Baseline definitions and methodological lessons from Zimbabwe

R.S. Maya

Southern Centre for Energy and Environment

31 Frank Johnson Avenue

Eastlea, Harare


1 Introduction

The present range of activities to determine global and national response strategies for reducing emissions of greenhouse gases has taken two specific routes. One represents activities carried out under the IPCC programme working on assessing the state of the knowledge on the scientific, technical, economic and social aspects of climate change. These have been published in a various of IPCC reports and are the subject of second assessment currently under way. The other route involves the Actual generation of such knowledge. This includes the development of methods for carrying out scientific, technical, economic and social assessments of issues related to climate change. While it has been demonstrated that some the knowledge being generated is shrouded in controversy including the scientific knowledge itself, it can be argued that there is even greater controversy on social and economic issues related to proposed responses to climate change particularly the financing of these responses and the management of the global concerted effort to carry out such responses.

The UNEP Greenhouse Gas Abatement Costing Studies carried out under the management of the UNEP Collaborating Centre On Energy and Environment at Riso National Laboratories in Denmark has placed effort in generating methodological approaches to assessing the cost of abatement activities to reduce CO2 emissions.

These efforts have produced perhaps the most comprehensive set of methodological approaches to defining and assessing the cost of greenhouse gas abatement. Perhaps the most importance aspect of the UNEP study which involved teams of researchers from ten countries is the mix of countries in which the studies were conducted and hence the representation of views and concepts from researchers in these countries particularly those from developing countries namely, Zimbabwe, India, Venezuela, Brazil, Thailand and Senegal. Methodological lessons from Zimbabwe, therefore, would have benefited from the interactions with methodological experiences from the other participating countries.

Methodological lessons from the Zimbabwean study can be placed in two categories. One relates to the modelling of tools to analyze economic trends and the various factors studied in order to determine the unit cost of CO2 abatement. The other is the definition of factors influencing the levels of emissions reducible and those realised under specific economic trends.

The latter is represented in the concept of baseline definitions.

Operationally, baseline definitions relate to the writing of economic scenarios which form the basis of assessing future emissions of greenhouse gases. Baseline definitions define the expected growth path of economic activities including production technologies, population trends, energy intensity of economic production and the carbon intensity of the energy supplied to the economy. The defined baseline influences first the level of emissions expected without intervention and also influence expectation about the levels of emissions reduction that can be achieved from a base case scenario.

2 Implications of baseline definitions for response strategies and assessed abatement costs

Baselines have no standard definition besides that they simply indicate a base case scenario for economic activity and greenhouse gas emissions from which emissions reduction departures can be drawn. In some instances, these have been described as trend cases where economic development and emissions are expected to continue on the present path or trend projected purely as time dependant extensions of presently observed patterns. Trend cases have also been termed "do nothing" scenarios. A variation of these scenarios has also been applied in emissions abatement studies. The variation includes a combination of the trend case and some policy or practice interventions to improve energy efficiencies and to reduce emissions of greenhouse gases.

This category of cases represents what have been termed business as usual scenarios. Even in these cases, it has been difficult to determine exactly what is fair or practical to include in this modified trend case. In building a business as usual scenario it becomes necessary to answer such questions as "what really is business as usual?" For practical purposes this may have two possible definitions as follows:

Trend case modified to include all currently planned and possible policy interventions to improve energy efficiency or alter the economic structure in a way that would affects energy efficiency.

Trend case modified to include planned policy and project interventions and in addition to include all known and proven efficiency improvement interventions in the economy that would have been carried out for their own economic benefits even if these do not appear in planned activities.

The former definition was applied in Zimbabwe and the latter appears to have been adopted in some developing country studies. The former definition inherently allows for negative costs within the cost curve since it does not account for the use of proven and economically viable efficiency options. It would follow therefore, that any baseline definition should be premised on the latter definition and not on the former as was done in the Zimbabwe study. There are good reasons, however, why the baseline defined for Zimbabwe was based on the narrower definition which excluded known and proven option relying only on planned options. This reason is that there are some serious limitation which constrain the implementation of some options even if they are known and economically proven. These include capital constraint, transaction costs and some distortion in macro-economic policies.

A study by the UNEP Centre and Southern Centre is formally assessing the nature of these costs and impediments.

Even with respect to planned activities, the Zimbabwean study sought to assess the extent to which some of the planned activities were practical or achievable. Those options which were not practical or achievable were excluded from the baseline. The result of this was to further increase the incidence of negative cost options in the abatement cost curve.

Specific lessons from the UNEP study

There are two important lessons leant during the UNEP Greenhouse Gas Abatement Costing Studies with respect to these baseline definitions as they relate to developing nations. The first is that developing nations appear to include only trend case and planned activities in developing baseline definitions. This is in contrast with industrialised country approaches (in most cases) where a clear distinction exists in that baseline definitions here assume that all known and practical efficiency improvements will be carried out in the base case and not in the abatement case.

3 Baselines assumptions and their consistency with developing country issues

Developing countries have allowed the present economic difficulties to dominate projections of future economic performance within their baselines. They assume low GDP growths or they peg their growth rates to what has been previously achieved or is assumed possible from the experiences of the past. Most of these experiences have been quite depressive particularly in Africa where negative growth rates of GDP became quite dominant over the past two decades. This has remained the case in the face of the common-place "Five Year National Development Plans" which have superseded each other with hardly an item in the preceding plan being achieved.

While it may be argued that these poor trends may be adopted as a true picture of what is possible in these economies, reliance on these poor performance trends in baseline definitions may also called into question on the grounds that the past performance of these countries does not reflect their aspirations. This indicates that baselines thus derived give only a depressed picture of the development path these countries would like to see. The differential between adopted history based growth paths and national growth aspirations are the reason why some developing countries continuously project GDP growth rates which are they rarely achieve. It is in light of this that in building baseline scenarios, we must query the validity of adopting depressed historical achievements and question the validity of adopting aspired growth trends in the face of ample historical evidence that these may not be achieved.

Without extending the discussion, the above arguments indicate that baseline definitions in developing countries facing difficult economic conditions and perhaps in some developed countries as well are inherently wrong unless an assessment is made to determine the extent to which the country can in fact drive the economy toward the path of the defined baseline. Without this assessment optimistic baselines can be misleading. However, for generating baseline scenarios for abatement costing exercises, adoption of depressed historical performance as a pointer to the future should be called into serious question. For if it is considered that developing nations cannot steer their economies toward a specified baseline or desired optimum path, it should also be considered that they have limited capacity to follow any other forecasted development route including the abatement path. It would not be logical to expect developing countries to be able to steer their economies toward an abatement path if we argue that they would unable to direct the same towards their own desired baseline paths.

In this debate lies a crucial aspect of the approaches adopted for the management of and support for abatement activities in developing nations. To be viable, abatement costing activities in developing countries require much more than direct project funding. They need significant support for capacity building. Even this capacity building if isolated (directed toward climate change projects alone) would yield poor results as abatement programmes would be taking place in the context of and as part of regular national economic activities. So capacity building for abatement programmes should be accompanied by an overall and generalised upliftment of broader national economic activities.

Results from the Zimbabwe case study indicate quite clearly that there are real and genuine negative cost abatement options particularly those in the housekeeping category which have not been taken up even though ample information through industrial energy audits has been provided. In a bid to understand this phenomenon, the UNEP Collaborating Centre on Energy and Environment in collaboration with Southern Centre for Energy and Environment are carrying out a follow up study to determine the barriers to implementing these options. Initial indications point to broader macro economic distortions including the existence of monopolies, inappropriate pricing for energy, limited access to technology and more importantly the cost of starting up these viable efficiency improvement options.

Because of this situation where the actual investor experiences perceives inhibiting costs to implementing the so called negative cost options and operates within a discouraging macro economic environment, it would not be practical to exclude these options from the abatement cost curve or it would not be correct to assume that they would be carried out in the base case. What has not happened in this light, is that these options show a negative cost on the curve. This impression has to be corrected both to give the proper perspective to the policy makers, investors, and to the global effort to reduce emissions.

Methodologically, however, it is still not clear how these issues should be incorporated into the costing exercise.

4 Baseline assumptions and their implications for the cost curves

The result of these assumptions are that for developing countries those options which are currently feasible but are excluded in the baseline and appear in the abatement scenario are reflected on the abatement cost curve as negative costs or win-win conditions whereby the country could invest in efficiency improvements and earn a profit from doing so (see Figure 1).

1Figure 1. Typical abatement cost curve with negative cost options.

The general approach in studies carried out in some developed countries has been different in that negative cost options would not as it were appear on the cost curve as abatement options. (See Figure 2). This is due to the fact that where economic conditions are not distorted, energy users are expected to have adopted practices and technologies which eliminate any slack in energy efficiency.

2Figure 2. Typical abatement cost curve without negative cost options.

It can also be further argued that by including negative cost options in the cost curve, developing countries are actually underestimating the average cost of emissions abatement. If these options were not included on the set of abatement options, the total or average cost of abatement activities would be much higher. The importance of this argument is more apparent in situations where emissions reduction targets are in place and a significant portion of the target is reduced through negative costs. In the Zimbabwean abatement cost curve, Figure 3, 71% of the emissions can be reduced through negative costs options before introducing positive cost options.

3Figure 3. GHG abatement cost curve for Zimbabwe showing the extent of reduction possible through negative cost interventions.

Further, in the absence of negative costs the abatement cost curve would be forced to extend higher on the Y axis while gaining very little on the X axis - indicating a much greater marginal cost of reduction as we seek additional reduction options.

The occurrence of negative cost options in the Zimbabwe study could further be queried on two arguments. The first is that the cost assessment boundaries used were narrow as they are based on financial parameters only. These include capital, interest, operation and maintenance costs and fuel costs. Introducing other non-financial costs would significantly increase the cost of those options appearing as negative cost options on the present cost curve. This would have the effect of increasing the total and average cost of reducing emissions but would not affect marginal cost of those options presently appearing as positive costs options.

The second is that the approach also excluded transaction costs associated with introducing abatement technologies. These costs would include the cost of identifying and preparing these negative cost technologies, the opportunity cost to the economy of assigning capital to the introduction of these options and the cost of eliminating economic distortions, for example the cost of a structural adjustment programme which it is argued would have the effect of eliminating inefficiencies in the economy including inefficiencies in resource utilization and price distortions. In developed countries, unlike in most developing nations, state and private sector mechanisms are in place to encourage reductions in resource inefficiencies. Developing these mechanisms in developing nations such as Zimbabwe represents a major transaction cost which does not appear in current methodologies adopted for abatement costing.

These costs would also include the cost of capacity building to enable developing nations to address the issue of emissions reduction. In the Zimbabwean study, for example, an attempt to calculate the economic costs of introducing abatement technologies beyond the financial costs was hampered by lack of input output tables for the country. These tables would have enabled the study to assess the impact of proposed abatement technologies on the rest of the economy and particularly to assess how realistic the options are when viewed from the new linkages they create with present economic base and its capabilities.

There are, therefore, three types of costs which are excluded from the present formulae for abatement costing. These are:

economic costs of abatement interventions

transaction costs for abatement projects

externalities of abatement activities

5 Baseline definitions and choice of response options

Response options for emissions abatement depend on the composition of the baseline. This is because the baselines outlines that development path which should not be followed unless the best possible abatement path fully overlays the best the baseline.

For example, in the Zimbabwean case, the baseline envisaged coal was representing a significant share of the primary fuels for generating electricity and the abatement case (the path to be followed) envisages some reduction of this source of power in favour of hydropower. This situation has two implications: The first is that the present projections envisaging high coal utilizations represents the best path for optimum economic development under presently known conditions.

But because presently know conditions and the decision on a certain baseline are subject to "presently known" conditions bring the validity or infallibility of the baseline into question. If new information and new capacity is introduced, the selected official baseline may not represent the best possible path. In this case the presently selected baseline would not be a useful basis for drawing a "departure" or abatement path.

The validity of an abatement path, therefore, is a function of the validity of the baseline. While present methodologies have succeed in indicating a the normal path, they have not been able to justify this normal or baseline case as a guide on which to "peg" abatement activities.

Second, the advent of new information under a quest for GHG emissions reduction reduces the chances of applying this information to formulate a more critical development path for the economy unless this path also overlays the desired abatement path. Under conditions where emissions reduction will be done predominantly through improving energy and resource use efficiency, chances for this overlay are much greater. If decarbonization of the energy base becomes the dominant instrument, chances for this overlay are much more reduced. See the section below on the "Kaya Identity".

The point here is that introducing the abatement strategy as a replacement for the baseline path redirects the applications of improvements in planning conditions from a national focus to a global benefit focus. This however, is not always the case. The Zimbabwean study avoided this situation by pegging national aggregates to those defined in the baseline in a way that ensured that national welfare aggregates were not compromised by the definition of the abatement path. This precaution, however, did not help with such aggregates as welfare redistribution and the redistribution of structural pressures among sectors of the productive economy.

Presently, intervention options appear to rest within the energy sector since the quest for greater GDP growth rates cannot be compromised and gains through the reduction in population growth are only long term. In this situation we have what has been termed the Kaya Identity. This econometric argument which is identified below, would form some basis for assessing the rate at which emissions would expand.

This identity is defined as follows:

i.e. CO2 = ((C/E) x (E/GDP) x (GDP/P)P x P

where E = Energy and P = Population.

This identity simply says that where population growth is given and GDP is predetermined, then CO2 reductions are achievable only through variations in energy intensity of production (E/GDP) and variations in the carbon intensity of energy supplied (C/E).

This argument is most appropriate for Africa for two reasons. The first is that Africa has maintained in international fora that it does not have a population density problem but rather its problem is that of constrained economic development. The second is that Africa like most developing nations will seek to expand GDP to achieve and maintain acceptable social welfare targets.

GDP and population will therefore remain givens at least in the short to medium terms.

Key interpretations of the Kaya identity

For Africa to be seen to be holding its position under the UNFCCC, the identity defined above should be seen to be infallible. It is this condition that will allow Africa to push for cooperation through the transfer of energy efficiency technologies without compromising welfare. Should this mean therefore, that treating (E/GDP) and (C/E) as the only variables ensures a safe position for Africa and forms the basis for defining emissions and abatement baselines for the continent?

This may not be the case particularly with respect to the factor (C/E). Carbon intensity can be adjusted only through fuel substitution from high carbon fuels such as wood, coal and petroleum to such fuels as hydro and solar. It also happens that the low carbon fuels are also the renewable ones toward which energy development policies have sought to tend but have been limited by lack of technology.

Using this factor as the primary variable for influencing emissions levels entails a transition from high carbon fuels to the low carbon ones. This transition would creates distortion in the natural resource access paradigm optimal to the Africans. It may make sense therefore for the Africans to stick to the (E/GDP) factor.

Using this as the variable factor allows Africans press for the transfer of efficient technologies to vary E/GDP downward as opposed to forcing distortion in the primary energy base simply to achieve a lower C/E.

The impact of a baseline predicated on C/E as the primary variable would be more acute if we took into account that the energy intensity of production in Zimbabwe as would be the case in most expanding developing economies, energy intensity of production will increase (against overall GDP) as economic activity moves from low energy agriculture to manufacturing which would have a higher energy intensity than agriculture.

The arguments presented above serve to show that baseline definitions can be a complex matter which has yet to be resolved on various fronts including technology, economics and international politics.

Emphasizing C/E would rely heavily on hydropower as the alternative fuel for generating electricity. In this case, the baseline would entail some risks associated with unstable hydrological patterns due to falling rates of precipitation being experienced in most African regions. An example of these patterns is shown below in Figure 1 for the Kariba dam which is the main hydropower reservoir for Zimbabwe.

Figure 4.Contribution of lower catchment and rainfall on lake to the inflow of Lake Kariba.

The transition to hydro in an abatement strategy therefore entails certain risks which present methodological capacity has not been able to address.

There are many such instances in the abatement scenario which require specific and focused methodological analysis than the present UCC country study for Zimbabwe has been able to address. With this kind of picture, it would not appear realistic to deviate from a coal dependant baseline since the reliability of the carbon option is limited. This constraint makes it important seriously investigate technological options for reducing energy intensity along side whatever effort may be made to reduce carbon intensity. Energy intensity reduction which in most cases is possible through housekeeping and minor process improvements is a short term option which could mope up existing inefficiencies. Longer term reductions should be achieved primarily through technological improvements requiring much larger investment costs and economic restructuring.

Methodological treatment of this mix of abatement categories includes introducing these options in parallel. This approach, however, requires a clear definition of trends in possible improvements in energy efficiency through technological advancements and trends in sectoral linkages such that proposed sectoral and technological restructuring in the economy remains internally consistent. Under conditions of limited technical data on the economy, introducing the argument of internal constancy of defined baselines and abatement scenarios dampens the validity of most scenarios build for assessing abatement cost. This was more the case in Zimbabwe where macro economic input output tables have yet to be developed. Without such information to check consistency, it is possible that baseline definitions remain unsubstantiated and reduced in validity. Either this or trying to follow them will lead to some distortions in the economy.


Baseline definitions save a useful purpose in that they enable us to develop emission trends for greenhouse gases for the future and provide a threshold from upon which we can build alternative scenarios for stabilising or reducing emissions of these gases. It would not be possible as it were to simply introduce abatement activities without a reference point. In carrying out the Zimbabwe study, however, it became apparent that these definitions are much more complex than just a projection of future economic activities. To be meaningful, they have to be based on as practical as possible an understanding of such factors as structural changes in the economy, trends in energy efficiency, possible and realistic fuel switches to reduce carbon intensity of supplied fuels. Above all it is important to have some reasonable confidence on the extent to which both the defined baseline and the proposed abatement cases can be achieved through the state machinery and through investment preferences.

An important aspect in defining baselines is assessing the internal consistency of these baselines. A systematic assessment of this consistency requires the use of detailed input/output tables defining the linkages among the various sectors of the economy such that we can determine if defined baselines are practical. This information would also determine what impacts proposed trends will have on such key national aggregates as GDP, employment and the balance of payments.

These relationships remain unresolved.

Back to Contents