Sunday, March 30, 2008

BEYOND STERN. FORECASTING, COST EFFECTIVENESS, CLIMATE CHANGE

John Rhys
April 2007

The following notes were prepared as the basis for written evidence to be submitted to the Environmental Audit Committee in April 2007. The comments are addressed to the themes of the Committee’s inquiry.

KEY POINTS

Targets should retain focus on the key objective - the control of the level of global cumulative emissions; annual emissions in particular years such as 2020 or 2050 are useful landmarks but are ultimately only secondary or intermediate targets.

Official forecasts sometimes do not appear to have the benefit of detailed end use analysis. This is essential to monitoring many aspects of policy, particularly those that depend on influencing consumer behaviour; a comprehensive research programme to remedy this could be established for a comparatively modest cost.

Statistics highlight the fundamental importance of the electricity sector, not only as the largest current source of CO2 emissions, but because of its potential future “carbon-free” contribution in buildings and transport; a successful long term strategy demands an essentially carbon-free power sector.

Cost effectiveness; a significant economic issue in the Stern analysis suggests Stern might have attached an even higher social cost to current CO2 emissions.

There are some limitations to the use of cost effectiveness analysis, including the danger of inconsistency between public policy and private sector decisions; a corollary is that the value of carbon should feed through to the supply chain and into consumer prices.

Short term incremental measures provide important but limited reductions in emissions. Ambitious longer term targets imply systemic change both in supply and in demand; this attaches key importance to more urgency in long term plans.

For the longer term, annual forecasts are likely to be significantly less important than the monitoring of actions against a credible pathway for each sector, with long lead times involved. Monitoring arrangements need to reflect this.

Competitiveness issues have been exaggerated, given the relatively small emissions from industrial fossil fuel use and the small or indirect impacts on competitiveness from policy initiatives in the key sectors.

FORECASTING AND TARGETS.

A first requirement of CO2 targets is that they should align with the objective. This is to minimise cumulative emissions, not to achieve a particular level by a given date. Targets such as 60% or 80% reductions in annual emissions by 2050 may be useful indicators of what is required but they should not obscure the primary objective, reinforced by Stern, of keeping cumulative emissions within “safe” limits.

This distinction is important for two practical reasons. First, the shape of the path from the baseline to a given 2050 annual emission level has a very large impact on cumulative emissions over 43 years. To illustrate the point, a 60% reduction over 45 years requires a 2% pa reduction. However a 3.5% pa reduction for 20 years followed by a 1% pa reduction for 25 years yields the same annual emissions after 45 years, but a cumulative emissions total that is lower by the equivalent of 9 years emissions at the end of the period, loosely speaking “gaining” an additional 9 years of time.[1]

Second, the undeniable primacy of cumulative emissions implies that a rational approach to the design of an international regime and associated market mechanisms is also likely to be based on cumulative emissions from a baseline, with carryover of emission rights/savings between time periods, not on rigid annual numbers. Aligning national targets consistently with the shape and structure of future international regimes, including the ETS, will be essential.

Larger early reductions, if they can be achieved and sustained, are disproportionately beneficial in reducing cumulative emissions, and hence in delaying adverse climate impacts and/or easing the pace of transition to low carbon in later periods. This is one of the factors behind the recent call by a group of UK energy economists for a greater urgency in UK climate change policy.[2]

Question 1. Government approach to forecasting.

The Government’s forecasting methods may be the best that is available for a high level view of energy trends. However there are issues in forecasting and monitoring targets, which are difficult to address without significant improvements to the knowledge base. This is particularly the case with “aspirational” targets or forecasts of non-specific “efficiency” gains.

Medium and long term forecasts normally fall down either because key economic assumptions prove to be wrong, for example on GDP growth or relative fuel prices, or because of new trends and relationships, including technical change. In a policy context this creates the risk of not very useful debates around the “counterfactual”, a hypothetical “what might have happened but for…”, rather than the policies themselves.

The real priority has to be the effectiveness of policy and the achievement of targets. Monitoring short term targets should be about evaluating the effects of particular initiatives and policies. This requires more attention to the detail and understanding of exactly how energy is used in particular sectors. Some aspects are relatively easy, such as modelling how changes in relative fuel prices for gas and coal impact on the carbon emissions from a given capital stock of electricity generation. By contrast it is much harder to assess the factors impacting on household use of electricity and gas.

I am not aware for example of any regular publication of estimates of how much energy is used by households for major sources of consumption such as space and water heating, cooking, refrigeration, lighting and other uses, let alone any monitoring of trends. One consequence of this may be exaggerated estimates of the carbon savings available from apparently trivial behavioural changes, for example in charging cell-phones. More seriously it makes it more difficult to assess accurately the impact of more significant changes, such as low energy lighting, or the impact of the housing stock on consumption for domestic heating. The less well founded any original savings estimates are in concrete assumptions about physical parameters, the more unreliable any monitoring will become.

An incidental but unfortunate by-product of the energy sector privatisations was the fragmentation and abandonment of the consumer research programmes carried out by the nationalised fuel industries. I would therefore recommend re-establishment of a comprehensive programme of load[3] and market research monitoring the nature of gas and electricity usage by consumers, and perhaps enhancing programmes aimed at better understanding of the dynamics of fuel use in the transport sector. Such a programme would be a relatively inexpensive form of “hard-edged”, quantified social research and could provide important inputs to policy formation across the board. Carried out on an annual basis it could provide considerable assistance in monitoring the effectiveness of policies to promote fuel efficiency, and give an early indication of, for example, the rebound effects that occur if extra efficiency is absorbed in extra consumption.

Question 2. Independent check, uncertainty and inclusion of international aviation.

The BIEE Climate Change Policy Group proposed[4] that “there should be an early and rigorous independent check on the feasibility of CO2 savings to 2020 projected to result from the enhanced ‘climate change programme’ of the July 2006 Energy Review”. Given the very limited real progress over two decades (as opposed to fortuitous reductions unrelated to carbon policy, notably the “dash for gas” or UK de-industrialisation), we need to be much more confident that this aspiration is achievable. Such an independent review might be commissioned by the new Carbon Committee.

Uncertainty should be set in the wider context of evaluating how targets are to be achieved. The correct response to uncertainty is emphasis on target achievement, not on hypothetical evaluations of excuses and counterfactuals.

International aviation and shipping are critical to a successful global outcome. Their exclusion reduces the credibility of other achievements. Hence they clearly need to be monitored, but as a distinct and separate component of the overall task, a component in which progress to international agreement will be a key factor.

Question 3. Projections to 2020 and 2050.

While alternative forecasts and scenarios for the longer term development of the energy sector can be instructive, the view of the BIEE group has been that the emphasis needs to shift to focus on the pathways required to achieve the overall target, and the policies that need to be put in place to implement them. The task of monitoring progress then becomes that of monitoring progress for a number of sectors, of which by far the most significant are electricity, buildings and transport, to ensure that the required changes are happening, and that the relevant infrastructure, market and regulatory arrangements are in place to support progress to sector targets.

To illustrate the point, I believe the arithmetic of current emissions makes it very clear that UK carbon targets can only be met if the electricity sector becomes virtually carbon free. If the position can be reached by (say) 2020 that no new generating capacity is being constructed other than renewables, nuclear or fossil with full carbon capture, then the range of possible outcomes for electricity demand in 2050, while still important, becomes much less relevant to carbon policy per se. If the sector is by then intrinsically carbon free, higher or lower levels of demand in 2050 may still have investment implications, but will be accommodated mainly by market-led adjustments to construction programmes.

SOCIAL COST OF CARBON. COST EFFECTIVENESS. APPROACH TO POLICY.

Question 4.
For the purposes of policy discussion we need to distinguish carefully between the incremental social cost, the incremental costs of mitigation, and the “market price” of carbon under particular sets of emission control arrangements, all of which are important measures. Stern makes it clear that the social cost of carbon emissions is very high, and has indicated a current value of around $85/ tonne of CO2. However Stern also makes it clear that in many sectors the cost of mitigation is well below this cost.

There is an important point on the social cost of carbon where the generally sound Stern analysis may be flawed. The Stern report suggests that the social cost of CO2 emissions rises over time because damage increases as the ppm limit is approached. I believe this is misleading. Evaluation of the economic cost of current emissions must reflect a central feature of the physical models of climate, namely the identification of the cumulative concentration of gases as the main driver of climate change and hence of negative economic consequences. If an incremental tonne of carbon emitted today results in an incremental tonne of carbon in 2050, then its future impact after 2050 is essentially the same as a tonne emitted in that year. [5] It follows that its total social cost impact will be greater by the extent to which it has already contributed to a higher concentration[6]. If this argument is correct, then $85/ tonne might be a significant understatement of the cost of current emissions.

As a very broad policy instrument, it would appear right for the government to recognise a high social cost of carbon, as part of the rational underpinning for a number of necessary policies both short term and long term, and for choices between options and sectors. However there are a number of caveats attaching that should inhibit uncritical use of the approach.

a. Regular review of carbon valuation will be needed as both the science and the economic analysis is developed. A more pessimistic outlook emanating from the science is likely to translate into higher valuations of social cost. Technical advance by contrast could reduce mitigation costs and market valuations of carbon.

b. Consistency should obtain between public policy and private economic decisions. In principle both should be based on the same valuation of carbon. Inconsistencies will arise if a high value of carbon is put into public policy making, but not into the price that consumers pay for fuel. It may for example be cost effective to subsidise a consumer to insulate a large heated garage or a heated but unoccupied second home, when the same householder, if forced to bear the cost of carbon directly in fuel prices, would choose a cheaper and more effective solution.

c. A corollary of the above is the general principle that an appropriate value of carbon should be allowed to feed through into consumer prices and into supply chains, so that consumers can make genuine economic choices informed by the high real cost of carbon. Potential hardship resulting for particular fuel-dependent sections of the community can and should be addressed through other policies.

d. Other aspects of policy choice are involved. Policy needs to be set in a context that compares relative costs of mitigation, particularly where these are in competition for the same market. Policy has to be assessed not only on cost effectiveness but also on consistency with overall delivery of carbon targets. For example a policy built around a new technology that cut by 25% the carbon emissions from fossil fired power generation might appear cost effective with a high valuation of CO2. But it would not be consistent with a CO2-free power sector; if the latter were a pre-requisite of achieving the overall target, then the technology would not pass muster as a long term solution.

Questions 5, 6. Approach to Policy.

The government needs to move quickly to put more emphasis on the policy measures that will be needed to achieve the longer term targets. We need to recognise that relatively “easy” short term measures, which include some “low hanging fruit”, can only deliver a small part of the total reduction that will be required. Longer term measures that require major systemic and infrastructure changes, or major investments, should be given at least as much urgency, and will deliver much more of the ultimate reductions required.

A sectoral approach should focus on electricity, transport and buildings.[7]

The Main Sources of UK Carbon Emissions

An estimate of current UK carbon emissions can be constructed from official fuel use data, using conventional assumptions of each fuel’s CO2 content, to show how sectoral policies and targets might be framed and prioritised. Excluding emissions within the oil and gas industries themselves, an admittedly imperfect but reasonable approximation, based on 2004 data, is the following

power generation (for final use in all sectors) xxxxxxxxxxxxxxxxxxx 34 %*
transport (mainly road but including rapidly growing aviation) xxx33 %**
domestic use of fossil fuels, mainly gas xxxxxxxxxxxxxxxxxxxxxxxx x17 %
general industrial use of fossil fuels (excl energy industries) xxxxxxx12 %
other, including commercial and public sector xxxxxxxxxxxxxxxxxx 4 %
total xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx x100%

* NB this excludes power already taken from nuclear and renewables
** aviation is often excluded from aggregate numbers for UK

Given the absence of proven near term alternatives in transport, and the inertia inherent in the building stock, this indicates very strongly that even a 60% reduction target implies that electricity has to be carbon free, and underscores the huge importance of getting electricity right. Electricity would constitute an even bigger share but for the 20% of generation that is already carbon free (some renewables but mainly nuclear). Its future importance is accentuated by its potential role to substitute in transport (electric vehicles or as a primary source for hydrogen), and to replace direct use of oil or gas in heating buildings.

Electricity is also significant because international experience shows that the capital stock could be turned over within comparatively short timescales, of one or two decades, to be virtually carbon free. The French experience (admittedly with nuclear power) is very instructive in this respect. This is also a market in which comparatively easy and very cost-effective policy initiatives to reduce uncertainty, such as a policy commitment or a floor price for CO2, could lead to significant increases in CO2 reducing investment.

More generally the BIEE group concluded[8] that the need for urgency in policy making requires that the government should demonstrate a singleness of purpose, at the earliest opportunity, by emphasising the importance of carbon targets within a “joined up government” approach. Wherever possible, policies to meet other objectives (eg fiscal, housing and other policies) should be consistent with and should not obstruct CO2 reduction. Where other policies do run counter to CO2 reduction, it should be made explicit that additional countervailing measures will be needed.

ACCOUNTABILITY TARGETS AND REPORTING.

Questions 7,8. The BIEE group[9] argued that there is a strong case for aligning the main sectoral targets with government departments and introducing ministerial responsibility for them. Annual cumulative emissions should be monitored and any excess over the target path should be justified and addressed convincingly in the evolution of policy.

It is important that reporting and monitoring should be viewed not only in terms of the annual reduction of emissions against the required trend, but also in terms of progress with measures necessary to secure sustainable momentum in future years. The use of an independent agency to monitor whether policies of individual ministries deliver on an annual basis, and also on mid-term targets, is attractive. Separating design and implementation from monitoring could increase the credibility of the monitoring agency and thus improve the monitoring and enforcement of targets.

The composition of the Carbon Committee should not be based on special interest groups, as this would weaken its independence and its credibility. It is important that it should include a substantial body of scientific, engineering and economic expertise to provide a good practical understanding of the many complex issues with which it will have to engage.

COMPETITIVENESS ISSUES.

Question 9. I believe that, in the context of measures by the UK to combat emissions, the significance of competitiveness, as an issue inhibiting unilateral action, has been greatly exaggerated. First, an analysis of the sources of emissions makes it clear that direct industrial use of fossil fuels (other than for power generation) is quite a small part of the total, so targeting industry is not a first priority. Second, the measures necessary to contain emissions in the most important sectors, electricity[10], housing and transport, will mostly have only small and indirect effects on industry costs, and in terms of competitiveness are dwarfed by the much more direct and overarching effects of exchange rate movements. Third, those fuel intensive industries which are subject to international competition account for only a small percentage of GDP. In this context it is useful to distinguish intra-EU and extra-EU competition. UK power costs for industry are already higher[11] than in much of the EU. In shaping the EU ETS we should certainly aim to prevent economic distortions that merely “export” emissions to countries outside the EU with lower energy efficiencies, but this is best done either by limited “ring fencing” for the few industries concerned, or by pursuing wider international agreement.

SETTING SHORT TERM TARGETS

Question10. Targets should be framed to reflect emphasis on the longer term objective of cumulative emissions; they should have a sectoral element to reflect individual ministerial responsibilities, be realistic on short term achievements and be capable of being monitored in fairly concrete terms; they should cover not only CO2 emissions but also progress with the fundamental longer term systemic aspects of policy.





[1] Likewise a back-end loaded reduction path adds a similar and substantial amount to cumulative emissions.
[2] BIEE Climate Change Policy Group. Bringing Urgency Into UK Climate Change Policy. December 2006
[3] For electricity, load research is based on sample recordings of load on individual circuits or appliances.
[4] BIEE Climate Change Policy Group. Bringing Urgency Into UK Climate Change Policy. December 2006
[5] A possible partial counter argument to this might apply if the extra emission were to increase the re-absorption rate; however with limited re-absorption or positive feedbacks it is perhaps more likely that an incremental tonne of carbon emission results in more than one incremental tonne in 50 years time.
[6]The comparison ignores time discounting in the context of this particular argument.
[7] Note that, if we exclude its share of electricity, industry is of less significance.
[8] BIEE Climate Change Policy Group. Bringing Urgency Into UK Climate Change Policy. December 2006
[9] Ibid.
[10] In the case of electricity, it may also be noted that French electricity for industry, already essentially carbon-free, has been very competitively priced.
[11] Arguably just part of an exchange rate issue.

No comments: