George Backus, Terry Barker, Department of Applied Economics, University of Cambridge, Sidgwick Avenue, Cambridge CB3 9DE, United Kingdom
This paper discusses some of the main macroeconomic linkages and feedbacks associated with policies for GHG abatement. The linkages in a global model are described when OECD carbon taxes or alternatively OECD/OPEC joint action raises world oil prices with a smaller OECD carbon tax. The results give a world perspective on the GHG abatement problem. The paper continues with a discussion of the linkages and feedbacks associated with three options: demand-side management (DSM), mainly to improve end-use energy efficiency, reductions in subsidies for fossil fuel production and use, and investment in renewable energy supplies, specifically in energy forestry and associated infrastructure. It concludes that DSM and investment in renewables are unlikely on their own to bring about reduction in GHG emissions, and that new supplies may even lead to increased emissions by driving down the price of energy. However, reductions in emissions may be compatible with increased efficiency and development, as well as with improvements in the quality of the local environment and in rural living, with the new supplies replacing fossil fuel supplies, provided that real fossil fuel prices are increased via removal of subsidies and if required carbon taxes.
This paper is concerned with national action to abate greenhouse gas (GHG) emissions using various policy instruments. It assumes that the problem of global warming is sufficiently serious to justify substantial global GHG abatement and that national governments, including those of developing countries, will pursue their own abatement strategies following international agreement. The paper is concerned with comparing the macroeconomic impacts and linkages of the technical and fiscal options available for abatement.
The main focus of the paper is on options in developing countries. It is recognised, however, that there is great diversity across countries in the structures of their economies and energy systems and in their economic performance. Nevertheless there are some common features of the policy options which may reduce the cost of abatement or even provide a net benefit. Improvements in economic welfare may well be compatible with reductions in GHG emissions.
The main options considered in the paper are as follows.
Demand-side management (DSM)
DSM policies are usually introduced by gas or electricity utilities, usually in the context of a regulatory regime, to encourage household, commercial or industrial energy-saving.
Reduction of energy subsidies
Many countries provide subsidies for coal and electricity use or force domestic prices below world levels. An option often favoured on the grounds of economic efficiency as well as GHG abatement is the phasing out of such subsidies and the use of the money saved by the government for other purposes. The introduction of a carbon or energy tax is an extension of this policy when the subsidies have been removed. A crucial issue for the effects on GHG abatement in the analysis of the reduced subsidy/carbon tax is the use of the revenues.
Investment in large-scale supply-side systems (energy forests; reduction in deforestation)
Funds for such investment may be provided by a world funding agency; otherwise, if they come from internal sources, some other development project will not be funded. The extra fuel when available will also change the demand/supply balance in the energy markets.
The macroeconomic effects of these policies will depend on whether there is external funding to cover costs. Several policies might well be justified without external funding; they will pay for themselves either directly in the form of government money saved which would otherwise have been spent on subsidies, or indirectly in the form or reductions in local pollution and associated costs. However, other policies might well be regarded as conditional upon external funds being made available from other countries on a bilateral aid basis or from a world agency.
The effects will also depend on whether energy markets are such that demand matches supply through changes in prices rather than through quantitative rationing in some form. Since the provision of electricity and gas is often by a single supplier, any liberalisation of the market implies that the government will have to regulate the industry to tax or restrict excessive monopoly profits.
There is a problem of monitoring the effects of national abatement whether or not external funding takes place, although such monitoring is likely to be a condition of funding. The problem is that in a developing economy baseline GHG emissions are likely to be growing and an abatement programme will only reduce that rate of growth. How is it possible to measure the success of the abatement? This is difficult in any event, but at least the countries signing up to the Rio convention have a clear target of stabilisation at 1990 levels of CO2 emissions by the year 2000.
The problem is not just one in monitoring programmes, it is much more general and has to be addressed in any analysis of policy intervention. In order to assess the effects of a policy, a base-line scenario or reference case must be postulated and, if quantitative results are required, the outcome must be measured or calculated. The base-line can be business-as-usual, or a projection of the economy without any abatement policies or the effects of abatement policies. (Note that even without a new policy being in place, economic agents may react to the possibility of future policies, by adjusting their investment spending: for example, power companies may favour new gas-fired generating stations over coal-fired ones because they expect GHG abatement policies to be introduced at some future date.) The base-line is important because it usually indicates that GHG emissions are expected to rise substantially over the next century and beyond, because world coal reserves are sufficient to meet greatly increased demands, although in doing so there is likely to be increasing costs of transportation and transmission. This implies that abatement policies designed to reduce emissions below 1990 levels have to be much stronger than might appear at first sight, because they have also to counter expected future increases in emissions.
When constructing models to explore GHG abatement, it is important at the outset to define clearly the objectives of the exercise. It is a mistake to try to construct a general all purpose model, but the objectives of a national model might be:
to explore the medium- to long-term consequences of economic growth and energy use on a range of indicators of environmental quality
to analyse the effects of a range of policy instruments on economic behaviour and these environmental indicators.
Such models should be general enough to include at least a rudimentary treatment of both population and technical change, if only to put the potential environmental impacts of these factors alongside those of economic growth. The model should treat the main economic variables, such as GDP and price inflation, as endogenous and should be capable of showing the effects of abatement policies on economic performance. Its time frame should be in terms of decades, since the accumulation of GHG into atmospheric concentrations. This is because of the time taken to reduce emissions at low costs and relates to the asset life of capital stock, particularly that in the energy supply industries, the time it takes for adjustment to new relative prices in the energy markets, found from estimated elasticities, and the time to develop new technologies.
The model also needs a bottom-up approach so that it can distinguish the main activities to give rise to substantial amounts of GHG emissions, the main industries those which produce different energy carriers, and the main technologies and potential technologies.
Finally, it should include the policy instruments by which GHG emissions may be reduced. Policy variables should include carbon taxes and charges, insulation programmes, energy conservation programmes, R & D. It is critically important to be able to say how the revenues of a carbon tax are to be spent.
This is a demanding list of requirements and it is not surprising that few if any actual models fulfils then. As a general approach to modelling abatement across a range of countries, a disequilibrium model is proposed because developing economies are far from equilibrium and the transient dynamics of GHG policies are a major interest. A general equilibrium modelling approach is not proposed because the equilibrium labour and technology assumptions simply do not hold for market economies in developing countries, let alone for economies in transition and centrally-planned economies.
GHG emissions are an unwanted byproduct of economic and social activity, mainly through the burning of fossil fuels and the clearing of forests. Anthropogenic GHG emissions include CO2, methane, CFCs, oxides of nitrogen and carbon monoxide. CO2 accounts for about 50% of the greenhouse effect, methane about 17%, CFCs about 12%. The methane and CFCs emissions problem is not considered in this paper; the solutions are likely to be more technical than economic. Anthropogenic emissions of CO2 are estimated to be some 28 billion tonnes (bt), of which 6 bt are from changes in land use (mainly deforestation in Africa, Latin America and East Asia) and 22 bt are from industrial processes including cement manufacture (World Resources Institute, 1992).
Considerable research has gone into the effects of a carbon tax on emissions of CO2 from industrial processes (see Boero et al., 1991, Grubb et al., 1993, and Barker et al., 1994, for reviews of costs) mostly at a global level and at a national level for industrialised countries.
One way of assessing the global macroeconomic linkages involved in abatement is to use a world econometric model which incorporates such linkages. Table 1 gives some results using OPEC's World Energy Model (OWEM) for a $10 per barrel oil-equivalent carbon tax, escalating each year 1992 to 2005 (Barker et al., 1992). The model is a macroeconomic model estimated on annual data 1973-1991. The table illustrates some general points about the global abatement problem.
Table 1.World CO2 emissions from industrial processes.
CO2 CO2 Base OECD carbon Joint (billio per tax action n capita $10/barrel tonnes) (tonnes) escalating World total 23.1 4 26.3 5.5 5.4 by fuel: oil 9.9 17.5 -3.1 -3.9 solid fuels 9.7 26.2 0.8 1.3 gas 3.5 51.6 43.2 43.2 by world area: North America 5.9 19 25.4 -25.1 -20.2 W. Europe 3.3 8 21.7 -20.1 -18.2 OECD Pacific 1.6 9 19.2 -12.4 17.9 OECD total 10.8 12 23.3 -21.7 -19.3 New market economies/CPES 8.2 6.5 6.5 6.5 (incl. China) 2.3 2 China 3.8 13 CIS 3.2 80.2 83.1 73.7 Other Developing countries 0.6 1 India
Sources: Barker, Miremadi, Walker and Brennand (1991), p. 5; World Resources Institute, 1992, pp. 346-348.
1) About 47% of 1988 CO2 emissions from industrial processes were in the OECD area, 35% in economies in transition and centrally planned economies and 14% in other developing countries.
2) There are huge variations in terms of emissions in annual tonnes per capita (tpc), ranging from 19 tpc for North America to under 1 tpc for many developing countries. If a 60% overall reduction in emissions was required (ie from about 4 tpc in 1988 to about 1 tpc given population growth), and if national action to mitigate climate change were related to emissions per capita, then some countries would have to reduce emissions by huge amounts, but others, mainly very low income countries, could still increase emissions several times over.
3) Emissions from other developing countries (although China in not included in this grouping) are expected to increase rapidly, although from a low base. The 83% increase from 3.2 bt in 1988 can only be an indicator of the order of magnitude of the increase, but it is striking.
4) Projections of emissions from economies in transition are very uncertain because increases arising from economic development are offset by decreases from removal of subsidies and liberalisation of the energy markets. Such liberalisation is happening anyway independently of any climate change policies. The projections assume that in the process of liberalising the Chinese economy over the next 10 years, the high subsidies on coal use will be removed. The baseline projections for this world area are only 6.5% 1988-2005 compared to 26% for the world total, a rather conservative estimate in view of the high growth rates expected for the Chinese economy.
5) Even if OECD emissions were to be cut by say 90% of 1988 levels by a combination of carbon taxes, energy-saving regulations and new technologies, world emissions would only be about 25% down on 1988 levels by 2005, since non-OECD emissions are on a sharply rising trend.
The table also illustrates some important linkages in the global macroeconomic response to an OECD carbon-tax abatement policy.
1) The table shows the effects of two policies: a $10 per barrel per year escalating carbon tax for OECD countries (ie rising from $10 in 1992 to $20 in 1993, $30 1994, etc), calculated to be sufficient to reduce OECD emissions to 20% below 1988 levels (the Toronto target); and joint OECD/OPEC action to combine a much smaller carbon tax with rises in the world crude oil price, but achieving the same global target. In both scenarios, carbon tax revenues are recycled within the OECD area by means of reductions in indirect taxes to keep the rate of inflation at baseline, so macroeconomic effects within the area are negligible. This is an important point because some of the earlier studies (see Boero et al., 1991) did not say whether or how the revenue from a carbon tax was spent by the government; the recycling option chosen makes a substantial difference to the GDP effects.
2) With an OECD tax, OPEC real revenues fall by 40%, which calls into question the assumption of a stable $21 per barrel for world oil prices in the scenario. If world oil prices were to fall, then non-OECD oil demand would rise, offsetting the OECD abatement; this is the carbon leakage problem. In the joint action scenario, world oil prices rise and CO2 emissions from developing countries are reduced compared to the baseline.
3) The results also show that world GDP is higher in the OECD tax scenario than in the base with the effect that GDP growth is higher for the developing country group. If world oil prices are reduced by the fall in world demand, then the terms-of-trade effect for oil-importing developing will further increase their GDP.
While the time-series, econometric studies find a correlation between economic and energy growth, consequent statements that argue for energy supply as an investment priority have tenuous validity. Energy is but one, albeit easily measurable, constraint to a rapidly growing economy. Historical evidence would indicate that the completion of a large energy-supply project, such as a hydro-electric dam, appears to either create secondary problems, or its usefulness becomes limited by secondary problems. These secondary problems are merely constraints that were waiting in-line to act as limits once another constraint was removed. Thus, the apparent link between energy and the economic growth must be reconsidered relative to development-policy implications.
Proper price signals would help investors and the government recognize the actual priority of potential constraint-relief measures. For example, transportation and telecommunications may prove to be much more critical than energy supply. Advocating transportation infrastructure in opposition to energy is not the incongruity it might appear to be, given the energy needs of transportation. Instead it highlights the crux of the issue: Where is the economy going? In the long term, the economy will demand a standard of living comparable to those of mature economies. It will require an infrastructure to support that economy. But for now, the day-to-day crises of transitional growth dominate daily life.
With rapid growth comes the opportunity to put in place an approach to infrastructure that best serves future needs. Actions based on relevant price signals would go a long way in the right direction. Nonetheless, the momentary path of least resistance may not reflect the path most suitable to a mature economy. To sell its goods and services, a country needs ports and telecommunications. It needs a means to get goods to the port. It does not need easier means for population to flood into the cities and add more strain to its infrastructure.
Despite the seeming hardship involved, governments confer a lesser hardship by letting the "cost" of doing business in city centres rise so that economic activity naturally migrates to other areas where constraints, costs and benefits are more balanced. Thus, during this stage of economic development, the type and function of transportation, for example, is more important than the energy efficiency of the service. An inefficient train, oil pipeline, or cargo vessel, will have much less energy consequence than promoting high efficiency automobiles whose engines idle in yet more oppressive centre city traffic.
Analyzing the implementation of DSM programmes in developing countries provides two important insights. First, it shows that developing economies are different from mature economies and experience in mature economies can be easily misapplied relative to GHG abatement. Second, a critical DSM analysis forces a change in perspective by showing that energy is just one of the many contributions and constraints to growth.
While the efficacy of implementing major DSM programmes for CO2 stabilization in mature economies remains suspect, the use of DSM programmes as the focus of CO2 stabilization in developing countries is even more questionable. Although DSM programmes do appear to increase the energy efficiency for the intended end-uses specified in the programmes, actual savings are often far different than the pre-programme expectations (Ford and Naill, 1985). Programmes can make the cost of using an end-use less expensive; this then provides an economic incentive to promote additional use of the supported end-use technology and so the expected savings will be reduced. DSM programmes may also change the relationship among competing fuels in the market place and thereby cause fuel switching into the DSM programme technologies with subsequent energy demand growth. Electric programmes that promote alternative fuels may cause participants to utilize fuels in a more polluting and inefficient manner than they would have under the original controlled energy production of electrical utilities (Smith, 1993).
Energy efficiency for its own sake is misguided. Energy is but one pressure on the economic dynamics. In many instances, what is not done is more important than what is done, if sustainable energy usage is to be achieved. Many investments for energy supply and demand may be better used to promote accounting procedures and mechanisms that let the market determine what response is most appropriate.
In mature economies, more limited economic growth prospects and marginal changes in economic demography allow well-designed DSM programmes to create persistent reductions in energy usage (Weisbrod, 1993). In developing economies, persistent, or even temporary, impacts may be a false hope. The dominant dynamics of DSM in developing economies bear almost no resemblance to the dynamics in mature economies. Developing economies currently fit into two categories: 1) those which are commodity based, or constrained, with a history of limited or negative growth, and 2) those which are transformation based, as in manufacturing, and with a history of high growth that slows to mature-economy growth when income per capita levels meet or exceed those of the mature economies.
For slow-growth economies, CO2 stabilization efforts, as seen from a global perspective, are of minimal importance because these economies will contribute an ever-decreasing fraction of the CO2 atmospheric burden. Further, their low standard of living indicates that they should be allowed to move toward energy-based conveniences while at the same time infers that they will unlikely be able to afford much more than the minimal complement of conveniences used on a limited basis and with limited CO2 implications. Low-carbon energy strategies must surely be justified from the point of view of development, not of reductions in potential CO2 emissions.
Rapid-growth economies, on the other hand, constitute a significant problem for CO2 stabilization strategies. Large-scale energy supply development is often a focus of the local government and foreign aid. Brown-outs and limited access to energy present major political and economic issues within the countries. But similar problems of rationing apply to water supplies, transportation, communication, financial, urban land, and a myriad of other factors in development in constrained supply. Rapidly-growing economies face one infrastructure and resource constraint after the other, with the implication that providing DSM, at best, only frees energy resources for immediate use elsewhere as the demand for energy services and for almost any other service, far out-strips supply. From a overly simplistic perspective, DSM might make even faster growth possible, and therefore an even greater need for new energy supply more immediate. In practice, however, if energy had been a constraint, then the DSM programme may mean that other physical constraints will then become limiting, and the well-intentioned DSM programmes may merely appear to be ineffectual.
This perspective of DSM in developing countries has several secondary implications that adds support for market-based approaches to energy and development policy. It reformulates the assumed linkage between energy and the economy, and it prescribes that the decisions on relaxing constraints to growth must be based on valid price information, rather than government intervention to prevent market clearing and foreign-aid investment schemes.
2) Reduction of energy subsidies
World Bank studies of abatement policies for developing countries (Anderson and Cavendish, 1992; Shah and Larsen, 1992a and b; Larsen and Shah, 1992) conclude that the removal of subsidies given for energy use and the introduction of a small carbon tax may be in the national interest of many developing countries. (The same is possibly true in most other countries - see Majocchi, 1994.)
Existing tax/subsidy regimes in the energy sector are likely to be a result of a complex set of social, political and economic forces: historical accident, pressures from vested interests and lobby groups, IMF requirement for loans, political promises, or pressures to reduce public borrowing. The need to reduce CO2 emissions has generally not been considered until fairly recently. The tax/subsidy system may also contain anomalies because of political pressures: for example the UK is still providing substantial subsidies to the coal industry in the UK at the same time as it is introducing policies to reduce CO2 emissions, eg valueadded tax on domestic fuel expenditure.
It appears that electricity has been and is regarded as a leading sector in many developing economies. In consequence, supplies have been subsidised and rationing of supply in several forms (black-outs, non-market allocation of supply) has been tolerated. The outcome of such subsidisation is extensive waste and inefficiency, with development biased towards electricity and energy-intensive industries. There are other candidates for the role of leading sector - eg telecoms linked with provision of information and training.
The reduction of energy subsidies may have serious macroeconomic and equity effects, although these can be ameliorated or removed through the use of the revenues. If the inflationary effect of reduced subsidies is offset by reductions in indirect taxes such that the general rate of inflation is unaffected, then the general adverse macroeconomic effects will be greatly reduced.
3) Investment in renewable energy supply: eg energy forestry
There has been considerable discussion in the literature on the potential for the development of biomass and a renewable energy source capable of replacing fossil fuels in supply. In this paper, the focus is on the planting of trees and energy crops, designed to replace fossil fuels and traditional unsustainable use of timber resources (Hall, et al., 1991; Read, 1994a and b; Edmunds et al., 1994). Note that simply cropping existing forests instead of burning coal has virtually no net effect of GHG emissions: new forests have to be grown, harvested and burned instead of coal to reduce emissions.
A large-scale investment in a renewable supply source, such as an energy crop, would have different macroeconomic effects depending on the pre-existing sources of supply. Assume that no changes are made to the existing and likely prices of energy carriers, except those arising from the new fuel source, ie existing subsidies remain in place and no new carbon taxes are introduced. Also assume that fossil and other energy supplies are in inelastic supply ie changes in prices received have little effect on supply because of sunk costs; this is essentially a short-term and medium-term assumption; in the long term, which may be over 20 years for fossil fuel supplies (coal mines) and longer for hydroelectricity, supply will contract since prices do not justify new investment.
During the start-up phase (the planting of the forests, the building of the transport infrastructure and the new power stations) the new investment will add demand to the system, using spare capacity is available and otherwise pre-empting other demand. Assume the investment is funded externally. It will then create new employment if the forests are in areas with surplus labour, and add to the areas income and general spending.
When new supplies of the energy carrier become available, they may be a close substitute for traditional fuel, which will then be replaced, depending on the relative costs of supply. However, this is likely to be a declining market since traditional fuels appear to be highly income inelastic in demand - ie as people get better off they tend to move away from traditional fuels such as wood for open fires, charcoal for cooking. The effect of the new supplies will then be to drive down the real price of energy, with the effect on the relative prices of the different energy carriers depending on substitution possibilities, and so encourage higher demand. For the policy to ensure that net GHG emissions are reduced, subsidies on fossil fuels will have to be reduced and/or a carbon tax introduced to maintain the overall price of energy and help to switch demand to the new supplies of renewable energy. Indeed there may be a real risk that the new supplies will not be able to compete with the existing fossil fuel supplies in the short term since fossil fuels typically have low marginal costs in supply. Indeed, faced with lower prices, the fossil fuel suppliers may well find themselves able to reduce costs; for example, the high-cost North Sea oil fields have been able to reduce costs substantially in the face of sharp cuts in the world oil price after their development. The removal of any subsidy and/or the introduction of a carbon tax may be essential to the long-term viability of the investment in the renewable supply.
The switch of resources from large-scale fossil fuel and hydro electricity projects to investment in biomass and traditional supplies of fuels could not only reduce net GHG emissions, it could lead to faster development. This is not such an unreasonable proposition as it appears at first sight. In fact substantial resources which are probably required for development have been allocated to large-scale energy supply systems based on fossil fuels and hydro resources. These are typically very capital intensive, very import-intensive and very skill-intensive. Therefore if the resources which would otherwise have gone into such systems were reallocated to other systems more instrumental in the growth process, characterised as more labour-intensive, less import-intensive and less skill-intensive, then the economy could grow and reduce GHG emissions.
Anderson and Cavendish (1992, p. 73) quote many examples of policies in developing and industrial countries which were biased against economic efficiency, economic growth and environmental quality, including GHG emissions. These include the absence of satisfactory tenurial arrangements in agriculture, defective (often corrupt) licensing policies for forestry, unnecessary subsidies for irrigation, for practically all forms of energy production and use, for land clearance and for water supplies. The reform of these policies will primarily improve efficiency, with small reductions in GHG emissions; however these are likely to become losses if energy demand is stimulated. Demand-side management has similar effects to reforms which improve the efficiency of energy use, again with probably small reductions in emissions which may easily be lost if economic growth and thus energy demand rises.
Economic development and reductions in GHG emissions may be compatible if reforms are combined with new policies to substitute away from the use of fossil fuels, via initially small carbon taxes and positive government encouragement of investment in broad-based renewable energy supplies such as energy forestry.
Anderson, Dennis and William Cavendish (1992). Efficiency and Substitution in Pollution Abatement, World Bank Discussion Paper 186, The World Bank, Washington, D. C.
Barker, Terry, Paul Ekins and Nick Johnstone (eds.) (1994). Global Warming and Energy Demand, Routledge, London, forthcoming.
Boero, G, R. Clarke and L.A. Winters (1991). The Macroeconomic Consequences of Controlling Greenhouse Gases: A Survey, UK Department of the Environment, Environmental Economics Research Series, HMSO, London.
Edmunds, Jae, M. Wise and C. MacCracken (1994). "Advanced energy technologies and climate change: an analysis using the Global Climate Change Assessment Model (GCAM)", paper presented to the IPCC Working Group III meeting, Milan, April 1994.
Ford, Andrew and Roger Naill, (1985). Conservation Policy in the Pacific Northwest, Bonneville Power Administration, Portland, Oregon.
Grubb, Michael, Jae Edmunds, Patrick ten Brink and Michael Morrison (1993). "The costs of limiting fossil-fuel CO2 emissions: a survey and analysis", Annual Review of Energy.
Hall, D.O., H.E. Myint and R.H. Williams (1991). "Carbon sequestration versus fossil fuel substitution: alternative roles for biomass in coping with greenhouse warming".
Larsen, Bjorn and Anwar Shah (1992). "Combatting the "Greenhouse Effect", Finance & Development, pp. 20-23, December.
Majocchi, A. (1994). "The employment effects of ecotaxes: a review of empirical models and results", OECD Workshop on the Implementation of Environmental Taxes, 1415 February 1994, Paris.
Read, Peter (1994a). Responding to Global Warming: the Technology, Economics and Politics of Sustainable Energy, Zed Books, London.
Read, Peter (1994b). "Carbon sequestration in forests: bio-fuel as the core technology in an effective response strategy", Massey University, Palmerston North, New Zealand.
Shah, Anwar and Bjorn Larsen (1992a). Carbon Taxes, the Greenhouse Effect, and Developing Countries, Policy Research Working Paper (WPS 957), World Development Office, The World Bank, Washington D. C., USA.
Shah, Anwar and Bjorn Larsen (1992b). World Fossil Fuel Subsidies and Global Carbon Emissions, Policy Research Working Paper (WPS 1002), World Development Office, The World Bank, Washington D. C., USA.
Smith, Bruce A. (1993). "Do energy-efficient appliance lead to a pot of environmental gold?", Proceedings: 6th National Demand-Side Management, Electric Power Research Institute, Report EPRI TR-102021.
Weisbrod, Glen (1993). "Persistence decay in a commercial retrofit program, results over a five year period", Proceedings: 6th National Demand-Side Management, Electric Power Research Institute, Report EPRI TR-102021.
Back to Contents