Except where noted, all GDP figures in this survey are based on OECD estimates expressed in constant 1995 dollars adjusted for purchasing power parity ($1995-PPP). ^[2]

Throughout the report, the data presented are highly aggregated. While these summary measures might not permit strong, specific conclusions, the broad review of international energy-use patterns presented here is useful in understanding global energy use and carbon dioxide emissions patterns.

These six regional categories closely adhere to those used in the Energy Information Administration’s International Energy Outlook 2003 (IEO), and are based on an historic view of these groupings. Thus, the category referred to as ‘developed’ or ‘the OECD’ excludes many countries that are currently included in that group (e.g. Turkey, South Korea, Poland and other former communist states). Turkey is placed with the Middle East and North Africa (MENA); South Korea is included in Developing Asia; and the former communist states are part of Eastern Europe and the Former Soviet Union (EE&FSU).

The definitions used in this survey diverge from the IEO in regards to North Africa. The IEO does not separate the fossil fuel rich countries of Egypt, Tunisia, Libya, and Algeria from the rest of Africa. This report does so, believing that those countries’ energy behavior and economic development are more consistent with the Middle East than with the rest of Africa.

A precise listing of the countries in each category is included in the Appendices.

In recent years, concern over these effects has been heightened by the possibility that the release of greenhouse gases -- the most important of which are carbon dioxide, methane, and nitrous oxide – may contribute to global warming. While other human activities promote emissions of each of these gases, adding to existing natural atmospheric levels of gases such as water vapor, the largest source of human-caused greenhouse gas emissions is the burning of fossil fuels, which releases carbon dioxide.

In considering the environmental ramifications of energy consumption, this report focuses only on emissions of carbon dioxide. It does not examine other greenhouse gases or aspects of energy consumption that impact the natural environment (e.g. oil spills).

This report defines “energy” strictly as commercial (or marketed) energy. Many developing economies remain dependent on “traditional” energy consumption (e.g. fuel-wood, waste, dung) in their residential and agricultural sectors. In some non-OECD countries, estimates suggest that traditional fuel consumption equals or surpasses commercial energy use. Unfortunately, despite its importance, traditional energy use cannot be included in this survey due to the lack of reliable data.

In part because consumption of non-carbon-emitting energy sources grew relative to fossil fuel consumption between 1980 and 2001, the world’s carbon dioxide emissions rose more slowly (1.2% per year) than energy use (1.7% per year) during the period. In the OECD, the share of non-carbon-emitting energy sources (nuclear, hydroelectricity, other renewables) rose from 12% to 18% between 1980 and 2001, largely as a result of increased consumption of nuclear power. During the same period, the shares of coal and oil declined, though consumption of both increased in absolute terms. Non-carbon-emitting energy sources also gained a progressively larger share in developing countries’ fuel mixes between 1980 and 2001.

One way of considering what the changes in the world’s energy mix mean is to compare actual world carbon dioxide emissions to a hypothetical scenario where the world retained its 1980 energy portfolio (with the implication that the per capita emissions rate also remained constant) and population rose to its 2001 level. In this hypothetical scenario, carbon dioxide emissions would have been 7.3% higher (25,900 million metric tons (MMT) as opposed to around 24,100 MMT) than actual 2001 emissions.

Economic growth tends to be directly correlated with increased energy consumption, at least to a certain point. Increased levels of household income, for instance, often lead to the purchase of larger housing units, raising residential energy consumption. Residential energy consumption levels also expand along with higher rates of appliance penetration and usage. In addition, increased wealth is often associated with “luxury” purchases, like larger, heavier vehicles that use energy more intensively.

Beyond a certain point, however, further economic development actually can lead to structural shifts in the economy that reduce the prominence of energy intensive industries. For example, more energy-intensive (and polluting) heavy industrial processes tend to move from developed to developing countries. Second, higher income levels can lead to the development and diffusion of more technologically sophisticated, but less energy intensive, machines. This trend has been especially prominent in the home appliances sector, where new models use only a small fraction of the electricity consumed by earlier models. In part, this reflects the imposition of efficiency standards by governments. Third, as economies develop beyond a certain point they may move into a “post-industrial” phase in which services become relatively more important as opposed to manufacturing, while capital and labor become more important compared to raw materials and energy.^[4]

The developing world has not experienced an analogous structural shift, despite the fact that the non-OECD’s real GDP also grew faster than energy use between 1980 and 2001. This apparent reduction in the non-OECD’s energy intensity does not take into account widening divisions within the developing world. Developing Asia, for example, has increasingly diverged from the rest of the non-OECD in terms of energy use patterns. When that region is excluded, it becomes clear that much of the developing world’s GDP growth is still tightly correlated with increased energy consumption (and carbon dioxide emissions).

Developing Asia’s divergence from other parts of the non-OECD stems in large part from the relatively rapid, export-led economic growth in that region. By the late 1990s and early 2000s, some East Asian countries achieved levels of output and consumption analogous to lower-tier members of the OECD.[**] It is important to note, however, that much of Developing Asia’s new wealth has been unevenly distributed among and within its countries.

The change in energy usage patterns in Developing Asia has a different underlying set of causes from that of the OECD. While some economies in the region also began to shift towards post-industrialism, much of the reduction in energy intensity was achieved through technological modernization. This is especially true for China, the region’s dominant energy consumer. There, the shift towards a more market-based economic system and away from central planning led to dramatic increases in energy efficiency as old, wasteful plants were shut down. The policies pursued during the previous era (i.e., the “Great Leap Forward” period under Mao Tse-tung), when rapid growth and self-sufficiency were prioritized, had led to extreme inefficiency.^[5]

Increases in population had a major influence, both directly on the demand for energy as well as indirectly through economic growth. Ceteris paribus (i.e. all else being equal), population growth is correlated with increased energy consumption and economic activity. Over the past few decades, the developed countries have experienced slow population growth and rising per capita incomes. The slow population growth reflects declining fertility rates. As noted above, increasing incomes beyond a certain point may actually lead to falling energy intensities or even lower absolute consumption levels.

Between 1980 and 2001, many developing countries experienced rapid population growth. While fertility rates are now declining in much of the developing world, growing populations in the region continue to create upwards pressure on demand for goods and services. This increased demand tends to exert upward pressure on energy consumption. In addition, many developing countries maintain energy price subsidies. These allow for higher energy consumption rates than market prices would imply.

A country’s factoral endowments exert yet another influence on national energy consumption and carbon dioxide emissions patterns. Some countries, like Canada and the United States for instance, based their development historically on relatively cheap and abundant natural resources, including energy. Comparative advantage theory, of course, states that those countries with relatively higher factor endowments in one particular area – labor, capital, or “land” -- are likely to specialize in industries utilizing that factor. Countries poor in natural resources, like Japan, tend to substitute for these with capital and highly skilled labor. Such differences in industrial mix (or industrial processes) can result in considerable variation in energy demand even among countries with similar levels of per capita income.

Government policy also has a major impact on energy use through taxes, subsidies, efficiency standards, industrial policies, etc. In formerly communist countries, for instance, decisions often were made to promote heavy industry and to subsidize the cost of energy. Today, more than ten years after the transition from centrally-planned to market-based economies, many of these former communist states continue to have comparatively high energy and carbon dioxide intensities. However, change is occurring rapidly as the countries integrate themselves increasingly in the globalized world economy.

Conclusion
Although no simple, unified theory of energy consumption and carbon dioxide emissions patterns is presented here, an examination of historical trends across countries and regions can help to increase understanding in this area. It also might potentially be of use in guiding future policy choices.

Data Sources
Most of the energy data used in this report come from the Energy Information Administration. Other significant sources include: 1) the OECD’s Energy Balances of OECD Countries 2002-2003; 2) the OECD’s Energy Statistics and Balances of Non-OECD Countries; and 3) the International Energy Agency’s Carbon Dioxide Emissions from Fuel Combustion (2003 Edition).

Economic and price information is taken mainly from the OECD’s National Accounts, Volume 2 (2003 edition). Other data sources for this report include: 1) Ward’s World Motor Vehicle Data (2003 Edition); 2) Ward’s World Automotive Yearbook (multiple years); and 3) specific journal and government publications.

Whenever non-EIA sources are used, these are noted in endnotes included within the Appendices. Data are available upon request.