1996; U N .CSD , 1996; M oldan and Billharz, 1997. In the last few years, the ecological foot-
print R ees, 1992; Wackernagel and R ees, 1996 has frequently been mentioned as one of the
indicators that could be used in this context. Wackernagel and R ees 1996 defined the ecologi-
cal footprint EF as the total amount of ecologi- cally productive land required to support the
consumption of a given population in a sustain- able way. In the N etherlands, most of the atten-
tion
for the
EF came
originally from
environmental and
educational organisations.
Currently, scientists, policy-makers and politicians are increasingly becoming interested.
1
The central metaphor of the EF is probably the most impor-
tant reason for its popularity: i.e. expression of the impacts of human consumption in terms of a
visible footprint made on the natural carrying capacity; EF refers to the continuing dependency
of human societies on nature in terms of the more obvious dependency of traditional societies on
their available land.
Several organisations have already calculated EF s at different scale levels, ranging from individ-
ual and urban to the global scale see for example Wackernagel et al., 1997; Bicknell et al., 1998;
M ilieudienst Amsterdam, 1998; Wackernagel and R ichardson, 1998. M ost of the EF work is still
rather exploratory. As well as support, the EF has also received criticism from both scientists and
policy-makers in particular U N .CSD , 1996; van den Bergh and Verbruggen, 1999. At the mo-
ment, explicit work still has to be done on the applicability and usefulness of the EF before it
can be used in more regular state-of-the-environ- ment
or sustainable-development
reporting. M eaningful criteria to judge the applicability of
EF as a sustainable development indicator are, for instance: 1 policy relevance and utility for users,
2 analytical
soundness, 3
measurability Bakkes et al., 1994 and 4 communication to a
broader public. In this article, we will discuss an application of
the EF concept for the N etherlands and Benin, Bhutan and Costa R ica — three developing coun-
tries with whom the N etherlands has a close relationship through development
cooperation see also van Vuuren and de K ruijf, 1998. By
assessing the EF of these totally different but small countries, we will also indicate current
possibilities and limitations of the footprint as indicator for a specific country and for interna-
tional comparison. D ata availability allows the focus on the N etherlands to be more detailed than
on the other three countries. In comparison to many of the earlier footprint applications, the
calculations presented here are more detailed by combining
national and
international data.
M oreover, we examined the way the EF for the four countries changes through time — while the
EF has so far almost always been applied as a static indicator.
The article starts with a short overview of some of the appealing aspects and limitations of the EF
indicator used so far. On the basis of this, we decided to use a slightly adapted methodology —
as will be discussed in the next section. N ext, the results of the four country applications will be
discussed and compared with earlier results of Wackernagel et al. 1997. F inally, we return to
the advantages of the EF and its current limita- tions — as we believe that a balanced and com-
prehensive discussion on this topic is still missing. Can the EF be used for policy-making or to
stimulate debate or are its results always debat- able? The conclusions also discuss the conse-
quences of the adapted definition of the EF used.
2. The ecological footprint concept
2
.
1
. W hat is the ecological footprint
?
The EF is intended to provide an overview of the use of resources that can be attributed to final
consumption.
2
The discussions in Wackernagel
1
In 1998 questions on the EF were raised in the D utch parliament. The M inister of Environment referred to the eco-
logical footprint in several speeches and has asked a consulting agency to look deeper into the issue.
2
Obviously, it is related to earlier concepts in this area such as ‘carrying capacity’ and ‘environmental utility space’ Weter-
ings and Opschoor, 1994.
and R ees 1996 show the EF can be regarded as both a conceptual model and derived from this
concept a calculation method. The most compre- hensive of the EF applications so far include the
use of six different resources: crop land and pas- ture land for production of food and goods,
built-up land to support infrastructure, forest for the production of wood products, fish food
production and carbon assimilating capacity for carbon dioxide emissions from fossil fuels. In the
EF defined by Wackernagel and R ees 1996 these different types of resources are aggregated in
terms of the amount of land. F or the first four resources, this is relatively straightforward.
3
F ish consumption is translated into surface area by
estimating the area of productive sea required for producing the fish. Carbon dioxide emissions are
accounted for by assessing the area of carbon-sink forest required to sequester the carbon dioxide
emissions associated with burning fossil fuels.
4
This is based on the assumption that increasing the carbon dioxide concentration in the atmo-
sphere cannot be regarded as sustainable.
2
.
2
. T he EF
:
appealing aspects and weaknesses Earlier, we have made an overview of both the
appealing aspects of the EF and its weaknesses based on the discussions in literature van Vuuren
et al., 1999a,b. An attractive aspect of the EF is that it highlights several interrelated topics di-
rectly relevant for sustainable development. These include: a its focus on the consequences of in-
creasing consumption patterns, b its focus on several key resources for sustainable development,
c its focus on the issue of distribution of access to natural resources, d its focus on the conse-
quences of trade and the issue of geographical re-allocation of environmental pressures, and e
the powerful communication of results. At the same, time there are also some weak points of
current EF calculation methods. Some of them are discussed here in more detail to indicate how
we will deal with them in the context of this article.
2
.
2
.
1
. A ggregation In the aggregated EF , an indirect weighting
system is used to translate different pressures into an amount of land.
5
This requires several subjec- tive assumptions, including that for sustainable
development no increase of greenhouse gas con- centrations can be allowed. Van den Bergh and
Verbruggen 1999 argue that by using land as an aggregated indicator, the EF creates a false con-
creteness. H ere, we will avoid the issue of aggre- gation as far as possible by not focusing on the
aggregated EF — but on its components in their own typical units.
2
.
2
.
2
. T he lack of common methodologies and definitions
We found the results of earlier studies looking at the issue of land use for the N etherlands to
differ substantially, varying from about 10 million hectares to about 23 million hectares R ietveld,
1985; H arjono et al., 1996; van den H oek et al., 1996; Wackernagel et al., 1997. Analysis of these
studies shows that differences are mainly caused by: 1 definitions used and 2 assumptions with
respect to productivity, imports and exports. In this article, we use the definitions as mentioned in
Appendix A.
3
Although not all land is equally productive; this aspect will be given more attention later on in the paper.
4
Wackernagel and R ees 1996 also propose alternative measures: 1 calculating the amount of land required to
produce a sustainable substitute biofuels and 2 calculating the land area required to rebuild a substitutable form of
natural capital at the same rate of fossil fuel being depleted. According to Wackernagel and R ees, these alternatives give
comparable results. It should be noted that the proposed methodology is no more than an accounting method. The fact
that there are also other methods to reduce carbon dioxide emissions such as energy efficiency is not ruled out.
5
As will be shown further in this article, the EF weighting system results in energy use accounting for over 50 of the
total ecological footprint for industrialised countries. D espite the claim of subjectiveness, the present weighting system is in
line with the conclusions of many environmental assessments indicating the important role of energy consumption in envi-
ronmental problems of the industrialised world for instance, R eddy et al., 1997.
2
.
2
.
3
. Producti6ity In international comparison of land use, differ-
ences in productivity = land per unit of produc- tion play a major role. Productivity is influenced
by human management factors, including types of product, technology and knowledge, but also nat-
ural circumstances, such as soils and climate. Com- paring land use among countries and global regions
is complicated by the fact that in some countries natural circumstances are less favourable for high
agricultural productivity we will return to this in the discussion of our results. In current EF work,
the issue is altogether avoided by relating consump- tion to the global average yield.
6
F or national governments, however, land use based on local
yields might be much more relevant since these can be influenced, for instance, by increasing productiv-
ity which might result in unsustainable land use practices. M oreover, using local yields means that
the calculated area is equal to the real, touchable, area used for the consumption of a specific country.
Based on the considerations above, we concentrate here on land use based on local yields. In addition,
land use based on global average yields is shown as a reference in international comparison and to
show the differences in results.
2
.
2
.
4
. Ecological deficits and surpluses M any EF studies also define an ecological capac-
ity and subtract the EF from this capacity to determine so-called ecological deficits or surpluses
for instance, Wackernagel and R ees, 1996; Wack- ernagel et al., 1997. This deficitsurplus indicates
whether a country, in principle, is able to supply itself with domestic resources or whether it has to
rely on ‘net imports of land’ — thus indicating net self-sufficiency in terms of land.
7
On a global scale, the notion of self-sufficiency and sustainability
coincide. F or other scale levels, self-sufficiency and the deficitsurplus is a function of, among others,
population density.
8
The use of information on national ecological deficits is in our view limited, as
we doubt whether a large, scarcely populated, self-sufficient country with a large per capita EF
should be regarded as more sustainable than a small, densely populated country with a lower per
capita EF . F or sustainable development, compari- son of the per capita EF is more meaningful. The
consequence of unequal footprints in an increas- ingly populated world is that for every country
exceeding the global average capacity, there should be countries with a lower EF either on the basis
of less consumption or more efficient production.
Table 1 D ifferences between the methodology of Wackernagel et al.
1997 and the methodology in this paper Wackernagel et al. 1997
This paper F ocus on the underlying
F ocus on the aggregated EF . indicators for land use and
carbon dioxide U se of local yields for
U se of global average yields agricultural products
for agricultural products U se of equivalence factors
for different types of land use
a
23 product categories for 35 product categories for
land use land use
D oes not include the use of Includes the use of fish
fish resources
b
resources
a
Wackernagel et al. 1997 realize that not all land is equally productive and that the impacts of current land use
can be different with regard to future potential of this land. Therefore they multiply their categories of land use by so-
called equivalance factors for pasture land, forest land, crop- land and built-up areas: 0.54, 1.14, 2.82 and 2.82, respectively.
These factors are chosen so that on the global scale total land use still equals total available land. The equivalance factors do
not deal with large differences in productivity within the land use types. We have decided not to include equivalance factors,
since we intended to assess the real amount of land used by each country.
b
The reason not to include fish is that we did not like to mix up the sea and land — this, again, requires weighting
factors. This means that for countries where fish is an impor- tant source of the food consumption, the EF will be typically
lower.
6
This could be taken to imply that the total global highly productive and marginal lands are equally distributed to all
global citizens.
7
‘N et’ self-sufficiency here indicates that more land is used for consumption of the population of a country than is
available within its national borders.
8
Among the industrialised countries, for instance, large countries with a low population density have an ecological
surplus e.g. Canada — while small countries with a high population density but comparable consumption patterns
have a considerable ecological deficit e.g. the N etherlands.
F ig. 1. The components of the EF differences in focus between Wackernagel and R ees and this paper are indicated close pattern.
3. Methodology
