MEASURING

CORPORATE SUSTAINABILITY

by

Giles Atkinson
Tannis Hett
Jodi Newcombe

CSERGE Working Paper GEC 99-01

 MEASURING CORPORATE SUSTAINABILITY

by
1
Giles Atkinson
2
Tannis Hett
1
Jodi Newcombe
1
Centre for Social and Economic Research
on the Global Environment
University College London
and University
of East Anglia.

2
Economics For The Environment
Consultancy (EFTEC) Ltd.

Acknowledgements

The Centre for Social and Economic Research on the Global Environment (CSERGE) is a
designated research centre of the UK Economic and Social Research Council (ESRC). We are
grateful to Jan Bebbington, Dan Colby, Rob Gray, Irene Lorenzoni, M.R. Mathews and David
Pearce for valuable comments on earlier drafts. Any remaining errors are the responsibility of
the authors alone.

ISSN 0967-8875
Abstract

Although sustainable development has been interpreted primarily as a national

(or global) goal, there is increasing discussion of the sustainable city, the
sustainable sector and the sustainable business. In this paper, the notion of
corporate sustainability is explored and developed. Intuitively, from societys
point of view this might be thought of in terms of the contribution of a given
entity (e.g. business or sector) to sustainability defined in the wider sense (e.g.
nation). From the entitys perspective, it is how this contribution has
implications for the sustainability of its own activity is also of concern. The key
to defining the sustainable business is to reconcile both of these interpre-
tations. It is the aim of this paper to provide such a framework. Issues to be
explored include of environmental responsibility, the extension of the polluter-
pays-principle to full cost accounting and valuing pollution in green accounts.
Defining the sustainable business is more than just an academic concern.
Corporate entities are increasingly under pressure to demonstrate how they
contribute to the national sustainability goals outlined by government. Hence,
this paper seeks to provide practical advice on how businesses might adapt and
improve current environmental accounting and reporting practice.
1. Introduction

As a result of official responses to both the Brundtland Commission (World

Conference on Environment and Development, WCED, 1987) and the United
Nations Conference on Environment and Development in 1992, most govern-
ments have adopted sustainable development as a national goal. An emerging
debate is how economic sectors or businesses contribute to this objective. This
has resulted in a number of concepts such as the sustainable business or
corporate environmental responsibility. However, the question of what
behaviour this is intended to embody needs to be addressed before this debate
can be used to inform practical concerns regarding the monitoring of progress
towards corporate sustainability.

There are two broad responses to this measurement problem. The first begins
with the proposition that there is little in the notion of the sustainable business
beyond defining a set of pragmatic guidelines whereby a corporate entity can
improve its environmental performance. The measurement issue here is to find
meaningful environmental indicators that capture the flavour of the broader
sustainability debate; for example, by conveying environment-economy
linkages. The second response is that lessons drawn from the green national or
macro- accounting literature allow us to define more formally what it means
for a business to be either sustainable or unsustainable. Common to both
approaches is an increased emphasis on accounting for external pressures or
impacts attributable to a corporate entity.

In particular, it is in exploring the latter response that we aim to contribute to the

current debate surrounding the meaning of full cost accounting. Indeed, the whole
issue of evaluating external costs associated with environmental change within
corporate accounts is judged to be undeveloped in a recent survey by de Koning-
Martens and van der Ende (1997). Hence, we outline in some detail one way in
which full cost accounting can be extended and developed. As an empirical
illustration we provide an example for air pollution, although, in principle, the
approach can be extended to other forms of pollution. Sustainable development is
undoubtedly a complex notion open to numerous interpretations and many issues
surrounding the usefulness and reliability of full cost accounting remain
unresolved. However, we argue that, far from being impossibly confusing, it is
possible to take practical steps towards measuring corporate sustainability.

1
2. Existing Practice

Corporate environmental accounting and reporting covers a diverse range of

activities with currently little standardisation of existing practice. If comparisons of
environmental performance are to be made across the corporate sector this will
require an overarching framework with which to evaluate these activities. For
example, within the domain of green national accounting this function is fulfilled
by the United Nations Satellite Environmental and Economic Accounts (UN
SEEA). This, in turn, is an adjunct to the conventional UN System of National
Accounts (SNA). The SEEA framework embodies natural resource accounts,
resource and pollutant flow accounts, environmental protection expenditure
accounts and the estimation of green accounting aggregates (United Nations, 1993;
Atkinson et al. 1997). The primary role of this framework is to lend some degree of
coherence to an otherwise impenetrable mass of data and indicators.

Within the corporate environmental accounting literature similar concerns can

be identified (see, for example, Bennett and James, 1997). Particular attention
has been given to those activities which explicitly identify and account for
environ-mental expenditures incurred by firms. Also relevant to the accounting
problem are those pressures on the environment such as pollution discharges
and emissions which at present go unabated. Hence, to the extent that a firms
environmental expenditure is not restricted to abatement effort that increases the
value of the economic output of the firm (e.g. clean-up of contaminated water
inputs) these two areas of accounting are related. In other words, expend-itures
actually incurred reflect a corporate decision to internalise, in some degree, its
external impact. At the national level, environmental expenditure accounts
typically have been used to address questions about the total economic burden
of environmental protection and the distribution of this burden between sectors.
Within individual corporations, a primary use is for monitoring which
production techniques and inputs will minimise internal environmental expend-
iture in order to attain a given protection goal. The separation of environmental
expenditures from general operating costs allows the former to be allocated to
processes, products or budgets and, where relevant, to apply appropriate
appraisal techniques to capital budgeting decisions (Bennett and James, 1997).

Work by the Intergovernmental Working Group of Experts for International

Standards of Accounting (ISAR), United National Conference on Trade and
Development (UNCTAD) and the European Union has sought to improve the
transparency, reliability and comparability of this information. It is argued that
recording environmental expenditures separately allows interested parties to gauge
the extent to which firms are sensitive to environmental concerns. Of course, high
levels of environmental expenditure do not necessarily correlate with good
2
environmental performance per se. Any appraisal thus requires a more subtle
demonstration of, for example, costs incurred per unit abated. In addition,
identifying environmental expenditures that are not of the end-of-pipe category
is far from trivial; e.g. separating the environmental expenditure from new
production technologies that jointly increase productivity and decrease pollution
emissions (Atkinson et al., 1997).

In many respects, the goal of gauging concern is better served by measuring external
costs. An external cost can be defined as a cost imposed by an entity - e.g. a firm - on
others; that is, other firms or households. For example, a firm emitting air pollutants as
a by-product of its economic activity may impose health costs on households. These
health costs are external because the firm takes no account of this outcome when
1
deciding on the amount of pollution that it should emit. Put another way, the firm
presently faces insufficient incentive to internalise the costs that it imposes on others.
Hence, it could be argued that, almost by definition, this firm also has little incentive
to monitor and report this activity. Countervailing reasons why these measurement
issues are relevant include the anticipation of potential future legislation to restrict
pollution, green consumerismand the environmental concerns of employees (de
Koning-Martens and van der Ende, 1997).

It is the complicated and often controversial process of quantifying external

costs that is primary concern of full cost accounting (Gray, 1992). Examples of
physical indicators that in some part reflect external effects can be found in
some company environmental reports. However, it is not possible in general to
comment on sustainability or the full costs of economic activity on the basis of
these data alone. Of course, this is not to say that these data are not useful. This
information is a crucial building block in the construction of more sophisticated
indicators. It may be also be the case that in the absence of acceptable methods
to construct more these indicators of simple physical pressures placed on the
environment are the only means of monitoring external costs. However, it is the
human welfare costs that these pressures ultimately generate that is, in principle,
the object of concern.

An indicator is not interesting in its own right and only acquires meaning when
supported by a convincing rationale or theory and balanced against policy
relevance and ease of interpretation (OECD, 1993). Moreover, it is usually
manipulations and aggregation of these data, consistent with such a theory, that
prove to be the most useful for many types of audience. For example, consider
the proposition that firms should aim to maintain output while decreasing
environmentally damaging by-products (Owen, 1997). In other words, for any

1
This definition can be found in Tietenberg (1996).
3
given level of output our goal is to minimise damage to the environment. What
this could suggest is indicators of the extent to which a firms economic activity
is being decoupled from its environmental consequences. Environ-mental
pressure indices, in themselves, do not capture this very well but can be linked
with economic indicators such as output (in physical or monetary terms). This is
one simple way in which economy-environment linkages can be brought to the
2
fore of environmental reporting.

In this way, any reduction in the amount of energy use or emissions per unit of
output could be approximated to a reduction in environmental impact per unit
of economic activity. (The analogue to this at the national level is an indicator of
say, the ratio of a countrys emissions of carbon dioxide to its GDP.) Efficiency
gains may come about naturally - that is, as a result of structural change in the
economy or at the firm level through investments that are for example output
augmenting and less polluting - or as a result of deliberate policy measures.
Disentangling the sources of efficiency gain can be complex and hence,
decision-makers may need additional guidance on a decomposition of causal
factors. Note, in addition, that this measure of efficiency is consistent with an
increase in the absolute level of environmental impact. This is because gains in
efficiency can be outweighed by the absolute growth of output.

In addition to the description of trends over time, comparisons of performance

across different economic entities is also important: i.e. benchmarking. Schemes
to benchmark (i.e. compare) environmental performance could be envisaged
either between (comparable) sectors or within a particular sector. In the case of
the latter, if we define some measure of average sector performance for a
particular attribute of environmental performance - e.g. resource use per unit of
output - then the distance between any individual firm and the average can be
evaluated. In this way, a firm is able to benchmark its own performance vis--
vis the average or individual firms. Bartolomeo (1996) provides several
examples of the increasing use of environmental bench-marking in Italy
(computer, semi-conductor and telecommunications sectors) and the United
States (where the Council on Economic Priorities produces an emission index).
Within the United Kingdom, both the Office for National Statistics (ONS) and
Department of Environment, Transport and the Regions (DETR) envisage
bench-marking to be an important, but as yet unexplored, use of the environ-
mental accounts and sustainability indicators currently being developed for use
in public policy in the UK.

2
An illustration of this can be found in National Power (1997) and PowerGen (1997).
4
Indicators of simple environmental pressures, if suitably constructed and refined,
can be a useful means of signalling environmental performance. However, much of
potential of full cost accounting would appear to lie in the possibility of
understanding corporate sustainability in terms of a better approximation of the
human welfare cost of pollution. It is interesting to note the handful of innovative
efforts by firms to account for external costs in this manner. One much cited
example is by BSO/Origin, a Dutch computer software consultancy, which
recorded a monetised imputation reflecting its external impact in an environ-
3
mental value added statement(Huizing and Dekker, 1992). Ontario Hydro
embraced a mode of accounting familiar from established academic studies of the
social costs of fuel cycles (see, for example, US EPA, 1996; Smith, 1996; European
Commission, 1995). This study not only modelled the dispersion of the firms
emissions but also quantified subsequent physical impacts and further used
economic valuation techniques to translate physical effects into monetary terms.
Tuppen (1996) notes that both AT&T and Dow Chemicals are considering a similar
approach to the costing of externalities.

The question we address in the remainder of the paper, is how full cost accounting
might be used to provide information about the sustainable business. The rest of
this paper is set out as follows. Section 3 provides an overview of the concept of
sustainable development both in the wider literature and in contributions to the
corporate accounting debate. Section 4 then goes on to outline a framework
whereby these distinct traditions and some directions for full cost accounting be
outlined. Section 5 provides empirical examples of a greeneconomic performance
indicator for the corporate sector. Finally, section 6 offers some conclusions.

3
This re-stated the conventional profit and loss account in value added form and attributed an
explicit monetary value to environmental impacts. These included all upstream and
downstream, direct and indirect costs associated with the firms activity. Impacts were valued
at the costs of abating the damage caused.
5
3. Defining Sustainable Development

The most celebrated formulation of sustainable development is that given by the

World Commission on Environment and Development (the Brundtland Commission)
(WCED, 1987). Over a decade on, substantial progress has been made in clarifying the
many controversial issues that have emerged since this early shaping of the problem in
the Brundtland Report. Nevertheless, significant diversity exists regarding what
actually constitutes sustainable development. Our objective in this section is to set out
briefly some of the aspects of this debate insofar as it has influenced the green national
accounting literature. This will provide the building blocks for our discussion of
corporate sustainability in the remainder of the paper. Hence, some aspects of the
theory described below are distinct from those concepts that have evolved in the
corporate accounting literature. However, we would argue that the distinction is one of
emphasis rather than a difference in substance.

Sustainable development can be defined as non-declining human well-being

over time (Pezzey, 1989; Pearce, Markandya and Barbier, 1989). If this goal is
accepted as a desirable objective for society, then current decisions need to be
4
made with some degree of concern for impacts on future generations.
Controversy arises when we consider how this goal is to be achieved; i.e. what
are the conditions for attaining sustainable development. It is only by resolving
this debate that progress towards sustainability can be monitored.

While there is no single unified theory of sustainable development, all theories

share a common theme in recognising that future welfare or well-being is
determined by what happens to wealth over time. Indicators of sustainability,
therefore, tend to emphasise either stocks of wealth or, more specifically, how a
portfolio of assets is managed over time. Changes in this portfolio include
investments in conventional forms of produced assets (infrastructure, buildings and
machines) and human capital (via training and educational expenditures and
primary health care). The contribution of the recent debate makes it clear that
portfolio management must also take account of, for example, the depletion of non-
renewable and living resources and changes in environmental liabilities arising as a
result of pollution. This research programme is constantly evolving as evidenced
by the recent focus on the link between social capital and sustainability (see, for a
discussion, World Bank, 1997). It is interesting to note that this emphasis on social
sustainability has long been a feature of the corporate accounting where, for
example, questions regarding quality of the work environment have been central
(see, for example, Mathews, 1997). Clearly, this

4
That is, unless we can rely on improvements in technology to take care of the future.
6
social dimension is important. However, it is the environmental aspect of the
sustainability problem upon which we wish to focus in this paper.

It is on the relative importance of components of the asset portfolio that

opinions diverge into two broad camps. The first is weak sustainability and
states that total wealth should not decrease over time. So long as adequate
compensation is made in the form of saving and investment it does not matter if,
say, produced capital or human capital are substituted for natural assets. Put
another way, it is the overall portfolio which is bequeathed to the future that
matters. On this view, there is nothing special about environmental assets as
such. However, if the liquidation of any asset is, unaccompanied by adequate
compensating investment - e.g. in produced, human or other forms of natural
capital - well-being will decline in the future. On the basis of the above
definition, this would not be considered to be sustainable. It is also worth noting
that weak sustainability does not permit overly rapid depletion of non-renewable
resources or imply that excessive environmental degradation does not matter.
Hence, in a world of over-pollution and over-depletion one way to increase
sustainability would be to implement policies that either prevent wasteful use of
resources or reduce pollution levels (Atkinson et al., 1997).

In contrast, strong sustainability suggests a greater emphasis on the conservation

of natural assets within the broader goal of prudently managing a portfolio of assets
over time. Specifically, it is argued that some classes of natural assets have no
substitutes and therefore cannot be replaced: i.e. critical natural assets. For
example, where critical assets provide life support services it has been argued that
allowing the stock size to fall below a certain critical level could result in serious
consequences (Pearce, Hamilton and Atkinson, 1996). If these outcomes are to be
avoided then sustainability requires that stocks of such assets are maintained above
these threshold levels. The key point is that catastrophic welfare losses could result
if the strong sustainability regime is not followed.

There would be little conflict between these two schools of thought if those
assets which embody critical services could be identified unambiguously. The
key indicators for such an economy would be twofold: are stocks of critical
natural assets declining? and are changes in the real value of all other assets
persistently negative in the aggregate? A positive answer to either of these
questions would be an indication of unsustainability. Of course, in the real world
matters are more complicated. Just as few would argue that all natural assets are
substitutable it is also unlikely that all assets are critical in the sense defined
above. However, the problems involved in firstly, identifying critical assets and
secondly, determining thresholds are far from trivial and are at the frontier of
interdisciplinary research (Hamilton, 1997).
7
Much of the formal discussion regarding the sustainability of particular
development paths has been conducted, either implicitly or explicitly, with the
national or global economy in mind. For example, it has been demonstrated how
changes in wealth lead to changes in average (e.g. per capita) conditions over
time and this has been conducted mostly in the context of greening national
accounts (Hartwick, 1993; Mler, 1991; Hamilton, 1996a). There has been,
however, increasing discussion which has looked at sustainability from the
perspective of smaller spatial scales (such as cities or regions) and economic
entities (such as sectors or firms). The predominant approach within the
corporate accounting literature is to take a definition of (usually, strong)
sustainability from the broader literature and redefine it in this specific context.
For example, Hawken (1993) argues that a firm is behaving sustainably if it
does not reduce the capacity of the environment to provide for future
generations. Similarly, Bebbington and Gray (1997) assert that at a minimum
the sustainable business is one that leaves the environment no worse off at the
end of each accounting period than it was at the beginning.

At first blush these definitions are enticing, serving both to focus attention on
responsibility for environmental impacts and to suggest that firms should be more
responsive to these social concerns. However, it is arguable that such statements
contain little that is operational. Firstly, it is unlikely that following this logic will
yield a more formal understanding of whether a business is sustainable in an way
analogous to that which exists regarding national or global sustainability. Clearly,
we would need to know what is happening elsewhere in the economy - e.g. the
behaviour of all other firms -, among other things, to make any evaluation of
corporate sustainability or otherwise. Secondly, there is an enormous range of
environmental assets that are affected by corporate activity. On one hand, if strong
sustainability is interpreted as meaning that all such assets should be preserved or
restored then, in the extreme, this is hardly conducive to any form of economic
activity. If, on the other hand, it does not imply such blanket support for such
actions then there is a huge research programme required to determine exactly what
actually should be conserved or restored.

A practical work programme would be to develop guidelines for improving

environmental performance while at the same time capturing the broad flavour of the
sustainability debate. Interestingly, Bebbington and Gray (1997) note that neither the
ICC Charter for Sustainable Business or the Business Council for Sustainable
Development offer an interpretation of the sustainable business. Instead, these bodies
outline a programme of action which is intended to be economically sound while
decreasing environmental impacts. While this is commendably pragmatic there are, of
course, many details to resolve such as the definition of what is economically
8
sound. Unless win-win opportunities are pervasive, which is unlikely (see, for
example, Cairncross, 1995) there will be a wide range of trade-offs between
economic performance and various aspects of what we have conveniently
pooled together as environmental quality. Hence, it would be desirable to
augment this pragmatic programme with a framework and rationale that can be
used to address explicitly this and other issues. What we argue is that an
exploration of the concepts of full cost accounting may offer the promise of a
better understanding of corporate sustainability.

9
4. Sustainable Development and Full Cost Accounting

Intuitively, from societys point of view the interesting question can be thought
of in terms of the contribution of a given entity (e.g. business or sector) to
sustainability defined in the wider sense (e.g. nation). From the entitys own
perspective, the extent to which its contribution impinges on the sustainability
of its own activity will also be of concern. The key to defining the sustainable
5
business is to reconcile, in a convincing way, these two outlooks. Thus, our
concern is how existing accounting information relating to economic perform-
ance can be augmented with indicators that provide signals regarding the overall
6
sustainability of an entity and its wider environmental performance.

Two approaches spring to mind. Firstly, there are those sets of environmental
accounts in non-monetary units which accompany conventional performance
indicators or accounts. Secondly, those accounts where environmental changes are
monetised and used to adjust conventional company accounting aggregates such as
profit or value-added - extended monetised accounts or full cost accounts. The idea
is not to replace existing accounting practices Clearly, existing records of financial
flows have well-defined and established uses essential to corporate decision-
making. In contrast, green accounting is still at an experimental stage and further
its uses have too often been poorly defined. There may well exist a corresponding
rationale for recording external costs associated with conventionally measured
7
flows. However, disclosure of this information is best viewed as an adjunct or
satellite to the existing accounts. The national accounting analogy is the UN SEEA
described above which leaves existing accounting aggregate measures such as GDP
intact. In the corporate context, an environmental report might be thought of as a
suitable medium for presenting such information.

The rationale for full cost accounting can be illustrated with reference to how a
country might account for transboundary pollution. Hamilton and Atkinson
(1996) propose an extension of the polluter pays principle (PPP) to the domain
of national accounting. If the property right to a clean environment lies with
5
Of
by course,
aview, about
decline
economic
consumers.
is it time
the
over should
output
conditions
For
definition some
an beresult
aof
(suchnoted
any
assectors as
or
impossibility. that,
given in
firm
offirms
any
changes
For thegeneral,
or
insector
number of there
(Pezzey,
reasons
costs)
possibility
example, there and is
isofno nothing
including
the
sustaining
such special,
1992).An from
asentitys
competitive
changing atastes
output
thing from societys
output mine.point
might
a pressures,
and
particular
sustainable of
exhibit
general
preferences
activity
6
See Mathews (1997) for a discussion of the disclosure of social information in company
accounts.

7This is in addition to the recording of internal environmental expenditures, e.g. mandated by

regulation or some other form of public pressure, which we take as given.
1
0
downwind or downstream countries, then damages caused by transboundary
externalities are attributable to the polluting country. In accounting terms this
damage is a notional liability that should be reflected in the accounts of the
polluting country. For two countries, A and B, any external effect on country B
that is directly identified with the generation of country As income should be
attributed to the national account of country A. This generalises to any decrease
in welfare suffered by any individual in the global population as a result of
emissions or discharges from country A. This explanation is very general. It says
nothing about how we might measure pollution damage; that is, in physical or
monetary terms. Nor does it say anything about how pollution should be valued
if money values are thought to be the relevant measurement unit. Thus, this can
support a wide range of activities from simple physical accounting for
environmental pressures to more ambitious forms of reporting. However, the
emphasis is squarely on those activities that describe external costs in some
shape or form.

To extend the above example further, if damage could be defined more formally
and could be expressed in money values then pollution damage in B attributable
to A should appear as a deduction from income in country A. In other words,
any countrys national income should be adjusted with an amount corresponding
to damages (specifically, externalities) directly associated with the generation of
that income. An equally strong case exists for debiting a given countrys total
savings by an amount which, at least notionally, should be set aside in order to
compensate the recipients of the pollution emitted and transferred across
international boundaries. This environmentally-adjusted net saving rate - or
genuine saving as it has become known - is a measure of the rate at which an
economy is on balance creating or liquidating its wealth or assets (Hamilton and
Atkinson, 1996, Atkinson et al., 1997, World Bank, 1997). Persistently negative
genuine saving is unsustainable in that welfare will have to decrease at some
point in the future.
8
This translates in an analogous way to corporate accounting. Any damage that
can be attributed to the polluting entity, such as a firm, is the magnitude that
corresponds, at least notionally, to that amount to be set aside in order to
compensate the victims of the pollution emitted. What then does this allow us
to say about the sustainability of the entity? We suggest the following which,
although hypothetical in nature, conveys the essence of what we mean when we
8
If anything the case for applying for methodology is even stronger here. Although the PPP is
an internationally accepted principle there is in practice no supranational institution to
implement and enforce it. In the case of an individual firm, sited within national jurisdictions,
credible mechanisms do exist. There are complications, however, if a firm is a global polluter
or if the migration of firms to a pollution-havenis a credible threat.
11
talk of an entity behaving sustainably or otherwise. If there existed some
mechanism to make a liability actual by enforcing property rights then this
would represent a tangible decrease in the entitys income available say for
investment in new assets. It is in this sense that, other things being equal, the
entity is less sustainable as a result of the damage that it causes. Conversely, an
entity can become more sustainable, other things being equal, by reducing the
amount of external damage that it causes. For example, Whitby and Adger
(1996) construct a land-use account extended to include imputations describing
externalities that farming activities cause not just in the land-use sector itself but
in a global context (for example, emissions of greenhouse gases). The imputed
items represent the value of damage caused by the polluter and furthermore are
directly associated with the generation of farming income.

The analogy for this type of pollution chargemight be a contingent liability. That is,
an expense that has not materialised by the end of the accounting period (Holmes and
Sugden, 1997). The difference is that such liabilities typically are applied in the
presence of accidental discharges of persistent pollutants (e.g. contaminated land)
rather than routine emissions of pollutants (Segerson, 1994; Devlin and Grafton,
1997). Hence, the liability analogy should not be taken too literally. Any monetary
estimate of damage done does not necessarily constitute what the firm would be liable
to pay in the event of a future public policy intervention. This would depend on what
form the policy takes and the precise allocation of property rights it embodies. (This is
an issue to which we return below when we consider the level of charge that a firm
might account for given the level of pollution that it emits.) Rather, the more specific
and practical aim underlying any imputation is that it provides the corporate entity
with valuable information regarding the external impact of its economic activities on
human welfare.

4.1 Upstream and downstream responsibility

Insofar as we have defined responsibility for pollution incidence, we have
understood this in a direct sense only. Clearly, this will not appeal to everyones
perception of ultimate moral responsibility (or blame) for damage caused. The
principle that the polluter should pay, which is so prevalent in the rhetoric of
environmental policy, does not necessarily explicitly define who the polluter
ultimately is. For example, some firms might argue that responsibility resides
with final consumers. In other words, much of the damage caused can be
attributed to the end of satisfying final demand in other sectors. It follows that
damage could be charged instead to the environmental account of these final
consumers. In contrast, Pearce and Newcombe (1998) describe a complex
notion of responsibility based on sharing of the blame between producers and

1
2
 9
consumers across a given product chain. Whether any one definition of
responsibility is correct in that it will satisfy all criticisms is doubtful. This
could only be achieved if there was a unequivocal social contract to inform
this complex question.

Hence, we do not claim that there is only one way to green corporate accounts.
However, there is a case for attributing damage where it is felt it would do most
to inform key decision-makers. In many instances this will be the entity that
10
directly emits or discharges pollution. Of course, many examples can be found
that do not strictly conform to this interpretation. Hence, it is quite plausible that
a furniture manufacturer may wish to demonstrate accountability to its
customers by documenting that its timber inputs come from sustainably
managed sources even though it is not directly responsible for unsustainable
forestry practices. Many other examples of comparable upstream concern
abound. US EPA (1995) notes that AT&T has acknowledged the role of
consumer pressure in driving its environmental performance targets such as the
reduction of packaging and increased recycling. Demand management schemes
initiated by electricity generators in the US also acknowledge the customer/ firm
linkage in finding least cost responses to achieve environmental targets. In
general, we argue, the primary accounting responsibility is with the firm that has
direct control over the externality arising from its activity and further a firm will
act only to the extent that it gets credit for any reduction of pollution.

The issue of upstream (and downstream) responsibility is clearly both highly

interesting and relevant to our everyday concerns. Indeed, a significant portion
of economic activity does directly generate pollution as such but relies on inputs
that are polluting. Energy use is the most obvious example. Accounting for these
characteristics of environmental change at the corporate level raises new and
challenging measurement issues. How corporate sustainability is affected by
such linkages is a key question for future research.

4.2 Full cost accounting and valuation

In principle, full cost accounting can accommodate a range of activities centred
around the measurement of external costs. There is a growing incentive to firms
to consider the environmental costs of their activities if, for example, the
direction of policy at both the national and the European level continues to
examining these linkages.

10For example, an efficient emission charge would be charged to the emitting entity rather
than its downstream or upstream cohorts. Of course, the firm may be able to pass on some of
these costs. Yet, this would depend, in a technical sense, on the supply and demand curves
that it faced rather than appeals to moral responsibility.
1
3
develop in line with the polluter pays principle. In the case of Ontario Hydro it
was explicitly stated that the anticipation of future policy and enhancement of
future competitiveness were the primary reasons driving its full cost accounting
11
exercise (US EPA, 1996). Indeed, one of the most interesting aspects of this
accounting challenge is the possibility that costs be accounted for using
monetary values along the lines of Ontario Hydro.

By monetary valuation we mean the value in dollar terms of the stream of

benefits that society derives from the environment. Measuring changes in this
stream could provide the linkage that is sought between human welfare and the
environment in discussions of sustainability at the corporate level and beyond.
Of course, these benefits are precisely those which typically are not bought and
sold in markets. In contrast, conventional accounting practice is founded on the
recording of transactions made in the market or, in the case of some items,
imputations based on prices arising from observed transactions.

It is worth noting the increasing tendency for the use of nonmarket valuation
methods to inform the environmental policy-making process. A number of
examples from the United Kingdom illustrate this point. The UK landfill tax
was introduced in 1996 to internalise some of the external costs associated with
the disposal of solid waste; thereby confronting waste generators with the full
costs of disposal (Brisson, 1996). An estimated value of damage caused by
disposal provided the policy target for setting the tax rate: i.e. tax per tonne of
waste disposed (Riley, 1996). The use of damage valuation in target setting is
currently being considered in a proposed aggregates tax. It is likely that the full
or social cost of externalities caused by this sector will be used to calibrate the
resulting tax charged on each tonne of aggregates extracted (ENDS Report 280,
1998). Similarly, the prospect of a tax on pesticide use in the UK may also use
valuation in this way (Pearce, forthcoming).

What additional data is required in order to value external costs? Clearly, this is
a question that relates to the corresponding price with which to weight an
environmentally related quantity such as a tonne of emissions. This can be
illustrated by the simple economics of pollution control (Hamilton, 1996b).
Figure 1 describes marginal damage (MD) and marginal cost of abatement
(MCA) curves associated with various levels of pollution emissions (e). The
vertical axis describes the costs and benefits of emissions reduction in dollar

11
Similarly and for much the same reason, within Europe there has been increasing interest
by the corporate sector in exercises that value the social costs of fuel cycles such as the
European Union funded ExternE work.
1
4
terms while the horizontal axis describes the level of emissions (increasing
levels of emissions from left to right) in physical units (e.g. tonnes).
Figure 1: Marginal Costs of Abatement and Marginal Damages

The MD curve is upward sloping: i.e. each successive unit of emission is

reckoned to be more damaging than the last. MAC, on the other hand, decrease
as pollution emissions increase. In this way, we can judge whether abating an
additional unit of pollution emission confers a net benefit on society. For
example, assume that, in the absence of any restrictions on its emissions, a firm
u
is operating at e . At this point, it is over polluting in that MD is greater than
MCA. Society would achieve a net benefit if the firm reduced its pollution
discharges to the efficient or desirable emission level, e*. From an accounting
u
point of view, for a firm emitting at e , what would be the corresponding
monetary charge to be made in its environmental account?

One response would be to impute the level of abatement costs that the firm
would incur if it actually reduced emissions to e*; i.e. the area corresponding to
u 12
ae e*. This appears to be the position advanced by Gray (1992). Note that this
constitutes an additional step to recording or identifying currently required
compliance costs - i.e. internal environmental expenditures (which are zero in
Figure 1). United Nations (UN, 1993) define maintenance costs as those costs
that would have been incurred in order to restore environmental quality to some

12
Pezzey (1988) notes that this was original spirit of the OECDs polluter pays principle.
1
5
predetermined level. If there was a potential policy intervention, e.g. some form
of regulation, that forced the firm to reduce its emissions to e* then this
imputation could provide valuable information on how policy might affect the
firms bottom-line. Note that this is a different issue to measuring pollution
damage. However, many feel that environmental quality should be improved
such that it is a natural corollary that the cost of pollution is the expenditure
required to take us back to some desirable point. To the extent that our interest is
in the link between emissions and human welfare then our reference point
should be the MD curve. Hence, basing an imputation on abatement costs may
grossly understate this impact on human welfare (Hamilton, 1996b).

The marginal damage curve conveys information about the marginal willingness to
pay for a small (i.e. a marginal or unit) improvement in environmental quality. A firm
u
which is emitting pollution at point e in Figure 1 the implicit pollution charge is the
u u
rectangle cbe 0 or c e . Note that marginal damages at c are an upper bound on the
value at the desirable pollution emissions level t. Only if the marginal damage curve
is horizontal over the relevant range for a firm, i.e. its emissions have no effect on
marginal damages, will t=c. Hence, if the firm were to impute its payment of a
u
hypothetical optimal tax, t, the charge would be t e . Further, if the tax, t, actually
was levied on every unit of pollutant that the firm emitted, it would be in the firms
interests to reduce actual emissions to point e*. In doing so, the firm would it in its
interests to incur compliance costs equivalent to be ue*. In other words, the firm would
pay t e* (i.e. tae*0) in tax. In the companys accounts, the former would be recorded
as an operating costs and the latter as indirect taxes paid.

The valuation question is not simply one of whether to use marginal damages or
marginal abatement costs. A firm might argue that the amount notionally owing to
victims to compensate them for the damage caused is the area under the marginal
u u
damage curve be 0 rather than c e . Again, unless the MD curve is horizontal, the
u
charge c e will overstate the amount that is required to compensate victims. This
still requires the firm to impute a damage estimate for any pollution that it emits,
including those that occur to the left of the desirable level e*. Other definitions of
the firms liabilityimply that a charge should be made only for emissions above
u
e*. In this case the charge should be abe e*. In summary, there is a range of
possible imputations that we could make to accounts based on monetary valuation.
Which we choose depends on the wider environmental policy context (e.g. taxes or
regulation), the interpretation of the full costs that we require (e.g. human welfare
cost or potential cost of compliance faced) and data available (e.g. do we know the
shape of the relevant MD and MCA curves).

1
6
Reticence on the part of corporate decision-makers to divulge environmentally
sensitive information cannot be discounted. Arguably, corporations have little
incentive to reveal environmental data not based on compliance driven
overheads (Bennett and James, 1997) or to reveal bad news in general (see, for
example, Gray, Bebbington and Walters, 1993, which provides evidence that
only 3% of companies report contingent liabilities in their annual company
reports). However, these remarks are generic to any environmental reporting
system regardless of the units in which data, reflecting external impacts, were
presented. The interesting question for the discussion here is the extent to which
placing a monetary value on environmental data raises new and distinct
concerns. We identify two main questions. Firstly, what is the value-addedin
providing indicators in monetary units or, put another way, what additional
information does the move from physical to monetary data yield? Secondly,
how reliable are monetised estimates of environmental damage?

Physical indicators provide important information for decision-making

purposes. For example, a downward trend in physical emissions of CO 2 in the
UK electricity generating sector provides a clear signal of an improvement in
environmental performance. For many pollutants there is no straightforward
relationship between emissions and subsequent damage. If damage doneis our
policy concern then information regarding physical discharges may be a poor
proxy. Furthermore, by moving beyond raw emissions data to quantify the
damage caused by incidence of pollutants this link to damage can be made
explicit. Of course, some of these links are relatively well-known - e.g. acid
rain impacts in the case of SO2. Other aspects of damage are probably less
familiar to a wide audience - the indirect effects of SO 2 as a precursor to the
formation of sulphates implicated in a range of human health problems ranging
from chronic mortality, asthma symptom days or itchy eyedays. Clearly, these
13
impacts are likely to be associated with significant human welfare costs. It is
also the case that in conveying these, and other, links in similar units (e.g.
dollars) priority to be assigned to control of particular pollutants, in terms of

13
Indeed, it may be that the controversial nature of accounting for health related damage, at
least as far as a corporation is concerned, is that this may be interpreted as an admission of de
facto responsibility for that damage. It is unlikely that any commercial entity would wish to
embrace such a framework if it were to cause adverse publicity or, in the extreme, encourage
victimsto pursue litigation. However, many studies now show that pollution related health is
an important, arguably the major, damage category particularly for routine emissions arising
from fuel cycles and transport (see, for example, European Commission, 1995; Maddison et
al. 1996). It is becoming clear that any discussion of externalities caused by many key
pollutants cannot be made without reference to this impact category. Hence, it is of the utmost
importance that further research clarifies the interpretation of any such imputations for health
related externalities in the context of corporate environmental accounting.
1
7
relative impact, can be determined. Moreover, it permits a relatively easy
comparison with other relevant magnitudes in the same units. In this way, the
costs and benefits of pollution abatement can be balanced.
This means that there is a valuable and desirable signalling property in
expressing damage in monetary terms. A tendency noted by Parker (1997) is for
environmental accounting techniques to be viewed as penalising firms for
tackling environmental problems. By incurring abatement expenditures the
company, in turn, incurs increased operating costs and, other things being equal,
reduced profits. Yet, if damages were expressed in comparable monetary units
then these expenditures could be demonstrated in the context of a comparable
decrease in damage done. Moreover, decreases in the value of damage can be
also reported over time. It is true, however, that much of this information can be
communicated with reference to the corresponding physical reduction in
pollution emitted. Valuing this pollution allows this information to be
aggregated. An overall picture of environmental performance can therefore be
communicated. This may be a useful service to certain audiences. Observation
of indicators used in everyday life confirms the status of headline or summary
indicators capable of relatively easy recognition and interpre-tation.. Thus, to
the extent that the welfare cost associated with pollution provides new and
useful information then valuation of pollution provides a meaningful
complement to data on physical emissions.

The main controversy is that most of these values lie outside of the market and thus
cannot be captured using methods familiar to the accounting profession. Hence,
demonstrating that methods used to value pollution exhibit an acceptable level of
reliabilityis clearly crucial. There are a number of aspects to. On one hand, there
are those who argue that we cannot value the environment with reference to the
14
concept of a critical natural asset. Recall that for such assets there is some
(possibly unknown) minimum level of the resource which must be maintained. If
the existence of criticality is pervasive across natural assets then there is no need,
in general, to monetise damage. A physical indication of the assets decline would
15
be enough to inform us that development is unsustainable. Indicators of physical
pressures on the environ-ment, such as emissions, do not measure this asset loss as
such but may provide a practical proxy in the absence of this information. For
example, in the case of ozone layer depletion, production or emissions of CFCs
(clorofluorocarbons) may offer acceptable proxies.

tradedmoral
nave off against
philosophy
other desirable
underlyingobjectives
this see, Pasek
(see, for
andexample,
Beckerman,
Hines,
1996).
1991; for a critic of the

15 Strictly speaking, unsustainability would require that the asset is drawn below its threshold
level.
1
8
Where there exists a trade-off between say, an environmental asset and other
goals valuation provides a potentially useful means to express this. The
reliability of the methods used for elicit this information is a source of much
discussion. In the US National Oceanic and Atmospheric Administration
(NOAA) concluded that the primary methodology for damage valuation (the
contingent valuation method) was reliable enough to be used in the judicial
process of natural resource damage assessment (Arrow et al., 1993). Such
studies, beginning with the litigation following the Exxon Valdez oil spill, are
increasingly being used as the basis to settle damage claims (Carson et al.,
1992). Nevertheless, many would argue that societys preferences for
environmental goods cannot be boiled down easily to a single money value
expressing willingness to pay or that existing techniques to do this are simply
not up to the task (see, for example, Hausman and Diamond, 1994).

In the case of routine emissions of pollution, the valuation task is additionally

complicated. The calculation of pollution emissions is only one required input for
downstream models of dispersion and impact. This results in a physical end-point
reflecting impact for example on health or living resources. It is only then once
these end-points have been evaluated adequately in physical terms that valuation
becomes possible. This is the basis of the approach taken by Ontario Hydro (US
EPA, 1996). The end-product of such studies is an unit value of damage caused per
tonne of pollution emitted. This may be a useful addition to the measurement of
direct emissions from relatively large corporate entities. It should be noted that
where pollution is largely accounted for by diffuse or non-point sources or where
interest is in the impact of upstream polluting activity (e.g. minimising polluting
inputs to production) in the absence of good data it is likely that physical proxies
for such impacts will have to be sought.

Reliability and usefulness have to be evaluated in the light of the real resource
cost required to implement this type of full cost accounting framework. Even for
significant direct emitters of pollution, it is likely that valuing pollution will
only be useful to the extent that use can be made of existing results from
dispersion-impact models. The idea underlying this is to take an estimate of the
unit value of the impacts of an emission source of a pollutant (e.g. SO 2) in a
different location and (perhaps after some adjustment) use this as an approxi-
mation of the unit value of the impacts of emissions (e.g. SO 2) from another
source. For example, there now exist numerous estimates of damage per tonne
for various pollutants (European Commission, 1995, forthcoming; Smith, 1996).
However, using the best of these estimates as the basis of prices with which to
value a given firms pollution requires careful evaluation.

1
9
The question is whether it is a legitimate procedure to transfer these values in
this way. In addition to the generic issue of whether damage can be valued at all,
there is a question of coverage of impact categories. Secondly, there are
technical difficulties associated with location issues. For example, the damage
caused by pollutants such as sulphur dioxide tends to vary depending on the
location of the emission source. This is, in turn, due to factors such as
meteorology and differences in population at risk (Krupnick and Burtraw,
16
1996). If any of these factors have a significant bearing on a unit value, it may
not be valid to transfer this value across emission source in wholly different
locations. Fortunately, some progress is being made in our understanding of this
problem by the extension of existing dispersion models to consider multiple
sources at different locations within and across countries. As a result, unit
damages for pollutants such as sulphur dioxide, nitrogen oxides and particulate
matter will emerge that take better account of these location issues (see, for
example, European Commission, forthcoming).

To summarise the above discussion, a number of problems or objectives to the

use of damage valuation in corporate environmental accounting studies can be
raised. These issues embody a range of issues regarding the reliability of
measurement of environment related prices and quantities. There is little to be
gained by glossing over such concerns. In particular, we have argued that
translating the theory of placing monetary values on corporate-specific
environmental data into practice will depend on research that further
investigates these anxieties. With this caveat in mind, we provide an illustration
of how a full cost account for various examples of economic activities from the
corporate sector in the next section.

16
For global pollutants such as carbon dioxide these location problems do not exist
marginal damage is the same no matter where the unit of pollutant was emitted. Of course,
there is significant uncertainty regarding the impacts of global warming arising as a result of
the accumulation of greenhouse gases in the atmosphere.
2
0
5. Valuation and Full Cost Accounting - Selected Examples

We wish to estimate the value of air pollution attributed to corporate entities

caused by five air pollutants. These are carbon dioxide (CO 2), methane (CH4)
sulphur dioxide (SO2), nitrogen oxides (NOX) and particulate matter (PM10). A
breakdown by sector of original of physical data relating to emissions of these
pollutants are beginning to emerge. Much of the impetus for this can be traced
to the on-going construction of detailed environmental accounts by national and
international statistical offices (see, for example, Office for National Statistics,
1998). Examples of similar data at the company level can also be found in some
environmental reports.

Estimates of unit damages per tonne of pollutant emitted are shown in Table 1.
These are drawn from an evaluation of the social cost of fuel cycles within
Europe (European Commission 1995). The CO2 and CH4 estimates are from
Fankhauser (1994). The rows of Table 1 represent receiving agents (broadly
conceived). These are the ultimate effects of polluting activities on human
health (e.g. respiratory problems), and non-health (forest damage, material and
buildings damage). The corresponding unit values indicate the damage done by
a tonne of pollutant vis--vis its impact on each receiving agent. For example,
for human health, the per unit values in Table 1 reflect both mortality and
morbidity effects per tonne of pollutant emitted. The former are based on studies
that estimate the willingness to pay for changes in risk of death while the latter
from studies of willingness to pay for reductions in episodes of illness.

Table 1: Marginal Damage Per Tonne of Pollutant Emitted ()

SO2 NOX PM10 CO CH4

2

Health 19500 - 44 800

Non-Health 300 - 670 200 - 280
Total 300 - 670 200 - 280 19500 - 44 800 12 80

Source: European Commission (1995); Pearce and Newcombe (1998); Fankhauser (1994)

Pollutants such as SO2, NOX and PM10 are implicated in significant damage to
health. However, these impacts are not simply additive as recent epidemiological
evidence appears to suggest that SO2 and NO X impacts occur via the formation
2
1
of sulphate and nitrate aerosols in the atmosphere (Pearce and Newcombe,
1998). Given that PM (which is itself a mixture of sulphate, nitrate and other
aerosols) adding impacts together risks double-counting. In view of this, Table 1
only attributes damage to health to PM10. The value of damage caused by SO 2
and NOX is necessarily restricted here to those direct effects associated with
non-health categories. Climate change, arising from emissions of CO 2 and CH4,
is likely to be characterised by a range of health and non-health impacts, mainly
resulting from the effects of sea-level rise. Table 1 reports a conservative
estimate, from Fankhauser (1994), of the unit damages, caused by the release
into atmosphere, for CO2 and CH4.

These valuation data can be combined with sector specific data of physical
emissions of pollutants to illustrate two indicators of environmental performance.
The first is the value of environment damage attributable to the activities of the UK
corporate sector. The second is a green measure of economic performance
comparable to conventional indicators of a sectors performance. A candidate
indicator is an adjusted or green value-added analogous to that constructed by
BSO/Origin (Huizing and Dekker, 1992). This indicator could signal how much
value-added a sector has generated after imputations reflecting environmental
17
damage have been charged. Table 2 combines published data on physical
emissions in ONS (1998) with the (central estimates of) monetary damage in Table
1. Green value-added is estimated for 10 sectors, in the years 1987 and 1994, by
18
deducting the value of environmental damage (row 3) from value-added (row 2).

more appropriate indicator than value-added.

18 We have left to one side the issue of how a firm should account for its use of finite
natural resources; e.g. fossil fuels used as energy inputs to production. To the extent that
energy sources are substitutable for one another, the only limiting factor is the costs of
switching to less scarce sources. For example, there has been a long-term decline in the real
international price of oil of some 1.5% per annum largely as a result of resource saving
technological innovation (Nordhaus, 1992).

2
2
23
Table 2. Environmental Damage and Green Value-Added in the UK Corpora
unless stated)

Agriculture Chemicals Electricity Const

1987 1994 1987 1994 1987 1994 1987
Value Added (VA) 5600 8810 10320 12870 5600 6830 29400
Env. Damage 865 462 228 173 2742 1955 688
% VA 15.4 5.2 2.2 1.3 49.0 28.6 2.3
Green VA 4735 8348 10092 12697 2858 4875 28712

Food Processing Air Transport Building Materials Post & T

1987 1994 1987 1994 1987 1994 1987
Value Added (VA) 14370 15530 2170 3620 4010 3890 12290
Env. Damage 241 241 212 296 341 210 4
(% VA) 1.7 1.6 9.8 8.2 8.5 5.4 0.0
Green VA 14129 15289 1958 3324 3669 3680 12286

Source: Office for National Statistics (1998); see also sources listed below Table 1.

2
4
Table 2 indicates that the level of damage caused varies significantly by sector.
Moreover, there is also significant variation in damage in proportion to sectoral
value-added. Not surprisingly, those sectors which are relatively damage-intensive
per unit of value-added include electricity, agriculture, air transport and building
materials. However, there is a marked decrease in environmental damage in some
sectors between the years 1987 and 1994. This trend is particularly notable in those
sectors where the estimate of damage in 1987 was relatively high. For example, the
result for the electricity sector is particularly striking. It appears to indicate that
damage decreased from almost 50% of the sectors value-added in 1987 to under
30% in 1994. This, in turn, is attributable to decreases in emissions of all major air
pollutants (except CH4). In the agricultural sector, much of the reduction in
environmental damage is caused by reductions in PM10 emissions. Emissions of
CO2 and SO2 from agricultural sources, in contrast, have increased while CH 4
releases have stayed roughly constant. Exceptions to this trend are air transport and
post & telecommuni-cations sectors.

An interesting extension would be a comparable analysis at the company level. At

present this may prove difficult due to either the absence of the prerequisite physical
data or the fact that few companies currently publish value-added (Pearce and
19
Newcombe, 1998). As regards the latter, there is no unequivocal reason why a green
indicator of economic performance need be based on value-added. Indeed, from the
point of view of commenting on the sustainability of a companys activities it may be
desirable to look for an alternative indicator. The analogy that we have in mind here is
an indicator that has proved to be useful at the macro-accounting level. This extended
net savings rate - genuine saving - was discussed earlier.

Any individual firm in generating its economic output incurs operating costs in
doing so. It is the value of resources left after these costs are deducted that
corresponds to potentially available resources for investment and saving.
Furthermore, a crude upper bound on this magnitude is the firms net profit (before
tax). The extension is that if, within an environmental of full cost account,
environmental damage is chargedagainst its net profits then a suitable indicator of
the company green economic performance is an adjusted or green profit rate.
Thus, other things being equal, the greater the value of environmental damage that
is charged to the firms profit the less resources there are, at least notionally, to
invest in new assets held by the firm or to distribute in the form of dividends to the
firms owners. It is in this sense that the firm is less sustainable as a result of the
20
pollution damage for which it is responsible.
19
Although, presumably this could be estimated by re-organising existing accounting data.

20 It should be noted that given that households are owners of businesses, the magnitude of a
business retained earnings will influence household saving decisions (and thereby the level of
2
5
We illustrate this using data available in the public domain for one of the UKs
largest electricity generating companies. In 1990, the company emitted
approximately 12% and 26% of the UKs total emissions of CO 2 and SO2
respectively. Nevertheless, considerable efforts are being made to reduce these
emissions within the company and the electricity sector in general. Specifically,
we are interested in how this improvement in environmental performance can be
described by comparisons of damage for an individual company over time. In
turn, this information is combined with conventional accounting data for this
same company in order to estimate an indicator of the companys green profits.
These values are estimated using physical data contained in various environ-
mental reports published by the company and, on a conservative assumption of
damage caused, using the lower-bound estimates of unit damages in Table 1.

Table 3 sets out this hypothetical environmental account for an electricity

generator over the period 1992 to 1996. Damage caused by all four pollutants
shown has declined over the period. In particular, damage caused by SO2
emissions has fallen significantly reflecting the companys move from coal-fired
electricity generation to gas-fired technologies. Nevertheless, according to the
bottom-line green profit indicator, the sum of the damage caused exceeds
profits in the years 1992 through to 1995. In other words, on our definition of
sustainability, the company appears to have been behaving unsustainably during
this period.

The signals provided by this indicator need to be interpreted with care. For
example, there is an apparent implication that companies can compensate for
increased pollution with higher profits. If we recall Figure 1 suggests that may
be situations whereby $1 of output can be generated at a social cost of less than
$1 of pollution damage. In situations such as these, where the economy is
21
operating to the left of e*, there is a net social gain to increasing output. How-

Table 3. Accounting for Pollution Damage: The Case of Electricity

Generation 1992-1996 ( millions, current prices)

personal saving) (Kauffman, 1993). In other words, changes in corporate saving and
household saving will move in opposite directions so that a change in the latter offsets the
former. If this offsetting process is entire then, other things equal, sustainability in the
aggregate is unaffected by the firms saving decision.
21
There remain outstanding distributional issues whereby a net gain is attained but where
pollution victims are left uncompensated.
2
6
 1992 1993 1994 1995 1996
1. Turnover 3097 3188 2932 2885 2933
- Operating costs 2771 2736 2455 2354 2240
= Operating profit 326 452 477 531 693
+ Other income 33 -27 8 14 -6
2. = Profit on ordinary activities 359 425 485 545 687
CO2 damage 256 227 229 219 212
+ SO2 damage 279 248 218 198 162
+ NOX damage 43 37 35 32 27
+ PM10 damage 491 338 311 264 215
3. = Pollution damage 1068 850 793 713 616
as % of profit 328 188 166 134 89
4. Adjusted profit (2.-3.) -709 -425 -308 -168 71

Source: Published Annual Reports and Environmental Reports of Company.

ever, if the economy is over-polluting, i.e. it is to the right of e*, the opposite
conclusion applies. Increasing output will lead to a net social loss and it is
reductions in pollution that will lead to net gains. It is crucial that a green
corporate indicator is able to provide this type of policy signal. In turn, this
places a greater onus on the reliability of damage valuation. Note that indicators
of physical emissions do not tell us anything about the nature of the trade-off
between pollution and output. These conclusions, of course, apply to situations
where standard cost-benefit rules are accepted. There could exist situations
where this is not the case such as for critical assets. In such events, society may
decide that some other decision-making rule is appropriate such as a safe
22
minimum standard (Farmer and Randall, 1998). This, in turn, would mean that
some new and additional indicator of progress would be required.

The term behaving unsustainably refers to a notional sustainability. The indicator

that we present is calculated as if the pollution damage magnitudes in Table 3 were
actually charged. Unless it was credible that these sums would be charged in reality
the sustainability of which we speak is of a hypothetical nature. Two points are
worth noting in this respect. Firstly, it is clear that the unit values that we use can
be disputed. Some impact categories could not be considered.
22
A safe minimum standard is defined by Farmer and Randall as representing the suspension
of standard arrangements for managing a given resource in order to prevent irreversible
outcomes providing that this policy switch itself does not incur intolerable costs.
2
7
There is ample scope to improve existing estimates as new data -based on better
knowledge of the pollution-human welfare link - become available. Secondly, it
may not be plausible that actual public policies to address excessive emissions
of pollutants will require that compensationtake this precise form.

Nevertheless, we would still argue the values in Table 3 offer a first

approximation of the welfare cost of polluting activities and that the concept of
corporate sustainability that we offer, however notional it may be, results in a
useful indicator of corporate environmental performance. Put another way, it
offers a headline indicator of environmental responsibility that will be
understood by a potentially wide audience. Finally, Table 3 indicates that the
companys green profits become positive in 1996. What this provides is an
overall signal that the company has made significant progress has been made in
reducing the environmental impacts of its economic activities. In other words,
this framework can be used to signal good newsto the extent that it exists.

2
8
6. Discussion and Conclusions

We have proposed that the notion of a sustainable businessdoes have meaning

beyond the provision of a unifying standard for diverse questions concerning the
corporate sector and the environment. This draws heavily on analysis familiar
from the green national accounting literature. It suggests that to the extent that
external costs or damages are directly associated with the generation of a
corporate entitys income, the environmental account of that entity should
reflect the magnitude of the damage caused. On our definition, a business is less
sustainable the greater the magnitude of that damage other things being equal.

Further, we have illustrated how these values can be incorporated into

conventional corporate accounting indicators such as company profit to derive a
green economic performance indicator. These are comparable to measures of net
(or genuine) saving that have been estimated at the level of the national
economy. An intuitive interpretation is that a business is, at least notionally,
unsustainable to the extent that this corporate net saving rate or green profit is
less than zero. When judged over time this indicator provides the desirable
signal that as the firms impact on the external environment diminishes the more
sustainablethat firm is indicated to be. This, we argue, provides one reasonable
link between an understanding of the sustainable business in terms of the
corporate entity itself and in the contribution of the entity to sustainability in a
wider sense. It should be noted, however, that defining corporate sustainability
by measuring the latter in terms of statements such as a given firm, in isolation,
decreases national sustainability by x% is a far less tractable measurement
problem.

We have illustrated our argument with reference to corporate green accounting

for air pollution and argued, that while there remain outstanding non-market
valuation concerns, this extension to environmental accounting is worthy of
further investigation. The appropriate treatment of living resources such as
forests and biodiversity is more problematic. The main concern is that living
resources provide many important, perhaps irreplaceable, non-marketed values.
In principle, the damage to these resources arising from corporate activity could
be handled within our framework. Yet, many of the ecological functions giving
rise to these values are poorly understood and so in practice crucial but
unresolved measurement issues are encountered. It should be noted, however,
that this would be true of any approach to corporate green accounting that
attempted to tackle these same concerns.

Of course, many questions concerning the application of valuation procedures to

this domain are yet to be resolved. Some of these are technical issues while other
2
9
questions are of a more qualitative nature. Many are a balance of the two. For
example, the controversial nature of accounting for health related damage, at least
as far as a corporation is concerned, is that this may be interpreted as an admission
of de facto responsibility for that damage. It is unlikely that any commercial entity
would wish to embrace such a framework if it were to cause adverse publicity or, in
the extreme, encourage victims - or their represent-atives - to pursue litigation.
However, many studies now show that pollution-related health damage is an
important, arguably the major, damage category particularly for routine emissions
arising from fuel cycles and transport (see, for example, European Commission,
1995; Maddison et al., 1996). It is becom-ing clear that any discussion of
externalities caused by many key pollutants cannot be made without reference to
this impact category. It is of the utmost importance that further research clarifies
the interpretation of any such imput-ations for health related externalities in the
context of corporate environmental accounting.

Other issues surround problems in establishing values of water pollution and

ecosystems in a systematic way. Lastly, corporate sustainability has been
defined narrowly here. It would be interesting to explore how downstream and
upstream effects can be dealt with using this framework and how this approach
can be extended in order to relevant to accounting for non-point sources of
pollution. What we have illustrated here is a starting point or building block
whereby a number of distinct but important issues can be addressed in a
consistent fashion. Assuming suitable progress can be made in addressing the
questions raised above, we have shown how monetary data reflecting the impact
of the corporate sector could be a practical green accounting tool. Many firms
are beginning the process of compiling environmental reports. For example, in
the United Kingdom, the electricity generating companies produce reports that
indicate progress made on reducing physical emissions of pollutants. Arguably
the usefulness of these indicators will be enhanced if economy-environmental
linkages are further explored (e.g. decoupling), criteria for selecting key
indicators established and uses such as bench-markingdeveloped.

Suitable use of monetary data on the environment could complement physical

data in several ways. Firstly, these data convey important information concern-
ing what is economic, and sometimes what is sustainable, and which embody
an evaluation of the human welfare cost of, for example, pollution. This in turn
can lead to improved decision-making. Secondly, these data can be consistently
aggregated such that an overall picture can be distilled. This, for example, has
an important communicative function where audiences find it difficult to
assimilate voluminous physical data in heterogeneous units.

3
0
References

Arrow, K. Solow, R. Portney, P.R. Leamer, E.E. Radner, R. and Schuman, H. (1993) Report
of the NOAA Panel on Contingent Valuation, Federal Register, January 15, 58(10):
4601-4614.

Atkinson, G. Dubourg, W.R. Hamilton, K. Munasinghe, M. Pearce D.W. and Young, C.E.F.
(1997) Measuring Sustainable Development: Macroeconomics and Environment,
Edward Elgar, Cheltenham.

Bartolomeo, M. (1996) About the Usefulness of Environmental Performance Evaluation,

FEEM Newsletter, February, 1996, pp 4-15.

Bebbington, J. and Gray, R. (1997) Sustainable Development and Accounting: Incentives

and Disincentives for the Adoption of Sustainability by Transnational Corporations, in
Hibbit, C. and Blokdjik, H. (eds.), Environmental Accounting and Sustainable
Development: The Final Report, Limperg Instituut, Amsterdam.

Bennett, M. and James, P. (1997) Environment-related Management Accounting in North

America and its Implication for UK Companies, in Hibbit, C. and Blokdjik, H. (eds.),
Environmental Accounting and Sustainable Development: The Final Report, Limperg
Instituut, Amsterdam.

Brisson, I. (1996) The UK Landfill Tax, paper presented at the workshop Environmental
Policy Targets and Non-market Valuation, Oslo, September 2nd to 3rd, 1996.

Cairncross, F. (1995) Green Inc. Earthscan, London.

Carson, R.T. Mitchell, R.C. Hanemann, W.M. Kopp, R.J. Presser, S. and Ruud, P.A. (1992) A
Contingent Valuation Study of Lost Passive Use Values Resulting from the Exxon Valdez
Oil Spill, report to the Attorney General of the State of Alaska, NRDA Inc. La Jolla.

Department of the Environment (1996) Sustainability Indicators for the United Kingdom,
HMSO, London.

Devlin, R.A. and Grafton, R.Q. (1997) Economic Rights and Environmental Wrongs:
Property for the Common Good, Edward Elgar, Cheltenham.

Diamond, P.A. and Hausman, J.A. (1994) Contingent Valuation - Is Some Number Better
Than No Number?, Journal Of Economic Perspectives, 8(4): 45-64.

ENDS Report (1998) Aggregate Tax Takes Shape, No. 280, May 1998.

European Commission (forthcoming) Maintenance, Improvement, Extension And Application

Of The Externe Accounting Framework, Directorate-General XII, European
Commission, Brussels.

European Commission (1995) Externalities of the Fuel Cycles: "ExternE Project,

Directorate-General XII, European Commission, Brussels.
3
1
Fankhauser, S. (1994) Evaluating the Social Costs of Greenhouse Gas Emissions The
Energy Journal, 15(2): 157-184.

Farmer, M.C. and Randall, A. (1998) The Rationality of a Safe Minimum Standard, Land
Economics, 74(3): 287-302.

Gray, R. Bebbington, J. and Walters, D. (1993) Accounting for the Environment, Paul
Chapman Publishing, London.

Hamilton, K. (1997) Preservation and Strong Sustainability, Environment and

Development Economics, 2(1): 72-76.

Hamilton, K. (1996a) Pollution and Pollution Abatement in the National Accounts, Review
of Income and Wealth, 42 (1): 13-33.

Hamilton, K. (1996b) Valuing Pollution in the National Accounts: Marginal Damages or

Maintenance Costs?, presented for the London Group on National Accounts and the
Environment, Stockholm, May 28-31th, 1996.

Hamilton, K. and Atkinson, G. (1996) Air Pollution and Green Accounts, Energy Policy,
24(7): 675-684.

Hartwick, J.M. (1993) Notes on Economic Depreciation of Natural Resource Stocks and
National Accounting, in Franz, A., and Stahmer, C., (eds) Approaches to
Environmental Accounting, Physica-Verlag, Heidelberg.

Hawken, P. (1993) The Ecology of Commerce, Harper Collins, New York.

Hines, R. (1991) On Valuing Nature, Accounting, Auditing & Accountability Journal, 4(3):
27-29.

Holmes, G.A. and Sugden, A. (1997) Interpreting Company Reports and Accounts, 6th
edition, Woodhead-Faulkner, New York.

Huizing, A. and Dekker, H.C. (1992) Helping to Pull Our Planet Out of the Red: An
Environmental Report of BSO/Origin, Accounting, Organisations and Society, 17(5):
449-458.

Kauffman, B. (1993) Microeconomic Aspects of Saving, in Heerje, A. (ed.) World Savings,

Basil Blackwell, Oxford.

de Koning-Martens, W. and van der Ende, N. (1997) Bridge Over Troubled Waters or
Chinese Walls: What Lessons can be Drawn from the EMAA Workshop?, in Hibbit,
C. and Blokdjik, H. (eds.), Environmental Accounting and Sustainable Development:
The Final Report, Limperg Instituut, Amsterdam.

Krupnick, A. and Burtraw, D. (1996) The Social Costs of Electricity: Do the Numbers Add
Up?, Resource and Energy Economics, 18(4): 423-466.
3
2
Maddison, D. Pearce, D.W. Johansson, P-O. Calthrop, E. Litman, T. and Verhoef, E. (1996)
Blueprint 5: The True Costs of Road Transport, Earthscan, London.

Mler, K.-G. (1991) National Accounts and Environmental Resources, Environmental and
Resource Economics, 1: 1-15.

Mathews, M.R. (1997) Towards a Mega-Theory of Accounting, Asia-Pacific Journal of

Accounting, 4(2): 273-289.

National Power (1997) Environmental Performance Review 1997.

Nordhaus, W.D. (1992) Lethal Model II: The Limits to Growth Revisited, Brookings
Papers on Economic Activity, 2: 1-43.

OECD (1993) Environmental Indicators, Organisation for Economic Cooperation and

Development, Paris.

Office for National Statistics (1998) UK Environmental Accounts 1998, HMSO, London.

Owen, D.L. (1997) A Critical Perspective on the Development of European Corporate

Environmental Accounting and Reporting, in Hibbit, C. and Blokdjik, H. (eds.),
Environmental Accounting and Sustainable Development: The Final Report, Limperg
Instituut, Amsterdam.

Owen, D. Gray, R. and Bebbington, J. (1997) Green Accounting: Cosmetic Irrelevance or

Radical Agenda for Change, Asia-Pacific Journal of Accounting, 4(2): 175-198.

Parker, L.D. (1997) Accounting for Environmental Strategy: Cost Management, Control and
Performance Evaluation, Asia-Pacific Journal of Accounting, 4(2): 145-174.

Pasek, J. and Beckerman, W. (1996) Plural Values and Environmental Valuation, CSERGE
Working Paper GEC 96-11, Centre for Social and Economic Research on the Global
Environment (CSERGE), University College London and University of East Anglia.

Pearce, D.W. (forthcoming) Cost-benefit Analysis and Environmental Policy, Oxford

Review of Economic Policy, 14(4).

Pearce, D.W. Hamilton, K. and Atkinson, G. (1996) Measuring Sustainable Development:

Progress on Indicators, Environment and Development Economics, 1(1): 85-101.

Pearce, D.W., Markandya, A. and Barbier, E. (1989) Blueprint for a Green Economy,
Earthscan, London.

Pearce, D.W. and Newcombe, J. (1998) Corporate Sustainability: Concepts and Measures,
CSERGE, mimeo.

Pezzey, J. (1992) Sustainability, Environmental Values, 1(4): 321-362.

3
3
Pezzey, J. (1989) Economic Analysis of Sustainable Growth and Sustainable Development,
Environment Department Working Paper No. 15, World Bank, Washington D.C.

Pezzey, J. (1988) Market Mechanisms of Pollution Control, in Turner, R.K. (ed.)

Sustainable Environmental Management: Principles and Practice, Belhaven Press,
London.

PowerGen (1996) Environmental Performance Report 1996.

PowerGen (various) Report and Accounts.

Proops, J.L.R. Faber, M. and Wagenhals, G. (1993) Reducing CO2 Emissions, Springer-
Verlag, Berlin.

Riley, C. (1996) The UK Landfill Tax, European Social Foundation, Dublin, January 1996.

Segerson, K. (1994) Liabilities and Penalties in Policy Design, in Bromley, D. (ed.)

Handbook of Environmental Economics, Basil Blackwell, Oxford.

Smith, V.K. (ed.) (1997) Environmental Costing for Electricity: Theory, Experience and
Prospects, Resource and Energy Economics, 18(4): 333-346.

Tietenberg, T. (1996) Environmental and Natural Resource Economics, 4th edition, Harper-
Collins, New York.

Tuppen, C. (ed.) (1996) Environmental Accounting in Industry: A Practical Review, British

Telecommunications, London.

United Nations (1997) Environmental Financial Accounting Guidelines, report prepared for
the UNCTAD Intergovernmental Working Group of Experts on International Standards
of Accounting and Reporting (ISAR).

United Nations (1993) Integrated Environmental and Economic Accounting. Series F No. 61,
United Nations, New York.

US Environmental Protection Agency (EPA) (1996) Full Cost Accounting for Decision
Making at Ontario Hydro: A Case Study, Environmental Protection Agency,
Washington D.C.

US Environmental Protection Agency (EPA) (1995) Green Accounting at AT&T,

Environmental Protection Agency, Washington D.C.

Whitby, M. and Adger, W.N. (1996) Natural and Reproducible Capital and the Sustainability of
Land Use in the UK, Journal of Agricultural Economics, 47(1): 50-65.

World Bank (1997) Expanding the Measure of Wealth, The World Bank, Washington, D.C.

3
4
Writh, M. (1997) Environmental Reporting, in Hibbit, C. and Blokdjik, H. (eds),
Environmental Accounting and Sustainable Development: The Final Report, Limperg
Instituut, Amsterdam.

3
5