Anda di halaman 1dari 25

Accounting, Organizations and Society 29 (2004) 447–471

The relations among environmental disclosure, environmental

performance, and economic performance: a simultaneous
equations approach
Sulaiman A. Al-Tuwaijria, Theodore E. Christensenb,*, K.E. Hughes IIc
Department of Accounting and Management Information Systems, College of Industrial Management,
King Fahd University of Petroleum and Minerals, KFUPM Box 1996, Dhahran 31261, Saudi Arabia
School of Accountancy and Information Systems, Marriott School of Management, Brigham Young University,
540 TNRB, Provo, UT 84602-3100, USA
Department of Accounting, E. J. Ourso College of Business Administration, Louisiana State University, Baton Rouge, LA 70803-6304, USA

This study provides an integrated analysis of the interrelations among (1) environmental disclosure, (2) environ-
mental performance, and (3) economic performance. Based on the argument that management’s (unobservable) overall
strategy affects each of these corporate responsibilities, we conjecture that prior literature’s mixed results describing
their interrelations may be attributable to the fact that researchers have not considered these functions to be jointly
determined. After endogenizing these corporate functions in simultaneous equations models, we obtain results that
suggest ‘‘good’’ environmental performance is significantly associated with ‘‘good’’ economic performance, and also
with more extensive quantifiable environmental disclosures of specific pollution measures and occurrences.
# 2003 Elsevier Ltd. All rights reserved.

Introduction common equity. This study provides an integrated

analysis of how management’s overall strategy
As managers scramble to compete in the global jointly affects (1) environmental disclosure, (2)
economy, they must do so within societal con- environmental performance, and (3) economic
straints characterized by ever-increasing environ- performance. Understanding these interrelations is
mental accountability. This accountability of increasing interest to both internal and external
includes heightened public scrutiny of both the stakeholders in an era in which corporate envir-
firm’s environmental performance and its public onmental costs have become a significant business
disclosure of that performance. These elements of expense.1
corporate environmental accountability jointly
impact the firm’s profitability and the value of its Whereas using a landfill to dump hazardous waste cost
only $2.50 per ton in 1978, this charge rose to over $200 per ton
by 1987 (Buchholtz, Marcus, & Post, 1992). Between 1972 and
* Corresponding author. Tel.: +1-801-422-1768; fax: +801- 1992, total annualized environmental protection costs for US
422-0621. firms tripled as a percentage of Gross Domestic Product
E-mail address: (GDP). Senior executives anticipate this trend to continue
(T.E. Christensen). (Walley & Whitehead, 1994).

0361-3682/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
448 S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471

Prior empirical and social-responsibility environmental performance, environmental dis-

research on the relations among environmental closure, and economic performance. If these cor-
performance, environmental disclosure, and eco- porate functions are endogenously determined,
nomic performance has, in general, considered the then piecemeal Ordinary Least Squares (OLS)
strength of pair-wise associations between two of estimation of pair-wise relations among these
these three factors, while not addressing the third. three functions will produce biased and incon-
Academic researchers from management, sistent results.
finance, and economic disciplines have focused on While accepting Ullmann’s premise that earlier
the environmental-performance–economic-perfor- research models used to explore these relations
mance relation and the root question: Is going may have been mis-specified, we consider incor-
green good for profits? Meanwhile, accounting porating an unobservable managerial strategy into
researchers have concentrated on the adequacy of an empirical model to be problematic. Instead, we
environmental disclosure in financial reporting implement Ullmann’s conceptual framework by
and its value relevance to investors: Do green dis- explicitly treating environmental performance,
closures adequately represent the firm’s exposure environmental disclosure, and economic perfor-
to future green regulation? And because these dis- mance as endogenous variables, jointly deter-
closures are largely voluntary, what factors deter- mined by the firm’s strategic management process.
mine their shade of green? While this research has In doing so, we advance the following research
advanced our knowledge in specific settings,2 no questions: First, how are the firm’s environmental
study has attempted to examine environmental performance, environmental disclosure, and eco-
performance, environmental disclosure, and eco- nomic performance interrelated after the endo-
nomic performance within a single inclusive model. geneity of these three corporate functions is
We propose a holistic approach to examine col- explicitly considered? Second, does joint estima-
lectively the relations among the firm’s (1) envir- tion of these relations significantly differ from
onmental performance, (2) environmental independent OLS estimation? If so, significant
disclosure, and (3) economic performance, using a methodological differences in estimating the coef-
conceptual framework first suggested by Ullmann ficients of the endogenous variables may be due to
(1985). Ullmann presents a descriptive analysis of bias in the OLS estimator. Documentation of such
prior social-responsibility studies that, in aggre- bias has implications for both interpreting prior
gate, report mixed empirical results of pair-wise research and planning future research designs.
associations between environmental performance To address these questions, we first specify
and economic performance, between environ- environmental performance, environmental dis-
mental performance and environmental dis- closure, and economic performance in three mul-
closure, and between environmental disclosure tivariate equations in which at least one of these
and economic performance.3 Ullmann posits that functions is an explanatory variable of another.
the inconsistent findings characteristic of these While our empirical proxies for economic perfor-
pair-wise studies reflect a common omitted vari- mance are market-based and our measure for
able—an inclusive management strategy. In environmental performance is a nonfinancial ratio
executing the corporation’s strategic business based on the relative quantity of hazardous waste
plan, management implements policies and initi- recycled, we feel that it is important to qualify our
ates decisions that simultaneously affect the firm’s measure of environmental disclosure and distin-
guish it from its more generic connotation. Within
the context of this study, environmental disclosure
For example, Barth, McNichols, and Wilson (1997) is the disclosure of specific pollution measures and
examine the determinants of environmental disclosure for firms
occurrences (toxic waste emissions, oil spills,
with Superfund liabilities.
Although Ullmann examined social-responsibility studies Superfund sites, etc.) that an investor might find
in general, approximately half (14 of 31) of these studies useful in estimating future cash flows. This defini-
focused on environmental concerns. tional constraint focuses on the disclosure of cost
S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471 449

drivers of future environmental costs and inten- growing body of interdisciplinary environmental
tionally excludes the ‘‘greenwash’’ commonly knowledge.
found in annual financial reports. Using a cross- The remainder of this study proceeds as follows:
sectional sample of 198 US ‘‘Standard & Poors the following reviews the pertinent literature and
500’’ firms, we first ensure that all sample firms frames the testable hypotheses; the next introduces
exceed a minimum threshold for exposure to the structural equations in the simultaneous equa-
future environmental costs. We then compare tions models and describes the variables along
independent OLS estimation of the relations with the data sources; next, provides the empirical
among the three corporate functions under inves- results; and the final section summarizes the
tigation, with joint estimation using two-stage study’s conclusions, implications, and limitations.
least squares (2SLS) and three-stage least squares
(3SLS) simultaneous equations models. After
controlling for endogeneity, we observe that Literature review and hypotheses development
‘‘good’’ environmental performance is positively
associated with ‘‘good’’ economic performance, The relation between environmental performance
and also with more extensive quantifiable environ- and economic performance
mental disclosures of specific pollution measures
and occurrences. Prior empirical research on the relation between
This research contributes to our understanding environmental performance and economic perfor-
of how societal concerns for the environment mance has reported mixed results. Bragdon and
affect corporate strategy and, ultimately, firm Marlin (1972) argued that pollution abatement
value. First, we recognize the endogeneity of the and profitability are compatible, and found a
firm’s environmental performance, environmental positive relation between profitability (earnings
disclosure, and economic performance, and find per share and return on equity) and the Counsel
that this research-design consideration sig- on Economic Priorities’ (CEP’s) environmental
nificantly affects the statistical significance of esti- performance ratings for pulp-and-paper firms.4
mated interrelations. Second, the significantly Spicer (1978) used firms in the pulp-and-paper
positive relation observed between environmental industry to measure the association between five
performance and economic performance suggests firm-specific variables—profitability, size, total
that managers should change their strategic out- risk, systematic risk, and the price-earnings
look regarding a firm’s environmental perfor- ratio—and the CEP’s pollution performance rat-
mance, from fixating on the deadweight costs of ex ings. His results indicated that all signs were in the
post regulatory compliance, to focusing on the ex directions hypothesized; however, only the corre-
ante opportunity costs represented by environ- lation coefficients for size, systematic risk, and the
mental pollution. Third, we find that good envir- price-earnings ratio were statistically significant.
onmental performers disclose (within the context Although both of these studies suffered from rela-
of our definition of environmental disclosure) tively low power due to small sample sizes and
more pollution-related environmental information
than do poor performers, which is consistent with
discretionary disclosure theory’s ‘‘good news’’ 4
Early US environmental studies have almost universally
explanation. Fourth, while providing additional relied on environmental performance data compiled by the
evidence regarding the determinants of environ- Council on Economic Priorities (CEP), a nonprofit corporation
mental performance, environmental disclosure, organized in 1970. Its purpose was to foster socially responsible
and economic performance, this research intro- business policies and practices. Environmental reports, of
varying quality and depth, were produced by the CEP between
duces new empirical proxies for environmental
1970 and 1977. Indices were formulated to measure the envir-
performance and environmental disclosure. onmental performance of firms from four highly polluting
Finally, this study spans the research agendas of industries: steel, oil, electric utilities, and paper and pulp
multiple academic disciplines and contributes to a (Abbott & Monsen, 1979).
450 S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471

measurement error, their findings were consistent The relation between environmental disclosure and
with the idea that good environmental and eco- environmental performance
nomic performance are complements. Adherents
to this concept of complementary association Establishing a relation between environmental
believe that acting on a firm’s social responsi- performance and environmental disclosure is
bilities (or externalities) reduces risk to which the important from a social responsibility perspective
capital markets are increasingly sensitive (Narver, in that a positive relation tends to validate the
1971). Additionally, if environmental pollution credibility of the latter. However, empirical
represents resources that have been inefficiently or research on the environmental-disclosure–envir-
incompletely used by the firm, the elimination of onmental-performance relation has generally
such waste and inefficiencies benefits both the found no significant association between the two.
environment and the bottom line (Porter & van Ingram and Frazier (1980) compared content
der Linde, 1995a, 1995b). analysis ratings of environmental disclosures that
Although these early empirical studies suggested appeared in corporate annual reports to CEP
a positive environmental-performance–economic- environmental-performance ratings. They did not
performance relation, later researchers have gen- find a significant association between environ-
erally found the association to be statistically mental disclosure and environmental perfor-
insignificant. Rockness, Schlachter, and Rockness mance. Freedman and Jaggi (1982) also reported
(1986) examined hazardous waste disposal in the insignificant results using disclosures made in
chemical industry using environmental-perfor- Forms 10-K. Using a different method to evaluate
mance data from a special site survey submitted to environmental disclosures in annual reports,
the US Congress in 1979. Testing the association Wiseman (1982) again found no significant
among two waste disposal variables and 12 finan- association between environmental disclosure
cial indicators representing economic perfor- and environmental performance. Freedman and
mance, Rockness et al. failed to document a Wasley (1990) used Wiseman’s method to evaluate
statistically significant relation. Freedman and environmental disclosures appearing in 10-Ks as
Jaggi (1992) examined the long-term relation well as annual reports. Again they observed no
between environmental performance and eco- significant association between environmental dis-
nomic performance, using the percentage change closure and environmental performance as
in three pollution measures and various account- measured by CEP performance ratings.
ing ratios as empirical proxies for environmental In addition to these archival studies, Rockness
performance and economic performance, respec- (1985) conducted a field experiment in which
tively. Again, Freedman and Jaggi’s results failed financial analysts, members of environmental pro-
to reject the null hypothesis of no significant tection organizations, environmental regulators,
association. Perhaps this observed lack of sig- and MBA students evaluated environmental dis-
nificance resulted from a negative association closures contained in annual reports. Rockness,
between the variables of interest that effectively then compared these subjects’ evaluations of
countered any positive association suggested by environmental disclosures to CEP environmental-
earlier studies. An inverse relation between envir- performance ratings, and reported negative corre-
onmental and economic performance is consistent lation coefficients, implying that subjects evaluate
with traditional economic thought that depicts the worst environmental performance as best, and
this relation as a tradeoff between the firm’s prof- vice versa. As a result, Rockness suggested that
itability and acting on its social responsibility subjects might have been misled by the disclosures.
(Friedman, 1962). In summary, the relation Li, Richardson, and Thornton (1997) arrived at a
between environmental performance and eco- somewhat similar result in providing empirical
nomic performance is founded on contradictory support for their game-theory model of environ-
theoretical support that prior empirical research mental disclosure. Using a sample of Canadian
has failed to clarify. firms, they found a significantly positive relation
S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471 451

between a firm’s decision to disclose and its pro- that strictly comply with SFAS 5. In summary,
pensity to pollute. This implied a negative relation prior research has not found a consistently
between environmental disclosure and environ- significant association between environmental
mental performance. More recently, Hughes, performance and environmental disclosure.6
Anderson, and Golden (2001) observed that
poorer US environmental performers tended to The relation between environmental disclosure and
make the most disclosures, consistent with their economic performance
responsibility to report contingent liabilities under
SFAS 5 (FASB, 1975). However, although these Prior research on the environmental-disclosure–
disclosures differed between groups, Hughes et al. economic-performance relation has used both
did not find them to be useful in classifying the market-based and accounting-based measures of
firms’ actual environmental performance. economic performance. Freedman and Jaggi
A negative environmental-performance–envir- (1982) tested the association of their measure-
onmental-disclosure relation appears to be incon- ments of environmental disclosure against six
sistent with Verrecchia’s (1983) discretionary accounting ratios used to measure economic per-
disclosure model.5 If we assume that good envir- formance. They found insufficient statistical sig-
onmental performance reduces the firm’s exposure nificance to reject the null hypothesis of no
to future environmental costs, then disclosure of association.7 However, Shane and Spicer (1983)
this information should be perceived as good news used an event study design and documented a
by investors. Therefore, firms with good environ- negative market reaction during the two days pre-
mental performance should disclose more envir- ceding the release of CEP environmental reports.
onmental information (in quantity and quality) Similarly, Stevens (1984) reported that a portfolio
than should firms with poorer environmental per- of firms that disclosed higher estimated future
formance. On the other hand, if greater disclosure pollution-abatement costs experienced monthly
provides information that may be used in litiga- returns consistently lower than did a similar port-
tion against the disclosing firm (presumably by folio of firms that disclosed lower estimates of
third parties with political or social agendas), future environmental costs.8
good environmental performers might elect to More recently, Richardson and Welker (2001)
minimize such disclosure (Li et al., 1997). A nega- observed that social disclosure (which subsumes
tive environmental-performance–environmental- environmental disclosure) behaved differently than
disclosure relation is also consistent with increased general financial disclosure in tests of association
disclosure from poor environmental performers with the firm’s cost of capital. These researchers
The lattitude afforded US firms to make environmental
disclosures under SFAS 5 (Accounting for Contingencies) led
researchers to classify such disclosures as largely voluntary
(Barth et al., 1997). Event studies have also established the informational rele-
6 vance of significant environmental events to investors while
We examine environmental disclosure as a function of
environmental performance. Gray, Jarad, Power, and Sinclair documenting their intra-industry effects. Bowen, Castanias, and
(2001) examine environmental disclosure as a function of turn- Daley (1983) examined the electric utility industry after the
over, capital employed, profit, and the number of employees. Three Mile Island nuclear accident; Blacconiere and Patten
For their sample of UK firms during an eight-year sample (1994) examined the chemical industry in the wake of the
period, they were unable to document any ‘‘unique and/or Bhopal chemical leak. Both studies documented a significantly
stable relationship between any measure of disclosure and any negative intra-industry effect. Ironically, Patten and Nance
corporate characteristic’’ (Gray et al., 2001, p. 349). How- (1998) found a positive intra-industry effect following the 1989
ever, their disclosure variable is significantly different from our grounding of the Exxon Valdez, as the spill triggered sub-
disclosure metric. stantial price increases in the wholesale and retail gasoline
7 markets. Additionally, Blacconiere and Patten (1994) observed
Partitioning their sample by firm size into quartiles,
Freedman and Jaggi observed a significantly negative corre- that firms with more extensive environmental disclosures
lation between environmental disclosure and economic indica- experienced less of a negative market reaction to the Bhopal
tors for the top quartile. disaster.
452 S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471

reported a significantly negative relation between H1. Economic performance is not associated with
the level of financial disclosure and the cost of environmental performance.
capital (consistent with prior research; Botosan,
1997), an association primarily driven by firms in H2. Environmental disclosure is not associated
less informationally rich environments. Con- with environmental performance.
versely, Richardson and Welker (2001) found the
relation between social disclosure and cost of H3. Economic performance is not associated with
capital to be significantly positive, with more environmental disclosure.
profitable firms being penalized more for their
social disclosures. Assuming an inverse relation
between cost of capital and share price as sug-
gested by the dividend discount model, Richard-
son et al.’s findings imply that increased social The empirical model and sample selection
disclosure is associated with lower share prices.
This evidence is not consistent with the notion Management’s grand strategy
that discretionary disclosure reduces asymmetrical
information costs or that increased social dis- Ullmann (1985) reviewed prior empirical studies
closure triggers a significantly favorable investor that investigated the interrelations among social
preference effect.9 performance, social disclosure, and economic per-
formance. He noted that the ambiguous results
Hypotheses reported in aggregate by this body of research
might be due to incomplete specification of the
Prior empirical research examining pair-wise empirical models measuring the statistical sig-
associations among environmental performance, nificance of pair-wise associations. Ullmann con-
environmental disclosure, and economic perfor- jectured that the ubiquitous omitted variable was
mance has often been based on contradictory the- management’s grand strategy, and reasoned that
oretical support. Additionally, prior empirical the firm’s social performance, social disclosure,
studies have often drawn on relatively small sam- and economic performance were jointly deter-
ples because of the limited availability of environ- mined by this common missing element. While
mental performance information. All of the recognizing that there have been several environ-
variables of interest are measured with error, mental studies examining one or more of these
especially environmental performance and envir- constructs since 1985 (e.g., Barth et al., 1997;
onmental disclosure. The resultant low-powered Hughes, 2000; Hughes et al., 2001; Richardson
tests have produced mixed results. Given the & Welker, 2001), we believe that the problem
absence of unambiguous theoretical or empirical described in Ullmann’s meta-analysis nevertheless
support for predicting the interrelations among persists. This body of empirical environmental
these three corporate functions, we test the fol- research may have produced mixed results by fail-
lowing nondirectional null hypotheses: ing to recognize the potential for endogenous
relations among these three constructs. We follow
Ullmann in suggesting that management’s overall
strategy affects economic performance, environ-
An example of this effect is that as of 1997, some 13 US mental performance, and environmental dis-
electric utilities had adopted some form of green pricing under closure. While we cannot directly represent a
which the customer is asked to pay a premium of up to 15% of firm’s unobservable strategy, we can accom-
the normal bill. In return, the utility acquires renewable energy
modate the joint determination process by esti-
sources according to a set formula. Additionally, surveys con-
sistently reveal that from 56 to 80% of respondents are willing mating the relations among these constructs using
to pay more for environmentally friendly energy sources a system of simultaneous equations defined in the
(Tietenberg, 1998). following structural form:
S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471 453

Economic Performance used measure is the Council on Economic Prio-

rities’ (CEP) company rating charts, which are
¼ f Environmental Performance and ð1:1Þ
 published quarterly. Corporate environmental
predetermined variables performance ranking provides insight into thirteen
environmental issues that affect US corporations.
Although prior researchers have used these ratings
Environmental Performance to measure corporate environmental performance,
¼ f Economic Performance and ð1:2Þ we elect not to use this measure, for the following
 reasons: (1) the unavailability of the CEP’s for-
predetermined variables
mulae to determine their environmental-perfor-
mance rating limits its interpretation; (2) weights
Environmental Disclosure assigned to different environmental factors are not
constant across industries, and we use an inter-
¼ f Environmental Performance and ð1:3Þ industry, cross-sectional sample; and (3) environ-

predetermined variables mental reporting (disclosure) is one of the factors
used to determine the CEP rating of a corporation’s
The empirical model environmental performance.
In contrast to studies that rely on qualitative
Having defined the theoretical model, we pro- rankings to measure environmental performance,
pose the following structural equations as an this study employs a quantitative measure: the
empirical model to test the study’s hypotheses: ratio of toxic waste recycled to total toxic waste
generated (ENVPERF). For example, if a firm
ECONPERF ¼ 0 þ 1 ENVPERF þ 2 UE introduces a pollution-abatement process,
þ 3 PREDISC þ 4 GROWTH þ 5 MARGIN decreasing the total amount of toxic waste gener-
ated, the denominator decreases and thus
þ 6 ENVEXP þ 1
ENVPERF increases. Or if the firm adopts pro-
ð2:1Þ cesses that recycle toxic waste (such as closed-loop
cooling systems), the numerator increases, again
increasing ENVPERF. Therefore as a rule, the
ENVPERF ¼ 0 þ 1 ECONPERF higher the ratio of recycled waste to total waste,
þ 2 PREDISC þ 3 GROWTH þ 4 ENVEXP the better the firm’s environmental performance.
þ 5 ENVCON þ 6 VISIBILITY þ 2 This environmental performance measure also
incorporates the first three principles of good
environmental performance as promulgated by the
Coalition for Environmentally Responsible
ENVDISCL ¼ 0 þ 1 ENVPERF þ 2 ENVEXP Economies (CERES): minimize pollutants, con-
serve resources, and reduce waste. Also, this
þ 3 ENVCON þ 4 SIZE þ 3 particular measure is sufficiently generic to be used
ð2:3Þ by virtually all polluting industries, avoiding the
problem of relying on industry-specific pollution
We define the variables included in these models measures in our cross-sectional, inter-industry,
below. research design. We obtain recycling ratio data
from the Corporate Environmental Profiles
Endogenous variables Directory, which is published annually (since

Environmental performance (ENVPERF) 10

See Ilinitch, Soderstrom, and Thomas (1998) for a discus-
Prior research has measured environmental per- sion of current problems in measuring and reporting environ-
formance in several ways.10 The most frequently mental performance information.
454 S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471

1992) by the Investor Responsibility Research (PRP) designation for the cleanup responsibility of
Center (IRRC).11 hazardous-waste sites; and (4) the occurrence of
reported oil and chemical spills. Thus our envir-
Environmental disclosure (ENVDISCL) onmental disclosure measure is based on informa-
Environmental disclosure measurement techni- tion reported in SEC Forms 10-K and focuses on
ques can be classified into two general groups. The pollution-related information in these four areas.
first group includes measures that quantify the This informational restriction allows a structured
level of environmental disclosure in the annual evaluation while capturing significant indicators of
report, such as the number of pages (Gray, the firm’s environmental exposure to future envir-
Kouhy, & Lavers, 1995; Guthrie & Parker, 1989; onmental costs.
Patten, 1992, 1995), sentences (Frazier, 1982; We believe that quantitative disclosures are
Ingram & Wiseman, 1980), and words (Deegan & more objective and informative to stakeholders
Gordon, 1996; Zeghal & Ahmed, 1990). Each of than qualitative information. Prior studies (e.g.,
these measures has its limitations. While pages Hughes et al., 2001) using quantitative disclosure
may include pictures that have no information on measures have assigned weights to different dis-
environmental or social activities, sentences and closure items based on the perceived importance
words may ignore necessary graphs and tables. of each item to various user groups. Similarly, we
This form of measurement is also susceptible to assign the greatest weight (+3) to quantitative
‘‘greenwashing,’’ in which management puts its disclosures related to the four environmental indi-
best ‘‘spin’’ on what otherwise might be a lack- cators described above. We assign the next highest
luster environmental performance. weight (+2) to non-quantitative but specific
The second measurement technique uses a dis- information related to these indicators. Finally,
closure-scoring measure derived from content general qualitative disclosures receive the lowest
analysis. Using this technique, researchers first weight (+1). Firms that do not disclose informa-
identify certain environmental issues, then analyze tion for a given indicator receive a score of zero
the environmental disclosure of each issue using a for that indicator.
‘‘yes/no’’ (or 1, 0) scoring methodology. After In constructing our disclosure measure, we rea-
individual issues are quantified, researchers lize that firms using production processes that are
determine the aggregate score for each firm.12 not associated with one of our polluting activities
We adopt a similar disclosure-scoring method- should not be penalized for failing to make envir-
ology based on content analysis that incorporates onmental disclosures for that activity. We accom-
disclosures of four key environmental indicators: modate this adjustment by factoring the firm’s
(1) the total amount of toxic waste generated and participation in a polluting process (as reported to
transferred or recycled; (2) financial penalties regulators rather than to shareholders) into the
resulting from violations of 10 federal environ- denominator of our disclosure score. These data
mental laws; (3) Potential Responsible Party are obtained from the IRRC database (i.e., the
firm’s pollution data reported to the EPA or some
other agency). For example, if a firm is designated
a PRP but does not disclose this information in its
The IRRC publishes various types of information on annual report, the numerator of the disclosure
S&P 500 firms that are deemed to be useful to institutional
investors, corporations, and other users. The IRRC uses three score would not be affected (0), while the denomi-
major environmental information sources: government agen- nator of the score would be increased (+1),
cies, corporations, and the media. ‘‘Much of this information thereby decreasing the firm’s overall disclosure
was obtained by filing Freedom of Information Act requests, score. This metric is the first environmental-dis-
often after extensive consultation with the numerous agencies
closure measure, to our knowledge, that attempts
involved’’ (Corporate Environmental Profiles Directory, 1995,
p. 3). to capture the ‘‘truth-telling’’ or transparency prop-
For example, Barth et al. (1997) used a variation of this erty of the firm’s environmental disclosure by mea-
scoring technique. suring the firm’s disclosure to investors conditioned
S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471 455

Fig. 1. Example of disclosure scoring technique—ENVDISCL=quality score/occurrence score. Examples of scored disclosures:
Qualitative: In addition, the Company has been identified as a potentially responsible party for investigation and cleanup costs at a
number of locations in the United States and Puerto Rico under federal remediation laws and is voluntarily investigating potential
contamination at a number of Company-owned locations (Abbot Laboratories, 1993). Qualitative specific: The Company had also
been named in several groundwater contamination suits, nearly all related to its former agricultural pesticide, ethylene dibromide
(EDB), which was banned by the US government in 1983. These cases, primarily in Florida, California, and the State of Washington, stem
from uses of EDB approved by all governmental regulatory agencies prior to its ban. All suits remaining in Florida and the State of
Washington were settled by 1991 in amounts not material to the Company’s financial results. The only significant remaining suit is in
California, where the Company is one of many defendants named. The Company is actively defending itself in this suit and does not now
believe that this case will result in material financial consequences for the Company (Great Lakes Chemical, 1991). Quantitative: Alcoa’s
remediation reserve balance at the end of 1994 was $329 million and reflects the most probable costs to remediate identified environmental
conditions for which costs can be reasonably estimated. About 28% of this balance relates to Alcoa’s Massena, New York plant site.
Remediation costs charged to the reserve were $79 million in 1994, $71 million in 1993, and $102 million in 1992. They include expendi-
tures currently mandated, as well as those not required by any regulatory authority or third parties (Aluminum Co. of America, 1994).

on its polluting activity reported to environmental using various economic performance metrics is
regulators.13 The total quality score (mini- that they tend to focus narrowly on one aspect of
mum=0, maximum=+12) is summed for the a firms’ economic performance. Net income mea-
four activities and then scaled by the total number sures a firm’s profitability without considering
of polluting activities associated with the firm’s firm size. This limitation can be addressed by
production processes, the occurrence score (mini- using measures such as return on assets (ROA),
mum=0, maximum =+4). Thus environmental- and scaling profitability by the firm’s investment
disclosure scores (ENVDISCL) range from 0 to in their asset base. However, this measure may be
+3. Fig. 1 provides an illustrative example of an biased if the sample (such as this study’s sample)
environmental-disclosure score calculation. includes firms from different industries with dif-
ferent industry-driven levels of fixed assets, and
Economic performance (ECONPERF) where there may be systematic differences across
Prior environmental studies have used both industries in the age of these assets. Given these
accounting-based and market-based measures to limitations, and assuming (a semistrong form of)
represent economic performance. For example,
Bragdon and Marlin (1972) used accounting- 13
Measuring environmental disclosure conditioned on the
based measures (earnings per share and return on
firm’s polluting activities as reported to the EPA is consistent
equity), while Spicer (1978) used both accounting- with the SEC and EPA’s consideration of formalizing their
based and market-based measures (profitability currently informal exchange of information programs (Slavich,
and the price-earnings ratio). One limitation in 1994).
456 S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471

market efficiency, we elect to use a market-based mental performance are conditioned on informa-
metric to represent economic performance. tion provided by prior environmental disclosures.
We measure the firm’s economic performance If a firm’s current environmental performance
using an industry-adjusted annual return. We cal- deteriorates (without disclosure) and reduces stock
culate this metric as the change in stock price price, then shareholders may have grounds for
during the year (adjusted for dividends), scaled by litigation. Also, management’s reputation for pro-
the beginning-of-year stock price minus the viding credible disclosures would suffer. We aver-
industry median return (based on two-digit SIC age ENVDISCL (the environmental-disclosure
codes). While other market-derived measures, score) over the three most recent years to proxy
such as debt ratings and the cost of capital, may for past environmental disclosure (PREDISC),
appear to be viable candidates for representing and we predict a positive relation between PRE-
economic performance, they are also associated DISC and ENVPERF.
with increased subjectivity and measurement We also use PREDISC rather than ENVDISCL
error. We believe that annual stock returns repre- in examining the relation between environmental
sent a more objective and comprehensive measure disclosure and economic performance. This is
of economic performance. The latter is due to the necessary because our proxy for economic perfor-
proposition that stock price should impound infor- mance—annual industry-adjusted return—is mea-
mation about the firm’s future prospects from a vast sured at the end of the fiscal year, while the
array of both financial and nonfinancial measures, ENVDISCL variable is derived from Forms 10-K
such as net income, ROA, operational data, etc. The financial disclosures that are released approxi-
annual industry-adjusted stock return, ECON- mately three months later. Therefore it is not possi-
PERF, represents a comprehensive measure of the ble to observe a contemporaneous relation between
firm’s current-period economic performance relative these variables. PREDISC approximates the aver-
to other firms in the same industry.14 We comple- age level of disclosure in prior years (ENVDISCL).
ment this ‘‘changes’’ (return) specification of
ECONPERF in subsequent sensitivity testing in Growth opportunities (GROWTH)
which a ‘‘levels’’ (share price) specification is used to We use the ratio of market value of equity to
represent economic performance. book value of equity as a proxy for future growth
opportunities (Gaver & Gaver, 1993; Smith &
Predetermined variables Watts, 1992) in the economic-performance equa-
tion. This ratio measures the difference between
Unexpected earnings (UE) the market’s appraisal of firm value and the esti-
Given our industry-adjusted-return measure of mate of value aggregated from GAAP-mandated
economic performance, we control for the unex- accounting transactions. For example, US firms
pected portion of earnings as the annual change in conducting research and development (R&D) must
earnings per share scaled by stock price at the expense this cost, whereas the market views (at least
beginning of the period (Christie, 1987). The some) R&D costs as an investment yielding future
association between returns and earnings is well benefits. Growth opportunities should be positively
documented in the earnings-response-coefficient related to economic performance (ECONPERF).
(ERC) literature (e.g., Collins & Kothari, 1989). We also include GROWTH in the environmental-
performance equation as a proxy for intangible
Predisclosure environment (PREDISC) assets associated with innovation (Porter & van der
A firm’s prior environmental disclosures may Linde, 1995a). We expect GROWTH to be posi-
represent a lower bound for current environmental tively related to environmental performance.
performance. Investors’ expectations of environ-
Profit margin (MARGIN)
Using industry-adjusted returns also permits us to hold the Profit margin, the ratio of net income to net
firm’s investment opportunity set (IOS) constant. sales, captures both profitability and the presence
S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471 457

of competitive markets. Firms may obtain a positively correlated with future environmental
higher profit margin by increasing sales prices, costs.
decreasing costs, or doing both. In competitive We expect ENVEXP to be negatively related to
markets, because firms have limited potential for economic performance, since future environ-
increasing prices, cost control becomes of primary mental costs should reduce future cash flows and
importance. We assume that the relatively large hence firm value. We posit that environmental
firms that comprise our sample operate in globally exposure is positively related to both environ-
competitive markets. Therefore, higher profit mental performance and environmental dis-
margins signal better cost control, and this indic- closure. Firms with greater exposures to future
tor should be positively associated with economic costs have greater incentives to perform well
performance (ECONPERF).15 environmentally in order to assure stakeholders
that their investments will not bear undue risk.
Environmental exposure (ENVEXP) Similarly, firms with greater ENVEXP are likely
Environmental exposure is defined as the firm’s to disclose more environmental information, both
exposure to future environmental costs. Since we to fulfill the disclosure requirements of SFAS 5
predict that a firm’s environmental exposure will and to reduce information asymmetry that
affect both environmental performance and envir- adversely affects their cost of capital (Richardson
onmental disclosure, we use it as a control variable & Welker, 2001).
in the environmental performance and environ-
mental disclosure equations. In addition, environ- Environmental concern (ENVCON)
mental performance and environmental disclosure Including the firm’s environmental concern as a
may affect economic performance if a firm has a predetermined variable in the environmental per-
significant level of exposure to environmental formance and environmental disclosure equations
costs. Consequently we also include our environ- is consistent with Ullmann’s (1985) conceptual
mental exposure measure as a control variable in emphasis on including management’s strategy in
the economic performance equation. models examining the firm’s social responsibility.
Because the firm’s production processes deter- We operationalize this variable by using a factor
mine environmental pollution, several researchers analysis of three firm characteristics that capture
have employed industry classification (SIC code) its concern for the environment, thereby extracting
as a proxy for environmental exposure (i.e., a single composite measure representing the firm’s
Brockhoff, 1979; Deegan & Gordon, 1996; Little, environmental concern (ENVCON).17 The first
Muoghlu, & Robison, 1995; Niskala & Pretes, firm characteristic we consider, COMMITTEE, is
1995; Wiseman, 1982). Given that SIC codes may an indicator variable that is coded one (1) if the
inadequately describe a firm’s unique portfolio of firm has a corporate environmental committee or
operations, we again use a quantitative measure. department charged with monitoring the environ-
We measure environmental exposure (ENVEXP) mental impact of the firm’s actions; otherwise it is
as the amount of toxic waste generated by the firm coded as zero (0). The second characteristic,
scaled by total revenues.16 The amount of toxic REPORT, measures the frequency with which the
waste generated per dollar of sales provides an firm publishes a stand-alone environmental report
indication of the pollution intensity of the firm’s (separate from the annual report). REPORT is
production processes, a measure we propose to be coded 1.0 if the report is published annually, 0.5 if

However, because environmental issues drive some costs,
we further isolate the profit-margin metric to capture the effi- Using factor analysis to construct the ENVCON variable
ciency of the firm’s cost control efforts, given a specific level of is consistent with the concept of a latent variable that cannot be
environmental exposure. To do this, we condition profit margin adequately described by a single metric. Also, this methodology
on environmental exposure (ENVEXP). reduces the potential for encountering multicollinearity
16 problems among ENVCON’s correlated component variables.
We also use SIC codes in the sensitivity analysis.
458 S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471

the report is published biennially, 0.33 if the report 1985) has found that firm size proxies for the
is published triennially, and zero if no report is firm’s information environment, where informa-
published. The third measure, PROGRAM, is a tionally rich environments are associated with lar-
discrete variable that measures the number of ger firms. Although we expect that larger firms
EPA voluntary programs in which the firm parti- have greater incentives to disclose environmental
cipates.18 Because management’s decision to information, our restrictive measure for environ-
engage in these three activities is completely mental disclosure (specific pollution information)
voluntary, these variables provide evidence of may distort the measurement of this relation.
management’s awareness and concern for the Therefore, we do not predict the sign of the
environment. We predict that ENVCON, the coefficient on the SIZE variable.
primary factor derived from factor analysis of
these variables, is positively related to both Sample selection and data collection
environmental performance and environmental
disclosure.19 This study uses a cross-sectional research design
and firm data for the year 1994. In order to be
Public visibility (VISIBLTY) included in our sample, a firm must
Firms exposed to greater public scrutiny are
more likely to incur political costs associated with  be listed in the IRRC’s 1994 Environmental
poor environmental performance. Consequently, Profiles Directory, which is limited to
we control for a firm’s public visibility by includ- Standard and Poor’s 500 companies;
ing the number of Wall Street Journal news  generate at least one pound of toxic waste
announcements about the firm during the year per $10,000 of revenue (in order to ensure
(VISIBLTY) in the environmental-performance some minimum level of environmental
equation. We expect higher-visibility firms to have exposure);
higher standards of environmental performance  have complete financial data reported in
because of increased public scrutiny. Compustat;
 have its annual reports accessable using the
Firm size (SIZE) LexisNexis database; and
We use market value of common equity (SIZE)  appear in the Wall Street Journal Index.
as a control variable for firm size in the environ-
mental-disclosure equation. Prior research (Atiase, Of the 531 firms included in the 1994 IRRC
Environmental Profiles Directory, 313 do not have
One example is the EPA’s ‘‘Green Lights’’ program, in sufficient environmental exposure to meet our sec-
which firms voluntarily scrutinize every path of electrical ond criterion. Four firms do not have complete
energy consumption. In return they receive EPA guidance on
data in the IRRC Directory, and 16 firms do not
efficient lighting, heating, and cooling operations (Porter & van
der Linde, 1995a). have complete Compustat data. The final sample
The eigenvalue for the first principle component (2.0351) is includes 198 firms that meet all of the selection
more than three times larger than that of the second principal criteria.
component (0.5256) and more than four times larger than that
of the third principal component (0.4393). Furthermore, the
first principal component explains 67.8% of the variation of the
three environmental concern proxies. Finally, the communality Results
estimates for the three environmental concern variables are
REPORT 0.70, PROGRAM 0.69, COMMITTEE 0.64 and the Descriptive statistics
correlations between the first principle component and the
three environmental concern variables are as follows: REPORT
Panel A of Table 1 provides descriptive statistics
0.83, PROGRAM 0.83, COMMITTEE 0.80. Therefore, we
conclude that the first principle component captures the core for the study’s dependent variables. The mean
construct common to these variables—management’s environ- (median) industry-adjusted stock return, ECON-
mental concern. PERF, for sample firms is 0.0076 (0.0064). The
S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471 459

Table 1
Descriptive statistics—cross-sectional data for 198 environmentally exposed firms

Variables Mean Standard deviation 25th Percentile Median 75th Percentile

Panel A: Dependent variables

ECONPERF 0.0076 0.3227 0.1166 0.0064 0.1091
ENVPERF 0.74 0.23 0.62 0.83 0.93
ENVDISCL 0.67 0.66 0.00 0.50 1.00

Panel B: Predetermined variables

PREDISC 0.47 0.54 0.00 0.17 0.67
ENVEXP 9.95 24.61 0.53 2.06 9.24
ENVCON 0.01 1.00 1.19 0.02 0.86
REPORT 0.26 0.41 0.00 0.00 0.50
PROGRAM 1.20 1.12 0.00 1.00 2.00
COMMITTEE 0.62 0.49 0.00 1.00 1.00
UE 0.0112 0.0461 0.0022 0.0092 0.0220
GROWTH 2.66 1.69 1.69 2.25 3.08
MARGIN 0.07 0.19 0.003 0.02 0.09
VISIBLTY 26.58 34.27 11.00 16.00 27.00
SIZE $8,534.31 $13,568.52 $1,781.56 $3,374.25 $7,842.94

ECONPERF=Industry-adjusted annual stock return. ENVPERF=Environmental performance measured as the percentage of total
waste generated that is recycled. ENVDISCL=Environmental disclosure score obtained from content analysis of the firm’s annual
report. PREDISC=Past environmental disclosure measured as the average ENVDISCL over the preceding three years. ENVEXP
=Environmental exposure measured as toxic waste generated scaled by total revenues. ENVCON=Environmental concern mea-
sured as the primary factor obtained from factor analysis of REPORT, PROGRAM, AND COMMITTEE. REPORT=Indicator
variable coded one if the firm publishes an annual environmental report separate from its annual report, 0.5 if the report is biennial,
0.33 if the report is triennial, and 0 if no report is published. PROGRAM=The number of voluntary EPA programs in which the firm
participates. COMMITTEE=Indicator variable coded one if the firm has an environmental committee and zero otherwise. UE=
Unexpected earnings measured as the annual change in earnings per share scaled by beginning of year stock price. MARGIN=
Profit margin (net income/net sales). GROWTH=Market-to-book ratio of common equity. VISIBLTY=The number of Wall Street
Journal news announcements about the firm. SIZE=Market value of common equity (in thousands of US dollars).

mean (median) percentage of waste recycled, Accounting Bulletin No. 92 in June 1993, and is
ENVPERF, 74% (83%) implies that sample firms also consistent with the findings of Barth et al.
recycle most of their generated waste. Finally, the (1997).21 Environmental exposure, ENVEXP, is
mean (median) disclosure score, ENVDISCL, 0.67 positively skewed as the mean, 9.95, and is higher
(0.50) suggests that on a scale of zero to three than the 75th percentile, 9.24. This suggests that,
(where three represents quantitative disclosures of on average, sample firms generate approximately
all significant environmental activities), sample ten pounds of waste for every one thousand dol-
firms, on average, disclose only qualitative infor- lars of sales generated. Because this variable is
mation at best.
Panel B of Table 1 reports descriptive statistics 20
Supplemental analysis reveals that, on average, the current
for the predetermined variables used in our system
level of environmental disclosure is 55% higher than the aver-
of equations. The environmental-disclosure score age of the past 3 years.
that represents the predisclosure environment 21
The SEC signaled a continuing interest in improving the
averaged over the three prior years, PREDISC, accounting for and disclosure of environmental liabilities with
(0.47) is significantly less (p < 0.01) than the cur- the 1993 release of Staff Accounting Bulletin No. 92 (SAB 92).
This bulletin (1) limited offsetting against insurance consistent
rent year’s (1994) environmental-disclosure score,
with GAAP, (2) limited the discounting of future environ-
ENVDISCL, (0.67).20 This result suggests that the mental liabilities while imposing a ceiling on the allowable dis-
higher 1994 environmental-disclosure level may count rate, and (3) attempted to elicit more meaningful
have been triggered by the SEC’s release of Staff information concerning environmental matters.
460 S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471

highly skewed, we use a log-transformation of does not present an econometric problem. Simi-
ENVEXP in the empirical model. larly, the proxies for SIZE (market value of
The firm’s environmental concern, ENVCON, is equity) and VISIBILITY (number of WSJ news
the primary factor from a factor analysis of three announcements) are highly correlated but again
different measures of management’s environ- do not appear in the same structural equation.
mental concern. Examining these three measures, Positive correlations of medium strength are noted
we first find that the frequency of separate envir- between environmental performance (ENVPERF)
onmental reports, REPORT, (0.26) implies that, and environmental exposure (ENVEXP), MAR-
on average, sample firms publish a distinct envir- GIN and SIZE, and between MARGIN and
onmental report approximately every fourth year. GROWTH. Other pair-wise relations depicted in
The mean PROGRAM shows that the average Table 2 are relatively weak. None of the depicted
sample firm voluntarily subscribes to 1.20 envir- relations represent obvious anomalies to theoretical
onmental EPA programs. Finally, the mean and explanation.
median COMMITTEE (0.62 and 1.00, respec-
tively) suggest that more than half of the sample The endogeneity problem
firms have environmental committees charged
with monitoring their environmental responsi- We posit that the mixed results of prior research
bilities. may be attributable to the fact that managers’
Panel B of Table 1 also reports that the financial overall strategies likely affect economic perfor-
determinants of a firm’s economic performance— mance, environmental performance, and environ-
unexpected earnings (UE), the market-to-book mental disclosure simultaneously. As depicted in
ratio (GROWTH), and profit margin (MAR- Table 2, our proxies for these constructs are all
GIN)—are all slightly positively skewed, which is positively correlated. We then examine the appro-
consistent with prior research. The variables priateness of using an OLS regression analysis to
representing public visibility (VISIBLTY) and estimate this system of equations by employing a
firm size (SIZE) are also positively skewed, sug- Hausman (1978) test. This statistical test should
gesting that many of the sample’s S&P 500 firms detect the presence of any endogenous relations
with significant environmental exposure are among our three dependent variables—ECON-
among the largest and most profitable public PERF, ENVPERF, and ENVDISCL.22 Using this
firms. Because VISIBLTY and SIZE are so highly procedure, we reject the null hypothesis of no
skewed, we again employ log-transformations for endogeneity with respect to ENVPERF in the
these two variables in the empirical model. third equation (t=2.910, p< 0.004). We therefore
Table 2 depicts both the parametric and non- conclude that OLS estimators are potentially
parametric, pair-wise correlation coefficients for
all variables included in the system of equations. 22
This particular Hausman test involves a two-stage proce-
The three variables of interest (ECONPERF, dure. In the first stage, each dependent variable is regressed on
ENVPERF, and ENVDISCL) are all positively all of the predetermined variables in the system, and predicted
correlated, but the strengths of these pair-wise values for the dependent variable are calculated using the
linear relations are relatively weak (coefficients estimated coefficients from the first stage regressions. In the
second stage, each dependent variable is regressed on the right-
< 0.4). This observation is consistent with the
hand-side dependent variables, the predicted values of the
mixed results experienced by prior researchers in right-hand-side dependent variables, and the respective pre-
evaluating these pair-wise relations. Of note, the determined variables for that equation. The significance of each
relation between the two environmental disclosure predicted right-hand-side dependent variable is then tested
variables, ENVDISCL and PREDISC approach against zero using a T-test or an F-test with the null hypothesis
of no endogeneity with respect to that variable. If a predicted
perfect correlation, consistent with the proposition
dependent variable is determined to have significant explana-
that a firm’s disclosure policy is relatively tory power, the dependent variable is presumed to be endo-
‘‘sticky.’’ As these variables do not appear in the genous. See Kennedy (1994, p. 169) for an intuitive explanation
same structural equation, this high correlation of this procedure.
Table 2
Correlation coefficients (P-values) for all variables included in the simultaneous equation model (Pearson coefficients in upper right and Spearman coefficients in lower


S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471

ECONPERF 1.0000 0.3166 0.2196 0.1868 0.2401 0.2392 0.4550 0.3658 0.4132 0.2290 0.2789
(0.0001) (0.0010) (0.0052) (0.0003) (0.0003) (0.0001) (0.0001) (0.0001) (0.0006) (0.0001)
ENVPERF 0.1850 1.0000 0.3418 0.2965 0.5402 0.3897 0.1983 0.1571 0.3874 0.3476 0.4259
(0.0057) (0.0001) (0.0001) (0.0001) (0.0001) (0.0030) (0.0192) (0.0001) (0.0001) (0.0001)
ENVDISCL 0.1771 0.3264 1.0000 0.9861 0.3909 0.4589 0.1915 0.0128 0.0853 0.2209 0.2483
(0.0082) (0.0001) (0.0001) (0.0001) (0.0001) (0.0042) (0.8494) (0.2054) (0.0009) (0.0002)
PREDISC 0.1771 0.3264 1.0000 1.0000 0.3444 0.4285 0.1649 0.0003 0.0720 0.1794 0.2098
(0.0082) (0.0001) (0.0001) (0.0001) (0.0001) (0.0139) (0.9970) (0.2857) (0.0074) (0.0017)
ENVEXP 0.1879 0.4990 0.4399 0.4399 1.0000 0.2525 0.3081 0.0678 0.2941 0.0577 0.2261
(0.0050) (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) (0.3146) (0.0001) (0.3926) (0.0007)
ENVCON 0.2283 0.3870 0.5063 0.5063 0.3276 1.0000 0.1384 0.0381 0.1494 0.5256 0.5263
(0.0006) (0.0001) (0.0001) (0.0001) (0.0001) (0.0394) (0.5720) (0.0260) (0.0001) (0.0001)
UE 0.2592 0.0831 0.1606 0.1606 0.2833 0.1278 1.0000 0.04476 0.2139 0.0616 0.0491
(0.0001) (0.2175) (0.0166) (0.0166) (0.0001) (0.0573) (0.4801) (0.0013) (0.3607) (0.4665)
GROWTH 0.3426 0.2028 0.0863 0.0863 0.1734 0.2033 0.0450 1.0000 0.5535 0.2860 0.4735
(0.0001) (0.0024) (0.2004) (0.2004) (0.0096) (0.0023) (0.5051) (0.0001) (0.0001) (0.0001)
MARGIN 0.2502 0.1824 0.0152 0.0152 0.1827 0.1892 0.0329 0.6243 1.0000 0.4079 0.5570
(0.0002) (0.0064) (0.8218) (0.8218) (0.0063) (0.0047) (0.6259) (0.0001) (0.0001) (0.0001)
VISIBLTY 0.1944 0.2738 0.2993 0.2993 0.1252 0.5935 0.0117 0.3491 0.3090 1.0000 0.8258
(0.0036) (0.0001) (0.0001) (0.0001) (0.0627) (0.0001) (0.8622) (0.0001) (0.0001) (0.0001)
SIZE 0.2184 0.3340 0.2794 0.2794 0.2351 0.5655 0.0349 0.5434 0.5173 0.8039 1.0000
(0.0011) (0.0001) (0.0001) (0.0001) (0.0004) (0.0001) (0.6051) (0.0001) (0.0001) (0.0001)

ECONPERF=Industry-adjusted annual stock return. ENVPERF=Environmental performance measured as the percentage of total waste generated that is recycled.
ENVDISCL=Environmental disclosure score obtained from content analysis of the firm’s annual report. PREDISC=Past environmental disclosure measured as the
average ENVDISCL over the preceding three years. ENVEXP=Environmental exposure measured as toxic waste generated scaled by total revenues. ENVCON
=Environmental concern measured as the primary factor obtained from factor analysis of REPORT, PROGRAM, AND COMMITTEE REPORT=Indicator
variable coded one if the firm publishes an annual environmental report separate from its annual report, 0.5 if the report is biennial, 0.33 if the report is triennial and
0 if no report is published. PROGRAM=The number of voluntary EPA programs in which the firm participates. COMMITTEE=Indicator variable coded one if the
firm has an environmental committee and zero otherwise. MARGIN=Profit margin (net income/net sales). GROWTH=Market-to-book ratio of common equity.
VISIBLTY=The number of Wall Street Journal news announcements about the firm.
SIZE=Market value of common equity
(in thousands of US dollars).

462 S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471

biased and inconsistent. After ensuring that each nificant determinant of environmental perfor-
structural equation passes a White (1980) test for mance. This result is not consistent with the
homogeneity and correct specification,23 we con- economic argument that profitability drives good
tinue our analysis using two-stage least squares environmental performance and that environ-
(2SLS) and three-stage least squares (3SLS) mental accountability is strictly a matter of
simultaneous equation models in order to control affordability. Furthermore, the firm’s predis-
for endogeneity while obtaining asymptotically closure environment (PREDISC) is positively
unbiased results. Because the results from the related to ENVPERF, consistent with the level of
2SLS and 3SLS models are largely similar, and disclosure in the prior periods serving as a lower
because 3SLS parameter estimates are the more bound for performance in the current period.24 In
efficient of the two, we report only 3SLS results. addition, the coefficients for both ENVEXP and
Our examination of this system using the 3SLS VISIBLTY are significantly positive, as predicted.
model incorporates all the available information This implies that firms with greater environmental
from all the equations in simultaneously estimat- exposure and greater public visibility respond with
ing the parameters. higher environmental performance standards than
other comparable firms. Coefficients for variables
Regression analysis (three-stage least squares) representing GROWTH and environmental con-
cern, ENVCON, are statistically insignificant.
The results of our 3SLS simultaneous equation The 3SLS results for Eq. (2.3) suggest that
model examining the relations among economic ENVPERF is significantly and positively asso-
performance, environmental performance, and ciated with ENVDISCL. This is consistent with
environmental disclosure, as specified in structural discretionary disclosure theory’s explanation that
Eqs. (2.1)–(2.3), appear in Table 3. The coefficients good environmental performers believe that dis-
for the explanatory variables in Eq. (2.1), which closure of their performance represents ‘‘good
specifies the determinants of economic performance, news’’ to market participants. These firms can
suggest a positive relation between economic per- therefore be more forthright in disclosing indica-
formance, ECONPERF, and environmental per- tors of environmental pollution. The positive
formance, ENVPERF. This positive relation relation between environmental performance and
(P=0.0739) is consistent with Porter and van der environmental disclosure is counter to that sug-
Linde’s (1995a) win–win scenario and the propo- gested by Li et al.’s (1997) gaming model. Addi-
sition that good environmental performance is tionally, this result is inconsistent with poor
rewarded in the market. This result is also con- environmental performers that have potentially
sistent with investors who view good environ- greater contingent environmental liabilities, dis-
mental performance as an intangible asset. In closing these contingent liabilities as required by
further examining the results of Eq. (2.1), we find SFAS 5. This implies that poor performers either
that both UE and GROWTH are significantly do not strictly follow the requirements of SFAS 5,
positively associated with economic performance, or that SFAS 5’s requirements provide sufficient
as predicted. We also find some evidence that latitude (e.g., materiality judgments) for poor
MARGIN is positively associated with ECON-
PERF, and that our proxy for environmental Because of the high correlation between PREDISC and
ENVDISCL as depicted in Table 2, an alternative explanation
exposure (ENVEXP) is insignificant. for this result is that the predisclosure environment variable is a
The results presented in Table 3 for Eq. (2.2) proxy for contemporaneous environmental disclosure. If we
suggest that economic performance is not a sig- substitute ENVDISCL for PREDISC in Eq. (2.2), we observe
similar results to those reported in Table 3. We acknowledge
White’s (1980) test fails to reject the null hypothesis of the research design’s inability to disentangle these variables
homoskedasticity and correct model specification for all equa- with regard to Eq. (2.2). Having said this, we elect to retain
tions at the 0.05 level. Furthermore, variance-inflation factors PREDISC, rather than ENVDISCL, as a determinant of
and collinearity diagnostics (Belsley, Kuh, & Welsch, 1980) are environmental performance because we believe the theoretical
well within acceptable ranges. support for PREDISC’s inclusion to be more persuasive.
S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471 463

Table 3
Three-stage least squares regression results (economic performance=industry-adjusted returns) coefficients (t-statistics)

Independent variable Predicted sign Dependent variable

ECONPERF [Eq. (2.1)] ENVPERF [Eq. (2.2)] ENVDISCL [Eq. (2.3)]

INTERCEPT 0.44b 0.37b 0.32

(3.24) (6.05) (0.81)
ENVPERF 0.40a 2.70b
(1.80) (3.01)
UE + 2.42b
PREDISC + 0.01 0.26b
(0.22) (10.44)
GROWTH + 0.05b 0.01
(3.47) (0.78)
MARGIN + 0.01a
ENVEXP /+/+ 0.02 0.06b 0.09
(1.24) (6.27) (1.31)
ENVCON + 0.01 0.10a
(0.35) (1.39)
VISIBLTY + 0.06b
SIZE 0.11a
System weighted R2 0.5483

ECONPERF=Industry-adjusted annual stock return. ENVPERF=Environmental performance measured as the percentage of total
waste generated that is recycled. ENVDISCL=Environmental disclosure score obtained from content analysis of the firm’s annual
report. UE=Annual change in earnings per share scaled by stock price at fiscal year end. PREDISC=Past environmental disclosure
measured as the average ENVDISCL over the preceding three years. GROWTH=Market-to-book ratio of common equity.
ENVEXP=Environmental exposure measured as the natural log of toxic waste generated scaled by total revenues. ENVCON
=Environmental concern measured as the primary factor obtained from factor analysis of REPORT, PROGRAM, and COM-
MITTEE. REPORT=Indicator Variable coded one if the firm publishes an annual environmental report separate from its annual
report, 0.5 if the report is biennial, 0.33 if the report is triennial, and 0 if no report is published. PROGRAM=The number of
voluntary EPA programs in which the firm participates. COMMITTEE=Indicator variable coded one if the firm has an environ-
mental committee and zero otherwise. MARGIN=Profit margin (net income / net sales) conditioned on environmental exposure.
VISIBLTY=The natural log of the number of Wall Street Journal news announcements about the firm. SIZE=The natural log of
market value of common equity.
Statistically significant at the 0.10 level (one-tailed test if the sign is predicted, otherwise two-tailed).
Statistically significant at the 0.01 level (one-tailed test if the sign is predicted, otherwise two-tailed).

performers to justify nondisclosure. On the not-so-good. We also find marginal evidence that
surface, it appears inconsistent that good envir- management’s level of environmental concern,
onmental performers would disclose more ENVCON, is positively related to environmental
quantifiable environmental information regarding disclosure. This suggests that managers who are
their polluting activities, which the market would concerned about the environment are more likely
then use to punish them by discounting their stock to disclose their environmental activities publicly.
prices. However, such disclosure may be viewed as In sum, the 3SLS results suggest two significant
a signal that the firm is in fact a good environ- relations among our dependent variables. First,
mental performer, and may be used by the market environmental performance is positively related to
to help differentiate between the good and the economic performance, suggesting that the market
464 S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471

rewards firms with higher levels of environmental could not include the same variable as both an
performance. Second, environmental performance endogenous and predetermined variable in the same
is positively related to environmental disclosure, system of equations. These are the only changes
suggesting that good environmental performers incorporated in the respecified system of equations.
are more forthright and disclose more with respect Table 4 presents results for the respecified sys-
to quantifiable pollution information. We observe tem of equations using stock price to represent
this last relation to be significant only after con- economic performance. The results are consistent
trolling for endogeneity (3SLS) and to be statisti- with those observed using annual stock return as
cally insignificant using OLS estimation (P=0.37). our economic performance proxy. In Eq. (2.1), we
This contrast in results addresses one of our observe a significantly positive relation between
research questions by providing evidence of the environmental performance and economic perfor-
importance for controlling for endogeneity in this mance, providing some assurance for inferences
research setting, and validates our adoption of a made based on weaker statistical results in our
joint-estimation research design. original returns-specified model. All other expla-
natory variables are significant and correctly
Sensitivity analysis signed, with two exceptions: our operating income
variable (OPINC) is insignificant, and the coeffi-
Model specification using stock price as a proxy cient on the variable representing the firm’s pre-
for economic performance disclosure environment is significantly negative.
In specifying our system of equations, we elected Because the latter variable was insignificant in our
to use annual industry-adjusted stock return as original returns model, we make no inferences
our dependent variable representing economic regarding the ECONPERF–PREDISC relation.
performance. While this decision is justifiable, In Eq. (2.2), we observe results identical to those
assuming efficient markets, researchers have often presented in our main analysis, except that the
preferred to use a valuation approach by measur- coefficient on VISIBLTY is no longer statistically
ing the economic performance construct using significant. Finally, the results for Eq. (2.3) are
market share price. We therefore provide a very similar to those presented in Table 3, with
robustness test of our model using market price one exception. While we still observe a highly sig-
instead of annual industry-adjusted returns as our nificantly positive relation between ENVPERF
dependent variable in Eq. (2.1). We elect to make and ENVDISCL, the coefficient on ENVCON is
two minor changes in the control variables in no longer statistically significant. In sum, the
order to reflect the change from returns to market results of the respecified system of equations with
price as a proxy for economic performance. First, regard to our variables of interest are similar,
since this new proxy for economic performance regardless of whether the economic performance
measures the level of performance as opposed to variable is specified in a ‘‘levels’’ (share price) or in
the change in performance, we substitute book a ‘‘changes’’ (returns) format. This provides some
value of common equity, BOOKVAL, for UE in assurance that our primary results are quite robust
Eq. (2.1). This acknowledges the theoretical link- to economic performance measurement.
age between market value and book value of
equity, on which accounting valuation researchers Industry effects
have heavily focused (e.g., Barth & McNichols, The study’s sample is composed of firms from
1994; Hughes, 2000) in attempting to measure 31 two-digit SIC codes. Given this sample size, it
environmental liabilities. We also add operating would be very inefficient to include an indicator
income, OPINC, which is intended to capture the variable for every industry to completely control
present value of expected future abnormal earn- for all industry differences. However, we isolate six
ings, as per Ohlson (1995). Finally, we substitute two-digit SIC codes containing at least 14 firms
net sales for market value of equity as a control each and comprising a total of 117 firms (SIC
variable for firm size (SIZE) in Eq. (2.3), since we codes 26, 28, 35, 36, 37, and 38). The remaining
S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471 465

Table 4
Three-stage least squares regression results (economic performance=market price per share) coefficients (t-statistics)

Independent variable Predicted sign Dependent variable

ECONPERF [Eq. (2.1)] ENVPERF [Eq. (2.2)] ENVDISCL [Eq. (2.3)]

INTERCEPT 103.79b 0.58b 2.95b

(6.44) (10.94) (2.65)
ENVPERF 135.07b 5.72a
(6.49) (2.53)
BOOKVAL + 1.51b
OPINC + 5.78
PREDISC + 14.12 0.22b
(7.41) (11.44)
GROWTH + 9.23b 0.00
(10.25) (0.12)
MARGIN + 0.30a
ENVEXP /+/+ 4.64b 0.03b 0.17
(4.01) (3.51) (1.49)
ENVCON + 0.00 0.03
(0.39) (0.39)
SIZE 0.06
System weighted R2 0.4968

ECONPERF=Market stock price per share. ENVPERF=Environmental performance measured as the percentage of total waste
generated that is recycled. ENVDISCL=Environmental-disclosure score obtained from content analysis of the firm’s annual report.
BOOKVAL=Book value of common equity per share. OPINC=Operating income per share. PREDISC=Past environmental dis-
closure measured as the average ENVDISCL over the preceding three years. GROWTH=Market-to-book ratio of common equity.
ENVEXP=Environmental exposure—the natural log of toxic waste generated scaled by total revenues. ENVCON=Environmental
concern measured as the primary factor obtained from factor analysis of REPORT, PROGRAM, and COMMITTEE. REPORT
=Indicator variable coded 1.0 if the firm publishes an annual environmental report separate from its annual report, 0.5 if the report is
biennial, 0.33 if the report is triennial, and 0 otherwise. PROGRAM=The number of voluntary EPA programs in which the
firm participates. COMMITTEE=Indicator variable coded one if the firm has an environmental committee and zero otherwise.
MARGIN=Profit margin (net income/net sales) conditioned on environmental exposure. VISIBLTY=The natural log of the
number of Wall Street Journal news announcements about the firm. SIZE=The natural log of net sales.
Statistically significant at the 0.05 level (one-tailed test if the sign is predicted, otherwise two-tailed).
Statistically significant at the 0.01 level (one-tailed test if the sign is predicted, otherwise two-tailed).

40% of our firms are distributed broadly that there may be systematic differences in eco-
throughout the other 25 two-digit SIC codes nomic performance across industries, but that
represented in our sample. Repeating our analysis, there are no systematic differences in environ-
including indicator variables to control for differ- mental performance or environmental disclosure
ences across industries, yields virtually identical related to industry-specific firm characteristics.
results. Only one of the industry control variables This result may also be driven by our selection of a
is significant in Eq. (2.1), while none is significant broad-based, generic metric for environmental
in the other two equations. This analysis suggests performance.
466 S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471

Changes in environmental disclosure execution of each of these corporate responsi-

To further explore the environmental-perfor- bilities is determined by management’s (unobser-
mance–environmental-disclosure relation, we note vable) overall strategy. We, like Ullmann,
that the SEC actively focused on increasing cor- conjecture that the mixed results reported by prior
porate environmental disclosure during the sample environmental empirical research may have arisen
period (see note 21). The effect of this increased because researchers did not allow for these con-
regulatory interest on environmental disclosure is structs to be endogeneous. By explicitly control-
noted in Table 1, as our measure for environ- ling for endogeneity in the study’s research design,
mental disclosure significantly increased ( p < 0.01) this research contributes to our understanding of
between the prior periods represented by PRE- how societal concerns for the environment affect
DISC (the disclosure score averaged over the corporate strategy and, ultimately, firm value. A
1991, 1992, and 1993 period) and the current per- more detailed description of how this study con-
iod in which ENVDISCL is measured (1994). If tributes to the current stream of environmental
the level of ENVDISCL of good performers is research follows.
greater than that of poor performers for the current First, we find that allowing for the potential
period, as the results suggest, then we assume that endogeneity associated with specifying our three
the same conditions existed in prior periods. corporate functions—economic performance,
If good performers disclose more than poor environmental performance, and environmental
performers do, then we believe that the increase in disclosure—makes a statistically significant differ-
environmental disclosure depicted in Table 1 must ence in estimating their interrelations. Finding our
have been predominantly made by good environ- proxy for environmental performance to be endo-
mental performers. We test this proposition by geneous, we provide evidence of the bias asso-
ranking firms by their environmental performance ciated with OLS estimation under such conditions
and measuring the pair-wise correlation between by using a system of simultaneous equations. The
their rankings and the corresponding changes in OLS results suggest that only the economic-per-
environmental disclosure (ENVDISCLPRE- formance–environmental-performance relation of
PREDISC). The correlation coefficient is sig- the potential interrelations among economic per-
nificantly positive in both parametric (0.1985, formance, environmental performance, and envir-
p< 0.005) and nonparametric (0.2262, p < 0.001) onmental disclosure is statistically significant in
tests. This result is consistent with good perfor- our cross-sectional sample. However, by using a
mers, on average, increasing their 1994 environ- joint-estimation research design, we also observe a
mental disclosure levels over their average significantly positive relation between good envir-
disclosure levels for the prior three years. If the onmental performance and more extensive quan-
increased SEC oversight of environmental dis- tifiable disclosure of environmental information.
closures during this period was intended to The contrast between independently estimated
increase disclosures from poor environmental per- OLS results and those obtained through joint
formers, these results suggest that it was not very determination highlight the importance of con-
effective. trolling for endogeneity in the research design.
Second, the significantly positive relation
observed between environmental performance and
Conclusions, limitations, and implications economic performance is consistent with Michael
Porter’s theoretical argument that innovative
This study investigates the relations among eco- solutions to reduce the inefficiencies associated
nomic performance, environmental performance, with pollution promote both environmentalism
and environmental disclosure, after explicitly con- and industrial competitiveness simultaneously.
sidering that these three corporate functions are Porter and van der Linde (1995a, 1995b) reject the
jointly determined. This specification is consistent economic-ecological tradeoff paradigm because it
with Ullmann’s (1985) argument that the assumes that everything except regulation
S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471 467

(i.e., technology, products, processes, and custo- Third, we find that good environmental perfor-
mer needs) is fixed.25 They argue that environ- mers disclose (within the context of our definition
mental pollution represents resources that have of environmental disclosure) more pollution-
been used incompletely, inefficiently, or ineffec- related environmental information than do poor
tively. Our results, consistent with this argument, performers. This finding is consistent with discre-
suggest that managers should change their strate- tionary disclosure theory’s ‘‘good news’’ explana-
gic outlook regarding a firm’s environmental per- tion. This result is also consistent with firms using
formance, from fixating on the deadweight costs voluntary disclosure to project a proactive envir-
of ex post regulatory compliance to focusing on onmental image by providing candid information
the ex ante opportunity costs represented by regarding their environmental performance, even
environmental pollution.26 though that information may be viewed as ‘‘nega-
Our finding that good environmental perfor- tive’’ on a situational basis. This disclosure policy
mance and economic profitability go hand-in- appears to be at odds with strategies that minimize
hand, in addition to supporting Porter’s theore- environmental disclosures because market partici-
tical advocacy of this relation, is also consistent pants may perceive such information as ‘‘bad
with the view that economic performance and news.’’ These results suggest that firms with
environmental performance are both related to the records of good environmental performance can
quality of management. Good managers, acting in be more forthright in disclosing that performance.
the firm’s long-term interest, accept the firm’s The study also documents a positive relation
social responsibility and adopt pro-active strate- between past environmental disclosure and cur-
gies for controlling environmental pollution. rent environmental performance. This association
Finally, because we use market-based proxies to is consistent with the notion that prior disclosure
represent economic performance, the observed establishes a lower bound for management’s
positive economic-performance–environmental- environmental performance. Not achieving this
performance relation is also consistent with inves- lower bound might adversely challenge the expec-
tors’ preferences for equities of environmentally tations of market participants, and potentially
responsible firms. Socially responsible investing is trigger shareholder litigation.
becoming more popular, with over $1.5 trillion Fourth, this research introduces new empirical
worldwide currently invested according to social proxies for environmental performance and envir-
or ethical criteria (Vogel, 2002).27 onmental disclosure. In contrast to studies that

Porter and van der Linde (1995a) cite the ‘‘false’’ tradeoff
between quality and cost as a similar misconception, induced ‘‘Good’’ corporate social responsibility does not neccess-
by the static manner in which the relation is framed. sarily carry-over to ‘‘good’’ accounting practices, as evidenced
This result is also consistent with a recent report (MCA, by the scandals appearing in today’s business press. Infamous
2000) released by the Management Consultancies Association Enron lobbied the Bush administration to accept the Kyoto
that examines the role of business in environmental issues. The Protocol on global warming, no doubt with the expectation of
report describes a ‘‘triple bottom line approach’’ that benefits profiting from trading in a market for carbon dioxide emis-
those firms that adopt proactive environmental strategies. First, sions. Merck, recently cited for misrepresenting revenues,
these firms see competitive advantage in attaining environ- received the prestegious Business Enterprise Trust award in
mental regulations that tend to crowd out competitors that are 1991 for developing and distributing Mectizan, a drug effective
accustomed to less stringent regulation. Second, these firms against river blindness, which treatens millions of the Third
have exploited alternative technologies to fit in with society’s World’s poor. And while Xerox has been lauded as an inter-
desire for a clean environment. Finally, these firms have national leader in enviro-management for its program of recy-
accepted the responsibility to minimize and, they hope, neu- cling copy cartridges, the firm recently paid a $10 million fine to
tralize the negative impacts of their polluting activities (MCA, settle a civil suit filed by the US SEC for overstating profits
2000, pp. 56–57). (Vogel, 2002).
468 S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471

rely on qualitative rankings to measure environ- observed relations among the variables of interest
mental performance, this study adopts a quantita- are relatively stable over time. We know of no
tive measure: the ratio of toxic waste recycled to external event limiting 1994 as a sample period,
total toxic waste generated. Whereas this measure and believe two of our three dependent vari-
may have certain limitations as discussed below, ables—environmental performance and environ-
this ratio also is sufficiently general to be a useful mental disclosure—do not change dramatically
environmental performance measure across around this temporal datum. The stability of
industries—a property that is essential in an inter- economic performance is more problematic. We
industry, cross-sectional research design, such as address this limitation by using two specifications
ours. This recycling ratio measure might be viewed for this variable: a changes specification (industry-
as a summary statistic, much like the earnings adjusted returns) and a levels specification (share
number. We also construct a new measurement of price).
environmental disclosure reported to investors The cross-sectional design presents another lim-
that is uniquely preconditioned on the firm’s pol- itation in our choice of a suitable proxy for envir-
luting activities as reported to regulators. This onmental performance. Because pollution is
disclosure metric not only controls for the degree determined by the production process, pollution-
of disclosure made by firms with limited polluting related measures of environmental performance
activities, but also extracts a ‘‘transparency’’ tend to be industry-specific. For example, mea-
dimension for polluters that file mandated envir- sures of air pollution (e.g., sulfur dioxide emis-
onmental information with regulators, yet do not sions) may be a relevant performance measure for
disclose this information in financial reports. the electric utility industry (Hughes, 2000),
Finally, this study spans the research agendas of whereas measures of water pollution may be more
multiple academic disciplines and contributes to a relevant in measuring the environmental perfor-
growing body of interdisciplinary environmental mance of firms in the pulp-and-paper industry
knowledge. In motivating our research design, we (Cormier & Magnan, 1997). Because our measure
call upon Ullmann’s (1985) meta-analysis of of environmental performance (the ratio of toxic
empirical studies that investigate the interrelations waste recycled to total toxic waste generated) is
among social performance, social disclosure, and one of the few metrics suitable for inter-industry
economic performance, which was published in comparisons (National Academy of Engineering,
the Academy of Management Review. Our results 1999), it is probably less representative of envir-
are also consistent with management guru Michael onmental performance for some firms than indus-
Porter’s win-win argument regarding the positive try-specific measures. Additionally, our measure
relation between environmental and economic for environmental performance does not consider
performance, which is widely cited in both man- the relative toxicity of the waste being recycled,
agement and economic literature. The market- and aggregates all waste into one medium (i.e., air
based proxies used to represent the firm’s eco- pollution is combined with water pollution).
nomic performance are common to both account- Although this summary statistic (recycling ratio) is
ing and finance research. While environmental suitable for inter-industry comparisons, it likely to
research questions may extend into the domains of be noisy relative to industry-specific metrics.
several academic disciplines, researchers are reluc- As explained above, one of the primary chal-
tant to violate these academic boundaries. How- lenges in conducting this research was measuring
ever, without such interdisciplinary trespassing, a each of the three dependent variables. Whereas we
holistic approach to investigating these inter- used two measures of economic performance and
disciplinary interrelations of interest would not be a recycling ratio common across industries for our
possible. measure of environmental performance, our
Like all cross-sectional studies, limitations to choice of a measure for environmental disclosure
interpreting our results apply regarding whether was, by its very nature, subjective. Although we
the time period examined is representative and the restrict our definition of environmental disclosure
S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471 469

to include only specific pollution-related environ- forthcoming and factual environmental disclosure
mental measures, and clearly highlight this should be good news for those questioning the
restriction throughout the discussion related to compatibility of corporate social responsibility
variable selection, our method of scoring dis- and economic profitability. Non-financial mea-
closure quality retains a subjective element.28 As a sures of environmental performance, such as the
final limitation, we recognize that our sample, recycling ratio presented in this study, may also be
drawn from S&P 500 firms, induces a size bias. leading indicators of future financial performance.
Whereas our results may be generalized for large These measures therefore may be suitable candi-
firms, inferring that small firms may behave simi- dates for incorporation in the firm’s ‘‘balanced
larly is overreaching. scorecard’’ used to evaluate managerial and firm
Although we acknowledge these limitations, we performance. Managers evaluated in this manner
also believe that we have used those econometric should be increasingly proactive in introducing
tests most relevant in assuring that our models of new processes that improve both production
simultaneous equations are correctly specified. efficiency and environmental sustainability.
Assuming correct specification, the simultaneous-
equation research design provides a significant
improvement in estimating the relations among Acknowledgements
environmental performance, environmental dis-
closure, and economic performance over that We appreciate the helpful insights and sugges-
obtained by independently estimated OLS models. tions provided by Julia D’Souza, Robin Dubin,
We conclude that the simultaneous-equation Tim Fogarty, Charlene Henderson, Dennis
approach used in this study provides a more Oswald, Larry Parker, Kenny Reynolds, Kay
coherent explanation regarding the relations Stice, and Kristina Zvinakis. This paper has also
among environmental disclosure, environmental benefited from comments made by participants at
performance, and economic performance than do the Brigham Young University, Case Western
those reported by prior studies using pair-wise Reserve University, and Louisiana State Uni-
tests of association. Further research might incor- versity accounting workshops, the Tenth Annual
porate different variables for these constructs over Financial Economics and Accounting Conference,
different sample periods. As the world becomes and the 2000 AAA Annual Meeting. The first
more environmentally conscious (e.g., the Kyoto author gratefully acknowledges the financial sup-
Protocol on climate change), additional exami- port of King Fahd University of Petroleum and
nation of the relations among economic perfor- Minerals.
mance, environmental performance, and
environmental disclosure is increasingly warranted.
The study’s primary results that good environ- References
mental performance is associated with good
economic performance and also with more Abbott, W., & Monsen, R. (1979). On the measurement of
corporate social responsibility: self reported disclosures as a
method of measuring corporate social involvement. Academy
Our measure of environmental disclosure appears to be of Management Journal, 22, 501–515.
consistent with information sought by philanthropic founda- Atiase, R. K. (1985). Predisclosure information, firm capitali-
tions and investment managers who recently called upon the zation, and security price behavior around earnings
SEC to enforce regulations requiring firms to more fully dis- announcements. Journal of Accounting Research, 23(1), 21–
close environmental liabilities. These investor groups appealed 36.
to the SEC that hidden environmental costs threaten their Bank, D. (2002). Group urges enforcing rules of environmental
portfolios similar to other corporate accounting ‘‘tricks.’’ SEC disclosure. The Wall Street Journal, 22 August 2002, B2.
guidelines require publicly held companies to disclose any Barth, M., & McNichols, M. (1994). Estimation and market
‘‘material’’ events and specifically require companies to disclose valuation of environmental liabilities relating to superfund
legal proceding that might result in sanctions of $100,000 or sites. Journal of Accounting Research, 32(Supplement), 177–
more (Bank, 2002). 209.
470 S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471

Barth, M., McNichols, M., & Wilson, P. (1997). Factors influ- financing, dividend and compensation policies. The Journal
encing firms’ disclosures about environmental liabilities. of Accounting and Economics, January/April/July, 125–160.
Review of Accounting Studies, 2, 35–64. Gray, R., Jarad, M., Power, D., & Sinclair, D. (2001). Social
Belsley, D., Kuh, E., & Welsch, R. (1980). Regression diag- and environmental disclosure and corporate characteristics: a
nostics: identifying influential data and sources of collinearity. research note and extension. Journal of Business Finance and
New York: John Wiley & Sons. Accounting, April/May, 327–356.
Blacconiere, W., & Patten, D. (1994). Environmental dis- Gray, R., Kouhy, R., & Lavers, S. (1995). Corporate social and
closures, regulatory costs, and changes in firm value. Journal environmental reporting. Accounting, Auditing, and Account-
of Accounting and Economics, 18(3), 357–377. ability Journal, 8(2), 47–77.
Botosan, C. (1997). Disclosure level and the cost of equity Guthrie, J., & Parker, L. (1989). Corporate social reporting: a
capital. The Accounting Review, 72(3), 323–349. rebuttal of legitimacy theory. Accounting and Business
Bowen, R., Castanias, R., & Daley, L. (1983). Intra-industry Research, 19(76), 343–352.
effects of the accident at three mile island. Journal of Finan- Hausman, J. (1978). Specification tests in econometrics. Econ-
cial and Quantitative Analysis, 18, 87–111. ometrica, 46, 1251–1271.
Bragdon, J., & Marlin, J. (1972). Is pollution profitable? Risk Hughes, K. (2000). The value relevance of nonfinancial mea-
Management, 19, 9–18. sures of air pollution in the electric utility industry. The
Brockhoff, K. (1979). A note on external social reporting by Accounting Review, 75(2), 209–228.
German companies. Accounting, Organizations and Society, Hughes, S. B., Anderson, A., & Golden, S. (2001). Corporate
4(1/2), 77–85. environmental disclosures: are they useful in determining
Buchholtz, R., Marcus, A., & Post, J. (1992). Managing envir- environmental performance? Journal of Accounting and
onmental issues: a casebook. Englewood Cliffs, NJ: Prentice- Public Policy, 3(20), 217–240.
Hall. Ilinitch, A., Soderstrom, N., & Thomas, T. (1998). Measuring
Christie, A. (1987). On cross-sectional analysis in accounting corporate environmental performance. Journal of Accounting
research. Journal of Accounting and Economics, 9(3), 231– and Public Policy, 17, 383–408.
258. Ingram, R., & Frazier, K. (1980). Environmental performance
Collins, D., & Kothari, S. P. (1989). An analysis of inter- and corporate disclosure. Journal of Accounting Research, 18,
temporal and cross-sectional determinants of earnings 612–622.
response coefficients. Journal of Accounting and Economics, Kennedy, P. (1994). A guide to econometrics. Cambridge, MA:
11, 143–182. The MIT Press.
Cormier, D., & Magnan, M. (1997). Research note: investors’ Li, Y., Richardson, G., & Thornton, D. (1997). Corporate
assessment of implicit environmental liabilities: an empirical disclosure of environmental liability information: theory and
investigation. Journal of Accounting and Public Policy, 16, evidence. Contemporary Accounting Research, 14(3), 435–
215–241. 474.
Corporate Environmental Profiles Directory. (1994, 1995). Little, P., Muoghlu, M., & Robison, D. (1995). An empirical
Investment. Responsibility Research Center: Washington analysis of the comparability of disclosure tendencies within
DC. and across industries. Research in Accounting Regulation,
Deegan, C., & Gordon, B. (1996). A study of the environ- 105–117.
mental disclosure practices of Australian corporations. Management Consultancies Association (MCA). (2000). Busi-
Accounting and Business Research, 26(3), 187–199. ness and the environment. London, UK: MCA Available:
Financial Accounting Standards Board. (1975). Statement of
financial accounting standards No. 5 Accounting for Con- Narver, J. (1971). Rational management responses to external
tingencies. Norwalk, CT: FASB. effects. Academy of Management Journal, March, 99–115.
Freedman, M., & Jaggi, B. (1982). Pollution disclosure, pollu- National Academy of Engineering. (1999). Industrial environ-
tion performance and economic performance. Omega, 10, mental performance metrics: challenges and opportunities.
167–176. Washington DC: National Academy Press Chapter 8..
Freedman, M., & Jaggi, B. (1992). An investigation of the Niskala, M., & Pretes, M. (1995). Environmental reporting in
long-run relationship between pollution performance and Finland. Accounting, Organizations and Society, 457–466.
economic performance: the case of pulp-and-paper firms. Ohlson, J. (1995). Earnings, book values, and dividends in
Critical Perspectives on Accounting, 3(4), 315–336. equity valuation. Contemporary Accounting Research, 11(2),
Freedman, M., & Wasley, C. (1990). The association between 661–687.
environmental performance and environmental disclosure in Patten, D. (1992). Intra-industry environmental disclosures in
annual reports and 10-Ks. Advances in Public Interest response to the Alaskan oil spill. Accounting, Organizations
Accounting, 3, 183–193. and Society, 471–475.
Friedman, M. (1962). Capitalism and freedom. Chicago: The Patten, D. (1995). Variability in social disclosure: a legitimacy-
University of Chicago Press, Phoenix Books. based analysis. Advances in Public Interest Accounting, 6,
Gaver, J., & Gaver, K. (1993). Additional evidence on the asso- 273–285.
ciation between the investment opportunity set and corporate Patten, D., & Nance, J. (1998). Regulatory cost effects in a
S.A. Al-Tuwaijri et al. / Accounting, Organizations and Society 29 (2004) 447–471 471

good news environment: the intra-industry reaction to the Spicer, B. (1978). Investors, corporate social performance and
Alaskan oil spill. Journal of Accounting and Public Policy, 17, information disclosure: an empirical study. The Accounting
409–429. Review, 53, 94–111.
Porter, M., & van der Linde, C. (1995a). Green and competitive: Stevens, W. (1984). Market reaction to corporate environ-
Ending the stalemate. Harvard Business Review, 73(5), 120–134. mental performance. Advances in Accounting, 1, 41–61.
Porter, M., & van der Linde, C. (1995b). Toward a new con- Tietenberg, T. (1998). Disclosure strategies for pollution control.
ception of the environment-competitiveness relationship. Environmental and Resource Economics, 11(3–4), 587–602.
Journal of Economic Perspectives, 9(4), 97–118. Ullmann, A. (1985). Data in search of a theory: a critical
Richardson, A., & Welker, M. (2001). Social disclosure, finan- examination of relationships among social performance,
cial disclosure and the cost of equity capital. Accounting, social responsibility, and economic performance of U.S.
Organizations and Society, 26(7/8), 597–616. firms. Academy of Management Review, 10, 540–557.
Rockness, J. (1985). An assessment of the relationship between Verrecchia, R. (1983). Discretionary disclosure. Journal of
U.S. corporate environmental performance and disclosure. Accounting and Economics, 5(3), 179–194.
Journal of Business Finance and Accounting, 12, 339–354. Vogel, D. (2002). Recycling corporate responsibility. The Wall
Rockness, J., Schlachter, P., & Rockness, H. (1986). Hazardous Street Journal, 20 August 2002, B2.
waste disposal, corporate disclosure and financial perform- Walley, N., & Whitehead, B. (1994). It’s not easy being green.
ance in the chemical industry. Advances in Public Interest Harvard Business Review, May–June, 46–52.
Accounting, 1, 167–191. White, H. (1980). A heteroskedasticity-consistent covariance
Smith, C., & Watts, R. (1992). The investment opportunity set matrix estimator and a direct test for heteroskedasticity.
and corporate financing, dividend, and compensation poli- Econometrica, 48, 817–838.
cies. Journal of Financial Economics, 32, 263–292. Wiseman, J. (1982). An evaluation of environmental disclosure
Shane, P., & Spicer, B. (1983). Market response to environ- made in corporate annual reports. Accounting, Organizations
mental information produced outside the firm. The Account- and Society, 7, 53–63.
ing Review, 58, 521–538. Zeghal, D., & Ahmed, S. (1990). Comparison of social
Slavich, J. (1994). Show and tell: new developments impacting responsibility information disclosure media by Canadian
the disclosure of environmental liabilities. Environmental firms. Accounting, Auditing, and Accountability Journal, 3(1),
Permitting, Summer, 327–343. 38–53.