transport across their countries for reasons of security, avoidance of congestion, and local environmental protection, Europe will over the next several decades create two very different transport systems (combined rail-road versus road-dominant) and will certainly have gone beyond a bifurcation point (Hourcade, 1992). A similar choice is to be faced in the near future in many large developing countries with far more dramatic consequences, as decisions are made about how to expand road, rail, and air networks to respond to large increases in the demand for personal mobility and freight transport.

8.3.1.2 Current and Future Socioeconomic Development Patterns

Combining the issues and considerations discussed above suggests that four factors are likely to give rise to differences in the development paths underlying different baseline emission levels and differences in the cost curves for emission reduction or adaptation. We will describe briefly the nature of these factors and point out to what extent they are amenable or not to conscious control or choice. The critical issue is that many of the decisions that are apt to control the long-term paths will be taken for reasons that have nothing to do with energy policy or climate change issues and are indeed very often taken "in passing" rather than as an explicit component of public policy.

8.3.1.2.1 Material and energy content of development in industrialized countries

An important determinant of greenhouse gas emissions in any economy is the raw material intensity of that economy: the amount of matter and energy used per unit of economic activity. Broadly speaking, for any given level of economic activity, the lower the raw material intensity, the lower the greenhouse gas emissions. Over the past several decades, the raw material intensity of industrialized countries has dropped significantly (Bernardini and Galli, 1993; Williams, 1987). An important question is whether such reductions in industrialized economies simply reflect the shifting of energy- and matterintensive activities to other countries, in which case the net effect may be to hold constant or even increase global raw material intensities, or whether there is a net reduction in the energy- and matter-intensive activities themselves (IAEE, 1993).

Some of the major socioeconomic factors that will effect the future raw material intensity of industrialized economies are:

structural shifts in the economy towards services;

increases in the "information intensity" of industrial processes, goods, and services (Chen, 1994);

the effects of telecommunications on travel and transportation energy use
(Selvanathan and Selvanathan, 1994);

saturation in the consumption of some goods and services and the emergence of

less energy- and material-intensive goods and services.

Clearly quite different configurations of these factors are possible, resulting in potentially large differences in raw material intensities, independent of the size of the economy. It is also clear that mitigation costs per unit of economic activity (e.g., dollars per unit GDP) will depend on both the mix and level of such intensities and will necessarily vary among economies with different structures.

8.3.1.22 Links among energy, transport, and urban planning

Transportation energy use accounts for a significant proportion of greenhouse gas emissions, 16 and its growth rate is typically higher than for other categories of energy demand. More critically it is the only form of energy use not drastically decoupled from economic growth after the oil shocks. This does not demonstrate that price effects are nonexistent in this sector but that they interact with other structural determinants. In the U.S., for example, energy efficiency gains in private automobiles over the 1970s and early 1980s were almost exactly offset by increases in distances driven per vehicle, so that total energy use remained flat (Schipper and Howarth, 1990). In the transport sector, the size and type of emissions are a function of the demand for transportation services (i.e., the amount of travel), the mode chosen (auto, air, bus, rail), the efficiency of the vehicles, and the types of fuel used. The first two of these, in turn, are greatly influenced by, and influence, the size and configuration of cities and towns (Newman and Kenworthy, 1989). Important factors are:

the location of housing relative to jobs, schools, and retail outlets;

the distribution of retail trade and industrial activities within the region;

the road and rail network within and among different cities and towns; and

the investment in, and choice of, public transit systems.

Beyond these specific urban planning issues are a whole set of questions related to the development of new transportation systems and technologies (Ross, 1989). The growth in personal mobility in industrialized countries over the past several centuries has been closely correlated with technological advances in the types of vehicles used, the availability of fuels, and the transportation infrastructure created. In turn, such factors interact in complex ways with urban form and development and directly affect both the prospects for, and the costs of, greenhouse gas mitigation. For example, the level of future automobile emissions will depend in part on the degree of saturation of automobile densities in urban areas, due to the availability of effective air and high speed ground transportation alternatives (Grübler et al., 1993). In the metropolitan area of Sao Paulo, for instance, most of the increase in atmospheric pollution has been attributed to increased congestion, as indicated by a reduction in the average vehicle

speed from 28 km/h during the 1980s to 20-24 km/h in 1993 (CETESB, 1994).

Clearly, quite different configurations of these factors can be envisaged, and transportation is typically a field in which bifurcations towards contrasted paths may occur in any region of the world. These alternatives would have major impacts on mitigation options and costs. As suggested earlier, the costs of transportation sector mitigation for a population that is dependent upon automobile transport in highly dispersed suburban developments would be very different than for a population located at the core of dense urban agglomerations with a well-developed urban transit infrastructure. Since towns and cities take shape, and change, over periods measured in decades and centuries, the full effect of alternative transportation developments will be manifest only in the relatively long term, and a high uncertainty stems from the long-run impact of short-term decisions (or nondecisions).

8.3.1.2.3 Land use and human settlements

This issue concerns all regions of the world, but its quantitative impact is more impressive in developing countries. Land use and human settlement pattern changes derived from agricultural and forestry activities as well as rural-rural and rural-urban migrations are among the main sources of greenhouse gas emissions in these countries. Deforestation in most developing countries is a complex phenomenon, mainly caused by the expansion of the road network, logging and cattle raising activities, agricultural production, and population growth fostered by rural-rural migrations due in some cases to the absence of guaranteed access to land (agrarian reform) in other regions.

These examples show that mitigation cost estimates that don't take account of such institutional factors can be very misleading. Hidden costs can be very important when one considers the institutional capacity building needed, on the one hand, to prevent the harvesting of a very low-cost natural resource such as a native forest, and, on the other, to enforce environmental protection, reforestation, and forest management practices. Moreover, even macroeconomic cost assessments don't tell the whole story: for instance, the whole GDP loss resulting from completely halting the development of the Amazon region would be very low compared with overall Brazilian economic output, but this scenario is clearly inconceivable for social, political, and cultural reasons. Deep structural changes in these underlying factors will often be required in order to make mitigation strategies feasible.

In this field it is even more important than elsewhere to look at the "greenhouse gas mitigation component" of more general development strategies instead of considering greenhouse gas mitigation measures on their own. Mitigation costs are only one part of a larger set of factors in such cases. For example, a first step towards increasing the relevance of a carbon sequestration objective might be to link it with the preservation of biodiversity. This could increase its political and social relevance, though it would make the task of quantifying economic costs and benefits much more difficult.

8.3.1.24 Development patterns in developing countries

The importance of development pattern assumptions for the assessment of mitigation costs is of particular importance in the case of developing countries. As a major part of the infrastructure needed for development is still to be built, the spectrum of future options is considerably wider than in industrialized countries. The traditional approach of using "business-as-usual" assumptions as the baseline is then particularly problematical.

Nor can one assume that developing countries will automatically follow the past development paths of industrialized countries. The significant transformations recently evident in the international economy and energy markets highlight the dangers of such an assumption. In this respect, we need to consider the structure of GDP in developing countries and how that structure might evolve given such global transformations. A crucial question concerns the developing countries' share of the world production of highly energy- and pollution-intensive goods, such as steel and aluminum. For example, as the recent shift of heavy industries from the industrialized countries towards the developing countries begins to slow, economic output may increasingly come from services and other less energy-intensive activities.

Moreover, technological choices in both the production and consumption sectors can substantially change critical parameters such as the elasticity of energy demand/GDP. Experience demonstrates that countries entering the path to development have had generally lower energy profiles than countries that developed sooner. This does not mean that the energy/GDP ratio will not increase in many of these countries and that increases in consumer purchasing power in many developing countries will not drive up energy demand growth rates. It does mean, however, that these countries are not bound to reach the same levels of energy intensity that countries such as the United Kingdom and the USA have; they are not bound, for example, to adopt the same model of energy-wasteful refrigerators that still are in operation in developed countries or to fail to include up-to-date technological improvements in factories, many of which will improve energy efficiency. More generally a growing interest in the preservation of cultural diversity might also favour less energy intensive housing, transportation, leisure, and consumption patterns, and less resource consumption per capita than is currently the case in the developed countries. One among other possible examples is related to avoiding low urban population densities that are coupled with long daily trips to work and to large shopping centres by car.

Finally, the spatial distribution of the population and of economic activities is still not settled in most developing countries, a situation that offers the possibility of adopting urban/regional planning and industrial policies directed towards rural development and a stronger role for small and medium cities, thus reducing the extent of rural exodus and the degree of demographic concentration in large cities." But, given current heavy trends towards huge conurbations, it will be difficult to shift them over the next century.

We are faced here with a large degree of socioeconomic inertia that can be overcome only gradually through steady long-term policies. Nevertheless, more decentralized development patterns than those typical of industrialized countries could allow developing countries to use modern technology (biotechnology, solar energy, wind, and small-scale hydro power) to tap their large reserves of natural resources. In the same ways, it could change their needs for transportation.

These examples show that developing countries might be in position to adopt anticipative strategies that could avoid in the long term some of the problems faced today by industrial societies. For example, in industrialized countries, energy demand/GDP elasticities first increased with successive stages of industrialization (with an acceleration during the 1950s and 1960s), but have sharply decreased since then, due to a number of different factors, such as the relative growth of services in the share of GDP and technical progress and energy conservation induced by higher oil prices (Martin, 1988; Darmstadter, et al., 1977). Developing countries could follow a path leapfrogging these intermediate stages and leading directly to less energy-intensive development patterns, thus avoiding a large increase in energy/GDP intensities in the short and medium terms - the so-called "tunnel effect" (Berrah, 1983).

Such possibilities for alternative development patterns highlight the technical feasibility of low carbon futures in developing countries that are compatible with national objectives. But, in the opposite direction, past and present experience suggest that these countries might not be in position to switch towards low greenhouse gas-emitting profiles. The recent experience of economic stagnation or recession in a substantial set of developing countries demonstrates, for example, that access to superior technologies may be limited by the increase of the share of second-hand equipment or the import of obsolete technology and products. In the same way, the prevailing indebtedness of many countries could prohibit the adoption of interurban transportation infrastructures or of the urban infrastructure policies necessary to reach balanced human settlement patterns. As a whole, barriers to more sustainable development in the developing countries can hardly be underestimated. These include insufficient capital stock thus preventing the use of the same technologies as in developed countries), tariff and nontariff trade barriers, the organization of international trade and the international financial environment, the distribution of national income, and very often the lack of appropriate institutions for using the indigenous technical and economic capabilities of these countries.

This means that switching to very different development patterns from those that have customarily been expected will depend upon the removal of these barriers, the setting up of an appropriate context by national public policies, and the evolution of international economic relationships. Development theory has not yet provided the analytical tools needed to enable modellers to discriminate among such different development patterns.