IPCC Special Report on Emissions Scenarios

Other reports in this collection

3.3.3. Scenarios of Economic Development

Unlike population projections, no long-term economic-development scenarios are available in the literature (for an earlier review, see Jefferson, 1983). In fact, for economic projections "long term" means time horizons of up to a decade (e.g., World Bank, 1997b), 1998b, far too short for the time frame addressed in this report. The longest time frames for economic growth projections available in the literature extend to 2015 (e.g., Maddison, 1998) and 2020 (World Bank, 1997b). The need for long-term economic growth scenarios has arisen primarily in connection with long-term energy and (1986), Gr�bler (1994), and Alcamo et al. (1995). An expert poll on uncertainty in future GDP growth projections is reported in Manne and Richels (1994). Recent scenario assumptions are reviewed in Chapter 2 above.

The current state of modeling long-term economic growth is not well developed, not least because the dominant forces of long-run productivity growth, such as the role of institutions and technological change (see Section 3.3.4), remain exogenous to models. As a result, productivity growth assumptions enter scenario calculations as exogenous input assumptions. The structural changes in the economy discussed in the previous paragraph result in additional difficulties; notably that service sector productivity growth is difficult to evaluate and project.

Figure 3-10: Per capita GDP or GNP growth rates, a review of the literature. Average annual growth rates for 1990 to 2020, 2020 to 2050, and 2050 to 2100 for world, industrial (IND), and developing (DEV) regions. Literature mean, median and ranges compared to SRES ranges (see Chapter 4).

Figure 3-10 summarizes the results of the analysis of available literature data on per capita economic growth, disaggregated into global as well as industrial and developing countries.

Overall, uncertainty concerning productivity and hence per capita GDP growth is considerable. Uncertainties in productivity growth rates become amplified because even small differences in productivity growth rates in all scenarios, when compounded over a time frame of a century or more into the future, translate into enormous differences in absolute levels of per capita GDP. For instance, in the scenarios reviewed in Alcamo et al. (1995) and Gr�bler (1994) per capita GDP growth rates range typically between 0.8 and 2.8% per year over the period 1990-2100. On the basis of an average global per capita income of US$4000 in 1990, global per capita GDP could range anywhere between about US$10,000 to about US$83,000 by 2100. Such uncertainties are amplified even more when regional disaggregations are considered, in particular future productivity growth in developing countries. The range of views spans all the extremes between developing countries that lag perennially behind current income levels in the OECD, to scenarios in which they catch up.

These ranges are reflected in the SRES scenarios shown in Figure 3-10. Exogenously assumed productivity growth rates correspond to alternative qualitative interpretations as to how the future could unfold, ranging from SRES low (all-min) to SRES high (all-max) rates. Extreme scenarios of productivity growth or lack of growth have not been explored because the SRES terms of reference cover a qualified range from the literature; methodological (and model) pluralism is mandatory (extreme scenarios can be reflected across a wide range of modeling approaches only to a limited degree). Furthermore, it is not possible to treat uncertainties of future demographic, economic, and technological developments as independent. This is shown by the conclusions of recent scenario evaluation exercises (Alcamo et al., 1995) as well as by theoretical and empirical evidence (e.g. Abramovitz, 1993; Barro, 1997). Thus, contrary to the previous IPCC IS92 scenario series (that varied salient scenario driving forces independently of each other), the SRES scenarios attempt to incorporate advances in the understanding of the relationships between important scenario drivers. From this perspective, uncertainties about future productivity and hence economic growth are not parametric, but rather are related to the uncertainties in current understanding and modeling of the interactions between demographics, productivity growth, and socio-institutional and technological change. These are addressed in Section 3.3.4.




Other reports in this collection