting at the Department of Business Administration of the University of Burgos. Financial support of the DGICYT of the Spanish Ministry of Education (Projects SEC2001-2633 and PB98-0415) is gratefully acknowledged. Manuel García-Ayuso also thanks the funding provided by the Spanish Association of Accounting and Business Administration to the articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 185 Carlos Cubillo Chair in Accounting and Auditing. We are indebted to Carmen Pineda for making her databases available for this study. The authors appreciate comments from two anonymous reviewers. ABSTRACT Social and environmental issues have become a major concern for accounting research over the past two decades. Social and Environmental Accounting has attracted the attention of a number of researchers attempting to understand, explain and predict the disclosure of information on the social and environmental implications of business activities. Empirical research has hypothesized that size, profitability and the potential environmental impact of the firm are the main factors explaining the amount of information disclosed. On the other hand, several studies have focused on the motivations for disclosing environmental information, hypothesizing that disclosures are aimed at building or sustaining corporate legitimacy. We test the main hypotheses developed to date by empirical research with regard to the disclosure of environmental information based on a sample of companies listed on the Madrid Stock Exchange. Results of a content analysis show that firms disclosing environmental information tend to be larger, have higher risk (measured by the beta coefficient) and operate in industries that have a high potential environmental impact. The environmental implications of the activities carried out by these companies also seem to receive more attention from print media. Our results also provide evidence that two factors directly associated with the amount of environmental information disclosed are the potential environmental impact of the industry and the extent of media coverage of the firms. KEYWORDS Environmental Disclosures, Corporate Characteristics, Content Analysis, Media Coverage, Spain. DATA AVAILABILITY Data used in this study was collected from publicly available sources and may be obtained from the authors upon request articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 186 1. INTRODUCTION The amount and the quality of environmental information disclosed by business companies has increased remarkably over the last few decades. Surveys have documented an increase in environmental disclosures in the late eighties and nineties in Australia (Deegan and Gordon, 1996), Spain (Archel, 2001; Moneva and Llena, 2000), the United Kingdom (Gray et al., 1995a) and elsewhere. According to an international survey conducted by KPMG (2002), 49% of the top 100 companies in 19 countries disclose social and environmental information as part of their annual reports, with 23% of them publishing a separate social and environmental report. Furthermore, there is increasing evidence on the usefulness of environmental information for the analysis of the financial position of companies as well as for other purposes (e.g. their inclusion in ethical funds or in sustainability indexes).The study carried out by Deegan and Rankin (1997) revealed that environmental disclosure is relevant for a wide variety of users of the financial reports, such as shareholders, individuals within the company and review organizations. Their results also indicate that potential users consider that the annual report is more important than any other source of information in order to understand the environmental performance of the corporation. The relevance of environmental accounting has also been recognized by standard setting bodies. Today, several European countries require companies to disclose environmental information as part of their financial reports. In Spain, an accounting standard issued in 1998 requires business companies to disclose details about their environmental expenses, assets and liabilities in the notes to the financial statements (Larrinaga et al., 2002). The increase of social and environmental accounting has attracted the attention of research which has attempted to codify and explain an area of corporate activity which lies outside the domain of conventional accounting (Gray et al., 2001). A major concern for social and environmental accounting research has been the analysis of the relationship between corporate characteristics and the extent and quality of the information disclosed by firms. articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 187 Empirical studies have hypothesized that large and profitable firms, as well as those listed in industries that have a greater potential impact on the environment (sensitive industries)1, tend to disclose more social and environmental information (see Gray et al., 2001 for a review). Another stream of research has focused on managerial incentives for disclosing environmental information. Particularly interesting are the research studies conducted within the framework of the so-called Legitimacy Theory (hereafter LT), according to which environmental information is disclosed in order to build or sustain corporate legitimacy, i.e. the societal acceptability of corporate actions (see Deegan, 2002 for a review). A good example of this approach towards the analysis of environmental disclosure is the study by Deegan and Rankin (1996) documenting that Australian companies only revealed information that was favorable to their corporate image and that the amount of positive information was even larger in the case of companies that had been successfully prosecuted for environmental offences. This paper tests in the Spanish context the hypotheses developed and tested in other countries by previous empirical studies. We first analyze the relationship between several corporate characteristics and the amount of environmental information disclosed by business companies and, then assess the extent to which environmental disclosures are driven by managers’ desire to gain or enhance corporate legitimacy. The remainder of the paper is organized as follows. The next section reviews previous empirical studies on environmental disclosures and develops the hypotheses tested. Section three describes the method and data analyzed and section four presents a discussion of our results. The fifth section contains a summary and some concluding comments. 1 Deegan and Rankin (1997, p. 81) define environmental sensitivity based on “a questionnaire administered to environmental lobby groups in which office bearers were required to rate industries (on a 0-5 scale) on the basis of whether the industry had been made the focus of action as a result of its environmental activities” articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 188 2. PREVIOUS RESEARCH Empirical evidence on environmental disclosure in Spain Because environmental accounting has not been a major issue in the agenda of Spanish researchers, the empirical evidence available in this area is scarce. However, there seems to be now a growing number of studies focusing on environmental disclosures. An early study was carried out by Carmona and Carrasco (1988). They analyzed social and environmental information disclosed by 61 Spanish firms in their 1985 annual report and found that the profitability was not consistently related to the amount of social and environmental information disclosed. More recently, Moneva and Llena (1996) analyzed social and environmental disclosures made in 1992 by 47 industrial companies. They reported that 33 firms (70%) disclosed some information on environmental issues, but only 6 firms (13%) included some reference in the notes to the financial statements. Not surprisingly, a greater amount of environmental information was found in the annual reports of electric utilities, and the firms in the pulp and paper and chemical industries. In general, social and environmental information was simply a description of their environmental activities. No relationship appeared to exist between the decision to disclose social and environmental information nor was there different measures of profitability or the existence of a specific market regulation. In a further survey, Moneva and Llena (2000) analyzed the environmental disclosure practices of 70 companies in 1992, 1993 and 1994. The information disclosed was found to be rather general, lacking details, included among the voluntary information in the annual reports and concentrated in regulated industries. They also reported a consistent increase in both, the number of disclosing companies and the extent of environmental disclosure. Finally, Archel and Lizarraga (2001) surveyed environmental disclosures in the 1995 to 1998 annual reports of a sample of Spanish companies. They consistently found that both, size and environmental sensitiveness are corporate characteristics that explain the extent of environmental information. articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 189 This paper extends this research by providing an analysis of the relationship between the disclosure of environmental information and several corporate characteristics, industry factors and the level of external scrutiny of the firm’s activities. Determinants and motivations of environmental disclosures In order to justify our empirical tests, we next discuss the nature and potential impact of the main factors driving the disclosure of environmental information by business companies. Size There is consistent evidence on the relationship between different measures of size and environmental disclosures (see Gray et al., 2001 for an international review and Archel and Lizarraga, 2001, for Spain). This evidence has received different interpretations depending on the theoretical framework adopted by the researcher (Hackston and Milne, 1996; Gray et al., 2001). However, it could be argued that larger firms are subject to stronger pressure from stakeholders and, consequently, they are expected to find more persuasive arguments (depending on the theory) to disclose environmental information. Accordingly, we test whether there is a positive relationship between the amount of environmental information disclosed and the size of the firm. Risk Financial accounting information provides investors with useful information that helps them reduce the possible errors when attempting to identify the best value creation opportunities. Reducing the risk of economic losses will likely reduce the cost of capital (Bushman and Smith, 2001). There is in the accounting literature a general agreement on the idea that by disclosing more information a company is able to lower its cost of capital at the possible expense of incurring losses due to the disclosure of proprietary information (Dye 2001, p. 224). More specifically, Verrechia (2001, p. 166), states that the disclosure of financial information by managers reduces the information asymmetry component of the cost of capital, that is, the factor by which investors discount firm equity offerings in anticipation of transaction costs that may arise from adverse selections in the event that they liquidate articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 190 their equity holdings at some future date. Li, Richardson and Thornton (1997) have provided empirical evidence supporting the hypothesis that firms are more likely to disclose environmental information when the risk of incurring proprietary costs decreases. In summary, based on the conclusions of previous theoretical and empirical studies on voluntary disclosure, firms with higher levels of risk are expected to disclose a greater amount of voluntary financial information. During the period analyzed in our study, Spanish accounting standards did not require the disclosure of information on environmental issues, with the exception of the allowances arising as a result of probable future environmental liabilities. The disclosure of voluntary information on environmental issues during that period suggests that the managers of Spanish firms were willing to reduce information asymmetries providing investors with the data they consider relevant regarding the environmental performance of the firm. In order to test for the existence of an association between risk and the disclosure of environmental information by Spanish companies we analyzed the extent to which the market model beta is associated with the decision to disclose and the amount of information disclosed. Trotman and Bradley (1981) correctly point out that it is far from clear whether systematic risk would be lower for firms that show environmental responsibility (i.e., disclose environmental information) or higher, in the understanding that managers could see environmental disclosure as a means of reducing risk. Therefore, we make no a priori prediction on the sign of the relationship between betas and the extent of environmental disclosure. Profitability Although Hackston and Milne (1996) suggest that that profitability allows management enough flexibility to detract funds from shareholders to undertake social and environmental programs and, therefore, there could be a direct relationship between corporate profitability and the relevance attached by the firm to environmental issues, published empirical studies have not documented the existence of a consistent relationship between profitability and the level of environmental disclosures (see Gray et al., 2001 for an international review and Carmona and Carrasco, 1988 and Moneva and Llena, 1996 for Spain). Gray et al. (2001) found a significant but weak association between profitability and articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 191 environmental disclosure in the U.K. In an attempt to provide further evidence on this issue within the Spanish context, we test whether the level of environmental disclosure is associated with corporate profitability without making any a priori prediction on the sign of such a relationship. Environmental sensitivity It has been argued that the more attention one industry receives from stakeholders, the greater the incentives for companies in that industry to make environmental disclosures (Deegan and Gordon, 1996; Bewley and Li, 2000). In that regard, environmental sensitivity is primarily driven by the potential (or actual) impact that the firms operating in a given industry may have (or have had) on the environment.2 With some exceptions (Alnajjar, 2000), empirical research has found that firms operating in more sensitive industries tend to disclose a greater amount of social and environmental information (Patten, 1991; Roberts, 1992; Walden and Schwartz, 1997; Bewley and Li, 2000; and Archel and Lizarraga, 2001). Therefore, we test whether the level of environmental disclosure is higher in the case of Spanish firms operating in sensitive industries. Deegan and Gordon (1996) assessed the environmental sensitiveness of different industries by means of a survey among environmental lobbying groups. According to the responses received, the top ten sensitive industries were: uranium mining, chemicals, coal, transport, oil/gas exploration, plastics manufacturing, oil/gas production, gas distribution, and paper and timber manufacturing. These results are consistent with the classifications used in other studies (Patten, 1991; Roberts, 1992; Hackston and Milne, 1996), except that none of the later studies include the transportation industry. In view of the companies included in our sample we identified four sensitive industries: oil and mining, electricity and gas, chemical and pulp and paper. Media exposure 2 One could argue that the potential to have a negative impact on the environment is a firm-specific characteristic, as companies within the same industry may carry out different activities, be settled in distant locations and have designed and implemented alternative production procedures. However, given the articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 198 Spearman rank-order correlation coefficients4 were computed for the LINE and SPACE variables obtained from the reports filed in the CNMV (line1 and space1) and from the brochures prepared for the shareholders’ general meeting (line2 and space2). The absolute amount of environmental disclosure (LINE) is greater in the brochures distributed by firms among their shareholders. However, the values of the SPACE variable are similar in the reports filed in the CNMV and the brochures delivered at the shareholders’ meeting. This indicates that the amount of environmental information included in the annual reports filed in the records of the CNMV is consistent with the amount disclosed in the brochures subsequently distributed by listed firms among their shareholders. This is confirmed by the statistics in panel B of TABLE 1. The amount of environmental information included by Spanish listed companies in the reports filed in the CNMV and in the brochures they disclose in their general meetings are highly and significantly correlated. Therefore, the financial reports available in the records of the CNMV provide a reliable basis for the analysis of the environmental disclosure policies of Spanish listed companies. TABLE 1 Comparative analysis of environmental disclosure in a randomly chosen subset a of the sample CNMV reports and their parallel annual brochures (1991-1995) b PANEL A: Descriptive Statistics N=116 line1 line2 space1 space2 Mean 7.89 12.81 2.88 2.22 Std. Dev. 17.72 27.86 6.43 4.17 Max. 82.00 151.00 40.71 21.51 Min. 0.00 0.00 0.00 0.00 PANEL B: Spearman rank-order correlation coefficients between environmental disclosure in CNMV annual reports and annual brochures line1 space1 line2 0.873*** space2 0.856*** a see footnote 3 b line1 is the total number of lines of environmental information, for the annual reports submitted by firms to the CNMV. 4 We decided to compute Spearman rank correlation coefficients as the variables analyzed were not found to be normally distributed. articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 199 line2 is the total number of lines of environmental information, for the annual reports published by firms. space1 is the number of lines of environmental information relative (per thousand) to the total information disclosed in the annual report, for the annual reports submitted by firms to the CNMV. space2 is the number of lines of environmental information relative (per thousand) to the total information disclosed in the annual report, for the annual reports published by firms. *** Significant at the 99% level. Profitability and other corporate characteristics Profitability was measured in accounting terms computing both, the Return on Assets (ROA) and the Return on Equity (ROE) ratios. The measure of earnings used (Xit) was earnings before extraordinary items and discontinuing operations. The book value of equity and the total asset figures were used in our calculations as shown in the balance sheet. We computed annual stock returns as the average monthly return observed during the twelve months following the third month after year-end. Market betas (âit) were estimated using a minimum of 36 consecutive stock returns and a maximum of 60 prior to the balance sheet date. Size and leverage are two corporate characteristics frequently considered as explanatory variables in empirical studies of accounting disclosure. We measured size as the logarithm of the total market capitalization of the firm. The measure of leverage used in our analysis was the total debt to equity ratio. Moreover, we used two other measures that have been thoroughly analyzed by recent research because of their relevance for investment decision making. The Earnings-to-Price and the Book-to-Market ratios were computed using stock prices at the end of the third month after year-end. Finally, firm size (Sit) was measured as the logarithm of the total market capitalization of the firm three months after the balance sheet date. Environmental sensitivity According to the discussion in section three, the environmental sensitivity variable (Dit) is assumed to be consistently associated with the pressure that society exerts the company because of the environmental visibility of the industry in which it operates. For the purposes of this study, environmental sensitivity is proxied by using a dummy variable that articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 200 takes on the value 1 for firms in the oil and mining, electricity and gas, chemical and pulp and paper industries and 0 for all other industries. Media exposure In order to develop a measure of the media coverage of the environmental impact of a given firm, we counted in each year the number of articles in the main Spanish newspapers and magazines containing any reference to the relationship between the company and environmental issues. Articles were obtained from the records of the BARATZ database, which includes all articles on economic, social and political issues published by 33 Spanish periodicals and journals since 1981. A search was carried out for each corporation, using as keywords the name of the company and environment (in Spanish medio ambiente). It is very unlikely that a Spanish article dealing with environmental issues will not include this word. Subsequently, the search results were carefully examined to exclude articles that did not specifically relate each company with environmental issues. The Media Articles variable is measured as the number of articles published in a given year dealing with the relationship between a firm and the environment. To make our results comparable with those of Brown and Deegan (1998), we selected negative articles, i.e. those providing the view that the activities carried out by the company have a negative impact on the environment. Therefore, the Negative Media Articles variable is measured as the number of articles that, for a given year, provide an unfavorable view of the corporate activities in relation to the environment. Finally, in order to evaluate the general evolution of the concern for environmental issues in Spain, we also considered the total number of articles dealing with the environment Total Environmental Articles for each year. Thus, we computed the ratio of the total number of articles including the term environment, to the total number of articles included in the database for each year. articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 201 4. RESULTS TABLE 2 presents descriptive statistics for the total sample of firms (panel A) and the sub-sample of firms disclosing environmental information (panel B). A comparison of the statistics in both panels provides some preliminary evidence on the existence of differences in the environmental disclosure policies of Spanish companies. TABLE 2 Descriptive Statistics a PANEL A: Full Sample (1991-1995) N=560 Mean Median St. Dev. Max Min Skewness Kurtosis Pit 2601.84 1687.50 2670.70 21460 37.22 2.212 10.573 Bit 2560.12 1867.08 2802.28 33254.3 -753.72 4.251 36.489 Xit 183.65 108.03 532.02 5444.02 -1535 4.062 36.582 Sit 23.801 23.813 1.758 28.246 19.472 0.046 2.882 βit 1.324 1.269 0.903 3.947 -0.369 0.385 2.442 ROE -0.030 0.066 0.949 3.349 -18.011 -15.172 280.66 ROA 0.052 0.066 0.098 0.842 -0.587 -1.008 20.475 L 0.824 0.541 3.297 28.571 -48.309 -4.519 126.40 MRG 0.107 0.086 0.383 4.884 -4.944 -0.737 117.27 ATR 0.837 0.751 0.574 3.243 0.022 1.386 5.187 LINE 5.45 0.00 16.00 145.00 0.00 4.508 24.756 SPACE 2.35 0.00 6.10 40.71 0.00 3.616 14.271 Media Articles 0.61 0.00 1.57 11.00 0.00 4.006 18.215 Negative Media Articles 0.13 0.00 0.48 4.00 0.00 4.515 23.025 PANEL B: Firms Reporting Environmental Information Between 1991 and 1995 N=142 Mean Median St. Dev. Max Min Skewness Kurtosis Pit 2915.15 2015 3018.19 21460 73 2.73 12.41 Bit 2279.58 2027.08 1567.98 6592.56 -163.53 0.913 0.400 Xit 153.63 127.78 291.79 1030.21 -735.71 -0.172 1.260 Sit 24.612 24.784 1.833 28.245 20.088 -0.257 -0.134 βit 1.624 1.592 0.776 3.557 -0.011 0.039 0.067 ROE -0.017 0.062 0.554 0.598 -5.451 -8.580 83.881 ROA 0.053 0.074 0.080 0.154 -0.274 -2.152 8.086 L 0.855 0.655 1.301 9.681 -2.643 3.505 24.801 MRG 0.106 0.108 0.174 0.508 -0.907 -2.120 12.627 ATR 0.709 0.615 0.427 2.659 0.134 1.706 6.775 LINE 21.492 10 25.844 145 2 2.182 5.491 SPACE 9.28 5.49 9.102 40.714 0.783 1.649 2.058 Media Articles 1.58 0.00 2.63 11 0 2.083 3.590 Negative Media Articles 0.36 0.00 0.78 4 0 2.493 6.091 a Pit is the end of the year stock price, Bit is the end of the year book value of equity, Xit is earnings before extraordinary items and discontinuing operations, Sit is the firm’s size measured as the logarithm of the total year end market capitalization value, βit is the market model beta estimated on the basis of 36 monthly returns, ROE is the return on common shareholders’ equity, ROA is return on articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 202 assets, L is the debt to equity ratio, MRG is earnings to sales ratio, ATR is the assets turnover: sales divided by total assets, LINE is the total number of lines of environmental information, SPACE is the number of lines of environmental information relative (per thousand) to the total information disclosed in the annual report. The median size of firms reporting environmental information is greater than the median size of the firms in the total sample. The difference in the average firm size between the two groups is significant at the 99% level. Although no causality relationship may be inferred from this result, it appears to be consistent with the view that environmental information is disclosed as a result of external pressure and that larger firms are likely to be affected by a greater pressure than smaller firms. Consistent with the findings of Trotman and Bradley (1981), who found that Australian companies reporting on social issues showed a higher beta, we found that the average market model beta was higher for firms reporting environmental information (see TABLE 2). This suggests that if there is a specific risk factor in firms disclosing environmental information (that could be related to the existence of probable future environmental costs and liabilities) beta either captures it or, at least, is directly associated with it. The mean (median) market model beta of the firms disclosing environmental information was 1.624 (1.592) and the average (median) beta of companies not reporting on environmental issues 1.230 (1.094). The difference in means was significant at the 99% level.5 The Earnings-to-Price and Book-to-Market ratios6 were lower for the sub-sample of firms reporting environmental information. However, no significant differences were found in ROE, ROA or gross margin. The mean and median asset turnover (sales-to-total assets) ratio was significantly lower for the firms disclosing environmental information and their average and median leverage (debt-to-equity) were higher than in the case of firms not disclosing information on environmental issues. This may be due to the fact that most firms in this subsample are capital intensive and operate with a high leverage. 5 As in all empirical studies in capital markets based research in accounting, the significance of the statistics used in our analysis does not imply the existence of causality. The fact that the average size or the mean beta are significantly greater for firms disclosing environmental information does not imply that size or risk are the factors driving the voluntary disclosure of information on environmental issues. articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 203 An examination of the print media coverage of the environmental implications of Spanish corporations (TABLE 3) reveals that the mean media coverage, total and negative, is larger for firms disclosing environmental information than for the full sample. This finding is consistent with the view (although it does not imply) that firms make environmental disclosures as a response to the societal environmental demands, mirrored in the media coverage of their environmental implications. The trend in environmental disclosure in Spain TABLE 3 provides a summary of the evolution of the environmental disclosure measures between 1991 and 1995. 25.4% of the companies analyzed were found to disclose some information related to environmental issues, with a significant increase from 1991 to 1993, and a decrease thereafter. Moneva and Llena (1996) found that 70% of Spanish companies disclosed some environmental information in 1992. However, their sample was limited to the industries in which we found a greater amount of companies disclosing environmental information. Both, the absolute and the relative number of lines of environmental information in the annual report were significantly higher (at the 95% level) in all years (except in 1991) for the firms reporting environmental information in all years than for those reporting less frequently. The results presented in TABLE 3 also indicate that the amount of environmental disclosure in Spanish annual reports varies significantly across years. TABLE 3 Average amount of environmental information in the annual reports, by year. 1991 1992 1993 1994 1995 TOTAL Number of firms disclosing environmental information 23 28 33 31 27 142 Percentage of the sample 20,5% 25,0% 29,5% 27,7% 24,1% 25,4% LINE 29,65 17,82 21,91 22,10 17,15 21,49 6 This statement is based on the ratio of average earnings to both, average book value of equity and average price in panels A and B of table 2. The statement also holds for the average and median Earnings-to-Price and Bookto-Market. Descriptive statistics for these ratios are not reported but are available from the authors upon request. articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 204 SPACE 10,82 8,73 9,37 9,12 8,60 9,28 LINEFREQ 45,92 30,17 40,92 38,58 24,17 SPACEFREQ 14,07 12,62 13,71 13,70 11,74 LINE is the total number of lines of environmental information. SPACE is the number of lines of environmental information relative (per thousand) to the total information disclosed in the annual report. LINEFREQ is the LINE variable computed only on the basis of the sample of companies that mentioned the environment regularly from 1991 to 1995 in their annual reports. SPACEFREQ is the SPACE variable computed only on the basis of the sample of companies that mentioned the environment regularly from 1991 to 1995 in their annual reports. Size In order to test whether larger firms disclose a greater amount of environmental information we retained all firms disclosing some environmental information and regressed our two measures of the absolute (LINE) and relative (SPACE) amount of environmental information in the annual report on a measure of firm size and the market model beta. Results in panel A of TABLE 4 indicate that firm size has only a marginal explanatory power for the total number of lines of environmental information included in the annual report. However, this result was only found in the pooled regression, and thus might be due to the existence of cross-sectional correlation in the residuals. To control for the size of the annual report, we regressed our SPACE variable on firm size and beta. Results in panel B of TABLE 4 indicate that firm size does not explain the cross-sectional variations in the space devoted to environmental information in the annual report. Overall, although there appears to be a positive relationship between the size of companies and the decision to disclose environmental information, the association was not found to be statistically significant. Consequently, although the average size of companies reporting environmental information is significantly greater than the average size of firms not disclosing information on environmental issues (TABLE 2), we cannot conclude that differences in size across firms explain the differences in the amount of environmental information included in their annual reports. TABLE 4 a b Explanatory power of firm size and market model beta for the total and relative amount of environmental information disclosed in the annual reports Panel A: Total amount of environmental information. articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 205 Model: LINEit = φ0 + φ1 Sit + φ2 βit + εit φ0 φ1 φ2 R2 F (Prob) Pooled Sample -55.549 (-1.65) 2.884** (2.12) 4.962 (1.55) 0.045 3.611 (0.03) 1991 -196.57 (-1.13) 8.363 (1.12) 11.611 (1.48) 0.001 0.933 (0.41) 1992 -36.649 (-0.61) 2.257 (0.92) 2.156 (0.40) 0.001 0.501 (0.61) 1993 -10.227 (-0.12) 0.878 (0.24) 8.567 (0.92) 0.001 0.536 (0.59) 1994 -81.021 (-1.13) 4.103 (1.39) 2.171 (0.29) 0.011 1.143 (0.33) 1995 -50.265 (-1.02) 2.585 (1.29) 1.985 (0.38) 0.005 1.060 (0.36) Panel B: Total amount of environmental information. Model: SPACEit = φ0 + φ1 Sit + φ2 βit + εit φ0 φ1 φ2 R2 F (Prob) Pooled Sample -0.707 (-0.05) -0.328 (-0.56) 0.449 (0.87) 0.000 0.825 (0.44) 1991 -47.401 (-1.00) 1.934 (1.09) 2.222 (1.21) 0.000 0.861 (0.44) 1992 16.764 (0.48) -0.633 (-0.66) -0.220 (-0.15) 0..000 0.228 (0.79) 1993 13.494 (0.41) -1.117 (-0.48) -0.074 (-0.05) 0.000 0.140 (0.87) 1994 -4.750 (-0.19) -2.712 (-1.37) 0.724 (0.72) 0.024 1.312 (0.28) 1995 -2.926 (-0.131) -2.680 (-1.12) 0.543 (0.63) 0.000 1.188 (0.32) a LINE is the total number of lines of environmental information. SPACE is the number of lines of environmental information relative (per thousand) to the total information disclosed in the annual report. b White heteroscedasticity consistent t-statistics in parentheses. *** Significant at the 99% level. ** Significant at the 95% level. Risk Our second conjecture is that the amount of environmental disclosure is some function of firm's perceived level of risk. In order to test for an association between these two variables articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 206 we regressed the two measures of disclosure on firm size and the market model beta. Our results (TABLE 4) reveal that the differences in the level of risk as measured by the beta coefficient do not explain the cross-sectional differences in the relative and absolute amount of environmental information in the annual reports of Spanish listed companies. The lack of explanatory power of market betas found in our analysis is consistent with the results reported by Trotman and Bradley (1981) for Australia. Corporate characteristics and environmental disclosure In order to determine whether there are significant differences in the corporate characteristics most commonly analyzed in the accounting literature between Spanish listed firms disclosing information on environmental issues and those not reporting environmental information, we regressed different variables representing corporate characteristics (return on assets, return on equity, stock returns, size, risk, assets turnover and dividend per share) on a dummy variable taking on the value 1 for firms reporting on environmental issues and 0 otherwise. In so doing, we tested the statistical significance of the difference in means across the two groups of companies. Results not reported here indicate that significant differences only existed in size, market model beta, assets turnover and dividend per share.7 Specifically, firms reporting on environmental issues appeared to be (on average) significantly larger in size, have higher betas, lower assets turnover, higher leverage and, lower dividends per share. Therefore, it appears that there is no consistent relationship between environmental disclosures and profitability or market performance. We found no significant relationship between the firms’ stock returns and the extent of environmental disclosure. Environmental sensitivity The sample included 112 industrial firms, 28 of which operated in the more sensitive industries, i.e. electric utilities and gas, oil and mining, chemical and pulp and paper. 17 (60%) of companies in more sensitive industries made environmental disclosures in any year in the period 1991-1995, whereas 32 (38%) of companies in other industries did make any environmental disclosure. We analysed 140 annual reports for companies in more sensitive industries, 71 of which (51%) included some environmental information. From the remaining 420 annual reports, 71 (17%) included environmental information. These results clearly 7 These results are available from the authors upon request. articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 207 indicate that environmental disclosures are more often found in the annual reports of firms operating in more sensitive industries. We regressed our two measures of environmental disclosure on D, a dummy variable taking on the value one for firms in the more sensitive industries. Results in TABLE 5 suggest that there is a positive and consistent relationship between the amount of environmental information disclosed and the environmental sensitivity of the company. Panel A indicates that, in all years, the total number of lines of environmental information in the annual reports is consistently related with the sensitivity of the industry. The average total number of lines is significantly higher for firms operating in environmentally sensitive industries. Similar results, excepted for 1991 and 1992, were found when the relative number of lines of environmental information was used as the dependent variable (see panel B). TABLE 5 a b Explanatory power of environmental sensitivity for the total and relative amount of environmental information disclosed in the annual reports Panel A: Total amount of environmental information. Model: LINEit = ψ1 + ψ2Dit + εit ψ1 ψ2 R2 F (Prob) Pooled Sample 1.152*** (5.07) 18.883*** (6.95) 0.236 145.21 (0.00) 1991 1.296 (1.49) 18.975*** (2.68) 0.173 18.863 (0.00) 1992 0.794* (1.83) 15.727*** (3.58) 0.277 37.465 (0.00) 1993 1.391*** (3.29) 23.216*** (3.04) 0.225 28.950 (0.00) 1994 1.297*** (3.09) 21.137*** (3.58) 0.285 39.407 (0.00) 1995 0.985*** (2.83) 14.814*** (3.38) 0.290 37.383 (0.00) Panel B: Relative amount of environmental information. Model: SPACEit = ϕ0 + ϕ1Dit + εit ϕ0 ϕ1 R2 F (Prob) Pooled Sample 5.061*** (9.22) 7.047*** (5.71) 0.158 21.88 (0.00) 1991 6.693 5.859 0.003 1.054 articles Manuel García-Ayuso and Carlos Larrinaga 184-214 ENVIRONMENTAL DISCLOSURE IN SPAIN: CORPORATE CHARACTERISTICS AND MEDIA EXPOSURE Spanish Journal of Finance and Accounting, nº 115 Special Issue for the 26th Annual Congress of the European Accounting Association, April 2003, Seville, Spain 214 ROCKNESS J.W. (1985): "An Assessment of the Relationship Between US Corporate Environmental Performance and Disclosure". Journal of Business Finance & Accounting, Vol. No. pp. 339-354. WISEMAN J. (1982), "An Evaluation of Environmental Disclosure Made in Corporate Annual Reports", Accounting, Organizations and Society, Vol. 7, No. 1, pp. 53-63. COWEN S.S., FERRERI L.B. Y PARKER L.D. (1987): "The Impact of Corporate Characteristics on Social Responsibility Disclosure: A Typology and Frequency-Based Analysis". Accounting, Organizations and Society, Vol. 12, No. 2, pp. 111-122. PATTEN D.M. (1991): "Exposure, Legitimacy, and Social Disclosure". Journal of Accounting and Public Policy, Vol. 10, No. 4, pp. 297-308. ROBERTS R.W. (1992): "Determinants of Corporate Social Responsibility Disclosures: An Application of Stakeholder Theory". Accounting, Organizations and Society, Vol. 17, No. 6, pp. 595-612. TROTMAN K.T. Y BRADLEY G.W. (1981): "Associations between Social Responsibility Disclosure and Characteristics of Companies". Accounting, Organizations and Society, Vol. 6, No. 4, pp. 355-362. ULLMANN A.A. (1985): "Data In Search of a Theory: A Critical Examination of the Relationship Among Social Performance, Social Disclosure". Academy of Management Review, Vol. 10, No. 3, pp. 540-557.