Costs of Mitigation
Economic analyses tend to produce differing
estimates on the economic implications of policies aimed at greenhouse
gas (GHG) mitigation. The main reasons appear to relate to differences
in underlying modelling approaches, in specific assumptions being
adopted and in policy scenarios being simulated.
There are broadly three types of modelling approach
that have been used to consider the costs of emission reductions.
(i) Macroeconomic. These models are
generally very country specific. They may allow for supply and
demand to be out of balance (for markets not to clear). Hence,
they are probably best suited to the consideration of the dynamics
of transition towards lower carbon futures and for applications
in the short to medium term. Results, in terms of GDP response,
show considerable variation across modelsthey can be very
model-dependent, according to the particular assumptions employed.
(ii) General equilibrium. These models
assume that markets clear (ie that demand always equals supply).
They cannot address transitional costs, but are better suited
to long run estimates, on the basis that in the long-run resources
are re-deployed and the economy reverts towards long-run trends.
(iii) Bottom-up. These models will
tend to represent technology and energy efficiency from a detailed
set of choices. The model will choose the technologies to deploy
depending, in particular, on their costs and the costs of energy
inputs. Depending on the particular model it may be possible to
constrain the choices in some way. But in general, like general
equilibrium models, this type of approach is better suited to
consideration of long-run impacts than transitional costs. The
MARKAL model we have used is one version of a bottom-up model.
It is generally considered that models of types
(i) and (ii), may overestimate costs. They start from a position
that deployment of resources in the base case is optimal. Such
an approach is criticised for underestimating the potential for
low cost efficiency improvement and ignoring gains that may be
tapped by non-price policy change. Worst case results come from
models using macro-economic models, with lump sum recycling of
revenues, no emission trading and no non-carbon backstop technology.
Bottom up models of type (iii), on the other
hand, assume that there is a lot of low or nil cost technology
or energy efficiency potential. Estimates from such models can
be criticised for under-estimating costs on the basis that they
ignore various hidden costs, transaction costs or other constraints
that in practice limit the take-up of what are, otherwise, cost-effective
In terms of specific assumptions and policy
scenarios, there are several factors that help explain differences
in cost estimates, including:
Choice of baseline emission
on the elasticity of substitution between different fuels and
Scope for international permit
trading; which help drive down costs by allowing firms to reduce
emissions in the most efficient way; 
If/how revenues from carbon
taxes are used to reduce distortionary taxes elsewhere in the
economic system, for example, reducing social security payments
or if the revenues are used to provide consumers with lump-sum
The time horizon over which
mitigation targets may be achieved;
The possibility of win-win policies
and low cost outcomes through increased energy efficiency uptake;
The role of further ancillary
benefits (for example, improvements in air quality and subsequent
reduction in health related illnesses).
In spite of these drivers of variation, we have
a relatively good idea of at least the order of magnitude of economy-wide
costs of mitigation action. Both IPCC's assessment and analysis
by the UK Government for the Energy White Paper suggest that the
deep cuts needed to put us on a path to stabilisation compatible
with limiting eventual temperature rise to 2ºC at around
500 ppm need not be largeof the order 0.5 to 2 per cent
GDP over 50 yearsa delay of a few months in reaching a
particular level of GDP.
It should be clarified that although economy-wide
costs are unlikely to be prohibitive this does not mean that reducing
emissions is going to be easy: industry will face great engineering
challenges as a wide range of technologies (from renewables, to
energy conservation, carbon capture and storage, hydrogen production
and decentralised energy networks) need to be developed further.
Keeping the cost manageable depends on the steady
introduction of measures, starting from now, which is why the
UK has a long-term policy. If we tried to get say 60 per cent
cuts in CO2 over a few years rather than the period to 2050, costs
would become much larger. Also, the risk if no early action is
taken is one of "lock-in" into an energy system that
is highly reliant on fossil fuels, which would make the transition
to a low carbon economy at a later stage (when the most worrying
climate change scenarios may well have been confirmed) extremely
expensive or even impossible.
A further caveat is that manageable economy-wide
costs can disguise higher costs for some sectors of the economyenergy
intensive industries for example. Careful policy design ensures
that the costs of taking action are fairly distributed across
the economy, taking account of what particular sectors can achieve,
given the technologies available. UK policy seeks to achieve this,
eg by flexible mechanisms such as emissions trading.
Finally, an important issue that has only recently
begun to be addressed by the economic models of climate change
policy is the role of the technological change that may be induced
by the adoption of mitigation policies (as opposed to models assuming
cost-reduction purely following precedents). According to several
models policy-induced technological change can make a significant
difference to cost estimates. For example, a recent literature
review by the Pew Center
concluded that technological change induced by early action aimed
at reducing emissions may significantly reduce the ultimate costs
of mitigation policies. The Pew Center report goes on to recommend
that the level of abatement should then increase over time following
the cumulative nature of technological change. Defra is currently
sponsoring a Cambridge-led Innovation Modelling Comparison project
on this issue, involving a large group of modelling teams from
Europe and beyond. Preliminary results should be available in
There are numerous examples from business that
suggest that significant reductions in emissions can be achieved
at zero or negative cost. Since 1990, through its aggressive actions
to save energy IBM has avoided 8.45 million tonnes of CO2 emissions
and achieved operating cost savings of $791.4 million. IBM reduced
global GHG emissions associated with energy consumption by 65.8
per cent between 1990 and 2003 (35.4 per cent due to energy conservation.
In 1998, BP set itself the target of reducing its greenhouse gas
emissions by 10 per cent within 12 years. It achieved this goal
inside just three years (Absolute reduction 18 per cent). The
company integrated emissions targets into its senior managers'
performance contracts. It also introduced an innovative emissions
trading scheme to minimise cost. The programme cost the company
$20 million to implement, but saved it $650 million over the three
year period. Executives are confident that there is at least another
$650 million in value to be realised.
Many economists would agree that in theory decision-making
frameworks that look at climate policy choices as a problem of
sequential decision-making under uncertainty are preferable to
deterministic cost-benefit analyses. Typically these decision-making
frameworks allow for "learning", ie the possibility
of acquiring new information on the climate change problem, which
tends to reduce uncertainty over time. Some have argued that the
benefit of acquiring new, better information is an argument for
delaying irreversible investment to reduce greenhouse gas emissions.
But the risk of committing resources to irreversible investment
in low carbon technology should be compared and contrasted to
the symmetrical risk of a "lock-in effect" into an energy-economy
system that relies excessively on fossil fuels. Also, a strategy
of waiting to learn more on the risks of climate change is likely
to reduce the opportunity of "learning by doing" through
investment in low-carbon technology.
A good review of the relevant theoretical empirical
literature is provided by Alistair Ulph and Alan Ingham (Ulph
and Ingham, 2003). Most
of the studies reviewed by Ulph and Ingham suggested that the
prospect of getting better information at some point in the future
should lead to a small reduction in current mitigation efforts.
However, Ulph and Ingham stress that these results are very much
dependent on specific model assumptions. Furthermore, none of
the empirical studies led to the recommendation of "doing
nothing" as a short-term strategy.
21 The EU ETS is a tangible illustration of using
such flexible mechanisms in order to allow firms to meet their
emission reduction commitments in the most cost effective and
efficient way. Back
LH Goulder, "Induced technological change and climate policy".
Prepared for The Pew Center on Global Climate Change, October
2004. http://www.pewclimate.org/global-warming-in-depth/all_reports/itc/index.cfm Back
see http://www.tyndall.ac.uk/publications/working_papers/wp37.pdf Back