Unformatted text preview: A Scoping Study on the
Macroeconomic View of
Sustainability Final report for the European Commission, DG
Environment 29 July 2010 Hector Pollitt, Anthony Barker, Jennifer Barton (CE)
Elke Pirgmaier, Christine Polzin, Stephan Lutter, Friedrich Hinterberger, Andrea
Stocker (Sustainable Europe Research Institute – SERI)
Web +44 1223 533100
+44 1223 533101
A Scoping Study on the Macroeconomic View of Sustainability Contents
Executive Summary iii 1 Introduction 1 2 Literature Review 4 3 Identification of Existing Models 39 4 What Should Models Take into Account? 67 5 The Potential for Integrated Modelling 90 6 Conclusions and Recommendations 104 7 References 111 Appendices
Appendix A: List of Existing models 124 ii A Scoping Study on the Macroeconomic View of Sustainability Executive Summary
Introduction This scoping study, carried out by Cambridge Econometrics (CE) and the Sustainable
Europe Research Institute (SERI), examines the links between macroeconomic
perspectives and sustainable development. It considers how the links are represented
in economic theory and asks if the macroeconomic modelling we use today is up to
the task of evaluating policy from a sustainable development viewpoint. If not, then
models risk missing out on the insights that sustainable development provides: the
strong (two-way) linkages between the economy and the environment, the importance
of the long term, the necessity of an integrated approach and the danger of thresholds.
If these issues are missed by our models, then they risk giving us the wrong answers
and leading us in the wrong direction.
The study has: examined the theoretical underpinning of macroeconomic analysis
from different theoretical traditions; reviewed the wide range of models that operate
within one or more of the sustainable development spheres; considered the
appropriateness of the existing model toolkit to address policy questions; and
identified gaps in the scope of existing models and their methods. We conclude with
recommendations for improving the models on which policy makers can call. The role of Macroeconomic analysis looks at the performance of the overall economy. How the
sustainability in macroeconomy and macroeconomic factors and their general conditions are examined
macroeconomic varies in different schools of economic thought.
In the neoclassical model of the economy (which is the home of environmental
economics) the environment and its natural resources have never found a strong
footing. The ecosystem is treated as a subsystem of the economy whose main
functions are the limitless extraction of resources and the free disposal of
The environment mainly features in microeconomics, where it is assumed that
the internalisation of negative externalities through the price mechanism can
solve our ecological problems.
Mainstream macroeconomic theory is profoundly oriented towards the goal of
continuous and exponential economic growth. It is assumed that economic
growth can increase innovation and efficiency and lead to decoupling of
economic growth from negative environmental impacts.
The welfare of future generations is safe because there is full substitutability
of natural capital so the depletion of natural resources can be compensated via
investments in other forms of capital (a concept known as ‘weak
sustainability’). From a neoclassical economics perspective, there is no need
for a new macroeconomic framework for sustainability. An alternative macroeconomic framework is being developed by ecological
economists by extending the neoclassical framework to explicitly include the
environment and its services to the economy. In ecological economics: The economic system is not only embedded in the larger environmental
system but is also completely dependent on it as both a source of inputs and as
a sink for the matter or energy transformations required by economic activity. iii A Scoping Study on the Macroeconomic View of Sustainability The assumption that capital can substitute for resources is rejected on the basis
that certain functions that the environment performs cannot be duplicated by
humans (known as ‘strong sustainability’).
Environmental constraints imply limits on economic scale and thus limits to
Ecological economists are sceptical about the possibility to dramatically
change technologies, investment and consumption patterns in a way that
decouples economic growth from environmental impacts. In effect, ecological economists argue for a serious rethinking of standard economic
assumptions and theories (although a complete macroeconomic model in tune with
ecological economists' thinking does not yet exist). This debate matters, as depending
on whether a neoclassical or ecological economics perspective is taken, different
conclusions can be drawn on how the macroeconomy and macroeconomic factors
affect sustainable development and vice versa.
Many of the models reviewed in this study are derived from neoclassical economics
and general equilibrium theory so they typically focus on economic relationships.
Environmental factors, including resource consumption and greenhouse gas emissions,
are typically considered as external to the economic system. The possibility of not
being able to substitute between input factors, or of the depletion of stocks of
resources, is largely ignored. Where external factors, such as environmental emissions
or human health effects, are included in the modelling framework they are often
assigned monetary values.
The macroeconomic models that we consider in this study (see below) could be used
to test features of the two different schools of thought, for example estimating
substitution elasticities or modelling scenarios without economic growth. The existing situation
Policy domain of The study has focused on 60 of the most widely-used existing macroeconomic models
existing models (defined as quantitative computer-based tools), and mapped their scope against the ten
policy themes from the Sustainable Development Indicators (SDI) framework, and
their sub themes. Table 1 summarises the extent to which individual models cover
more than one SDI theme by counting the number of linkages in the models that were
reviewed. From this it is clear that: The vast majority of existing models focus on the link between economic
development and energy use.
Other linkages that are well-represented are economic development and
climate change; energy use and climate change; energy use and sustainable
transport; and economic development and global partnerships.
Perhaps the most notable omissions are links involving good governance or
public health, suggesting that the interactions between these policy areas and
other areas of sustainable development are not well covered. Figure 1 illustrates the dominant direction of the main links between SD policy areas
in existing models. There are some clear feedback loops between policy areas, for
example environmental Integrated Assessment Models (IAMs) cover the themes in the
iv A Scoping Study on the Macroeconomic View of Sustainability bottom-right part of the diagram (economy to energy to climate change to economy),
and integrated energy-transport models those in the mid-right section (economy to
sustainable transport to energy use). Nevertheless, it is clear that the dominant
direction of impact is from the economy, with few policy domains impacting back on
to the economy. This is consistent with the neoclassical economic theory, which
focuses on the economy as a system and treats other factors as external to this system.
For example, while the models may give outputs for consumption of finite fossil fuels,
they are less likely to include this as a driving factor of economic growth. TABLE 1: HOW EXISTING MODELS LINK ASPECTS OF SUSTAINABILITY
Sustainable C & P
Good governance 1 1
10 2 3 4 5 6a 6b 7 8a 8b 0 0 2 0 2 0 0 0 1 23 2 0 1 0 5 0 0 0 1 8 2 1 0 0 0 5 0 2 0 0 0 0 4 4 0 1 0 0 0 0 1 1 0 1 1 0 0 0 0 1 4 1 2 2 6 0 0 0 0 3 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 Sustainable Consumption
and Production Demographic Change
Natural Resources Sustainable
Public Health 9 2 Figure 1: Model Linkages
Global Partnership 8c Socio-Economic
Development Energy Use Social Inclusion
Good Governance v 0 A Scoping Study on the Macroeconomic View of Sustainability Overall, existing models appear only to provide limited coverage of SD policy areas
(though greater coverage could in principle be achieved through linking models) and
little role for developments in non-economic policy areas to influence economic
outcomes. Key gaps in The key gaps in the scope of existing models are:
existing analyses The good governance theme is not covered by any of the models, although
some separate quantitative analysis has been carried out, for example the value
of democracy in determining long-term growth rates.
The direction of linkages within a particular model is often only one-way,
especially in the case of models that include socio-economic development.
The socio-economic development and energy themes are well connected to
other themes. However, fewer direct linkages exist between the other aspects
of sustainable development.
Consumption of (non-energy) resources is not very well covered in existing
models, but this is an area of potential development (see Box 1).
Finally, there are a number of issues relating to the technical structure or
functional form of existing models which could be improved upon, particularly
the treatments of technology, uncertainty and non-linear relationships. Do the current In our view this leads to three major constraints in macroeconomic modelling analysis,
deficiencies which in certain circumstances are material for policy makers:
matter? The often one-way linkages from economy to environment can distort the
results of analysis and therefore also policy recommendations. For example,
environmental factors are usually only allowed to influence the economy
through price-based measures, such as taxation. Other environmental impacts,
such as loss of tourism due to degradation or loss of ecosystem services, are
Physical limits, such as stocks and maximum carrying capacities are not
usually covered in the analysis (see Box 1). This means that models miss out
on a wide range of factors, such as how the price of scarce material inputs
changes, or non-linearities in impact.
There is a loss of accuracy using conventional modelling approaches to
consider the effects of large-scale change. This makes a proper assessment of
‘extreme’ scenarios, for example where fossil fuel stocks dwindle, or of largescale change such as an 80% carbon reduction, difficult with existing tools. Key areas for development
Scope for linking
widening areas of
coverage It is not possible to have a single model that covers every aspect of sustainable
development. While it is desirable to have a number of models in the tool kit, it is
important to recognise boundaries and to ensure that the key relationships are
determined endogenously. In some cases this may mean linking individual models to
provide a more complete coverage, although it is often a resource-intensive exercise to
set up two-way linkages between different models.
Our analysis identifies the following improvements as important to adding value to
model-based analyses of sustainable development: vi A Scoping Study on the Macroeconomic View of Sustainability The treatment of natural resources could be improved, using existing model
structures (see Box 1).
Migration: This is clearly important in all areas of sustainable development. It
is usually treated as exogenous, but an endogenous treatment may be possible. We also identified areas where exogenous factors should be identified more explicitly: Government actions: It is difficult to model this endogenously, but qualitative
checks need to be carried out to make sure that model-based scenarios are
politically feasible. For example, if a scenario predicted very high rates of
unemployment, it is likely that the government would take action.
Models should acknowledge explicitly impacts on stocks of natural resources,
even if there is no feedback to behavioural patterns.
Implicit assumptions, such as financial stability or availability of energy
supplies, should be made much clearer so that they can be assessed in the
context of any set of model simulations.
Box 1: Incorporating resource use into the explanation of economic
development This study finds that it would be possible to integrate demand equations for the
physical consumption of materials (minerals and biomass) and water into existing
macroeconomic frameworks (including feedback to economic sectors such as
agriculture, mining and water supply). Eurostat provides relatively detailed data
sets on which such an analysis could be based. However, this is only one step in
setting up a system that is capable of carrying out a comprehensive analysis. A
more complete list of steps is: Identify and define the most important groups of resources. Measure the available stocks (eg fossil fuels) or maximum carrying capacities
(eg annual fresh water supply). Include the demands for these resources and, where possible, available stocks or
carrying capacities in macroeconomic models. Allow supplies to influence behaviour, for example (but not limited to) in price
formation in the model structures.
The last of these steps requires a much larger research input as the behavioural
responses to extreme outcomes are unpredictable. However, the other steps are all
possible with given model frameworks and supplementary analysis, and the
modelling approach required is close to that already applied for energy use.
A possible extension to this exercise would be to incorporate a treatment of
physical waste that is linked into the consumption of physical inputs. Recommendation 1: All models need to recognise that technology, and its development, is an important
The role of factor; an endogenous treatment of technology should therefore be seen as a
technology requirement. However, there is no agreed treatment at present and the focus of the
modelling is usually on the development rather than the dissemination of technology.
The two most common treatments of technology are based on engineering principles,
where individual technologies in a sector are explicitly defined (for example gas,
vii A Scoping Study on the Macroeconomic View of Sustainability renewables or CCS in the energy sector), and economic theory, where technology is
implicitly linked to the capital stock. These two approaches affect efficiency, unit
costs and product quality and each has its own advantages. However, when
considering future scenarios, both approaches are limited by the fact that it is not
possible to predict future trends in technological development. A good assessment of
current technological options is therefore seen as a priority and a methodology that
combines both approaches is possible. Recommendation 2:
thresholds, limits The standard modelling approach is based on linear (or log-linear) elasticities, for
example an elasticity of -0.5 would mean a 10% increase in price leads to a 5%
decrease in demand. Although there are cases where this assumption is relaxed, our
view is that this type of relationship is often forced on model parameters. This
potentially negates several important features and leads to the following issues and
potential developments: Linear approximations of curved relationships may be reasonably accurate for
small changes, but will become less accurate as the model moves further from
Threshold effects and physical limits should be taken into account (see
example in Box 1). However, problems arise in defining thresholds that have
not previously been reached (eg mineral resources becoming scarce), or where
thresholds vary over time (eg annual rainfall).
A proper treatment of asymmetrical relationships could be a relatively easy
improvement to make, with separate parameters for positive and negative
relationships. For example, if high commodity prices lead to investment in
new equipment, this equipment will still be used if prices fall again.
The assumption that model elasticities do not change over time should be
examined more closely. Recommendation 3: There are two separate issues related to uncertainty: the treatment of uncertainty
Uncertainty within the model; and uncertainty resulting from the model itself and its various
development Uncertainty clearly affects human behaviour, particularly in the case of
investment decisions, and should therefore be included in a model of
sustainability as an explanatory factor. An empirical exercise to identify the
main sources of uncertainty would be helpful in this respect.
Attempts to address the issue of uncertainty in model outputs are important for
putting results in context. They should be carried out outside the structure of
the model, for example by using different input assumptions. This is an area
where research is already under way. As well as better modelling of uncertainty, non-linear relationships and technology,
there are a number of key priorities for model developments to better assess the role of
sustainability in macroeconomic development. The report discusses longer-term
aspirations, but we consider the following steps to be practical developments given the
current state of the art.
1. Models should take biophysical data into account (for example the ecological
footprint or physical consumption of resources) as it provides a link to
problems that refer to the quality of the environment. This means moving
away from only including monetary values to also providing physical outputs. viii A Scoping Study on the Macroeconomic View of Sustainability 2. Models should be adapted to include issues of resource use. Modelling of the
supply and demand for material inputs (biomass and minerals) and water could
be integrated into existing frameworks relatively easily. This treatment could
be extended to include waste, though other issues such as biodiversity are more
difficult. The modelling would also need to be disaggregated as looking at
(bio)physical data at macro level may not be very informative: materials used
in the construction sector, for example, may be bulky but have little
environmental impact. The models would need to provide outputs capturing
the importance of the environmental impacts of the (non-renewable) resource
use, and perhaps also the option loss for future unknown use.
3. Building on this point, models should take into account the stocks of available
resources, at least given the available disaggregation. Even if no behavioural
responses are included, this would give an indication of whether model results
fit inside the limits imposed by bio-physical constraints.
4. The analysis identified gaps in existing data, specifically in data for incomes
and spending for different household groups, and a definition of the ecoindustries in standard classifications.
5. Population movement inevitably has an effect on sustainability, so the impact
of migration should be included in the models. A more immediate exercise
would be to use the models to quantify the effects of population changes.
6. For each set of model simulations, the model user should assess the accuracy
of assuming that exogenous factors remaining unchanged.
Overall, these developments would move macroeconomic models towards a more
systematic way of modelling the two-way linkages between the environment and
the economy, and would allow for ...