jeudi 20 mars 2008


17 Environmental and resource accounting Glenn-Marie Lange

1. Overview of environmental accounts
Sustainable development is the stated objective of many countries and the
search for operationalizing this concept has focused in part on the system
of national income accounts (SNA) (UN et al., 1993). The SNA is crucial
because it constitutes the primary source of information about the econ-
omy and is widely used for assessment of economic performance and policy
analysis throughout the world. However, the SNA has a number of well-
known shortcomings regarding the treatment of the environment. For
example, while the income from extracting minerals is recorded in the
national accounts, the simultaneous depletion of mineral reserves is not.
Uncultivated fisheries and forests receive similar treatment. This can result
in quite misleading economic signals about sustainable national income.
Indeed, one of the primary motivations for the early environmental
accounting efforts in the mid-1980s was concern that rapid economic
growth in some developing countries was achieved through liquidation of
natural capital, a practice that appears to boost GDP in the short run, but
is not sustainable in the long run.
Equally important, ecosystems provide non-marketed goods and ser-
vices that are often not fully included in national accounts, or are wrongly
attributed to other sectors of the economy. For example, the harvest for
own use of firewood and wild foods, so critical to livelihoods in many devel-
oping countries, is often underestimated. Forests provide recreation and
tourism services, which are not attributed to the forest industry. Forests
may also provide watershed protection benefiting agriculture, hydroelectric
power, municipal water supply and so on, but the value of these services is
not recognized and, hence, not attributed to the forestry sector. Thus the
total benefits from sustainable forestry are underestimated, and other
sectors of the economy are not fully aware of their dependence on the
health of this natural resource.
Over the past few decades, many natural scientists and social scientists
have worked to develop environmental accounts as a tool to promote sus-
tainable development. This effort resulted first in the publication of an
interim handbook in 1993, the System of Environmental and Economic
Accounting (SEEA), under the aegis of the UN’s Statistical Commission
(UN, 1993), followed by a substantially revised and expanded SEEA
Handbook of sustainable development
Handbook in 2003 based on more than a decade of additional conceptual
work and empirical applications by national and international agencies,
academics and NGOs (UN et al., 2003).
The SEEA provides a comprehensive and broadly accepted framework
for incorporating the role of the environment and natural capital in the
economy through a system of satellite accounts to the SNA. As satellite
accounts, the SEEA has a similar structure to the SNA, consisting of both
stocks and flows of environmental goods and services. The SEEA has four
major components, which are constructed, wherever possible, in both phys-
ical and monetary units:

Asset accounts, which record the volume and economic value of
stocks and changes in stocks of natural resources.
Flow accounts for materials, energy and pollution, which provide
information at the industry level about the use of energy and mater-
ials as inputs to production and final demand, and the generation of
pollutants and solid waste. The flow accounts also make explicit the
input of non-market environmental services to production and final
consumption that may be implicitly included in the production
values of other sectors.
Environmental protection and resource management expenditure
accounts, and other environmentally related transactions. These
accounts reorganize information already in the SNA to make more
explicit 1) expenditures incurred to protect the environment and
manage natural resources and 2) taxes, fees and other charges, and
property rights related to the environment.
Environmentally-adjusted indicators of macroeconomic performance,
which include indicators of sustainability such as environmentally-
adjusted GDP and NDP, Adjusted Net Savings (genuine savings: see
Chapter 18), and a broader measure of national wealth that includes
natural capital in addition to manufactured capital
Environmental accounts are now constructed regularly by many
developed countries and some developing countries (Table 17.1). Of
course, environmental accounts are a broad undertaking and countries
have implemented them on an incremental basis, compiling the parts of
the accounts that are most useful for their environmental priorities.
Environmental accounts improve policy making by providing aggregate
indicators for monitoring environmental–economic performance, as
well as a detailed set of statistics to guide resource managers toward
policy decisions that will improve environmental–economic performance
in the future.
This chapter describes some of the policy applications for each compon-
ent of the environmental accounts; a more detailed review of applications
can be found in (Lange, 2003a; 2004a; Lange et al., 2003; World Bank,
forthcoming 2005). For technical aspects of the environmental accounts,
the reader is referred to (UN et al., 2003).
2. Asset accounts: monitoring total wealth
Theoretical work (by for example Arrow et al., 2003a; Dasgupta and
Mäler, 2000; Heal and Kristrom, 2005; Kunte et al., 1998; see also
Chapter 18) has demonstrated that sustainable development requires
non-declining per capita wealth, where wealth is defined in the broadest
sense to include produced, natural and human (including social) capital.
This implies that economic development can be viewed as a process of
‘portfolio management’ seeking to optimize the management of each
asset and the distribution of wealth among different kinds of assets
(World Bank, 2002). The particular challenge for resource-rich economies
is to transform natural capital into other forms of productive wealth, a
process that requires good policy in three critical areas: 1) promotion of
efficient resource extraction that maximizes resource rent, 2) recovery of
the rent by an agency capable of reinvesting rent and 3) efficient reinvest-
ment of rent.
Environmental accounts provide information to monitor sustainable
development by measuring total wealth (produced + natural capital)1 over
time, which indicates whether depletion of resources is compensated for by
investment in other assets; for example is development sustainable or not?
The environmental accounts also provide more detailed information to
assess the environmental and natural resource policies guiding this process:
the amount of resource rent being generated from each resource, the
amount of rent recovered by various agencies, and the share of that rent, if
any, that is invested in other assets.
The SEEA asset accounts in physical units provide indicators of ecolog-
ical sustainability and information for resource management. The volume
of mineral reserves, for example, is needed to plan extraction paths and
indicates how long a country can rely on its minerals. A more complete
assessment of sustainability requires calculation of the monetary value of a
resource stock as well. From the monetary accounts, trends in per capita
national wealth – a measure of sustainable development – can be derived.
These trends can also be analyzed to assess characteristics important to
economic development, such as the diversity of wealth, ownership distribu-
tion, and volatility due to price fluctuations, an important feature for
economies dependent on primary commodities (see Lange, 2003a, for a
discussion of this issue and some examples).
Environmental and resource accounting
Among the developed countries, only Australia (ABS, 2004a) and
Canada (Smith and Simard, 2001) regularly include natural capital in the
balance sheets of their annual national income accounts, and a number of
other countries calculate the value of some assets, particularly subsoil
assets. In the developing countries, figures for total wealth including natural
capital have been compiled for Botswana and Namibia (Lange, 2004b), and
are shown in Table 17.2 and Figure 17.1.2
Both Botswana and Namibia have significant natural capital: diamond
mining accounts for roughly a third of Botswana’s GDP; mining and
fishing account for over 20 per cent of Namibia’s GDP. But only Botswana
has been successful in using its natural capital to increase national wealth,
pushing it into the ranks of upper middle-income countries. Namibia has
not used its natural capital to build wealth.
The rapid growth of national wealth in Botswana is consistent with its
development policy, which set a goal of improving living standards and
reducing poverty based on investment of mineral revenues (see Lange and
Wright, 2004). Botswana has recovered much of the resource rent gener-
ated by its minerals and has consistently reinvested virtually all of it (see
below). Namibia, occupied by South Africa until 1990, has not based its
development strategy on reinvestment of resource rent. Namibia has not
recovered as much of the resource rent, partly due to external factors such
as the lack of control over its marine fisheries before 1990, but partly due
to domestic policy decisions even after independence. Not surprisingly,
Namibia has failed to build national wealth. The effect is significant:
Botswana’s per capita, real GDP has grown at an annual rate of 5 per cent,
while Namibia’s per capita, real GDP has stagnated, declining at an annual
rate of –0.025 per cent (Lange, 2004b).
Table 17.2 shows the breakdown of wealth by asset type. For a small
country like Botswana with limited capacity to absorb capital quickly, the
importance of net foreign financial assets has been particularly important.
Recovery of resource rent and reinvesting it in alternative assets is key to
sustainable development. Regarding recovery of resource rent, Botswana
has been rather successful, recovering on average 76 per cent of rent.
Namibia has had much more volatile rent in both mining and fishing. The
Namibian mining industry, dominated by diamonds, uranium and gold,
has paid on average at least 50 per cent of the rent in taxes. By contrast,
government has not recovered much rent from fisheries, partly because rent
taxes (fishing quota levies) were set rather low and not adjusted for infla-
tion, but also because of poor enforcement of rent collection.
As long as fisheries are not being depleted, recovery and reinvestment of
resources is not necessary for sustainable development. When managed
sustainably, fisheries will continue to generate income and employment for
future generations. However, exploitation of fisheries cannot be sustainably
increased as the human population grows. For a country with a growing
population and aspirations for higher standards of living, failure to rein-
vest resource rent represents a lost opportunity to build national wealth.
Furthermore, the recent collapse of the pilchard industry calls into ques-
tion whether the fisheries are being managed sustainably.
Regarding the final requirement for using natural capital to build
national wealth – reinvestment of resource rent – the policies of Botswana
and Namibia are quite different. Botswana developed an explicit policy of
reinvestment of all resource rent from mining and an indicator to monitor
this policy, the Sustainable Budget Index (SBI). (See Lange and Wright,
2004, for discussion of the SBI). Namibia has had no explicit policy regard-
ing reinvestment of revenues from natural capital.
3. Flow accounts for materials, services and pollution
The flow accounts of the environmental accounts are compiled and used for
economic analysis much more extensively than the asset accounts. They
provide macroeconomic indicators of sustainability as well as more detailed
information to support economic analysis of sources of environmental
pressure and options for change that can be used to improve sustainability.
The aggregate indicators provide an overview of the relationship between
economic development and the environment; the more detailed accounts
help explain the overview.
The flow accounts consist of three components: use of material and
energy resources, resource degradation and emission of pollutants, and pro-
duction and use of ecosystem services. The flow accounts are compiled in
both physical and monetary units. The physical accounts help set priorities
for policy based on the volume of resource use, pollution and so on while the
monetary accounts identify the relative costs and benefits of reducing pollu-
tion, resource use and so on. The flow accounts are also used in economic
models to evaluate options for development and specific policy instruments
for implementing a given development strategy, such as green taxes.
Physical accounts
At their simplest, the flow accounts are used to monitor the trend over time
of environmental goods and services, and pollution emissions, both total
and by industry. An example for wastewater and water pollution from the
Netherlands’ accounts is shown in Figure 17.2.
The construction of environmental–economic profiles, or ‘eco-efficiency’
indicators has become a common way of monitoring sustainability, and is
also used for benchmarking industry performance. These descriptive
statistics provide a first approach to identifying major users of resources
and sources of emissions, and provide a comparison of each sector’s rela-
tive environmental burden and economic contribution. Typically, eco-
efficiency indicators report an industry’s percentage contribution to the
national economy (value-added, employment) alongside its environment
impact such as emissions of various pollutants. A similar sector-level indi-
cator is the ‘resource productivity indicator’ calculated as materials (energy,
water and so on) or pollution per unit of value-added. (See example from
the water accounts for Australia in Table 17.3 and a more extensive example
for two industries in Sweden in Figure 17.3.)
While the eco-efficiency indicators report the direct generation of pollu-
tion associated with production, it is useful for policy makers to understand
the driving forces that result in such levels of pollution. The driving forces
for economic production are the final users. Input–output analysis has been
used to measure the total impact (direct + indirect) of a given final use. This
approach is especially useful in understanding the effects of different pat-
terns of household consumption or trade on the environment. An example
for SO2 air pollution in Sweden is given in Figure 17.4.
Monetary accounts
Effective environmental management is based not only on an understand-
ing of the volume of environmental goods and services and pollution, but
also an understanding of the economic implications. Policy makers need to
know, for example, what the welfare loss of pollution is (damage costs) and
where limited financial resources will be most effective in reducing envir-
onmental pressure, that is, the relative benefits and costs of reducing
different forms of environmental degradation from different sources.
Similarly they need to know the value of damages from deforestation in
terms of reduced productivity or increased production costs in other
sectors of the economy.
One of the most important applications of environmental accounting in
developing countries has been to identify goods and services from ecosys-
tems such as forests that are not adequately represented in the SNA. Many
non-market forest goods for example, (fuelwood, wild foods, medicines,
construction materials and so on) are, in principle, included in the SNA,

but due to measurement problems countries may underestimate the harvest
of these goods. In South Africa, for example, the value of non-market
forest goods, timber and non-timber, is greater than the commercial timber
harvest, but it is not included in the national accounts of South Africa
(Table 17.4).
In addition, forests provide environmental services that are often not rec-
ognized explicitly in the SNA. In Sweden, the value of recreation services
from forests is equal to the value of the timber harvest, but this service is
not attributed to forests. Similarly, forests in South Africa contribute sub-
stantially to agriculture (providing livestock grazing services and habitat
for wild bees that provide pollination services): a conservative estimate is
1907 million Rands in 1998; again, greater than the commercial timber
harvest, which is the only explicit value for forests in the national accounts.
In the case of these forest services, the value is included in the national
accounts, but as part of the livestock and crop activities, not as forest input
to those activities.
The issue of ecosystem services and undercounted non-market goods is
particularly important for many developing countries that may be overex-
ploiting their forests (or other natural resources, for example fisheries and
marine resources, wildlife and so on) for short-term economic growth. They
may have calculated that the revenues received compensate for the defor-
estation. But if the cost–benefit calculation does not also take into account
the loss of forest services to other sectors, such as tourism, agriculture,
hydroelectric power, fisheries, municipal water supply and so on, it is quite
possible that the losses from deforestation may outweigh the benefits.
The monetary flow accounts have also been used to address other policy
issues that are important for resource management, for example the subsidy
for water or wastewater treatment. The monetary accounts for water report
both the cost of delivery and the market price charged for water and waste-
water; the difference between the two is the subsidy. Figure 17.5 shows figures
for wastewater treatment in the Netherlands at a national level. Calculation
of subsidies from the monetary accounts for water have been compiled at the
industry level for three southern African countries; Botswana, Namibia
and South Africa (Lange and Hassan, forthcoming 2006). The accounts
for all three countries show extensive cross-subsidization, especially of
In many other countries, developed and developing, the cost of air and
water pollution is a major concern. After some initial experimentation with
valuation of pollution, many countries have not continued efforts to incor-
porate these into their environmental accounts. In large part, this is because
of a lack of consensus over alternative methods of valuation, and partly
because accurate valuation is quite difficult. There are two broad approaches

to valuation recommended in the SEEA: the cost of actions to prevent or
remediate degradation, and the benefit of actions to reduce pollution mea-
sured in terms of the value of the damages prevented.
In the absence of efficient markets, the cost and benefit measures are
likely to be quite different. The damage cost is the theoretically correct
approach for measuring changes in economic well-being and adjusting
macroeconomic aggregates, although both measures provide useful infor-
mation for environmental management. Until the SEEA provides more
concrete guidelines about valuation, most countries are unlikely to include
them in their environmental accounts. An example of monetary accounts
for air pollution in Sweden, based on the damage cost approach, is shown
in Figure 17.6.
Economic modeling with environmental accounts
Assessment of trade-offs in a partial equilibrium framework is a first step
towards understanding the policy impacts on the environment. But under-
standing the impact of broader changes, such as trade liberalization,
population growth, agricultural and industrial policy, energy pricing and
so on usually requires an economy-wide environmental–economic model.
One of the most important applications of the flow accounts is for eco-
nomic planning. Planning for sustainable development requires an integra-
tion of environmental and economic modeling. In the past, it was difficult

to integrate environmental and economic planning because the underlying
database for such models did not exist. The contribution of environmental
accounting is to provide the economist with a consistent, systematic, and
reliable set of accounts that are linked to the economic accounts. While this
topic is too broad to review in detail here, examples of widespread model-
ing applications include: modeling of environmental taxes and resource
user fees, modeling trade and the environment, modeling environmental
impacts of long-term development strategies, energy modeling.
4. Environmental protection and resource management expenditure
This component of the environmental accounts takes figures that are
already included in the SNA and rearranges them to make them more
useful for policy. There are three major parts: accounts for environmental
protection expenditure, accounts for natural resource management, and
environmental taxes and related fees. Two examples are provided here that
are both relevant to all countries, developed and developing.
a resource that is commercially exploited paying at least enough in taxes to
cover the costs of its management? In this case, only taxes and fees directly
related to the resource are included, not any corporate business income
taxes, which all companies may pay, regardless of what industry they are
in. In the Namibian fishing industry, the taxes contributed have always
covered the costs to government of managing the industry.
In the second example, Sweden has compared the share of carbon emis-
sions by industry to the share of carbon taxes that a given industry pays
(Figure 17.7). If a carbon tax is administered equally, on the basis of CO2
emitted, the two shares should be the same for an industry. Surprisingly,
there seems to be little relationship between the two. Households pay a
much greater share than the share of CO2 they are directly responsible for,
while manufacturing pays much less.
5. Economy-wide indicators of sustainable development
A wide range of macroeconomic indicators can be derived from the
asset and flow accounts of the SEEA; the major ones are listed in Table
17.6. The role of economic valuation in accounting, and the border
between accounting and economic analysis are unresolved issues in the
SEEA. Consequently, the SEEA does not make a recommendation for any
particular indicators and presents both physical and monetary macroeco-
nomic indicators. The Netherlands has been the major proponent of phys-

1997 (percentage of total)
ical NAMEA indicators for main environmental ‘themes’ determined by
national emission targets.
Within the monetary macro-indicators, there is further controversy over
whether sustainability is more accurately monitored from a national
income approach (for example environmentally adjusted GDP) or from a
wealth approach (for example genuine savings). These issues are addressed
in more detail in Chapter 18. There is also a view reported in the SEEA that
hypothetical national income calculated through modeling exercises should
also be included in the environmental accounts. However, most practition-
ers recognize that such indicators, while quite useful, belong firmly in the
realm of economic analysis rather than statistics.
6. The future of environmental accounting
Environmental accounts make a great contribution to further integrating
environmental and economic analysis by providing a single database that is
consistent for both sets of information. The SEEA, as an official handbook
endorsed by the UN Statistics Committee, provides the basis for viewing
environmental accounting as simply a more thorough way of doing
national accounts. However, the SEEA is far from a complete handbook
providing clear standards on all issues, and the problem is both conceptual
and empirical. The three most urgent issues are the following:
Asset valuation, depletion and degradation. At this time, the SEEA pre-
sents several alternative approaches to measuring the value of assets and
depletion/degradation but makes no recommendation for which approach
to use, even though the approaches can give widely differing results. The
issue of constant price asset values is not even discussed in the SEEA. This
situation is not one that will encourage countries to implement the asset
Macroeconomic indicators (monetary). Ministries of Finance need to
know whether their development strategy is laying the basis for long-term
economic growth or not. In developing countries, PRSPs (Poverty
Reduction Strategy Programs) have been widely adopted as a planning tech-
nique to promote sustainable economic growth and poverty reduction.
However, PRSPs use GDP and other conventional macro indicators in their
monitoring framework; consequently, policy makers receive information
about only half of the objective, short-term economic growth, but not sus-
tainability of that growth. The long-term cost of soil erosion, for example,
is enormous in many countries and may undermine any short-term gains in
Handbook of sustainable development
GDP. There is a great need for a complementary indicator of sustainability,
such as Genuine Savings, that can be used in PRSPs. The SEEA does not
make clear whether countries should be monitoring stocks (wealth and
changes in wealth/savings) or flows (national income).
Ecosystem accounting. Some ecosystem values, notably for forests, have
been incorporated in environmental accounts, but much of this work has
not yet been systematically incorporated in the SEEA. Accounting for
ecosystem services is especially important for developing countries for
several reasons. Developing countries contain most of the world’s
biodiversity; biodiversity protection services benefit not only local com-
munities but also the global community. Ecosystem services, such as water
and soil protection, are often under greatest threat in developing countries,
but these countries often have fewer resources to cope with loss of ecosys-
tem services (flood control, water purification, increased health care and so
on). In addition, the well-being of developing countries may be more vul-
nerable to loss of these services as a majority of people depend directly on
ecosystem health (for example soil stability for subsistence farming,
fisheries habitat and so on), and often have limited alternative sources of
livelihood. Noting that the poor are often those most vulnerable to deteri-
oration of natural systems, the Millennium Ecosystem Assessment states
that ‘development policies aimed at reducing poverty that ignore the
impact of our current behavior on the natural environment may well be
doomed to failure’ (Millennium Assessment Board, 2005).
1. There is no consensus yet about how to measure human capital.
2. The most important natural capital is included here: minerals for both countries and fish-
eries for Namibia. The value of other important natural capital, notably land and water,
has not yet been estimated, but this is not expected to seriously affect the trends in per
capita wealth. This is discussed further in (Lange, 2004b).
Ahlroth, S. (2000), ‘Correcting NDP for SO2 and NOx emissions: implementation of a theo-
retical model in practice’, National Institute for Economic Research (NIER): Stockholm.
Arrow, K., P. Dasgupta and K. Mäler (2003a), ‘Evaluating projects and assessing sustainable
development in imperfect economies’, Environmental and Resource Economics, 26: 647–85.
Arrow, K., P. Dasgupta and K. Mäler (2003b), ‘The genuine saving criterion and the value of
population’, Economic Theory, 21: 217–25.
Australian Bureau of Statistics (ABS) (2004a), Australian System of National Accounts,
Consolidated Balance Sheet, Canberra: ABS.
Australian Bureau of Statistics (ABS) (2004b), ‘Water Accounts Australia 2001–2001’,
Canberra: ABS.
Dasgupta, P. and K. Mäler (2000), ‘Net national product, wealth, and social well-being’,
Environment and Development Economics, 5: 69–94.
Hamilton, K. and M. Clemens (1999), ‘Genuine savings rates in developing countries’, World
Bank Economic Review, 13(2): 333–56.
Environmental and resource accounting
Heal, G. and B. Kriström (2005), ‘National income and the environment’, in K. Mäler and
J. Vincent (eds), Handbook of Environmental Economics, Volume 3, Amsterdam: North-
Hellsten, E., S. Ribacke and G. Wickbom (1999), ‘SWEEA – Swedish environmental and
economic accounts’, Structural Change and Economic Dynamics, 10(1): 39–72.
Kunte, A., K. Hamilton, J. Dixon, and M. Clemens (1998), ‘Estimating national wealth:
methodology and results’, Environment Department Papers, Environmental Economics
Series No. 57, Washington: The World Bank.
Lange, G. (2002), ‘Trade and the environment in Southern Africa: impact of the user pays
principle for water on exports of Botswana, Namibia, and South Africa’, paper presented
at the Conference of the International Input–Output Association, 10–15 October,
Montreal, Canada.
Lange, G. (2003a), ‘Environmental accounts: uses and policy applications’, Environment
Department Paper No. 87, Washington, DC: World Bank.
Lange, G. (2003b), ‘Fisheries accounts; management of a recovering fishery’, in G. Lange et
al., Environmental Accounting in Action: Case Studies from Southern Africa, Cheltenham,
UK and Northampton, MA, USA: Edward Elgar.
Lange, G. (2004a), ‘Manual for environmental and economic accounts for forestry: a tool for
cross-sectoral policy analysis’, FAO Forestry Department Working Paper, March.
Lange, G. (2004b), ‘Wealth, natural capital, and sustainable development: the contrasting
examples of Botswana and Namibia’, Environment and Resource Economics, November,
29(3): 257–83.
Lange, G. (2005), ‘Introducing environmental sustainability into the Ugandan national
accounts’, report to IUCN and the Environment and Natural Resources Sector Working
Group, Kampala, Uganda, March.
Lange, G. and R. Hassan (forthcoming, 2006), The Economics of Water Management in
Southern Africa: An Environmental Accounting Approach, Cheltenham, UK and
Northampton, MA, USA: Edward Elgar.
Lange, G. and M. Wright (2004), ‘Sustainable development in mineral economies: the example
of Botswana’, Environment and Development Economics, August, 9(4).
Lange, G., R. Hassan and K. Hamilton (2003), Environmental Accounting in Action: Case
Studies from Southern Africa, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
Millennium Assessment Board (2005), ‘Millennium Ecosystem Assessment’, available from
Sjölin, M. and A. Wadeskog (2000), Environmental Taxes and Environmentally Harmful
Subsidies, Report prepared for DG Environment and Eurostat, available at http://www.
Smith, R. and C. Simard (2001), ‘A proposed approach to sustainable development indicators
based on capital’, paper presented by Statistics Canada at the National Conference of
Sustainable Development Indicators, 27 March, Ottawa, Canada.
United Nations (1993), Operational Manual for the System of Integrated Environmental and
Economic Accounts, New York: UN.
United Nations, European Commission, International Monetary Fund, Organization for
Economic Cooperation and Development and World Bank (1993), System of National
Accounts, New York: UN.
United Nations, European Commission, International Monetary Fund, Organization for
Economic Cooperation and Development and World Bank (2003), Integrated
Environmental and Economic Accounting 2003, New York: UN.
Van der Veeren, R., R. Brouwer, S. Schenau and R. van der Stegen (2004), ‘NAMWA: a new
integrated river basin information system’, Voorburg, The Netherlands: Central Bureau of
World Bank (2002), World Development Report, Washington, DC: World Bank.
World Bank (forthcoming, 2005), Where is the Wealth of Nations?, Washington, DC: World
18 Genuine saving as an indicator of sustainability Kirk Hamilton and Katharine Bolt
1. Introduction
Choosing sustainable development is an ethical position adopted by
society, reflecting a desire to ensure that future generations enjoy at least as
much welfare as the current generation. Because sustainability is inherently
about the future, measuring it has been a challenge. Without indicators,
promises to achieve sustainability risk being largely empty.
A common thread in the literature on sustainable development concerns
the treatment of the environment and natural resources within the System
of National Accounts (SNA). This is important because the SNA has an
incomplete treatment of resource issues. To give one example, commercial
natural resource stocks are supposed to be measured in the national
balance sheet accounts of the SNA, but there is no corresponding adjust-
ment to net national income or net saving to reflect the consumption of
capital that occurs when these stocks are exploited. Similarly, there is no
explicit accounting in the SNA for the damages to economic assets that
result from pollution emissions. The consequence is that SNA measures of
income and saving are overstated, substantially so for the most resource-
dependent economies. In many countries finance ministries are simply
working with the wrong figures.
If depletion of the environment is ignored in the most common and
powerful set of indicators used to guide economic development, then the
threat to sustainability is obvious. Decisions to exploit natural resources
now may harm future generations if the depletion of one asset is not offset
by investment in another – the fact that this depletion is occurring would
be completely invisible in standard national accounting.
To correct this flaw in the national accounts, measures of ‘genuine’ saving
account for the change in real wealth in an economy after due account is taken
of the depreciation and depletion of the full range of assets in the economy.
Pearce and Atkinson (1993) laid the conceptual foundation for such an
extended measure of saving, as well as presenting some of the first empirical
estimates using results from the green national accounting literature.
In a series of papers, Hamilton and Clemens (1999), Dasgupta and
Mäler (2000) and Asheim and Weitzman (2001) have established the growth
Genuine saving as an indicator of sustainability
theoretic basis for the linkage between saving and sustainability. While the
main result from this literature will be presented below, the intuition is
straightforward. If we conceive of wealth – the value of all assets in an
economy – as the basis of future welfare, then current changes in wealth
must have future welfare consequences. It is at least conceivable that a
decline in wealth now will lead to falls in future levels of welfare – such an
economy would not be sustainable by Pezzey’s (1989) definition. Growth
theory makes this connection concrete.
The focus in the sustainable development literature is on genuine saving
rather than ‘genuine income’ (that is consumption plus genuine saving) for
good reason – adjusting the level of income to reflect the depreciation of a
wider array of assets does not in itself indicate whether an economy is on
a sustainable path. However, the fact that genuine income would typically
be lower than the standard measure of Net National Income does send an
important message – that we should not be treating asset consumption as
Genuine saving is more than a theoretical construct. In addition to the
empirical results in Pearce and Atkinson (1993) and Hamilton and
Clemens (1999), the World Bank has been publishing estimates of ‘adjusted
net’ saving (the formal name for genuine saving at the Bank) for 140 coun-
tries since 1999 in the World Development Indicators (World Bank, 2005).
The plan of the chapter is the following. The next section will lay out the
theoretical basis and measurement issues for genuine saving. This will be
followed by presentation of some of the published saving estimates from
the World Bank. Recent extensions of the saving analysis in the literature
will be presented. Finally, the chapter concludes with some thoughts on
current challenges.
2. Theory and measurement
Pearce and Atkinson (1993) made a first attack on the problem of measur-
ing sustainable development by employing basic intuitions concerning
assets and sustainability. They argued that sustainability can be equated to
non-declining values of all assets, including natural resources. The conse-
quence of this conceptualization is that changes in asset values, measured
by net saving, should signal whether an economy is on a sustainable path.
Pearce and Atkinson presented empirical results on net saving for a range
of developed and developing countries using values published in the green
accounting literature.
More recent theoretical work on savings has firmly established the
linkage between net savings, social welfare and sustainable development.
Hamilton and Clemens (1999) tackle the problem for an optimal economy,
and Dasgupta and Mäler (2000) for non-optimal economies (with suitable
Handbook of sustainable development
definition of shadow prices). Asheim and Weitzman (2001) show that
growth in real NNP (where prices are deflated by a Divisia index of con-
sumption prices) indicates the change in social welfare in the economy.
Genuine saving is defined as,
Here the Ki are the stocks of assets in the economy, and the pi are their
shadow prices. The expression says that genuine saving is measured as the
change in real wealth. To measure sustainability it is important that genuine
saving span as wide a range of assets as possible, including assets with nega-
tive shadow prices such as pollution stocks. In principle changes in the
stocks of produced, human, natural, social and institutional capital should
all be measured in saving – in practice there are data and conceptual prob-
lems associated with the measurement of assets like social capital.
The basic theoretical insight of Hamilton and Clemens (1999) is to show
that genuine saving G, utility U, social welfare V, marginal utility of con-
sumption , and pure rate of time preference are related as follows
This says that social welfare is equal to the present value of utility, and that
genuine saving is equal to the instantaneous change in social welfare mea-
sured in dollars.1 The utility function can include consumption C and any
other set of goods and bads to which people attribute value.
Hamilton and Clemens (1999) go on to show that negative levels of
genuine saving must imply that future levels of utility over some period of
time are lower than current levels – that is negative genuine saving implies
unsustainability. Similar implications hold for the approaches of Dasgupta
and Mäler (2000) and Asheim and Weitzman (2001).
These approaches to greening the accounts, and the models that under-
pin them, are agnostic on the question of the degree of substitutability
between different assets, in particular between produced and natural assets.
An important strand of the sustainability literature, dating back to Pearce
et al. (1989), looks at the question of strong versus weak sustainability
(see also Chapter 4). Weak sustainability assumes that there are no funda-
mental constraints on substitutability. If, however, some amount of nature

must be conserved in order to sustain utility – the strong sustainability
assumption – then these saving models need to be modified to incorporate
the shadow price of the sustainability constraint.
A formal approach to the strong vs weak sustainability problem has been
explored in the ‘Hartwick rule’2 literature. Dasgupta and Heal (1979) and
Hamilton (1995) show that if the elasticity of substitution between pro-
duced capital and natural resources is less than 1, then the Hartwick rule is
not feasible – eventually production and consumption must fall, implying
that the economy is not sustainable under the rule.
The question of ecological thresholds is potentially important in
measuring sustainable development. Crossing certain boundaries may
produce catastrophic results, such as the re-routing of the Gulf Stream as
a result of global warming, or the death of most plankton in the ocean as
a result of ozone layer destruction. In environmental economic terms we
may think of a threshold as a point where the marginal damage curve is
unbounded. As long as marginal damages are smooth as a threshold is
approached, the saving approach will give correct signals concerning
sustainability, since approaching the threshold will eventually result in
negative savings. If the marginal damage curve is not smooth and
becomes vertical at the threshold, then the saving rule may not indicate
unsustainability as the threshold is approached. There is clearly an
important question of the science of threshold problems, since we do not
know a priori what the shape of the marginal damage curve is for many
important problems.3
Pezzey (2004) makes the point that genuine saving provides a one-sided
sustainability test: if saving is negative, then there must be future declines
in utility. The opposite is not true in general – positive saving at a point in
time does not indicate that future utility is everywhere non-declining.
However, Hamilton and Hartwick (2005) show that making positive
genuine saving an element of a policy rule can yield sustainability – this
result is described below.
3. Empirical experience
Each year the World Bank publishes genuine saving estimates in the World
Development Indicators (World Bank, 2005).4 The following summarizes
how the saving estimates are constructed:
Genuine saving

There are a number of points to note about the calculation. First, genuine
saving as published by the World Bank is not just a ‘green’ indicator – it
includes investment in human capital (as proxied by education expenditure)
as a part of saving. Carbon dioxide damages, a global issue representing
damages inflicted on other countries, are included in national savings on
the assumption that a certain property right holds: that countries have the
right not to be polluted by their neighbours. Finally, damages from parti-
culate matter in air are based on the value of damage to health – health-
fulness is treated as an asset, part of human capital.
In any given year 10–30 countries actually have negative genuine saving.
As Figure 18.1 shows, aggregate savings for the developing regions of the
world show distinctive levels and trends.

The Middle East and North Africa stands out for its consistently
negative saving rate, reflecting high dependence on petroleum extrac-
tion. Regional genuine saving rates are highly sensitive to changes in
world oil prices. This is clearly shown in Figure 18.1 – genuine saving
rates dropped in 1979, largely owing to the consumption of sharply
increased oil rents following the Iranian revolution.
East Asia and Pacific stands in stark contrast, with recent aggregate
genuine saving figures nearing 30 per cent, driven largely by China.
The boom in economic performance from the second half of the
1980s until the Asian financial crisis in 1997 is reflected in the genuine
saving numbers, largely driven by increases in gross national saving.
Genuine saving rates have been hovering around zero in sub-Saharan
Africa. Positive saving in countries such as Kenya, Tanzania and
South Africa is offset by strongly negative genuine saving rates in
resource-dependent countries such as Nigeria and Angola, which
have genuine saving rates of 30 per cent in 2003.
South Asia displays consistently strong genuine saving rates, fluctu-
ating between 10 and 15 per cent since 1985, with India dominating
the aggregate figure.
Latin American genuine saving rates have remained fairly constant
throughout the 1990s. The large economies in the region, Mexico and
Brazil, have positive genuine saving rates in excess of 5 per cent.
However, like many oil producers, Venezuela’s genuine saving rate
has been persistently negative since the late 1970s.
Genuine saving data for Eastern Europe and Central Asia are only
available from 1995. Saving rates have fallen from over 7.7 per cent in
1995 to 1.7 per cent in 2003, largely driven by dissaving in the oil
states of Azerbaijan, Kazakhstan, Uzbekistan, Turkmenistan, and
the Russian Federation.
One of the themes that suggests itself in the analysis of regional trends in
saving is the link between high resource dependence (typically on oil) and
genuine saving rates. Figure 18.2 looks more specifically at this question by
scattering genuine saving rates against rates of dependence on exhaustible
resources in 2003 (only mineral and energy rent shares greater than 1 per
cent of GNI are shown).
The tendency in Figure 18.2 is clear. If mineral- and energy-dependent
economies were diligently investing their rents in other types of capital, as
the Hartwick rule suggests, then there should be no apparent link between
resource dependence and genuine saving. Instead we see a clear downward

Figure 18.2 Genuine saving vs exhaustible resource dependence, 2003
trend, which suggests a tendency to consume rents that increases with
resource dependence.
Genuine saving lends itself to a variety of empirical applications beyond
the analysis of sustainability. Recent examples include Atkinson and Hamil-
ton (2003) who explore the extent to which genuine saving can explain the
‘resource curse’, while de Soysa and Neumayer (2005) look at the impact of
trade openness and other liberalization measures on genuine saving.
4. Extensions
Reference was made above to the Hartwick rule, a rule for achieving sus-
tainability that is built around genuine saving. Under this rule an economy
will achieve maximal constant consumption forever (or constant utility in
a more general formulation) if genuine saving is set to zero at each point in
time. This holds even in the canonical exhaustible resource economy of
Dasgupta and Heal (1979) with fixed technology, a single produced capital
stock and a finite resource stock that is essential for production – in this
economy the rule reduces to ‘invest resource rents’.
Hamilton and Hartwick (2005) point toward a generalization of the
Hartwick rule by deriving the following relationship between consumption,
saving and the interest rate for an optimizing Dasgupta–Heal economy:

Here C is consumption and r the (time-varying) interest rate. This expres-
sion relates growth in consumption to the sign of genuine saving and the
difference between the interest rate and the growth rate of genuine saving.
Dixit et al. (1980) showed that a slightly generalized version of the
Hartwick rule holds in any economy that is competitive – an economy
where producers maximize profits and households maximize utility. A
competitive economy is not necessarily PV-optimal (the path defined by
solving the growth problem where the present value (PV) of utility is max-
imized), so a variety of policy rules can potentially be applied. Hamilton
and Withagen (2007) show that expression (18.4) holds in competitive
economies, which means that it is possible to define a more general rule for
sustainability: in a competitive economy, maintaining genuine saving rates
that are (i) positive and (ii) growing at a rate less than the interest rate, will
lead to increasing consumption at each point in time.
Ferreira and Vincent (2005) use World Bank historical data on con-
sumption and genuine saving to test a basic proposition linking current
saving to future welfare. They start with a result from Weitzman (1976): if
the economy is PV-optimal and the interest rate is constant then,

Genuine saving is equal to the difference between a particular weighted
average of future consumption and current consumption. This relationship is
tested econometrically using per capita data from 1970 to 2000. Ferreira and
Vincent find that the relationship holds best for non-OECD countries, and
that there is a better fit as more stringent measures of saving are tested, that
is when going from gross saving to net saving to genuine saving (but excluding
the adjustment for education expenditure, which performs very badly).
Hamilton and Hartwick (2005) note that expression (18.4) can be inte-
grated to yield,5

So genuine saving is equal to the present value of changes in future
consumption. Hamilton (2005) uses historical data to test whether this
expression holds. Figure 18.3 displays the right-hand side of expression
(18.6) scattered against genuine saving in 1980. The broad conclusion is
similar to Ferreira and Vincent (2005) – using data for all countries, genuine
saving fits expression (18.5) better than other measures of saving, while the
fit is extremely poor in OECD countries.
Hamilton et al. (2006) show that a particularly simple saving rule yields
sustainability in a competitive Dasgupta–Heal economy: if genuine saving
is positive and constant then consumption will rise without bound. This
rule and the standard Hartwick rule are then used to test the counterfac-
tual: how rich would countries be if from 1970 to 2000 they had followed
either the standard Hartwick rule or had maintained genuine saving at
a constant value equal to 5 per cent of 1987 GDP? Figure 18.4 compares
the two counterfactual estimates of fixed capital (it is assumed that all
savings are invested in produced assets) with the observed level of fixed
capital in 2000 for selected countries.
The results of following either policy rule are dramatic for the oil pro-
ducers: Venezuela, Trinidad and Tobago and Gabon would all be as rich as
South Korea if they had followed the constant genuine saving rule. Nigeria
would not be rich, but it would be five times richer than it is today. It is no
simple matter for a resource-dependent developing country to maintain
positive savings through financial crises, civil unrest and natural disasters –
but the payoffs are potentially huge.
Finally, World Bank (2006, ch. 5) extends the empirical work on genuine
saving to examine the effects of population growth. The net change in
wealth per capita GN is calculated as For population N, this says that the net change in wealth per capita is equal
to total genuine saving per person minus a Malthusian term, the popula-
tion growth rate g times total tangible wealth W per person. Dasgupta
(2001) shows that this expression measures the change in social welfare
when (i) the population growth rate is constant, (ii) per capital consump-
tion is independent of population size, and (iii) production exhibits con-
stant returns to scale.
Figure 18.5 scatters the net change in wealth per capita against GNI per
capita (logarithmic scale) in 2000. The upward trend and the fact that most
low income countries (GNI of less than $750 per capita) face net declines
in wealth per capita means, roughly speaking, that the rich are getting
richer while the poor are getting poorer. However, Hamilton (2005) pre-
sents evidence that the Malthusian adjustment tends to overstate the
impact of population growth on future changes in consumption.
World Bank (2006) also calculates the saving gap – the increase in saving
that would be required to bring a country’s net change in wealth per capita
back to zero. For many African countries in particular this gap is huge,
from 10–70 per cent of GNI, suggesting that economic and environmental
policy alone will not suffice to bring sustainability in per capita terms to
these economies.
5. Challenges for the future
A conceptual challenge for the work on genuine saving concerns the ques-
tion of optimality. Hamilton and Clemens (1999) derive expression (18.3)

in an optimal economy, so the application of the theory to the real world
becomes an important question. Dasgupta and Mäler’s (2000) solution is
to derive the parallel expression for a non-optimal economy, but they are
required to use accounting prices that are defined as the partial derivatives
of the value function V for the non-optimal path – to define the prices it is
therefore necessary to define the path. Arrow et al. (2003) explore this ques-
tion in some depth.
If we assume that world prices for resources do reflect scarcities and are
therefore relatively undistorted, then the derived shadow prices should be
a reasonable reflection of the user costs associated with resource extraction.
Whether genuine saving measured using these prices truly reflects the
change in social welfare is still an open question, although there is a huge
amount of literature on cost–benefit analysis of projects which would
suggest using precisely these prices. More work on this topic is required.
The new results on saving rules in competitive economies offer promise
in this regard – there is no underlying assumption of optimality, and it is at
least a reasonable proposition that many economies are competitive. One
obvious conclusion follows from expression (18.4): if genuine saving is neg-
ative and constant then the economy is on an unsustainable path. The
general rule for sustainability was stated above: maintain positive saving
and ensure that it does not grow faster than the interest rate. These saving
rules for competitive economies offer scope for actually using the concept
of genuine saving in designing policies for sustainability.
There is no shortage of empirical questions when it comes to measuring
genuine saving. Among the challenges that appear the most urgent are:
Identifying non-linearities in the natural world that may not be cap-
tured in any simple way in measures of genuine saving. We do not
want to be assuring ministers that all is well because saving is pos-
itive, only to discover that a major flip in natural systems has severe
consequences for human welfare.
Valuing truly difficult assets such as biodiversity.
Inventorying and valuing the environmental services that underpin
so much economic activity, whether it is pollination or regulation of
flow in a watershed. While many of these values are captured indir-
ectly in other asset values – the value of farmland, for example – the
fact that there is no explicit valuation means that there are opportu-
nities for unpleasant policy surprises.
Estimating elasticities of substitution for resources. The availability
of databases of natural resource stocks and flows, in quantity and
value terms, means that there should be more scope for exploring this
important question – World Bank (2006, Chapter 8) estimates the

elasticity of substitution between land and fixed capital to be close to
one, an important result.
The policy challenges involved in increasing genuine saving are closely
linked to the components of saving. The ‘bottom line’, genuine saving, will
be affected by fiscal and monetary policies that influence gross saving effort.
In addition, increasing human capital investments and making them more
effective will boost the bottom line. Achieving efficient levels of resource
extraction and pollution emissions will also increase genuine saving – note,
however, that this does not imply reducing resource extraction or pollution
emissions to zero.
While the focus of this chapter has been on saving, the profitability of
investments financed by this saving is of paramount importance. If gov-
ernments invest in ‘showcase’ projects with low or negligible social returns,
then savings have in effect been consumed, with consequent effects on
future welfare.
Finally, for the poorest economies, increasing saving could be taken to
imply decreasing consumption, not a palatable policy option in countries
where consumption is already at subsistence levels. For these countries a
better alternative will be to focus on boosting the efficiency of the economy
through economic reforms, raising growth and potentially leading to a vir-
tuous cycle of increasing saving and consumption.
1. This result is foreshadowed in Aronsson et al. (1997, expression 6.18) who show that net
saving measured in utility units is equal to the present value of changes in utility for a
general (possibly time-varying) pure rate of time preference.
2. Hartwick (1977) showed that consumption is sustainable (in fact constant) in a fixed tech-
nology economy with an essential exhaustible resource if: (i) net saving is everywhere 0;
(ii) the elasticity of substitution between resources and produced capital is 1; and (iii) the
elasticity of output with respect to produced capital is greater than the corresponding
elasticity for the resource.
3. See also Pearce et al. (1996).
4. The formal name of the saving indicator is ‘adjusted net saving’. Genuine saving is the
informal name.
5. This is also proved, in a more general framework, in Dasgupta (2001) Ch. 9, appendix A.7.
Aronsson, T., P.-O. Johansson and K.-G. Löfgren (1997), Welfare Measurement, Sustainabiliy
and Green National Accounting: A growth theoretical approach, Cheltenham, UK and
Northampton, MA, USA: Edward Elgar.
Arrow, K.J., P. Dasgupta and K.-G. Mäler (2003), ‘Evaluating projects and assessing sustainable
development in imperfect economies’, Environmental and Resource Economics, 26(4): 647–85.
Asheim, G.B. and M.L. Weitzman (2001), ‘Does NNP growth indicate welfare improvement?’,
Economics Letters, 73(2): 233–9.
Atkinson, G. and K. Hamilton (2003), ‘Savings, growth and the resource curse hypothesis’,
World Development, 31(11): 1793–807.
Handbook of sustainable development
Dasgupta, P. (2001), Human Well-being and the Natural Environment, Oxford: Oxford
University Press.
Dasgupta, P. and G. Heal (1979), Economic Theory and Exhaustible Resources, Cambridge:
Cambridge University Press.
Dasgupta, P. and K.-G. Mäler (2000), ‘Net national product, wealth, and social well-being’,
Environment and Development Economics, 5, Parts 1&2: 69–93, February and May.
de Soysa, I. and E. Neumayer (2005), ‘False prophet, or genuine savior? Assessing the effects
of economic openness on sustainable development, 1980–99’, International Organization,
59(3): 731–72.
Dixit, A., P. Hammond and M. Hoel (1980), ‘On Hartwick’s rule for Regular Maximin Paths of
Capital Accumulation and Resource Depletion’, Review of Economic Studies, XLVII: 551–6.
Ferreira, S. and J. Vincent (2005), ‘Genuine savings: leading indicator of sustainable develop-
ment?’, Economic Development and Cultural Change, 53(3): 737–54.
Hamilton, K. (1995), ‘Sustainable development, the Hartwick rule and optimal growth’,
Environmental and Resource Economics, 5(4): 393–411.
Hamilton, K. (2005), ‘Testing genuine saving’, Policy research Working Paper no. 3577,
Washington: The World Bank.
Hamilton, K. and M. Clemens (1999), ‘Genuine savings rates in developing countries’, The
World Bank Economic Review, 13(2): 333–56.
Hamilton, K. and J.M. Hartwick (2005), ‘Investing exhaustible resource rents and the path of
consumption’, Canadian Journal of Economics, 38(2): 615–21.
Hamilton, K. and C. Withagen (2007), ‘Savings growth and the path of utility’, Canadian
Journal of Economics, 40(2), forthcoming.
Hamilton, K., G. Ruta and L. Tajibaeva (2006), ‘Capital accumulation and resource deple-
tion: a Hartwick rule counterfactual’, Environmental and Resource Economics, 34: 517–33.
Hartwick, J.M. (1977), ‘Intergenerational equity and the investing of rents from exhaustible
resources’, American Economic Review, 67(5): 972–4.
Pearce, D.W. and G. Atkinson (1993), ‘Capital theory and the measurement of sustainable
development: an indicator of weak sustainability’, Ecological Economics, 8: 103–8.
Pearce, D.W., K. Hamilton and G. Atkinson (1996), ‘Measuring sustainable development:
progress on indicators’, Environment and Development Economics, 1: 85–101.
Pearce, D.W., A. Markandya and E.B. Barbier (1989), ‘Blueprint for a green economy’,
London: Earthscan.
Pezzey, J. (1989), ‘Economic analysis of sustainable growth and sustainable development’,
Environment Dept Working Paper No. 15, The World Bank.
Pezzey, J. (2004), ‘One-sided sustainability tests with amenities and changes in technology,
trade and population’, Journal of Environmental Economics and Management, 48(1):
Weitzman, M. (1976), ‘On the welfare significance of national product in a dynamic economy’,
Quarterly Journal of Economics, 90(1): 156–62.
World Bank (2005), World Development Indicators, Washington: The World Bank.
World Bank (2006), Where is the Wealth of Nations? Measuring Capital for the 21st Century,
Washington: The World Bank.

19 Measuring sustainable economic welfare Clive Hamilton
1. Introduction1
It has long been recognized that, above a threshold, GDP growth does not
correlate well with changes in national well-being (for example Layard, 2005
and Chapter 16 of this volume). That threshold has been well and truly
passed by OECD countries. The principal shortcomings of GDP as a
measure of changes in national well-being are: the failure to account for
how increases in output are distributed within the community; the failure to
account for the contribution of household work; the incorrect counting of
defensive expenditures as positive contributions to well-being; and the
failure to account for changes in the stocks of both built and natural capital.
There have been several attempts to construct indicators of changes in
well-being that are more comprehensive than GDP. A well-known earlier
index was built by Nordhaus and Tobin (1972). In more recent years Daly
and Cobb have constructed the Index of Sustainable Economic Welfare
(ISEW) in an influential appendix to their book, For the Common Good
(1990). The Daly and Cobb index has led to a lively debate on a series of
methodological and measurement issues (much of which was presented in
Cobb and Cobb, 1994), and construction of similar indexes for several
other countries.2
These later efforts have placed a particular emphasis on accounting for
environmental costs in the new measure of welfare. The initial Daly and
Cobb index for the USA has been refined and developed by Cobb, Halstead
and Rowe (1995) and renamed the Genuine Progress Indicator (GPI), the
name that has increasingly replaced ISEW and that will be used here.
2. Welfare and sustainability
The key to understanding the attempts to develop the GPI lies in the notion
of sustainability. The best starting point is John Hicks’ 1939 definition of
income. ‘Hicksian income’ is defined as the maximum amount that a person
or a nation could consume over some time period and still be as well off at
the end of the period as at the beginning (Hicks, 1946: 172).3 Thus income
is maximum sustainable consumption. Sustaining consumption over a
given period depends on maintaining the productive potential of the
capital stocks that are needed to generate the flow of goods and services
that are consumed.

to define and measure ‘consumption’ in a way that provides a better
approximation of actual well-being than the simple measure of mar-
keted goods and services that appears in the national accounts; and
to account for the sustainability of consumption by incorporating
measures of changes in the value of capital stocks

Taking account of these two classes of influence on welfare over time, we
may end up with a situation in which GDP is increasing while consumption
(more broadly defined) is rising or falling, and while capital stocks are
growing or declining.
The GPI combines changes in the value of stocks and the values of flows
of current consumption. Consistent with the definition of Hicksian
income, capital stocks perform two functions in the GPI method of mea-
suring changes in welfare – they yield an annual flow of services and they
contribute to the sustainability or otherwise of levels of consumption in the
future. In order to prevent the depreciation or depletion of capital stocks,
a portion of current consumption needs to be ‘set aside’ to replenish the
stocks. The implication of this is that, unlike the way in which changes in
GDP are used, year-on-year changes in the GPI are not very meaningful.
The purpose of the GPI is to illustrate trends over time.
We now look more closely at the two tasks that the GPI sets itself and
then consider some of the further methodological issues it gives rise to.
3. Measuring ‘consumption’ comprehensively
For individuals or households, consumption may be defined as annual
flows of marketed and non-marketed goods and services. Perhaps the
biggest category of non-marketed goods and services comprises those pro-
duced in the home by unpaid household work. Non-marketed goods and
services also include services provided by the natural environment, such as
the aesthetic and recreational services of old-growth forests and the health-
sustaining properties of clean air.
A more comprehensive definition of consumption that takes account of
non-marketed goods and services is particularly important because mea-
sured GDP growth may reflect nothing more than the transfer of activity
from the non-market to the market sector, a problem long recognized in the
development literature. This is most apparent in the case of household
work, but applies equally to any other ‘free’ service. Just as, in the well-
known observation, GDP declines ‘if a man marries his housekeeper’,
GDP rises if an entrance fee is levied on visits to a national park or a family
decides to eat out more often.

Measuring sustainable economic welfare
Consumption includes negative flows or ‘bads’. Some monetary expen-
ditures by final consumers – which are therefore included as expenditures
in GDP – represent not additions to welfare but attempts to offset some
change in social, environmental or individual circumstances which is
causing a decline in welfare. These are known as defensive expenditures and
are deducted from the value of personal consumption expenditure, which
provides the starting point of the GPI.
These observations apply to consumption by individuals. At a national
level it is important to take account of differences in the welfare impact of
consumption between households or individuals. One of the most fre-
quently heard criticisms of the use of GDP growth as a measure of national
welfare is that it assumes that an extra $1 million of consumption by wealthy
households has the same impact on national welfare as an extra $1 million
of consumption by impoverished households. The GPI rejects this assump-
tion and adjusts consumption flows by a measure of income distribution.
The GPI assumes that personal consumption spending by individuals on
marketed goods and services is the major component of welfare and that
an increase in this spending represents, ceteris paribus, a corresponding
increase in welfare. There is a large amount of literature critical of the
assumption that there is a close relationship between changes in consump-
tion spending and changes in individual welfare (see for example Layard,
2005; Frey and Stutzer, 2002). Many studies have shown that, above a
certain level of income, perceived well-being depends more on the level of
one’s income relative to other people’s incomes, or to previous or expected
levels, than on absolute levels.4 But the purpose of the GPI is to demon-
strate that, even using conventional economic methods, a more compre-
hensive attempt to account for changes in welfare may show large
deviations from GDP over time. Consequently, we adopt the assumption
that increases in personal consumption (adjusted for the distribution of
income) reflect increases in welfare. It is important to keep this ‘consump-
tion framework’ in mind because, if it is accepted, many of the criticisms of
the GPI and ISEW are neutralized.
4. Accounting for changes in the value of capital stocks
Sustaining levels of consumption requires that the productive potential of
capital stocks be maintained. Capital stocks can be divided into five forms,
which we discuss in turn. While GDP accounts for changes in none of them,
the GPI attempts to incorporate changes in the value of the first three.
Built capital This covers the stocks of physical machinery, buildings
and infrastructure that are essential to sustaining levels of GDP. These
stocks deteriorate and a portion of income must be set aside each year to
invest in them to maintain and improve their productive potential. This is

Handbook of sustainable development
a long-recognized problem and has led periodically to attempts by statist-
ical agencies to construct measures of net national product (NNP). The
GPI adjusts consumption spending to take account of net capital growth
which, if positive, adds to sustainable economic welfare. (In principle, it
should take account of changes in annual flows of services from the stock
of built capital.)
Financial assets A nation’s ability to sustain investment in built capital
assets is diminished if it is accumulating foreign debts, since some part of
future income must be devoted to repaying the debts.5 But if those loans are
being invested productively then future income will be higher and it will be
possible to repay the debts without additional burden. To the extent that
foreign debt has been invested productively in the past, current consump-
tion will be higher. But if foreign borrowing is dissipated on consumption
goods it represents a drain on future consumption. The GPI adjusts con-
sumption spending to account for net foreign liabilities.
Natural capital Maintaining the stocks of natural capital is essential to
sustaining consumption in the future, especially when consumption is
defined more broadly. These stocks take two forms. The first are stocks of
renewable and non-renewable resources used as inputs in production, such
as minerals, fossil fuels and soils. The second take the form of waste sinks
that are provided by the natural environment and are essential for dissipat-
ing waste products so that they do not represent a danger to humans. The
GPI takes account of the depletion of both types of natural capital.
However there are some difficult methodological issues concerning the sub-
stitutability of built for natural capital that are discussed in the next section.
Human capital This represents the accumulation of health, skills, know-
ledge and experience in humans that makes them more productive than brute
labourers. Technology is partly embodied in humans. The GPI does not
account for human capital because of the conceptual and measurement
difficulties involved. If it did, the GPI would ideally be adjusted to account
not for annual investments in human capital but for the annual services
provided by the stock of human capital. This is an area for future work.
Social capital A nation that possesses sound and stable political, legal
and commercial institutions and cohesive, supportive and trusting com-
munities will be in a better position to generate flows of goods and services
than one that does not. However, this form of ‘capital’ is difficult to define
precisely and to measure and is for that reason excluded from the GPI.
Substitutability among capital assets
The depletion of one form of capital does not represent a decline in sus-
tainable consumption if other forms of capital are accumulating and can
be substituted for the disappearing asset. Thus the issue of substitutability

Measuring sustainable economic welfare
within and between these classes of assets is critical. For instance, the run-
down in physical capital is not necessarily a problem if financial wealth that
could be used to rebuild it (or could be used to invest in assets in other coun-
tries) is being accumulated outside of the country.
More controversially, the run-down of one type of natural asset will not
necessarily impose a cost if built capital or another type of natural asset can
perform, at the same or similar cost, the same functions. The question of the
degree of substitutability of built for natural capital is perhaps the most
strongly contested issue in the economics of the environment (see Chapters
3, 4 and 6 in this volume and, for example, Neumayer, 2003). We have taken
the view that for three classes of natural assets complete substitutability
between built and natural assets is not a valid assumption. These classes are:

Certain natural resources that are irreplaceable and form essential
inputs to continued productive activity – soils and supplies of fresh
water are examples;
Waste sinks, that is those components of the natural environment that
absorb or process wastes and render them benign, particularly the
atmosphere (covering the climate system and the ozone layer) and the
oceans; and
Assets whose services are consumed directly by final consumers
and which are valuable because of their unique natural features –
old-growth forests and coral reefs are examples.
In addition to these, there may be some natural resources for which there
are, or probably will be, substitutes, but for which the substitutes are likely
to be significantly more expensive. Fossil fuel-based energy is the most
pertinent category here. Energy is essential for economic activity, yet the
evidence suggests that the market for energy may not adequately reflect the
likely scarcity of fossil fuels (especially oil and natural gas).
Neumayer (2000) has argued that the fact that changes in the value of
these ‘non-substitutable’ assets are added in the GPI to other consumption
goods makes them substitutable, so that the GPI is an indicator of weak
sustainability only. But adding the value of haircuts to the value of oranges
in calculating GDP does not make them substitutes for each other. He also
argues, correctly, that some ISEWs or GPIs use an erroneous method to
value the depletion of natural resource stocks, which tends to exaggerate
the difference between GDP and the adjusted welfare measure.
Defensive expenditures
Whereas GDP counts them as additions to output, the GPI deducts defen-
sive expenditures undertaken by consumers and governments because, by Handbook of sustainable development
definition, they are undertaken to offset some decline in social welfare. In
principle, most defensive expenditures are reactions to a decline in the value
of the stock of social, human or natural capital, as long as they are broadly
defined. This applies to private defensive expenditures on health and per-
sonal security and public defensive expenditure on social welfare. If we
could adequately account for changes in stocks of human and social capital
then it would not be necessary to deduct defensive spending.
A more difficult question is that of how much of a given expenditure is
defensive and how much makes a net contribution to welfare (Neumayer,
1999). This is particularly relevant to some public expenditures, on social
security and law and order for instance. An increase in spending on polic-
ing, courts and prisons due to a crime wave is clearly defensive, yet some
basic level of spending on crime prevention and punishment is essential and
makes a large contribution to national well-being. Ultimately judgements
about how much spending is defensive and how much makes a positive con-
tribution to welfare will be somewhat arbitrary.
Time accounting
The GPI attempts a systematic approach to valuing time.6 The value of time
is a very important aspect of various components of the GPI, including the
value of household and community work and the costs of unemployment
and of overwork. In the Australian GPI we have adopted the principle that
the value of time devoted to voluntary activities counts as a positive in the
GPI and the value of time engaged in involuntary activities counts as a
negative. The following voluntary activities contribute to our welfare:
paid work (except the involuntary component referred to below as involuntary leisure,7 that is the times when we are unemployed but
want to be employed; and involuntary work, that is the times when we are doing paid work but would prefer not to be.
The GPI is a measure of sustainable consumption. Thus in addition to
measures of currently consumed goods and bads – including the costs of
crime, the costs of commuting, the benefits of household work, and the dis-
tribution of income – it considers the future implications of present con-
sumption (and production) activities. Thus it incorporates an estimate of
the unsustainability of foreign debt, indicated by the proportion of total
foreign borrowing that finances consumption rather than investments that
can generate revenues to be used to repay the debt. It also considers the
long-term impact of economic activities on the stocks of irreplaceable
natural capital assets. In this way, future costs are in a sense brought
As a result, while graphing GPI per person over time illustrates the direc-
tion of change, caution must be exercised in interpreting the GPI measure
in any one year as a measure of national welfare in that year. Just as a con-
sumer can increase their consumption levels and thus ‘welfare’ by spending
up on a credit card, credit binges must be paid for by lower consumption in
future years. Neumayer (1999) has observed that the GPI/ISEW cannot
function simultaneously as an indicator of current welfare and as an indi-
cator of sustainability. While there is some confusion in the GPI literature
about what it does measure, it seems agreed that the GPI does not function
as an indicator of current welfare and of sustainable income but as an indi-
cator of sustainable welfare. In other words, it measures what we might call
‘Hicksian welfare’, the maximum amount of welfare that a nation can
enjoy over some time period and still be as well off at the end of the period
as at the beginning.
The GPI therefore engages in a type of smoothing process. As a result,
we take the view that it may be misleading to construct the GPI on an
annual basis (and even more misleading to do so on a quarterly basis) if
the impression were given that an increase, say, in the GPI from one year
to the next indicated that national well-being had risen by that amount. On
the other hand, many of the items included in the GPI are current rather
than capital items and do indicate year-on-year changes in well-being.
The results of three GPI/ISEW calculations are shown in Figure 19.1.
5. Conceptual problems in the GPI
While many people have welcomed the GPI, others have raised objections.8
There are four objections to the GPI that have been raised, the first three of
which are misconceived.
‘Subjective weighting’
It is often claimed that the ‘weighting’ of various components in the GPI is
subjective. In fact, the GPI uses a range of techniques to attach dollar
values to the various components, thus converting every component into a
common unit of measurement. For instance, the value of household labour
is arrived at by multiplying the number of hours worked in the household
by the hourly wage rate of a housekeeper. The value of the loss of ozone is
arrived at by assessing the health costs of the damage caused. These are not
subjective ‘weights’ but are dollar values generated in markets of one sort
or another – actual markets, related markets or hypothetical markets.
Everything is expressed in dollar values via prices generated in markets, so
that the weights look after themselves.
Arbitrariness of components
Some critics argue that the GPI lacks a sound theoretical foundation; as a
result the inclusion of various components is arbitrary (Neumayer, 1999).
While the rationale has not always been clearly stated in previous GPIs and
ISEWs, the selection of components is not arbitrary but follows some rules.
The process begins by identifying the deficiencies of GDP as a measure of
welfare and asks how it would need to be changed to make it a better
measure. In so doing, it builds a framework for measuring sustainable
Thus the GPI is not ‘arbitrary’ in the sense that its authors simply add in
components at random. In each case, there is an identified problem with
GDP as a measure of welfare, and an attempt is made to fix it so far as is
permitted by availability of data. When statisticians calculate NNP by sub-
tracting an estimate of the depreciation of built capital from GNP and say
that it is a better measure of changes in output, we do not accuse them of
being arbitrary; they are correcting for a known problem.
Quality of goods
It is sometimes argued that the GPI fails as a measure of changes in
national well-being because it does not account for the improvement in the
quality of goods. Thus real consumption spending may double over a given
period, but the utility derived from that spending may more than double
because the quality of goods has improved. For example, in real terms we
may pay the same amount for a TV today as we did 20 years ago, yet the
benefit we derive is much higher because the set has a bigger and flatter
screen, and the quality of picture and sound are better.
This is true; however, exactly the same criticism applies to the use of GDP
as a measure of national well-being, so it should be no surprise that since
the GPI begins with the final consumption component of GDP, all of the

Handbook of sustainable development
problems in it will be carried over to the GPI. Arguably, the quality problem
‘cancels out’, so that if we focus our attention on the gap between GDP and
GPI then it is perfectly feasible to maintain that the GPI is a better measure
of changes in economic well-being. This does, however, temper the useful-
ness of conclusions drawn on the basis of changes in the GPI over time.
Ethical versus economic values
There is one serious problem with the GPI as a measure of national well-
being that should be acknowledged. Aggregating all of the factors into a
single monetary index strikes many people as being invalid. By converting
everything into dollars, doesn’t the GPI fall into the same trap as GDP, that
of reducing well-being to economics? This is perhaps the major flaw in the
GPI. The problems with the approach become apparent when we attempt,
for example, to estimate the costs of climate change, since the greatest costs
will be associated with loss of life, which must be given a dollar value if it
is to be included. Should the life of a person in a poor country be worth less
than the life of someone in a rich country? Placing dollar values on many
things converts ethical values into economic ones, a process that for many
people actually devalues the environment and human life (see Chapter 2 on
this issue). These profound problems with the GPI are acknowledged. For
some, constructing the GPI is the most effective way of pointing to the fail-
ings of current systems of measurement. Moreover, refusing to value some
things means they must be left out of the GPI, even though it is generally
agreed they affect our well-being.
6. Areas for future refinement of the GPI
There are a number of areas of future work that will help refine and resolve
difficulties in the GPI method. They include:
Employing better measurement of changes in income distribution
over time, including more robust estimates of the social preference
for equality, or aversion to inequality;
Development of a more comprehensive natural resource accounting
framework for incorporating environmental and resource use
impacts in to the GPI;
Securing the collaboration of various government agencies in pro-
viding the best and most consistent data on a number of variables
(for example those components affected by transport including
urban air pollution, costs of noise, costs of accidents); and
Using a full capital depreciation framework for the GPI components,
that is evaluation of the elements of human and social capital and
valuing changes in these stocks.
1. I would like to thank the editors for very helpful comments on an earlier draft of this
2. Including the UK (Jackson and Marks, 1994; New Economics Foundation, 2004),
Canada (Coleman, 1998), Germany (Diefenbacher, 1994), Sweden (Jackson and Stymne,
1996) and Australia (Hamilton, 1997; Hamilton and Denniss, 2000).
3. Hicks also wrote that ‘the practical purpose of income is to serve as a guide for prudent
conduct’ (Hicks, 1946: 172), a comment that has particular relevance for today’s concern
with ecological sustainability.
4. For a formal treatment of the roles of relative incomes, aspirations and environmental
quality in welfare see Ng and Wang (1993).
5. In the case of debts owed to domestic creditors, increased consumption now by the debtor
is offset by a decline in consumption now by the creditor, a situation that is later reversed.
6. The most systematic attempt to sort out the problem of time valuation in the GPI appears
in Hamilton and Denniss (2000).
7. Some GPIs include the cost of (lost) leisure. Others include the costs of overwork instead.
8. See for example Castles (1997) and Neumayer (1999, 2000).
Anielski, M. and J. Rowe (1999), The Genuine Progress Indicator – 1999 update, San Francisco:
Redefining Progress.
Castles, I. (1997), ‘Measuring wealth and welfare: why HDI and GPI fail’, paper to a sympo-
sium on Wealth, Work and Well-being, Academy of the Social Sciences in Australia
(10 November).
Cobb, C. and J. Cobb (1994), The Green National Product: A Proposed Index of Sustainable
Economic Welfare, Maryland: University Press of America.
Cobb, C., T. Halstead and J. Rowe (1995), The Genuine Progress Indicator: Summary of Data
and Methodology, San Francisco: Redefining Progress.
Coleman, R. (1998), ‘Measuring sustainable development: the Nova Scotia genuine progress
indicator’, report published by GPI Atlantic, Nova Scotia, Canada.
Daly, H. and J. Cobb (1990), For the Common Good, Boston: Beacon Press.
Diefenbacher, H. (1994), ‘The index of sustainable economic welfare in Germany’, in C. Cobb
and J. Cobb (eds), The Green National Product, Lanham, MD: University of Americas Press.
Frey, B. and A. Stutzer (2002), Happiness and Economics, Princeton: Princeton University Press.
Hamilton, C. (1997), ‘The genuine progress indicator: a new index of changes in well-being in
Australia’, Australia Institute Discussion Paper No. 14 (October) (with contributions from
Hugh Saddler).
Hamilton, C. (1999), ‘The genuine progress indicator: methodological developments and
results from Australia’, Ecological Economics, 30: 13–28.
Hamilton, C. and R. Denniss (2000), ‘Tracking well-being in Australia: the genuine progress
indicator 2000’, Australia Institute Discussion Paper No. 25 (December).
Hicks, J. (1946), Value and Capital, 2nd edn, London: Oxford University Press.
Jackson, T. and N. Marks (1994), Measuring Sustainable Economic Welfare – A Pilot Index:
1950–1990, Stockholm: Stockholm Environment Institute.
Jackson, T. and S. Stymne (1996), Sustainable Economic Welfare in Sweden: A Pilot Index
1950–1992, Stockholm: Stockholm Environment Institute.
Jackson, T., N. Marks, J. Ralls and S. Stymne (1997), Sustainable Economic Welfare in the UK
1950–1996, Guildford, Surrey: Centre for Environmental Strategy, University of Surrey.
Layard, Richard (2005), Happiness: Lessons from a new science, New York: Penguin.
Neumayer, E. (1999), ‘The ISEW: not an index of sustainable economic welfare’, Social
Indicators Research, 48: 77–101.
Neumayer, E. (2000), ‘On the methodology or ISEW, GPI and related measures: some con-
structive suggestions and some doubt on the “threshold” hypothesis’, Ecological Economics,
34(3): 1–34.

Handbook of sustainable development
Neumayer, E. (2003), Weak versus Strong Sustainability: Exploring the Limits of Two
Opposing Paradigms, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
New Economics Foundation (2000), ‘Chasing Progress: Beyond measuring economic growth’,
Ng, Yew-Kwang and Jianguo Wang (1993), ‘Relative income, aspiration, environmental
quality, individual and political myopia’, Mathematical Social Sciences, 26: 3–23.
Nordhaus, W. and J. Tobin (1972), ‘Is growth obsolete?’, in National Bureau of Economic
Research, Economic Growth: Fifth Anniversary Colloquium, New York: NEBR.
Pannozzo, L. and R. Colman (2004), ‘Working time and the future of work in Canada: a Nova
Scotia GPI case study’, Report published by GPI Atlantic Nova Scotia,
20 Environmental space, material flow
analysis and ecological foot printing Ian Moffatt
1. Introduction
The terms ‘sustainability’ and ‘sustainable development’ are often used
interchangeably in both academic research and policy making. They are,
however, different, and should be defined clearly and used carefully. To
sustain an activity or process is to ensure that the system runs for a long
time. In environmental and ecological economics a sustainable resource is a
potentially renewable resource which can be used indefinitely. The word
‘sustain’ is often used in the context of maximum sustainable yield (MSY)
and has been used for understanding and contributing to resource policy in
areas such as multi-species forestry and fisheries management (Clark, 1976;
Christensen, 1995). Sustainable development is a broader concept than sus-
tainability and stresses both the idea of sustaining activity for a long time
for current and future generations as well as linking such activity to devel-
opment rather than economic growth per se. It is also vital for development
to be sustainable that the life support systems of the planet are protected
(WCED, 1987). One thing is certain, you cannot have continuous growth of
economies, population, resource consumption and pollution generation on
a planet with finite biophysical stocks and limited assimilative processes
(Daly, 1972). This was noted over three decades ago at the Stockholm con-
ference on the Human Environment (Ward and Dubos, 1972) and at the
summits in Rio de Janeiro (1992) and in Johannesburg (2002). Sustainable
development is an on-going process integrating ecological, economic, equity
and ethical considerations for current and future generations of people and
other living creatures, without endangering the life support systems of the
planet upon which ultimately all life depends (Moffatt, 1996a).
This chapter examines environmental space, material flow analysis and
ecological footprints as contributions to the processes of achieving the goal
of sustainable development. The next section discusses weak and strong
sustainable development issues and resource use. Sections 3 to 5 examine
environmental space, material flow analysis and ecological footprinting
respectively. Each section defines the concept, briefly describes the method-
ology including for brevity a ‘master equation’ for the concept, and illus-
trates its application with examples. Section 6 then subjects the three

Handbook of sustainable development
methods to a critical assessment with regard to contemporary research
problems and policy relevance. The final section provides a summary of the
positive findings.
2. Weak and strong sustainable development and resource use
The three approaches described in this chapter are all based on the idea of
strong as opposed to weak sustainable development. Whilst we accept that
if a country is unable to pass a weak test for sustainable development then
it is unlikely to pass a stronger test (Pearce and Atkinson, 1993; Atkinson
et al, 1997; Neumayer, 2003) – the weak test is underpinned by very ques-
tionable and debatable assumptions of resource use (Beckerman, 1998;
Daly and Cobb, 1989). These include the assumption of perfect substi-
tutability between man-made (Km) and natural capital (Kn); setting the
correct price for specific resource use which is often not included in the
market and the role of technical change in areas where there may be no
technical solutions. Weak sustainable development is generally based on
neo-classically derived marginal analysis at the resource frontier rather
than on absolute limits (Mirowski, 1990). Furthermore, it could be argued
that the weak sustainability argument assumes that the ecology is sub-
servient to economics. If, however, we are to assume that economics is a
subset of ecology then we must consider strong sustainability.
Strong sustainability is based on several principles of classical science.
These recognize the fact that we only have one earth and that for sustain-
able living we have to live within its absolute biophysical limits. From the
principles of conservation of matter we cannot make matter but we can
change its form. From the laws of thermodynamics we cannot get any more
energy from a machine than we put into it. The earth ecosystems derive the
bulk of their energy from solar radiation, and in open living systems energy
consumption is hierarchically organized to maintain higher-level organ-
isms in ecosystems. From ecology we cannot expect a receiving environ-
ment to exceed its assimilative capacity without increasing levels of
pollution above a natural level. The proximity and precautionary principles
are also included in strong sustainability arguments. The differences
between the weak and strong perspectives are shown in Table 20.1.
Resources are a term of cultural appraisal (Kirk, 1963) and depend in
part on a society’s technology and on the political choices to use resources
or leave them untouched as part of nature. The indigenous Aboriginal
peoples of North Australia, for example, did not use metal as mining the
earth was seen by some tribes as desecrating the land in which their God
resides. They feared that such activity could result in divine retribution.
During the Roman occupation of Britain ( 120) coal, formed in the
Carboniferous period about 350 million years ago, was used for making jewellery. Later in the Industrial Revolution (circa 1790) the Carboniferous
capitalists used coal to fuel industrial production (Rees, 1985). In this sense
natural resources are neutral stuff which may become useful for different
purposes (Zimmermann, 1951).
When examining sustainable development it is conventional to describe
resources as a stock (that is a physical quantity) or as a flow (that is rate of
use). It is also essential to note that the use of any resource leads to waste
which, in the earth’s closed and inter-related biogeochemical cycles, gener-
ally impacts on other ecological cycles. Most of the potentially living
resources – the life support systems of the planet – depend on incoming
radiation from the sun and matter from the earth. If we are to use poten-
tially renewable resources in a sustainable manner then we must ensure that
the rate of harvesting (or fishing or hunting) is much less than the natural
rate of reproduction. Next, that the rate of pollution and waste generation
is less than the natural assimilative capacity of the receiving environment.
For strong sustainable development we need to ensure that the man-made
capital resulting from the use of the non-renewable resources (for example
minerals, fossil fuels) are set aside to fund renewable alternatives (Daly,
1990). We should also strive to minimize the damage to the environment
which always accompanies resource use. The methods underpinning envir-
onmental space, material flows and ecological footprints assume that these
ideas are well understood.
3. Environmental space
Environmental space is defined as a share of the planet and its resources that
the human race can sustainably take without depriving future generations
of the resources they would need. The idea of environmental space was first
put forward in 1994 (Opschoor and Weterings, 1994). It describes the quan-
tity of energy, non-renewable (for example minerals) and potentially renew-
able resources (for example water, food, wood, farmland) that we can use in
a sustainable fashion without exceeding environmental limits (McLaren et
al., 1998, p. 6). It was argued that at the current rates of use non-renewable
resources would have a short life, that the use of potentially renewable
resources would result in overexploitation and that the assimilative capacity
for waste would be exceeded unless reductions in resource use occur. The
second major assumption underpinning environmental space is the idea of
equity for current and future generations. This was, of course, noted in the
Brundtland definition of sustainable development (WCED, 1987). In envi-
ronmental space equity is defined as an equal per capita share of resources.
Environmental space was used in both national and European-wide
studies. The original Netherlands study defines environmental space as
estimating the global resource (such as wood energy, water, raw materials,

Environmental space and ecological footprinting
arable land) and dividing it by the number of world citizens, to produce an
average figure for each resource per capita for a given date. By comparing
the global average per capita figure for a given resource with the total of
that resource consumed in a particular country then the amount of envi-
ronmental space consumed by a nation can be observed. The test for sus-
tainable development using environmental space is ‘the use of resources
and pollution of that country can be compared to the environmental space
belonging to that country’ (Buitenkamp et al., 1991, p. 18). The calculation
for environmental space is simple and is given in equations 20.1 and 20.2.
The environmental space for country i is the amount (Q) of consumption
of resource x per capita. Then a country’s consumption of one resource
(ESi,x,t ) can be compared to the environmental space of global resource use
(ESx,t). The policy prescriptions which follow from the calculations of envir-
onmental space are based on a comparison of one nation’s resource use of
type x with the global average. Naively, if ESi,x,t ESx,t then policies should
be implemented to reduce resource use of type x in country i. If ESi,x,t
ESx,t then presumably no reductions are necessary. If ESi,x,t ESx,t do policy
makers increase resource consumption in country i?
The environmental space concept was actively pursued by Friends of the
Earth groups across Europe (Friends of the Earth, 1995a, 1995b). In a
series of national reports the environmental space required for countries in
a sustainable Europe as described (Buitenkamp et al., 1991; McLaren et al.,
1998). By 1996 reports on sustainable Europe and some nations within
Europe had been published (Tables 20.2 and 20.3). Generally, these studies
argued that Europeans are consuming more than our fair earth share of
environmental space and that we would have to undertake massive cuts in
resource use by 2050. To achieve these large cuts a per capita reduction
target was established for each resource for 2010. Whilst such ideas are
useful as a guide to policy they are only useful if the underlying basis for
such proposed cuts is sound; alas, even in the important example of atmos-
pheric carbon dioxide reductions, this was not the case (see section 6).
It will be observed that in both the European-wide and national studies most
of the resources have to be drastically reduced. In one sense this research effort
by numerous groups was to be welcomed as a bold statement of the degree of
unsustainability different European countries exhibited. The original
Netherlands study was set up to encourage debate over sustainable develop-
ment. This debate must not, however, ignore the technical details in the
methods used. It is wrong to assume that these technical details ‘should in no
case to be allowed to delay the debate on the consequences of the concept of
limited and finite environmental space for daily life in society’ (Buitenkamp et
al., 1991, p. 181). This is methodologically unacceptable because if the method
is wrong then the policy prescriptions offered would carry very little or no con-
viction. Quite simply if you divide the resource consumption by the global
population you obtain ‘environmental space per capita’ for a given time. As the
global population grows, the ‘environmental space per capita’ share is reduced
and, on this basis, it could be argued that countries need to control global pop-
ulation growth as well as reduce resource consumption. We will return to crit-
icisms of the environmental space method in section 6. Finally, whilst the idea
of environmental space was poorly conceived, it did point the way to more
rigorous methods such as material flow analysis and ecological footprinting.
4. Material flow analysis
The purpose of material flow analysis is to track and quantify the flow of
materials including energy in a defined area over a set time period. It is
obvious that any economy takes in raw materials from the environment
including imports from foreign nations, for further processing, manufac-
turing, production and consumption (Linstead and Ekins, 2001). Some
materials such as the construction of buildings and infrastructure add to
the stock of man-made capital. Eventually, the products become waste and
may be recycled, but finally have to be disposed via landfill or incineration.
Since any resource input sooner or later becomes an output, it is possible
to account for resource flows and use them in material balance modelling
(Figure 20.1).
The mass balance equations used in material flow analysis (MFA) can be
written as follows:
Obviously, collecting all the relevant data for each of items A, B, C and D
is a difficult and time-consuming task. Fortunately, the Statistical Office of
the European Communities has developed national economy-wide mater-
ial flow accounts (Eurostat, 2001). These accounts exclude water and air
but include energy flows through the national economy. Several studies
have been undertaken at the national scale to give an empirical account of
resource use (Linstead et al., 2004). In the United Kingdom, for example,
a material flow analysis using resource use in agriculture, forestry and
fishing together with mining of minerals, fossil fuels and other aggregates
was undertaken in 2002. The calculation also includes ‘hidden’ flows of
materials such as mining wastes which are moved during extraction but are
not used directly in the economy. In the UK for the period 1970–2000 it was
shown that the total resource use rose during the 1970s as oil and gas pro-
duction from the North Sea reserves started to flow, but eased off during
the early 1980s. Generally, there has been an increase in material flows, in
line with economic growth, in the latter part of the 1980s, but from 1990
onwards resource use has stabilized despite a considerable increase in the
size of the UK economy (Sheerin, 2002). From 1990–2001 GDP in the UK
has increased by 28 per cent yet the Total Material Requirement (TMR)
increased by 7 per cent, the Direct Material Input (DMI) remained con-
stant and the Direct Material Consumption (DMC) fell by 10 per cent
(DEFRA, 2003).
Material flow accounting can also be used at sub-national scales at either
the level of individual business enterprises, or at specific sectors of the
economy such as mineral resource use in North-West England (NCBS,
undated) or at the city and regional scales (Ravetz, 2000). In 2002 a study
of material flow analysis and ecological footprint (see later) in York was
published (Barrett et al., 2002). Although the researchers acknowledge that
both the fossil fuel carriers and hidden flows (such as the overburden left at
the site where minerals are mined) may have been underestimated (30 per
cent and 35 per cent less than the UK average respectively) they give a good
account of the material flow in the urban economy. In 2000 the total mate-
rial requirement of York was 3 387 000 tonnes; an average of 18.8 tonnes
per person for each York resident.
Just under half of this was material that entered the city, the rest being either
energy carriers (579 000 tonnes) or hidden flows (1 231 000 tonnes). The major-
ity of the material flows into York are due to the construction of houses and
roads (approximately 67 per cent). The stock of materials in York increased by
over 1 million tonnes. On the output side, over 250 000 tonnes of materials left
York or were deposited in landfill sites and nearly 70 000 tonnes were recycled.
Over 4.5 million tonnes of greenhouse gases were produced (Barrett et al.,
2002, p. xiv).
An imaginative study of South-East England has used material flow
analysis to explore different scenarios of development and waste reduction.
In 2000 the South-East region generated 36.8 million tonnes of waste
(53 per cent construction and demolition, 19 per cent industrial and com-
mercial, 16 per cent agricultural, 11 per cent household and 1 per cent
other). Whilst the different sectors do use the waste hierarchy (recycle,
recover and reuse some of the resources) it is estimated that waste is
growing at 1–3 per cent per year and could double in 25 years. This growing
problem was examined by a material flow analysis combined with an explo-
ration of four different scenarios. The four scenarios of waste generation
were: a high growth of 3 per cent per year; a Business as Usual 2 per cent
per year, a zero growth and a ‘factor four’ rapid minimization scenario
beginning with 3 per cent growth and tapering to 3 per cent, giving a net
decline in waste of 14 per cent by 2020.
Unsurprisingly, the waste minimization scenario results in less waste but
implementing such a strategy is a major task especially as economic and
demographic growth is forecast for the South-East region of the UK
(Anon, undated).
One of the policy drivers in material flow analysis is the idea of ‘Factor
Four’ reductions in resource use to half resource use and double output
(Ayres, 1978; Weizsacker et al., 1997). The scientific basis for this factor X
(where X is any positive real number) argument is very suspect (Robert
et al., 2000). It will be noted that Figure 20.2 simply shows a hypothetical
monotonically declining function for resource use. Obviously, if you
increase non-renewable resource consumption by any amount then the
quantity of resources will decline. In a series of papers Schmidt-Bleek
asserts, without any proof, that we need to make a 50 per cent cut in mate-
rials inputs advanced economies (or more if population growth is taken
into account) (Schmidt-Bleek, 1992; 1993a; 1993b). It is this assertion,
coupled with the view that technical solutions to resource efficiencies can
be implemented, that colour the thinking in this area of material flow
analysis. As a policy instrument this untested idea has had some support in
the advanced industrial nations. Whilst it is good to see innovative ideas
being produced to address the problems of unsustainable consumption of
commodity production we must, however, temper this enthusiasm for every
new idea with careful criticisms (see section 6 below).
The ecological footprint concept has captured the imagination of academics,
decision-makers and the public because it can be measured, is easy to under-
stand and it has a resonance with different scientists, policy makers and other
members of the public. There is a large and rapidly growing literature
concerned with ecological footprinting. It has been the focus of academic
scrutiny (Ayres, 2000; Haberl et al., 2004); the basis for many empirical
studies as an indicator of strong sustainability (Wackernagel and Rees, 1994)
and is being examined by both governments in Europe and businesses as a
sustainability indicator (Chambers and Lewis, 2001). The question for aca-
demics and policy makers is not just whether the footprint is attractive but
whether it is internally consistent and whether it helps as an input into poli-
cies which are designed to make development sustainable in the early years
of this century.
The ecological footprint concept can be defined as the total area required
to indefinitely sustain a given population at the current standard of living

Handbook of sustainable development
and at an average per capita consumption rate. The original idea of
ecological footrpinting was proposed by Rees in a study of cities which
consume vast amounts of resources (Rees, 1992). In 1994 this concept was
developed and illustrated from work in Canada (Wackernagel and Rees,
1994). Over the last ten years the early methodology of ecological foot-
printing has been substantially altered partly in response to well inten-
tioned criticisms (see Ecological Economics, 2000, Vol. 32). Essentially it is
a measure of land per person – not a density – but an expression of how
much of the earth’s surface is required to support an average person in a
specific area. More precisely, ecological footprint accounts measure the
amount of the earth’s biological productivity that a human population –
global population, a country, a city or an individual – occupies in a given
year using prevailing technology no matter where that land is located. This
methodology has now been established by setting up a global forum so that
standard methods can be used in substantive studies. The National
Footprint Accounts (NFA) constitute the underlying methodology with
which ecological footprints have been calculated for 149 countries of the
world (WWF, 2004). A detailed description of the NFA methodology has
been presented in 2004 (Monfreda et al., 2004) and also from the Global
Footprint Network (Wackernagel et al., 2004a). The unit of measurement
is the biologically productive area, termed the global hectare (gha), which
represents an equal amount of biological productivity. The gha is normal-
ized so that the number of actual hectares of bioproductive land and sea
on the earth is equal to the number of global hectares on this planet. To
calculate the biocapacity of a nation, each of six different types of biopro-
ductive areas within a nation are multiplied by both an equivalence and
yield factor for that land type. The six bioproductive areas are:
Crop land for food and animal feed, fibre oil crops and rubber;
Grazing land for animals for meat, hides, wool and milk;
Forest area for harvesting timber or wood fibre for paper;
Fishing grounds for catching fish;
Built-up areas for accommodating infrastructure for housing, trans-
port and industrial production;
Land for sequestering the excess CO2 from burning fossil fuels to
replace it with biomass, for harvesting fuelwood, for nuclear energy
and for hydropower (Wackernagel et al., 1999).
The hectares for each type of bioproductive area are converted into
global hectares (gha) by multiplying an equivalence factor (to represent the
world’s average potential productivity of a given bioproductive area or land
cover type) with yield factors (to capture the difference among local and
global average productivity). The biocapacity for an area constitutes the
supply side of the equation and the aggregate human demand (ecological
footprint) can then be compared. Whilst an individual nation’s area
demand can exceed supply it is obvious that to live ecologically sustainably
on the earth we must live within the earth’s biocapacity. If we exceed this
limit then we do so by depletion of natural capital (Kn). An individual
nation can also exceed its biocapacity by depletion of Kn and by imports
(ecological trade deficit). Obviously all nations cannot continue to live in
ecological deficit and be ecologically sustainable.
Essentially the ecological footprint ‘master equations’ for ecological
footprinting can be written as a supply and demand identity .The supply is
given as:
The right-hand side of equation (20.4) is the summation of each area in
hectares multiplied by the equivalence factor multiplied by the yield for
each land cover class. Equation (20.5) represents the ecological footprint or
demand side of the identity. Again in a general form the area given to a land
cover type is multiplied by the equivalence factor divided by yield per ha
and then summed for all six bioproductive areas. The equivalence factor
represents the world average potential productivity of a given bioproduc-
tive area relative to the world average potential productivity of all biopro-
ductive areas (Wackernagel et al., 2004b, p. 262; For full details of the
method see Wackernagel et al, 2004a).
Once the ecological footprint is calculated by summing all the resources
used in a country or area it can be compared to the demand with the actual

Handbook of sustainable development
country or area also expressed in global hectares. Three policy prescrip-
tions follow from an ecological footprint analysis. First, if a country’s foot-
print is greater than its area measured in global hectares then some
reductions in resource consumption are required. If the ecological foot-
print demand is equal to or less than the supply, then one condition for
ecological sustainability is being met and presumably the socio-economic
practices are within the ecological footprint and are therefore contributing
to sustainable development.
Currently, ecological footprints can be calculated using an aggregate or
compound approach or alternatively a component approach using an
index. The compound approach uses national data to determine the
average person annual consumption (national data /total size of popula-
tion) whilst the component approach builds up the economy by different
sectors using an index (Simmons et al., 2000) and can be more useful for a
range of policies. These two approaches are complementary and have been
used in many studies (Wackernagel and Rees, 1994; Chambers et al., 2000;
Haberl et al., 2004). Studies have been undertaken including global scale
with the Living Planet Index (Loh, 2002; WWF, 2004; Wackernagel et al.,
2002). National studies of the economy of Australia (Lentzen and Murray,
2001), Austria (Haberl et al., 2004), UK (Barrett and Simmons, 2003)
Canada, Chile, Italy, the Philippines and South Korea (Wackernagel et al.,
2004b), Scotland (Best Foot Forward, 2004) and Wales (Best Foot Forward,
2002b; WWF, 2005a) as well as Benin, Bhutan, Costa Ricas and the
Netherlands (van Vuuren and Smeets, 2000) have been completed. Urban
studies including London (Best Foot Forward, 2002a), Liverpool (Barrett
and Scott, 2001), as well as regional studies of Guernsey (Barrett, 1998), Isle
of Wight (Best Foot Forward, 2000c), the South-East England (Anon.,
undated) and Tuscany (WWF, Italia, 2004) have been published. The
redesigned ecological footprints methodology now consists of a 2000 rows
by 100 columns spreadsheet for each country. The integration of ecological
footprinting accounting into standard economic models allows systematic
evaluation of policy options as extensive scenario analysis becomes possi-
ble. It opens up possible links with UN National Statistical accounting
which offers a consistent time series from 1970 for all UN member states
and all other countries in the world. The relevant data can be found
at ( This permits the
integration of ecological footprinting with input–output analysis to allocate
ecological footprints and material flows to final consumption (Wiedmann et
al., 2005; Moffatt et al., 2005). It also opens up the prospect of integrating
input–output and dynamic modelling to explore future scenarios at different
geographical and hierarchical scales (Moffatt et al., 2001; Kratena, 2004;
Faucheux and O’Connor, 1998; and Faucheux et al., 1999).

Environmental space and ecological footprinting
The original ecological footprint was a one-shot or static review but it is
important to explore scenarios of different development paths in a dynamic
context. This approach has been used in the study of North America using
the ecological footprint scenario model (EFSM). The researchers note that
if North Americans want to maintain their lifestyles and their correspond-
ing levels of consumption while avoiding ecological deficits, then the pro-
ductive capacity of all ecosystems would have to at least double and be
coupled with a reduction in economic growth or its accompanying spend-
ing (Senbel et al., 2003, p. 90). The summary results from their work indi-
cate that reducing consumption has the most significant impact on the
ecological footprint. Without such changes in reduced consumption and
increased resource productivity, then ‘North Americans will increasingly
live in a continent of ecological deficits’ (Senbel et al., 2003, p. 92). Given
that natural and agricultural ecosystems can not continue to double their
productivity, the future looks dismal. A second study using the IMAGES
model to examine responses to global warming has indicated that the
global ecological footprint will not exceed 15 billion gha in 2050. This sce-
nario assumes that there are changes to production and reduced consump-
tion in the rich nations and economic growth in low-income areas even with
population growth (van Vuuren and Bouwman, 2005). Unfortunately, the
planet has only 11.3 billion gha, which means that even if this scenario is
followed we are still overshooting the ecological limit by approximately 33
per cent (WWF, 2005a). The message from these recent ecological footprint
studies is clear: we must reduce resource consumption if we are to live
within the ecological limits of this earth.
6. A constructive critique of the methods
The three methods described in this chapter are being developed by a variety
of individuals and groups and are being promoted as contributions to mea-
suring sustainable development. This is an important aspect of research, as
measurement can help avoid self-deception and can contribute to policy.
Indeed the UK sustainable development strategy notes the need for innova-
tive ways to measure sustainable development as well as for ways of con-
tributing to policy initiatives (Cm 6467, 2005). Ideally, these methods should
be applicable at different spatial scales and through different organizations,
for example Governmental, businesses and local communities so that we can
all contribute to the process of making development sustainable. Given the
importance of the issues we are involved in, such as maintaining the life
support systems and improving the quality of life for all inhabitants on the
planet, it is essential that we develop robust methods. This section offers
some constructive criticisms of each of the methods and then some more
general comments on how we can make progress in this area of research.

Handbook of sustainable development
Environmental space
There are numerous criticisms that can be made of the environmental space
concept (Moffatt, 1996b). First, from a scientific perspective it is difficult to
know with any precision the amount of non-renewable resources available
at current technologies and prices. Next, it is generally agreed that human
economic activity is putting strains on many environmental systems. It
would therefore seem sensible to reduce resource consumption but Friends
of the Earth Europe have relied on a weak argument for environmental
space to propose draconian reductions in resource use. Third, environmen-
tal space assumed greater certainty about waste assimilation processes than
is currently scientifically known. It is exceedingly difficult to establish such
limits in an accurate manner. The major exception is the atmospheric
assimilation of carbon dioxide (CO2). In this case it can be shown scientif-
ically that anthropogenic emissions to the stock of atmospheric CO2 have
grown from 276 ppmv in 1790 to over 360 ppmv by 2000 (Gorshkov, 1995).
Yet, applying the environmental space concept to this major problem yields
misleading results and, if accepted, could give rise to misguided policies. In
2000, for example, some environmentalists suggested that
assuming a global target of 11.1 gigatonnes CO2 emissions is required to main-
tain global stability by 2050, and assuming the global population in 2050 is 9.8
billion, the per capita ‘environmental space’ for energy is 1.1 tonnes per year. UK
per capita production of CO2 is in the region of 9 tonnes, thus implying a reduc-
tion in UK emissions by about 85%’. (Chambers et al., 2000, p. 21)
Fourth, statistically environmental space is simply an average number, but
no standard deviations around the mean are given. Obviously if you divide
a finite resource by a large (and growing) number of people then year on
year the ‘fair share per capita’ becomes smaller. Hence, the policy to reduce
resource use in one country simply because the global population has
grown is naive and best ignored. Fifth, the idea of a fair share of envir-
onmental space is ethically naive as it confuses inequality with inequity (Le
Grand, 1991, p. 11) this vitally important issue will be discussed below. So
what should be done?
With regard to the serious problem of humanly induced global climate
change the scientific community would like a reduction of 60 per cent put
in place by 2050 (IPCC, 1992). The reductions are based on good science
and accurate measurements and not on the poor methodology underpin-
ning environmental space. The good science refers to identifying the correct
causal processes such as burning fossil fuels and land use change as part of
the feedback loops which contribute substantially to the complex processes
known as global climate change (Moffatt, 1991; Moffatt, 2004). The accu-
rate measurement of atmospheric CO2 concentrations establishes a time

Environmental space and ecological footprinting
series of data from the pre-industrial to today. Given the processes and the
measurements, then it is possible to calculate the reductions in CO2
required by each country (Moffatt, 2004; Owen and Hanley, 2004). In the
case of the UK, and using the figure presented by Friends of the Earth
(McLaren et al., 1998), a 60 per cent reduction would mean a reduction
from its current 9 tonnes to 3.6 tonnes per capita by 2050 and not the 1.1
tonnes per capita required from environmental space arguments. It could
be argued that larger reductions should be preferred on the basis of inter-
national equity. So how would such reductions come about? There are at
least two strategies: the first is to reduce emissions by some international
agreement and the second is not to produce them in the first place! The
Kyoto agreement, signed by 141 nations in 2005, has started the process of
CO2 reductions. It could be argued that the Kyoto reductions are a small
step in the right direction, but from a scientific perspective they are, in
themselves, insufficient to prevent a further global warming phenomenon.
The reductions proposed at the Kyoto meetings are a first step towards 60
per cent reductions. In the political arena the ideals of CO2 reductions have
not gone far enough. Obviously, given the small nature of the proposed
reductions at Kyoto, there is much more to do politically and diplomati-
cally to get all nations (especially the USA) to agree to the proposals.
The second approach to reduce greenhouse gas emissions is to introduce
new technologies and refine older ones. The new technologies would
include hydrogen power, and more efficient electrically driven engines. The
energy for the latter would come from the use of older technologies – the
so-called alternative technologies of waves, wind and solar power. The use
of these potentially renewable resources could maintain the quality of life
and improve conditions for many of the poorer nations without causing
further damage to the earth’s life support systems. These developments
would need to be encouraged both by changes to the macroeconomics of
the global economy and by ensuring, if necessary by international law, that
large companies do not prevent these developments.
Material flow analysis
We can all agree that reductions in resource use are a good thing so that
environmental waste is all but eliminated (Jacobs, 1991). It is, however,
difficult to ensure that if resource productivity is raised, this will be enough
to offset extra demands on the environment arising from economic and
population growth. Pearce has conducted a ‘thought experiment’ to show
that resource productivity would have to increase by 1.8 per cent per annum
to offset the potential rising environmental impact from economic and
demographic growth, 1975–1998. If past trends in resource productivity
occur, rather than Factor Four or greater resource efficiencies, then the

Handbook of sustainable development
world ‘will be worse off environmentally in 50 years’ time’ (Pearce, 2001,
p. 12). Recent research has also reported on a rebound effect in resource
efficiencies. The rebound effect occurs when an improvement in energy or
material use is offset by an increase in the number of units consumed (for
example video recorders, washing machines, new cars) (DEFRA, 2003).
Even if a nation apparently achieves some resource efficiencies one must,
however, be very wary of assuming such elimination has been accomplished
by increases in efficiency in resource use (that is dematerialization), and by
restructuring (that is decoupling) the economy.
It is acknowledged that business has caused major ecological damage and
that this cannot continue if we are to safeguard the environment for current
and future generations as well as for other life forms. In the UK, for example,
indices of TMR alongside GDP and population all increased between
1970–1999. While GDP grew by a large amount there was only a relatively
modest increase in material flows through the economy. Some might argue
that these results show an increase in resource productivity (dematerializa-
tion) and a de-coupling of the economy from resource requirements
(DEFRA, 2002). Similarly, in the OECD nations the energy intensity (the
amount of energy used to generate 1 unit of GDP) has fallen, but the total
energy consumption has increased by over 30 per cent in 20 years. This would
indicate that resource efficiencies are taking place. The argument over dema-
terialization of the economy of the OECD countries needs to be set in
context. Relative dematerialization has occurred in certain sectors of the
OECD economies. Yet to a significant degree this has been brought about ‘by
the net transfer of energy and resource intensive industries to the developing
world, in effect displacing rather than solving the environmental problems of
production’ (Robins and Trisoglio, 1995, p. 164).
Apart from the debatable scientific basis for Factor X reductions, mater-
ial flow accounting does permit detailed analysis of resource use and pollu-
tion generation at different geographical scales and over time. Obviously,
when using MFA there is a need to be clear about whether or not the reduc-
tions proposed are an important contribution to reducing our ecological
footprint or, alternatively, they are simply applied as fine tuning to reduc-
tions in the emissions of waste to the receiving environment. Currently,
MFA is used in both contexts. The MFA approach has its limitations and
its merits. The major limitation in MFA is that the materials flowing from
the ‘cradle to the grave’ are simply measured as a mass. The impact of one
tonne of arsenic on a receiving environment such as a river system would be
more lethal than one tonne of sewage. Yet the differential impacts of the
resource flows into the receiving environment are rarely stated – simply
adding up the total mass moved is insufficient from an ecological perspec-
tive. Clearly, the determination of MFA in an economy is a good first step

Environmental space and ecological footprinting
towards monitoring the links between the economy and ecology. It should,
however, be noted that a blanket reduction in MFA is ‘not guaranteed to be
ecologically effective, but is guaranteed to be highly economically ineffective
with respect to whatever reduction in environmental damage that might be
achieved’ (Neumayer, 2003, p. 181). Neumayer also doubts that MFA can
be used as a strong sustainability measure. Others, however, see MFA as one
way to demonstrate the ways in which increases in efficiency as well as reduc-
tions in material and energy flows (sufficiency) can be modelled (Barrett
et al., 2002). Furthermore, MFA does allow decision-makers to examine
scenarios to assist in choosing the best options to encourage reductions in
both resource consumption and waste generation and to consider the appro-
priate technical changes as a contribution to sustainable development.
Ecological footprinting
As the ecological footprint concept evolved it has been subjected to many
criticisms (Ecological Economics, 2000). First, the unit of measure ‘global
hectares’ has been criticised as too crude an indicator for detailed policy
proposals. Initially this was true, but recent developments have integrated
footprints with more conventional national accounts. Next, the idea of
trade flows being difficult to account in ecological footprinting has been
raised, but this has also been partially answered by integrating ecological
footprinting with material flows and input–output analyses (Moffatt et al.,
2005; Wiedmann and Barret; 2005; Wiedmann et al., 2005). Clearly, it is
physically impossible for every country to be a net importer of biocapacity
as this will lead to global overshoot of resource use (Wackernagel et al.,
2002). It should, however, be noted that the ecological footprint does not
address all environmental issues involved with pollution and species loss. It
should also be realized that the earth-share of ecological footprinting is
open to the same criticism of environmental space. To argue for a fair earth
share by simply dividing the amount of land expressed as global hectares
(gha) by the total global population is meaningless as it again confuses
equity with equality (Le Grand, 1991). If, however, emphasis is placed on
absolute limits and not per capita ratios then the footprint can still be a
useful indicator of environmental sustainability. As noted in the previous
section we cannot live beyond the 11.3 gha of this planet. We need there-
fore to encourage each nation to reduce conspicuous consumption, control
population growth and introduce ecologically friendly technology. It
should, however, be noted that the variations in the ecological footprint in
different countries may be due to socio-economic rather than ecological
processes (Kooten and Bulte, 2000).
There is, however, a partial solution to the problem of living ecologically
sustainably on the earth. Numerous ecological footprinting studies have

Handbook of sustainable development
shown that waste, food and energy make up a large portion of the footprint.
In the case of Scotland, for example, these three components make up 38
per cent, 29 per cent and 18 per cent respectively (Best Foot Forward, 2004).
Clearly, these are sensitive parameters in the ecological footprinting
methodology and each one could be reduced. In the case of energy, ignor-
ing the nuclear option as too high a risk and potentially very damaging,
then the use of renewable resources for energy production can be an
effective way of reducing the footprint. Assuming that we move to renew-
able energy sources we could, in theory, have reduced the size of the eco-
logical footprint of anthropogenic CO2 emissions by over 50 per cent for
the period 1961–99. This was technically and practically possible but would
have to overcome the vested interests of the powerful energy lobby. It would
also require some alterations to macroeconomic policy to encourage both
the development and market for renewable energy sources. Nevertheless,
this simple example illustrates that if energy were obtained from non-
carbon resources then the ecological footprint would drop automatically
and substantially (Ayres, 2000). Similarly, if waste could be substantially
reduced then this would also have a major impact on reducing the size of
the footprint. If policies were pursued to increase renewable energy supplies
and reduce waste substantially then this would substantially reduce the
ecological footprint well below the biocapacity limit.
If the global adoption of alternative energy has the potential of reducing
the ecological footprint to well below the biocapacity limits, does this then
mean we are living sustainably? Clearly the answer to this question is a qual-
ified ‘yes’. If policies are implemented to promote reductions to zero in
nuclear and fossil fuel energy over, say, ten years, and simultaneously increase
the input from renewable energy sources to meet demand, then we have the
necessary conditions for sustainable living. In order to attain the necessary
and sufficient conditions for sustainable development we need also to address
other environmental problems (such as different pollutants and biodiversity
loss) as well as economic and social justice issues. This inevitably raises ide-
ological and ethical problems concerning contemporary globalization with
its use and abuse of the earth’s environment and its inhabitants. Harvey sug-
gests that a globalized world is one of ‘class oppression, state domination,
unnecessary material deprivation, war and human denial and that we should
strive to create our environments in a state of liberty and mutual respect of
opposed interests’ (Harvey, 2001, p. 120). He suggests that we need to adopt
a holistic, dialectical approach to understanding the dynamics of the current
trajectory that we are locked into in order to break free from the present
situation. Alternatively, those in favour of natural capitalism argue that,
‘natural capitalism is not about fomenting social upheaval. On the contrary
that is the consequence that will surely arise if fundamental social and

Environmental space and ecological footprinting
environmental problems are not addressed’ (Hawken et al., 1999, p. 322).
Clearly, there are ideological differences underpinning these diametrically
opposed perspectives. The current generations have a difficult choice because
making the wrong decisions can result in universal misery and the collapse
of civilizations (Diamond, 2005). Closely associated with this vital issue is
the related question of a just distribution of resources. We have suggested
that the fair share approach (equal resources per capita) confuses equality
with equity. This is an important normative issue with major environmental,
economic and social policy implications. It should, however, be noted that,
‘Policies should be equitable and that distributional consequences of policies
should, so far as possible, be just or fair. These are considerations that policy
makers ignore at their peril’ (Le Grand, 1991, p. 175). Policy making should
be transparent, accountable, just and based on sound methodologies. It is the
integration of multi-disciplinary research into ecological, economic and
equity issues that poses the fundamental methodological challenge for
making development sustainable.
7. Conclusions
Environmental space, material flow analysis and ecological footprinting
are discussed in the literature on sustainability and sustainable develop-
ment. From its inception environmental space was based on some very
suspect scientific premises. We do not know the amount of non-renewable
resources remaining in the earth’s crust nor, with the exception of atmos-
pheric CO2, do we know the global assimilative capacity of the earth’s
receiving environments. In order to contribute to sustainable development
Friends of the Earth proposed the use of environmental space as a blunt
policy tool. This method is very suspect and has resulted in the environ-
mental space concept being ignored by both the scientific community and
most policy makers.
Material flow analysis is based on sound concepts of the conservation of
matter and the laws of thermodynamics. This sophisticated form of analy-
sis allows researchers and others to explore different ways in which mater-
ials and energy flow through a system. The systems under investigation can
be at different geographical scales including natural and man-made ecosys-
tems such as factories and businesses, cities, regional and national
economies, and globally. Ideally it would be interesting to tie these models
of material flow analysis with similar models of ecosystems but this has
rarely happened (Odum and Odum, 1976). The problem of using material
balances (MFA) and conventional neo-classical economics still remains.
Nevertheless, the development of material flow analysis including the use
of scenarios has the potential to contribute to action leading to sustainable
development at different geographical scales.

Handbook of sustainable development
The ecological footprint concept has developed a large and rapidly
growing amount of literature. As noted earlier this methodology has now
been standardized and this also has been applied at different spatial scales.
It has also been argued that by committing to alternative energy supplies it
is possible to reduce the ecological footprint substantially. Furthermore,
recent work on integrating ecological footprinting with material flow analy-
sis and input–output analysis shows that the allocation of resource use
sector by sector can be achieved. Despite this progress it should be noted
that the problem of linking masses with monetary measures still remains –
although some attempts to bridge this gap are being made but a firmer
theoretical basis for combining mass and monetary measures is required.
At present the integration of MFA with ecological footprinting via
input–output analysis is a step along the way to a more formal solution to
this ecological–economic problem. Current work permits the examination
of scenarios and allows policies to be targeted at different sectors of the
economy as a contribution to do more with less. It would also assist in sus-
tainable consumption.
If we are to live within the ecological possible, as proponents of strong
sustainable development urge, then it can be seen that many nations are
currently living well beyond what the world’s ecosystems can withstand. At
a global level, we are exhausting the earth’s renewable and non-renewable
resources on the untenable assumption that current economic and demo-
graphic growth and resource consumption processes, together with waste
generation, can continue indefinitely on a planet of finite size. The mea-
sures described in this chapter are beginning to address these problems.
This raises issues over the radical restructuring to our economic system so
that individuals and their organizations can begin to live as part of the
ecology of the planet rather than trying in a futile manner to live apart from
it. These changes are not impossible to achieve, and there are signs of hope
as outlined in the UK Government Sustainable Development Strategy (Cm
6467, 2005). The challenge is up to the political will and determination of
our elected leaders to encourage business and individuals to behave as
citizens rather than consumers (Dobson, 2003). In this sense, individuals,
as citizens of a global community, can contribute to the creation of an eco-
logically sound and socially just economy and it is through the collective
political processes that sustainable development will be achieved.
Anon (undated), ‘Taking stock managing our impact an ecological footprint of the South East
Region’,, accessed 25 February 2005.
Atkinson, G., R. Dubourg, K. Hamilton, M. Munasinghe, D. Pearce and C. Young (1997),
Measuring Sustainable Development: Macroeconomics and the Environment, Cheltenham,
UK and Northampton, MA, USA: Edward Elgar.

Environmental space and ecological footprinting
Ayres, R.U. (1978), Resources, Environment and Economics: Applications of the
Materials/Energy Balance Principle, New York: Wiley.
Ayres, R.U. (2000), ‘Commentary on the utility of the ecological footprint concept’,
Ecological Economics, 32: 347–9 (special issue on ecological footprinting).
Barrett, J. (1998), Sustainability Indicators and Ecological Footprints: The case of Guernsey
School of Built Environment, Liverpool: Liverpool John Moores University.
Barrett, J. and A. Scott (2001), ‘An ecological footprint of Liverpool: developing sustainable
scenarios’, Sweden: Stockholm Environment Institute.
Barrett, J. and C. Simmons (2003), ‘An ecological footprint of the UK: Providing a tool to
measure the sustainability of local authorities’, York: Stockholm Environment Institute,
University of York.
Barrett, J., H. Vallack, A. Jones and G. Haq (2002), ‘A material flow analysis and ecological
footprint of York: technical report’, Sweden: Stockholm Environment Institute.
Beckermann, W. (1998), ‘Sustainable Development: Is it a useful concept?’, Environmental
Values, 3: 191–209.
Best Foot Forward (2002a), ‘City limits: a resource and ecological footprint analysis of
Greater London’, Oxford: Best Foot Forward.
Best Foot Forward (2002b), ‘Ol-troed Cymru: the footprint of Wales’, Oxford: Best Foot
Best Foot Forward (2004), ‘Scotland’s footprint a resource flow and ecological analysis of
Scotland’, Oxford: Best Foot Forward.
Best Foot Forward and Imperial College London (2000c), ‘Island state an ecological footprint
analysis of the Isle of Wight’, Oxford: Best Foot Forward.
Buitenkamp, M., H. Verner and T. Wams (eds) (1991), ‘Action plan: sustainable Netherlands’,
Netherlands: Friends of the Earth.
Chambers, N. and K. Lewis (2001), Ecological Footprint Analysis: Towards a Sustainability
Indicator for Business, Certified Accountants Educational Trust, Research Report No. 65,
Chambers, N., C. Simmons and M. Wackernagel (2000), Sharing Nature’s Interest, London:
Christensen, V. (1995), ‘A model of the trophic interactions in the North Sea in 1981, the year
of the stomach’, Dana, 11(1): 1–28.
Clark, C.W. (1976), Mathematical Bioeconomics: the Optimal Management of Renewable
Resources, London: Wiley.
Cm 6467 (2005), ‘The UK Government sustainable development strategy’, London: HM
Daly, H.E. (1972), Steady State Economics: The Economics of Biophysical and Moral Growth,
San Francisco: W.H. Freeman.
Daly, H.E. (1990), ‘Towards some operational principles for sustainable development’,
Ecological Economics, 2(1): 1–6.
Daly, H.E. and J.B. Cobb (1989), ‘On Wilfred Beckerman’s critique of sustainable develop-
ment’, Environmental Values, 4(1): 49–55.
DEFRA (2002), ‘Changing patterns: UK Government framework for sustainable
consumption and production’, London: Department of Environment Food and Rural
DEFRA (2003), ‘Sustainable consumption and production indicators: Joint DEFRA/DTI
consultation paper on a set of “decoupling” indicators of sustainable development’,
London: Department of Environment Food and Rural Affairs.
Diamond, J. (2005), Collapse: how Societies Choose to Fail or Survive, London: Allen Lane.
Dobson, A. (2003), Environmental Citizenship, Oxford: Oxford University Press.
Ecological Economics (2000), ‘Commentary forum: the ecological footprint’, Ecological
Economics, 32(3): 341–94.
Eurostat (2001), ‘Economy-wide material flow accounts and derived indicators’,
Luxembourg: Eurostat.
Faucheux, S. and M. O’Connor (eds) (1998), Valuation for Sustainable Development,
Cheltenham, UK and Northampton, MA, USA: Edward Elgar.

Handbook of sustainable development
Faucheux, S., D. Pearce and J. Proops (1999), Models of Sustainable Development,
Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
Friends of the Earth (FoE) (1995a), Towards Sustainable Europe: The Study, Brussels: Friends
of the Earth.
Friends of the Earth (FoE) (1995b), Towards Sustainable Europe: A Summary, Brussels:
Friends of the Earth.
Friends of the Earth Scotland (1996), Towards a Sustainable Scotland, Scotland: Friends of
the Earth.
Friends of the Earth England, Wales and Northern Ireland (1996), ‘Draft report on sustain-
able UK’, London: Friends of the Earth.
Gorshkov, V.G. (1995), Physical and Biological Bases of Life Stability, Berlin: Verlag.
Haberl, H., M. Wackernagel and T. Wrbka (2004), ‘Land use and sustainability indicators’,
Land Use and Policy, 21(3): 194–320, (Guest Co-editor Ian Moffatt).
Harvey, D. (2001), Spaces of Capital Towards a Critical Geography, Edinburgh: Edinburgh
University Press.
Hawken, P., A.B. Lovin and L.H. Loven (1999), Natural Capital: the next Industrial
Revolution, London: Earthscan.
IPCC (Intergovernmental Panel on Climate Change) (1990), Climate Change – The IPCC
Scientific Assessment World Meteorological Organisation and the United Nations
Environmental Program, Cambridge: Cambridge University Press.
IPCC (Intergovernmental Panel on Climate Change) (1992), The Supplementary Report to the
IPCC Scientific Assessment, edited by J.T. Houghton, B.A. Callander and S.K. Varney,
Cambridge: Cambridge Univesity Press.
Jacobs, M. (1991), The Green Economy: Environment, Sustainable Development and the Politics
of the Future, London: Pluto.
Kirk, W. (1963), ‘Problems in geography’, Geography, 48: 357–71.
Kooten, Van, G.C. and E.H. Bulte (2000), ‘The ecological footprint: useful science or poli-
tics?’, Ecological Economics, 32: 385–9.
Kratena, K. (2004), ‘ “Ecological value added” in an integrated ecosystem–economy model –
an indicator for sustainability’, Ecological Economics, 48(2): 189–200.
Le Grand, J. (1991), Equity and Choice: An Essay in Economics and applied Philosophy,
London: Harper Collins.
Lenzen, M. and S.A. Murray (2001), ‘A modified ecological footprint method and its appli-
cation to Australia’, Ecological Economics, 37(2): 262–71.
Linstead, C. and P. Ekins (2001), ‘Mass balance UK: mapping UK resource and material
flows’, London: Forum for the Future.
Linstead, C., C. Gervais and P. Ekins (2004), ‘Mass balance: an essential tool for under-
standing resource flows’, London: The Royal Society for the Conservation of Nature.
Loh, J. (2002), ‘Living Planet Index’, Gland, Switzerland: WWF International.
McLaren, D., S. Bullock and N. Yousef (1998), Tomorrow’s World: Britain’s Share in a
Sustainable Future, London: Earthscan.
Mirowski, P. (1990), ‘Smooth operator: how Marshall’s demand and supply curves made
neo-classicism safe for public consumption but unfit for science’, in R.M. Tullbeg (ed.),
Alfred Marshall in Retrospect, Edward Elgar, Aldershot, UK and Brookfield, US:
pp. 61–90.
Moffatt, I. (1991), The Greenhouse Effect: Science and Policy, in the Northern Territory,
Darwin: Australia Australian National University, NARU.
Moffatt, I. (1996a), Sustainable Development Principles, Analysis and Policy, Carnforth and
New York: Parthenon Press.
Moffatt, I. (1996b), ‘An evaluation of environmental space as the basis for sustainable
Europe’, International Journal Of Sustainable Development and World Ecology, 3: 49–69.
Moffatt, I. (2004), ‘Global warming and its relationship to the economic dimensions of
policy’, in A.D. Owen and N. Hanley (eds), The Economics of Climate Change, London:
Routledge, pp. 6–34.
Moffatt, I., N. Hanley and M.D. Wilson (2001), Measuring and Modelling Sustainable
Development, Carnforth and New York: Parthenon Press.

Environmental space and ecological footprinting
Moffatt, I., T. Wiedmann and J. Barrett (2005), ‘The impact of Scotland’s economy on the envi-
ronment: a note on input–output and Ecological Footprint analysis’, Quarterly Economic
Commentary, Fraser of Allende Institute, University of Strathclyde, 30(3): 37–44.
Monfreda C., M. Wackernagel and D. Deumling (2004), ‘Establishing national natural capital
accounts based on detailed ecological footprint and biological capacity assessments’, Land
Use Policy, 21(3): 231–46.
NCBS (National Centre for Business and Sustainability) (undated), ‘Rocks to rubble: Building
a Sustainable Region’, National Centre for Business and Sustainability, Manchester
(, accessed 22 February, 2005.
Neumayer, E. (2003), Weak versus Strong Sustainability, Cheltenham, UK and Northampton,
MA, USA: Edward Elgar.
Odum, H.T. and E.C. Odum (1976), Energy Basis for Man and Nature, New York: McGraw-
Opschoor, J.B. and R. Weterings (1994), ‘Towards environmental performance indicators
based on the notion of Environmental space’ Rijswijk, Netherlands: Advisory Council for
Research on Nature and Environment (RMNO).
Owen, A.D. and N. Hanley (eds) (2004), The Economics of Climate Change, London:
Pearce, D. (2001), Measuring Resource Productivity, London: DTI and Green Alliance.
Pearce, D. and G. Atkinson (1993), ‘Capital theory and the measurement of sustainable deve-
lopment: an indicator of weak sustainability’, Ecological Economics, 8(2): 103–8.
Ravetz, J. (2000), City Region 2020: Integrated Planning for a Sustainable Environment,
London: Earthscan.
Rees, J. (1985), Natural Resources Allocation, Economics and Policy, London: Methuen.
Rees, P. (1992), ‘Ecological footprint and appropriate carrying capacity: what urban econo-
mics leaves out’, Environment and Urbanisation, 4: 121–30.
Robert, K.-H., J. Holmberg and E.U. von Weizsacker (2000), ‘Factor X for subtle policy-
making’, Green Management International, 31(Autumn): 25–37.
Robins, N. and A. Trisoglio (1995), ‘Restructuring industry for sustainable development’, in
J. Kirkby, P. O’Keefe and L. Timberlake (eds), The Earthscan reader in Sustainable
Development, London: Earthscan, pp. 161–73.
Schmidt-Bleek, F. (1992), ‘Materials flow and eco-restructuring’, Fresenius Environmental
Bulletin, 1: 529–34.
Schmidt-Bleek, F. (1993a), ‘MIPS – a universal ecological measure’, Fresenius Environmental
Bulletin, 2: 306–11.
Schmidt-Bleek, F. (1993b), ‘Towards universal ecology disturbance measures’, Journal of
Regulatory Toxicology and Pharmacology, 18: 456–62.
Senbel, M., T. McDaniels and H. Dowlatabadi (2003), ‘The ecological footprint: a non mone-
tary metric of human consumption applied to North America’, Global Environmental
Change, 13: 83–100.
Sheerin, C. (2002), ‘UK material flow accounting economic trends 583’, Office of National
Statistics, London, 140, accessed 25 February 2005.
Simmons, C., K. Lewis and J. Barrett (2000), ‘Two Feet – two approaches: a component based
model of ecological footprinting’, Ecological Economics, 32: 375–80.
Vuuren, D.P. van and E.M. Smeets (2000), ‘Ecological footprints of Benin, Bhutan, Costa
Rica and the Netherlands’, Ecological Economics, 34(1): 115–30.
Vuuren, D.P. van and L.F. Bouwman (2005), ‘Exploring past and future changes in the eco-
logical footprint of world regions’, Ecological Economics, 52(1): 43–62.
Wackernagel, M. and W. Rees (1994), ‘Ecological Footprints and Appropriated Carrying
Capacity’, in A.-M. Jansson, M. Hammer, C. Folke and R. Costanza (eds), Investing in
Natural Capital: the Ecological Economics Approach to Sustainability, Washington: Island
Press, pp. 362–90.
Wackernagel, M., L. Lewan and C.B. Hansson (1999), ‘Evaluating the use of natural capital
with the ecological footprint’, Ambio, 28(7): 604–12.
Wackernagel, M., C. Monfreda, D. Moran, S. Goldfinger, D. Deumling and M. Murray
(2004a), ‘National footprinting and biocapacity accounts, 2004: the underlying

Handbook of sustainable development
calculation method’, Global Footprint network, Oakland California, USA, pp. 1–32,, accessed 5 April 2005.
Wackernagel, M., C. Monfreda, K.-H. Erb, H. Haberl and N.B. Schulz (2004b), ‘Ecological
footprint time series of Austria, the Philippines, and South Korea for 1961–1999: compar-
ing the conventional approach to an “actual land” approach’, Land Use Policy, 21: 261–9.
Wackernagel, M., N.B. Schulz, D. Deumling, A.C. Linares, M. Jenkins, V. Kapor,
C. Monfreda, J. Loh, N. Myers, R. Noorgaard and J. Randers (2002), ‘Tracking the
ecological overshoot of the human economy’, Proceedings of the National Academy of
Science, 99(14): 9266–71.
Ward, R. and R. Dubos (1972), Only One Earth, Harmondsworth: Penguin.
WCED (1987), Our Common Future, Oxford: Oxford University Press.
Weizsacker, E.V., A.B. Lovins, and L.H. Lovins (1997), Factor Four: Doubling Wealth, Halving
Resource Use, London: Earthscan.
Wiedmann, T. and J. Barrett (2005), ‘The use of input–output analysis in REAP to allocate
ecological footprints and material flows to final consumption categories’, REAP Report
number 2, Stockholm Environment Institute, University of York.
Wiedmann, T., J. Minx, J. Barrett and M. Wackernagel (2005), ‘Allocating ecological footprints
to final consumption with input output categories’, Ecological Economics, 56(1): 28–48.
WWF (2004), ‘Living Planet Index’, World Wide Fund for Nature International (WWF),
Global Footprint Network, UNEP World Conservation Monitoring Centre, WWF, Gland,
WWF (2005a), ‘Living Planet Report 2004’, Gland, Switzerland: WWF.
WWF (2005b), ‘Reducing Wales’ ecological footprint’, Report Summary, Cardiff: WWF
WWF, Italia (2004), Ecological Footprint of the Tuscany Region, Rome: WWF.
Zimmerman, E.W. (1951), World Resources and Industries, California: University of
California Press.

Aucun commentaire: