56563 REV Climate Change and Fiscal Policy: A Report for APEC Office of the Chief Economist East Asia and Pacific Region The currency unit is US Dollars unless mentioned otherwise. Vice President James W. Adams (EAPVP) Chief Economist and Sector Director, Vikram Nehru (EASPR) PREM and FP, EAP Task Team Leader and Lead Ahmad Ahsan (Office of the Chief Economist Economist and EASPR) Preface This report was prepared as part of the APEC Finance Ministers` Policy Initiatives of 2008. Under this initiative, the World Bank was asked to prepare studies on the current state of economic policies concerning climate change and recommendations for strengthening these policies. This is one of the background studies which have been prepared. The background studies and a synthesis report based on these studies were presented to APEC bodies such as the Senior Finance Officials Meetings in September 22, 2010. The final versions are being tabled at the Finance Ministers` meetings in November, 2010. The authors of the report are Stephen Howes and Leo Dobes from the Australian National University. Other members of the report team are introduced on the next page. Ahmad Ahsan, Lead Economist, Office of the Chief Economist, East Asia and Pacific Region, World Bank provided much practical help and guidance throughout the preparation of this report. Carter Brandon, Lead Environmental Specialist in the Beijing Office of the World Bank, helped with regard to research on China. Other World Bank staff, especially in the Vietnam and China offices, assisted in various ways in the course of the preparation of the report. The authors would also like to thank their colleagues in various research institutions, government ministries, diplomatic missions and aid agencies in both Hanoi and Beijing. They are too numerous to mention here, but provided invaluable assistance. Valuable comments on the first draft of the report were provided by: Professor David Victor, Director, Laboratory on International Law and Regulation, School of International Relations and Pacific Studies, University of California, San Diego; Mike Toman, Research Manager, World Bank; Apurva Sanghi, Team Leader, Joint World Bank - UN Project on Economics of Disaster Risk Reduction; Jim Brumby, Acting Director, Governance & Public Sector, Poverty Reduction and Economic Management, World Bank. John Roome, Director, Sustainable Development Network, East Asia and Pacific Region, and Vikram Nehru, Director, Poverty Reduction and Economic Management, Finance, Private Sector Development, and Chief Economist, East Asia and Pacific Region, chaired the Review Meetings for the report, and gave useful guidance. Syud Amer Ahmed, Development Economics Research Group, Katherine Patrick and Mildred Gonsalvez, East Asia and Pacific Region, helped to process this document for publication. This Report was circulated to the APEC Senior Finance Officials` Meetings in September 2010 in Tokyo and its findings were presented there as part of the presentation on the Climate Change and Economic Policies in APEC Economies: Synthesis Report. Valuable comments were provided by participants in that meeting. Funding from the Environmental Economics Research Hub financed the visit of Leo Dobes and Nguyen Van Kien to Vietnam. Project Team This report was written by Stephen Howes (team leader and main author of Chapters 1-4) and Leo Dobes (main author of Chapter 5) of the Crawford School of the Australian National University. Important contributions were made by Peter Downes of Outlook Economics (to Chapters 1 and 2), Eric Knight (to Chapter 2) and Kurnya Roesad (to Chapter 3). Excellent research assistance was provided by Seungwon Chung, Huw Slater, Feng Shenghao, Nguyen Van Kien, and Sabit Otor. Useful advice, suggestions and feedback were provided by Frank Jotzo. Professor Stephen Howes is Director, International and Development Economics at the Crawford School of the Australian National University. He has spent the last 20 years at the London School of Economics, the World Bank, AusAID, and now the Australian National University working on the economies of the Asia-Pacific region, with a particular focus on the energy sector and on fiscal policy. He has been active on climate change mitigation policy since 2008, when he worked on the Australian Garnaut Climate Change Review. Dr Leo Dobes is Adjunct Associate Professor at the Crawford School of the Australian National University. Following a DPhil (Oxford), he worked for almost 30 years in public service positions ranging from the diplomatic service to the Australian Treasury. In 1992 he established an Environment Branch within the Australian Bureau of Transport Economics, publishing a number of important reports on the costs and benefits of mitigating emissions in the transport sector. He is now one of Australia`s leading economists on climate change adaptation, and serves as an occasional advisor to the Australian Department of Climate Change and Energy Efficiency. Peter Downes is one of Australia`s leading macroeconomic modelers. After a twenty-year career with the Australian Treasury, Department of Prime Minister and Cabinet and the Organisation for Economic Co- operation and Development in Paris, he now runs his own economic modeling and forecasting consultancy, Outlook Economics, and is a Research Associate with the Centre for Climate Economics and Policy at the ANU Crawford School. Dr Frank Jotzo is a Senior Lecturer at the Crawford School of the Australian National University and Director of the Centre for Climate Change Economics and Policy at the ANU. He is an environmental economist specializing in the economics and policy of climate change. He has worked and published on these and other aspects of international and development economics since 1998. He has worked and consulted for several governments and international organizations, and was lead consultant for the 2009 Indonesian Ministry of Finance Green Paper on Climate Change. Eric Knight, Kurnya Roeasad, Feng Shenghao and Nguyen Van Kien are PhD students in various aspects of climate change at Oxford (Knight) and the ANU. Seungwon Chung, Huw Slater, and Sabit Otor are postgraduate students, past or present, of the Australian National University. Table of Contents Overview ....................................................................................................................................................... i Executive Summary .................................................................................................................................... ii Chapter 1 Goals and targets: climate change mitigation and related policy objectives ....................... 1 1.1 Introduction ........................................................................................................................................ 1 1.2 Climate change mitigation and clean energy targets among APEC economies ................................. 5 1.3 Goals behind the climate change and clean energy targets ................................................................ 8 1.4 From targets to instruments .............................................................................................................. 15 Chapter 2 The instruments: fiscal policies for mitigation ..................................................................... 18 2.1 Types of instruments ........................................................................................................................ 18 2.2 Carbon pricing ................................................................................................................................. 19 2.3 Technology-based policies ............................................................................................................... 27 2.3.1 Technology policy rationales ........................................................................................................ 27 2.3.2 Technology policy options ............................................................................................................ 30 2.4 Conclusion ........................................................................................................................................ 36 Chapter 3 The context: energy sector and other important characteristics relevant to mitigation instrument choice. ..................................................................................................................................... 37 3.1 Introduction ...................................................................................................................................... 37 3.2. The energy sector in developing economies .................................................................................... 37 3.2.1 Rapid growth ................................................................................................................................. 38 3.2.2 Ongoing importance of traditional energy .................................................................................... 39 3.2.3 Energy as a luxury good................................................................................................................ 40 3.2.4 Energy subsidies ........................................................................................................................... 42 3.2.5 Price setting ................................................................................................................................... 49 3.2.6 Energy rationing............................................................................................................................ 51 3.2.7 Captive power ............................................................................................................................... 52 3.2.8 Flexibility in dispatch.................................................................................................................... 52 3.2.9 Dominance of the sector by vertically-integrated state-owned enterprises. .................................. 53 3.2.10 Reliance on central planning to guide generation expansion ...................................................... 54 3.2.11 Lack of commercial orientation .................................................................................................. 55 3.2.12 Difficulty of reform ..................................................................................................................... 56 3.3 Other relevant features of some developing countries...................................................................... 57 3.3.1 Factor market distortions .............................................................................................................. 57 3.3.2 Compensation instruments ............................................................................................................ 58 3.3.3 Institutional capacity ..................................................................................................................... 59 3.4 Conclusion ........................................................................................................................................ 59 Chapter 4 Choices: mitigation policies for developing countries ......................................................... 61 4.1 Implications for carbon pricing ........................................................................................................ 61 4.1.1 Is carbon pricing desirable? .......................................................................................................... 62 4.1.2 Is carbon pricing feasible? ........................................................................................................... 63 4.1.3 How important are and what are the mitigation implications of energy sector reforms? ............. 66 4.1.4 How important are broader economic reforms for mitigation? ..................................................... 67 4.2 Technology-based policies ............................................................................................................... 68 4.2.1 Technology-based policies and carbon pricing ............................................................................. 68 4.2.2 Research and development ............................................................................................................ 69 4.2.3 Financing....................................................................................................................................... 71 4.2.4 Policy rigour.................................................................................................................................. 71 4.3 Conclusion ........................................................................................................................................ 72 Chapter 5 Fiscal aspects of adaptation to climate change ..................................................................... 74 5.1 Adaptation policy processes in China, Vietnam and Indonesia....................................................... 75 5.2 Costing adaptation ........................................................................................................................... 76 5.3 Decision tools and criteria for adaptation fiscal policy-making ..................................................... 77 5.3.1 Multi-criteria analysis .................................................................................................................. 77 5.3.2 Vulnerability`, adaptive capacity` and resilience` indexes ...................................................... 79 5.3.3 Cost-effectiveness analysis .......................................................................................................... 79 5.3.4 Cost-benefit analysis .................................................................................................................... 80 5.4 Incorporating uncertainty into adaptation decision-making ............................................................ 82 5.4.1 Extreme value distributions and Monte Carlo methods ................................................................ 82 5.4.2 The real options` approach ......................................................................................................... 85 5.5 Instruments for adaptation .............................................................................................................. 89 5.5.1 Public goods .................................................................................................................................. 90 5.5.2 Pricing .......................................................................................................................................... 91 5.5 3 Financial instruments .................................................................................................................... 92 References .................................................................................................................................................. 96 Appendix Tables...................................................................................................................................... 109 Figures Figure .1: Developed and developing economies in APEC ........................................................................ 1 1 Figure .2: Per capita emissions and income are closely correlated across APEC ....................................... 2 1 Figure .3: China: rise of a steel giant .......................................................................................................... 3 1 Figure .4 :In most developing economies, CO2 emissions track GDP ........................................................ 4 1 Figure .5: Energy intensities are converging across APEC ........................................................................ 4 1 Figure .6 : Many APEC Economies have set ambitious emissions reduction targets ................................. 5 1 Figure .7: Achieving national targets will mean a very different emissions future for APEC ................... 7 1 Figure .8: Most APEC developing economies are already or will become energy importers .................... 9 1 Figure .9: Apart from Russia, Australia and the US, APEC countries have little coal and/or it won`t last 1 long ............................................................................................................................................................. 10 Figure .10: China, long and increasingly dependent on oil imports, is now also a net coal importer....... 11 1 Figure .11: World energy prices have been volatile and rising over the last decade ................................ 12 1 Figure .12: Korea`s Green Growth Strategy ............................................................................................. 13 1 Figure .13: Steel, cement, electricity and electricity from coal production, as well as coal consumption, 1 have all risen relative to GDP since 2005 .................................................................................................. 17 Figure .1: Developed countries have low energy prices or high energy efficiency, but not both ............. 20 2 Figure .2: Most models show that a given carbon price will impose proportionately greater economic 2 costs on a developing economy than a devloped economy......................................................................... 23 Figure .3: A $20 carbon tax in 2020 would raise significant government revenue, especially for the 2 poorer APEC economies ............................................................................................................................. 24 Figure .4: A global increase in gasoline tax rates would make energy importers better off ..................... 26 2 Figure .5:The long innovation chain: how a new invention gets to the market ........................................ 29 2 Figure .6: Energy R&D funding by OECD countries has stagnated since the second oil crisis, but is now 2 starting to pick up........................................................................................................................................ 32 Figure .1: Energy and especially electricity is by far the biggest source of CO2 emissions worldwide. .. 37 3 Figure .2: Poorer countries experience faster energy growth ................................................................... 38 3 Figure .3: In most APEC economies, electricity grows faster than both GDP and energy ....................... 39 3 Figure .4: Almost half of Indonesia`s fuel subsidies benefit the richest 10% of households ................... 41 3 Figure .5: APEC developing economy oil and gas subsidies: increasing in absolute terms but not 3 necessarily relative to GDP ......................................................................................................................... 42 Figure .6: Pump prices in APEC countries show a lot of variation, but are not strongly related to income 3 .................................................................................................................................................................... 43 Figure .7: The era of cheap coal in China is over ..................................................................................... 44 3 Figure .8: Despite rapidly rising coal prices, electricity prices for industry and household ..................... 45 3 Figure .9: Coal fuel costs are squeezing margins in the electricity sector ................................................ 45 3 Figure .10: Indonesia`s industry pays close to prices prevailing in developed countries for their 3 electricity, but its households pay only half ................................................................................................ 46 Figure .11: Developing economies show a lower ratio of industrial to household tariffs than developed 3 ones ............................................................................................................................................................. 47 Figure .12: Tariffs are uniform across Indonesia's regions but costs vary by factor of almost three ....... 47 3 Figure .13: Low electricity prices in APEC developing economies manifest themselves in low prices for 3 households not industry .............................................................................................................................. 48 Figure .14: Indonesia's energy subsidies have become a major and persistent claim on the budget ........ 49 3 Figure .15: Petrol prices in China follow world prices except when world prices are very high ............. 50 3 Figure .16: Petrol prices in APEC`s developing countries have diverged over the last decade ............... 51 3 Figure .17: China`s 2020 generation targets aim to reduce the dominance of coal-fired generation, while 3 doubling total capacity ................................................................................................................................ 55 Figure .18: Surging investment in China and Vietnam ............................................................................. 57 3 Figure .19: Some developing economies are characterized by financial repression ................................. 58 3 Figure .20: Poorer countries tend to have lower government effectiveness, regulatory quality and control 3 of corruption................................................................................................................................................ 59 Figure .1: China's wind industry is still far from the technological frontier ............................................. 71 4 Figure .1: Illustration of annual exceedance probabilities (AEP) for rainfall and floods ......................... 82 5 Figure .2: Uncertainty illustrated by a range of distributions for each year under study .......................... 83 5 Figure .3: Illustration of cost functions generated from flood probability distributions .......................... 83 5 Figure .4: Two different approaches to building a dike to respond to increased flooding risk ................. 86 5 Figure .5: New housing design in coastal areas in Indonesia ................................................................... 87 5 Tables Table .1: Climate change mitigation and renewable energy targets for APEC economies ......................... 6 1 Table .2: Most developed countries show growth from fossil fuel emissions despite their Kyoto 1 commitments to reduce emissions .............................................................................................................. 15 Table .1: Classification of climate-change mitigation instruments ........................................................... 18 2 Table .2: Progress across APEC in introducing carbon pricing ................................................................ 22 2 Table .3: Long-run economic effects from revenue-neutral tax reductions: an example from Canada .... 25 2 Table .4: Top 12 inventors in climate change mitigation technologies, with average percentage of total 2 global inventions across different mitigation technologies ......................................................................... 28 Table .5: APEC economies with feed-in tariffs ........................................................................................ 30 2 Table .6: Examples of biofuel targets in APEC economies ...................................................................... 31 2 Table .7: Goals, targets and instruments in climate change and related areas .......................................... 36 2 Table .1: The great majority of households in APEC developing countries have electricity, but over half 3 continue to rely on biomass and coal for cooking and heating ................................................................... 40 Table .2: The share of expenditure on modern energy (biomass) rises (falls) with income: energy 3 expenditure shares for the poorest and richest 20% of households for various Asian countries ................ 41 Table .3: Power sector structure and ownership in APEC economies ...................................................... 54 3 Table .4: Characteristics of the energy sector and the broader economy of developing economies ......... 60 3 Table .1: Illustrative example of simplified multi-criteria analysis: hypothetical dike building project .. 78 5 Boxes Box .1: Taxing fossil fuels and pricing carbon improves the terms of trade for energy importers ........... 13 1 Box .2: What has happened to energy intensity in China since 2005? .................................................... 17 1 Box .1: China`s instruments for achieving a 20% reduction in energy intensity ...................................... 21 2 Box .2: Real-world experience with carbon pricing ................................................................................. 22 2 Box .3: Feed-in tariffs versus Renewable Energy Certificates ................................................................. 31 2 Box .4: Green Investment Banks for the UK and the US?........................................................................ 33 2 Box .5: Same end, multiple means: the promotion of wind energy in China and the United States ........ 34 2 Box .6: Energy efficiency gone wrong: Australia`s aborted home insulation program ............................ 35 2 Box .1: Does reducing energy subsidies always reduce emissions? ......................................................... 67 4 Box .2: Why Indonesia`s feed-in tariff for renewable energy hasn`t worked. .......................................... 69 4 Box .3: Cheap but not easy: reducing emissions from deforestation and forest degradation. .................. 73 4 Box .1: Costing adaptation: lessons from the Millennium Development Goals ....................................... 77 5 Box 5.2: Cost-benefit analysis of dikes in the Mekong Delta region ........................................................ 80 Box 5.3: Cost-cost` versus cost-benefit` ................................................................................................. 81 Box 5.3: Application of Monte Carlo analysis to adaptation to coastal inundation .................................... 84 Box 5.5: Real options in Vietnamese housing ........................................................................................... 88 Box 5.6: Adaptation in rural Ningxia Hui Autonomous Region ................................................................ 90 Box 5.7: Agricultural responsiveness to increased water prices in China .................................................. 91 Box 5.8: Small rural loans in China: redressing the balance ..................................................................... 93 Overview APEC economies display large variation in terms of income per capita. The richest APEC economies have an income per capita about twenty times higher than the poorest ones. So far most work on fiscal policy and climate change has been written with developed economies in mind. This report corrects that bias with a particular focus on the developing economies of APEC. It draws on examples from three developing economies in particular, China, Indonesia and Vietnam. It also plays close attention to lessons that could be learnt from the advanced economies of APEC and elsewhere. On mitigation, the report notes that many developing as well as developed economies have now proposed or adopted emissions reduction targets. These targets are ambitious, and their achievement will not be easy. Instrument choice will be the difference between success and failure, and is a major theme of the report. The key message is that the characteristics of developing economies, particularly of their energy sectors, matter a lot when it comes to policy choice in this area. Simply mimicking what has been adopted in or recommended for developed countries is unlikely to work. Mitigation in developing countries requires a broad-based response with four key components. First, carbon pricing is critical, but on its own will not suffice, and in some economies and some sectors may have little or no impact due to pre-existing distortions. Second, energy sector reforms in many countries will be a prerequisite for effective mitigation, though they may on their own increase emissions. Of particular importance are policies to allow for cost pass-through in the energy sector ­ so that subsidies do not re-appear, and so that carbon prices can be passed on. Without this, carbon pricing will lack both signaling power and credibility. Enabling cost pass-through will require liberalization of energy markets and the establishment of independent regulators, both formidable tasks. Third, broader economic reforms may also be important. Broader policy settings in some economies may bias economic growth to be more capital and energy intensive than is optimal. Fourth, technology-based mitigation policies will also be needed, but, given the mixed track record in this area, must be chosen with care. Given the many uncertainties involved, and the multiple reforms needed, a verifiable quantity anchor for mitigation policy is recommended for developing economies, such as the energy-intensity target recently adopted by China. Fiscal analysis of adaptation has so far largely focused on cost projections, but for policy makers adaptation instruments and decision-making tools are as or more important. Adaptation instruments include the provision of public and club goods (such as infrastructure), public sector pricing reform (in particular of water) and financial instruments (microcredit and insurance) which can be cost-effective alternatives to subsidies. Key to the right choice of instruments (which will vary from location to location) will be the correct use of appropriate decision-making tools. In particular, the social costs and benefits of alternative strategies need to be analyzed under conditions of uncertainty, in many ways the hallmark of climate change. Popular tools such as multi-criteria analysis, vulnerability indexes, and cost- effectiveness analysis are inadequate to the task. A combination of Monte Carlo and real options` analysis within a cost-benefit framework is recommended. Examples from a range of economies are provided to demonstrate the utility of such an approach. i Executive Summary 1. The reports main focus is on mitigation, the subject of Chapters 1-4 of this report. Adaptation is covered in Chapter 5. This allocation of space does not reflect any judgment on the relative importance of the two topics. Rather, the field of fiscal policy and adaptation is in its infancy. The chapters follow in logical sequence, and each is summarized below in the five sections which follow. 2. Space constraints inevitably force various limitations on the scope of this analysis. Two are particularly important. First, in relation to mitigation, the focus is on carbon dioxide (CO2) from fossil fuels. Though other greenhouse gases and sources are also important, CO2 from fossil fuels is the largest single source of greenhouse gases, and the fastest growing. Almost 80% of greenhouse gas emissions are in the form of CO2. About one-quarter of this is through deforestation. The rest is from fossil fuels. Electricity generation is the largest single source of fossil fuel emissions, and a special focus of this report. Other sectors also burn fossil fuels, including transportation, industry, and services as well as households. Second, in relation to both mitigation and adaptation, the focus is on domestic fiscal policy, not on international issues. Of course, both are important, but international funding and technology transfer have been well covered by a number of reports (such as World Bank, 2010d). The Copenhagen Green Climate Fund and Technology Mechanism, agreed to as part of the Copenhagen Accord, would provide a natural basis for international cooperation within APEC. 1. Goals and targets: climate change mitigation and related policy objectives 3. APEC economies as a group are responsible for almost two-thirds of annual global CO2 emissions. Richer economies have much higher per capita and also cumulative emissions. While the USA is no longer the world`s largest emitter, it is responsible for the grea test volume of accumulated emissions. This report`s focus on developing economies notwithstanding, without leadership from the US and other developed economies, there can be no effective global response to climate change. China`s rise to industrial power status has accelerated global emissions growth. China is now the largest emitter of CO2 from fossil fuel, with 25% of global emissions in 2009 compared to 17% from the USA. Consequently, China`s leadership will also be critical. 4. More generally, developing country emissions have been growing rapidly and, absent the introduction of policies to mitigate climate change, will continue to do so. Fortunately, many economies, both developed and developing, have decided that business as usual` is no longer an option, and have adopted targets to constrain their emissions growth. Recognizing the urgency of a global response to climate change, many developing economies have recently made self-imposed, non- internationally-binding commitments through the Copenhagen Accord to constrain the growth of their own emissions. Most of the targets adopted are ambitious ones. There is little difference between the ambition of APEC developed and developing economies; if anything, the latter are actually more ambitious (Figure 1). ii Figure 1 Many APEC Economies have set ambitious emissions reduction targets 2020 APEC emissions control targets expressed as a reduction relative to business as usual (BAU) Notes: Emissions reduction targets as reported in the Copenhagen Accord except for Taiwan. Where a target range has been committed to, the mid-point of that range is selected. Source: Table 1.1 and Figure 1.7. 5. If implemented, the targets would fundamentally alter the emissions trajectories of APEC economies (Figure 2). Under business as usual, by 2020 China will have not only Korea`s per capita income of today but also Korea`s current per capita emissions, which at 10 tonnes of CO2 per person are more than double China`s current level. However, things will be very different if the APEC economies are able to achieve their national targets. There will be significant convergence between developed and developing economies; developed economies will have shown that significant absolute emissions reductions are possible, and developing economies will have avoided a massive ramping up of emissions: China`s per capita emissions at the end of the decade will be 7 rather than 10 tonnes, a saving of about 4 billion tonnes of CO2 (more than 10% of current annual global emissions of CO2 from fossil fuels). iii Figure 2: Adhering to national commitments will mean a very different emissions future for APEC Per capita CO2 emissions, historical (1971 to 2007) and projected to 2020, both (under business as usual, and assuming targets are met) for various APEC economies Notes: Solid lines are historical. Upper dotted lines show business as usual. Lower dotted lines show trajectories assuming national commitments (from the Copenhagen accord, except for Taiwan) are adhered to. The historical period covered is from 1971 to 2007; the projection period is from 2007 to 2020. Emissions per capita are assumed to grow/decline in a linear manner over the projection period. These graphs are based on the simplifying assumption for most economies that targets announced for all greenhouse gases will be adhered to for CO2 from fossil fuels as well (the graph shows the latter only). Where a target range is provided, the mid-point of that range is selected. Source: Historical data from IEA (2009a); Garnaut et al. (2009), Table 1.1, and own calculations for projections. 6. An even larger number of APEC economies have embraced renewable or clean energy targets. Worldwide, renewable energy targets are more popular than mitigation targets, with the former having been established by at least 73 economies globally as of 2009. Fourteen APEC countries now have renewable energy targets. Some of them are also ambitious. 7. While the focus of this report is on climate change mitigation, it is important to understand how mitigation objectives and targets sit within the broader policy goals. Why have so many APEC economies adopted both climate change and clean energy targets? Of course, leaders are increasingly worried about the impact of climate change. But three other goals are also driving action. First, policy- makers are seeking to tackle national environmental problems, in particular air pollution and acid rain. Second, energy security is a growing concern, as dependency on energy imports and global energy prices rise. Apart from Russia, Canada, Australia, and Brunei, APEC economies either already are or will become energy importers. Third, economies are also seeking technological advantage, and see low- carbon technologies as a growth opportunity for the future. iv 8. There are synergies between these goals. By pushing down global energy prices, global action on climate change would improve the terms of trade for most APEC developing economies, and thus improve energy security. As demonstrated in the report, improvements in the terms of trade, as well as the use of carbon revenue to offset other taxes, could partially or perhaps fully offset the costs of mitigation. Lower emissions also means cleaner skies. Promoting renewable energy addresses national and international pollution concerns, as well as energy security, and also, it is hoped, might give economies first-mover technological advantages. 9. But there are also trade-offs. Energy security concerns focus mainly on oil, climate change concerns mainly coal. An emissions reduction target on its own could increase reliance on oil relative to coal. Likewise, some measures to improve energy security could increase emissions via the reverse substitution. Conversely, some measures to improve energy security (by reducing oil use) could increase emissions via the reverse substitution (into coal). A combination of targets ­ both to reduce emissions and to expand clean energy ­ therefore can enable APEC economies to pursue the mix of goals that they have. 10. The key risk with APECs climate change mitigation and renewable energy targets is not that they might not be ambitious enough to prevent dangerous climate change, but rather that they might not be achieved. The difficulty of achieving emissions reduction targets is underlined by the failure of many developed economies to reduce emissions in line with their Kyoto Protocol targets and by the resumption of rapid emissions growth in China. Despite the global financial crisis, and despite the strenuous policy efforts of the Chinese government, emissions growth in China continues to exceed the business as usual projections of analysts. 11. The instruments chosen will be the difference between success and failure. So far, APEC economies have used a mix of regulatory and fiscal technology-specific measures to achieve their goals. But, as shown in the next section, carbon pricing is increasingly on the agenda. Given the ambitious targets now in place, now is the time to consider the full range of instruments, and to adopt the most effective. 2. The instruments: fiscal policies for mitigation 12. Climate change mitigation instruments can be divided into carbon-pricing and technology-based policies. Carbon pricing policies include a carbon tax, emissions trading schemes, and hybrids of these two approaches. All other policies are technology-based (or, simply, technology) policies because they are all, to some extent, technology-specific. A feed-in tariff can only be set for a type (or types) of technology, such as solar or wind power. Clean energy targets and research and development subsidies, however broad their scope, have to be defined in relation to a set of clean` technologies. By contrast, carbon pricing instruments are technology neutral: they do not require the government to pick winners`. Note too that whereas carbon pricing instruments are by definition fiscal, technology-based instruments can be fiscal or regulatory. Table 1 provides an illustrative classification of policy instruments along these lines. v Table 1: Classification of climate-change mitigation instruments Carbon pricing Technology-based Fiscal Fiscal Regulatory Emissions trading Demonstration grants Technology performance Carbon tax Public R&D standards Hybrid trading-tax Investment subsidies Renewable fuel/energy schemes Preferential tax standards treatment Building regulations Government investment Automobile regulations in venture capital Information standards Public investment vehicles Feed-in tariffs Tax credits Public procurement Renewable energy certificate trading Subsidies for energy- efficiency purchases 13. Broader energy policies and structural reforms are also important for climate change mitigation. All parts of the economy use energy and emit CO2. Energy sector reforms in particular, and economic reforms more generally can have a powerful impact on emissions trajectories. This is a theme picked up later in the report. This section discusses both carbon pricing and technology-based policies. Carbon pricing 14. Carbon pricing is essential for effective climate change mitigation. By pushing up the relative price of emissions-intensive goods, a carbon price reduces emissions in four ways. First, it pushes consumer demand in the direction of goods which are less emissions intensive (e.g. to wear extra clothing and turn down the heating). Second, it induces suppliers to make their goods less emissions intensive (e.g. to make electricity with gas instead of coal). Third, it leads investors to invest in less emissions- intensive projects (e.g. to build an aluminum smelter powered by hydro rather than thermal electricity). And, fourth, carbon-pricing gives a financial incentive for innovators to develop new products, which are less emissions-intensive (e.g. to invent a hydrogen or electric car). 15. International comparisons of energy prices and usage point to the importance of pricing as a determinant of energy efficiency. To simplify somewhat, the message from Figure 3 is that the US and Canada have electricity and gasoline prices at 50% below the levels prevailing in Japan and Europe, and energy per unit of output at 50% above. No doubt the relationship runs both ways (with higher energy intensity in North America leading to political resistance to tax hikes), but it is equally clear that higher energy prices will encourage energy efficiency, and that, as a special case of this, the introduction of carbon pricing will discourage fossil fuel use. The European Union`s experience with emissions trading confirms that putting a price on emissions does lead to abatement. vi Figure 3: Developed economies have low retail energy prices or high energy efficiency, but not both Electricity prices, gasoline prices, and energy intensity (ratio of energy use to GDP) for US, Canada and Japan relative to the OECD member economies of Europe. Notes: Prices measured in current USD, using market exchange rates. For energy intensity definitions, see Figure 1.4. Energy efficiency is defined as the inverse of energy intensity. All OECD Europe values are normalized to one. Sources: IEA (2009a, 2010b) 16. To date, however, little use has been made of carbon pricing. Technology-based policies have been much more popular. Among APEC economies, only New Zealand, and some American and Canadian states have actually introduced a price on carbon. In some economies, carbon pricing is still politically controversial, especially because it has not been adopted by the world`s largest economy, the United States. APEC economies, including developing ones, are however showing increasing interest in carbon pricing. An emissions trading scheme is being prepared and/or debated in the US, Australia, Japan, and Korea. China and Indonesia are both contemplating the introduction of carbon pricing. Both a carbon tax and emissions trading seem to be under consideration in China. Some developing economies (in and out of APEC) have already introduced carbon-price-like levies, including India, Vietnam and China. 17. A carbon price can be introduced either through a price-based approach (e.g., a carbon tax) or a quantity-based approach (e.g., an emissions trading scheme), or a combination of the two. Under conditions of certainty and perfect information, a carbon tax and emissions trading scheme are equivalent. In the real world, there are pros and cons to both approaches. 18. Most models show that, absent international transfers, carbon pricing will be more costly to developing than to developed economies. However, this modeling typically ignores the revenue benefits of carbon pricing which could be substantial, especially in developing economies. A $20 carbon price applied across fossil fuels could fetch China in excess of 2.5% of GDP by 2020. If this revenue is used to reduce other taxes, or to support productive spending, then the costs of carbon pricing would be reduced. The evidence presented in this report suggests that, with terms of trade gains as well, carbon pricing could vii possibly be a win-win reform, with short-term economic gain as well as long-term environmental protection, at least for energy importers. Technology-based policies 19. Governments have a wide-range of reasons for introducing technology-based policies. This report enumerates seven rationales. As discussed earlier, many governments have renewable energy targets and industrial policy objectives. Technology-based policies are as important for clean energy targets and industrial policy objectives as carbon prices are for emissions reduction ones. Inventions and discoveries have public good characteristics. Firms under-invest in research and development because of the fear that their competitors will benefit. Trading off the need to provide incentives to invent (prior to the invention) and the need to make maximum use of any inventions (once they are made) is an important responsibility for governments which they discharge not only through the creation of patent systems but also through public funding for research and development (R&D). APEC, led by Japan, dominates global innovation in climate change technologies. 8 of the top 12 most inventive economies in the area of climate change mitigation are in APEC, with Japan alone being responsible for 37% of the world`s climate change mitigation inventions. The US is in second position, and China, South Korea and Russia in 4th, 5th and 6th positions respectively (Dechezleprêtre et al., 2008). Public promotion of new technologies, beyond support for research, can be justified by consideration of dynamic increasing returns generated by learning-by-doing, learning-by-using and network externalities. Successful innovation is a long and arduous process. The average period for taking a new energy technology to market ­ to traverse the valley of death` as it is often called ­ is 20 to 30 years (Lee, Iliev and Preston, 2009). In such an environment, early- movers generate spill-over effects which are of benefit to society but cannot be privately appropriated. Policy risk is unavoidable for renewable energy. Given that no renewable energy technology has yet reached price parity in its production costs with coal-fired electricity, a profitable return on the development of these technologies may be dependent on many years of government subsidy, whether through a carbon price or other means. This once again makes renewable energy projects quite different from other areas of high innovation, such as information technology. Technology-based policies can help reduce policy risk by providing support upfront (e.g., through capital subsidies) rather than over time and/or by embedding support that is provided over time into legally binding contracts (e.g., through feed-in tariffs). A fifth rationale for technology policies rests on capital market failures. Research has shown that a combination of large upfront costs and high risk profiles can make renewable energy demonstration projects unsuitable for both venture capital and commercial financing and therefore leave them with inadequate market financing. One option which governments are increasingly considering is to intervene directly through loans or guarantees. Technology-based policies are not only about new technologies. A range of information and agency barriers can restrict the use of known technologies, including those for greater energy efficiency. In the most well-known case, landlords, who pay for capital costs, will under- invest in energy efficiency, since the benefits in terms of reduced energy bills will flow to their tenants. Or consumers, because they lack the relevant information, will judge durables on the viii basis of upfront rather than lifetime costs. The policies to respond to these market failures are especially important for efforts to promote energy efficiency. They are not analyzed in detail in this report, which focuses on clean energy, but the general lessons which are drawn about technology-based policies apply more generally. Finally, technology-based policies can sometimes be considered substitutes for rather than complements to carbon pricing. As noted, carbon pricing is still controversial in many economies. Some economies have been unable to generate political support for carbon pricing but have for technology-based policies. In the same year in which the Australian Parliament voted down an emissions trading scheme, it provided bipartisan support for a renewable energy credit trading scheme. 20. Some of the outcomes sought by technological policies, such as cost reductions and technological breakthroughs, are inherently difficult to observe, and harder still to attribute to policies. However, experience suggests both supply-side and demand-side technology policies can be effective in promoting the advance and spread of particular technologies. 21. However, more than a strong rationale and the likelihood of an impact are required to justify the introduction of any specific technology-based policies. Costs and risks also need to be borne in mind. The high feed-in tariff offered for solar power in Germany has been successful at getting greater use of PV solar in Germany, but at great expense. The cost of emissions savings brought about through this scheme is estimated to be in excess of 716 Euro per tonne of CO2 (Frondel et al., 2010). Likewise, the biofuel policies adopted by many economies have resulted in serious concerns about their environmental and economic impact, leading to the Consultative Group on International Agricultural Research Science Council (2008) calling for governments to scale back their support for and promotion of biofuels until better technologies are available. Another example of a technology-based policy gone wrong is a large home insulation scheme in Australia designed to promote energy efficiency which had to be abandoned amidst growing concerns relating to safety and fraud. 22. To date, most discussion of fiscal policy for climate change mitigation has been in the context of developed economies. A number of reviews (such as the Stern Review, 2006, the Garnaut Review, 2008, or Burniaux, Chateau, Duval and Jamet, 2008) provide a good idea of the policies needed for developed economies. Carbon pricing is universally recommended as essential, and then a range of other complementary, technology-based policies is also recommended. There is (not surprisingly, given the mixed experience summarized above) less agreement on this range of complementary policies, but typically they include policies to support research and development, and possibly to address other market failures which might otherwise hinder mitigation action. 23. In other areas of economic policy, structural differences between developed and developing economies are given appropriate recognition, and provide the basis for differences in policy recommendations. A similar approach is needed for climate policy. The reviews mentioned above are essentially written for developed economies, and make little mention of the special characteristics and policy objectives of many developing economies. The argument of this report is that, although no generalizations are foolproof, there are important characteristics of many developing economies, in particular in the energy sector, which need to be taken into account when making policy choices in the area of mitigation. 3. The context: energy sector and other important economic characteristics relevant to instrument choice. 24. The theories which support the introduction of fiscal policies to mitigate climate change typically assume that energy prices already reflect market costs, and that the energy sector, comprised of profit- ix maximizing companies, runs along commercial lines. Research undertaken for this report in relation to the three case-study economies of China, Vietnam and Indonesia, as well as through broader cross- country studies shows that in many developing economies the situation is quite different. 25. The results of the research in this report are described in detail in Chapter 3, and summarized briefly in Table 2 below, which compares China, Vietnam and Indonesia with a typical developed economy. (Developed economies show little variation in regard to these characteristics, so in most areas generalizations can be safely made, and the table notes where this is not the case.) A rating scale of low`, moderate` and high` is used, and characteristics are worded so that the typical developed economy scores a low` for most of the characteristics investigated. The report has a particular focus on the energy sector, and in particular electricity, since this is the largest and most rapidly growing source of CO2 emissions. However, Chapter 3 also looks at some broader economic features of developing economies. Table 2 analyzes 15 characteristics in all ­ 12 in relation to the energy sector, and another 3 more broadly. The characteristics selected for analysis are all those which have to be kept in mind when advising on instrument choice in developing economies. 26. The simple message from Table 2 is that developing economies do tend to differ from developed ones in some important ways with respect to their energy sector and more broadly. Developing economies tend to experience high energy, and especially electricity, growth. Despite this, they still have up to half their population dependent on traditional fuels for cooking and heating. The benefits from access to the modern electricity sector are regressively distributed (not progressively as in developed economies). Energy subsidies are prevalent in developing economies. Just as important, but less frequently commented on, price setting is still political rather than market-oriented, and mechanisms for cost pass-through in the energy sector are underdeveloped. In many developing economies, the energy sector is neither liberalized nor subject to independent regulation. Due to a combination of rapid growth in demand, and financial constraints associated with subsidies, energy shortages are common, and in some economies the private sector responds to this through the construction of captive power. There is often little flexibility in electricity dispatch, not only because of shortages (so that everything that can be will be dispatched) but also because of transmission constraints and policy settings. The energy sector in developing economies tends to be dominated by vertically-integrated state-owned utilities. Investment expansion, especially in the electricity sector, is planned centrally. The relationship between the government and the dominant utility or utilities lacks a commercial orientation. Many developing economies have tried to reform their energy sector, but have found reforms to be difficult, especially in electricity. 27. Turning to relevant characteristics outside of the energy sector, some developing economies suffer from factor market distortions, such as financial repression. This perhaps makes their growth more capital and energy intensive than it would otherwise be. They have limited instruments to compensate households for energy price increases, and they have weaker institutional capacity than developed economies. x Table 2: Characteristics of the energy sector and the broader economy of developing economies (China, Vietnam, and Indonesia) and a typical developed economy. Characteristics relating to the energy sector (with emphasis on the power sector) Typical Developing economies developed economy China Vietnam Indonesia 1. Rate of energy growth Low High High High 2. Importance of traditional Low Moderate Moderate Low energy sector 3. Likelihood that energy Low High High High expenditure share rises with consumption 4. Presence of subsidies Low Moderate Moderate High 5. Degree of political Low Moderate Moderate High discretion in price setting 6. Degree of rationing Low Moderate High High 7. Reliance on captive power Low Low Moderate High 8. Constraints on flexibility in Low High Moderate Moderate dispatch 9. Dominance by state-owned Low with High High High vertically-integrated utilities some exceptions 10. Reliance on central Low with High High High planning in the electricity some sector exceptions 11. Divergence from Low Moderate Moderate High commercial orientation 12. Political difficulty of Mixed Moderate N/A (Just High reform starting) More general characteristics 1. Distortions in factor Low High High Moderate markets (as indicated by degree of financial repression) 2. Degree of difficulty to find Low Moderate Moderate Moderate instruments to compensate low-income households for price changes 3. Institutional weaknesses Low Moderate Moderate High relating to quality of regulation, levels of government effectiveness, and absence of corruption xi 4. Choices: mitigation policies for developing economies 28. Mitigation policy choices for APECs developing economies have to take into account several considerations. These include: policy objectives (Chapter 1); fiscal policy instruments available and the global experience with these instruments (Chapter 2); and the particular features of developing economies, especially in the energy sector (Chapter 3). Based on all the analysis summarized in the preceding sections, this final section on mitigation provides some preliminary policy guidance and points to some new directions for research and analysis. It first considers carbon pricing (its desirability, its feasibility, and its importance relative to energy sector and broader reforms), and then technology-based policies. 29. The first conclusion is that carbon pricing is desirable for developing economies with emissions reduction targets, but should be accompanied by measures to reduce reliance on traditional energy sources, and to prevent substitution from coal to oil. Carbon pricing, if effective, would help achieve emissions reduction targets, and would raise revenue in a progressive, low-cost manner, and help improve the terms of trade for energy importers. Increasing electricity prices could have adverse consequences for removing households from the traditional energy sector. These can be managed, however, for example through electricity connection subsidies. In any case, fuel subsidies have been shown to be ineffective for reducing household usage of traditional fuels for cooking. Direct measures to wean households off biomass, or at least to make biomass reliance safer, would seem to be a better route to tackle this problem than an avoidance of carbon pricing. Carbon pricing on its own could also perhaps lead to substitution from coal to oil, which would be undesirable from an energy security perspective. To prevent this, higher taxes should be imposed on oil and/or incentives given to encourage substitution instead from coal to nuclear and/or renewable, for example, through clean energy targets. 30. Carbon pricing might be desirable (with the above qualifications), but would it work? Feasibility needs to be assessed from both political and economic perspectives. Increasing energy prices is politically difficult in developing economies. The fact that modern energy is a luxury good in developing economies might improve the welfare consequences of energy price increases, but it might also raise the political costs. The limited availability of compensation instruments also makes the politics more difficult, although a few economies have shown that, even with limited instruments, compensation packages can be put together to help poor households adjust to higher energy prices, and ease the political pain. Another problem from a political perspective is the failure of most developed economies to introduce carbon prices. This means that any carbon prices, if introduced in developing economies, are likely to be on the low side, as current discussions confirm. That said, it is important to take a longer- term perspective. Climate change will not go away as an issue. Over time, as climate change becomes more evident, more economies will act and this particular constraint will weaken. 31. From an economic perspective, some features of the energy sector of a developing economy into which a carbon price is to be introduced may limit its impact, at least in those important areas which are regulated (typically, electricity and liquid fossil fuels but not, say, industrial consumption of coal). For example, there is often limited flexibility in the electricity sector to change the fuel mix, due to both power shortages and other dispatch constraints. Moreover, given the dominance of central planning in the electricity sector in developing economies an explicit carbon price may have little impact on the decisions of planners who are already trying to diversify away from coal. 32. The main risk however, given the political control exercised over energy prices in most developing economies, is that carbon prices may not be passed through. The recent spike in energy prices has revealed the limits on the extent to which many developing economies are willing to pass through increases in energy costs to the consumer. The report`s analysis suggests that in China for example, electricity prices fell in real terms in recent years even as the market price of coal doubled and more. If governments are likewise unwilling (or in some economies simply unable ­ in Indonesia, the xii Parliament has to approve tariff increases) to allow carbon prices to be passed through to final users, this will clearly limit their ability to signal to consumers to make more efficient (or simply less) use of energy. More importantly, the prospect of limited cost pass through will limit the credibility of any introduced carbon prices and thus their ability to influence investment decisions. If utilities are not confident of being compensated in the market for making cleaner energy investments, experience suggests that they will limit such investments, even if they are mandated by a central planner. (Note that promises of compensation through the budget, that is, by the government effectively paying the carbon price, will not typically be credible.) 33. This argument should not be misinterpreted as being opposed to carbon pricing in developing economies. Despite these important limitations, the introduction of carbon pricing could, nevertheless, still be a useful measure as part of a broader mitigation program. Carbon pricing could have a powerful impact on the unregulated segments of the energy sector. And, in any case, no developing economy is actually contemplating making carbon pricing the central plank of its mitigation policy. 34. The point of this argument is rather to caution against unrealistic expectations and to emphasize the importance for mitigation in developing economies of a broad-based response. Carbon pricing could be important but will not suffice on its own. As is argued immediately below, energy sector reforms, and broader economic reforms can also be important. And technology-based instruments will also be needed. 35. With respect to energy sector reform, the need for subsidy removal has been the main focus of attention. The G20 in September 2009, agreed to eliminate fossil-fuel subsidies over the medium term. More important from the perspective of enabling carbon pricing to have an impact, but also more difficult to put in place, are policies to allow for cost pass-through ­ so that subsidies do not re-appear, and so that carbon prices can be passed on. This will require liberalization of energy markets, and the establishment of independent regulators, both formidable tasks. 36. While carbon pricing will not be effective without energy sector reforms, it is also important to recognize that the impact of such reforms on their own could be to increase rather than reduce emissions. For example, if subsidies are reduced in a context of energy shortages, it is possible, and has happened, that energy use goes up rather than down (as the financial and thus investment constraints on the sector are relaxed). 37. With regard to broader reforms, the report focuses on the case of China and the need for financial sector, land and social security reforms which between them would probably make Chinas growth not necessarily less rapid but less capital and therefore less energy intensive. Ultimately, successful mitigation will require the decarburization of the energy sector. At that stage, whether growth is energy-intensive or not will have little influence on total emissions. Until that time, in some developing economies, broader economic reforms will have an important role to play in any mitigation strategy. 38. Technology-based policies are inevitable given the policy mix prevailing in many economies. As discussed earlier, developing economies want to promote clean energy not only or even primarily to reduce emissions, but also to reduce air pollution, promote energy security and gain technological advantage. Policies which specifically target particular technologies are therefore an inevitable part of the mix. But which policies are best to achieve clean energy targets is more difficult to say. The evidence cited earlier (in Chapter 2) is both mixed and limited. Four points can be made. First, the characteristics of the energy sector in developing economies which will reduce the impact of carbon pricing might also hinder the efficacy of technology-based policies. xiii Second, there is a clear case for much more R&D funding in relation to mitigation technologies, and some more advanced developing economies will want to participate in this global technological revolution as research partners and innovators. Third, there may be a case for government-provision of finance and guarantees to address the capital market failures which seem to haunt the renewable energy sector (for the reasons discussed in Chapter 2). Fourth, the mixed experience to date points to the importance of policy rigor and simplicity to encourage the adoption of good technology-based policies and weed out bad ones in a field that, in part because of its complexity, experience has shown is susceptible to rent-seeking, populism and government failure. 39. The relative importance of these four reform fields ­ carbon pricing, technology-based policies, energy sector reforms, and broader reforms ­ is a matter for judgment, and will vary from economy to economy, and over time. Sequencing is an even more difficult issue. Which should be introduced first: a price on carbon, or the mechanisms in the energy sector to allow that price to be passed through? There is no clear and certainly no universal answer to that question. Where an economy chooses to move first will depend as much on political judgment as on economic analysis. 40. There are high uncertainties around mitigation. In developed economies, the main uncertainties around mitigation are political (will carbon pricing be introduced, and when) and cost. In developing economies, uncertainty also attends to both the implementation of reforms (for example, will energy sector reforms, even if announced, by carried through), and their impact (for example, how much will carbon pricing change the emissions intensity of an economy). There is little cross-country experience, but major risks, as outlined above. 41. Given the uncertainties, a quantity anchor for climate change fiscal policy is recommended. The uncertainty around the impact of a carbon price, whether fixed by government or determined by the market, makes a strong case for thinking of carbon prices as instruments to achieve explicitly-stated environmental outcomes. The multiple fronts on which governments may need to move for an effective mitigation response ­ carbon pricing, technology-based policies, energy sector reforms, and broader economic reforms ­ also make the case for an underlying quantity target, since it will make it easier for economies to judge progress, and adjust the policy mix accordingly. 42. Developing economies have started to put quantity targets in place, but most have some way to go. All of the economies submitting domestic commitments to the Copenhagen Accord have nominated quantity rather than price targets. It must be noted, however, that of the APEC non-OECD economies, only Russia and China have submitted targets which are verifiable. The majority of developing economies have submitted targets relative to business as usual which is, by definition, only observable if there is no mitigation. If economies want a yardstick by which to judge their mitigation progress, they should convert targets defined relative to business as usual into absolute emissions or emissions-intensity targets. xiv 5. Fiscal aspects of adaptation to climate change 43. Adaptation is the deliberate effort to obviate or ameliorate the bio-physical effects of a changing climate. Adaptation to climate change necessarily occurs in the context of ongoing competition for a limited set of a society`s resources. Choices need to be made between different adaptation projects, and between adaptation projects and other socially desirable investments such as education. Fiscal policy therefore has an essential role to play, though, as noted at the start, the study of fiscal policy in relation to adaptation is in its infancy. 44. The fiscal approach to adaptation which has received most prominence to date is its costing. A range of studies have been produced by the United Nations and the World Bank. For example, the World Bank (2010e) has estimated the global costs of adaptation to be $70-100 billion annually out to 2050 (in 2005 prices). Costings are important for the purpose of international negotiations. Though their calculation faces various conceptual and empirical challenges, they also draw attention to the importance of adaptive responses, and in some instances help with planning for them. But policy choices need not only to be costed but to be made on the basis of sound decision-making tools. 45. Costings usually revolve around the provision of public goods, such as infrastructure. This will certainly be critical to support effective adaptation, and the report highlights in particular the provision of information, agricultural R&D, and infrastructure. But public good provision is only one of three types of adaptation instruments governments need to consider. A second is public-sector pricing reform. Water pricing reform in particular will be essential for successful adaptation and there is a clear parallel to the importance for mitigation of the reform of energy pricing. A third important set of instruments are financial in nature, such as microcredit and insurance. These can be effective, and fiscally much cheaper, alternatives to the provision of subsidies to help households adapt. There are clear challenges in making such instruments accessible to rural households, but some important examples of successes. 46. The right choice of adaptation instruments will vary from location to location, which limits the utility of general guidance in this regard, and emphasizes the importance of decision-making tools. Key to the right choice of instruments will be the correct use of appropriate decision-making tools. As climate change and awareness of climate change grows, Finance Ministries can expect to be increasingly confronted with requests for funding and possibly for fiscal reforms, such as changes in taxes, subsidies, and prices. How are fiscal decisions about adaptation to be made? Four decision rules are reviewed, all of which have been used for adaptation and/or broader environmental decisions. The purpose of multi-criteria analysis is to assess the relative contribution of a selected group of impacts or attributes to the achievement of an overall objective or goal. Apart from the arbitrary and atheoretical nature of selecting attributes and assigning scores and weights, the process is fundamentally flawed. As various commentators have noted, it is equivalent to adding apples and oranges. Vulnerability, adaptive capacity and resilience indexes are increasingly calculated in relation to climate change. Such indices suffer from several of the same methodological weaknesses as multi-criteria analysis. They also provide no clear guidance for action, including the timing of any action, when considering potential adaptation measures. Cost-effectiveness analysis is used in everyday life, and is easily presented to and understood by policy makers. A measure of technical efficiency, it expresses a result in terms of the cost of achieving it: for example, the number of lives saved for the cost of each kilometer of a 5 meter dike constructed. At its most simple, it can reveal projects that generate the biggest bang for the buck.` However, this is also its limitation. Comparisons can only be made between projects of a very similar nature. xv Cost-benefit analysis (CBA) has a number of well-known drawbacks. It generally requires monetization of both costs and benefits, assumes that the marginal utility of money is equal for everyone (unless distributional weights are used), and is expensive to conduct. Nevertheless, CBA remains the only rigorous analytical tool available in terms of assessing issues such as the relative merits of different adaptation projects and strategies. In particular, it affords policy makers an unambiguous decision tool in requiring that the present value of benefits to society as a whole exceed the present value of costs incurred. It alone permits comparison of adaptation measures not only with each other, but also with alternatives that are not necessarily associated with climate change effects. 47. A key advantage of the cost-benefit approach is its capacity to incorporate the inevitable uncertainty around climate change impacts. Uncertainty is in many ways the hallmark of climate change. Knowledge is lacking ­ particularly at the local level ­ as to both the intensity and the timing of future climate change. Much work on adaptation has been based on mean values, but adaptation, by its very nature, requires consideration of extremes. The intensity of future weather events might remain much as it is today, or change along a spectrum that includes both the benign and the catastrophic. The timing of any additional intensity is equally uncertain. 48. The report canvases two methods for incorporating uncertainty within a CBA framework: Monte Carlo analysis and the ,,real options approach. 49. Most countries have sufficient historical data to generate models based on extreme value distributions. Probability distributions which reflect historical experience can be augmented` or shifted to take into account climate change over the coming century on the basis of predictions obtained from physical climate models. Further, to reflect the uncertainties involve, a series of distributions can be generated for, say, each year in the future period under examination. By then applying a damage function to each of the augmented` probability distributions, a set of cost functions can be generated for each year. Applying Monte Carlo analysis by generating random numbers to choose flood heights (from among the different distributions), a probability distribution of costs can then be generated for each year, giving a range of possible costs, rather than single point estimates 50. Options are typically thought of in relation to financial assets but they can also be identified, or developed, for physical (real) assets. The real options` approach is useful not only for assisting decision-makers to cope with uncertainty within a cost-benefit framework, but also to encourage planners to incorporate options into the design of adaptation projects. The flexibility incorporated into real options, due to the opportunity to gain better knowledge of climatic changes by delaying full investment, can expand the number of potentially viable projects. The flexibility of real options is particularly advantageous for seriously resource-constrained governments. It enables expenditure on climate-relevant projects can be spread out, to be made when required over time. A greater number and range of climate- relevant projects can thereby be funded. This aspect makes real options analogous to a just-in-time` technology. 51. Panels A and B of Figure 4 below provide an overview of the real options approach, using the building of dikes as an illustration. In this case, the decision is whether and when to construct a dike intended to prevent future inundation resulting from more intense or more frequent floods due to climate change. xvi Figure 4 Two different approaches to building a dike to respond to increased flooding risk 52. While it is possible to build a dike immediately (along the lines of the popular ,,precautionary principle), an alternative approach is probably more appropriate given the nature of climate change impacts Panel A illustrates the case in which the full cost of construction is incurred immediately, with regular maintenance costs thereafter, but the benefits are realized only at some uncertain time in the future. The alternative shown in panel B is to expend only a relatively small amount today on surveying or preparing the land on which a future dike could be built. This amount ­ smaller than the full cost of a dike ­ is analogous to the price paid to acquire a financial option, because it establishes the opportunity, but not an obligation, to invest in the full asset when required. A full investment can be made when the benefits of countering the effects of climate change increase the value of the asset above its costs. The net present value of the panel B approach is higher than that in panel A, irrespective of the discount rate used. 53. The key difference between the two approaches is that panel B incorporates flexibility. As time goes by, it may be found that climate change is occurring less quickly than originally anticipated, perhaps due to reduced global emissions. By waiting until better information becomes available, a dike can be built with a height more closely matched to the actual climate of the future. The greater the uncertainty, the more valuable is the flexibility associated with being able to wait for better information. 54. The ,,real options framework can be incorporated into a social cost-benefit framework, but it will require innovation Uncertainty about climate change precludes specification of actual dates when benefits from a dike will begin to flow. The application of Monte Carlo analysis to the results of climate modeling can provide some guidance to the analyst in the form of a predicted optimal construction date, which can be updated with new information over time, as long as the analyst accepts that the future predicted by a suite of climate models and/or projections is approximate. While the real options approach thus has clear strengths, it also demands a good degree of creative thought, and a willingness to move away from seemingly risk-averse deterministic solutions, even though such projects, executed without delay, are better suited to established budgetary precepts and are more closely aligned with current project management practices. 55. The tools presented and advocated in the report ­ cost-benefit analysis, Monte Carlo analysis, and real options ­ will only be adopted by policy-makers if they are confident that these are practical methods able to guide decision-making in real-world circumstances. A large part of Chapter 5 is dedicated to providing examples from a range of economies ­ from dikes and houses in Vietnam to coastal inundation in Australia ­ where these techniques have been used, and found useful. xvii Chapter 1 Goals and targets: climate change mitigation and related policy objectives 1.1. This chapter provides an initial profile of APECs economies in relation to emissions (Section 1.1) and then outlines the ambitious targets which they have recently adopted to reduce emissions and promote clean energy (Section 1.2). These targets serve a variety of important economic, security and environmental goals (Section 1.3). Policy instrument choice, the main theme of this report, will be the difference between success and failure in relation to their achievement (Section 1.4). 1.1 Introduction 1.2. APEC economies show enormous diversity in terms of per capita income. While any precise cut- off will be arbitrary, the economies of APEC can easily be divided into developed (advanced or rich) and developing (middle- and low-income).1 As Figure 1.1 shows, New Zealand, the poorest of the rich economies, has a per capita income almost twice that of Russia, the richest of the poor economies. Figure 1.1: Developed and developing economies in APEC GDP per capita for APEC economies in US$ (PPP) Note: 2008 GDP per capita income expressed in current US$ using purchasing power parities for the conversion. Sources: World Bank (2010a); Brunei and Taiwan from IMF (2010a) 1.3. APEC economies as a group are responsible for almost two-thirds of annual global CO2 emissions from fossil fuels (PBL, 2010). In APEC, CO2 emissions and GDP per capita are closely correlated (Figure 1.2). Richer economies have much higher per capita and also cumulative emissions. 1 Under the UN Framework Convention on Climate Change, developed countries are those which belong to the OECD in 1992, the year in which the UNFCCC was signed. Our analysis is instead in economic terms. The World Bank uses $11,906 as the cut off for high-income using market exchange rates. The economies we classify as developed are those the World Bank classifies as high-income. The Bank does not classify Taiwan but if it did it would be as a high-income economy. See http://data.worldbank.org/about/country-classifications/country-and-lending-groups#High_income. 1 Figure 1.2: Per capita emissions and income are closely correlated across APEC Sources: IEA (2009a) for CO2 (reference approach data used), World Bank (2010a); IMF (2010a). 1.4. The US is no longer the worlds largest emitter, but it is responsible for the greatest volume of accumulated emissions. Without leadership from the United States and other developed economies, it will not be possible for the world to mount an effective mitigation response to climate change. 1.5. China is also a country of great significance. China`s rise to industrial superpower status (illustrated by comparative steel production volumes in Figure 1.3 below) has accelerated global emissions growth. China is now the largest emitter of CO2 (from fossil fuels), with 25% of the global total in 2009, considerably ahead of the second largest annual emitter, the US with 17% (PBL, 2010). China has been responsible for 72% of the world`s growth in CO2 emissions (from fossil fuels) between 2000 and 2009, a period during which China`s emissions grew at an annual average rate of 9.4%, and the rest of the world`s at 0.8% (PBL, 2010). Consequently, China`s leadership will also be critical. 2 Figure 1.3: China: rise of a steel giant Production of steel in China, Europe, Japan and the United States, 1994-2010 Notes: Steel production in thousands of tonnes per month, seasonally adjusted. Source: OECD Main Economic Indicators, World Steel Institute; compiled by Outlook Economics (2010a). 1.6. Developing country emissions more generally have been growing rapidly and, absent the introduction of policies to mitigate climate change, will continue to do so. Emissions growth is a function of GDP growth, and changes in energy intensity (the ratio of energy consumption to GDP) and the carbon intensity of energy. Total CO2 emissions have been growing much faster in developing economies than in advanced ones. So too has output, which emissions have tracked in most developing economies, but lagged in advanced ones. To understand this, one needs to consider China and other developing economies separately, as Figure 1.4 does. The carbon intensity of energy among APEC`s developing economies (with and without China) has been on a slight upward trend, reflecting the move away from biomass and increasing dominance of coal-fired generation. Energy intensity has been remarkably flat among APEC`s developing economies other than China (see Sheehan, 2008 for this result more generally among developing economies). 1.7. China shows different trends before and after 2000. Emissions in China grew more slowly than GDP pre-2000, but have since roughly tracked GDP. China`s energy intensity fell over the 1980s and 1990s, as a result of market liberalization, the relative decline of heavy industry and perhaps power shortages, but has flattened out since 2000. Figure 1.5 suggests that whereas China entered its reform era with an extraordinarily high energy intensity, it has now become a normal` developing country in terms of both the level and trend of its energy intensity.2 2 Though note that more recent PPPs give a lower GDP for China and thus would give it a higher energy intensity. 3 Figure 1.4: In most developing economies, CO2 emissions track GDP CO2, GDP, emissions intensity of energy, and energy intensity of GDP for three groups of APEC economies, 1971-2007 (2000=1): (a) mature developed economies, (b) developing and newly developed economies excluding China and Russia, and (iii) China, Notes: Mature developed economies are Australia, Canada, Japan, NZ and US. Developing and newly developed are all others apart from China and Russia. Data for PNG is missing. GDP is measured in billions of constant year (2000) US$, using purchasing power parities (PPPs) to convert from local currency. Energy is measured in Mtoe (million tonnes of oil equivalent), and CO2 is measured in millions of tonnes. Source: IEA (2009a). Figure 1.5: Energy intensities are converging across APEC The ratio of energy consumption to GDP for selected APEC economies, 1971-2007 Notes: See notes to Figure 1.4. Source: IEA (2009a). 1.8. Emissions volumes and growth are also driven by the regions high and increasing reliance on coal, the most CO2-intensive fuel. China has the fifth highest carbon intensity of energy of the world`s large economies (with populations above 20 million). Vietnam`s baseline projections for the electricity sector include a rise in coal-fired generation from 21% of total generation today to 51% in 2020 (World Bank, 2010, p.36). Indonesia`s Government`s baseline projections are for annual growth in CO 2 emissions of 5.9% up to 2020, slightly above projected GDP growth of 5.7%, with the share of emissions from coal rising from 25% in 2003 to 51% in 2050 (Ministry of Finance, Republic of Indonesia, 2009, 4 Table 2.3, p.39).3 More generally, Burke (2010) demonstrates that the carbon intensity of energy of a developing economy increases as it develops before it starts to decline. 1.2 Climate change mitigation and clean energy targets among APEC economies 1.9. Fortunately, most developed and many developing economies have decided that ,,business as usual is no longer an option, and have adopted emissions targets. Recognizing the urgency of a global response to climate change, many developing economies have recently made non-internationally-binding commitments through the Copenhagen Accord to constrain the growth of their own emissions. Table 1.1 summarizes the national targets of APEC economies. Some, but not all, of the targets of developing economies are conditional on financial assistance. 4 1.10. Most of the targets adopted are ambitious ones. There is little difference between the ambition of APEC developed and developing economy members; if anything, the targets of the latter are actually more ambitious. Figure 1.6 provides one set of estimates of the reduction in emissions implied by the targets relative to business as usual. Figure 1.6 : Many APEC Economies have set ambitious emissions reduction targets 2020 APEC emission control targets expressed as a reduction relative to business as usual (BAU) Notes: National targets as recorded in the Copenhagen Accord or other statements. Where a target range has been committed to, the mid-point of that range is selected. Source: Table 1.1 and Figure 1.7 3 Note that this excludes emissions from deforestation. 4 While some of the developing economy targets are conditional on developed country support, others are not, and even some of those that are conditional promise a minimum of unconditional action (e.g., Indonesia). Table 1.1 provides the detail. 5 Table 1.1: Climate change mitigation and renewable energy targets for APEC economies Climate change mitigation target (for 2020, unless specified to the contrary) Clean energy target Summary Details Target Australia 5% to 25% below Australia will unconditionally reduce emissions by 5% below 2000 levels by 2020, and by up to 15% or 20% by 2020 up from 8% in 2007. 2000 emissions 25% conditional on a global agreement and action by others Canada 17% below 2005 17%, to be aligned with the final economy-wide emissions target of the United States in enacted legislation. 90% (non-emitting energy sources) by 2020. emissions Chile 5% by 2014, 10% by 2024. China 40% to 45% below China will endeavor to lower its carbon dioxide emissions per unit of GDP by 40-45% by 2020 compared to 15% by 2020 up from 8% in 2006. 2005 in emissions the 2005 level, increase the share of non-fossil fuels in primary energy consumption to around 15% by 2020 intensity and increase forest coverage by 40 million hectares and forest stock volume by 1.3 billion cubic meters by 2020 from the 2005 levels. Indonesia 26% to 41% below 26% reduction unilaterally, up to 41% reduction with international assistance 15% by 2025 (inc. nuclear). BAU Japan 25% below 1990 25% reduction, which is premised on the establishment of a fair and effective international framework in 16.0 TWh by 2014. emissions which all major economies participate and on agreement by those economies on ambitious targets. Korea 30% below BAU To reduce national greenhouse gas emissions by 30% from business as usual emissions at 2020. 6.08% by 2020 up from 2.7% in 2009. Malaysia Target to be introduced from 2011. Mexico 30% below BAU Mexico aims at reducing its GHG emissions up to 30% with respect to the business as usual scenario by 40% by 2014. 2020, provided the provision of adequate financial and technological support from developed countries as part of a global agreement. New Zealand 10% to 20% below 20% target subject to comprehensive global agreement. 90% by 2025 up from 73% in 2009. 1990 emissions Russia 15% to 25% below The range of the GHG emission reductions will depend on the following conditions: Appropriate accounting 4.5% by 2020. 1990 emissions of the potential of Russia`s forestry in frame of contribution in meeting the obligations of the anthropogenic emissions reduction; - Undertaking by all major emitters the legally binding obligations to reduce anthropogenic GHG emissions. Singapore 16% below BAU 16% from projected 2020 business as usual levels Taiwan 2000 level by 2025 12% by 2020 up from 6% in 2009. Thailand 30% below BAU 30% reduction by 2020 in energy sector 20% by 2022. Interim target of 15.6% by 2011. The 100% increase from 2005 to 2015. Philippines United States 17% below 2005 In the range of 17%, in conformity with anticipated U.S. energy and climate legislation, recognizing that the No national target. About 30 states have mandatory targets. emissions final target will be reported to the Secretariat in light of enacted legislation. Vietnam 5% by 2020 up from 3% in 2010. Notes: BAU` is business as usual. Climate change targets are ones submitted in relation to the Copenhagen Accord, except for Taiwan. Source: Summaries of targets taken from various national documents, from submissions to the UNFCCC, and from Ölz and Beerepoot (2010). 6 1.11. If achieved, the targets would fundamentally alter the emissions trajectories of APEC economies. Figure 1.7 shows historical and projected fossil-fuel-CO2 emissions and GDP per capita (with the latter plotted in logs). Two sets of projections are shown: those under business as usual (as estimated by Garnaut et al., 2009) and those assuming national emissions reduction targets are achieved. The table shows that APEC`s poorer developing economies are entering into a period of rapid emissions growth similar to what Taiwan and Korea experienced over the last 40 years. Under business as usual, China by 2020 will have not only Korea`s per capita income of today but also Korea`s current per capita emissions, which at 10 tonnes of CO2 per person are more than double China`s current level. However, things will be very different if the APEC economies are able to adhere to their national targets. There will be significant convergence between developed and developing economies, developed economies will have shown that absolute emissions reductions are possible, and developing economies will have avoided a massive ramping up of emissions: China`s per capita emissions at the end of the decade will be 7 rather than 10 tonnes, a savings of about 4 billion tonnes of CO2 (more than 10% of current annual global emissions of CO2 from fossil fuels). Figure 1.7: Achieving national targets will mean a very different emissions future for APEC Per capita CO2 emissions, historical (1971 to 2007) and projected (to 2020, both under business as usual, and assuming targets are met) for various APEC economies Notes: Solid lines are historical. Upper dotted lines show business as usual. Lower dotted lines show trajectories assuming national commitments (from the Copenhagen accord, except for Taiwan) are adhered to. The historical period covered is from 1971 to 2007; the projection period is from 2007 to 2020. Emissions per capita are assumed to grow/decline in a linear manner over the projection period. These graphs are based on the simplifying assumption for most economies that targets announced for all greenhouse gases will be adhered to for CO2 from fossil fuels as well (the graph shows the latter only). Where a target range is provided, the mid-point of that range is selected. Source: Historical data from IEA (2009a); Garnaut et al. (2009), Table 1.1, and own calculations for projections. 7 1.12. An even larger number of APEC economies have embraced renewable or clean energy targets some of which are also ambitious. Worldwide, renewable energy targets are more popular than mitigation targets, with the former having been established by at least 73 economies globally as of 2009.5 Fourteen APEC economies now have renewable energy targets (Table 1.1), defined either in absolute terms or as a percentage of capacity (easier to monitor and therefore harder` than mitigation targets defined relative to an unobservable business as usual). Some of these clean energy targets are also ambitious. For example, Ölz and Beerepoot`s (2010) review of renewable energy targets in ASEAN singles out Thailand and Indonesia for their level of ambition. 1.13. While the focus of this report is on climate change mitigation, it is important to understand how mitigation objectives and targets sit within the broader policy mix. The next section argues that clean energy targets have been adopted not only, or even primarily, as a means to achieve climate change targets, but rather as important policy targets in their own right. 1.3 Goals behind the climate change and clean energy targets 1.14. Why have so many APEC economies adopted both climate change and clean energy targets? It is important to know the reasons in order to understand how these targets relate to each other, and to guide the choice of instruments selected to achieve the targets.6 To start with, APEC economies are increasingly worried about the impact of climate change. But three other goals are also driving action. 1.15. First, many developing economies are seeking to tackle national environmental problems. China has 13 of the world`s 20 most polluted cities, 30% of its land is damaged by acid rain as a result of sulphur emitted from coal, and health damages from air pollution are expected to reach 13% of GDP in 2020 (see, for example, Zissis and Bajoria, 2008). Millions in Asia face serious indoor as well as outdoor air pollution problems. A meta-analysis of epidemiological studies concluded that indoor air pollution from solid fuel use in China is responsible for approximately 420,000 premature deaths annually, more than the approximately 300,000 attributed to urban outdoor air pollution in the country. (Zhang and Smith, 2007, abstract). For Asia as a whole, UNDP estimates that 1.3 million people died in 2004 as a result of indoor air pollution from solid fuel use (biomass plus coal) (UNDP and WHO, 2009, p.23).7 Many APEC economies have programs and targets directed at reducing local air pollution. 1.16. Second, energy security is a growing concern, as dependency on energy imports rise. APEC comprises a mix of energy exporters and importers. Australia, Canada and Russia will be long-term energy exporters, but other major APEC economies are either importers of energy already or are moving towards being net importers (Figure 1.8). Vietnam is an exception, but is likely to follow China, Mexico, Malaysia and Indonesia and move towards becoming a net importer as its domestic oil production declines and its economy continues to grow. 8 (See the Appendix Tables for more data on the energy production and consumption mix and import dependency for different APEC economies.) 5 http://www.renewableenergyworld.com/rea/news/print/article/2009/09/renewables-global-status-report-2009-update 6 The aim is to provide the reasons for the introduction of mitigation and clean energy targets. The aim is not to provide a complete listing of all goals related to the energy sector. This is why fiscal goals, for example, do not get a mention. They are of course important more broadly in relation to the energy sector (for example, an important driver of subsidy removal efforts) but they do not seem to be behind the adoption of mitigation or clean energy targets. For a discussion of the revenue implications of carbon pricing, see Section 2.2. 7 For East Asia, the estimate is 665,000. 8 One important factor that needs to be considered in formulating projections for future energy dependency is the tendency for real exchange rates to rise as economies develop. This is due to the Balassa-Samuelson effect. The Chinese real effective exchange rate in the CHN_TRYM model (see Box 1.1) for example needs to appreciate by around 5 per cent per annum in equilibrium to maintain the steady state exchange rate. A rising exchange rate means that over time high cost domestic sources of coal, gas and oil will become less competitive with imported sources. As low cost domestic sources are run down and as internal demands for energy rise there will be a tendency for import dependency on fossil fuels to rise. 8 Figure 1.8: Most APEC developing economies are already or will become energy importers Energy imports over time as a share of total energy use for various APEC economies Source: World Bank (2010a) 1.17. Some APEC developing economies are reliant on coal, and Indonesia exports coal, but none except Russia is rich in coal resources relative to production. Figure 1.9 shows the proportion of the world`s coal reserves held by various APEC economies. Only the US, Russia and Australia have shares in excess of their population. At current production rates, Indonesia`s currently proven coal reserves will only last another 19 years, and China`s another 41. 9 Figure 1.9: Apart from Russia, Australia and the US, APEC economies have little coal and/or it won't last long 30% 500 450 25% Share of w orld's total coal reserves (%, left-hand-side axis) 400 Reserves to production ratio (years, right-hand-side axis) 350 20% 300 15% 250 200 10% 150 100 5% 50 0% 0 US Russia China Australia Canada Indonesia Thailand Mexico Share of world's total coal reserves and reserves to production ratios for selected APEC economies, 2008 Notes: Reserves are proven reserves at the end of 2008. They include both anthracite and bituminous coal, and sub-bituminous and lignite coal. Proven reserves are Generally taken to be those quantities that geological and engineering information indicates with reasonable certainty can be recovered in the future from known deposits under existing economic and operating conditions. (BP, 2010). South Korea, Vietnam and Japan all have less than 0.1% of the world`s total coal reserves. The reserves to production ratio is the number of years proved reserves (at the end of 2008) would last if production at 2008 rates continue. Sources: See Appendix Tables A.1-A.3. 1.18. While this implies that APEC economies will increasingly become coal importers, coal does not give rise to the same security concerns as oil. China has long been reliant on oil imports, but has just recently become a significant coal importer (Figure 1.10). However, oil imports will remain much more important to China than coal imports. China already imports almost 60% of its oil needs, but in 2010 will import less than 10% of its coal. Worldwide, there is a lot of coal left. Garnaut (2008, Table 3.3) reports that, at 2007 production rates, the world has 139 years of coal left in its reserve base, as against only 60 years of gas and 40 years of oil. Also, unlike oil, much of that coal is in secure locations, such as Australia. 10 Figure 1.10: China, long and increasingly dependent on oil imports, is now also a net coal importer Oil and coal imports and exports, and oil consumption and production, China, 1990-2010. Notes: Series are seasonally adjusted. Updated trade data taken directly from the customs web site and cross checked against data from the World Trade Atlas. June quarter is a forecast based on data up to May. Data source: CEIC Database, Chinese National Bureau of Statistics, Chinese customs, OECD Main Economic Indicators Database, World Trade Atlas, Outlook Economics CHN-TRYM database; compiled by Outlook Economics (2010a). 1.19. Energy insecurity is a function not just of quantities but also of prices. The volatility and overall upward trend in world energy prices over the last decade (Figure 1.11) have also heightened energy security concerns. As many APEC economies increasingly become energy importers, these higher fossil fuel prices will impose an increasing economic cost. 11 Figure 1.11: World energy prices have been volatile and rising over the last decade World prices for crude oil, LNG, and coal, adjusted for inflation (1990=1) Notes: Data is quarterly. Series are converted from US$ to more currency neutral SDRs, divided by the G7 CPI to convert them to constant price terms and then indexed to 1990s levels (average of the decade) for the purpose of comparison. Oil prices are the IMF indicator series (which is an average across a number of types of crude). LNG prices are the price of Russian LNG in Europe. Coal prices are based on Australian thermal coal export prices. Source: IMF International Financial Statistics, OECD Main Economic Indicators; compiled by Outlook Economics (2010a) 1.20. Third, economies are also seeking technological advantage. Economies around the world increasingly view clean energy and more broadly low-carbon technology as the next source of innovation, the next big thing`, and they want a share of the action. This often influences policy choices. For example, even though it doesn`t have good conditions for wind generation, Korea is promoting wind power, especially off-shore wind, to give its companies experience and help them become global players. By one estimate, the wind energy market will be worth $90 billion by 2015, as big as the global ship- building industry is today. Korean companies and government figure that, with their engineering prowess and reputation, they should be able to dominate the wind generation sector tomorrow, just as they do ship- building today (Kang, 2010). Many economies have similar objectives. Denmark started promoting wind power in the mid-1970s. 30 years later, Denmark had about 3,000 megawatts (MW) of installed wind power capacity, and Danish firms had installed about 20,000 MW globally, about 40 percent of the world total (Brandt and Svendsen, 2006). 1.21. There is thus in many economies a mix of objectives driving policy action, not just one. The Government of Korea has gone further than most through its Green Growth strategy in making its multiple objectives explicit. As Figure 1.12 shows, climate change mitigation and adaptation feature prominently in the Korean strategy but within a broader policy framework. The Korean framework in fact incorporates all four objectives discussed in this section: mitigating climate change, enhancing energy security, creating new growth engines, and reducing pollution. 12 Figure 1.12: Korea's Green Growth Strategy Source: Government of Korea (2009) 1.22. There are synergies between these goals. By pushing down global energy prices, global action on climate change would improve the terms of trade for most APEC economies. Modeling undertaken for this report illustrates this for China (Box 1.1). Box 1.1: Taxing fossil fuels and pricing carbon improves the terms of trade for energy importers IMF staff using the Global Economic Model (GEM) have estimated that a 25 percentage point increase in global oil taxes, offset by reductions in labour taxes, would reduce pre-tax oil prices by 17% in the first year, and lead to a 0.5% increase in aggregate household consumption both in the United States and in emerging Asia after around 5 years (Elkedag et al. 2008): see Figure 2.4. Consumption and GDP are both higher in oil importing economies as a result of a policy that both reduces emissions and import dependency and hence the susceptibility to oil price shocks. This report builds on this work by examining the terms of trade effect for China of the introduction of a carbon tax. Unilateral imposition by China will have little impact on its terms of trade, but China would be the beneficiary of global mitigation action. The CHN-TRYM model and a simple global demand-supply framework for global fossil fuels, calibrated to the IMF results above, are used to simulate the impact on China of a coordinated global 9 introduction of a $20 a tonne CO2 tax on carbon in 2010 rising to $100 by 2030. Tax revenues are recycled to offset impacts on real household disposable income and business investment. The imposition of the carbon tax leads to an initial rise of $7.50 per barrel in the after-tax price of oil or 9% on the price projected in the baseline, and a $46 per tonne (or 41%) increase in the after-tax price of coal. The tax margins rise to $24 per barrel (22 %) and $149 per tonne (150%) for oil and coal respectively by 2020.10 The increased after-tax prices lead to a substitution away from coal, oil and gas and a fall in pre-tax prices. Coal 9 The CHN-TRYM model is a quarterly time-series econometric model of the Chinese economy in the style of the Australian Treasury Macroeconomic (TRYM) model. It incorporates industry detail using input output linkages, is estimated where possible on the basis of historical time series data, and is designed for forecasting and policy analysis. For more information on the model contact :info@outlookeconomics.com 10 Prices are in real terms at constant 2007 dollars. The percentage increases are baseline dependant. For a given tax on carbon the lower the baseline price the higher the tax margin in percentage terms. 13 consumption and prices fall by 40% and 20% relative to baseline by 2025, while oil consumption and prices fall by 15% and 9% respectively. China`s terms of trade increase by around 1% by 2025, giving a boost to national income and consumption of about 0.3%. The exchange rate appreciates slightly relative to baseline and the current account balance is relatively unchanged. As a cross check, these results can be compared with those of Downes (2007) using the Global Trade Analysis Project (GTAP) model. A $9 dollar a barrel reduction in the price of oil and a 18 per cent fall in the price of coal were simulated. (By comparison, the pre-tax oil price falls by $9 per barrel and coal prices by 29% in the global carbon tax simulation above.) The impacts on China`s terms of trade (0.9% improvement) and private consumption (0.4% improvement) are similar to those reported in the CHN-TRYM simulation above. Terms of trade for other energy importers in APEC also improve. Source: Outlook Economics (2010b) 1.23. Positive terms of trade effects point to the possibility of action on climate change being an economic benefit for energy importers. Normally, climate change mitigation is seen as imposing a short- term economic cost to achieve longer-term environmental goals. But, if mitigation is pursued in an efficient manner ­ in particular, as discussed in more detail in Chapter 2, through a carbon price, the revenue from which is used to offset other taxes ­ then it is possible that climate change mitigation will in fact bring short-term economic benefit as well as environmental gain (see Section 2.2 for further elaboration of this point in the context of an analysis of carbon pricing). 1.24. But there are also trade-offs. Energy security concerns focus mainly on oil, climate change mainly on coal. Reducing the use of coal has to be a prime target for climate change mitigation and air pollution, but not necessarily the other two goals. If an economy is concerned primarily about energy security, or promoting technological advantage, it will give more weight to promoting clean energy, and reducing oil consumption, and not worry as much about coal. 1.25. An emissions reduction target on its own might undermine energy security goals. In the scenario described above (Box 1.2), the global carbon tax reduces China`s coal, oil and gas imports, but oil and gas take a larger share of the fuel mix and coal a smaller share. This is because oil and gas are more efficient from an emissions perspective. 11 But a rising oil share in the fuel mix is sub-optimal from an energy security perspective. Carbon capture and storage is another example: it will help reduce emissions, but will also worsen local air pollution and weaken energy security, since it will significantly reduce the efficiency of coal plants. 1.26. Likewise, some measures to improve energy security can increase emissions. Coal-to-liquid conversion (currently under consideration and/or development in several Asia-Pacific economies) will reduce reliance on oil imports, but will increase emissions. 1.27. A combination of targets works well for most economies. Most leaders want to reduce reliance on both oil (to promote energy security) and coal (for climate change and local pollution reasons). Their climate change targets discourage coal, and total energy use as well. They can use clean energy targets to dampen demand for oil as well as coal and to promote their competitive advantage in what they see as a future growth area.12 11 Oil and coal are substitutes in some industrial processes, in heating, and in electricity. A supercritical coal plant will generate about 800 grams of CO2 per kWh of electricity, an oil combined cycle plant 500 grams, and a gas combined cycle 400 grams (IEG, 2009, Box 2.1). In the scenario, oil imports still fall in absolute terms, but given energy security goals they should fall in relative terms as well. Put differently, if the energy security target (of reduced oil reliance) was expressed through an instrument (such as an oil levy) then the levy could be set to prevent this substitution into oil. 12 This caricature says nothing about gas. Most countries would like to increase gas consumption, which is low polluting and doesn`t present the same energy security worries as oil. However, gas is a tiny percentage of oil use in most APEC developing countries, and growth is supply-constrained. Also note that a similar result could be obtained by simply taxing oil at a higher rate 14 1.4 From targets to instruments 1.28. The key risk with APECs climate change mitigation and renewable energy targets is not that they lack enough ambition to prevent dangerous climate change, but that they might not be achieved. It is one thing to commit to emissions reductions, another to achieve them. In 1997, developed countries as a group committed under the Kyoto Protocol to reduce greenhouse gas emissions (by about 5% relative to 1990). But fossil fuel emissions (the largest single source of greenhouse gas emissions, and the only one for which up-to-date data is available) increased by 14% by 2007 over the 1990 base. The global financial crisis reduced emissions in the subsequent two years, the first two of the 2008-12 Kyoto commitment period, but the increase since 1990 is still 7% (Table 1.2). Countries could get closer to their targets by trade in permits and offsets, permitted under Kyoto, and by reductions in other domestic sources of greenhouse gases. Overall, however, the prospects for hitting Kyoto targets are bleak. The reason for this is simply that developed countries did not do enough to put in place policies to achieve the targets they signed up to. Table 1.2: Most developed countries show growth from fossil fuel emissions despite their Kyoto commitments to reduce emissions Countries Emissions (Mt of CO2) Growth from Growth from 1990 to 2007 1990 to 2008-09 1990 2007 2008-09 (%) (%) average U.S 4.97 5.90 5.51 18.7% 10.8% Canada 0.45 0.59 0.56 30.2% 24.4% Australia 0.27 0.41 0.40 50.2% 44.7% Japan 1.16 1.33 1.25 14.8% 8.0% EU 15 3.32 3.36 3.17 1.3% -4.4% Total 10.17 11.58 10.89 13.9% 7.0% Notes: Kyoto targets cover emissions from other greenhouse gases and some CO2 emissions from land-use change. Kyoto targets can also be met through trading and the purchase of international offsets. Thus, some countries can achieve their Kyoto targets despite growth in fossil fuel emissions. The table focuses on these emissions because of the availability of the most recent data on them, which is important given the reduction in emissions as a result of the global financial crisis. Source: PBL (2010). 1.29. The difficulty of achieving emissions reduction targets is underlined by the resumption of rapid emissions growth in China. China`s reported 8.4% annual average growth in CO2 emissions (from fossil fuels) between 2005 and 2009 is well above most business as usual` or reference case` scenarios for that country (PBL, 2010). The 2008 World Economic Outlook (IEA, 2008a) has CO2 emissions in China growing at an annual average of 5.1% between 2006 and 2015 under its reference case. In projections from the World Bank (2010b), annual average emissions growth between 2010 and 2015 is 4.6%. Even Garnaut et al. (2009), who assume continued rapid growth in developing economy emissions under business as usual, project 7.1% annual average emissions growth for China between 2005 and 2015. It is remarkable that China is exceeding these projections in a period which encompasses a global downturn. It is not that China has not already made a policy effort; to the contrary, it has stringent energy efficiency and renewable energy targets in place. UNDP (2010, p.82) notes that There are few, if any, developing economies that have promulgated as many laws, policies and other measures to support low carbon development as China. (And this is probably true not only in relation to developing economies.) But, as discussed in Chapter 3, China has powerful forces pushing its economy in the opposite, capital-, energy-, and emissions-intensive direction. These forces are making it hard for it to achieve its 2010 energy than its carbon content alone would justify. A market-based approach would also price local air pollutants. The only role left for clean energy targets in such an approach would be the technological advantage one. However, as argued in Chapter 4, there need to be major reforms before price-based instruments can work effectively in most developing countries. 15 intensity reduction target (Box 1.2), and will equally make it difficult to achieve its 2020 emissions intensity target. 1.30. The instruments chosen will be the difference between success and failure. Whether or not APEC`s ambitious climate change and renewable energy targets will be met will depend on the instruments chosen to achieve them. So far, APEC economies have used a mix of regulatory and technology-specific fiscal measures to achieve their goals. There is increased interest, however, in carbon pricing. New Zealand and some American states have recently introduced emissions trading schemes. Several APEC economies, developed and developing, are considering the introduction of a carbon price (Table 2.1). 1.31. How well different fiscal instruments will reduce emissions and promote clean energy in developing economies, and which ones are more appropriate, is the main focus of this report. To date, most discussion of fiscal policy for climate change mitigation has been in a developed country context. The policies needed for developed economies has been discussed in a number of reviews (such as the Stern Review, 2006, or the Garnaut Review, 2008, or Burniaux, Chateau, Duval and Jamet, 2008) and which can be consulted. Carbon pricing is universally recommended as essential, and then a range of other complementary policies is also recommended. There is less agreement on this range of complementary policies, but typically they include policies to support new technologies and to address other market failures, which might block action. 1.32. In other areas of economic policy, structural differences between developed and developing economies are given appropriate recognition, and provide the basis for differences in policy recommendations. See, for example, the treatment of tax policy in Newbury and Stern (1987) or of regulation in Estache and Wren-Lewis (2009). This sort of analysis has not yet been undertaken in relation to climate change mitigation. The argument of this report is that, although no generalizations are foolproof, there are important characteristics of many developing economies, in particular in the energy sector, which need to be taken into account when making policy choices in the area of mitigation. 1.33. The next three chapters pursue this theme. Chapter 2 sets out the options. It lists and classifies fiscal policy instruments available for use in relation to climate change mitigation and clean energy. Chapter 3 provides the context in which the policy objectives articulated in this chapter will be pursued. It outlines characteristics of developing economies, especially in the energy sector, which are relevant to mitigation policy design and impact. Based on this analysis and on the experience with these instruments in developed economies, Chapter 4 draws out the implications for fiscal instrument choice and mitigation policy for developing economies. 16 Box 1.2: What has happened to energy intensity in China since 2005? As part of its Eleventh Five Year Plan, China announced it would target a 20% reduction in energy intensity by 2010 relative to 2005 levels. China has recently announced it has achieved a 15.6% reduction in the first four years, which puts it well on track to hit 20% by the end of this year. However, published data suggest a reduction in energy intensity of only 10% up to 2009 (updated from Howes, 2010). Measurement changes introduced as a result of the 2009 Economic Census make published 2005 and 2009 energy figures non-comparable. But the fact that total electricity production, electricity from coal, steel and cement production, and coal consumption have all risen significantly relative to GDP since 2005 makes it unlikely that energy intensity has declined by very much. As Figure 1.13 shows, only China`s oil intensity continues to decline rather than rise. Between the first half of 2005 and the first half of 2010, China`s steel production increased by 94%, cement by 86% and electricity by 75% (first quarter comparison), all above the extraordinarily rapid rate of GDP growth of 71% (an annual average growth in GDP of 11.3%). And this in a period which encompassed the global financial crisis. The global slowdown in the second half of 2008 hit China`s energy-intensive industries particularly hard, and no doubt caused a slump in energy use. However, the recovery has removed that effect. In May 2010, Chinese authorities announced that energy intensity had increased by 3.2% in the first quarter of 2010 relative to the first quarter of 2009, and also announced an iron hand crackdown, with new measures to ensure the 20% target is met. Electricity production grew by 20% in the first six months of 2010 (relative to the first half of 2009) and steel production by 26%, double or more the rate of GDP growth. The Chinese authorities more recently announced that energy intensity in the first half of 2010 was equal to the intensity from 12 months earlier. It is unclear what has happened to energy intensity since 2005. But the underlying trends in Figure 1.13 are certainly of concern. Figure 1.13: Steel, cement, electricity and electricity from coal production, as well as coal consumption, have all risen relative to GDP since 2005 1.25 1.25 Coal consumption 1.20 1.20 (includes stock- building) 1.15 Steel 1.15 Production Electricity 1.10 1.10 from Coal Index 2005=1 1.05 1.05 Total 1.00 Electricity 1.00 Cement Production 0.95 0.95 0.90 0.90 Oil consumption 0.85 0.85 Jun-05 Jun-06 Jun-07 Jun-08 Jun-09 Jun-10 Notes: Series show production series divided by GDP at constant prices and indexed to 2005=1. Series are seasonally adjusted. Coal production and import and export data are updated from the NBS and Chinese customs websites before seasonal adjustment. (June quarter 2010 data is a forecast based on data to May.) Note that coal consumption as derived implicitly includes stock-building which would probably have been positive in the last quarter of 2008 as the GFC hit. Also the NBS coal production source data has probably not been adjusted for changes in the number of coal mines in the economic census between 2005 and 2009. Data source: OECD Main Economic Indicators Database, China NBS web site, ADB database, Chinese Customs, Outlook Economics CHN- TRYM database; compiled by Outlook Economics (2010a). 17 Chapter 2 The instruments: fiscal policies for mitigation 2.1. This chapter sets out the options: the various instruments which can be used to achieve the national targets set out in Chapter 1. The approach is selective rather than comprehensive, with a focus on fiscal instruments and renewable energy. Section 2.1 classifies instruments into different types. Section 2.2 considers carbon pricing and Section 2.3 technology-based fiscal instruments. For each of these, the rationale, different instruments of this type which have and could be used, and experience with them to date are considered. 2.1 Types of instruments 2.2. Climate change mitigation instruments can be divided into carbon-pricing and technology-based policies.13 Carbon pricing policies include a carbon tax, emissions trading schemes, and hybrids of these two approaches. All other policies are labeled technology-based (or, simply, technology) policies because they are all, to some extent, technology-specific. A feed-in tariff has to be set for a type (or types) of technology. Clean energy targets have to be defined in relation to a set of clean` technologies. A research and development (R&D) subsidy, however broad its scope, has to be restricted to particular types of technologies. In this sense carbon pricing instruments are, by contrast, technology neutral: they do not require the government to pick winners.` 2.3. Carbon pricing instruments are by definition fiscal, whereas technology-based instruments can be fiscal or regulatory. A non-exhaustive classification of policy instruments along these lines can be found in Table 2.1. A comprehensive analysis of climate change mitigation policies is beyond the scope of this report. Burer and Wustenhagen (2009) list 23 renewable energy policies alone. A similar number of possible energy-efficiency policies could be enumerated. When it comes to technology-based policies, the focus is on fiscal ones designed to promote renewable energy. However, the points made hold more broadly, in relation to all technology-based policies. Table 2.1: Classification of climate-change mitigation instruments Carbon pricing Technology-based Fiscal Fiscal Regulatory Emissions trading Demonstration grants Technology performance standards Carbon tax Public R&D Renewable fuel/energy standards Hyrid trading-tax Investment subsidies Building regulations schemes Preferential tax treatment Automobile regulations Government investment in venture capital Information standards Public investment vehicles Feed-in tariffs Tax credits Public procurement Renewable energy certificate trading Subsidies for energy-efficiency purchases 13 Stern (2006, p.308) has a similar, but somewhat different classification. He divides policies into those which put a price on carbon, those which promote technologies, and those which remove the barriers to behavioural change. Our categorization is similar, but Stern`s is on the basis of the aim of the policy (which market failure is being tackled), whereas ours is in terms of the nature of the policies (whether they require government promotion of particular technologies). This enables us to combine the second and third categories into one (since both require some technological specification). In practice, however, we give little attention to policies in Stern`s third category (such as the provision of information) due to space limitations. 18 2.4. Broader policies and structural reforms are also important for climate change mitigation. All parts of the economy use energy and emit CO2. Energy sector reforms in particular, and economic reforms more generally can have a powerful impact on emissions trajectories. This is a theme that is returned to in Chapter 4. 2.2 Carbon pricing 2.5. Carbon pricing is universally regarded as essential for effective climate change mitigation. Professor William Nordhaus, one of the world`s leading climate change economists, writes: Whether someone is serious about tackling the global-warming problem can be readily gauged by listening to what he or she says about the carbon price. Suppose that [a] person proposes regulating the fuel efficiency of cars, or requiring high-efficiency lightbulbs, or subsidizing ethanol, or providing research support for solar power--but nowhere does the proposal raise the price of carbon. You should conclude that the proposal is not really serious and does not recognize the central economic message about how to slow climate change. To a first approximation, raising the price of carbon is a necessary and sufficient step for tackling global warming. (Nordhaus, 2008, p.22) 2.6. While not all would agree that carbon pricing is sufficient for tackling global warming, nearly all economists would assent that it is necessary.14 Emitting greenhouse gases causes global damage. Putting a price on carbon (and other greenhouse gases) is the most effective way to get individual actors to internalize the environmental damage they are causing. 2.7. By pushing up the relative price of emissions intensive goods, a carbon price reduces emissions in four ways. First, it pushes consumer demand in the direction of goods which are less emissions intensive (e.g. to wear extra clothing and turn down the heating). Second, it induces suppliers to make their goods less emissions intensive (e.g. to make electricity with gas instead of coal). Third, it leads investors to invest in less emissions-intensive projects (e.g. to build an aluminum smelter which runs on hydro rather than thermal power).15 And, fourth, carbon-pricing gives a financial incentive for innovators to develop new products, which are less emissions-intensive (e.g. to invent a hydrogen or electric car). 2.8. International comparisons of energy prices and usage point to the importance of pricing as a determinant of energy efficiency. To simplify somewhat, the message from Figure 2.1 is that the US and Canada have electricity and gasoline prices at 50% below the levels prevailing in Japan and Europe, and energy per unit of output at 50% above. No doubt the relationship runs both ways (with higher energy intensity in North America leading to political resistance to tax hikes), but it is also clear that higher energy prices will encourage energy efficiency, and that, as a special case of this, the introduction of carbon pricing will discourage fossil fuel use. The experience with the European emissions trading scheme, though fraught in various ways, confirms that putting a price on carbon leads to abatement (see Box 2.2 and Ellerman et al., 2010). 14 Even those economists who argue that technology-based policies are more important than carbon pricing still call for carbon pricing as part of the mix (see Montgomery and Smith, 2007) 15 In a world of partial mitigation, suppliers and investors may also respond by moving emissions-intensive production off-shore. This is the problem of carbon leakage. 19 Figure 2.1: Developed economies have low energy prices or high energy efficiency, but not both Electricity prices, gasoline prices, and energy intensity (ratio of energy use to GDP) for US, Canada and Japan relative to the OECD member economies of Europe. Notes: Prices measured in current USD, using market exchange rates. For energy intensity definitions, see Figure 1.4. Energy efficiency is defined as the inverse of energy intensity. All OECD Europe values are normalized to one. Sources: IEA (2009a, 2010b) 2.9. To date little use has been made of carbon pricing outside of Europe. Technology-based policies have been much more popular. As an example, China`s largely regulatory approach to reducing its energy intensity is described in Box 2.1. Among APEC economies, only New Zealand, and some American and Canadian states have actually introduced a price on carbon (Table 2.2). In many economies, carbon pricing is still politically controversial, and it will remain so until adopted by the world`s largest economy, the United States. 2.10. APEC economies, including developing ones, are, however, showing increasing interest in carbon pricing. Table 2.2 reports on the status of carbon pricing plans and discussions across APEC. Within APEC, only New Zealand and some American states have actually introduced an emissions trading scheme. Emissions trading schemes are currently being prepared and/or debated in the US, Australia, Japan, and Korea. China and Indonesia are both contemplating the introduction of carbon pricing. Both a carbon tax and emissions trading seem to be under consideration in China. Some media reports suggest that China is contemplating the introduction of a carbon tax. Rates mentioned are 20 Yuan per tonne of CO2 (about $3) rising to 50 Yuan by 2020.16 Other media reports suggest that China will in fact introduce emissions trading. Indonesia`s Green Paper (Ministry of Finance, 2009) proposes the introduction of a carbon tax at Rs 80,000 per tonne of CO2 (about $8), rising at inflation plus 5% to 2020. 16 http://www.businessgreen.com/business-green/news/2262857/reports-china-impose-carbon-tax 20 Box 2.1: China's instruments for achieving a 20% reduction in energy intensity China`s main method for pursuing its national goal of a 20% reduction in energy intensity by 2010 compared to 2005 (see Box 1.2) has been to cascade this target down, and convert it into targets for provinces and large enterprises. Thus each province has been given its own energy intensity reduction target (above, below or equal to 20%). Similarly, each province has translated its goal into targets for its cities and counties. The top 1000 enterprises have also been required to show energy savings, and have each been given their own, tailored target. To ensure these targets are taken seriously, each provincial leader has been required to sign a contract taking responsibility for its energy intensity target with the central government. City and country leaders have likewise been required to sign contracts with their provincial government. Contracts have also been signed with enterprise managers. Progress of both provinces and enterprises is reported annually at the highest political level, and is taken into account during the performance evaluation of provincial, local and enterprise managers. China has also set detailed quantitative targets for closing down old and inefficient plants and factories in major industrial sectors. It has also established 10 new technological projects or thrust areas, to introduce more efficient coal-fired power plants, and to promote energy-efficient lighting and more efficient buildings, and so on. These projects have been backed by high levels of government spending, expected to exceed $10 billion by the central government alone in the current year. A number of fiscal policies have also been used. Energy-intensive industries have been subjected to higher electricity prices and reduced export rebates. Source: Wang and Chen (2009); Zhou, Levine, Price (2010). 2.11. Some developing economies (in and out of APEC) have already introduced carbon-price-like levies. China has introduced a levy (set at 0.002 CNY/kWh) on electricity to subsidize renewable energy. In July 2010, India introduced a Rs 50/tonne (about $1) levy on coal to support renewable energy. Thailand has a Baht 0.04/liter (USc 0.1) levy on petroleum products which goes to an energy conservation promotion fund. Vietnam has also proposed an environmental levy to come into effect in 2012 on petrol, diesel and coal and a number of other potentially harmful products (such as plastic bags). The rate proposed in the Draft Law on Environmental Taxes for petrol and diesel is in the range of VND 500 to 4000 per liter (between 3 and 20 USc), and between VND 6000 and 30,000 (30c and $1.60) per tonne of coal. 2.12. While there is consensus among economists that carbon needs to be priced, there is no agreement about how it should be done. A carbon price can be introduced either through a price- or a quantity-based approach. A carbon tax is an example of a price-based approach, and an emissions trading scheme an example of a quantity based approach. In the former, the price of carbon is fixed by government, and the market determines the resulting quantity of carbon emissions. In the latter, the government fixes the quantity of permitted emissions, and the market determines the price at which permits will trade to clear the market. Hybrid approaches combining features of the quantity and price based approach are also possible. 2.13. Under conditions of certainty and perfect information, a carbon tax and emissions trading scheme are equivalent. In the real world, there are pros and cons to both approaches. A carbon tax provides cost certainty, and reduces price risk for investors. An emissions trading scheme gives certainty in regard to the environmental outcome, but may result in a volatile price. A tax and emissions trading scheme may also have different revenue effects (depending on whether permits are auctioned or given away), and different political economies. Actual experience with the two approaches is summarized in Box 2.2. 21 Table 2.2: Progress across APEC in introducing carbon pricing Country Status US Emissions trading legislation passed by House of Representatives June 2009, pending in Senate. Under the Regional Greenhouse Gas Initiative (RGGI), ten Northeastern and Mid-Atlantic states have capped and will reduce CO2 emissions from the power sector 10% by 2018 compared to 2009 through an emissions trading scheme they jointly established in 2009. (See also Western Climate Initiative, under Canada.) Japan Emissions trading scheme approved by Cabinet but not yet passed by the Japanese Parliament. Some voluntary ETS markets already established Australia ETS defeated by Senate, deferred until 2013 or later. NSW has energy efficiency trading scheme. NZ ETS started on July 1, 2010 (two years earlier for forestry) Canada Legislation proposed but delayed, pending action in the US. Under the Western Climate Initiative (involving 6 American and 4 Canadian states), an emissions trading scheme will commence in 2012 with the aim to reduce greenhouse gas emissions by 15% from 2005 levels by 2020. Quebec and British Columbia have introduced carbon taxes. Korea Basic Law on Low Carbon Green Growth passed on April 14, 2010 commits government to introduce legislation to establish emissions trading scheme; possible starting date 2012 or 2013 China Government is reported to be considering both a carbon tax and an emissions trading scheme. Energy intensity cap-and-trade scheme commenced in Tianjin in June 2010. Indonesia 2009 Ministry of Finance Green Paper on Climate Change proposed carbon tax starting at Rs 80,000 per tonne of CO2 or about $8, rising at inflation plus 5% to 2020 Mexico Emissions trading scheme under consideration. Sources: Own research. Box 2.2: Real-world experience with carbon pricing The European Emissions Trading Scheme (ETS) is the largest GHG emissions trading scheme in the world. Set up in 2005 in line with the GHG emissions reduction targets under the Kyoto protocol, the ETS caps emissions of about 11,500 power and industrial installations in 25 countries and six major industrial sectors. The ETS has been plagued by several problems, and has resulted in volatile prices, but analysis does indicate it has helped reduce emissions (by 2-5% against business as usual according to Ellerman et al., 2010). Sijm et al. (2006) have showed that utilities passed through a proportion of their carbon costs under Phase I of the EU Emissions Trading Scheme in the form of higher electricity prices. The extent to which carbon costs were passed through to power prices depended on the merit order of the supply curve due. Their research found pass through rates of between 60-100% for wholesale power markets in Germany and the Netherlands despite almost full issuance of free allowances. As a result, many utilities under the ETS enjoyed high windfall profits. Outside of the ETS, the Regional Greenhouse Gas Initiative (RGGI) was set up in 2005 by seven Northeastern states to form the first mandatory cap-and-trade program for reducing CO2 in the USA. Unlike in the EU, permits are mainly auctioned. Prices have been low, in the range of $2-3. New Zealand has passed legislation for an ETS in 2008 and has initiated a transition scheme starting July 1, 2010. In this transition period (until 31 December 2012), participants in the ETS will buy emission units at a fixed price of NZ$ 25 per unit, with energy, industrial and liquid fossil fuel sectors being able to buy at half the price. The global experience with carbon taxes is mainly limited to European countries. Many European countries (notably Germany, all Scandinavian countries, France) have applied fuel taxes, energy and emissions taxes partly contingent on carbon content. Denmark imposed the world`s first carbon tax on fossil fuels in the early 1990s, and this is held partly responsible that for the fact that its emissions have fallen by 5% since 1990 despite 22 annual average GDP growth of just over 2%. Germany has a tax on electricity. In these countries, various exemptions are typically provided for industries. In the electricity sector, both Finland and Germany tried to impose taxes based on the type and source of electricity production, but the EU Court found this discriminated against imported energy. Sources: World Bank 2010d, Sterner, 2003. 2.14. Most models show that, without international transfers, carbon pricing will be more costly to developing than to developed economies. Figure 2.2 illustrates this finding using the EMF-22 multi- model exercise, which used a range of models to analyze the costs of mitigation to achieve a 550 ppm atmospheric stabilization target for carbon dioxide equivalent (CO2e) with a single global carbon price. The models show that the resulting costs as a percentage of GDP are typically much higher for China and India and much lower for the EU and the US. In terms of major country groupings, they tend to be highest for emerging economies (the BRICs: Brazil, Russia, India and China), lower for other developing economies, and lowest of all for developed economies. This is related to the fact that in general poorer economies tend to be more emissions intensive and so their economies adjust more in response to the introduction of a carbon price (Howes, 2009a and Stern and Lambie, 2009). Figure 2.2: Most models show that a given carbon price will impose proportionately greater economic costs on a developing economy than a developed economy. 6% 6% China Global average EU 5% 5% Developed countries India BRICs US Other developing countries 4% 4% % GDP loss % GDP loss 3% 3% 2% 2% 1% 1% 0% 0% Cost to GDP in 2030 of a global carbon price according to a number of different economic models. Notes: The different models (shown on the horizontal axis) are those used in the EMF-22 modeling exercise to estimate the cost to GDP in 2030 of a global carbon price set at a price high enough, with universal participation, to stabilize the concentration of greenhouse gases at 550 ppm CO2-e.The price ranges from $40-85 per tonne of CO2 in 2005 prices. GDP and carbon prices are valued at market prices. BRICs are Brazil, Russia, China and India. All models are shown for which cost to GDP results are available. No international transfers are assumed. Source: Lee (2010) and EMF (2009). 2.15. However, these costings typically ignore the revenue benefits of carbon pricing which could be substantial, especially in developing economies. Carbon pricing could generate substantial revenue for a significant period of time. Figure 2.3 shows that most APEC economies could by 2020 generate revenue of about 0.5% of GDP or more if they placed a $20 price per tonne of CO2 (in 2005 prices) through either a tax or auctioned permits. (Probably a higher price would be needed to meet the 2020 targets.) The revenue implications are especially significant for APEC`s developing economies, since their emissions 23 are higher relative to GDP when measured using market exchange rates. A $20 carbon price applied across fossil fuels could fetch China in excess of 2.5% of GDP by 2020. Figure 2.3: A $20 carbon tax in 2020 would raise significant government revenue, especially for the poorer APEC economies Revenue raised by a $20 tax per tonne of CO2 emissions from fossil fuels in 2020 for various APEC economies, assuming that national 2020 targets are met Notes: 2005 prices; American dollars. Revenue raised is as a percentage of GDP measured at market prices; GDP per capita is measured using PPPs. The assumption that 2020 targets are met is used simply to derive an emissions volume by which the $20 price can be multiplied to derive revenue. It is not being claimed that a $20 price is what is needed to achieve these targets; likely, a higher price would be needed. Sources:.National targets and GDP data from Figures 1.6 and 1.7. 2.16. If the revenues raised from carbon pricing are used to reduce other taxes, this will reduce the economic cost of mitigation, and possibly turn a cost into a gain. Most modeling assumes that carbon pricing has no revenue effects (with revenue simply being returned in lump-sum form to the economy rather than being used to reduce other taxes). However, if the revenue raised was used in place of other taxes, this would offset at least part of the costs of pollution control, and raise the prospect of a double dividend` from the introduction of a carbon tax. 17 2.17. There are several factors on the revenue side which could help offset the economic cost of a carbon price: 17 Strictly, in the jargon of the literature, a strong` double dividend, i.e. a net gain. A weak dividend is assured: that is, the costs of a carbon price will be lower with non-lump-sum revenue recycling than without. Note, the argument here is not that revenue recycling gives price instruments an advantage over non-price ones, such as command and control (it doesn`t; see Fullerton and Metcalf, 1997), but simply that revenue recycling will reduce the costs of carbon pricing. 24 In general, analysis finds that increasing broad-based commodity taxation and reducing personal and corporate taxes are efficiency-enhancing. The difference can often be large as Table 2.3 illustrates for Canada. Recent cross-country econometric work by the OECD indicates that a shift of 1% of tax revenues from income taxes to consumption and property taxes would increase GDP per capita in the long run by between ¼-1% (Johansson et al., 2008, and Arnold, 2008). Table 2.3: Long-run economic effects from revenue-neutral tax reductions: an example from Canada Percentage change in Welfare gain (in dollars) GDP per 1% of GDP Tax Measure: per dollar of lost revenue reduction in revenue A cut in personal capital income taxes 1.30 3.36 A cut in sales taxes on capital goods 1.29 3.05 A cut in corporate income taxes 0.37 1.94 A cut in personal income taxes 0.32 1.29 A cut in payroll taxes 0.15 0.66 A cut in consumption taxes 0.13 0.19 Notes: The revenue loss is assumed to be recovered through non-distortionary lump-sum taxation Source: Baylor and Beauséjour (2004). Economic theory provides some support for taxing relatively inelastic goods, such as energy, at a higher rate.18 In developing economies, modern forms of energy are consumed proportionately more by the rich than the poor (Section 3.2.3). Carbon pricing in developing economies would therefore be progressive. 2.18. A range of studies examining the welfare impacts of carbon pricing taking into account revenue impacts have been carried out for developed economies:19 In an early paper, Goulder (1992) found that in the US some but not all of the cost of a carbon price would be offset by tax cuts elsewhere: the welfare costs of the introduction of a carbon tax would be 25-32% lower when it was used to reduce income and corporate tax than when it was redistributed in lump-sum form. More recently, Takeda (2007) found in Japan that if revenue from a carbon tax was used to reduce capital taxes, then its introduction would be welfare-enhancing (but not if it was used to replace labour or consumption taxes). Bor and Huang (2010) find that a tax on energy offset by a reduction in income tax would be welfare enhancing for Taiwan. 2.19. Recent work suggests the possibility of a double dividend from global action for energy importers. The Goulder (1992) paper was written at a time when fuel prices were much lower than they are today. And none of the above papers consider the case of a global carbon price, which would give rise not only to domestic revenue recycling effects but also to a softening of fossil fuel prices. Recently, Elekdag et al. (2008) have investigated the impact of a 25% global gasoline price increase. As discussed in Box 1.1, the combination of revenue recycling through reductions in labour taxation and positive terms of trade effects 18 This is the Ramsey Rule. Whether or not the Ramsey Rule applies in the presence of an income tax depends on the functional form assumed to represent preferences. See Stern (1987). 19 For other studies, see Bor and Huang (2009). 25 for oil importers leads to an increase in consumption in the world`s oil-importing regions. The results are summarized in Figure 2.4.20 Figure 2.4: A global increase in gasoline tax rates would make energy importers better off Deviation from baseline for a simulated global increase in gasoline tax rates by 25%. Source: Figure 25 in Elekdag et al. (2008). 2.20. More work is needed on the fiscal impact of carbon prices both in the context of developing economies, especially to take into account the progressivity arguments, and given high fuel prices now being experienced. For a higher fuel price, the same proportional carbon tax will raise more revenue, making a double dividend more likely . 20 Elekdag et al. (2008) note marginal negative welfare costs in the long run, of 0.4% or less, due to negative wealth effects, but they qualify this result as partial. Overall global GDP is higher. 26 2.21. For political reasons, the fiscal gains from a carbon price might only accrue in the medium term. In the shorter term, Pezzey and Park (1998) argue that non-revenue-raising instruments are much more likely to be politically acceptable, and therefore implemented, precisely because Pure emission taxes or auctioned tradable emission permits transfer large sums of money away from emitters. (p.547). But this is a short-run argument. In the first and second phases of the EU ETS most permits were given away for free. In the third (post-2012) phase, however, the proportion of permits which is expected to be auctioned will rise significantly. 2.3 Technology-based policies 2.22. There is a much greater range of, and experience with, technology-based policies than carbon pricing. However, the experience with these policies is often conflicting, and the lessons uncertain. It is not clear what constitutes optimal policy for developed economies, let alone for developing ones. This section first surveys the rationale for technology policies (2.3.1), and then summarizes the experience to date with specific policies (2.3.2). 2.3.1 Technology policy rationales 2.23. The rationale for technology-based policies cannot simply be that they promote technological development. Of course, innovation is critical for effective and cheap mitigation ­ and for solving environmental problems more generally (a point recognized since Kneese and Schultze, 1975). But carbon- pricing would typically be expected to induce innovation on its own. Supplementary technology-specific policies require a more convincing rationale. 2.24. Governments in fact have a wide-range of reasons for introducing technology-based policies. This section enumerates seven. It is important to stress at the outset the difference between a rationale and a justification. However, since the former is a necessary but not a sufficient condition for the latter (which requires consideration of arguments against as well as for), it provides a good starting point. 2.25. The first is simply that many governments have renewable energy targets and industrial policy objectives. As previously noted (in Section 1.3), most governments don`t only want to reduce emissions. Many also want to reduce air pollution, improve energy security, and develop technological and industrial advantage. As Table 1.1 showed, many embody these goals through renewable energy targets. Technology-based policies are as important for clean energy targets and industrial policy objectives as carbon prices are for emissions reduction targets. 2.26. A second reason for technology policies is the public good nature of inventions and discoveries. Firms under-invest in research and development because of the fear that their competitors will benefit. Trading off the need to provide incentives to invent (prior to the invention) and the need to make maximum use of any inventions (once they are made) is an important responsibility for governments which they discharge not only through the creation of patent systems but also through public funding for research and development (R&D). If climate change mitigation and clean energy are important social goals, then they will also be claimants on the public R&D budget.21 2.27. APEC, led by Japan, dominates global innovation in climate change technologies. Dechezleprêtre et al. (2008) measure technological innovation in respect of climate change mitigation using patent filings. 8 of the top 12 most inventive economies are in APEC, with Japan alone being responsible for 37% of the world`s climate change mitigation inventions (Table 2.4). The US is in second position, and China, South Korea and Russia in 4th, 5th and 6th positions respectively. China`s recent Medium and Long-term Development Plan for Renewable Energy in China (2007) explicitly identifies the deployment of Chinese 21 For a recent restatement of the importance of R&D subsidies for environmental problems within a theoretical framework, see Acemoglu et al. (2009). 27 intellectual property domestically as a future policy objective. Other middle-income APEC economies, such as Malaysia, also give heavy emphasis to innovation, which they see as critical for escaping the so- called middle-income trap.` Table 2.4: Top 12 inventors in climate change mitigation technologies, with average percentage of total global inventions across different mitigation technologies Country Rank Average % of world's inventions Japan 1 37.1% USA 2 11.8% Germany 3 10.9% China 4 8.1% South Korea 5 6.4% Russia 6 2.8% Australia 7 2.5% France 8 2.5% UK 9 2.0% Canada 10 1.7% Brazil 11 1.2% Netherland 12 1.1% Total - 87.2% Notes: Inventions are measured based on patent count data. The percentages shown average over 13 different climate change mitigation technology areas. These include not only renewable energies, but also relevant inventions in the area of building, lighting, CCS and cement. Source: Dechezleprêtre et al. (2008). 2.28. Renewable energy R&D is important for all economies, not only global technology leaders. Renewable energy technologies have to be tested for environmental conditions which vary from economy to economy (Knight, 2010). 2.29. The public good nature of information is relevant not only to inventions but also to discoveries. Whether to make discoveries public or proprietary is another important government decision, with considerations parallel to those involved in setting up a patent system for inventions. Consider the challenge Indonesia, host to 40% of the world`s geothermal resources, faces in trying to exploit those resources, only 3% of which are currently being utilized as an energy source. Currently, firms negotiate geothermal power supply contracts prior to exploration of the geothermal resource they will utilize under the contract. Indonesia`s Climate Change Green Paper (Ministry of Finance, 2009) argues that it would make better sense to negotiate these power supply contracts after exploration, so that firms don`t have to build a risk premium into their bid to protect against the eventuality that the resource base they have just been awarded turns out to be sub-standard. If individual firms undertake pre-bidding exploration, then either (if just one firm explores) there will be asymmetric information which the better-informed firm will exploit, or (if multiple firms explore) there will be waste. Therefore, while resource exploration is usually a private sector activity, the Green Paper argues that the government should undertake exploration itself and make the results public. Firms would then have a better idea of the asset they were buying and, facing less risk, would be prepared to accept lower tariffs. Of course, that exploration would cost the government fiscally, but these costs would more than offset by a reduction in geothermal tariffs, and the government could recover costs by charging successful bidders upfront fees. 2.30. Third, public promotion of new technologies, beyond support for research, can be justified by consideration of dynamic increasing returns generated by learning-by-doing, learning-by-using and network externalities. These concepts (explored in Jaffe, Newell and Stavins, 2002) formalize the simple idea that the more people use a technology the cheaper it will be. Successful innovation is a long and arduous process, as Figure 2.5 illustrates. Lee, Iliev and Preston (2009) estimate that the average period for taking a new energy technology to market ­ to traverse the valley of death` as it is often called ­ is 20 to 28 30 years. Throughout this period, early-movers can generate spill-over effects which are of benefit to society but cannot be privately appropriated. This not only deters R&D, already discussed, but also discourages firms from taking investments through the valley of death to the point of commercialization. The disadvantages of being a first-mover are manifold. Actions to commercialize new technologies give rise to spill-over benefits for competitor firms relating to skills training, regulatory development, support capacity, and social acceptance (Garnaut, 2008). If product differentiation is possible, first-movers can charge a premium, but product differentiation is impossible in the case of electricity production, a homogenous good. Figure 2.5:The long innovation chain: how a new invention gets to the market Source: Garnaut (2008, Figure 15.1) based on Grubb (2004) 2.31. Fourth, policy risk is unavoidable for renewable energy, but can be reduced by technology polices. Given that no renewable energy technology has yet reached price parity in its production costs with coal- fired electricity, a profitable return on the development of these technologies may be dependent on many years of government subsidy, whether through a carbon price or other means. This makes renewable energy projects quite different from other areas of high innovation, such as IT and biotech projects. 22 Technology-based policies can help reduce policy risk by providing support upfront (e.g., through capital subsidies) rather than over time and by embedding support that is provided over time into legally binding contracts (e.g., through feed-in tariffs). 2.32. A fifth rationale for technology policies rests on capital market failures. The large upfront costs and high risk profiles of renewable energy demonstration projects can lead to inadequate market financing. Renewable energy projects typically have high capital and low operating costs. The upfront cost of a renewable energy technology demonstration project is upwards of $100 million per project (Knight, 2010), a multiple of what it costs to demonstrate a typical biotech or IT project. Such a sum is typically too big for venture capitalists who have relatively small funds under management, often in the range $100-500 million (Shellenberger et al. 2008). Project financiers and commercial banks have the capital base to lend for deals of this size but are much more risk-averse and unwilling to take on the technology and policy risks involved (as per the previous paragraph). They will finance less risky deals (hospitals, toll roads, airports) ahead of renewable energy projects because the risk/return premium is more predictable and favorable. Power utilities themselves are also traditionally risk-averse and spend little on R&D. The result is limited private sector financing at a high risk premium for breakthrough renewable energies (Justice, 2009). Technology policies can address this either by reducing policy risk or by capital market interventions. 2.33. Sixth, technology-based policies are not only about new technologies. A range of information and agency barriers can restrict the use of known technologies. In the most well-known case, landlords, who pay for capital costs, will under-invest in energy efficiency, since the benefits in terms of reduced energy bills will flow to their tenants. Or consumers, because they lack the relevant information, will judge durables on the basis of upfront rather than lifetime costs. The policies to respond to these market failures 22 Biotech has higher policy risks than IT, but mainly relating to regulatory approval, whereas renewable energy projects will face policy risks throughout their life. 29 (discussed in Chapter 17 of Garnaut, 2008) are not analyzed in detail in this chapter, which focuses on innovation, but they are technology-based policies. If governments want to step in to solve these market failures, they have to be prescriptive and back particular technologies and technical standards, e.g., for building or appliance efficiency. 2.34. Seventh and finally, technology-based policies can sometimes be considered substitutes for rather than complements to carbon pricing. As noted earlier, carbon pricing is still controversial in many economies. Some economies might be unwilling to introduce carbon pricing but willing to implement technology-based policies. In the same year in which the Australian Parliament voted down an emissions trading scheme, it provided bipartisan support for a renewable energy credit trading scheme. Moreover, as is argued in Chapter 4, in some economies and sectors carbon pricing might be ineffective even if implemented, which would be another reason for looking at other emissions-reducing policies. 2.3.2 Technology policy options 2.35. The vast array of technology-based fiscal measures can be divided into two types. Demand-side (or demand-pull) measures are policies put in place to increase demand for a technology. Supply-side (or technology-push) measures refer to policies which aim directly to reduce the (private or social) cost of a technology. The paragraphs following outline the main examples of each type of policy used in practice to promote renewable energy. 2.36. Renewable energy certificates (RECs). A market mechanism for meeting renewable energy targets can be established by issuing renewable energy generators with credits (in proportion to the clean energy they generate) which can then be traded and purchased by utilities in fulfillment of the renewable portfolio standard mandate imposed on them. This stimulates private investment because investors in renewable energy are now required to compete with each other rather than with coal-fired electricity where they have a price disadvantage. Such an approach was pioneered in the United States, and has also been used by Australia, among APEC economies. 2.37. Feed-in tariffs have been widely used in Europe, especially to promote wind and solar power, and have become increasingly popular elsewhere, having been adopted by 45 economies in total, including by several in APEC (Table 2.5). Feed-in tariffs vary by technology and offer price certainty in the end- market to renewable energy generators. Typically, feed-in tariffs guarantee grid access for renewable energy suppliers under a long-term contract. The costs of feed-in tariffs can be borne either by the purchasing utility, the consumer, or the taxpayer. Table 2.5: APEC economies with feed-in tariffs Country Technology Australia* Solar Canada* Solar China Wind Indonesia Renewable Japan* Solar The Philippines Wind, solar, biomass, small hydro, and ocean Republic of Korea Solar Thailand Wind, solar, biomass, and micro-hydro Note: * indicates that the feed-in tariffs are used in at least some regions, but not nationwide. Source: Own research. 2.38. Feed-in tariffs and RECs are different ways to achieve the same goal of boosting demand for renewables. There is now considerable experience with both but an unresolved debate, which parallels that of carbon taxes versus emissions trading, as to which is the superior approach (Box 2.3). 30 Box 2.3: Feed-in tariffs versus Renewable Energy Certificates Much as there is an ongoing debate about the relative merits of a carbon tax and an emissions trading scheme, there is a similar discussion around feed-in tariffs versus Renewable Energy Certificate (REC) schemes for promoting renewable energy. Some of the contours of the debate are familiar: feed-in tariffs (like a carbon tax) fix a price; RECs (like an ETS) fix a quantity. But there are added complications: feed-in tariffs are often more technology specific, with different tariffs for different types of renewable energy. Also there is a lot more real-world experience: feed-in tariffs have been the tool of choice in mainland Europe, while REC schemes have been used in the US and Australia. Germany pioneered the use of feed-in tariffs as early as 1991. They are now used in some 45 countries. The development of REC schemes in the US has been at the state level. As of July 2010, 30 states and Washington DC have REC schemes in place, and a further six states have voluntary targets (DSIRE, 2010). The strength of an REC scheme is that it does not require governments to choose which specific renewable technologies are appropriate to subsidize. It will also provide an assured clean energy outcome, in accordance with the goal set by government. However it may also lead to higher levels of risk for investors, due to exposure to fluctuating levels in both the electricity market and REC prices (Mitchell et al., 2006). Not surprisingly, RECs advantage more competitive technologies, while feed-in tariffs provide incentives to innovate in technologies which are less competitive today but which may in the long run hold more potential (Johnstone, Hascic and Popp, 2010). A recent review undertaken by the Commission of the European Communities (CEC, 2008) concluded that it was not possible to say which of the two approaches was better. 2.39. Biofuel targets and policies. Targets also play an important role in the biofuels sector. Biofuels are controversial because the environmental integrity of this production process is contentious. The controversy has grown as the targets have spread. In 2003, the EU set a target for biofuels to reach 5.75% of transport fuels by 2010. India has just set an ambitious target of 20% for biofuel blending for both gasoline and diesel by 2017. Table 2.6 provides examples of biofuel targets for a number of APEC economies. Policies implemented to pursue these targets include biofuel mandates and preferential tax treatment. Some economies (e.g., United States, Australia) have imposed trade barriers to protect national biofuel producers. Table 2.6: Examples of biofuel targets in APEC economies Economy Target Australia 350 million liters of biofuels by 2010 (less than 1% transport fuel demand) Canada 3 billion tonnes biofuels by 2012. China 3 million tonnes by 2010 and 10 million by 2020 Indonesia 2% biofuels by 2010, 3% 2015 and 5% by 2025. Japan 500 million liters of ethanol by 2010 Korea B3 mandated by 2012. Peru E7.8 mandate to be introduced in 2010 Philippines 5% ethanol by 2009 Taiwan E3 mandate in 2011 Thailand 10% biodiesel by 2012. Vietnam 500 million liters by 2020 United States 7.5 billion gallons of ethanol by 2012 (nearly 1/3 US gasoline now contains ethanol) Notes: Targets are for biofuels expressed either in absolute terms or as a percentage of the fuel mix. E10,` for example, is a target of 10% ethanol blended into petrol. Targets are for ethanol only, except where it is specified they are for biodiesels of for biofuels (both). Most economies with ethanol targets also have biodiesel targets. Source: http://www.biofuels.apec.org/index.html 2.40. Research and development subsidies. Turning to supply-side policies, the most direct way governments can promote innovation is by funding it. R&D funding can be paid for out of the budget, or through R&D tax subsidies. The subsidies can extend beyond R&D to cover demonstration and 31 commercialization. Often matching-fund programs are used for these latter phases of the product development cycle. Most publicly funded R&D is undertaken by OECD economies. Unfortunately, and perhaps surprisingly, there has been little real increase in public R&D spending on energy by OECD economies over the last two decades, though the recent uptake is encouraging (Figure 2.6). Figure 2.6: Energy R&D funding by OECD economies has stagnated since the second oil crisis, but is now starting to pick up R&D energy spending in billions of US dollars at constant prices Note: Purchasing power parities used to convert to USD Source: IEA (2010c) 2.41. Policies to address financing risks. Governments have a number of routes to address the capital market failures associated with renewable energy innovation. For example, in the United States, conditional loan guarantees are offered by the Department of Energy to promising venture-backed clean tech companies which were struggling to raise capital. The technologies covered include biomass, solar, wind, hydropower, carbon sequestration, and advanced fossil energy coal. 2.42. The multilateral agencies also offer international guarantees. The World Bank and Global Environment Fund have recently set up the Geothermal Energy Development Fund, which offers partial guarantees for risks such as the short-term upfront geological risk of exploration. The multilaterals are also increasingly playing a direct carbon-financing role. 2.43. Many Asian economies have a tradition, with mixed results, of directed credit favoring ,,priority sectors.` Already in some Asian economies, such as Korea, government-owned banks have been given lending targets for the green energy sector, and are providing guarantees. 2.44. Both the United Kingdom and the United States are considering proposals to establish ,,Green Banks.` The report of the UK Green Investment Bank Commission makes a strong case for a direct- financing role of government (as discussed in Box 2.4). 32 Box 2.4: Green Investment Banks for the UK and the US? The United Kingdom`s government-established Green Bank Commission delivered its report in June 2010. It recommended the creation of a Green Investment Bank, as an independent organization capitalized with £1 billion government investment matched with £1 billion of private sector investment. The Commission`s report starts with two estimates. One is that to achieve its national emissions reduction target, the UK will need £55 million of low-carbon investment annually to 2020. The other is that actual low-carbon annual investment is only £9 million, despite the existence of a strong emissions reduction target (with legislative backing), a Europe-wide carbon price and a whole host of other technology-based policies, including renewable targets and feed-in tariffs. The Commission argues that various market failures as well as a lack of policy credibility are holding back the required investment, and recommends the establishment of the Green Investment Bank (GIB) to fill the massive gap. The GIB, like any other bank, would have two financing roles. First, it would act as a financing intermediary, and in particular try to tap into the huge pool of funds managed by institutional investors. The Commission argues (p.9) that Institutional investors, with their long-term liabilities and vast pools of capital could provide a significant proportion of the funds needed to fill the investment gap. UK pension funds, for example, manage £1.5 trillion of assets. Such funds will not invest in individual projects, but they do buy bonds, and the Commission argues they could be interested in investing in Green Bonds which the Bank would use to finance low-carbon investments. Second, the GIB would invest. The Commission proposes a range of ways the Bank allocate its capital from the provision of guarantees and insurance mechanisms to the taking of equity stakes to the aggregation of small energy- efficiency projects. In particular, the Commission argues that the bulk of new clean energy projects will be led by the ten large utilities that already operate in the UK. However, these incumbents are, it is argued, income-focused and risk-averse and have little interest in expanding their capital base. By co-investing equity in projects managed by the utilities, the Commission reasons that the GIB will make renewable energy projects more attractive to these conservative utilities, and so increase their investment in clean energy. Although the previous UK government had committed itself to the establishment of the Green Investment Bank, recent press reports indicate that the new government may have shelved the idea. Several legislative initiatives before the US Congress in recent years would also establish a Green or Clean Energy Bank. Under the Green Bank Act of 2009, introduced by Congressman Chris Van Hollen, the Bank would be established as an independent, tax-exempt, wholly owned corporation of the United States, with an initial capitalization of $10 billion through the issuance of Green Bonds by the U.S. Department of Treasury. The Green Bank would provide loans, loan guarantees, debt securitization, insurance, portfolio insurance, and other forms of financing support or risk management to qualified clean energy projects and qualified energy efficiency projects, all in cases where it was judged that the private credit market was not providing adequately low-priced financing. However, provision for such a Bank does not appear in the main climate change legislation which the House and Senate has considered, and has not yet been endorsed by the US Administration. Sources: Green Investment Bank Commission (2010), Koo (2010) 2.45. Some governments offer loans or guarantees to home owners or building occupants seeking to finance climate-related installations. Mexico, for example, has established green mortgages` at favorable interest rates for home owners seeking to finance solar water installations. 2.46. Policies to reduce investment costs. Fiscal policies targeting investment costs aim to reduce the high start-up costs of renewable energy technologies. Options include investment tax credits,23 accelerated depreciation, tax holidays, and low tariffs on capital equipment. For example, the United States has set a federal 30% investment tax credit for households and utilities until 2016 for solar PV, solar thermal power, solar hot water, small wind and geothermal. In India, the government has allowed accelerated depreciation for a number of key renewable technologies. In China, the high upfront costs of clean energy projects are 23 An investment tax credit allows a company to write-off an investment against tax. A production tax credit allows a company which generates renewable energy to write-off a portion of their tax obligations based the volume of energy (cents/ kwh) generated. 33 reduced by tax holidays and import duty exemptions. For a selection of renewable technologies, the standard VAT rate of 17% is reduced to 13% for biogas, 8.5% for wind, and 6% for small hydro-projects. Import duties are also lower in China for imported renewable energy technologies. 2.47. Policies to reduce operating costs in the USA have used production tax credits for some time (see footnote 23). The American Recovery and Reinvestment Act (2009) extended the eligibility of companies generating wind, solar, geothermal energy and closed-loop` bioenergy for production tax credits (until 2013, and 2012 for wind). These tax credits last for the first ten years of the facility`s operation. 2.48. Both supply-side and demand-side policies can be effective. Some of the outcomes sought by technology-based policies (such as cost reductions and technological breakthroughs) are inherently difficult to observe, and harder still to attribute to policies. However, experience suggests both supply- side and demand-side policies can deliver large boosts to renewable energy use. Wind power capacity has expanded rapidly in recent years in both China and the US, in response to a mix of policy tools, as Box 2.5 explains. Box 2.5: Same end, multiple means: the promotion of wind energy in China and the United States The impressive growth in wind power in both China and the US has been driven by a number of government initiatives in both economies. In China, the Wind Power Concession Programme (implemented from 2003) provided a guarantee for connection and power purchase for large-scale wind projects. The winning bidder had its bid price guaranteed as a feed-in tariff for the first 30,000 hours of electricity produced (Wang 2010, p.706). (China has recently placed caps on the feed-in tariff.) Following the Annual additions to wind power capacity (MW) in China and Renewable Energy Law in January the US 2006, and the Interim Measure of Renewable Energy Tariff and Cost Sources: China Wind Energy Association (2010), and Sharing Management, a national http://www.windpoweringamerica.gov/wind_installed_capacity.asp electricity surcharge of 0.002 CNY/kWh was put in place in order to subsidize renewable energy. The Medium and Long-term Renewable Energy Development Plan (implemented August 2007) provided for a mandatory renewable market share, of 10% by 2010 and 15% by 2020. Tax incentives have also been used, including VAT rebates and import duty concessions. Finally, international support has also played an important role, with nearly all wind power in China supported financially under the Clean Development Mechanism of the Kyoto Protocol. The United States has Renewable Portfolio Standards in 28 states, and wind power has been a beneficiary of these policies. The rapid growth in wind power in the US has also been driven by production tax credits for renewable energy projects which provide credits against taxes for each kilowatt hour of energy produced in a particular year. An optional 30% investment tax credit was also introduced on a one-off basis in 2009- 2010. Other policies include a new Department of Energy (DOE) renewable energy loan guarantee program. The main challenge for the US has been policy consistency. Production tax rules are currently defined by The American Recovery and Reinvestment Act 2009 Earlier production tax credits arrangements were short- lived and expired in 2000, 2002 and 2004 (only to be put back in place). This explains the irregular growth in wind capacity in these years shown in the figure above. Current arrangements are set to expire in 2012. 34 The main challenge in China appears to be grid connection. One estimate is that up to 60% of wind capacity is not connected to the grid. Some reports suggest that grid companies are reluctant to connect wind power plants because of their high cost. 2.49. Technology-based policies seek to correct market failures but can result in government failure. More than a strong rationale and the likelihood of an impact are required to justify the introduction of any specific technology-based policies. Costs and risks also need to be borne in mind. The high feed-in tariff offered for solar in Germany has been successful at getting greater use of PV solar in Germany (which now has the highest installed PV capacity of any OECD economy),24 but at great expense. Frondel et al. (2010) calculate the cost of emissions savings brought about through this scheme at 716 Euro per tonne of CO2, suggest only modest benefits in terms of cost-reduction, and in fact note some perverse incentives which might have actually pushed costs up. Likewise, the biofuel policies adopted by many economies have resulted in serious concerns in about their environmental and economic impact, leading to the Consultative Group on International Agricultural Research Science Council (2008) calling for governments to scale back their support for and promotion of biofuels until better technologies are available. Australia`s recently-abandoned home insulation scheme provides another vivid illustration of what can go wrong with technology-based policies (Box 2.6). 2.50. Clearly, it is not easy to generalize on the basis of lessons learnt from the application to date of technology-based policies. For carbon pricing, while there are still debates over aspects of design, there is reason to be confident that it works, or would work, in developed economies and ask whether it will in developing economies. For technology-based policies the picture is clearly more complex. The report returns to this topic in Chapter 4. Box 2.6: Energy efficiency gone wrong: Australia's aborted home insulation program Energy efficiency policies have the attraction of being good for the environment and good for the economy. The difficulty with improving energy efficiency through technology-based policies is not the expense (which should be negative) but the complexity: typically, the decisions of thousands of businesses or millions of households need to be influenced. The Australian Home Insulation Program provides an instructive case study of how things can go wrong. Introduced partly as an emissions mitigation strategy, and partly as an economic stimulus measure due to the global financial crisis, the program was introduced in early 2009. It seemed to make a lot of sense: insulation could employ the low-skilled, most at risk from unemployment, and the famous McKinsey (2007) global cost curve for mitigation had put home insulation right at the top of the list of the cheapest (negative cost) ways to cut emissions. Certainly in terms of numbers, the program was a huge success. By the time of its closure in February 2010, over one million homes had been insulated, providing employment for thousands. However, the program`s short duration was symptomatic of its troubled life, which was dogged by concerns around the quality of workmanship and fraud. A series of house fires, and possibly up to four deaths, were suspected to be the result of inappropriate installation techniques. With an increasing number of critical stories in the media, the program became toxic for the government, and was swiftly terminated. What went wrong? The program provided an effectively 100% subsidy directly paid by the government to the insulators. The response was overwhelming. Prior to the program, there were less than 75,000 retrofit insulations of houses in Australia per year. By November 2009, the number of claims had peaked at nearly 180,000 per month. This massive growth in demand (much of it elicited by companies contacting households) transformed the industry. The number of installers rose from the hundreds to around ten thousand. For most of the period of the program, in order for an installation company to be eligible for rebates, it had to have just one person undertake a short training course. Yet insulation is a risky business, and shoddy insulation can reduce energy-saving benefits as well as increase safety risks. The ceilings into which insulation goes often contain electric wires, which pose particular risks if foil insulation is used. Households were in no position to supervise the work done. They lacked the expertise, and the rapid growth of the 24 IEA (2009b) 35 market meant that firm reputation could not be used as an effective screening device. The 100% subsidy and the absence of any requirement that households obtain quotes from more than one potential supplier meant that the government was bearing nearly all the costs and financial risks of the program. Yet, the government was not in a position to exercise effective supervision either. The federal government had never managed a program of this type before (which would typically be a state government program in Australia), let alone one of this speed and scale. There was no requirement for sign-off by inspectors before payment, and the establishment of an audit system was delayed. Prior to September 2009, 100,000 installations took place, but only 172 roof inspections were undertaken (Hawke, 2010, p.48). Yet 16% of these inspections revealed problems with quality and 8% with safety (ibid, p.56). This sad story has several lessons. First, most technology-based policies are complex and ill-suited to serve short- term stimulatory fiscal objectives. Second, sound programs have clear rationales. The program was not targeted at rental properties where there is a clear split-incentives problem which could justify government intervention. In fact, most of the beneficiaries were owner-occupiers. Third, easy money attracts low-quality operators. If governments back technological winners, they have to take responsibility for technical standards and risk management. Fourth, 100% subsidies are ill-advised. Fifth, poor management costs political support. It will take a brave government to mount another insulation support program for Australian households, even though one for rental households could certainly be justified. Source: The main source for this box is the Hawke (2010) Review set up by the Australian government to investigate into the management of the program. 2.4 Conclusion 2.51. Table 2.7 summarizes the discussion so far. Many APEC economies have adopted both emission reduction and clean energy promotion targets in pursuit of four high-level goals: the mitigation of climate change, the promotion of energy security, the reduction of air pollution, and the pursuit of competitive advantage. To achieve these targets, economies have a range of instruments to choose from. They can choose carbon pricing and/or technology-based instruments, both fiscal and regulatory. They can also pursue structural reform in pursuit of their targets. Table 2.7: Goals, targets and instruments in climate change and related areas Goals Climate change mitigation (Reasons for introducing the targets) National pollution reduction Energy security enhancement Search for competitive advantage Targets Emission reduction targets Clean energy targets Others (e.g., national pollution standards, energy efficiency targets) Instruments Carbon pricing (Means to achieve the targets) Technology-based fiscal polices Regulatory instruments Structural reforms 2.52. This report takes as given the targets economies have selected, and the goals which have motivated their selection. The key question for the report is which instruments should be selected, and, more precisely, what role fiscal instruments should have. This chapter has presented the rationale for various fiscal instruments, and the experience with them to date. In the next chapter, the report changes tack to examine the context within which whichever instruments are selected will operate in a developing economy. Only with that information can judicious policy choices be made. 36 Chapter 3 The context: energy sector and other important characteristics relevant to mitigation instrument choice. 3.1 Introduction 3.1. The theories which support the introduction of fiscal policies to mitigate climate change have been constructed for developed economies. They assume that energy prices already reflect market costs, and that the energy sector, comprised of profit-maximizing companies, runs along commercial lines. In such an environment, a focus on carbon pricing to internalize the environmental externalities associated with greenhouse gas emissions, and on other technology-based policies to correct the various market failures along the innovation chain is clearly warranted. 3.2. In developing economies, however, the energy sector often has distinctive features. Often, pricing is politicized, prices are set below economic cost, and the sector is dominated by state-owned companies which do not run on commercial lines. What this implies for the effectiveness of carbon pricing and other fiscal policies for climate change has been little considered. The two sections below outline structural features of the energy sector in developing economies (Section 3.2), and articulate a few important and distinctive structural features of their economies more broadly (Section 3.3). They include all, and only, features which, as argued in Chapter 4, are relevant for the question of instrument choice. 3.3. The main focus of this chapter on the energy sector is on electricity. The most important, and by far the most rapidly growing source of CO2 emissions is from the electricity sector as Figure 3.1 shows. This is now double the two next largest sources, direct emissions from industry and transportation. Figure 3.1: Energy and especially electricity is by far the biggest source of CO 2 emissions worldwide. Global CO2 emissions by source Notes: 1. Includes fuelwood at 10% net contribution, large scale biomass burning and decomposition and peat fires. 2. Other domestic surface transport, non-energetic use of fuels, cement production and venting/flaring from gas production. 3. Including aviation and marine transport. Source: Rogner et al. (2007), Figure 1.2. 3.2. The energy sector in developing economies 3.4. This section outlines twelve features which distinguish the energy sector, and in particular the electricity sector, in many developing economies from that in most developed ones. 37 3.2.1 Rapid growth 3.5. Energy growth is much faster in developing economies than in developed ones. This is illustrated for APEC economies in Figure 3.2 below. As discussed in Chapter 1, this reflects faster GDP growth, and a higher energy-GDP elasticity (i.e., a more constant energy intensity). Figure 3.2: Poorer economies experience faster energy growth Average energy growth 2000-2007 and GDP per capita for APEC economies Source: IEA (2009a), World Bank (2010a), IMF (2010a) 3.6. Growth in the electricity sector is particularly rapid. In most economies, rich or poor, electricity grows faster than GDP (Figure 3.3). For the period 1971-2006, this is true for all the APEC economies shown in the figure, except for three (Canada, US and Russia). Since economies grow faster when they are poorer (at least in APEC), developing economies have very rapidly growing electricity sectors. Annual average electricity sector growth for the still developing and newly-developed economies of APEC between 1971 and 2006 is 7.6%. 38 Figure 3.3: In most APEC economies, electricity grows faster than both GDP and energy 14% 12% GDP 10% electricity 8% energy 6% 4% 2% 0% Peru Korea Brunei Mexico Chile Singapore Vietnam China Japan Canada United States Taiwan New Zealand Russia Australia Indonesia Thailand Philippines Malaysia -2% Long-term (1971-2006) annual average growth in real GDP, and electricity and energy consumption (%) Notes: GDP is measured in billions of constant year (2000) US$, using purchasing power parities (PPPs) to convert from local currency. Energy is measured in Mtoe (million tonnes of oil equivalent). Electricity is measured in GWh. The time period for Russia is from 1990 to 2006. Source: IEA (2009a) for GDP and energy, IEA (2009e and 2009d) for electricity. GDP data from World Bank (2010a); Brunei and Taiwan from IMF (2010a). 3.2.2 Ongoing importance of traditional energy 3.7. Many households in many developing economies still rely on traditional energy sources. Globally, 1.5 billion people still do not have access to electricity. APEC developing economies do better than most, though there are still 120 million in APEC who lack access to electricity, two-thirds of them in Indonesia (Table 3.1). Moreover, electrification on its own does not signify exit from the traditional energy sector. Many households in developing economies use electricity for lighting, but still use biomass (wood or dung) or coal for cooking and heating. Table 3.1 presents estimates for some individual economies and for APEC as a whole. Around the middle of the last decade, 54% of the population of APEC developing economies, or almost one billion people, still relied on biomass and coal for cooking. This includes 730 million in China, 120 million in Indonesia and 55 million in Vietnam. 39 Table 3.1: The great majority of households in APEC developing economies have electricity, but over half continue to rely on biomass and coal for cooking and heating Electrification Reliance on biomass and coal for cooking charcoal, wood & dung coal Population without %of % of Electrification electricity national Population national Population Reported rate (%) (m) population (m) population (m) year APEC developing economies China 99.4 8.0 26.7 350 28.9 379 2007 Indonesia 64.5 81.1 53.8 119 0 - 2007 Vietnam 89 9.5 60.3 51 5.2 4 2006 Thailand 99.3 0.4 36.9 24 0 - 2005 Malaysia 99.4 0.2 0.9 0 2.5 1 2003 Chile 98.5 0.3 0.00 0 - 2007 PNG 10.0 87.0 5 0 - 2004 Peru 76.9 6.5 34 10 3 1 2007 Philippines 86.0 12.5 48.6 40 0 - 2004 All APEC 94.3 119 33 599 21 385 All developing economies 78.2 1,456 2,564 436 % in APEC 8.1 23.4 88.2 Notes: Reliance on biomass and coal is judged on the basis of primary fuel usage. Data is for approximately 2008. Sources: IEA (2006a), UNDP and WHO (2009) 3.8. Weaning households off traditional/solid fuels is not easy. While APEC economies have made great progress in reducing the absolute number of people without electricity, the International Energy Agency forecasts that by 2030 there will still be about 900 million APEC citizens using biomass as their primary fuel for cooking, about as many as there are today (IEA, 2006). 3.9. This reliance on biomass and coal for cooking and heating comes at a high health, environmental and social cost. As noted in Chapter 1, in China, solid fuels are responsible for more premature deaths (through indoor air pollution) than modern energy is through outdoor air pollution (Zhang and Smith, 2007). Household burning of biomass and coal is also an important contributor to climate change due to incomplete combustion. Ramanathan and Carmichael (2008) argue that the deposition of black carbon resulting from the incomplete combustion of biomass and coal (in large part through domestic burning) is contributing to the melting of the Himalayas with potentially catastrophic impacts within a few decades. Tollefson (2009) argues that black carbon is responsible for half of the melting of the Arctic ice seen to date. Both black carbon and other aerosols are implicated in the weakening of the Indian monsoon and the north-south shift in Asian rainfall (Ramanathan and Carmichael, 2008). Finally, collection of biomass imposes a heavy burden on women, who are typically responsible for collecting firewood, and also bear a disproportionate share of the health burden because they do the cooking (UNDP and WHO, 2009, p.27). 3.2.3 Energy as a luxury good 3.10. Whereas in developed economies modern energy is a necessity, in developing ones it is a luxury. In poor economies, richer households are more likely to drive cars, have an electricity connection, and cook using a modern energy source. Therefore in developing economies, unlike in developed ones, the 40 share of expenditure on modern energy rises with income. 25 The one exception is kerosene, which tends to be consumed by poorer households for both lighting and cooking. Predictably, biomass (traditional energy) shows the opposite trend: as households rise out of income poverty they transit away from biomass to the modern energy sector. Table 3.2 illustrates these points with data from various Asian economies. Table 3.2: The share of expenditure on modern energy (biomass) rises (falls) with income: energy expenditure shares for the poorest and richest 20% of households for various Asian economies richest 20% Natural gas Gasoline & Electricity Poorest or Kerosene Country Biomass Modern energy diesel LPG Bangladesh Poorest 0.4 ND 1.5 0.0 0.0 1.9 6.2 Richest 1.9 ND 0.5 0.3 0.9 3.6 2.5 Cambodia Poorest 0.1 0.0 1.6 ND NA 1.7 6.6 Richest 2.2 0.7 0.3 ND NA 3.2 2.4 India Poorest 1.3 0.0 2.1 0.0 NA 3.4 8.7 Richest 3.5 2.5 0.8 2.2 NA 9.1 1.4 Indonesia Poorest 2.6 0.0 2.3 0.2 0.0 5.2 3.5 Richest 3.7 0.5 1.6 1.8 0.0 8.1 0.3 Pakistan Poorest 3.4 0.1 0.5 0.1 0.3 4.4 4.4 Richest 4.2 0.4 0.1 2.5 0.9 8.2 1.7 Thailand Poorest 3.2 0.3 0.0 4.6 -- 8.2 1.6 Richest 2.9 0.4 0.0 7.0 0.0 10 0.0 Vietnam Poorest 2.4 0.3 0.4 1.2 NA 4.3 5.3 Richest 3.7 4.7 0.1 4.7 NA 13 0.9 Notes: NA indicates that the fuel was not available; indicates that no household in the quintile used the fuel. ND indicates that survey did not ask for information about this fuel. Source: Bacon, Bhattacharya, and Kojima (2010). 3.11. A corollary is that energy subsidies, discussed immediately below, are often regressively distributed. Figure 3.4 illustrates with the case of fuel subsidies in Indonesia. Figure 3.4: Almost half of Indonesia's fuel subsidies benefit the richest 10% of households 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% 1 2 3 4 5 6 7 8 9 10 Poor Household Consumption Decile Rich Share of fuel subsidy received by household consumption deciles Note: For the year 2007. Source: Agustina et al. (2008). 25 Strictly speaking, in the jargon of economics, this means modern energy in developing countries is a superior good (that is, one whose share rises with income). However, we use the word luxury` since it carries with it this meaning in its everyday use. 41 3.2.4 Energy subsidies 3.12. Many developing economies subsidize petroleum products, while some of the large APEC economies provide the largest oil and gas subsidies in the world. The IEA estimates that subsidized consumption of oil and gas by non-OECD economies is on the increase, rising, after inflation, from $161 billion in 2005 to $218 billion in 2007 to $439 billion in 2008.26 Some of the large APEC economies provide the largest oil and gas subsidies in the world. In 2008, 4 were in the top 10 subsidy providers: Russia (3rd, $44 billion), China (5th, $28 billion), Mexico (7th, $21 billion) and Indonesia (9th, $15 billion).27 3.13. Petroleum subsidies have risen in recent years in many APEC economies, although they are still moderate and in many cases not rising relative to GDP. As Figure 3.5 shows, domestic prices did not keep pace with rapidly rising international prices in the run up to the global financial crisis. APEC`s subsidies appear much more moderate when measured in terms of GDP rather than in dollars. In none of the economies do they exceed 5% of GDP, whereas in the Middle East they are in excess of 10% or even 20%. The highest APEC subsidies are in Indonesia, Russia, and Malaysia, all energy exporters. In several APEC economies, even though subsidies have increased in dollar terms, they have not risen relative to GDP. As Figure 3.5 shows, oil and gas subsidies as a percentage of GDP have actually fallen in some APEC developing economies (Indonesia, Russia, Vietnam) and remained stable in others (China). Figure 3.5: APEC developing economy oil and gas subsidies: increasing in absolute terms but not necessarily relative to GDP USD billion (2005 prices) % GDP 50 6% 2005 45 2005 2007 5% 40 2007 35 2008 2008 4% 30 25 3% 20 15 2% 10 1% 5 0 0% a nd na a m a o a an nd na a m a o an si si si ic si si na si ic la na hi ne ay la iw us hi ex ne ay iw us ex ai C et ai C et Ta al do R M Ta al Th do R M Vi Th Vi M M In In Oil and gas subsidies for major APEC developing economies for 2005, 2007 and 2008 in billions of USD and as a percentage of GDP. Notes: Data is provided for those economies which are included in the IEA graphs of top (about 20) non-OECD energy subsidy providers for the various years. All subsidies are reported by the IEA in billions of current USD using market exchange rates. These are converted into 2005 prices, using the US GDP deflator. They are calculated as a percentage of GDP by dividing the subsidy value by the value of GDP also measured in current USD. IEA presents its data only in graphical form. Estimates from the graphs are obtained by Koplow (2009) for 2005 and 2007, and by us for 2008. Source: Koplow (2009) based on IEA figures and IEA (2010d). 3.14. Pump prices are not particularly low in most APEC developing economies compared to developed economies, but are in some oil exporters. The low prices for both diesel and petrol in Indonesia, Malaysia and Brunei are evident from Figure 3.6. Some other APEC developing economies keep diesel prices below petrol, but for petrol only the three ASEAN economies just mentioned had lower pump prices than the United States in 2008. 26 All figures for the top 20 non-OECD economies for which IEA presents data for these years, and in 2005 US$, using the US GDP deflator from the World Bank (2010a). See Figure 3.5 for subsidy data sources. 27 See notes to Figure 3.5 for details. 42 Figure 3.6: Pump prices in APEC economies show a lot of variation, but are not strongly related to income Diesel and petrol retail prices in 2008 plotted against income per capita Notes: Russia`s petrol and diesel prices are above those of the US, but it still has large fuel subsidies (Figure 3.5) due to large gas subsidies. 2008 fuel prices are converted to USD using market exchange rates, but GDP using PPPs. Sources: IEA (2010b) 3.15. Coal and electricity subsidies are more difficult to calculate. Subsidies for oil and gas are easy to measure because there is a clearly identified international price which can be used to benchmark economic costs. Subsidies for coal (which is partially traded) and electricity (non-traded) are harder to measure, and require more focused country studies.28 3.16. There are significant coal subsidies in Vietnam and China. In Vietnam, the World Bank (2010, p.14) estimates that the Government`s 2009 decision to price coal for electricity production following market principles will double the price of coal. Though China has been liberalizing the price of coal over recent years, a study by Mao et al. (2008) concludes that the price of coal would be higher in China by 18% if all government subsidies in the production and distribution of coal were removed. In recent years, the price of coal in China has risen sharply, as illustrated by Figure 3.7, which plots the spot or market price for coal. A lot of coal (about 70%) is sold under long-term contract, but in 2007 price controls for long-term contracts were removed (Rosen and Houser, 2007 p.25). Contract prices are significantly lower than market prices, but should over time follow the latter upwards. Market coal prices spiked in the middle of 2008, with some prices reaching in excess of 1000 Yuan per tonne. At this point, the Chinese government capped the market price at 800 Yuan. 28 The IEA present some estimates of coal and electricity subsidies for some years, but the most recent year is 2007, little detail is provided concerning method, and they appear to be on the low side (as argued by Koplow, 2009). 43 Figure 3.7: The era of cheap coal in China is over Current and constant spot market prices of coal, 1994 to 2010. Notes: Six-month average FOB prices per tonne (1000 kg) of coal at the Qinghaungdo Port for 3 types of coal (where available): Datong Premium Mix 6k, Shanxi Premium Mix 5.5k, and Shanxi and Datong Mix 5k. The CPI deflator is used to obtain the constant price series, using 2009 as the base. Source: National Bureau of Statistics of China (2010) and national Chinese coal data. 3.17. Rising coal prices have led to the re-emergence of electricity subsidies in China. Coal is not the dominant fuel for electricity in Vietnam, but it is in China. Through a series of electricity tariff increases, China greatly reduced electricity subsidies over the 1990s. However, China has found it difficult to pass on the increase in coal costs it has recently experienced. In nominal terms, coal prices rose 40% between the first six half of 2006 and 2010, but electricity prices only by about 15%. In fact, over the last few years, electricity selling prices have not kept pace with inflation, as Figure 3.8 shows. In 2003, coal costs were less than half of the price at which grid companies purchased power from generators. In 2008, they were over 100%, and, despite some relief from falling coal prices, at the end of 2009 coal costs still consumed over 70% of coal-fired generator revenue. How is the sector managing to survive? Much coal is still sold under contract, and contract prices would lag spot prices, given the doubling of the latter since mid-last-decade. Generator profits are also being squeezed and the margin between the final selling price and the wholesale power purchase (generation) price has also fallen. Morse, Rai and He (2009) report that Chinese power companies lost an estimated 70 billion Yuan in 2008. 44 Figure 3.8: Despite rapidly rising coal prices, electricity prices for industry and household in China have not kept pace with inflation in recent years Average electricity selling price for industry and households, all of China, 2006-2010 Notes and sources: National prices are calculated by taking weighted averages for provinces from SERC (2006-2009) using total consumption in the province as the weight. Data for Tibet is missing. Average prices (revenue per kWh) are calculated in this way for all consumers and for households. The industry price is then calculated as the average for all categories other than households and agriculture (assumed to pay household prices) using 2007 national electricity consumption data by sector from National Bureau of Statistics of China (2010) for 2007. Figure 3.9: Coal fuel costs are squeezing margins in the electricity sector Yuan 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 power purchase price (per kWh) cost of coal (per kWh) 0.1 coal price (per ton/1000) household selling price (per kWh) industry selling price (per kWh) 0 jan-jun, jul-dec, jan-jun, jul-dec, jan-jun, jul-dec, jan-jun, jul-dec, jan-jun, jul-dec, jan-jun, jul-dec, jan-jun, jul-dec, jan-may, 03 03 04 04 05 05 06 06 07 07 08 08 09 09 10 Average coal price (yuan/ton divided by 1000), estimated coal fuel cost, average generation price, and electricity selling price for industry and households (all Yuan/kWh), all of China, 2006-2010 Notes and sources: Current prices used. For household and industry prices see Figure 3.8. For coal prices, see Figure 3.7. For assumptions on coal efficiency, see Zhao (2008) and China Electricity Council (2010). Values for 2008 and 2009 are interpolated. For the (wholesale or generation) power purchase price, information is used from SERC (2006-2009) and public information on tariff increases, averaged across provinces. 45 3.18. Electricity subsidies in Indonesia are larger still. IEA (2008b, p.176) reports that in 2008 the average revenue received by Indonesia`s electricity utility, PLN, was USD 6 cents per kilowatt hour, while costs on average were about USD 12 cents per kilowatt hour. Total electricity supply estimates for 2008 of 140 TWh imply a total subsidy of $97 billion or 1.4% of GDP. The persistence of Indonesia`s electricity subsidies (documented as far back as McCawley, 1970) is illustrated in Figure 3.10, which uses average developed economy electricity prices as a benchmark. Note that electricity prices for industry in Indonesia are comparable with those in developed economies, except in the aftermath of the Asian Financial Crisis when the Indonesian currency collapsed. However prices for households have been consistently and increasingly below those in developed economies, and are now only about half of developed country levels. Figure 3.10: Indonesia's industry pays close to prices prevailing in developed countries for their electricity, but its households pay only half. Electricity selling prices for industry and households for Indonesia and the developed country average, 1979-2007 Note: The developed country average is the unweighted average. Electricity prices converted to USD using current market exchange rates. No adjustment for inflation is made. Sources: IEA (2010b), World Bank (2010a), IMF (2010a) 3.19. Cross-subsidies are common in the electricity sector in developing countries, both by sector and by region. In an undistorted electricity market, residential prices will exceed industrial ones. But developing countries tend to cross-subsidize households by pushing industrial tariffs up. This can be seen from Figure 3.11, which shows that on average the ratio of industrial to household electricity prices increases across APEC with income per capita.29 Regional cross-subsidies are also common. For example nationally uniform tariffs are in place in Indonesia, but the costs of its different regional grids differ by a factor of close to 3 (Figure 3.12). This can lead to strong disincentives to expand or even sustain supply in loss-making regions. 29 We could not obtain this data for Vietnam. But the World Bank (2010c, p. 16) estimates that the industrial and commercial sectors provide cross-subsidies of $370 million to residential consumers. 46 Figure 3.11: Developing economies show a lower ratio of industrial to household tariffs than developed ones. The ratio of industrial to household electricity selling prices in 2007 plotted against income per capita in 2008 Source: IEA (2010b), World Bank (2010a), IMF (2010a), own calculations for China from Figure 3.8 (prices for 2008) Figure 3.12: Tariffs are uniform across Indonesia's regions but costs vary by a factor of almost three Electricity sales revenue and supply costs in Indonesia by region (in US cents per kWh) Notes: The source does not provide the year for this data but it is presumably not long before 2005. Source: World Bank (2005) 47 3.20. The net result of all these subsidies and cross-subsidies is that industrial electricity prices in APEC developing economies are slightly below the developed economy average and household electricity prices significantly below. The comparison for 2007 is 7.9 and 7.8 USc/kWh for industries in developed and developing economies respectively, and 12.3 v 7.8 USc/kWh for households.30 (Figure 3.13) illustrates for individual economies. Figure 3.13: Low electricity prices in APEC developing economies manifest themselves in low prices for households not industry Average electricity selling prices for industry and households for APEC economies in 2007 plotted against GDP per capita for the same year. Notes: Electricity prices converted to USD using market exchange rates, GDP using PPPs. Sources: IEA (2010b), World Bank (2010a), IMF (2010a), own calculations for China from Figure 3.8 (for 2008). 3.21. Indonesias subsidies are exceptional for their size, but not for their presence. In both Vietnam and China, the electricity utilities and petroleum companies normally return a profit, and are not usually reliant on budget support.31 By contrast, Indonesia`s petroleum and electricity companies could not survive without budget support. Indonesia`s government has budgeted for subsidies for petroleum products since the 1970s, and for electricity for the last decade or so (Figure 3.14). The average energy subsidy since 2000 is 16% of total spending, roughly equal to government spending on education. While Indonesia is an extreme case, the preceding analysis suggests that petroleum, coal and/or electricity subsidies, even if they are not on budget are significant in both China and Vietnam, and that electricity subsidies have re-emerged in China. 30 Comparison of unweighted averages between countries shown in Figure 3.13 using categorization of countries as per Figure 1.1, excluding Russia for which only the industrial selling price is available. 31 This is not to say that these countries never provide budgetary support. Kojima (2009, p.3) reports that The government of China paid partial compensation to two refiners between 2005 and 2008, with the largest compensation paid in the latter year. The country`s biggest refiner, Sinopec, alone was paid $7.5 billion in 2008, but the two refiners still suffered a combined loss of more than $20 billion. 48 Figure 3.14: Indonesia's energy subsidies have become a major and persistent claim on the budget % central government expenditure 30 Not-specified 25 Electricity Fuel 20 15 10 5 0 01 04 07 10 8 1 4 7 0 3 6 9 97 98 98 98 99 99 99 99 20 20 20 20 /1 /1 /1 /1 /1 /1 /1 /1 77 80 83 86 89 92 95 98 19 19 19 19 19 19 19 19 Central government spending on energy subsidies (decomposed into electricity and fuel, where possible) as a share of total central government spending. Notes: In some years, no decomposition is available for energy subsidies into electricity and fuel. In these years, the subsidies are recorded in the figure under non-specified` energy subsidies. Source: Indonesia budget documents. 3.2.5 Price setting 3.22. Price-setting in the energy sector in developing economies is often ad hoc, and therefore politicized. In the petroleum sector, prices can either be set by the market, or by the government in an automatic manner (using an established formula or procedure), or by the government in an ad hoc manner (i.e., without an established formula or procedure). All OECD developed economies have liberalized domestic petroleum markets. According to a sample of 44 developing economies undertaken by the IMF (2007), 15 have liberalized domestic petroleum markets, 8 have automatic price-setting by government, and 21, almost half, have ad hoc price setting by government. APEC member developing economies have a mix of mechanisms in place. Price setting for petroleum is liberalized in Russia and the Philippines, but an ad hoc approach by governments is still used in all three of the case-study economies, despite various reform efforts. China announced it was moving to market-based pricing in December 2008, but in May 2009 the government announced that it would set prices to protect consumers when world oil prices exceed $80 a barrel (Kojima, 2009, p.4). As a result, China`s petrol prices track world prices closely, but not when world prices spiked in 2008 (Figure 3.15). Similarly, the government of Vietnam was to move to a market-based mechanism in April 2007 but postponed against the backdrop of rising oil prices. When oil prices started to fall again, the Ministry of Finance announced in September 2008 that a newly established committee would monitor oil product prices and approve requests for price changes submitted by importers. (Kojima, 2009, p.27)32 32 Although Vietnam exports crude, it imports refined oil. 49 Figure 3.15: Petrol prices in China follow world prices except when world prices are very high CNY per litre 10 9 8 7 6 5 4 3 cost-covering retail price 2 actual retail price 1 0 2004.1 2004.7 2005.1 2005.7 2006.1 2006.7 2007.1 2007.7 2008.1 2008.7 2009.1 2009.7 2010.1 2003.1 2003.7 Actual retail prices for petrol in China, and what they would need to be to cover costs given world prices for crude. Note: Cost-covering retail prices are based on the world price for crude and include margins for refining and distribution. In January 2009, China increased the fuel tax from 0.2 CNY/liter to 1 CNY. The cost-covering retail price does not include taxes. The price is for Beijing, and for 93-octane gasoline. Source: Li (2009), updated. 3.23. In the electricity sector, some regulation of prices will always be required, though some economies allow full pass through of generation costs. While some developing economy governments have attempted to introduce independent regulators and pass-through formulae to make price setting less ad hoc and political, price setting is still a discretionary process in many economies. No survey is available, but in all three of the case-study economies, price setting in the electricity sector is ad hoc or politicized. Indonesia attempted to introduce independent regulation but this was struck down by the courts as unconstitutional. Price-setting is now with government, and without the guidance of any formulae, and tariff increases have in fact to be approved by Parliament. Not surprisingly, they are few and far-between: Indonesias last two electricity tariff increases were in 2004 and 2010. In Vietnam, electricity prices are promulgated by decision of the Prime Minister, following a protracted and unclear process of negotiation between EVN [the dominant vertically-integrated monopoly] and the government and then within the government itself. (World Bank, 2010c, p.16).33 China has a formula in place for adjusting the electricity price every six months if the coal price changes by more than 5%. However, since the end of 2004, when the formula was introduced, although this condition has been met 10 out of 12 times (in relation to coal market prices), the price of electricity has only been changed thrice, and by much less than the formula mandated, as the previous sub-section showed. 3.24. Ad hoc price setting mechanisms make cost pass-through less likely. IMF (2008) finds that in the period between 2003 and 2007 ­ during which international oil prices fluctuated and rose significantly (Figure 1.11) ­ economies relying on ad hoc petroleum price settings on average passed on 40-70% of the 33 Recent reforms will allow the relevant Minister to approve tariff increases of less than 5%. 50 rise in international prices (depending on the product: gasoline, diesel or kerosene), whereas those with automatic mechanisms or liberalized pricing passed on 100% of the increase or more.34 3.25. The mixed ability of APEC governments to pass energy costs through to their economies is illustrated by Figure 3.16 which plots domestic petrol prices over the course of the recent commodity boom and bust. US retail prices are used as a benchmark. The star performer is Vietnam which started in 2002 with petrol prices at just $0.03 per liter, but managed to reach price parity with the United States despite rapidly rising crude prices. The worst performers are Mexico which started with petrol prices above US levels and ended with them significantly below (and largely unchanged in nominal terms), and Malaysia where prices have collapsed from parity with the US to half of US levels. China and Thailand have both bought their prices to above those of the US, whereas Indonesia, despite its subsidy removal efforts, still struggles to get prices to US levels. Philippines, with liberalized prices, has tracked US prices over this volatile period. Figure 3.16: Petrol prices in APEC's developing economies have diverged over the last decade Petrol pump prices in current USD for the United States and various APEC developing economies. Source: Coady et al. (2010) database www.imf.org/external/pubs/ft/spn/2010/data/spn1005.csv 3.2.6 Energy rationing 3.26. Sometimes government control of prices leads to petroleum fuel shortages. Kojima (2009) notes that shortages have occurred in recent years in 8 of the 49 developing economies she studies, including Thailand (for LPG) and China in APEC. 3.27. Power rationing is common in developing economies. Not only is it difficult to keep up with rapid growth in electricity consumption, but the prevalence of subsidies often means that utilities are in poor financial shape and sometimes lack incentives to expand or restore supply. Systematic data is hard to come by, but power rationing is a problem in all the three of the case-study economies. In Indonesia, the current electricity network suffers from blackouts, brown-outs and enforced supply cuts (IEA, 2008b). Unserved demand is estimated at 4,000 MW compared to an installed grid capacity of 28,000 MW (IEA, 2008b, pp. 177-179). 15 of Indonesia`s 26 grids are reported to be in deficit, with customers being disconnected because demand exceeds supply (Jakarta Updates, 2010). 34 The presence of ad valorem taxes gives rise to the possibility of market pricing giving rise to more than 100% pass-through of crude oil costs. 51 3.28. In Vietnam, the World Bank (2010c, p.12) reports that high demand growth "has eroded reserve margins to the point where power shortages leading to load shedding have been an intermittent problem since 2005 especially in dry years. Current estimates are that Vietnam could face a shortage of about 1,200MW of capacity in 2010, again about 10% of installed capacity. The seriousness of the problem is indicated by the fact that Vietnam`s goal is to push up its reserve margin (the excess of maximum generating capacity over peak demand, expressed as a ratio) up only to 10%, well below the international standard of 25% (World Bank, 2010c). The state-owned utility, EVN, predicts severe power shortages for the next few years (VOVNews, 2009) 3.29. China suffers from intermittent power shortages. Wang, Qui and Kuang (2009) describe two severe electricity shortages in China. In 2004, a total of 24 provincial areas experienced power brownouts, the power deficit amounting to 10% of the installed capacity. China suffered another power shortage in 2008, a total of 19 provincial areas experiencing power brownouts. More recently, the Chinese Government reported that high demand resulting from harsh weather conditions led to power rationing in 13 provinces in January 2010. Press reports indicate that industrial consumers also had their gas supply rationed in some provinces. The reason behind the more recent electricity shortages in China lies in low contract prices for coal sales to generators, which give incentives to coal-suppliers to divert coal to the open market (Wang et al., 2009). There is also evidence that, when coal prices increased, and electricity prices didn`t, some generators withdrew from the market rather than selling at a loss (Wang et al., 2009). 3.30. Globally, in low-income economies, more businesses (almost 30% on average) rate electricity as their most pressing concern than they do any other of the ten possible business environment constraints they were surveyed about, and by a considerable margin.35 For lower-middle income it is equal first (about 13% on average, tied with concerns over access to finance). Electricity is less of a concern for business in upper-middle income economies, but still ranks in the top half of constraints.36 3.2.7 Captive power 3.31. Captive power generation is common in some developing economies. Unreliable power supply and high tariffs due to cross-subsidies often drive industry in developing economies towards self-reliance through captive electricity generation. Accurate estimates of captive power are not surprisingly hard to come by. In Indonesia, captive power capacity in 2004 was estimated at 15,000 MW37 not much below PLN capacity of 21,000 MW for the same year (IEA, 2008, p.178). Captive power is much less in evidence in China, perhaps because rationing is only intermittent and the quality of supply is good. Vietnam has also seen growth in captive power, as indicated by the fact that the government has put in place a scheme whereby captive generators can sell into the grid.38 3.2.8 Flexibility in dispatch 3.32. Flexibility in dispatch is often limited in developing economies. One reason for this is weaknesses in transmission grids. The Chinese power system consists of seven interconnected grids, with limited interconnectivity. Zhongfu et al. (2008) note of China that The relatively low investment in power 35 In the case of Vietnam, although electricity doesn`t rate as one of the top concerns on www.enterprisesurveys.org, other informed observers suggested that it was a major constraint. A recent forum of Chinese investors into Vietnam singled out inadequate electricity supplies and infrastructure as major challenges facing them (Chinese investors voice concerns at forum, Vietnam News, p. 15, 17 July 2010) 36 See www.enterprisesurveys.org. 37 http://www.nema.org/gov/trade/briefs/indo_elecsurvey.pdf 38 ...Hiep Phuoc, Bourbon Tay Ninh and Formosa are examples of successful 'captive' power production schemes, allowing private companies to generate their own electricity for use in their factories or industrial zones and to sell excess output to EVN. http://www.eurochinacom.eu/culture-region/current-publications/infrastructure-fdi-vietnam/ 52 transmission and distribution has led to a lagged grid system with weak major grid network and disconnected regional coordination.39 In Vietnam, according to the World Bank, there are limits to the transfer capacity of the 500kV system connecting the north and south of the country. (2010c, p.36), and the grid system is more accurately described as three weakly-connected separate systems. Indonesia`s archipelagic geography poses particular challenges, and its system in fact consists of one large and 25 small unconnected grids. The main grid serves the main island Java and the nearby island of Bali. [T]he backbone of the transmission grid across Java requires urgent upgrading to enable it to transport power across the island as well as to improve system security (IEA, 2008b, p.179). 3.33. Limits to dispatch flexibility can also be policy-induced. In China, all generators sell to the grid, but all payments to generators are based solely on dispatch. Unlike in developed country power markets, no separate payments are made for providing capacity, as important a service as this is (for stand-by reasons). This policy constraint has led to a quota system for dispatch, whereby each generator is given a proportion of total forecasted demand, perhaps linked to total capacity, and any deviation between actual and forecast demand is shared across all generators. Such a system is persisted with despite its obvious economic inefficiencies because, within the constraint of making all payments based on dispatch, it ensures that even the less efficient generators are viable, and so gives them an incentive not to close down but to provide spare capacity. (See Section 4.1.2 and Mercados (2010) for further discussion of attempts to reform this system.) 3.34. A similar, and final, constraint on dispatch is the presence of long-term take-or-pay contracts under which private generators (independent power producers) are guaranteed dispatch, regardless of merit order considerations. 3.2.9 Dominance of the sector by vertically-integrated state-owned enterprises. 3.35. The energy sector in developing economies is dominated by vertically-integrated state-owned companies. The focus below is on the electricity sector, but the point holds more broadly. As Table 3.3 shows, the dominant model among APEC developing economies is the vertically integrated (VI) monopoly with a single state-owned company with responsibility for generation, transmission and distribution, supplemented by independent (private) power producers, who run power stations and sell electricity to the VI monopoly. China has separated generation from transmission and distribution to create a single-buyer model. A bidding system is prescribed for generators, but is not functioning. The major generation and grid companies are all state-owned. Some APEC economies have moved on to implement third-party access or even a wholesale power market, but they are a minority. Developed economies have a mix of sectoral structures, but are much less dominated by state-owned monopolies, except for Canada. 39 Our translation. 53 Table 3.3: Power sector structure and ownership in APEC economies Vertically integrated Single monopoly (VI) VI + IPPs buyer Third party access Power market Largely Indonesia China Australia (some states) government Malaysia owned Mexico Vietnam PNG Taiwan Thailand Largely Japan Russia Australia (some states) private- US (some states) Peru owned Chile US (some states) NZ Mixed ownership Singapore Korea Philippines Notes: IPPs are independent power producers who sell power either to the vertically integrated monopoly or to the single buyer. An economy in which there is, say, one generation/transmission company and one distribution company falls into the VI model. A single buyer structure is one in which there are several generators and several distributors, but all generators sell to and all distributors buy from a single buyer. Under third party access there is still a single buyer, but the single buyer has lost its monopoly. Under a power market, there is no single buyer. Note that Malaysia`s electricity company has been partially privatized, though the government is still the majority-owner. Source: Besant-Jones (2006), own research. 3.2.10 Reliance on central planning to guide generation expansion 3.36. Investment decisions in the electricity sector in developing economies are often made through a central planning process. In an increasing number of developed economies, investment decisions in the power sector are decentralized. They are the responsibility of private companies and are made on commercial lines. In many developing economies, especially where the power sector is dominated by a small number of government-owned companies, it is not surprising that investment decisions are often coordinated through sectoral plans. In the last decade, Indonesia has announced two generation expansion plans called crash programs.` The first announced in 2004 targeted 10 000 MW of coal-fired generation. The second announced in mid-2009 involved another 10 000 MW, but with more diversified fuel sources, including 12% from hydropower and 48% from geothermal (Ölz and Beerepoot, 2010). China and Vietnam also expand their generation sector using central planning with targets for different generation types. As Wu, Wen and Duan (2004, p.4) write in relation to China: Although the generation sector has already been separated from the utilities in China, investment and construction of new power plants are still under strict control of the government. Private-investment is allowed, but under expansion plans laid down by the central and provincial governments.40 The Chinese government has 2020 capacity targets for all major generation types. They imply not only a rapid aggregate expansion, but significant continued diversification away from coal to gas, nuclear and wind (Figure 3.17). 40 We do not suggest China runs a pure and monolithic planning system for electricity. The reality is no doubt much more complex. There have been significant reforms in the structure of the power sector. But these have not ended central planning. As Zhang and Heller (2004, p.35) write : Although Beijing no longer aims to control how many restaurants will emerge in the next five years, it does see a need to continue to plan and decide how many power plants to build, where to site them, what fuel they should tap and what prices they will charge. As a result, instead of partially withdrawing from business, the government merely switched its role from directly controlling the power industry via repatriation of all revenues and direction by ministerial fiat to indirectly controlling utility SOEs` access to financial markets and project approval. SOEs in the power sector are not substantially more independent than they were before the reform in terms of power project development. 54 Figure 3.17: China's 2020 generation targets aim to reduce the dominance of coal-fired generation, while doubling total capacity. Total generation capacity (GW) Generation capacity mix (%) 1800 100% 1600 90% 1400 80% 1200 1000 70% 800 Non-hydro renewable 60% Nuclear 600 Hydro Natural gas 400 50% Oil Coal 200 40% 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 0 1980 1985 1990 1995 2000 2005 2010 2015 2020 Electricity generation capacity by fuel type (%) in China, historical (1980-2009) and projected (2020) Notes and sources: Target year is 2020; last year of historical data is 2009; for intermediate years, linear interpolation used. Historical capacity data comes from EIA (2010). However, this source doesn`t distinguish between coal, oil and gas capacity. The subdivision of thermal capacity into these three types is done using electricity generation data from IEA (2009d), up to 2006 and own sources for 2009 (2007 and 2008 are interpolated). Note vertical axis truncated from below at 40% to magnify the changes envisaged. At the time of writing, there was no public, up-to-date and comprehensive generation expansion plan. Targets were therefore compiled from various sources. (The 12th Five Year Plan, due shortly, is expected to contain an updated generation expansion plan.) 3.2.11 Lack of commercial orientation 3.37. Often energy companies in developing economies do not run on commercial lines. Reliance on central planning does not in itself signify a lack of commercial orientation. However, the relationship between the energy sector and the government in developing economies is often one in which credible commitments are very difficult to make. 3.38. The credibility problem in the energy sector in some developing economies goes further than a soft-budget constraint. In some economies, public enterprises operate in a soft budget constraint` environment, that is they know they will be bailed out if they fail (Kornai, Maskin and Roland, 2003). Energy sector companies operating under politicized pricing, while confident that the government will not let them go broke, often have no confidence that governments will make them whole either. They are aware of the risk (or sometimes the reality) that prices will not match costs, and also know that in such cases governments will be reluctant to put all (or sometimes any) of the resulting subsidy on budget, and will rather try to push as much as that burden as possible back on to the utility. Universal Service Obligation payments, or indeed any kind of explicit subsidy formulae, are rare in developing economies. Governments worry not only that they will not be able to afford the resulting subsidy bill, but also that any automatic subsidy payment will simply encourage inefficiency and a cost-pass-through mentality on the part of the utility. 3.39. As mentioned earlier, of the three case-study countries, Indonesia is the only one where the government provides consistent budget support. Both the timing and the amount of the subsidy it receives are uncertain. Arrears often build up between PLN, the electricity company, and Pertamina, the 55 oil company, as both struggle to survive with below-cost tariffs.41 Commercial discipline is clearly higher in China and Vietnam, where arrears and government subsidies are rarer. However, the presence of politicized pricing and soft-budget constraints in both China and Vietnam (see Li (2008) and Ito (2006) for the case of China) means that commercial discipline in the energy sector in these two countries probably lags behind the level in developed countries. 3.40. In some developing economies, though not generally in APEC ones, the lack of commercial discipline extends to customers as well. In India, for example, it can be difficult to disconnect households for non-payment, and illegal connections are rampant. 3.2.12 Difficulty of reform 3.41. Energy sector reforms in developing economies are difficult, and often make slow progress. Most developing economies are aware of the weaknesses of their energy sector, and have been attempting to reform. But reforms are not easy. While there are some success stories, a World Bank (Besant-Jones, 2006) review of power sector reforms concludes that overall political forces are difficult to align for reform (p.14), that interest groups constitute a major impediment to reform (p.16), and that successful reform requires sustained political commitment. (p.2) Not surprisingly therefore, Power market reforms in developing economies are generally tentative and incomplete, and are still works in progress. (p.4). Power sector privatization is particularly challenging. Most privatization-focused power sector reforms in developing economies have stalled, and some have been abandoned in all but name (Rosenzweig, Voll and Pabon-Agulado, 2004, p.16). 3.42. Indonesia has certainly found power-sector reforms very hard going. Consideration began in the mid-90s. In 1998, a power sector reform strategy was launched and in 2002 a new electricity law was passed, providing for full cost recovery, the development of an electricity market with private players, the unbundling of the utility, PLN, and the introduction of an independent electricity regulator. However, the law was annulled by Indonesia`s constitutional court in 2004, which ruled that the new law was in conflict with the provisions in Indonesia`s constitution which required that the state control vital sectors. 42 Today, once again, PLN is the single provider for the provision of electricity. Private participation is limited to generation. A new Energy Law (No.15/2007) has been put in place, but it does not address the core issues of restructuring, market competition, tariffs and independent regulation. 3.43. China has also made slow progress with electricity reform. In 2002, China split its single, vertically integrated utility into two grid companies (a large one covering most of the country, and a small one in the south) and a number of generation companies (including five large ones). It experimented with wholesale electricity markets in 2002, but that was short-lived and generators no longer bid for dispatch, but sell at centrally-fixed prices. China also established in 2002 a State Electricity Regulatory Commission, but it focuses on technical rather than economic regulation. Prices are still set by government (though the SERC can offer its advice) and, as noted earlier, mechanisms for cost pass- through have been established but are not used. Central planning is still used to guide generation expansion The IEA`s conclusion that China is caught between the old planning mechanisms and a new approach (2006b, p.16) is probably as relevant today as when it was written. 3.44. Vietnam has just embraced power sector reforms, and is planning to move to a single-buyer model as a first step. 41 See, for example, http://jakarta.usembassy.gov/econ/energy_highlight_jan07.html 42 Article 33.2 of the Indonesian Constitution reads Production sectors that are vital to the state and that affect the livelihood of a considerable part of the population are to be controlled by the state. http://en.wikisource.org/wiki/Constitution_of_the_Republic_of_Indonesia In their verdict, the Court took the view that control meant to regulate, facilitate, and operate. (IEA, 2008b, p.33) 56 3.3 Other relevant features of some developing economies 3.45. This chapter has so far focused on the energy sector. This last section turns to three other broader aspects of developing economies which, as argued in Chapter 4, are also relevant to policy choices in relation to climate change. 3.3.1 Factor market distortions 3.46. Developing economies sometimes have distorted factor markets. Both China and Vietnam have spectacularly high and, especially in the case of Vietnam, rapidly rising levels of investment (Figure 3.18). Indonesia`s investment levels were high, but investment collapsed with the Asian financial crisis of the late 1990s, and have only partially recovered. Figure 3.18: Surging investment in China and Vietnam Investment/GDP ratios for China, Vietnam and Indonesia compared to the average for APEC developed economies Notes: Developed` is the unweighted average for investment/GDP for Japan, US, Canada, Australia and NZ. Source: World Bank (2010a) 3.47. Huang (2010) explains Chinas high investment rate by his characterization of it as a country which has liberalized its product markets but not its factor markets. Limited liberalization of the labour, land, financial and energy sector markets have repressed wages, land prices, interest rates and, as analyzed earlier, energy prices. Low interest rates and land and energy prices encourage capital-intensive production. Low wages should push in the other direction, but Huang argues that the lack of social security increases savings which further pushes down the cost of capital. Vietnam has traced a similar path of reform away from central-planning, and its financial sector is similarly state-dominated. The Government also dominates land transactions at least outside of agriculture. In Indonesia, there is less government dominance of both the financial sector and land. Indeed, land can be difficult to obtain for development. Note, for example, the difficulties Indonesia`s electricity company is having in obtaining land to upgrade its transmission network (IEA, 2008). 57 3.48. To investigate the problem of factor market distortions further, this chapter considers the financial sector and uses the IMF database of financial liberalization created by Abiad, Detragiache, and Tressel (2008). Figure 3.19 shows there is not a tight link between financial repression and GDP per capita across economies in general. But note the very low score (indicating high repression) for Vietnam and China, the two APEC economies with the highest energy growth for 2000-2007 (Figure 3.1).43 3.49. In some developing economies, financial repression has been a drag on growth, but moderate financial repression has also traditionally been part of the so-called ,,East Asia Miracle (World Bank, 2003). The argument here is that financial repression along with other distortions, such as in the land market, has pushed up investment and thus energy use both via accelerating economic growth and by making that growth more capital and therefore energy intensive (see also Rosen and Housen, 2007, who argue this for China). Figure 3.19: Some developing economies are characterized by financial repression Plot of an index of financial liberalization (0 to 12) against GDP per capita, 2005 Notes and sources: The graph uses four of the variables provided by Abiad, Detragiache, and Tressel (2008) in the database published on the web with their working paper. Each is scored by them on an index from 0 to 3, where a lower score indicates more financial repression, or less liberalization. Thus the highest aggregate score possible is 12 and the lowest zero. The four variables are: the extent of credit controls (measured by reliance on directed credit and credit ceilings), the extent of interest rate controls, the degree of state ownership of banks, and the level of barriers to entry faced by new banks. 3.3.2 Compensation instruments 3.50. Developing economies have fewer instruments to compensate poor households for energy price increases. The poor in developed economies either receive welfare payments and/or pay taxes. Not so in developing economies. This makes compensation of poor households for energy price increases more difficult in developing economies. That said, a few developing economies have shown that compensation packages are possible. Indonesia provided temporary cash handouts to some 20 million poor families in 2005 and 2008 to help offset the rise in fuel and food prices. Families were selected on the basis of a number of observable characteristics (not income), and payments were made through the post office (in cash). Other economies have used a range of instruments, including government expenditure on services that are important to the poor (e.g., spending on schools), offsetting price changes (e.g., the introduction/strengthening of lifeline tariffs), and the promotion of alternatives (e.g., promoting LPG when kerosene subsidies are reduced) (IMF, 2010b). Compensation can be more difficult in federations, such as China, where existing welfare payments are provincial rather than central responsibilities. 43 See Du, Tao and Lu (2008) for more on financial repression in China. 58 3.3.3 Institutional capacity 3.51. Developing economies generally have lower levels of regulatory quality and government effectiveness, and higher levels of corruption. The World Bank (2010e) releases cross-country indicators of governance measured along six dimensions using surveys of firms, households and analysts. Three of these seem of particular relevance to the challenge of effective environmental regulation: regulatory quality (which measures the ability of the government to formulate and implement sound policies and regulations that permit and promote private sector development), government effectiveness (which measures the quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government`s commitment to such policies), and control of corruption (which measures the extent to which public power is exercised for private gain, including petty and grand forms of corruption, as well as capture of the state by elites and private interests). 3.52. These three governance indicators against GDP per capita for APEC member economies are plotted in Figure 3.20. Not surprisingly, developed economies do significantly better in all three dimensions. Figure 3.20: Poorer economies tend to have lower government effectiveness, regulatory quality and control of corruption Three dimensions of governance plotted against GDP per capita in 2008 for APEC economies Note: The vertical axis measures percentile rankings, where a higher ranking indicates a higher score. Source: World Bank (2009a) 3.4 Conclusion 3.53. Table 3.4 summarizes the long preceding discussion for the three case-study economies and for a typical developed economy. Developed economies show little variation in their characteristics, so in most areas generalizations can be safely made, and the table notes where this is not the case. A rating scale of low`, moderate` and high` is used, and characteristics are worded so that the typical developed economy scores a low` for most of the 15 characteristics. 3.54. The simple message from Table 3.4 is that developing economies do tend to differ in some important ways both with respect to the energy sector and more broadly. The next chapter turns to the implications of all this for climate change policy. 59 Table 3.4: Characteristics of the energy sector and the broader economy of developing economies (China, Vietnam, and Indonesia) and a typical developed economy. Typical Developing economies developed economy China Vietnam Indonesia Characteristics relating to the energy sector (with emphasis on the power sector) 1. Rate of energy growth Low High High High 2. Importance of traditional energy Low Moderate Moderate Low sector 3. Likelihood that energy expenditure Low High High High share rises with consumption 4. Presence of subsidies Low Moderate Moderate High 5. Degree of political discretion in price Low Moderate Moderate High setting 6. Degree of rationing Low Moderate High High 7. Reliance on captive power Low Low Moderate High 8. Constraints on flexibility in dispatch Low High Moderate Moderate 9. Dominance by state-owned Low with High High High vertically-integrated utilities some exceptions 10. Reliance on central planning in the Low with High High High electricity sector some exceptions 11. Divergence from commercial Low Moderate Moderate High orientation 12. Political difficulty of reform Mixed Moderate N/A (Just High starting) More general characteristics 1. Distortions in factor markets (as Low High High Moderate indicated by degree of financial repression) 2. Degree of difficulty to find Low Moderate Moderate Moderate instruments to compensate low-income households for price changes 3. Institutional weaknesses relating to Low Moderate Moderate High quality of regulation, levels of government effectiveness, and absence of corruption 60 Chapter 4 Choices: mitigation policies for developing economies 4.1. This chapter builds on the foundation of the last three and brings this part of the report, focused on mitigation, to a close. It asks: given the climate change mitigation and related policy objectives of developing economies (Chapter 1), given the fiscal policy instruments available to them and the experience with these instruments to date in both developed and developing economies (Chapter 2), and given the particular features of their energy sectors and broader economies (Chapter 3), what policy choices should they make? 4.2. The question is an ambitious one, and the aim here is only to derive some tentative conclusions and point in some new directions. Section 4.1 focuses on carbon pricing, and then Section 4.2 looks at other technology-based fiscal policies. Section 4.3 concludes. 4.1 Implications for carbon pricing 4.3. So far the focus of the literature, in so far as it has considered the application of carbon pricing to developing economies, has been on the implications of limited institutional capacity (covered in 3.2.13) for the choice of instruments, and in particular on the choice of price-based versus quantity- based approaches to carbon pricing. Thus, SEI (2010, p.26) advocate a carbon tax over an emissions trading system (ETS) in China on the grounds that a cap-and-trade system [or ETS] requires effective administration as well as a mature legal system. Emissions must be measured and monitored, individual emitters identified and charged correctly, and trading of permits requires a well-functioning market with a high degree of sophistication. Likewise, Indonesia`s Ministry of Finance argues for a carbon tax initially, on the basis that it is potentially simpler to set up and administer, since it can be collected through existing taxation institutions. (2009, p. 27). The Ministry argues that if coverage is limited, as it might be initially, any permit market could be manipulated. It suggests that the tax be subsequently replaced by emissions trading when carbon measurement and accounting systems are sufficiently developed and the number of market players enlarged (p.6). World Bank (2010d, Box 6.3) also argues that a carbon tax would impose lower monitoring and institutional costs on developing economies. While these are all valid arguments, the analysis of Chapter 3 suggests that limited institutional capacity is only one of 15 relevant ways in which developing economies can differ from developed economies. This chapter does not take a position on whether developing economies should introduce a carbon tax or an emissions trading scheme, but rather argues that, taken together, the features outlined in Chapter 3 have more profound implications for carbon pricing in developing economies than have been recognized in the literature to date. 4.4. What are the implications of the characteristics outlined in the previous chapter for the desirability and utility of carbon pricing? This section attempts to answer the following four questions, all posed for developing economies: Is carbon pricing desirable? Is carbon pricing feasible? How important are and what are the mitigation implications of energy sector reforms? How important are broader economic reforms for mitigation? 61 4.1.1 Is carbon pricing desirable? 4.5. Economic theory suggests that carbon pricing should be the first instrument countries with an emissions reduction target turn to. A carbon price is an efficient instrument for reducing emissions, and one which minimizes the extent of government intervention in the economy. However, recall that most developing economies have multiple goals. They want to reduce emissions, but also control local pollution, promote energy security, and pursue competitive advantage with new energy technologies. Given all these goals, does carbon pricing still make sense? 4.6. It has been suggested that carbon pricing will threaten the transition of households into the modern energy sector, including their exit from biomass reliance. Recall that APEC is still home to over 100 million people yet to have access to electricity and one billion still reliant on traditional fuels and coal for cooking and heating (Section 3.2.2). Reducing these numbers should be an urgent policy priority, and is in many economies. But carbon pricing on its own will tend to increase the prices of electricity and fossil fuels (kerosene and gas) which households need to switch to if they want to reduce reliance on biomass (still the main cooking fuel for 600 million APEC citizens). Is this a reason not to introduce carbon pricing in low-income economies, as Somanathan (2008) argues? 4.7. This problem is relatively easy to solve in the case of electricity, but more difficult for biomass. To promote electrification, subsidies can be directed to the capital costs of connections and through lifeline tariffs. Weaning households from using biomass for cooking and heating is more difficult, as is borne out by the slow progress many economies have made and are projected to make with regard to this problem (Section 3.2.2). Experience suggests that subsidized connection programs (e.g., giving out free or subsidized gas tanks and stoves) on their own may not induce switching (Wadhwa et al., 2003). This presents a case for recurrent cost subsidies. But kerosene and gas subsidy programs are notorious for high leakage to the non-poor, precisely because it is impossible, or at least very difficult, to develop a lifeline- type scheme which provides only a minimum quantity at a subsidized price. A review for India, based on broader international experience, concludes that subsidies have been found to be ineffective in expanding the uptake of LPG and kerosene as primary household fuels among the poor and that No effective subsidy mechanism for kerosene or LPG seems to exist.(Wadhwa et al., 2003, p.6). Taking this argument to its logical conclusion, if subsidies are not the answer to the biomass reliance problem, then there should be no argument against extending carbon pricing to gas and kerosene, even if they are biomass substitutes. 4.8. In summary, biomass reliance should not be used as an argument against carbon pricing. Continued reliance on traditional fuels is a serious problem, but the solution does not seem to include abandonment of carbon pricing (and of course carbon pricing of household coal would help discourage it). Indeed, as shown by the projections in Table 3.1, biomass reliance will likely be reduced only very gradually over time. Factors driving continued biomass reliance include the free availability of firewood and dung, the lack of information about health costs, cultural inertia, and the lack of alternatives. Solutions include information campaigns, improved cooking stoves,44 and liberalized energy markets to encourage entrepreneurs to develop market solutions. Though clearly more research is needed (and one compromise could be to neither subsidize nor tax kerosene and LPG), the solution does not appear to lie in subsidies or even in avoiding carbon pricing. 4.9. Another potential drawback of carbon pricing is that, on its own, it might lead to substitution of oil for coal, with adverse implications for energy security. As coal is more emissions-intensive than oil, there is a risk that a carbon price will lead to substitution from coal to oil (Section 1.3). To prevent this, economies would need to place additional taxes on oil or promote renewable or nuclear energy to ensure 44 Though note the finding of Slaski and Thurber (2009, Abstract) that Programs to distribute improved biomass stoves have traditionally been unsuccessful... China`s program is a notable exception to this generalization. 62 that any substitution away from coal is towards renewable or nuclear rather than oil. APEC economies are moving in this direction. As noted in Chapter 1, most have renewable energy targets. China has recently increased its per liter tax on petrol from 0.2 to 1 Yuan. 4.1.2 Is Carbon Pricing Feasible? 4.10. Feasibility has two dimensions, one that could be considered broadly political (influencing the likelihood of introduction and level of any carbon price) and one broadly economic (influencing the impact of any carbon price once introduced). Increasing energy prices is politically difficult in developing economies. The fact that modern energy is a luxury good in developing economies might improve the welfare consequences of energy price increases (Section 3.2.2), but it might also raise the political costs, since it means that any energy price increase will disproportionately hit the rich who are also likely to be the politically powerful.45 The limited availability of compensation instruments (Section 3.3.2) also makes the politics more difficult, though, as noted in that discussion, a few economies have shown that, even with limited instruments, compensation packages can be put together to help poor households adjust to higher energy prices, and ease the political pain. 4.11. The other political barrier to carbon pricing is international rather than domestic. The tardiness of developed economies outside of the European Union to introduce carbon pricing (see Table 2.2) will inevitably bound the ambition of developing economies. If carbon prices are introduced, they will be at a low level: current proposals in Indonesia and China suggest a $5-10 range (Section 2.2). Clearly, the lower the price, the lower the impact. Europe`s carbon price under its ETS has been volatile. At the time of writing, prices are about 14 Euro (about $20). Prices are expected to rise in the coming years, but neither historical nor projected prices have been sufficient to deter Germany from engaging in a rush to coal: Germany currently has some 20-29 GW of coal and lignite generation plants under planning or construction (about 30-40% of current peak demand), and only 3-6 GW of gas (Pahle, 2010). Pahle shows that, at current fuel prices, a carbon price of 40 Euro is needed to favor gas over coal, and argues that investors either don`t believe that carbon prices will reach this level, or are worried that gas prices will continue to rise relative to coal, and so require an even higher carbon price. (Worries about the security of future gas supplies from Russia only add to the preference for coal.) 4.12. That said, it is important to take a long-term perspective. Given the science, climate change is unlikely to go away as an issue. Over time, as climate change becomes more evident, more developed economies will introduce carbon prices, and this political constraint to carbon pricing for developing economies will weaken. 4.13. Even setting aside the question of level, carbon pricing might have a limited impact in developing economies, especially in the electricity sector. Now consider the economic feasibility of carbon pricing. A carbon price on coal would send a strong signal to commercial consumers of coal, such as steel manufacturers. But the features outlined in Chapter 3 might limit the various ways outlined in Section 2.2 in which a carbon price can reduce emissions from transport and especially electricity. The paragraphs following explore this point.46 4.14. Developing economies often lack mechanisms for carbon-pricing cost-pass-through. Given the politicized nature of energy price setting in many developing economies (Section 3.1.4), which frequently leads to subsidies (Section 3.1.3) there is clearly no guarantee that any carbon price will be passed on to consumers. As discussed earlier, most developing economies have not yet liberalized fuel markets, and 45 Interestingly, SEI (2010, p.24) notes that China`s main political concern over any energy-tax reform is that it will be regressive. Our analysis would suggest such a concern is probably misplaced, unless coal used by households for cooking is taxed. 46 They don`t look explicitly at innovation, but the extent to which carbon pricing encourages innovation depends on the extent to which carbon pricing is allowed to influence consumer, producer and investor behaviour, which is the subject of the analysis following. 63 economies with regulated prices have struggled to transmit rising global fuel costs in the last few years into retail fuel prices. Electricity prices cannot be fully liberalized. Generation prices can be, but most developing economies control both generation and retail prices. Some economies are making progress (India has just deregulated retail fuel prices), but others have such elaborate procedures for energy price setting that cost pass-through seems unlikely at best. One thinks here of Indonesia with its requirement that electricity tariff increases be approved by Parliament. The higher the carbon price, the greater the risk of non-pass-through of costs becomes. 4.15. The credibility of any carbon pricing system will suffer if carbon prices are not passed through. If carbon prices are not passed through, then energy prices will not rise, and consumers will not have the incentive to reduce energy use. More than that, if carbon prices are not passed through, then electricity and oil companies will be unable to pay their carbon bill to government. So there will be no offsetting positive fiscal impact from the introduction of carbon pricing. Rather, utilities will look to government to cover the cost of their carbon bill, either through a subsidy payment or through the ongoing free allocation of permits (if a trading system is used rather than a carbon tax). It is hard to predict what the results would be. Given the lack of an arms-length relationship between the government and the energy sector in most developing economies, it is hard to see how a carbon price that isn`t passed through would be taken seriously, a point explored further below in relation to investment decisions. 4.16. Lack of cost-pass-through could also have perverse consequences. Say that cost pass-through isn`t allowed, and that generators are given free permits to compensate for this. Then their incentive will be not to generate, but to sell their permits on the market. The result will be power-shortages, such as China has experienced in recent years on account of coal costs not being fully passed through. If neither permits can be sold, nor costs passed through, then essentially the government will end up regulating the electricity sector not by a price but by its allocation of permits. 4.17. Some developing economies will have limited capacity to respond to the incentives carbon pricing provides to change the dispatch order in the electricity sector. For a carbon price to influence the fuel mix of an electricity sector in the short run (i.e., prior to new investment), economies need to have the capacity to change the dispatch order. This capacity is limited if not absent in the many developing economies which suffer from supply shortages, and which therefore simply dispatch whatever capacity is available (see Section 3.2.6). Some developing economies also have limited dispatch flexibility due to transmission weaknesses, contractual obligations, and policy commitments (Section 3.2.8). 4.18. Considering the impact carbon pricing would have on investments in the electricity sector in many developing economies requires separating out the role of the government and the utilities. As noted in Section 3.2.10, most developing country governments use central planning to determine electricity sector investments. Price signals are crucial for getting decentralized agents to adjust their actions to meet national targets. But central planners can directly incorporate national targets into their decision making, without any price signal at all. China is already trying to boost the share of gas, renewable and nuclear energy, and reduce the share of coal (Figure 3.17). Indonesia is trying to increase the share of geothermal. These targets already incorporate a range of environmental and energy-security goals. For a carbon price to have an impact on the investment decisions it would need to be higher than the implicit carbon price already guiding the planners` move away from coal. Given that any carbon prices that are likely to be introduced will be low, this might be judged unlikely. 4.19. However, it is one thing to plan, and another for that plan to be implemented. To understand the prospects of implementation, attention needs to be shifted from the government to the (usually state- owned) utility. Utilities have significant discretion in implementing plans, given the complexity of constructing new plant or entering into new long-term power purchase contracts. The features of relevance here are the pressures to increase supply to meet rapid growth (Section 3.2.1) and avoid shortages (Section 3.2.6), and the lack of a commercial orientation (Section 3.2.11). The key issue is whether any carbon price (whether explicit or implicit) would be taken seriously or, put differently, 64 whether an investment plan which deviates from least financial cost will be implemented. For this to happen, the carbon price would need to be credible. It was argued above that if there is a significant risk that carbon price will not be passed through, then it will not be credible (i.e., will not be taken seriously). Given that governments often fail to meet their subsidy obligations to energy companies (Section 3.2.11), the latter will not want to take on commitments which only make sense if there is a carbon price if they perceive a high risk that they will have to collect that carbon price not from consumers but from governments. The point here is simply that central planning decisions which significantly deviate from financial realities are unlikely to be implemented.47 And pass through of carbon pricing would be essential to such a divergence. 4.20. The case of Indonesia is instructive in this regard. As noted earlier, Indonesia`s most recent generation expansion plan has a very large geothermal target: some 48% of the total planned expansion. Considering that only 5% of Indonesia`s electricity comes from geothermal, this is clearly a very radical plan. To date, the electricity utility, PLN, has only paid 6c per kWh on average for geothermal, but the average geothermal cost is closer to 12c (Ministry of Finance, Republic of Indonesia, 2009). PLN`s average revenue, however, is only 6c per kWh, and the cost it faces for coal is about 9c per kWh. This makes it unlikely that the geothermal target will be achieved, as PLN will be unwilling for financial reasons to offer a high-enough purchase price to make geothermal attractive. A sizeable carbon price would be needed to make geothermal financially attractive at 12c: a $30 per tonne of CO2 price would approximately push up PLN`s coal fired price to 12c per kWh price. But there would be a significant risk that a significant carbon price would not be passed through. Developers might be willing invest in geothermal if they were given a government guarantee and offered a price of 12c. But, without being sure of its capacity to pass through costs, and unable to get a credible guarantee of budget subsidies in the future, PLN will likely be unwilling to buy geothermal from the developers at this price, even with a high carbon price. In other words, even a high carbon price might not impact on investment decisions if cost pass-through is not assured. 4.21. Quantification of these effects is not possible, or at least beyond this report. Their combined effect could be large, however. Based on the three case-studies, and wider research, the typical developing economy would seem to have politicized energy price setting, power shortages, a limited commercial orientation, and heavy reliance on central planning. A carbon price introduced into such a setting might not be passed through, might have no impact on electricity dispatch, and might have no influence on investment outcomes. 4.22. That the impact of carbon pricing might be limited due to pre-existing distortions is more than a theoretical proposition, as a real-world example from China shows. Carbon pricing in developing economies is an unknown quantity, so any argument about its impact has to be proceed largely on the basis of theorizing. But there is at least one real-world example to draw on. China has not yet introduced a carbon price, but it has tried to introduce a reform to its electricity dispatch system which mimics a carbon price.48 Under the Energy Saving and Emissions Reduction in Power Generation or ESERD pilot introduced into 5 provinces, provinces have been instructed to dispatch generators, not on an across-the- board basis as in the past, but rather according to a mix of economic and environmental criteria. To simplify, the dispatch order is: renewable, nuclear, gas, and then coal, with coal plants ordered by their thermal efficiency, from highest to lowest. Note that this is roughly the order that one would expect with a high-enough carbon price, and, indeed, simulations show implementing ESERD would cut emissions by 10%. However, the pilot provinces have only been able to partially implement this decree, because of the negative financial implications full implementation would have for less-efficient coal-fired units. These units are still valuable as reserve capacity, but, under the Chinese on-grid tariff system, plants only 47 The pressure to increase supply and avoid shortages means that even loss-making utilities will try to expand capacity. They know that the task of expansion is important enough that the government will keep them going. But, as argued in Section 3.2.11, they also know that government will not make them whole, and they will therefore strongly favour financial least-cost expansion. 48 This paragraph draws on Mercados (2010). 65 receive a payment if they are dispatched, and so have no incentive to provide stand-by capacity (see Section 3.2.8). Instead, if not regularly dispatched, they would simply shut down, thereby depriving the system of valuable spare capacity, in case of an emergency or a spike in demand. Or, put differently, the policy-induced lack of flexibility in dispatch has undermined the impact of the introduction of a carbon price (or, in this case, equivalent). 4.1.3 How important are and what are the mitigation implications of energy sector reforms? 4.23. Energy sector reform will heighten the impact of carbon pricing. Many of the various characteristics which, it was suggested above, may limit the impact of carbon pricing are those which energy sector reforms, and power sector reforms in particular, aim to address. Core reforms typically include the establishment of cost-pass-through mechanisms (through price liberalization or the establishment of independent regulators), and unbundling, privatization, and the introduction of competition (to strengthen the commercialization of the sector and harden budget-constraints). These reforms, if successful, would certainly facilitate carbon pricing pass-through and would also make investment and dispatch decisions more price sensitive. To return to the earlier example from China, introducing a two-part tariff for wholesale electricity (a power sector reform) would enable the introduction of carbon pricing to have a much greater influence on the dispatch order. 4.24. However, many of these reforms will be difficult, and may, on their own, lead to an increase in emissions. While energy sector reforms are important for mitigation, experience shows that they will be difficult to implement, especially in the power sector (Section 3.2.12). There are many reasons to undertake energy sector reforms. The main driver is economic. From a climate change perspective, the main reason for introducing energy or power reforms would be to increase the impact of carbon pricing. However, energy sector reforms would have their own impact on emissions, the direction of which is unclear. If there is no rationing, price increases will reduce demand and therefore emissions. But energy and especially electricity rationing is often endemic in developing economies (Section 3.2.6). Price increases often relax rationing constraints (for example, by improving the financial health of the sector, and thereby enabling more investment) and this can lead to increased usage and emissions (for examples, see Box 4.1). In addition, reforms can move the energy sector away from the plan, towards the market, but not all the way. This can introduce inefficiencies, both economic and environmental. The Chinese dispatch problem discussed earlier is yet again a case in point. If China hadn`t separated generation from transmission, it could have implemented optimal dispatch (taking into account environmental considerations) through central planning. But separating the two, and yet not establishing separate markets for energy and capacity, has led to inefficient dispatch. In this case, partial reforms may have had a negative environmental impact. Overall, while it is tempting to claim that energy or power sector reforms are win-win` (Independent Evaluation Group, 2009), this will vary from case to case. 66 Box 4.1: Does reducing energy subsidies always reduce emissions? Of all energy sector reforms, reducing energy subsidies appear the most likely to be win-win`: good for the economy, and good for the environment. However, whenever subsidies co-exist with rationing no a priori conclusion is possible. The increase in price will reduce (unconstrained) demand, but any concomitant relaxation in rationing will increase supply. A recent World Bank Independent Evaluation Group (2009) report concedes this point, noting that Where low prices have led to inadequate investment, removal of subsidies could result in expanded supply of grid-based power. Unfortunately, the same report doesn`t allow this rider to qualify its conclusion that reducing subsidies will reduce emissions. The one case where the report allows a subsidy-removal- induced reduction of rationing to influence its conclusions is in the case of gas where it argues that the reduction of price controls will increase the use of gas ­ and therefore reduce emissions because the gas will displace coal. In the end, whether subsidy reduction will lead to a reduction in emissions is an empirical matter which will vary from economy to economy. Few economies have successfully got out of a subsidy-rationing equilibrium, so it is difficult to get a sense of which effect will dominate. However, the experience of both India and China suggest the rationing-relaxation effects could be large. Some have suggested that India`s success in services rather than manufacturing is due to its weak infrastructure, including electricity (World Bank, 2006). Smaller electricity subsidies could have led, and could still lead, to a very different and much more emissions-intensive economic structure. China is one country that has, over time, moved away from both subsidies and rationing. World Bank (2007) reports that in the early 1980s electricity tariffs barely covered operational costs and that Severe energy shortages persist[ed] until the mid-1990s. After a succession of above-inflation price hikes, by 2000 average prices started to exceed supply costs, and shortages had been eliminated. Does this explain China`s 10% annual rate of growth in emissions since 2000? It certainly seems likely that if China hadn`t eliminated its electricity subsidies, it wouldn`t have been able to finance the expansion in supply required to overcome its shortages. As World Bank (2007) notes, increasing tariffs was crucial for China to meet its rapidly expanding electricity demand (p.33). In the case of China, reducing energy subsidies increased rather than reduced emissions. 4.1.4 How important are broader economic reforms for mitigation? 4.25. In some economies, important mitigation reforms might lie outside the energy sector. This could be the case for China. Section 3.2.1 pointed to distortions outside of the energy sector which might be pushing up emissions in some developing economies. Rosen and Hauser (2007, p.38) conclude, in relation to China: the pervasive revealed comparative advantage of heavy industry manufactured goods from China is generally rooted in distortions other than energy inputs. This is not to say that there are not energy subsidies. As documented in Section 3.2.4, there are. But petroleum and electricity prices for industry are already higher in China than in the United States (see Figures 3.6 and 3.13). It is not cheap energy that is driving China`s massive expansion of energy-intensive goods, such as steel. The search for what Rosen and Houser call the root causes of [China`s] structural over-allocation into energy-intensive industry (p. 37) must extend beyond the energy sector. Section 3.3.1 suggested that responsibility might lie with the factor-market distortions identified by Huang (2010), which make savings too high, interest rates too low, and land too cheap. The result is a record-high investment rate, which drives energy use both through high growth and through the capital-intensive nature of that growth.49 He and Kuijs (2007) argue that a rebalanced Chinese economy would not necessarily grow more slowly (since it would use capital more efficiently), but it would use less energy due to more rapid growth of services and less of industry. 49 SEI (2009, p.23) makes the same point about the importance of macroeconomic restructuring and financial reform to rebalance the allocation of capital and lending away from heavy industry 67 4.26. If this is correct, then Chinas current focus on improving efficiency within sectors is inadequate. China`s current approach (see Box 2.1) stresses the need to have, for example, more efficient power generators and steel-producing plants. This analysis suggests a broader approach is needed. It might seem tenuous to link social security reforms in China to climate change mitigation, but that is the direction the analysis of this report points us in. It is a daunting finding because it suggests that a successful mitigation response will require economy-wide reform (since such reforms would reduce savings and thus investment). But it is also perhaps an encouraging one, because China appears convinced of the need to undertake these rebalancing reforms in the coming years, and the climate change effort will then be a beneficiary of this broader reform effort. As Rosen and Houser write (2007, p.38): By coordinating the energy and environment imperative with existing calls to rebalance external and macroeconomic distortions, it may be possible to break through these conflicts sooner than would otherwise occur. 4.2 Technology-based policies 4.27. Technology-based policies are critical given economies multiple policy objectives. It was noted in Chapter 1 that developing economies have a range of overlapping policy goals ranging from energy security to clean air to climate change to industrial policy, some of which make technology-based policies essential. 4.28. Technology-based policy making is, however, a complex area which defies clear-cut conclusions. A discussion of technology policies cannot hope to be as definitive as one in relation to carbon pricing. Unlike carbon pricing, it is difficult to derive strong conclusions about the desirability of specific technology policies even assuming feasibility. As Section 2.3 noted, while no doubt technology policies can induce strong results, whether they are worth the costs of the intervention is more difficult to say. The discussion of Chapter 2 gave examples of prima facie success and failure. Nevertheless, it is not that nothing can be said. This section emphasizes four points, starting where the previous section left off in relation to carbon pricing. 4.2.1 Technology-based policies and carbon pricing 4.29. The weaker the case for carbon pricing, the stronger that for technology-based policies. If carbon pricing is a less effective mitigation instrument in some developing economies than in developed economies, for the reasons explored in the previous section, then prima facie this strengthens the case for technology-based policies. Australia`s experience has been that, with a clean energy target but without a price on carbon, renewables have displaced gas rather than coal. However, this is unlikely to be the case for developing economies, for two reasons. First, they have very little gas to start with, and the expansion of gas is constrained by supply shortages. Second, their electricity systems operate under central planning, with targets to expand gas. 4.30. However, the same structural weaknesses in the energy sector which can undermine carbon pricing can also undermine some technology policies. The implications are technology- and policy- specific, but one can think of several examples: Section 3.2.7 showed the prevalence of captive power in some economies. The costs of complying with renewable energy mandates will typically fall only on the grid company, and so may promote reliance on (polluting) captive power, which won`t be subject to any such levy. Politicized pricing can also undermine feed-in tariffs. Economies establish feed-in tariffs with the assumption that the incumbent utility will buy up all electricity renewable generators are prepared to supply at the mandated price. However, as discussed in Section 4.1.2, the incentives of the utility may not be aligned with this policy. Say that the utility is loss-making. Buying renewable energy at a high feed-in tariff will only add to its losses, and the subsidy it must request from government. While in theory the government might have committed itself to cover the additional 68 costs a feed-in tariff would impose, in practice the utility knows that the overall subsidy bill is subject to negotiation. The utility will therefore resist, and the feed-in tariff will be little used. Box 4.2 illustrates this point with an example from Indonesia. Box 4.2: Why Indonesia's feed-in tariff for renewable energy hasn't worked. Indonesia has had feed-in tariffs in place for renewables for over a decade, but with little effect. The IEA concludes that this is because the feed-in tariffs are based on PLN`s avoided costs, and so are not transparent, leading to long negotiations (IEA, 2008b, p.103). However, it can`t be that difficult to work out what PLN`s avoided costs are. The experience of one small-scale hydro project in Central Java, recounted in detail in ADB (2008), suggests a different and deeper reason. Initially, in March 2005, PLN offered a purchase price of IDR 246 (US 2.5c) per kWh for electricity supplied from this project, but, eventually, two years later, after consultations with the provincial government, PLN increased its offer price to IDR 618 (US 6c). This was in line with the avoided cost regulations, but PLN was only prepared to sign a one-year contract which it finally did in January 2008, almost three years after the project started. That it took a project with multilateral bank backing three years to negotiate a contract, and then to receive only a one-year contract, suggests that Indonesia`s feed-in tariff scheme is not working well. The ADB project review is written from the perspective of the organization which constructed the hydro project, but it is not hard to infer from it what was driving PLN`s tardiness, especially when one considers that its average revenues are only 6c per kWh. Yet this is what is what it was being asked to pay for the hydropower, leaving it with no margin for transmission, distribution and retail costs (which can be about half of total costs). Essentially, the feed-in tariff is requiring PLN to add to its losses. While the economic rationale is sound (as willingness to pay is far above 6c), PLN`s behavior is also rational, as there is no guarantee that the government will increase its subsidy to match the increase in costs. While no doubt the scheme could be tweaked to reduce the discretion of the utility, and so increase take-up, the review document is full of technical and legal reasons for delay. As long as the feed-in scheme is contrary to the utility`s financial interests, PLN will minimize take-up, and thus undermine the scheme`s chances of success. A similar problem may arise even if the utility is not dependent on government subsidies, as long as electricity pricing is politicized and cost pass-through is not guaranteed. Then again utilities may be reluctant to incur expensive obligations. One of the mysteries about Chinese wind power policy is why so much of the wind power is not connected to the grid (Box 2.5). It has been suggested that China`s grid companies are not keen to connect wind power because they know that by doing so they will add more to their costs than their prices. 4.2.2 Research and development 4.31. There is good evidence that research and development is both effective and underfunded. Japan`s position as the world`s dominant renewable energy innovator (Table 2.3) is no doubt related to the fact that its government invests more in renewable energy R&D than the United States and Europe combined.50 High prices for fossil fuels (Figure 2.1) must also provide a strong push factor, but do not explain Japan`s leadership role relative to Europe. The importance of research and development is also illustrated by the history of innovation in solar power. Nemet (2006, pp. 2226-2227) reports that government and university R&D programs produced 10 of the 16 breakthroughs in cell efficiency post 1980. He also notes that the rapid rise in laboratory cell efficiency from 1983 to 1990 immediately followed the unprecedented $1.5 billion investment in worldwide PV R&D in the previous 5 years. Not all public R&D projects will generate a return; in fact, only a minority might. There are risks of rent- seeking (Cohen and Noll, 1991). Australian government support for green cars seems more likely to 50 In 2004, Japan`s government spent $200 million on renewable energy R&D, the United States` government $70 million and the EU15 governments $50 million (IEA, 2010c). 69 benefit the local car industry than to make a contribution to climate change mitigation. But, overall, R&D support shows results. An evaluation of energy efficiency and fossil energy research at the U.S. Department of Energy (DOE) over the last two decades suggests that the benefits of [the few] successes justified the overall portfolio (Newell, 2009, p.21). 4.32. Of course, demand-pull policies are also essential. R&D will push renewable costs down, but probably not below fossil fuel costs. So demand-side subsidies (of whatever type) will remain critical for take-up, and will themselves stimulate technological improvement. The work of Johnstone et al. (2009) shows that Europe`s renewable energy targets and feed-in tariffs have stimulated technological innovation, as measured by patent count. 4.33. The point made here is rather the need for a balanced approach. The IEA has estimated global clean energy research, development and demonstration spending of $18-37 billion is needed annually between now and 2050 to achieve global climate change goals. This compares to the $5.2 billion actually being spent, leaving an annual gap of $13-32 billion (IEA, 2009c, p.51).51 Although energy research and development spending is now increasing, it is off a low base and the increase to date has been modest (Figure 2.6). Economies have shown a strong preference for demand-side policies over R&D funding. Frondel et al. (2010) estimate that the costs of Germany`s solar feed-in tariff over the period 2000-20010 was 65.5 billion Euro in net present value terms. By contrast, German government expenditure in research and development for all clean energy was 211 million Euro (p. 4050). (Private R&D expenditure on clean energy in the same year was only 139 million Euro.) This ratio of about 30:1 for annual spending on demand-side incentives relative to R&D implies a stark neglect of the latter. More generally, it can be noted that while many economies have renewable energy targets, very few have targets to expand public clean energy R&D spending. 4.34. How much developing economies will want to expand R&D will vary from case to case. Chapter 2 noted that all economies need some clean energy R&D, even if it is only to investigate which renewables are best suited to their climatic conditions. It also noted that special country conditions are driving some to make much bigger investments in public R&D in the future than they have in the past: Indonesia is planning a much bigger public-led exploration its enormous geothermal resources, for example. The larger and more advanced among the APEC non-OECD economies such as China, Russia and Malaysia already see themselves as incumbent or prospective technological leaders, and could be important players in any global scaling-up of clean energy R&D. Technological leadership will, however, require greater R&D spending on the part of these economies. To take one example, despite substantial cost reductions and indigenization, Chinese wind-power is still a considerable distance from the technological frontier (Figure 4.1). 51 This is less than the amount shown in Figure 2.6 which is all energy R&D. IEA in its study only considers 8 clean technologies. 70 Figure 4.1: China's wind industry is still far from the technological frontier Source: UNDP (2010) Box 3.3 4.2.3 Financing 4.35. The difficulties renewable energy projects might have getting financing were outlined in Section 2.3.1. Renewable energy projects are unusual among high tech ventures for their large capital size (and so upfront cost) and for their long-term policy dependence: the best-case scenarios have renewable technology costs coming down but not to levels at which they can hold their own against established fossil-fuel technologies. As a result, these projects might have difficulty attracting funds from venture capital funds (too small) and other sources of financing (too conservative). There is a case that the state should confront this failing of the financial markets by stepping into the financing role itself through the provision of loans and guarantees for low-carbon projects. One of the arguments for this is that it will strengthen the credibility of carbon policies if the government is seen to have backed those policies with its own funding. By making the value of that money hostage to the continuation of renewable policies, the credibility of the renewable policies it itself enhanced. Section 2.3.2 outlined some of the initiatives and proposals in this area, in both APEC and non-APEC economies. While the dangers of state-backed lending are self-evident, the risks of refusing to contemplate this option are also real. 4.2.4 Policy rigor 4.36. Countries need processes to encourage the adoption of good technology-based policies and weed out bad ones. As Chapter 2 made clear, there is no guarantee that technology-based policies will work well. Experience to date suggests that there is a real risk of governments choosing bad policies which advantage vested interests but do not provide sufficient environmental benefits to justify their economic cost. Independent policy review processes, and the sharing of lessons and experience will both increase the chances that good policies are adopted and bad policies avoided or dropped. Requiring that policies pass cost-benefit tests, even if there can be no rigorous proof, will encourage contestability and raise the barrier against bad policy. 71 4.37. Keeping technology policy as simple as possible will help promote coherence and resist rent- seeking. As Section 2.2 showed, there is a vast array of policies which governments are increasingly deploying to promote renewable energy. Governments can provide tax and tariff credits, accelerate depreciation, subsidize R&D, and set feed-in tariffs and/or performance standards, all at different rates for different technologies. It is tempting for a government that wants to do more to introduce more instruments. But complexity comes at a price. Australia, for example, provides feed-in tariffs in some states for household solar power, but also includes the energy produced (using imputation) against the national renewable energy target. The risk, which has become evident over the last year, is that the policies will be out of synch. The feed-in tariffs (set by state governments) seem to be encouraging solar to such an extent that the price of the renewable energy credits arising from the federal government`s renewable energy target is falling, reducing support for the other technologies which the renewable energy standard was intended to encourage. Complexity also makes rent-seeking and lobbying more likely, as it reduces transparency and increases the scope for deal-making. 4.3 Conclusion 4.38. The main point of the arguments developed in this chapter is to caution against unrealistic expectations and to emphasize the importance for mitigation in developing economies of a broad-based response. Carbon pricing is important but its impact may be limited and it will not, on its own, suffice. Energy sector reforms and broader economic reforms can also be important. Technology-based instruments will also be needed. 4.39. The relative importance of these four reform fields ­ carbon pricing, technology-based policies, energy sector reforms, and broader reforms ­ is a matter for judgment, and will vary from economy to economy, and over time. Sequencing is an even more difficult issue. Which should be introduced first: a price on carbon, or the mechanisms in the energy sector to allow that price to be passed through? There is no clear and certainly no universal answer to that question. Where an economy chooses to move first will depend as much on political judgment as economic analysis. 4.40. There are high uncertainties around mitigation. In developed economies, the main uncertainties around mitigation are political (will carbon pricing be introduced, and when) and cost. In developing countries, uncertainty also attends to both the implementation of reforms (for example, will energy sector reforms, even if announced, by carried through), and their impact (for example, how much will carbon pricing change the emissions intensity of an economy). There is little cross-country experience, but major risks, as outlined above. 4.41. Given the uncertainties, a quantity anchor for climate change fiscal policy is recommended.52 The uncertainty around the impact of a carbon price, whether fixed by government or determined by the market, makes a strong case for thinking of carbon prices as one among several instruments to achieve explicitly-stated environmental outcomes. The multiple fronts on which governments may need to move for an effective mitigation response ­ carbon pricing, technology-based policies, energy sector reforms, and broader economic reforms ­ also make the case for an underlying quantity target, since it will make it easier for countries to judge progress, and adjust the policy mix accordingly. 4.42. Developing economies have started to put quantity targets in place, but most have some way to go. All of the economies submitting domestic commitments to the Copenhagen Accord have nominated quantity rather than price targets. It must be noted, however, that of the APEC non-OECD economies, 52 Note that a quantity anchor need not be an absolute emissions cap. It could be an emissions intensity target, or even a target defined relative to business as usual. 72 only Russia and China have submitted targets which are verifiable. The majority of developing economies have submitted targets relative to business as usual (BAU) which is, by definition, only observable if there is no mitigation. If economies want a yardstick by which to judge their mitigation progress, they should convert targets defined relative to BAU into absolute emissions or emissions- intensity targets. 4.43. While the focus of the report is on CO2 and the energy sector, several of the key issues which have emerged from the preceding discussion are also important for other sources and types of greenhouse gases. Box 4.3 illustrates with a discussion of forestry sector mitigation issues. Box 4.3: Cheap but not easy: reducing emissions from deforestation and forest degradation. The global importance of reducing emissions from deforestation and forest degradation has come to be stressed in recent years. The size of these emissions means that they are simply too large to ignore. The IPCC`s Fourth Assessment Report puts emissions from deforestation and related sources at about 17% of total global greenhouse gas emissions (IPCC, 2007, Chapter 1). The costs of reducing LUCF emissions are thought to be on average very low. An influential study carried out for the Stern Review (Grieg-Gran, 2006) examined the cost of reducing emissions from deforestation for eight economies that together are responsible for 70% of land-use emissions. The study finds that If all deforestation in these economies were to cease, the opportunity cost would amount to around $5-10 billion annually (approximately $1-2/tCO2 on average). (Stern, 2007, p. 245). The general message that forestry can make a very significant contribution to a low-cost global mitigation portfolio is one endorsed by the latest IPCC Fourth Assessment Report on the basis of its survey of forestry-related cost studies (IPCC, 2007, Chapter 9) which found that 60% of all economic potential for avoiding deforestation would carry costs below $20/tCO2. But, as Howes (2009b) argues, even if reducing emissions from deforestation and forest degradation is cheap, that doesn`t mean that it will be easy. Howes points in particular to two problems which will face many developing economies. (This in the context of a study of Papua New Guinea, but the analysis is easily generalized.) First, monitoring of forestry emissions is difficult. Grainger (2008, p. 818) comments that while the planet has been monitored by remote-sensing satellites since 1972, estimates of the annual deforestation rate are still inaccurate, and the appearance of each new estimate generates debate. Second, governments might have few handles to respond to the drivers of deforestation, which can include subsistence agriculture and illegal logging. Forestry, which by definition occurs in remote locations, is a difficult activity to monitor. PNG would not be the only country where observers perceive a lack of effective governance in the logging business (Shearman et al., 2008, p.7). In the energy sector, emissions monitoring is much easier, but, as discussed at length in this report, there is a similar problem of finding an appropriate policy handle for reducing emissions. In general, the key mitigation issue for many developing economies will not be the cost of mitigation but the choice of instruments. There is no magic bullet, and much experimentation will be needed. 73 Chapter 5 Fiscal aspects of adaptation to climate change 5.1. Adaptation is the deliberate effort to obviate or ameliorate the bio-physical effects of a changing climate.53 Building sea walls to control flooding from rising sea levels or storm surges is an example of adaptation. Responses to mitigation measures such as reduced petrol consumption due to the introduction of carbon taxes are not. Afforestation may serve to both mitigate greenhouse gas emissions and to assist adaptation by lowering local ambient temperatures.54 5.2. Adaptation is not the same as ,,climate-proofing. Farmers may prefer to protect fishing activity in the local river from the effects of climate change, while accepting occasional flooding of roads or other hard infrastructure (see Dobes, 2009). Similarly, people may prefer more developmental investment in schools or hospitals, combined with acceptance of some degree of flooding, rather than an expensive structure that totally prevents local inundation. They may even adapt through migration (Dun, 2010; Gemenne 2010). The benefit of any adaptation measure should ideally be measured on the basis of willingness to pay to avoid the consequences of specific aspects of climate change, rather than the production-oriented damages avoided` approach. 5.3. Adaptation to climate change necessarily occurs in the context of ongoing competition for a limited set of a societys resources. Choices need to be made between different adaptation projects. Adaptation projects also compete with other socially desirable investment such as education, health or defense. Fiscal policy therefore has an essential role to play when it comes to adaptation. 5.4. The study of the fiscal dimensions of adaptation is in its infancy. Fiscal analysis in relation to adaptation has three dimensions: the costs of adaptation; fiscal instruments to facilitate adaptation; and decision-making tools to guide adaptation responses. The literature to date has focused on costings. While useful for advocacy and awareness-raising, adaptation costings provide little policy guidance to resource-constrained governments. Analysis of possible instruments which can be brought to bear to help societies adapt is clearly critical, but because adaptation challenges are so location-specific there is a natural limit as to how much general guidance can be given in this area. Therefore, while this chapter covers both adaptation costings and instruments, the main focus is on decision-making tools for adaptation, especially under uncertainty, which is the hallmark of climate change. 5.5. The chapter begins with a summary of national adaptation policy processes in the three case- study economies (Section 5.1). It then turns to adaptation costing (Section 5.2). Section 5.3 presents a range of decision tools, and argues for the use of cost-benefit analysis in adaptation decision-making, against more popular techniques such as multi-criteria Analysis. Section 5.4 explains how uncertainty can incorporated into cost-benefit analysis. The final section of the chapter, Section 5.5 discusses some important fiscal tools for adaptation: funding of public goods, public pricing reform and financial instruments. 53 The Intergovernmental Panel on Climate Change (IPCC) considers adaptation to refer to responses to climate change where the climate has been altered due to anthropogenic causes, an approach that is consistent with its focus on mitigating anthropogenic emissions of greenhouse gases. The approach taken here is that of the United Nations Framework Convention on Climate Change (UNFCCC), which treats adaptation as a response to climatic change in general. No distinction is made here between adaptation to man-made climate change alone, climate change that may occur due to natural phenomena such as sunspot cycles, or some combination of both. 54 Conceptually, adaptation and mitigation are imperfect substitutes in terms of climate change policy. Lower levels of mitigation can be compensated by more adaptation: if global temperatures increase because atmospheric greenhouse gas concentrations have not been sufficiently reduced, more air conditioners may be fitted to houses. Similarly, the destruction of an ecosystem may be ameliorated or prevented by increased mitigation, and hence reduced climate change, but adaptation (e.g., through relocation of the entire ecosystem) may not be feasible, even if conceptually possible. 74 5.1 Adaptation policy processes in China, Vietnam and Indonesia 5.6. All three case-study countries have put in place high-level bodies and processes to guide their adaptive response to climate change. 5.7. In China a National Coordination Committee on Climate Change, established in 1998, was transformed into the National Leading Group to Address Climate Change in 2007, coordinating about 20 government agencies. China`s 2008 White Paper (State Council, 2008) on climate change announced that the National Development and Reform Commission (NDRC) had assumed responsibility for general work and administration in respect of the National Leading Group, effectively giving the NDRC a national coordinating role. 5.8. The National Climate Change Programme issued by NDRC (2007) is a comprehensive but largely aspirational rehearsal of climate change issues that need to be addressed, including adaptation in key sectors such as agriculture, forestry, water resources and coastal eco-systems. A more concrete review document on policies and actions for addressing climate change was published as a progress report by NDRC (2009), primarily canvassing achievements in 2008 and foreshadowing future targets in a range of developmental and energy-related fields. 5.9. Except for major projects, responsibility for implementing adaptation measures appears to rest at the level of provincial government and below. In some areas, there is an expectation of in-kind contributions of labour and resources by local citizens. In the nationwide afforestation campaign, for example, every citizen volunteers to plant trees, every department undertakes a certain amount of afforestation, every city creates forests ...` (NDRC 2009, p. 31). 5.10. The key document setting out Vietnams approach to adapting to climate change is the National Target Program to Respond to Climate Change, proclaimed by Prime Minister Nguyen (2008). Key features of which are the establishment of a National Steering Committee, Executive Board and Standing Office to oversee implementation. Significantly, but unusually, Prime Minister Nguyen Tan Dung chairs the Steering Committee. This has been interpreted by some as deliberately providing the Ministry of Natural Resources and Environment with important backing relative to the other, more powerful Ministries that comprise the Executive Board. 5.11. The hallmark of the National Target Program itself is the enunciation of three phases. The initial two phases (2009-2010 and 2011-2015) are intended to develop and consolidate knowledge about potential climate effects on Vietnam in a range of sectors, including updating climate change scenarios, especially sea level rise over the course of the century, with initial implementation actions. Subsequent actions envisaged include development of response measures, public awareness-raising, establishment of cooperative mechanisms with international donors, mainstreaming climate change issues into development plans, etc. An annex sets out a tentative budget for various ministries and tasks for the period 2009-2015. 5.12. The Government of Indonesia issued a National Action Plan Addressing Climate Change in December 2007, and established a National Council on Climate Change by Presidential Decree in July 2008. Indonesia`s adaptation action plans focus on the core issues of water management, rice production, coastal management, and disaster preparedness. In 2009, the Indonesian National Development Planning Agency, Bappenas, issued the Indonesia Climate Change Sectoral Roadmap. The Roadmap, which is intended to be an input to the 5-year Medium-term Development Plan 2010-2014, catalogues the main climate impacts for Indonesia and lists broad policy approaches and strategies (e.g., population management` in coastal areas subject to inundation) to address them. Strategies listed are regulatory or managerial in nature, rather than fiscal or financial. Initial reforms have stressed the mainstreaming of adaptation (on which, see Jotzo et al., 2009) and institutional development. A number of new bodies 75 have been created including a National Water Resource Council, and 13 provincial water resources councils, and irrigation asset information management system, a National Disaster Management Agency, and a National Secretariat for the Coral Triangle Initiative. 5.2 Costing adaptation 5.13. The fiscal approach to adaptation which has received most prominence to date is its costing. A range of studies have been produced by the United Nations and the World Bank. Most recently, World Bank (2010e) has estimated the global costs of adaptation ($70-100 billion annually out to 2050, in 2005 prices), and also produced costs for a number of individual economies, including Vietnam. The total undiscounted cost of adaptation measures in agriculture in Vietnam is estimated to be about $210 million per annum at 2005 prices over the period 2010-2050. Construction of sea dikes and other flood defenses for urban infrastructure and the most valuable agricultural land is separately estimated at $20-50 million per annum. General equilibrium modeling indicates that damages avoided through implementation of adaptation measures would be clearly cost-effective under all three climate scenarios examined. 5.14. While the costings contain a wealth of useful information at the sectoral level, they are of less value as a decision-making tool. Parry et al. (2009) provide a detailed critique of the methodology and results of UNFCCC estimates, while the World Bank (2010f) study identifies the major methodological drawbacks of its own approach, as well as those of others. Apart from all the empirical difficulties of estimating a variety of locally-incurred costs, the fundamental conceptual problem is that costing adaptation to future climate change requires an ability to distinguish clearly between the components of expenditure that are attributable respectively to adaptation and to development. This is more difficult than it sounds, since most expenditures that can be classified as adaptive can also be classified as developmental. For example, introduction of drip irrigation techniques by farmers along China`s Yellow River might be attributed to increased water prices, lower water availability due to an increased upstream population, or lower run-off into the river due to long-term climate change, or to cyclical drought conditions. There appears to be no satisfactory method in general of distinguishing adaptation measures from those intended to foster economic development. In some sectors the calculations seem straightforward. Estimating the additional maintenance that will be required to keep road quality at a given level with increased temperatures seems a reasonable way to cost road infrastructure adaptation expenditure requirements, provided that sufficient technical information is available. But costing adaptation to extreme weather events through calculation of the cost of educating women to neutralize the cost of the more extreme weather events seems contrived (World Bank, 2010f). In any case, costs alone can provide no basis for policy action. Comparison of both the costs of implementation of adaptation measures and benefits is required if welfare is to be maximized. That is, as the next section argues, standard cost-benefit analysis is required at the local level. 5.15. A focus on costs also reduces the importance of uncertainty and policy reform. Despite the inherent difficulties, the uncertainties associated with future climate change must be front and centre of any analysis (Section 5.4). With regard to policy reform, incurring fiscal costs can often be an unnecessarily expensive alternative to implementing reforms. For example, introducing rational pricing policies for water in drought-affected areas might lead to savings on the construction of new dams. 5.16. Global and national adaptation costing exercises are useful to draw attention to the need for adaptation and are required to gauge the extent of compensatory assistance that may need to be transferred to less developed economies in the future. However, the experience with the costing of other international goals ­ ones conceptually far less challenging to estimate than adaptation ­ should serve as a cautionary tale, as Box 5.1 on the Millennium Development Goals illustrates. Adaptation costings may be useful for international fund-raising campaigns, and may in some cases help with the planning of adaptive responses, but they do not serve as a good guide for policy-making. 76 Box 5.1: Costing adaptation: lessons from the Millennium Development Goals The idea that development challenges should be costed and used is an old one. It has long been used both at the national level (e.g., to calculate investment requirements` to achieve target growth rates) and at the international level, in relation to various global goals. Typically, there is a close link to fund-raising, with such need-assessments providing the basis for aid mobilization campaigns. A recent prominent example is the Millennium Development Goals (MDGs), the set of 8 global goals relating to poverty and human development adopted by the UN in 2001 for achievement by 2015. A number of costings by the World Bank and the UN were attempted for these goals in the first half of the last decade, arriving at the range of $35-76 billion. One of the striking features of the MDG costings is the skepticism of the authors who derived them. Some have described this exercise as highly speculative (Devarajan et al., 2002, p.15). A UN high-level panel likewise provided some estimates, but at the same time conceded that it lacked the data to put even a rough price tag on the cost of the MDGs let alone a convincing one (Zedillo et al., 2001, p.31). If the MDGs are difficult to cost, adaptation is even more so. This is consistent with the wider range of costing estimates for global adaptation: from $4-$166 billion annually (Agrawala et al., 2008). The other noticeable feature about the MDG costing exercises is the explicit recognition given by the authors of these costings to the primacy of domestic policy. Both the World Bank and the UN explicitly defined their costings as necessary but not sufficient for the achievement of the MDGs. They noted, in the words of the UN panel, that investment in developing countries is unlikely to promote ... human development if domestic policy fails to attend to the fundamentals (Zedillo et al., 2001, p.4). In the same spirit, there is a need for more focus on policy reforms. The World Bank (2006b, p.33) concludes that Initially, public finance is likely to be the main driver of adaptation. There is a need to look at a wider range of policy settings and instruments rather than just likely government spending needs. 5.3 Decision tools and criteria for adaptation fiscal policy-making 5.17. As climate change and awareness of climate change grows, Finance Ministries can expect to be increasingly confronted with requests for funding and possibly for fiscal reforms, such as changes in taxes, subsidies, and prices. How are fiscal decisions about adaptation to be made? This section reviews four tools. 5.3.1 Multi-criteria analysis 5.18. Multi-criteria analysis is generally approached by developing a goals achievement matrix. A simplified example is shown in Table 5.1. As its name suggests, the purpose of multi-criteria analysis is to assess the relative contribution of a selected group of impacts or attributes to the achievement of an overall objective or goal. The basic methodology is presented below, but there are many related variants and accompanying nomenclatures, including multi-attribute utility theory, analytical hierarchy process, multi-criteria decision analysis, weighted product model, etc. World Bank (2009b) presents a study using multi-criteria analysis to set priorities for adaptation decisions for Mexico, Peru and Uruguay. Although it has not commonly been used to analyze adaptation, multi-criteria analysis has been widely used in relation to environmental problems more broadly. Eakin and Bojorquez-Tapia (2008) apply it to categorize household vulnerability in rural Mexico. 5.19. The first step in a multi-criteria analysis is the selection by an analyst or policy-maker of a range of attributes or impacts they consider to be relevant to the issue at hand. In Table 5.1, the hypothetical analyst has chosen to represent the construction of a flood-proof dike in a fairly common way. Increased production, easier planning by the commune authorities, reduced drownings and employment prospects are seen as the main benefits, and are offset by the cost of the project. 77 Table 5.1: Illustrative example of simplified multi-criteria analysis: hypothetical dike building project attribute units impact Score weight weighted (-4 to +4) % score additional rice crop each year kilograms 13,000 2 10 20 easier planning of production -- -- 4 40 160 reduction in people drowned number 4 3 10 30 employment jobs 23 3 20 60 cost of project RMB, VND 89 bill -4 20 -80 total 100 190 5.20. Because any number of other impacts might have been equally reasonably considered to be relevant, this stage of the analysis is clearly highly subjective, irrespective of the analysts intent or desire to be objective. There is no theoretical basis to guide the choice of criteria. An ecologist or an official from the Ministry for Environment would most likely have chosen a quite different set of criteria. Orlove (2009), for example, reports a striking divergence` between indigenous herders in highland Peru and NGOs regarding time horizons and the scope of projects. Another common problem is that of overlapping criteria, resulting in double-counting.55 This contrasts with cost-benefit analysis, discussed below, where the analyst is required to address all relevant impacts from the perspective (the standing`) of society as a whole, and double-counting is specifically eschewed. 5.21. In the next step, each of the selected attributes is given a score ­ usually on the basis of an ordinal Likert scale56 ­ generally by the analyst or a focus group. In Table 5.1, a score of 2 has been given to the additional rice production, half the score for easier planning of production by the commune. A farmer may well have reversed the scores. Again, there is no clear theoretical basis involved, although various methods have been devised to increase the sophistication of the procedure. Ultimately, the scores remain subjective no matter who determines them, or to what degree of sophistication. 5.22. The analyst or a focus group then assigns weights to each of the impacts, in order to reflect their relative importance within the analysis. In Table 5.1, production-oriented aspects (rice and planning) have been given an importance that totals half the index. Cost can only influence the index to the extent of 20 per cent. The weights are again subjective and can be subject to a considerable degree of arbitrariness. 5.23. The weighted scores are aggregated to provide a single figure that is used by decision-makers to assess the desirability of proceeding with the project. In other words, multi-criteria analysis begins by taking cardinal values (the physical, numerical or financial units in which impacts have been specified), reinterpreting them within an ordinal scoring system and then multiplying by an interval scale (the weights). In essence, the technique aggregates incommensurable quantities such as kilograms of rice and lives saved from drowning through the artifice of attaching scores and weights. 5.24. Apart from the arbitrary and atheoretical nature of selecting attributes and assigning scores and weights, the process is fundamentally flawed mathematically. It is equivalent to adding apples and oranges (see Dobes & Bennett, 2009, Fuessel (2009, p. 14) and Cox, 2009). 5.25. The final result of 190 in Table 5.1 is a unitless quantity that can only be compared with alternative projects of the same kind, with the same set of impacts, and which cannot provide guidance on the social value of undertaking the project. It is not possible to judge how significant the final figure might be, or to compare it to a result derived for different circumstances, such as a drought-affected area. 55 An example is the inclusion in World Bank (2009b, p. 78) of the impact criteria economic benefits of the response option` and capacity to reduce damage caused by extreme events`. 56 Likert scales are commonly used in questionnaires to measure attitudes or preferences of respondents. They are generally used to represent choices between 5 alternatives such as strongly disagree` ­ disagree` ­ neither agree nor disagree` ­ agree` ­ strongly agree`, although the range can be extended or compressed, for example to 7 or 10 or 3 or 4 point scales. 78 5.26. Multi-criteria analysis therefore has little, if any use in policy terms. Worse, the subjective choice of impacts and assignment of scores and weights is open to abuse by vested interests. Faced with a serious housing shortage, for example, a Minister for Housing may understandably downplay longer-term climate change risks such as loss of life in populating a floodplain. Criteria such as reduced homelessness are likely to attract comparatively high scores and weights, with little or no consideration of issues of greater consequence to other Ministers (e.g., disaster management), and hence society overall. Because any arbitrariness tends to be masked by the superficial credibility afforded by the quantification inherent in the multi-criteria technique, the choice of criteria may not be subjected to appropriate scrutiny. 5.3.2 `Vulnerability', `adaptive capacity' and `resilience' indexes 5.27. The climate change literature is replete with discussion of concepts, case studies and indexes of ,,vulnerability, ,,adaptive capacity and ,,resilience. Such indexes are used increasingly in areas such as identifying the world`s most livable city`, the greenness` of buildings, environmental vulnerability, etc, to summarize or capture diverse attributes or characteristics. Examples of climate-related indexes include the comparison of the vulnerability of beach tourism in 51 economies to climate change (Perch-Nielsen, 2010), a comparison by Hahn et al. (2009) of two districts in Mozambique, a similar approach by Mohan and Sinha (undated) for the Ganga River Basin in India, and the application of a Climate Vulnerability Index with an emphasis on water-related impacts by Sullivan and Meigh (2005) to four developing economies. While vulnerability indices at one level are just descriptive summaries, there is an inevitable tendency to treat them as decision-making tools, with implicit or explicit assumptions that, for example, the most vulnerable areas should have the greatest claim on adaptation funds.57 5.28. At the most basic policy level, terms such as ,,vulnerability can be analytically trite. Low-lying coastal areas with no dikes or mangroves to protect them against storm surges are likely to be subjected to flooding. The transformation of basic facts like this into indexes tells us nothing about what decisions should be taken, or when. 5.29. In reviewing indexes of vulnerability, Fuessel (2009, p. 8) concludes that they "show substantial conceptual, methodological and empirical weaknesses." Cox (2009) further points out that impacts that are chosen for inclusion in indexes may not accord with the essential condition of additive independence if the impacts are interactive. Where risks are correlated, additive indexes can perform even worse than setting priorities randomly (Cox, 2009, p. 942), with obvious implications for setting funding priorities. 5.30. Vulnerability indexes have much in common with the approach used in multi-criteria analysis. Lack of an established theoretical basis leads analysts to choose the impact attributes that are used to constitute the index subjectively, with weights and scores used to derive aggregated values. Such indexes therefore suffer from the same weaknesses as multi-criteria analysis, and provide no clear decision rules when considering potential adaptation measures. 5.3.3 Cost-effectiveness analysis 5.31. Cost-effectiveness analysis is often used in everyday life, and is easily presented to and understood by policy makers. A measure of technical efficiency, it expresses a result in terms of the cost of achieving it: for example, the number of lives saved for the cost of each kilometer of a 5 meter dike constructed. At its most simple, it can reveal projects that generate the biggest bang for the buck`. Although generally used only for a single output or effect, cost-effectiveness analysis can be extended to multiple outputs and inputs through data envelopment analysis (see, for example, Coelli et al., 2005), a technique based on linear programming. 57 Although the conclusions and recommendations of the report are carefully expressed, Preston et al. (2008) produce mean vulnerability scores for 15 coastal councils in the Sydney region, using weights to aggregate key risk categories like bushfires, sea-level rise and extreme heat. The intention is to provide a basis for further work required to develop a comprehensive understanding of risk that may guide future management decisions (p. 2) by Australian governments. 79 5.32. Because it does not require the monetization of effects such as number of lives saved, policy makers find cost-effectiveness an attractive means of ranking projects with similar objectives. However, the very lack of a common numeraire means that comparisons can only be made between projects of a very similar nature. It is not possible to compare a dike project with a water project, for example, if the comparison made is between number of lives saved per dollar and kilograms of additional rice grown per dollar. Assessing the effect of climate change on the cost-effectiveness of nutrient management in a eutrophic lake, Gren (2010), for example, presents separate results for target reductions of nutrients (nitrogen and phosphorous) and water quality (minimum sight depth). Boardman et al. (2006, Ch. 17) review other issues that limit the usefulness of cost-effectiveness analysis. 5.33. Cost-effectiveness analysis also cannot be used to assess which projects will generate the largest benefits for an economy or its society as a whole. It is therefore of only limited use as a policy decision tool for comparing different adaptation projects and programs. 5.3.4 Cost-benefit analysis 5.34. Cost-benefit analysis has a number of well-known drawbacks. It generally requires monetization of both costs and benefits, assumes that the marginal utility of money is equal for everyone (unless distributional weights are used), and it is relatively expensive to conduct. It also requires a strong knowledge on the part of the analyst of micro-economic concepts and practical estimation techniques. 5.35. Nevertheless, cost-benefit analysis remains the only rigorous analytical tool available in terms of assessing issues such as the relative merits of different adaptation projects and strategies. In particular, it affords policy makers an unambiguous decision tool in requiring that the present value of benefits to society as a whole exceed the present value of costs incurred. In other words, of all the analytical tools available, it alone permits not only comparison of adaptation measures with each other, but also with alternatives that are not as closely associated with climate change effects. Boardman et al. (2006) provide a comprehensive exposition of the technique, as does Mishan (1988) in his classic version. 5.36. Cost-benefit analysis has been little used to date for adaptation analysis, but there are some good examples of its utility. Most analysis has been in terms of the various vulnerability indices discussed above, with some use made of cost-effectiveness analysis. However, the number of contributions to the literature is growing. Thang and Bennett (2007) find that a lack of information on environmental protection values, especially non-market values, has contributed to wetland degradation in the Mekong River Delta. World Bank (2010e) analysis suggests that irrigation projects planned for Bolivia are robust to future climate change. Nguyen`s (2006) cost-benefit analysis notably questions whether the increasing reliance on dikes in the Mekong delta is socially optimal (Box 5.2). Box 5.2: Cost-benefits analysis of dikes in the Mekong Delta region Examples of cost-benefit analyses related to the climatic conditions of Asia are not numerous. However, Nguyen (2006) compares three alternative flood adaptation systems in the Mekong Delta. A third rice crop can be grown each year during the flood season after the construction of a permanent, but expensive high dike that totally prevents flooding. Construction of a temporary, low seasonal dike can be used to protect the (second) summer rice crop from floods in August each year, with the option of allowing floodwaters into the paddy field afterwards for fertilization with river silt and to undertake flood-based farming activities such as raising prawns within net fences, fish in mosquito net` cages and neptunia, a local aquatic vegetable. Fish, crab and wild vegetables can also be harvested naturally from the floodwaters. Complete absence of a dike permits two rice crops per year, but flooding ensures dumping of fertilizing sediment, capture of natural fish, collection of snails that are sold to eel farmers, and other products. However, absence of a dike would mean damage to the summer rice crop and to other property due to annual flooding. Using available data, Nguyen (2006) concluded that the net economic benefit of seasonal August` dikes exceeded that of permanent dikes, with the no dike` method being least valuable. This is an important conclusion, since the practice of permanent dike construction is growing rapidly in the Mekong Delta region. 80 Despite the apparently greater social and environmental benefits to be gained from employing new flood-based farming practices using a temporary, seasonal dike, there has been strong growth in the area of the Mekong Delta that has been put under permanent dike systems (Nguyen, 2006, Figure 3, p. 5) Anecdotal evidence is that permanent dikes are encouraged by local authorities (provincial, district and communal governments) whose priority is to produce rice, a key Vietnamese export commodity. It is also likely that construction of dikes holds some appeal for aid agencies because they incur lower administrative costs in managing single, large-scale projects. Nor does there appear to be a significant degree of coordination between aid agencies providing assistance in different provinces. It is therefore likely that externalities may not be fully taken into account, especially in the absence of rigorous cost-benefit analysis. Construction of permanent dikes in upstream provinces like An Giang, for example, is likely to increase the intensity of flooding downstream due to the loss of natural floodplains. This creates new risks for the population whose livelihood was not previously threatened by flooding. The presence of acid sulphates in areas not flushed out with regular flooding is also likely to increase. Such problems are likely to intensify with further deterministic responses to future climate change. Further research to investigate the important results of Nguyen (2006) is urgently needed. 5.37. Many studies of adaptation actions present results in terms of so-called ,,cost-benefit analysis, but are more strictly characterized as ,,cost-cost studies, because they compare the cost of implementing an adaptation measure with the cost avoided of climate change effects.58 While some times there is no alternative to using damage costs avoided`, as noted in Box 5.3, they can only be a rough proxy for benefits. Box 5.3: `Cost-cost' versus `cost-benefit' Current studies of adaptation actions generally present results in terms of cost-benefit` analysis. Despite their cost- benefit label, most such studies are more strictly characterized as cost-cost` studies, because they compare the cost of implementing an adaptation measure with the cost avoided of climate change effects. Damage costs avoided` can only ever be a rough proxy for benefits. Benefits are more correctly measured as willingness to pay to avoid damage, or willingness to accept compensation for the damage. A householder may be willing to pay much more than simply the avoided damage to their furniture in a flood, for example, because they also place a (negative) value on the general inconvenience caused by the flood. In this case, using only a damage avoided` measure would underestimate the benefits of adaptation. On the other hand, a farmer who values a traditional lifestyle that includes the fertilization benefits of regular flooding may be only just willing to pay an amount represented by the damaged furniture to avoid the effects of the flood. Thus damage avoided` is probably best thought of as a lower bound estimate of benefits. 5.38. Cost-benefit analysis can be a challenge for economies with limited analytical resources. However, even limited back of the envelope` analysis can provide a useful indication whether an adaptation measure is likely to improve the well-being of society as a whole. Moreover, simplified and standardized approaches can be taken. World Bank (2010i) provides a toolkit for community-based cost- benefit analysis of adaptation projects in the energy sector, which has been successfully piloted in Albania and Uzbekistan. Resort to cost-effectiveness analysis also provides a fall-back option in cases where even a rudimentary cost-benefit analysis is not possible. 5.39. Increasing demand for use of cost-benefit analysis remains a challenge, in developed and developing economies alike. Continued advocacy and demonstration of the utility of the cost-benefit approach is important. The examples in this and the next section are provided with this in mind. 58 See for example table 1.1 (p. 23) in Agrawala and Fankhauser (2008). 81 5.4 Incorporating uncertainty into adaptation decision-making 5.40. Uncertainty is the hallmark of climate change. Knowledge is lacking ­ particularly at the local level ­ as to both the intensity and the timing of future climate change. Much work on adaptation has been based on mean values, but adaptation, by its very nature, requires consideration of extremes. The intensity of future weather events might remain much as it is today, or change along a spectrum that includes both the benign and the catastrophic. The timing of any additional intensity is equally uncertain. 5.41. A particular advantage of the cost-benefit framework is that it allows readily for consideration of uncertainty in at least two ways. Monte Carlo analysis permits probabilistic modeling of estimated net present values, taking into account probabilities associated with climate change and other variables (Section 5.4.1). The real options` approach discussed below is also readily assimilated within the cost- benefit framework (Section 5.4.2). 5.4.1 Extreme value distributions and Monte Carlo methods 5.42. Most economies have sufficient historical data to generate models based on extreme value distributions. Because historical data are used, they effectively represent a basecase` scenario of extreme events. Figure 5.1 illustrates an extreme value distribution for flood-inducing rainfall now and in the future. It plots the Annual Exceedance Probability (AEP) for various flood heights. Only the fat tail` of the distribution59 is shown because minor flooding (less than 1 meter in Figure 5.1) is assumed to be internalized by the local population. Figure 5.1 shows that a 5 percent AEP (1 in 20 year flood) is currently associated with flood heights of about 1 meter. Floods that occur once in a hundred years (AEP = 1%) are obviously less frequent, but cause more damage because they may reach heights of around 4 meters. Figure 5.1: Illustration of annual exceedance probabilities (AEP) for rainfall and floods 5.43. The basecase probability distribution for 2010 can be ,,augmented to take into account climate change over the coming century by shifting it to the right on the basis of predictions obtained from physical climate models. In Figure 5.1, for example, the current 1-in-20 year flood may be expected by the year 2100 to be 4 meters high, a level currently experienced only once in a hundred years. The approach as described so far is suggested by Repetto and Easton (2009) for hurricane losses, and by World Bank (2010g, Section 3.1.3). While it is clearly in the right direction, it ignores the fact that 59 Gaussian distributions such as the Normal are not particularly suited to modelling extreme events because the tails of the distributions rapidly approach zero probability. The tails of distributions with greater kurtosis, such as the Cauchy, are said to be fat` or heavy` because they indicate higher probabilities of occurrence of very low or very high values compared to the Normal distribution. 82 physical climate models will provide different, often contradictory results. Further, climate models may assume significantly different scenarios about greenhouse emissions. 5.44. The uncertainty surrounding climate change can be incorporated into the ,,augmented extreme value distributions by using a series of climate models to generate a series of distributions for each year or period under examination. Each of the distributions shown in Figure 5.5 illustrates the results of simulating climate change for a single with different models, or using different scenarios. (Only the extreme value distributions for one year are shown in Figure 5.2, for illustrative purposes.) The range of results indicated by the climate models reflects the uncertainty inherent in the knowledge of future climatic conditions. Figure 5.2: Uncertainty illustrated by a range of distributions for each year under study 5.45. By applying a damage function to each of the ,,augmented probability distributions for each year between 2010 and 2100, a set of cost functions can be generated for each year. This is illustrated in Figure 5.3. Applying Monte Carlo analysis by generating random numbers to choose flood heights (from among the different distributions in Figure 5.3), a probability distribution of costs can then be generated for each year, giving a range of possible values of cost for each year, rather than single point estimates. This approach was applied in cost-benefit work commissioned by the Australian Department of Climate Change and Energy Efficiency (2010): see Box 5.4. World Bank (2010g, p. 22) also endorses this approach. Figure 5.3: Illustration of cost functions generated from flood probability distributions 83 Box 5.4: Application of Monte Carlo analysis to adaptation to coastal inundation at Narrabeen Lagoon in Australia Narrabeen Lagoon is one of about 70 intermittently closed and open lakes and lagoons (ICOLLs) along Australia`s eastern coast. Storms can block ocean entrances to lagoons by depositing sand, but, in combination with flood waters from creeks that feed into a lagoon, they will occasionally also flush away deposits in the entrance. When its entrance is blocked, rain and floodwaters will generally fill a lagoon like a bathtub, and can therefore flood the land and houses around it. Because climate change is expected to increase the frequency and intensity of storms and rainfall in the Narrabeen catchment over the coming century, as well as raising sea levels, the Australian Department of Climate Change (2010) commissioned a study of the social costs and benefits to the community of adaptation measures such as levee banks to protect major access roads, widening the lagoon entrance, flood awareness programs, and planning controls. Two observations on historical data (a one in 20 years rainfall extreme event and one in a hundred years) obtained from local authorities were used to estimate the two parameters of a Gumbel extreme value distribution for the year 2009. Eleven runs of climate model simulations supplied by CSIRO were used to generate sets of distributions of rainfall probabilities for the years 2055 and 2090 (like Figure 5.2), with intervening years estimated by interpolation. Probability distributions were transformed into cost functions using damage estimates for different flood heights. Using readily available @Risk software, Monte Carlo analysis was applied by sampling from the 11 cost functions for each year from 2009 to 2010 to generate a single probability distribution for costs in each year. (An optimization model was also applied to assess the effect of interdependencies between different adaptation measures.) The study found that a flood awareness program, increasing the minimum height of new buildings and a levee at one site next to the lagoon would generate benefits greater than costs if implemented immediately. However, the benefits of widening the lagoon entrance would not exceed the costs until 2035. Source: Department of Climate Change and Energy Efficiency (2010). More information is available at http://www.climatechange.gov.au/~/media/publications/adaptation/coastal-flooding-narrabeen-lagoon.ashx 5.46. The methodology sketched out here would enable estimation of likely future costs of damages avoided due to climate change, taking into account the associated uncertainties. However, it would be feasible to extend this methodology to the estimation of willingness to pay. For example, insurance companies may be able to provide estimates of likely market premiums that would be commercially viable, both in terms of price and quantity demanded. Alternatively, some form of hedonic pricing, based on comparison with existing insurance policies in analogue` economies or regions, might be used. 5.47. Further refinement may be required of this outline approach to ensure that it is relevant to conditions in countries like Vietnam or China. For example, subsistence farmers who do not actually have access to insurance may be highly risk averse in order to ensure sufficient food for the next season. Rather than growing the most profitable crop, they may instead grow low risk and low yield crops. As their income grows over the course of the century, or if insurance becomes more readily available, perhaps in the form of micro-insurance, they may switch production patterns. Damage cost functions would also need to be altered to take this into account. Similarly, as their incomes grow, the general population may become willing to pay for environmental goods` (Stage, 2010). 5.48. Decision-making costs could be reduced if climate model-,,augmented extreme value distributions were produced for each economy, either by governments or aid agencies. Generation of augmented distributions such as those in Figure 5.2 requires considerable modeling expertise and information about each economy`s climate. There would be a credible public good` argument for the production of such functions for major regions within each economy. Because analyses of adaptation projects would be able to draw on the data generated, cost savings would be realized at all levels of government in commissioning future cost-benefit studies. 84 5.4.2 The `real options' approach 5.49. The real options approach is useful not only for assisting decision-makers to incorporate uncertainty within a cost-benefit framework, but also to encourage planners to incorporate options into the design of adaptation projects. The flexibility incorporated into real options, coupled with the delay of full investment, can expand the number of potentially viable projects.60 5.50. The flexibility of real options is particularly advantageous for seriously resource-constrained governments. Expenditure on climate-relevant projects can be spread out, to be made when required over time. This aspect makes real options` analogous to a just-in-time` technology. Because budgets can be spread out more efficiently over time, a greater number and a greater range of climate-relevant projects can be funded. The beneficial fiscal effects are obvious. Techniques and examples follow below. Addressing uncertainty in project investment 5.51. A financial option to buy a share (stock) gives the owner a contractual right to purchase the share fully at a specified price on or before a specified date in the face of uncertainty about the future price of the share. If the value of the share increases because its price rises above the specified level, it can generate a benefit to the purchaser in the form of a profit. If the price falls below the agreed level, the financial option, like a losing lottery ticket, becomes worthless, but there is no obligation to buy the share. 5.52. Many commercial enterprises regularly make use of the real options concept to provide managerial flexibility. Patents are a good example of a real option: for a fee, the buyer of a patent acquires the right, but no obligation, to commercialize an invention when conditions are propitious to making a profit. Similarly, a firm engaged in pharmaceutical R&D may build only a small pilot plant until commercial viability can be confirmed. Dixit and Pindyck (1994, pp. 15-16) give the example of American firms in mid-1993 not hiring permanent workers in the face of uncertain economic conditions. At the same time, the firms in question were willing to pay wage premiums for overtime work and to use agencies that supplied temporary workers and charged fees of 25 percent or more of the wage. The option to expand output when conditions improved was maintained, but without committing fully and immediately to hiring a larger permanent workforce. 5.53. Options can also be identified, or developed, by governments for physical (real) assets. A simplified example might be a dike or a sea wall built to prevent expected future flooding, but where the timing and intensity of the floods cannot be predicted with certainty due to insufficient knowledge about the future climate. 5.54. Panels A and B of Figure 5.4 below provide an overview of a ,,real options approach to constructing a dike intended to prevent future inundation resulting from more intense or more frequent floods due to climate change. 60 The real options approach is also relevant to mitigation policies. Anda et al. (2009) explore the benefits of flexibility in the context of climate change uncertainty, and Lambie (2010) analyses the influence on investment decisions by electricity generators of the design of emission trading schemes. 85 Figure 5.4: Two different approaches to building a dike to respond to increased flooding risk 5.55. It is, of course, possible to build a dike or a sea wall immediately (along the lines of the popular ,,precautionary principle) perhaps to a height that will provide protection against a devastating 1-in- 10,000 year flood. Panel A illustrates this case. The full cost of construction is incurred immediately, with regular maintenance costs thereafter, but the benefits are realized only at some uncertain time in the future. 5.56. The alternative shown in panel B is to expend only a relatively small amount today on surveying or preparing the land on which a future dike could be built. 61 This amount ­ smaller than the full cost of a dike ­ is analogous to the price paid to acquire a financial option, because it establishes the opportunity, but not an obligation, to invest in the full asset when required. A full investment can be made when the benefits of countering the effects of climate change increase the value of the asset above its costs. The net present value of the panel B approach is higher than that in panel A, irrespective of the discount rate used. 5.57. The real options framework is particularly attractive analytically because it can be readily incorporated into a social cost-benefit framework (Dixit and Pindyck, 1994). However, even the alternative project shown in panel B needs to meet the conventional cost-benefit test of a positive net present value if the welfare of society is to be increased. In a budget-constrained situation, priority should be given to projects with greater net present values, irrespective of whether they are adaptation` or development` projects, or some alternative such as national defense. 5.58. An important question raised by both panel A and panel B above is how a cost-benefit analysis is to be performed in practice. Uncertainty about climate change precludes specification of actual dates when benefits from a dike will begin to flow. Provided that one is prepared to approximate the future by a suite of climate models and/or projections, the application of Monte Carlo analysis to the results of climate modeling (Section 5.4.1 above) can provide some guidance to the analyst. 5.59. The key benefit of the approach taken in panel B is that it incorporates flexibility. As time goes by, it may be found that climate change is occurring less quickly than originally anticipated, perhaps due to reduced global emissions. Unlike the situation in panel A, where the dike has already been built, the option remains to further delay construction based on new analysis. Conversely, if climate change occurs faster than expected, it may be beneficial to proceed with construction sooner. It is also possible to abandon the dike construction project completely in the future if little discernible climatic change occurs. 61 In market economies, it might also be necessary for a government to purchase (or compulsorily acquire with appropriate compensation) the land to provide greater certainty of being able to build the dike when it may be required. An even cheaper alternative would be to only purchase options from landowners to acquire their land in the future. 86 5.60. By waiting until better information becomes available, a dike can be built with a height more closely matched to the actual climate of the future. The greater the uncertainty, the more valuable is the flexibility associated with being able to wait for better information. Because the community`s resources are used more efficiently, a greater number of adaptation projects (or even general developmental projects) can be initiated out of a given budget. Similarly, adaptation assistance provided by international organizations can be made available to more economies. Building options into projects 5.61. Wang and de Neufville (2005) provide a number of examples of real options that have been incorporated at the design stage of infrastructure projects, and could be usefully considered by economies in the APEC region. For example, citing Gesner and Jardim (1998), they report the construction of a bridge over the Tagus River at Lisbon where the original bridge was built to be strong enough to carry a second level, if required. The Portuguese government exercised the option in the mid- 1990s, building a second deck for a suburban railroad line (p. 5). Dobes (2009) suggests, as an example, that the real-options approach could be applied to the construction of a new airport runway. In a hotter climate, longer runways may be required to allow planes to develop sufficient lift to take off safely with full loads It would be expensive to build a long runway immediately, and may turn out to be an unnecessary cost if temperatures do not increase as much as initially anticipated. In this situation of uncertainty about future climate-change impacts, a real option` could be the construction of a normal runway, but accompanied by the purchase of additional land (or just an option to buy the land) at the end of the runway to allow for a possible extension later, if required. In other words, the airport operator does not need to commit the full extent of funding immediately. 5.62. It would be possible for governments or aid agencies to assist rural residents in flood-prone areas by constructing solid houses on stilts. An example is shown in Figure 5.5 below of houses constructed in coastal areas of Indonesia on the basis of a new housing design introduced by the Indonesian Ministry of Marine Affairs and Fisheries. The design is reportedly intended to meet future threats by raising houses 160cm above the ground. Figure 5.5: New housing design in coastal areas in Indonesia Source: Asian Development Bank (2009), Figure 6.7, p. 113. 5.63. While it would represent a sensible solution to a known and quantifiable threat, the Indonesian solution lacks flexibility to respond to emerging risks that may not have been foreseen at the time of implementation. The house is evidently solidly built, but it would be difficult to raise its floor height, or to move it elsewhere, if required. In this sense, it is similar to the Vietnamese houses that have concrete or wooden stilts (Box 5.4), but is almost certainly more expensive to construct. 87 5.64. A practical example of a real option that is embedded in housing manufactured in Vietnam that can be quickly assembled and disassembled is provided in Box 5.5. There are at least three options embedded in the Vietnamese version: upgrading the cladding, raising the floor level to address any future increase in flood levels, and relatively simple relocation if required. Box 5.5: Real options in Vietnamese housing Houses on wooden stilts are a traditional form of construction in flood-prone areas, including in the Mekong delta. An analogous approach has been adopted by the BlueScope Buildings (Vietnam) company, which supplies relatively light, easily assembled zinc-coated steel frames for houses and community facilities. There are at least three different real options` embedded in the manufactured, steel-frame version. Most important is the fact that the floor of the steel-framed house can be easily raised because the beams and joists are attached to each other with nuts and bolts that can be adjusted with a spanner. Adaptation to increased flood levels is therefore an option over time at minimum cost. Houses built on wooden or concrete stilts do not have this flexibility. Another embedded option is that the wall cladding can be changed. Poor farmers may initially use bamboo or wood, but can later switch to concrete or other more permanent materials as their wealth increases. Where total relocation is necessary ­ for example, where increased salinity precludes viable agriculture ­ a third option exists of dismantling the steel frame and transporting it to a new location. The frame members are light enough to be carried in boats along canals, or even by means of the ubiquitous motor cycle. (a) Traditional houses in the Mekong Delta 1. low income house, no stilts, mixed cladding 2. concrete stilts 3. wooden stilts (b) Bluescope houses in the Mekong Delta with embedded options 4, display house with wood cladding 5. house with floor raised in flood-prone area Sources: 1&4: BlueScope Buildings (Vietnam); 2: Nguyen Van Kien; 3: Leo Dobes 5: North Sullivan. 5.65. Another example of a real option implemented in a commune in Cho Moi district of An Giang Province relates to the planned future raising of the height of a dike. The commune authorities have located electricity power lines away from the existing dike, and have banned planting of eucalypts and melaleuca (so-called permanent` trees) next to the existing dike. These measures will ensure quick 88 exercise of the option of increasing the height of the dike, once a decision has been taken on timing of implementation. On the other hand, the same commune has made provision to embed in the dike only a limited number of sluice gates (to allow irrigation from a nearby canal), based on existing farmer land allocations: there is no flexibility to change future irrigation patterns if land use changes. 5.66. In drought-affected areas farmers and urban water authorities may seek greater certainty in production or consumption by building dams. Where water trading markets exist, farmers or water authorities may purchase rights to water for the coming season if they anticipate specific needs. However, a cheaper and more flexible alternative that achieves the same result is to enter into contractual agreements that provide an option to purchase water from other owners of water rights. Michelsen and Young (1993) model a case study of farmers in northeastern Colorado selling options over water rights to municipal and industrial users. In a similar analysis using Wangaratta, a large regional town in Australia, as a case study, Leroux and Crase (2010) consider that the use of options contracts can reduce political friction in the allocation of water. ABARE (2005) explores a similar approach with environmental managers purchasing water options from irrigators as a means of ensuring adequate environmental flows in rivers. 5.67. Measures incorporating real options can also be applied on a regional, transnational basis. Cross-border sharing of water in systems like the Mekong is an obvious case where the use of water options might be helpful once water markets are established. Another example might be the purchase by a country such as Vietnam of an option to acquire weather information and storm warnings from other regional economies during the typhoon season: a cheaper alternative fiscally compared to the development of an indigenous system. 5.68. Applying the real options approach to the built environment requires a degree of innovation and creativity. In particular, a willingness to move away from obvious deterministic solutions is essential. The real options approach is unlikely to be fully compatible with established budgetary precepts or current project management practices, so complementary institutional change may be required to implement it. The New York City approach of iterative Flexible Adaptation Pathways acknowledges the ineffectiveness of instituting an inflexible set of climate policies, no matter how stringent (Yohe and Leichenko, 2010, p. 33). Instead, adaptation policies are to be monitored and interim targets adjusted at regular intervals as new information about impacts comes to light. The flexibility embodied in not committing fully to a single strategy at the outset results in a form of hedging that is akin to a real options approach. 5.5 Instruments for adaptation 5.69. In discussing real options, this report has already started to shift the discussion from decision- tools to instruments since the real-options approach is useful for both: as a way of making decisions under uncertainty, and as a principle for project design. The last section of this chapter considers three types of instruments governments will have available to them to help their societies adapt to climate change: the provision of public goods, public pricing policies, and financial instruments. The discussion is not exhaustive. Governments might also want to provide cash transfers and subsidies to households that have been disadvantaged by climate change, and planning and land-use regulations are obviously all of critical importance for successful adaptation. But the three which are focused on are all important instruments, with applicability to all economies. While public goods funding has received considerable attention, pricing reform and financial instruments have been less prominent in adaptation discussions (Box 5.6). Public goods funding and pricing reform are clearly fiscal instruments. Financial instruments are often substitutes for subsidies (and vice versa), which are fiscal in nature. 89 Box 5.6: Adaptation in rural Ningxia Hui Autonomous Region Situated in northwest China and contiguous with Inner Mongolia and Gansu and Shanxi Provinces, the Ningxia region straddles three climatic zones. Most rain falls in the mountainous south where annual precipitation is above 400mm, less falls in the flat central part (250-400mm), and the northern section (<250mm) relies on irrigation from the upper reaches of the Yellow River which runs through it. Irrigation in the north allows production of a range of grain crops and animals, the arid central part supports only corn, spring wheat, potatoes and some animals, and the potatoes are the main output of the rain fed southern mountains. In collaboration with a number of number of Chinese and English government agencies, the Chinese Academy of Agricultural Sciences and the University of East Anglia, undertook detailed modeling of future climatic conditions and expected agricultural output, as well as making recommendations for adaptation measures based on surveys of farmers. Xiong et al. (2008) report that modeling suggests that lower cereal production due to water constraints will be largely offset by increased (atmospheric) carbon fertilization and adaptation measures. On the other hand, the PRECIS model used to provide regional climate inputs appears to simulate wetter conditions than a multi-climate model average for China. The various reports issued as a result of the Ningxia study recommend a number of adaptation measures, including cloud seeding, water capture in terraced fields and water cellars, increasing the area of irrigated land, more efficient irrigation, migration of farmers to peri-urban areas, reduction in loss of soil moisture, training to increase skills of farmers seeking off-farm income, increased government support for research and development, and direct government cash handouts as income support are the responses considered (Li et al., 2008, and Lin et al., 2008). Little attention is given in the Ningxia study to fiscal instruments of adaptation. Provision of public goods is recommended (through increased R&D), but pricing water to reflect its scarcity value does not appear to be discussed, nor are other potential measures ­ such as index insurance ­ that would foster experimentation and encourage flexible responses by individual farmers given any prominence. Consistent with this, in a study of willingness to pay for water in Chongqing in south-western China, Wang et al. (2008) conclude that there was clearly room for a water price increase in the surveyed areas but that special attention would need to be given to the needs of the poor and those opposed to price increases. 5.5.1 Public goods 5.70. Markets will under-provide public goods, and there is a clear role for government provision. There are three major types of public goods62 relevant to adaptation in countries such as China and Vietnam: the construction of dikes and sea walls; disaster management, including warnings of extreme weather events; and general research, particularly into new crop varieties. The first of these has already been discussed in relation to cost-benefit analysis in Sections 5.3 and 5.4. This section focuses on the latter two. 5.71. China is already using pamphlets and text messages to convey information to the populace about climatic conditions. According to the Ministry of Agriculture (2009, p. 69), more than 100 million pamphlets containing information about techniques for fighting the freezing weather [in southern China for an unusually prolonged period in early 2008] and changing irrigated farming to dry farming were handed out and more than 100 million farmers were trained. The Chinese Ministry of Agriculture (2009, p. 65) also sent about 200 million short text messages to better guide farmers in areas affected by the low temperatures, heavy rains and snow and freezing weather in South China and in the earthquake-hit 62 Pure public goods are non-rival, non-excludable, and provide services to additional users at zero marginal cost. Although the term public goods` is used here, more accurately one should say these are goods with public-good characteristics. Research, dikes and storm warnings may be closer in nature to club goods because there may be some element of excludability present. 90 Wenchuan area of Sichuan Province ... These are relatively cheap measures: simple estimates suggest a cost of 100m Yuan for the pamphlets, and perhaps 5m Yuan for the text messages. 5.72. Vietnams system of weather warnings does not appear to be as developed as Chinas. Anecdotal material suggests that one reason is that weather forecasting is not particularly advanced. Warnings are issued by public broadcasting via radio and television. 5.73. Agricultural research would seem to warrant more attention. Neither the Chinese progress report on actions for addressing climate change (NDRC 2009) nor the Vietnamese blueprint (Nguyen 2008) appear to specify research into local agricultural crops. In both cases, research appears to be focused more on general climate change impacts, sea level rise, warning systems, meteorology, health issues, etc. While these are undoubtedly of greater consequence in the short term, longer-term adaptation policy would merit devotion of some resources to agricultural research specific to each economy. 5.5.2 Pricing 5.74. Water pricing reform will be essential for successful adaptation. There is a clear parallel to the importance for mitigation of pricing reform in the energy sector (as discussed in Chapter 4). 5.75. China faces a particularly challenging task in reconciling climate change impacts on water supply, a growing total demand for water, and equity issues associated with its distribution. Xie et al. (2009) review in detail many of the key issues to do with water scarcity in China, including inefficient usage rates in terms of GDP (p. 35) compared to developed economies. Although water tariffs in China have been increasing since the early 1990s, water appears to be seriously underpriced in most regions and cities and subsidies from general government revenues are still the norm (Xie et al., 2009, p. 84). Low tariffs are partially responsible for the lack of finance that regional governments can generate to improve and expand water and sewerage treatment infrastructure. However, China is experimenting with higher water charges, and the results are encouraging (Box 5.7). Box 5.7: Agricultural responsiveness to increased water prices in China Until relatively recent times, farmers in irrigation districts of China have been able to access sufficient water, invariably at very low prices. More realistic pricing of water would be highly likely to encourage structural adjustment, with water being used in production activities where it adds the most value. A study by Liu et al. (2008) of three very different villages on or near the Yellow River that experienced water shortages is instructive. After an initial stage of one-time adjustments such as watering crops with saline or polluted water, farmers turned to sinking wells and trucking in water. Although partly subsidized by local authorities, farmers were faced for the first time with significantly increased costs. As well as being expected to contribute their labour to collective water projects, they were faced with the costs of pumping water from irrigation channels and wells, and, in some instances, transporting it by truck. As water shortages worsened in the three villages, the cost to farmers of extracting water increased further. Enterprising farmers responded by switching production to more water-efficient alternatives such as growing tomatoes in greenhouses, breeding pigs, and collecting winter dates from neighboring villages for processing and resale to urban areas. Although more profitable, these activities required substantial capital investments for greenhouses, deeper wells and processing equipment (Liu et al., 2008, p. 549). The success of the three villages suggests that better pricing of water in future ­ underpinned by access to credit facilities (see Section 5.5.3 below) ­ would be a realistic and efficient means of fostering autonomous adaptation to climate change. 91 5.76. While significantly increased water tariffs would adversely affect the poor, current low standards of water quality and service levels also disproportionately affect less affluent areas (Xie et al., 2009, p. 84). Two-part water tariffs offer a potential solution, with a first, low-priced block allocated for daily needs and the second reflecting full marginal social costs. Practical issues of migrant worker populations and differing household size would, however, need to be addressed in such a scheme. Poor households, for example, are generally larger in size than rich ones and would therefore require a larger block allocation for daily needs. 5.77. Another possibility that may be worth exploring in greater detail is that of central government loans to provincial and city governments that wish to invest in improved infrastructure. For example, replacement of leaking pipes would reduce supply costs, with resultant savings available to repay loans. Better infrastructure may also attract industry, thus increasing local government revenues. A demonstrably improved water service is also likely to make any increase in tariffs socially and politically more acceptable, with additional revenues being made available for loan repayments. 5.5 3 Financial instruments 5.78. Credit facilities and insurance are important for both adaptation and development. The ability to borrow allows consumption smoothing over the period of the loan, and facilitates investment for the generation of future income. Insurance, which can be characterized as a financial put option, provides the right to compensation where damage is incurred for a contractually specified activity or asset. In the context of climate change in the developing economies of APEC, insurance and credit are generally most relevant to the agricultural sector, where the vicissitudes of weather and climate are currently clearly evident. 5.79. These financial instruments are potential alternatives to fiscal ones, and in particular alternatives to subsidies from the government to private individuals. Many governments help farmers who face uncertain weather by providing them with subsidies when the weather is bad. However, the fiscal implications can be significant, especially if charity hazard` (the Samaritan`s Dilemma`) leads to an entitlement mentality regarding government assistance. In Australia, for example, drought relief for farmers has effectively resulted in a degree of long-term income dependence in some areas. Rural credit and insurance schemes are obvious alternatives to cash hand-outs. Any subsidies to rural credit and insurance will obviously incur fiscal costs, but there will be savings in terms of reduced income assistance. Credit and income-contingent loans 5.80. Many developing countries struggle to get an adequate access of credit to rural households. The World Bank (2010h) notes of China, for example, that Access to medium and long-term loans continues to be a considerable constraint for rural households. And Marsh et al. (2004, Abstract) conclude that small household farms in Vietnam, and the rural sector in general, are recognized as facing severe credit restrictions.63 5.81. Rural banks faced with many small borrowers wanting loans are likely to incur high administrative costs because of the number of credit checks and monitoring inspections that are 63 Anecdotal material suggests that the nature of Vietnamese land reforms in recent years is partly responsible for the relatively low level of commercial lending to farmers. Land continues to be owned by the state, but farmers now produce crops under contractual agreements with provincial governments, in return for leases over arable land. However land that is leased to farmers is provided in parcels that are as homogenous as possible, in order to facilitate estimation of the value of leases (on the basis of potential income generation). Homogeneity within land parcels means that the parcels are smaller than they might otherwise have been. Administrative costs to commercial banks that lend to farmers are therefore relatively high because of the larger number of different holdings. 92 required. Further, income derived from small loans is generally not high, especially if interest rates are subject to government control or suasion. Where this is the case, commercial banks may use administrative measures ­ such as understaffing rural branches, or forcing potential rural borrowers to apply in larger towns ­ to minimize the number of small, unprofitable loans. 5.82. Microcredit is being increasingly explored as a means of providing greater opportunity for farmers and small businesses in developing economies. It is generally used to provide small loans for short periods to the economically active` poor. Its key distinguishing feature compared to conventional banking practice is that security is not required against a loan. However, repayment schedules have traditionally been strict, with subsequent loans dependent on a borrower`s repayment record. Microcredit is increasingly spreading throughout Asia. Bank Rakyat Indonesia (BRI) is the world's largest and most profitable microfinance network in the world with some 30 million savers catered to by 4000 branches (BWTP, undated) In China the Agricultural Bank of China (ABC) now extends small loans to 15 million farmers (Box 5.8). Box 5.8: Small rural loans in China: redressing the balance An 18 July 2010 Xinhua item illustrates the imbalance between rural and urban lending in China by a major bank, as well as the important institutional initiative of allowing farmers (as individuals) to act as guarantors: Agricultural Bank of China (ABC) loaned 82.9 billion Yuan (12.2 billion US dollars) in micro-credit to Chinese farmers in the first three months of the year, compared with 67.3 billion Yuan in the whole of 2009. Starting in April 2008, ABC's micro-credit business, which makes loans of 3,000 to 50,000 Yuan in size, has helped 15 million farmers across the country, Lu Chuan, an ABC official told Xinhua Saturday. Farmer Bu Nianhua, a grape grower, is happy these days. On his 8.5-acre grape farm he has seven breeds of grape. They all ripen at different times, between July and November and he earns an annual income of 200,000 Yuan. But thanks to a 30,000 Yuan micro-credit loan from the local ABC branch at the end of 2009, Bu has planted a new high-yield grape breed - Minicure Finger. The half acre of Minicure Finger grapes alone will bring him 75,000 Yuan in income this year. In a break with the past, Bu's loan was guaranteed by two other farmers, instead of by enterprises. "Like Bu, many other farmers have received financial support from ABC through an innovative credit guarantee system," said Lu. .. ... However, the loans' value is only two per cent of the four trillion Yuan ABC loaned in 2009, said Zhao Baige, deputy director of the National Population and Family Planning Commission. Given there are 720 million residents in rural China, the beneficiaries of the loans merely account for two per cent of the country's rural population, Zhao added. Up until now, almost 3,000 rural townships in China ­ around nine per cent of the country's total ­ have not had a banking branch, said Tang Renjian, deputy director of the Central Finance Leading Group office. 5.83. Microfinance can also be used to facilitate adaptation. Microfinance is obviously not suitable for large adaptation projects such as building dams or sea wall. But research has revealed that in some economies microfinance is already delivering credit to areas that will be critical for adaptation. For example, in Bangladesh agriculture, disaster relief and preparedness, and water and sanitation ­ which are all particularly affected by climate change ­ constitute almost 70% of the existing microfinance portfolio (Agrawala and Carraro, 2010, p.9). They suggest (p. 32) that microfinance could play a greater role in disaster preparedness and early warning systems, in promoting crop varieties that might be more resilient to the anticipated impacts of climate change, and in technical training and education programs related to community level adaptation. 5.84. An important problem common to both commercial bank lending and microfinance is that strict, time-based repayment schedules are enforced. But specified periods for repayment may not match rural crop cycles, so that a farmer may be forced to repay at least part of a loan before harvesting and selling a 93 crop. If farmers are forced to sell crops before harvest for lower prices, they may experience problems generating sufficient income. 5.85. Where conventional lending proves to be too risky in terms of expected income, or income streams are uncertain over time, income contingent loans may provide a pragmatic alternative. Unlike conventional loans, income contingent loans do not entail fixed repayment schedules. Nor do they require security against default as a condition of the loan, as required by conventional banks. 5.86. Normally provided by governments, repayment of income contingent loans depends only on the borrower achieving a specified minimum income level. Loans are repaid only when income levels are considered sufficient; otherwise no repayment is required (Chapman, 2006). An example proposed for, but not (yet) implemented in Australia is drought relief for farmers (Botterill and Chapman, 2006). However the classic example is lending to tertiary education students in Australia (and other economies), with graduates repaying their loan through the taxation system once their earnings reach a certain level. 5.87. There may be scope for the creation of commercial or subsidized financial instruments that draw on the concept of income contingent loans to assist rural development and adaptation to climate change. However, income contingent loans require careful design and monitoring to avoid potential problems of moral hazard and adverse selection. Equally important, it would be important to ensure that hybrid income contingent loan instrument did not inhibit the growth of existing commercial credit markets. Insurance 5.88. Well-designed insurance can be an ideal adaptation instrument because it creates an incentive for the individual to reduce their exposure to a hazard until the premium they pay just equals the expected residual value of any loss. Where possible, it is therefore preferable to other financial instruments such as grants or low interest loans that are often provided as part of disaster relief. Subsidized assistance may discourage individuals from taking preventative action to reduce the hazards they face. 5.89. However, the extent to which insurance can be used to adapt to climate change is not yet clear. Effects such as coastal inundation involve correlated risk, often for many households simultaneously. Whether such risks can be diversified, even on a global basis, is not yet entirely clear. However, Schanz (2010) argues that insurers and re-insurers will be able to cope with continued offerings of weather- related insurance, even in the face of climate change. He argues that there is sufficient regional diversity and independence of risk, and, if insurers are able to charge premiums accordingly, with no government intervention or rate-capping, insurance will continue to be viable. 5.90. Another challenge that does not yet appear to have been faced explicitly by insurers and governments is the ,,slow onset nature of some aspects of climate change. For example, riverine salinity or coastal inundation or erosion may occur relatively slowly, in decadal timescales. Such events may be difficult to classify as being fortuitous`, a cardinal principle of insurance cover, because they are already apparent and represent a so-called loss in progress`. 5.91. But even if some climate change risks cant be insured against, many households are already underinsured. Closing this gap would help them adapt. Where feasible, index insurance (also called group risk insurance` in the USA: e.g., Barnett 1999) can be used to assist individual farmers to adapt to the vagaries of climate change. Because such insurance is based on pre-agreed weather indexes such as the amount of rainfall over a defined area during a given period, and payouts are not linked to crop survival or failure, individual farmers retain an incentive to maximize crop output. Some developing economies have started to pilot index insurance schemes (Barnett et al., 2008). Catastrophe insurance programs are also starting to get more attention (Cummins and Mahul, 2008). 94 5.92. The main role for government in relation to both rural credit and insurance is the development of a sound policy framework. For example, macroeconomic stability is critical, and sound regulatory frameworks are also important. But government funding can also be important, for example, in starting- up or expanding microfinance and insurance schemes. Whether subsidies are needed or warranted can only be decided on a case by case basis. Where they are provided, care should be taken to keep the product at arms` length from government to prevent politicization and the undermining of a repayment and commercial culture. 95 References Abiad, A, Detragiache, E, and Tressel, T 2008, A new database of financial reforms, IMF Working Paper WP/08/266, IMF Institute and Research Department. Acemoglu, D, Aghion, P, Burstyn, L, and Hemous, D 2009, The environment and directed technical change, NBER Working Paper No. w15451, National Bureau of Economic Research. Agrawala, S, and Carraro, M 2010, Economic aspects of adaptation to climate change: assessing the role of microfinance in adaptation, Working Party on Global and Structural Policies, OECD Environment Policy Committee, NV/EPOC/GSP(2009)12/FINAL, 2 February. Agrawala, S, and Fankhauser, S, 2008, Economic aspects of adaptation to climate change: costs, benefits and policy instruments, OECD, Paris. Agrawala, S, Crick, F, Jetté-Nantel, S and Tepes, A 2008, Empirical estimates of adaptation costs and benefits: key insights and limitations, presentation to OECD Workshop on Economic Aspects of Adaptation, Paris, April 7-8, 2008, available at http://www.oecd.org/dataoecd/21/48/40896691.pdf Agustina, CRD, Granado, JA, Bulman, T, Fengler, W, and Ikhsan, M 2008, Black hole or black gold? The impact of oil prices on Indonesia`s public finances, Policy Research Working Paper no. 4718, World Bank, Washington, DC. Anda, J, Golub, A, and Strukova, E, 2009, Economics of climate change under uncertainty: benefits of flexibility, Energy Policy, vol. 37, p. 1345-1355. Arnold, J, 2008, Do tax structures affect aggregate economic growth? Empirical evidence from a panel of OECD countries, OECD Economics Department Working Paper No 643, OECD, Paris. Asian Development Bank 2009, The economics of climate change in Southeast Asia: a regional review, Manila. Australian Bureau of Agricultural and Resource Economics (ABARE) 2005, Water options for environmental flows, November, Commonwealth of Australia, Canberra. Bacon, R, Bhattacharya, S and Kojima, M 2010, Expenditure of low-income households on energy: evidence from Africa and Asia, Extractive Industries and Development Series, World Bank, Washington, DC. Badan Perencanaan dan Pekbangunan Nasional (Bappenas) 2009, The Indonesia climate change sectoral roadmap, December, Republic of Indonesia, Jakarta. Baig, T, Mati, A, Coady,D, and Ntamatungiro, J 2007, Domestic petroleum product prices and subsidies: recent developments and reform strategies, IMF Working Paper WP/07/71, Washington, DC. Barnett, B 1999, US Government Natural Disaster Assistance: historical analysis and a proposal for the future, Disasters, vol. 23, no. 2, pp. 139-155. Barnett, BJ, Barrett, CB, Skees, JR 2008, Poverty traps and index-based risk transfer products, World Development, vol. 36, pp. 1766-1785. Baylor M, and Beauséjour, L 2004, Taxation and economic efficiency: results from a Canadian CGE model Canadian Department of Finance Working Paper 2004-10 , Ottawa, November, available at http://www.fin.gc.ca/wp/2004-10-eng.asp. 96 Besant-Jones, JE 2006, Reforming power markets in developing countries: what have we learned?, Energy Mining Sector Board Discussion Paper no.19, World Bank, Washington DC. Boardman, AE, Greenberg, DH Vining, AR and Weimer, DL, 2006, Cost-benefit analysis: concepts and Practice, 3rd edition, Pearson Prentice Hall, New Jersey. Bor, YJ and Huang, Y 2010, Energy taxation and the double dividend effect in Taiwan`s energy conservation policy ­ an empirical study using a computable general equilibrium model, Energy Policy, vol. 38, pp. 2086-2100. Botterill, L and Chapman, B 2006, Turning grants into loans: income contingent loans for drought relief, Ch. 7 in Chapman, B. (ed.) 2006, Government managing risk: income contingent loans for social and economic progress, Routledge, UK. BP 2010, Statistical review of world energy 2010, June, available at http://www.bp.com/liveassets/bp_internet/globalbp/globalbp_uk_english/reports_and_publications/statisti cal_energy_review_2008/STAGING/local_assets/2010_downloads/statistical_review_of_world_energy_f ull_report_2010.pdf Brandt, U, and Svendsen, G 2006 Climate change negotiations and first-mover advantages: the case of the wind turbine industry`, Energy Policy, vol. 34, pp. 1175­1184. Burer, MJ, and Wustenhagen, R 2009 Which renewable energy policy is a venture capitalist's best friend? Empirical evidence from a survey of international cleantech investors, Energy Policy, vol. 37, pp. 4997-5006. Burke, P 2010, The energy supply ladder and its carbon implications, Australian National University, mimeo, August. Burniaux, J.-M. Chateau, J, Duval, R and Jamet, S 2008, The economics of climate change mitigation: policies and options for the future, Economics Department Working Paper, ECO/WKP(2008)66, OECD, Paris. BWTP (Asia Resource Centre for Microfinance) undated, Bank Rakyat Indonesia, available at http://www.bwtp.org/arcm/indonesia/II_Organisations/MF_Providers/Print_Versions/BRI_Indonesia.pdf CGAP 2004, The role of governments in microfinance, Donor Brief No. 19, available at http://www.cgap.org/gm/document-1.9.2371/DonorBrief_19.pdf China Electricity Council 2010. 20159.33 (China's coal-power generation capacity will reach 933GW in 2020), available at http://www.cec.org.cn/news/showc.asp?id=132468. China Wind Energy Association, 2010, China wind power installed capacity, (2009 statistics), Available at: http://www.cwea.org.cn/upload/201006102.pdf. Coady, D, Gillingham, R, Ossowski, R, Piotrowski, J, Tareq, S and Tyson, J 2010, Petroleum product subsidies: costly, inequitable, and rising, IMF Staff Position Note SPN/10/05, International Monetary Fund, Washington DC. Cohen LR, Noll RG 1991, The technology pork barrel, Brookings Institution. Washington DC. Commission of the European Communities 2008, The support of electricity from renewable energy sources, Commission Staff Working Document, Brussels. Consultative Group on International Agricultural Research Science Council 2008, Bio-fuels research in the CGIAR: a perspective from the Science Council, A CGIAR Science Council Policy Statement on Biofuels Production, 24 April. 97 Cox, LA 2009, What`s wrong with hazard-ranking systems? An expository note, Risk Analysis, vol. 29, no. 7, pp. 940-948. Cummins, D, Mahul, O 2008, Catastrophe risk financing in developing countries: principles for public intervention, World Bank, Washington DC. Dechezleprêtre, A, Glachant, M, Hascic I, Johnstone, N, and Ménière, Y 2008, Invention and transfer of climate change mitigation technologies on a global scale: a study drawing on patent data, CERNA Working Series, available at: http://ssrn.com/abstract=1414227. Department of Climate Change and Energy Efficiency 2010, Coastal inundation at Narrabeen Lagoon: optimising adaptation investment, Report prepared for the Department by AECOM, Commonwealth of Australia, Canberra, available at http://www.climatechange.gov.au/~/media/publications/adaptation/coastal-flooding-narrabeen- lagoon.ashx. Devarajan, S, Miller, M and Swanson. E 2002. Goals for development: history, prospects and costs. Policy Research Working Paper 2819, Washington DC. Dixit, AK and Pindyck, RS, 1994, Investment under uncertainty, Princeton University Press, New Jersey. Dobes, L 2008, Getting real about adaptation to climate change: using real options` to address the uncertainties, Agenda, vol. 15, no. 3, pp. 55-69. Dobes, L 2009, People versus planners: social preferences for adaptation to climate change, Research Report no. 41, Environmental Economics Research Hub, Crawford School of Economics and Government, The Australian National University, Canberra. Dobes, L and Bennett, J 2009, Multi-criteria analysis: good enough` for government work?, Agenda, vol. 16, no. 3, pp. 7-29. Downes P 2007, ASEAN fiscal and monetary policy responses to rising oil prices, ASEAN Regional Economic Policy Support Facility Report 06/004, May, available at http://www.asean.org/aadcp/repsf/publications.html. DSIRE, 2010 Database of state incentives for renewables and efficiency, http://www.dsireusa.org. Du, J, Tao, Z and Lu, Y 2008, Bank loans and trade credit under China's financial repression, Working Paper, available at SSRN: http://ssrn.com/abstract=1495600. Dun, O. 2010 (forthcoming), Migration and displacement triggered by floods in the Mekong Delta, International Migration [accepted for publication 7 December 2009]. Elekdag, S, Lalonde, R, Laxton, D, Muir, D and Pesenti, P 2008, "Oil price movements and the flobal economy: A model-based assessment," IMF Staff Papers, Palgrave Macmillan Journals, vol. 55(2), pp. 297-311. Ellerman, D Convery, F and de Perthuis, C 2010, Pricing carbon: the European Union Emissions Trading Scheme, Cambridge University Press, Cambridge. Energy Information Administration (EIA) 2010, International energy statistics, US Energy Information Administration, available at http://tonto.eia.doe.gov/cfapps/ipdbproject/IEDIndex3.cfm?tid=2&pid=2&aid=7 Energy Modelling Forum (EMF) 2009, international database available at http://emf.stanford.edu/events/emf_briefing_on_climate_policy_scenarios_us_domestic_and_internationa l_policy_architectures Estache, A and Wren-Lewis, L 2009, Toward a theory of regulation for developing countries: following Jean-Jacques Laffont`s lead, Journal of Economic Literature, vol. 47, no. 3, pp. 729-770. 98 Frondel, M, Ritter, N, Schmidt CM, and Vance, C 2010, Economic impacts from the promotion of renewable energy technologies: the German experience, Energy Policy, vol. 38, pp. 4048­4056. Fuessel, H-M 2009, Review and quantitative analysis of indices of climate change exposure, adaptive capacity, sensitivity, and impacts, Background Note to World Development Report 2010: Development and Climate Change, Potsdam Institute for Climate Impact Research. Fullerton, D and Metcalf, G 1997, Environmental taxes and the double-dividend hypothesis: did you really expect something for nothing?, NBER Working Paper No. 6199, Cambridge MA. Garnaut, R 2008, The Garnaut climate change review, Cambridge University Press, Cambridge. Garnaut, R, Howes, S, Jotzo, F, and Sheehan, P 2009, The implications of rapid development for emissions and climate-change mitigation, Ch. 5 in Helm, D and Hepburn, C (eds.), The Economics and Politics of Climate Change, Oxford University Press, Oxford Gemenne, F, 2010, Migration, a possible adaptation strategy?, Syntheses, no. 3, 10 June, Institut du Developpement Durable et des Relations Internationales, Paris. (Translated from an article first published in Liaison Energie-Francophone, no. 85, October 2009). Gesner, GA, and Jardim, J 1998, Bridge within a bridge, Civil Engineering, vol. 68, no. 10, pp. 44-47. Goulder, LH 1992 Do the costs of a carbon tax vanish when interactions with other taxes are accounted for?, NBER Working Paper No. 4061, Cambridge MA. Government of Korea Presidential Committee on Green Growth 2009, National strategy`, available at http://www.greengrowth.go.kr/english/en_policy/en_strategy/en_strategy.cms Grainger, A 2008, Difficulties in tracking the long-term global trend in tropical forest area, PNAS, vol. 1095, no. 2, pp. 818­823 January 15. Green Investment Bank Commission 2010, Unlocking investment to deliver Britain's low carbon future, London. Gren, I-M 2010, Climate change and the Water Framework Directive: cost effectiveness in the Swedish Malar region, Climatic Change, vol. 100, pp. 463-484. Grieg-Gran, M 2006, The cost of avoiding deforestation: report prepared for the Stern Review on the Economics of Climate Change, International Institute for Environment and Development. Grubb, M 2004, Technology innovation and climate change policy: an overview of issues and options, Keio Economic Studies, vol. 41, pp. 103­32. Hahn, M, Riederer, AM, and Foster, O, 2009, The livelihood vulnerability index: a pragmatic approach to assessing risks from climate variability and change ­ a case study in Mozambique, Global Environmental Change, vol. 19, pp. 74-88. Hawke, A 2010, Review of the administration of the home insulation program, available at http://www.climatechange.gov.au/~/media/publications/energy-efficiency/Home-Insulation-Hawke- Report.ashx He J and Kuijs, L 2007, Rebalancing China`s economy ­ modelling a policy package, World Bank China Research Paper No.7, September, World Bank, Washington, DC Howes, S 2009a, Sustaining growth and mitigating climate change: are the costs of mitigation underestimated?, paper presented to the July 2009 East-West Center Conference on Climate Change and Green Growth: Korea`s National Growth Strategy. Howes, S 2009b, Cheap but not easy: the reduction of greenhouse gas emissions from deforestation and forest degradation in Papua New Guinea, Pacific Economic Bulletin, vol. 24(1), pp. 130-143. 99 Howes, S 2010, China`s energy intensity target: on-track or off?, East Asia Forum, March 31, 2010, available at http://www.eastasiaforum.org/2010/03/31/chinas-energy-intensity-target-on-track-or-off/ Huang, Y 2010, China`s great ascendancy and structural risks: consequences of asymmetric market liberalization, Asian-Pacific Economic Literature, vol. 24, pp. 65-85. Independent Evaluation Group (IEG) of the World Bank (2009) Climate change and the World Bank Group: Phase I ­ an evaluation of World Bank win-win energy policy reforms, World Bank, Washington, DC, available at http://lnweb90.worldbank.org/servlet/OEDSearchServletV656?SearchType=byText&PerPage=20&DbU RL=oed/oeddoclib.nsf&Form=OEDSearch&SearchText=Phase%20I:%20An%20Evaluation%20of%20 World%20Bank,%20Win-Win%20energy%20policy%20reforms International Energy Agency (IEA) 2006a, World energy outlook 2006, Paris. International Energy Agency (IEA) 2006b, China`s power sector reforms: where to next?, Paris. International Energy Agency (IEA) 2008a, World energy outlook 2008, Paris. International Energy Agency, 2008b, Energy policy review of Indonesia, Paris. International Energy Agency (IEA), 2009a, CO2 emissions from fuel combustion, vol. 2009 release 01 -, available at http://oberon.sourceoecd.org/vl=15178979/cl=15/nw=1/rpsv/statistic/s26_about.htm?jnlissn=16834291. International Energy Agency (IEA), 2009b, Trends in photovoltaic applications: Survey report of selected IEA countries between 1992 and 2008, Report IEA PVPS T1-18:2009, available at: http://www.iea-pvps.org/products/rep1_18.htm. International Energy Agency (IEA), 2009c, Global gaps in clean energy research, development, and demonstration, prepared in support of the Major Economies Forum (MEF) Global Partnership, Paris. International Energy Agency (IEA) 2009d, Energy balances of Non-OECD Member Countries ­ Extended Balances, Vol. 2009 Release 1, Paris. International Energy Agency (IEA) 2009e, Energy balances of OECD Countries ­ Extended Balances, Vol. 2009 Release 1, Paris. International Energy Agency (IEA) 2010a, The electricity access database, available at: http://www.iea.org/weo/database_electricity/electricity_access_database.htm International Energy Agency (IEA) 2010b, Energy prices and taxes ­ energy end-use prices, (US/toe, PPP/unit), Vol. 2010 release 02 -, available at http://puck.sourceoecd.org/vl=16663138/cl=15/nw=1/rpsv/statistic/s29_about.htm?jnlissn=1683626x International Energy Agency (IEA) 2010c, Energy Technology RD&D 2010 Edition, available at: http://www.iea.org/stats/rd.asp International Energy Agency (IEA) 2010d, Global fossil fuel subsidies and the impacts of their removal, Office of the Chief Economist, available at http://www.iea.org/weo/docs/energy_subsidies_slides.pdf International Monetary Fund (IMF), 2010a, World economic outlook database, available at http://www.imf.org/external/ns/cs.aspx?id=28. International Monetary Fund, 2008, Fuel and food price subsidies: issues and reform options, prepared by the Fiscal Affairs Department, available at: http://www.imf.org/external/np/pp/eng/2008/090808a.pdf IPCC (Intergovernmental Panel on Climate Change) 2007, Climate Change 2007: mitigation. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on 100 Climate Change. Metz, B, Davidson, OR, Bosch, PR, Dave, R, Meyer, LA (eds), Cambridge University Press, Cambridge. Ito, J. 2006, Economic and institutional reform packages and their impact on productivity: a case study of Chinese township and village enterprises, Journal of Comparative Economics, vol. 34(1), pp. 167- 190. Jaffe, AB, Newell, RG, and Stavins, RN 2002, Environmental policy and technological change, Fondazione Eni Enrico Mattei Nota di Lavoro, 26.2002. Jakarta Updates 2010, Indonesian government is resolute in fixing its broken electricity system, April 18, available at http://www.jakartaupdates.com/2010/04/indonesian-government-is-resolute-in-fixing-its- broken-electricity-system Johansson, A, Heady, C, Arnold, J, Brys, B, and Vartia, L 2008, Tax and economic growth, OECD Economics Department Working Paper No.620, Paris, available at http://www.oecd.org/officialdocuments/displaydocumentpdf/?cote=ECO/WKP(2008)28&doclanguage=e n Johnstone, N, Hascic, I, Popp, D 2010, Renewable energy policies and technological innovation: evidence based on patent counts, Environmental and Resource Economics, European Association of Environmental and Resource Economists, vol. 45(1), pp. 133-155 Jotzo, F, Resosudarmo, IAP, Nurdianto, DA, and Sari, AP 2009, Climate change and development in Eastern Indonesia, in Resosudarmo, BP and Jotzo, F (eds.) 2009, Working with nature against poverty: development, resources and the environment in Eastern Indonesia, Institute of South-East Asia Studies, Singapore. Justice, S 2009, Private financing of renewable energy: a guide for policymakers, UNEP, SEFI, Bloomberg New Energy Finance, Chathem House, available at: http://www.energy- base.org/fileadmin/media/sefi/docs/publications/Finance_guide_FINAL-.pdf. Kang, H 2010, Korea`s wind energy industry eyeing overseas markets, SERI Quarterly, July 9, available at www.seriquarterly.com/srch/wldSrch_1.html?kw=Korea%A1%AFs%20Wind%20Energy%20Industry% 20Eyeing%20Overseas%20Markets. Kneese, AV and Schultze, CL 1975, Pollution, prices, and public policy, a study sponsored jointly by Resources for the Future, inc. and the Brookings Institution Washington, DC. Knight, E (2010) The economic geography of clean tech venture capital, Working Paper Series in Employment, Work and Finance, School of Geography and the Environment, University of Oxford. Koh, D 2004, Illegal construction in Hanoi and Hanoi`s wards, European Journal of East Asian Studies, vol. 3, no. 2 Kojima, M 2009, Government response to oil price volatility: experience of 49 developing countries Extractive Industries for Development Series #10, World Bank, Washington, DC. Koo, J 2010, Supporting green companies and industries through banks, Korea Institute of Finance, draft. KOPENINDO: Indonesian Renewable Energy Cooperative, 2008, Final report: Wangan Aji micro hydro power project Indonesia, Asian Development Bank, Manila, available at: http://www.adb.org/documents/produced-under-ta/37715/37715-04-reg-dpta.pdf Koplow, D 2009. Measuring energy subsidies using the price gap approach: what does it leave out?, International Institute for Sustainable Development, Global Subsidies Initiative, Geneva, available at www.iisd.org/pdf/2009/bali_2_copenhagen_ff_subsidies_pricegap.pdf. 101 Kornai, J, Maskin, E and Roland, R 2003, Understanding the soft budget constraint, Journal of Economic Literature, vol. 41, pp.1095-1136. Lambie, R, 2010, Understanding the effect of an emissions trading scheme on electricity generator investment and retirement behaviour: the proposed Carbon Pollution Reduction Scheme, Australian Journal of Agricultural and Resource Economics, vol. 54, no. 2, pp. 203-217. Lee, B, Iliev, I, and Preston, F 2009, Who owns our low carbon future? Intellectual property and energy technologies, A Chatham House Report, Royal Institute of International Affairs, available at http://www.chathamhouse.org.uk/publications/papers/view/-/id/775/ Lee, H (2010) Regional mitigation costs from EMF22 models, Australian National University, mimeo. Leroux, A and Crase, L. 2010 (forthcoming, September), Advancing water trade: a preliminary investigation of urban-irrigation options contracts in the Ovens Basin, Victoria, Australia, Economic Papers, vol. 29, no. 3. Li, J 2010, The economic effect of raising petrol product prices ­ application of SIC-GE, paper presented to the International Conference on the SIC-GE Model and Policy Simulation, 15-16 April, 2010, Beijing. Li, L. 2008, Employment burden, government ownership and soft budget constraints: evidence from a Chinese enterprise survey, China Economic Review, vol. 19(2), 215-229. Li, Y, Wu, Y, Conway, D, Preston, F, Lin, E, Zhang, J, Wang, T, Jia, Y, Gao, Q, and Shifeng, JH, 2008, Climate and livelihoods in rural Ningxia: final report, AEA Group, UK. Lin, E, Conway, D, Li, Y, and Calsamiglia-Mendiewicz, S, (eds.) 2008, The impacts of climate change on Chinese agriculture ­ phase II. Climate change in Ningxia: scenarios and impacts. Technical report. Final report, AEA Group, UK. Liu, C, Golding, D, and Gong, G, 2008, Farmers` coping response to the low flows in the lower Yellow River: a case study of temporal dimensions of vulnerability, Global Environmental Change, vol. 18, pp. 543-553. Mao, Y, Sheng H, and Yang F 2008, The true cost of coal, Greenpeace, the Energy Foundation and WWF, available at www.eu-china.net/web/cms/upload/pdf/materialien/TCOC-Final-EN-08_10-28.pdf. Marsh, SP, MacAuley, TG and Anh, LH 2004, Credit use in farm households in Vietnam: implications for rural credit policy Contributed paper presented at the 48th Annual Conference of the Australian Agricultural and Resource Economics Society, 11-13 February, 2004, Melbourne http://www.aares.info/files/2004_marsh1.pdf McCawley, P 1970, The price of electricity, Bulletin of Indonesian Economic Studies, vol. 6, pp. 61-86. McKinsey 2007, A cost curve for greenhouse gas reduction, The McKinsey Quarterly, February. Mercados­Energy Markets International (2010), China power dispatch efficiency improvement: final technical report, prepared for Energy Sector Management Assistance Program, World Bank. Michelsen, AM and Young, RA 1993, Optioning agricultural water rights for urban water supplies during drought, American Journal of Agricultural Economics, vol. 75, November, pp. 1010-1020. Ministry of Finance, Republic of Indonesia, 2009, Indonesia green paper: economic and fiscal policy options for climate change mitigation in Indonesia, Jakarta. Mishan, EJ 1988, Cost-Benefit analysis. An informal introduction, 4th edition, Unwin Hyman, London. 102 Mitchell, C, Baukech, D, and Connor, PM 2006, Effectiveness through risk reduction: A comparison of the renewable obligation in England and Wales and the feed-in system in Germany, Energy Policy, vol. 34, pp.297­305. Mohan, D, and Sinha, S, undated, Vulnerability assessment of people, livelihoods and ecosystems in the Ganga Basin, World Wildlife Fund ­ India, New Delhi. Montgomery, D and Smith, AE 2007, Price, quantity and technology strategies for climate policy, in Schlesinger, M. et al. (eds.), Human-induced climate change: an interdisciplinary assessment, Cambridge University Press, Cambridge. Morse, RK, Rai, V and He, G 2009, Real drivers of carbon capture and storage in China and implications for climate policy, Program on Energy and Sustainable Development, Working Paper no. 88, Stanford University. National Bureau of Statistics of China 2010, 2009 China statistical yearbook, China Statistics Press, Beijing. National Development and Reform Commission (NDRC) 2007, China's National Climate Change Programme , People`s Republic of China, June, available at http://www.ccchina.gov.cn/WebSite/CCChina/UpFile/File188.pdf National Development and Reform Commission (NDRC) 2009, China's policies and actions for addressing climate change: the progress report 2009, Beijing, November, available at http://www.ccchina.gov.cn/WebSite/CCChina/UpFile/File571.pdf Nemet, GF 2006, Beyond the learning curve: factors influencing cost reductions in photovoltaics, Energy Policy, vol.34, pp. 3218­3232. Netherlands Environmental Assessment Agency [PBL] (2010) Database associated with Olivier, J, and Peters, J. No growth in total global CO2 emissions in 2009, Netherlands Environmental Assessment Agency, Bithoven, The Netherlands, available at http://www.pbl.nl/en/publications/2010/No-growth-in- total-global-CO2-emissions-in-2009.html Newbury, D, and Stern, N (eds.) 1987, The theory of taxation for developing countries, World Bank Research Publication, Oxford University Press, Oxford. Newell, RG 2009, Literature review of recent trends and future prospects for innovation in climate change mitigation, OECD Environment Working Papers no. 9, OECD, Paris. Nguyen, TD (Prime Minister) 2008, National target program to respond to climate change, Decision No. 158/QD-TTg, 2 December. Nguyen, VK 2006, An economic evaluation of dyke construction alternatives in Vietnam`s Mekong Delta: a case study in An Giang province, thesis in fulfilment of the degree of Master of Environmental Management and Development, Crawford School of Economics and Government, The Australian National University, Canberra, ACT, Australia. Nordhaus, W 2008, Question of balance: weighing the options on global warming policies, Yale University Press, New Haven, Connecticut. Ölz, S, and Beerepoot, M 2010, Deploying renewables in Southeast Asia: Trends and potentials`, IEA Working Paper no. 2010/6, International Energy Agency, Paris, France. Orlove, B, 2009, The past, the present, and some possible futures of adaptation, in Adger, WN, Lorenzoni, I, and O`Brien, K (eds.), 2009 Adapting to climate change: thresholds, values, governance, Cambridge University Press, Cambridge. Outlook Economics (2010a) China Energy Graphs, Canberra. 103 Outlook Economics (2010b) Macroeconomic and terms of trade impacts of carbon taxes, Canberra. Pahle, M 2010, Germany`s dash for coal: exploring drivers and factors, Energy Policy vol.38, pp.3431­ 3442. Parry, M, Arnell, N, Berry, P, Dodman, D, Fankhauser, S, Hope, C, Kovats, S, Nicholls, R, Satterthwaite, D, Tiffin, R, and Wheeler, T 2009, Assessing the costs of adaptation to climate change: a review of the UNFCCC and other recent estimates, International Institute for Environment and Development and Grantham Institute for Climate Change, London. PBL (see Netherlands Environmental Assessment Agency) Perch-Nielsen, S.L, 2010, The vulnerability of beach tourism to climate change ­ an index approach, Climatic Change, vol. 100, pp. 579-606. Pezzey, J and Park A Reflections on the double dividend debate: the importance of interest groups and information costs, Environmental and Resource Economics, vol. 11 pp. 539-555. Preston, BL, Smith, F, Brooke, C, Gorddard, R, Measham, TG, Withycombe, G, Beveridge, B, Morrison, C, McInnes, K, and Abbs, D 2008, Mapping climate change vulnerability in the Sydney Coastal Councils Group`, Report prepared for the Sydney Coastal Councils Group, CSIRO and Sydney Coastal Councils Group Inc, available at http://www.csiro.au/resources/SydneyClimateChangeCoastalVulnerability.html Ramanathan, V, and Carmichael, G 2008, Global and regional climate changes due to black carbon, Nature Geoscience , vol.1, pp. 221 ­ 227. Repetto, R and Easton, R 2009, Climate change and damage from extreme weather events, Working Paper no. 207, October, Political Economy Research Institute, University of Massachusetts, Amherst. Rogner, H.-H., D. Zhou, R. Bradley. P. Crabbé, O. Edenhofer, B.Hare (Australia), L. Kuijpers, M. Yamaguchi, 2007: Introduction in Climate Change 2007: Mitigation. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [B. Metz, O.R. Davidson, P.R. Bosch, R. Dave, L.A. Meyer (eds)], Cambridge University Press, Cambridge, United Kingdom and New York. Rosen, DH, Houser, T 2007, China energy: a guide for the perplexed, The Centre for Strategic and International Studies, and Peterson Institute for International Economics, Washington, DC. Rosenzweig, MB, Voll, SP, and Pabon-Agudelo, C 2004, Power sector reform: experiences from the road, The Electricity Journal, pp. 16-18, doi:/10.1016/j.tej.2004.10.002. Schanz, K-U 2010, Defining the limits of insurability ­ a continuous strategic challenge`, Middle East Insurance Review, June, p. 50. SEI: See Stockholm Environmental Institute. SERC (2006), Electricity regulation report 2006 [Dianli jianguan baogao (2006)], State Electricity Regulatory Commission, Beijing, China. SERC (2007), Electricity regulation report 2007 [Dianli jianguan baogao (2007)], State Electricity Regulatory Commission, Beijing, China. SERC (2008), Electricity regulation report 2008 [Dianli jianguan baogao (2008)], State Electricity Regulatory Commission, Beijing, China. SERC (2009), Electricity regulation report 2009 [Dianli jianguan baogao (2009)], State Electricity Regulatory Commission, Beijing, China. 104 Shearman, P, Bryan, J, Ash, J, Hunnam, P, Mackey, B, Lokes, B 2008, The state of the dorests of Papua New Guinea: mapping the extent and condition of forest cover and measuring the drivers of forest change in the period 1972-2002, University of Papua New Guinea. Sheehan, P (2008), The new global growth path: implications for climate change analysis and policy, Climatic Change, vol. 91, nos. 3-4, pp. 211-231. Shellenberger, M, Nordhaus, T, Naviin, J, Norris, T, and Noppen, AV 2008, Fast, clean, and cheap: cutting global warming's Gordian Knot, Harvard Law and Policy Review, vol.2, pp. 93-118. Sijm, J, Neuhoff, K and Chen, Y (2006) CO2 cost pass through and windfall profits in the power sector, Climate Policy, 6(1), pp 49-72. Slaski, X and Thurber, M 2009, Research note: cookstoves and obstacles to technology adoption by the poor, Program on Energy and Sustainable Development, Working Paper no. 89, Stanford University. Somanthan, E 2008, What do we expect from an international climate agreement? a perspective from a low-income country, The Harvard Project on International Climate Agreements Discussion paper 08-27, Harvard Kennedy School, Harvard, Cambridge. Staehr, K, 2006, Risk and uncertainty in cost-benefit analysis: toolbox paper, Environmental Assessment Institute (Danish Ministry of the Environment), Copenhagen. Stage, J 2010, Economic valuation of climate change adaptation in developing countries, Annals of the New York Academy of Sciences, pp. 150-162. Stern, D and Lambie, R 2010, "Where is it cheapest to cut carbon emissions?," Environmental Economics Research Hub Research Reports 1063, Environmental Economics Research Hub, Crawford School, Australian National University. Stern, N 1987 The theory of optimal commodity and income taxation: an introduction, Ch. 2 in Newbury and Stern (eds.), op. cit. Stern, N 2006, The Stern review: the economics of climate change, Cambridge University Press, Cambridge. Sterner, T 2003, Policy instruments for environmental and natural resource management, Resources for the Future, Washington DC. Stockholm Environment Institute (SEI) and the Chinese Economists 50 Forum Stockholm 2009, Going clean ­ the economics of china's low carbon development, Gang, F, Stern, LN, Edenhofer, O, Shanda, X, Eklund, K, Ackerman, F, Li, L, and Hallding, K (editorial board), SEI, Sweden. Sullivan, C, and Meigh, J 2005, Targeting attention on local vulnerabilities using an integrated index approach: the example of the climate vulnerability index, Water Science and Technology, vol. 51, issue 5, pp. 69-78. Takeda, S 2009, The double dividend from carbon regulations in Japan, Journal of the Japanese and International Economies, vol. 27 pp. 336-364. Thang, N.D, and Bennett, J, 2007, Estimating wetland biodiversity values: a choice modelling application in Vietnam's Mekong river delta, Economics and Environment Network Working Paper EEN0704, The Australian National University, Canberra. Tollefson, J 2009, Climate`s smoky spectre, Nature, vol. 460, pp. 29-32. United Nations Development Program (UNDP) 2010, China and a sustainable future: towards a low carbon economy and society, China Human Development Report 2009/10, China Publishing Group Corporation, Beijing. 105 United Nations Development Program (UNDP) and World Health Organization (WHO) 2009 The energy access situation in developing countries: a review focusing on the least developed countries and Sub- Saharan Africa, UNDP, New York. VOVNews 2009, Vietnam faces severe power shortage for next few years, 12 August http://english.vovnews.vn/Home/Vietnam-faces-severe-power-shortage-for-next-few- years/20098/106748.vov Wadhwa, W, Gangopadhyay, S, Bacon, R, Kumar, P, Kojima, M, Lvovsky, K, and Ramaswamy, B 2003, India: access of the poor to clean household fuels, UNDP and World Bank Energy Sector Management Assistance Programme, available at: http://vle.worldbank.org/bnpp/en/publications/energy- water/india-access-poor-clean-household-fuels Wang, H, Xie, J, and Li, H, 2008, Domestic water pricing with household surveys: a study of acceptability and willingness to pay in Chongqing, China, Policy Research Working Paper 4690, the World Bank, Washington DC. Wang, Q 2010, Effective policies for renewable energy ­ the example of China`s wind power ­ lessons for China`s photovoltaic power, Renewable and Sustainable Energy Reviews, vol. 14, pp.702-712. Wang, Q, and Chen, Y 2009, Energy saving and emission reduction revolutionizing China`s environmental protection, Renewable and Sustainable Energy Reviews, vol, 14, pp 535­539. Wang, Q, Qiu, H, and Kuang, Y 2009, Market-driven energy pricing necessary to ensure China`s power supply, Energy Policy, vol. 37, pp 2498­2504. Wang, T and de Neufville, R 2005, Real options in` projects, prepared for the 9th Real Options Annual International Conference in Paris, France. June 2005. Viewed 31 July 2010, available at http://www.realoptions.org/papers2005/Tao_Wang_RO_in_projects.pdf World Bank 2003, The East Asian miracle: economic growth and public policy, A World Bank Policy Research Report, Oxford University Press, New York. World Bank 2005, Electricity for all: options for increasing access in Indonesia, East-Asia and Pacific Region Energy and Mining Unit and Indonesia Country Program, World Bank, Washington DC. World Bank 2006b, Clean energy and development: towards an investment framework, Development Committee meeting paper, April 23. World Bank 2007, China and the World Bank: a partnership for innovation, Washington DC. World Bank 2009a, Aggregate governance indicators 1996-2008 available at http://info.worldbank.org/governance/wgi/pdf/wgidataset.xls. See also http://info.worldbank.org/governance/wgi/faq.htm#2 World Bank 2009b, Building response strategies to climate change in agricultural systems in Latin America, Washington DC, available at http://siteresources.worldbank.org/EXTLACREGTOPRURDEV/Resources/503766- 1225476272295/PDF_Agriculture_Climate_change.pdf World Bank 2010a, World development indicators and global development finance, available at http://databank.worldbank.org/ddp/home.do?Step=1&id=4 World Bank 2010b, Winds of change: East Asia's sustainable energy future, East Asia and Pacific Region, East Asia Infrastructure Unit, Washington DC. World Bank 2010c, Vietnam: first power sector reform development policy operation, Report 53 1 19- VN, World Bank, Washington DC 106 World Bank 2010d, World Development Report 2010: development and climate change, Oxford University Press, New York. World Bank 2010e, The economics of adaptation to climate change: a synthesis report, final consultation draft, August, Washington DC. World Bank 2010f, The costs to developing countries of adaptation to climate change: new methods and estimates. The global report of the economics of adaptation to climate change study, consultation draft, unpublished. World Bank 2010h Rural Development and Agriculture in China, available at http://web.worldbank.org/WBSITE/EXTERNAL/COUNTRIES/EASTASIAPACIFICEXT/CHINAEXT N/0,,contentMDK:20534329~menuPK:318981~pagePK:1497618~piPK:217854~theSitePK:318950,00.ht ml World Bank 2010i, Hands-on Energy Adaptation Toolkit, available at http://esmap.org/esmap/node/312. See also http://climate-l.org/2010/08/26/world-bank-presents-hands-on-energy-adaptation- toolkit/?referrer=climate-l.org-daily-feed World Bank, 2006, India inclusive growth and service delivery: building on India's success, Development Policy Review, Washington DC. World Bank/International Bank for Reconstruction and Development 2010g, Economic evaluation of climate change adaptation projects: approaches for the agriculture sector and beyond, Washington DC. Wu, FF, Wen, F, Duan, G 2004, Generation planning and investment under deregulated environment: Comparison of USA and China, Power Engineering Society General Meeting, 2004. IEEE, vol.2, p.1324-1328. Xie, J, Liebenthal, A, Warford, J, Dixon, JA, Wang, M, Gao, S, Wang, S, Jiang, Y, and Ma, Z 2009, Addressing China's water scarcity. A synthesis of recommendations for selected water resource management issues, The World Bank Group, Washington DC. Xinhua 2010 Agricultural Bank of China`s Micro-Credit Loan Programme in Rural Areas Expands Rapidly, July 18 (reported by BBC Monitoring, July 23) Xiong, W, Conway, D, Lin, E, Xu, Y, Ju, H, Jiang, J, Holman, I, and Li, Y, 2009, Future cereal production in China: the interaction of climate change, water availability and socio-economic scenarios, Global Environmental Change, vol. 19, pp. 34-44. Yohe, G and Leichenko, R 2010, Adopting a risk-based approach, Annals of the New York Academy of Sciences, no. 1196, New York City Panel on Climate Change 2010 Report, pp. 29-40. Zedillo, E et al. 2001 Report of the high-level panel on financing for development, United Nations, available at: http://www.un.org/reports/financing/full_report.pdf Zhang, C, and Heller, TC 2004, Reform of the Chinese electric power market: economics and institutions, working paper no.3, Program on Energy and Sustainable Development, the Center for Environmental Science and Policy, Stanford Institute for International Studies, Stanford university, Stanford. Zhang, J and Smith, KR 2007, "Household air pollution from coal and biomass fuels in China: measurements, health impacts, and interventions, Environmental Health Perspectives, vol. 115, pp. 848­855. Zhang, X, Chang, S, Huo, M, and Wang, R 2009, China`s wind industry: policy lessons for domestic government interventions and international support, Climate Policy, vol. 9, pp. 553-564. 107 Zhao, L 2008, Research on issue of electrovalent composing in China, Masters` Dissertation (in Chinese), Dongbei University of Finance and Economics. Zhongfu, T Wang,M, Jiang, H and Wang, C 2008, Theory and methodology for price chain design of China's power industry (wo guo dian li chan ye jia ge lian she ji li lun ji fang fa ), Economic and Management Press, Beijing. Zhou, N, Levine, M, and Price, L 2010, Overview of current energy efficiency policies in China, Energy Policy, vol. 38, corrected proof, available at http://www.sciencedirect.com/science/article/B6V2W-4X5YWTS- 3/2/e6e86707f0355f59689532b9639f933e) Zissis, C, and Bajoria, J 2008, China`s Environmental Crisis, Council on Foreign Relations, available at: http://www.cfr.org/publication/12608/chinas_environmental_crisis.html. 108 Appendix Tables Table A.1 Oil, coal and gas, production consumption and reserves, selected APEC countries 2009 Increase in GDP Oil Coal Gas Energy Use at MER Prod Cons Reserves Prod Cons Reserves Prod Cons Reserves 2010-30 % United States 14,256 2,627 6,820 28,396 973 898 238,308 542 589 6,356 9 Japan 5,068 .. 1,605 .. 1 197 355 .. 79 .. -3 Canada 1,336 1,173 801 33,172 63 51 6,578 145 85 1,610 1 Korea 833 .. 849 .. 3 152 133 .. 30 .. 19 Australia 997 204 344 4,225 409 91 76,200 38 23 2,823 5 New Zealand 118 .. 54 .. 5 3 571 4 4 .. 19 Mexico 875 1,088 710 11,692 11 14 1,211 52 63 437 56 China 4,909 1,383 3,148 14,832 3,050 3,020 114,500 77 80 2,253 79 Singapore 177 .. 366 .. .. .. .. .. 9 .. na Malaysia 191 270 171 5,520 .. 7 .. 56 28 2,184 na Indonesia 539 373 491 4,403 252 50 4,328 65 33 2,922 95 Thailand 264 120 356 454 19 50 1,354 28 35 330 66 Phillipines 161 .. 97 .. .. 12 .. .. 3 .. 179 Vietnam 92 126 28 4,500 45 4 150 7 1 626 91 Brunei Darussalam 11 61 22 1,078 .. .. .. 10 1 321 -1 Chile 162 .. 122 .. .. 7 .. .. 3 .. 77 Peru 127 53 69 1,121 .. 1 .. .. 3 291 88 Russia 1,229 3,662 984 74,204 298 176 157,010 475 351 40,716 52 APEC 31,346 11,139 17,035 183,597 5,129 4,733 600,698 1,499 1,419 60,870 43 World 57,925 29,181 30,688 1,333,127 6,941 6,675 826,001 2,696 2,653 172,027 44 Notes: Oil production and consumption is measured in million barrels per year. Coal production and consumption are measured in million tonnes per year. Gas production and consumption is in million tonnes of oil equivalent. Proved reserves are those quantities that geological and engineering information indicates with reasonable certainty can be recovered in the future from known deposits under existing economic and operating conditions. Data source: Outlook Economics (2010b) Table A.2 Oil, coal and gas, production consumption and reserves, selected APEC economies 2009 ­ expressed as a percentage of world production consumption and reserves GDP Oil Coal Gas at MER Prod Cons Reserves Prod Cons Reserves Prod Cons Reserves United States 24.6 9.0 22.2 2.1 14.0 13.4 28.9 20.1 22.2 3.7 Japan 8.7 5.2 0.0 3.0 0.0 3.0 Canada 2.3 4.0 2.6 2.5 0.9 0.8 0.8 5.4 3.2 0.9 Korea 1.4 2.8 0.0 2.3 0.0 1.1 Australia 1.7 0.7 1.1 0.3 5.9 1.4 9.2 1.4 0.9 1.6 New Zealand 0.2 0.2 0.1 0.0 0.1 0.1 0.1 Mexico 1.5 3.7 2.3 0.9 0.2 0.2 0.1 1.9 2.4 0.3 China 8.5 4.7 10.3 1.1 43.9 45.2 13.9 2.8 3.0 1.3 Singapore 0.3 1.2 0.3 Malaysia 0.3 0.9 0.6 0.4 0.1 2.1 1.1 1.3 Indonesia 0.9 1.3 1.6 0.3 3.6 0.7 0.5 2.4 1.2 1.7 Thailand 0.5 0.4 1.2 0.0 0.3 0.8 0.2 1.0 1.3 0.2 Phillipines 0.3 0.3 0.2 0.1 Vietnam 0.2 0.4 0.1 0.3 0.6 0.1 0.0 0.3 0.0 0.4 Brunei Darussalam 0.0 0.2 0.1 0.1 0.4 0.0 0.2 Chile 0.3 0.4 0.1 0.1 Peru 0.2 0.2 0.2 0.1 0.0 0.1 0.2 Russia 2.1 12.5 3.2 5.6 4.3 2.6 19.0 17.6 13.2 23.7 APEC 54.1 38.2 55.5 13.8 73.9 70.9 72.7 55.6 53.5 35.4 World 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Notes: see Table A.1 Data source: Outlook Economics (2010b) 109 Table A.3 Oil, coal and gas, dependency, relative intensity and years of reserves, 2009 Share Oil Coal Gas of GDP Depend- Year of Depend- Years of Depend- Years of at MER ency Intensity Reserves ency Intensity Reserves ency Intensity Reserves United States 24.6 61 90 11 -8 55 245 8 90 12 Japan 8.7 100 60 99 34 277 100 34 Canada 2.3 -46 113 28 -24 33 105 -70 139 11 Korea 1.4 100 193 98 159 53 100 80 Australia 1.7 41 65 21 -349 79 186 -65 51 74 New Zealand 0.2 100 87 -64 21 125 0 66 Mexico 1.5 -53 153 11 22 14 109 16 156 8 China 8.5 56 121 11 -1 534 38 4 36 29 Singapore 0.3 100 390 100 107 Malaysia 0.3 -58 168 20 100 33 -99 323 39 Indonesia 0.9 24 172 12 -409 80 17 -96 133 45 Thailand 0.5 66 255 4 63 166 72 21 292 12 Phillipines 0.3 100 113 100 66 100 41 Vietnam 0.2 -354 57 36 -1,177 33 3 -735 20 87 Brunei Darussalam 0.0 -181 390 18 -1,884 107 31 Chile 0.3 100 142 100 39 100 40 Peru 0.2 23 102 21 100 6 100 54 Russia 2.1 -272 151 20 -70 124 527 -35 623 86 APEC 54.1 35 103 16 -8 131 117 -6 99 41 World 100.0 0 100 46 0 100 119 0 100 64 Notes: Dependency is defined as (consumption - production)/consumption * 100. For example the 61 per cent figure for oil dependency for the US indicates that it imports 61 per cent of the oil it consumes. Relative intensity is defined as the share of the countries consumption in world consumption divided by its share of world GDP times 100. Years of reserves are proved reserves in 2009 divided by annual production at 1009 levels. Data source: Outlook Economics (2010b) 110