12.2.4.1 Energy

The implications of four broad categories of energy policies on emissions are discussed: provision of affordable energy services to the poor; liberalization; energy efficiency; and energy security. Policies that support the penetration of renewable energy - which are often introduced for non-climate reasons, but are also obvious tools for climate mitigation in the energy sector - are discussed in Section 4.5.

Access to Energy: Access to energy is critical for the provision of basic services such as lighting, cooking, refrigeration, telecommunication, education, transportation or mechanical power (Najam et al., 2003). Yet, an estimated 2.4 billion people rely on wood, charcoal or dung for cooking, and 1.6 billion are without access to electricity (IEA, 2004c). Providing access to commercial fuel and efficient stoves would have highly positive impacts on human development by reducing child mortality, improving maternal health, and freeing up time used to collect fuel wood, especially for women and girls (Najam and Cleveland, 2003; Modi et al., 2006). For example, indoor air pollution, mainly from cooking and heating from solid fuels, is responsible for 36% of all lower respiratory infections and 22% of chronic obstructive pulmonary disease (WHO, 2002). See also Chapter 4 and Chapter 6. It is estimated that a shift from crop residues to LPG, kerosene, ethanol gel, or biogas could decrease indoor air pollution by approximately 95% (Smith et al., 2000). The impact on GHG emissions depends on the nature of the biomass resources and the carbon intensity of the replacement. Providing reliable access to electricity would also have highly positive impacts on human development, by providing preconditions for the development of new economic and social activities, for example, allowing for education activities at night and employment generating business initiatives (World Bank, 1994; Karekezi and Majoro, 2002; Spalding-Fecher et al., 2002; Toman and Jemelkova, 2003).

The implications of improved access to commercial fuels for cooking on GHG emissions are ambiguous. On the one hand, emissions increase, albeit by a small amount globally. Smith (2002) estimates that providing LPG as fuel for roughly two billion households would increase global GHG emissions by about 2%. On the other hand, unsustainable use of fuelwood and related deforestation decreases. For example, the ‘butanization’ programmes adopted in Senegal in 1974 to support LPG use through a combination of subsidies to LPG, support for the development for stoves suitable for local conditions and removal of tax on imported equipment, is estimated to have resulted in a 33-fold increase in LPG use, and in a 15% drop of charcoal consumption (Davidson and Sokona, 2002). Similarly, the implications of electrification programmes for GHG emissions are ambiguous. Energy demand is likely to increase as a result of easier access and induced economic benefits. However, emissions per unit of energy consumed might decrease, depending on the relative carbon content of the fuel used in the baseline (typically kerosene) and of the electricity newly provided (de Gouvello and Maigne, 2000). Public policies have a strong influence on this technology choice. In some cases, the technology is set directly by public decision-makers. But even where left to private entities, public policies, such as the choice between centralized or decentralized models of electrification, or the nature of the fiscal system, strongly constrain technology choices.

One example of such indirect impact is documented by Colombier and Hourcade (1989). They found that the “equal price of electricity for all” principle embedded in French law has generated vast implicit subsidies from urban to rural areas and discouraged, over time, the development of cost-effective decentralized electrification alternatives to grid expansion. The expanded grid the country is locked into, however, is the source of very high maintenance and upgrading costs to accommodate increased demand from rural households and companies – much higher than would have occurred had decentralized solutions been implemented at the onset. The implications for GHG emissions (not studied in the paper) are probably limited given the share of nuclear power in France. But similar dynamics could have more important GHG emissions implications in countries with fossil-fuel dominated power grids.

Liberalization: Many countries have embarked on liberalization of their energy sector over the past two decades. These programmes with the objective to reduce costs and improve efficiency of energy services include privatization of the energy producers, separation between production and transmission activities, liberalization of energy markets, and lifting restrictions on capital flows in the sector. Overall, liberalization programmes aim at improving the efficiency of the energy sector, and should, therefore, lead to reduced emissions per unit of output. Effective privatization programmes, however, differ markedly from country to country (Kessides, 2004), depending on prior institutional arrangements. In addition, privatization programmes are often sequentia. See, for example, Jannuzzi (2005) for a discussion on how the Brazilian regulator progressively adapted policies to elicit sufficient resources for energy efficiency and R&D from private utilities. These policies are often ‘incomplete’, in the sense that former public power generators remain dominant by combining features from both the public and private sector, an outcome very different from the ideal private energy markets (Victor and Heller, 2007: see also Section 12.2.3.1). It may, therefore, not be surprising that there is little literature drawing general lessons on the implications of privatization programmes on GHG emissions.

A great deal of literature, however, deals with the emission implications of some components of privatization programmes, particularly removal of energy subsidies. Energy subsidies removal may also be adopted as a stand-alone policy, independent from privatization. Conversely, subsidies may remain even within competitive markets. Government subsidies in the global energy sector are in the order of US$ 250-300 billion per year, of which around 2-3% support renewable energy (De Moor, 2001). Removing subsidies on energy has well-documented economic benefits. It frees up financial resources for other uses and discourages overuse of natural resources (UNEP, 2004). But, reducing energy subsidies might have important distributional effects, notably on the poor, if not accompanied by appropriate compensation mechanisms. The impact of policies to reduce energy subsidies on CO₂ emissions is expected to be positive in most cases, as higher prices trigger lower demand for energy and induce energy conservation. For example, econometric analyses have shown that price liberalization in Eastern Europe during the period 1992-1999 was an important driver of the decrease in energy intensity in the industrial sector (Cornillie and Fankhauser, 2002). Similarly, removal of energy subsidies has been identified as instrumental in reducing GHG emissions compared with the baseline in China and India over the past 20 years (Chandler et al., 2002). Overall, an OECD study showed global CO₂emissions could be reduced by more than 6% and real income increased by 0.1% by 2010, if support mechanisms on fossil fuels used by industry and the power generation sector were removed (OECD, 2002). Yet subsidies removal may actually result in increased emissions in cases where poor consumers are forced off-grid and back to highly carbon intensive fuels, such as non-sustainable charcoal or diesel generators. For example, removal of the subsidies for LPG in Senegal under the ‘butanization’ programmes discussed above is expected to increase charcoal and unsustainable fuelwood use (Deme, 2003). For additional discussion on energy subsidies, see Section 4.5.1 and Section 6.8.3.2 and Section 13.2.1.5.

Energy Efficiency: Policies that increase energy efficiency – both on the demand and on the supply side – are pursued to reduce demand for energy without affecting, or while increasing, output at very low costs. This is the case even though some of the direct efficiency gains might be offset by increased demand due to lower energy costs per unit of output. Efficiency also increases competitiveness, relaxes supply constraints and, therefore, enhances the range of policy options and space, and lowers expenditure on energy thereby freeing up more resources for other development goals. The impact on CO₂ emissions, in turn, tends to be positive, but depends heavily on the carbon content of the energy supply. For example, Brazil National Electricity Conservation Program (PROCEL), created in 1985, has saved an estimated 12.9 TWh and an estimated R$ 2.6 billion from 1986 to 1997. This is 25 times as much as the amount invested in the programmes, while reducing emissions by an estimated 3.6 Mt CO₂ over the same period of time (La Rovere and Americano, 1999; Szklo et al., 2005). Similarly, Palmer et al. (2004) estimate that the annual energy savings generated by all current Demand-Side Management programmes (DSM) in the USA represent about 6% of the country’s non-transportation energy consumption. This leads to reductions in CO₂ emissions equivalent to (at most) 3.5% of the country’s total. DSM programmes are also discussed in Section 6.8.3.1 and Section 5.5.1.Over the period 1973-1998, the International Energy Agency (IEA, 2004b) estimates that energy efficiency - driven both by policies and by autonomous technical improvements - have resulted in energy savings corresponding to almost 50% of 1998 energy consumption levels. Without these savings, energy use (and CO₂ emissions) in 1998 would have been almost 50% higher than observed.

Energy Security: Energy security is broadly defined as ensuring long-term security of energy supply at reasonable prices to support the domestic economy. This is a major concern for Governments worldwide, and it has taken new prominence in recent years with the political instability in the Middle East, increased oil prices, and tensions over gas in Europe (Dorian et al., 2006; Turton and Barreto, 2006). Energy security concerns, however, can translate into very different policies depending on national and historical circumstances (Helm, 2002). Their impact on carbon emissions is ambiguous, depending on the nature of the policies and, in particular, on the fuel sources being favoured. For example, in response to the first oil shock, Brazil launched in 1975 the National Alcohol Fuel Program (PRO-ALCOOL) to increase the production of sugarcane ethanol as a substitute for oil, at a time when Brazil was importing about 80% of its oil supply^[5]. The programmes resulted in reduction of oil imports and expenditure of foreign currency and job creation, as well as in large emission reductions, estimated at 1.5 Mt CO₂/yr (Szklo et al., 2005). Brazil also provides an example where emissions actually increased as a result of energy security considerations. During the 1990s, Brazil faced lack of public and private investment in the expansion of the power system (both generation and transmission) and a growing supply-demand imbalance, which culminated in electricity shortage and rationing in 2001. This situation forced the country to install and run emergency fossil-fuel plants, which led to a substantial increase in GHG emissions from the power sector in 2001 (Geller et al., 2004). Hourcade and Kostopoulo (1994) show how reactions to the first oil shock by France, Italy, Germany, and Japan led to very different emissions with relatively similar economic outcomes (see Box 12.5).

Box 12.5: Differentiated reactions to the first oil shock in France, Italy, Germany and Japan

An example of how different development paths can unfold in relatively similar countries is given by Hourcade and Kostopoulou (1994) for France, Italy, Germany, and Japan - countries with similar levels of GDP per capita in 1973 – in their response to the first oil shock. France moved aggressively to develop domestic supply of nuclear energy and a new building code. Japan made an aggressive shift of its industry towards less energy-intensive activities and simultaneously used its exchange-rate policies to alleviate the burden of oil purchases. Germany built industrial exports to compensate the trade balance deficit in the energy sector. Much of the variations of CO₂ emissions per unit of GDP from 1971 to 1990 can be attributed to these choices (Figure 12.1 left). Yet, while this indicator diminished by half in France, by a third in Japan, and ‘only’ by a quarter in Germany (IEA 2004b).

Figure 12.1: (left) CO₂ emissions from fossil-fuel combustion per unit of GDP; (right) Evolution of GDP per capita

Note: GDP in US$ in constant prices at market exchange rates.

Source: IEA, 2004a

Hourcade and Kostopoulou (1994) also observe that the macro-economic performance of these countries was relatively comparable between 1973 and 1990 (Figure 12.1 right), suggesting that widely different environmental outcome can be obtained at similar welfare costs in the long-run. In addition, they observe that the responses were for a large part driven by the country’s pre-existing technologies and institutions (thus providing an illustration of the general observations about decision-making made in Section 12.2.3.1).