Emissions Scenarios

Other reports in this collection A1 Scenarios

By design (see Section 4.3) the "High Growth" scenario family A1 explores a world in which future economic development follows the patterns of the most successful historical examples of economic development catch-up. Free trade, continued innovation, and a stable political and social climate enable developing regions to access knowledge, technology, and capital. Combined with a rapid demographic transition, this is assumed to lead to acceleration in time and space of economic growth compared to the historical OECD experience since the 19th century. The global economy is projected to expand at an average annual rate of 2.9% to 2100 (see Table 4-5), roughly in line with historical experience over the past 100 years (of 2.7% per year, see Chapter 3). Such growth rates are considered high by the current scenario literature (see Chapter 2). Compared to historical experience, however, the broad-based nature of economic development catch-up (i.e., no region "is left behind") is without precedent. The 2.9% per year economic growth rate translates into a 25-fold expansion of global GDP that would reach US$529 trillion by 2100.

As a byproduct of rapid economic development and fast demographic transition, income inequities between industrial and developing countries are virtually eradicated. Per capita income ratios are 1:1.6 in 2100, compared to a ratio of 1:16 in 1990 in terms of the GDP/capita difference between current developed (IND) and developing (DEV) regions across the four SRES regions. However, even if relative income differences are reduced drastically, absolute differences remain large, not least because of the high incomes characteristic of the A1 scenario family (per capita income differences are also larger when considered at a more disaggregated level). When measured across the four SRES regions in 1990, income per capita differences are nearly 1:40 (between ASIA and OECD90). Per capita income differences are yet higher for differences across countries or between different social strata. The poorest 20% of Bangladesh's population, for instance, earn per capita incomes that are a factor of 700 lower than that of the 20% richest Swiss population (UNDP, 1993). A distinguishing feature of the A1 scenarios is to explore pathways of reductions in present disparities. In A1, per capita income in industrial countries (IND) increases to about US$107,300 and in now developing countries (DEV) to US$66,500. Non-OECD GDP growth rates rise to a peak of about 8% between 2010 to 2030 in scenario A1, and decline once the industrial and infrastructural bases of their economies are established. By and large, the A1 scenario implies a replication across all developing regions of the post-World War II experience of Japan and Korea or the recent economic development of China. Harmonized and Other A1 Scenarios

The high economic growth characteristics of the A1B marker scenario reproduce well in the scenarios calculated with different models. The A1B-MESSAGE scenario tracks the A1B marker scenario closely at the global and regional levels ("fully harmonized" input assumptions). At the global GDP level, most growth trajectories agree within a range up to 15%, except the A1v2-MiniCAM scenario (see Box 4-5)22. Differences at the regional level are larger. All models with comparable regional aggregation levels agree well for ASIA (except A1v2- MiniCAM). Differences in economic growth prospects also agree well for OECD90 (except A1v2-MiniCAM). For REF the A1B scenarios group into two clusters - one group reproduces the GDP growth scenario of the A1B marker, while the other group suggests a GDP level by 2100 about one-third lower than that of the A1B marker. Mostly this reflects different assumptions used in the models on future labor productivity growth that have not been harmonized with the values adopted for the A1B marker scenario. For ALM, again one group of scenarios tracks closely the A1B marker, whereas other groups indicate either higher (A1B-ASF, A1B-MiniCAM) or lower (A1v2-MiniCAM, A1T-MARIA) GDP growth. The reasons are similar to those discussed above. For the SRES region REF, the different regional aggregations across the models required complicated "inverse" calculations on regional growth rates for harmonization with the respective marker scenario at the level of the aggregated SRES region. For ALM, such calculations were neither possible for all models nor considered desirable by various modeling teams, which preferred to emphasize the inherent uncertainties in regional economic growth perspectives even for an otherwise shared vision of rapid global economic growth and development catch-up. A2 Scenarios

As compared to the other SRES scenario families, the A2 world is characterized by relatively slower productivity growth rates and resultant lower per capita incomes (see Table 4-6). Yet, the global average (1990-2100) growth rate in per capita income of 1.3% is still somewhat higher than that observed from 1970 to 1995 (1.2%; World Bank, 1998). The comparatively conservative assumptions on per-capita-income growth reflect both the more fragmented economic outlook of the A2 storyline (see Section 4.3) and the slow pace of the demographic transition that underlies A2's high population growth trajectory. The fastest growth in per capita incomes (on average over 2.3% per year) occurs in the ASIA region, while the slowest growth is observed in the OECD90 region (on average 1.0% per year). In a reversal of current short-term trends, the REF and ALM regions experience a stable increase of their per capita income levels over the 21st century at a rate that is almost twice as high as in the OECD90 region (see Table 4-7). The A2 world is also characterized by a slow convergence of incomes among regions. Nonetheless, present income disparities become narrower, from a factor of 40 difference in 1990 per capita income levels between the richest and the poorest of the four SRES regions, to a factor of seven or eight by 2100. The increase of global population from 6 to 15 billion by 2100 translates into an increase of global GDP by a factor of 12 over a century. The average (1990 to 2100) annual growth rate of total GDP is 2.2%, which is lower than the 2.9% average annual growth rate observed between 1970 and 1995 (World Bank, 1998) and the 4% rate observed from 1950 to 1990 (see Table 4-5). Harmonized and Other A2 Scenarios

Four non-marker A2 scenarios (A2-AIM, A2-IMAGE with the exception of 2050, A2-MiniCAM, and A2-MESSAGE) have global GDPs within 5% of the A2 marker. The A2-A1- MiniCAM scenario has much lower global GDP than the A2 marker, reflecting a different viewpoint on future labor productivity growth (see Box 4-5) and a different interpretation of the A2 scenario storyline altogether (see Box 4-6). While this particular scenario illustrates important uncertainties with respect to economic growth and development catch-up for developing countries, it remains controversial within the writing team, especially as to whether it reflects the overall development tendencies captured in the A2 scenario storyline. Nonetheless, in view of the spirit of the SRES open process this contrasting scenario is presented in Box 4-6. Regional GDP growth differs more than global values across the different model interpretations of the A2 scenario storyline. A2-MESSAGE and A2-AIM scenarios are harmonized, based on the marker scenario, at the regional level also, while the A2- IMAGE and A2-MiniCAM scenarios have significant regional deviations from the A2 (ASF) marker. In particular, these scenarios assume stronger GDP/capita growth in the ASIA region and slower growth in the ALM and REF regions. B1 Scenarios

The B1 scenario storyline assumes high levels of social consciousness and successful governance that result in strong reductions in income inequalities and social inequity. Growth in GDP, while being substantial, is qualitatively different compared to that of other scenarios, as social activities and environmental conservation are emphasized. Concepts of "green" GDP, including socially desirable activities such as childcare, apply in particular in the B1 scenario and qualify its similarity to other scenarios in terms of monetary value of GDP. In contrast to the world of scenario A1, the reduction of income inequalities is not a byproduct, but rather the result of constant domestic and international efforts. Global GDP reaches US$328 trillion, which corresponds to an average annual growth rate of 2.5%, slightly less than the long-term historical average. Per capita income differences between the IND and DEV regions are reduced from 1:16 in 1990 to 1:1.8 by 2100; income disparities within particular regions are assumed to be even further reduced, consistent with the thread of the B1 storyline described in Section 4.3. For the IMAGE model simulation, convergence assumptions were applied in the following domains:

  • Technology convergence was toward the level of the productivity frontier region (either Japan or the US).
  • Economic structure convergence was toward long-term sectoral shares of OECD economies (e.g., the US), which over the long-term implies a decline of the share of manufacturing sectors, and hence convergence to a service-oriented economy.
  • Education convergence was to the OECD ratio of highly skilled workers within the total workforce. Harmonized and Other B1 Scenarios

The global GDP trajectories are all within the proposed harmonization intervals, except for the B1High-MiniCAM (see Box 4-5) and the B1-MARIA scenarios in the middle period of the 21st century. On the regional level, differences in GDP trajectories from different models are larger because of the differences in regional aggregations outlined above in the discussion of the A1 scenario family. B2 Scenarios

Global GDP in B2 is assumed to increase by more than a factor of 10 during the 21st century, or at an average annual growth rate of 2.2%. This growth rate of GDP is similar to the median GDP growth in the scenario database reviewed in Chapter 2. Stabilization of global population at less than double current levels, as projected in the UN median scenario adopted for B2, combined with a sustained pace of development implies that a B2 world generally achieves high levels of affluence. Average per capita income reaches about US$18,000 by 2100 in the developing countries, which exceeds the current OECD averages. In comparison, average per capita GDP reaches US$54,400 by 2100 in the developed regions, which corresponds to an income ratio of about 3:1 between industrial (IND) and now developing (DEV) regions, a considerable improvement in interregional equity by 2100 (Riahi and Roehrl, 2000). Nonetheless, given the nature of the B2 scenario storyline (Section 4.3), per capita income differences among the world regions are higher than those in the A1 and B1 scenarios, but much smaller than those in A2. Harmonized and Other B2 Scenarios

The economic growth paths described by the B2-MESSAGE marker scenario are closely tracked by the B2 quantifications derived from alternative models at the global level, with the exception of the B2-IMAGE scenario, which is slightly below23. Differences at the regional level are larger. For B2- ASF and B2-MiniCAM, GDP is higher in the ALM and lower in the ASIA regions, respectively, compared to the B2 marker scenario and its "fully harmonized" companion B2-AIM. The latter - like all AIM scenarios - was developed by an interdisciplinary group of researchers from different countries in the ASIA region and therefore helps to guide readers as to which scenario better reflects regional perspectives.

Box 4-5: Labor Productivity Growth Rates in the SRES Scenarios

The high income growths assumed in the scenarios, especially in the A1 and B1 scenario families, imply large labor productivity increases. According to one member of the writing team such increases might not be plausible, but the other members find the assumptions plausible, especially in view of historical precedents. In line with IPCC practice, the dissenting view is elaborated here. Its corresponding implications on scenario quantifications are discussed with one example in Box 4-6.

Long-term economic growth rates can be expressed as the sum of the labor force growth rate and the growth rate of labor productivity. This framework can be used to understand more clearly the kind of technological and demographic assumptions present in the SRES scenarios. A simplified measure of labor productivity, the average economic output per member of the labor force, is used to examine the SRES scenarios.

Productivity assumptions are not comparable across the different models used to quantify the SRES emissions scenarios, so these issues were examined by running quantifications of the SRES scenarios with the MiniCAM model. Population projections were taken as exogenous inputs and regional GDP growth paths were taken to be similar to those in the SRES marker scenarios. From these, the implied labor force productivity growth rates are determined using assumptions about labor force participation rates. In these calculations, total labor force participation rates were assumed to be asymptotic to a participation rate of 80% for a working age population of all persons aged 15 to 65.

The result of this exercise is that increases in labor force productivity range between 0.79% and 5.85% per year (calculated for the periods 1990 to 2020, 2020 to 2050, and 2050 to 2100) for the four scenario families and four SRES macro-regions considered separately. This compares with historical experience between 1970 and 1995 of growth in regional labor force productivity of 0.69 to 4.13% per year and the longer-term productivity growth rates of between 1.1 and 7.7% per year. The implied future growth in global labor productivity for the MiniCAM scenario calculations ranges between 1.12% and 3.49% per year for the four scenarios (again from 1990 to 2020, 2020 to 2050, and 2050 to 2100). The historical growth in global labor force productivity between 1970 and 1995 was 1.04% per year. Historical rates of GDP per capita growth (a macro-economic proxy for labor productivity; see Table 3-2 and Table 4-7) were 1.1% per year between 1980 and 1992 and 2.5% per year between 1950 and 1980.

At the upper end of the these ranges, the SRES scenarios exhibit a growth in global labor force productivity that is higher than recent historical global experience, particularly for the SRES regions REF and ALM. This indicates that none of the four SRES scenario families envisions a recurrence of the current economic crisis in Eastern Europe and Russia or a recurrence of the "lost decade" of negative GDP per capita growth in Africa or Latin America. In addition, the period over which some developing regions exhibit high growth rates in these scenarios is longer than any historical record of high growth rates. However, there is limited analogous historical experience, as Japan is the only country that can be said to have completed such an economic "catch-up."

The assumptions on labor force participation used in the calculations reported in this box result in a decline in the growth of the labor force through the 21st century. The total labor force actually declines in all regions for the last simulation period. This reflects the demographic pattern of an overall decline in the population and the assumed stability in labor force participation rates. However, the scenarios describe generally affluent worlds in which people live longer. Thus, the labor participation patterns are likely to change with respect to current practices. People may work much longer over their lifetime, but this trend is countered by the probable need for an increase in education levels, which act to delay entry into the labor force. Pushing the asymptotic labor participation rates upward from the value assumed here results in only a small increase in total productivity, on the order of 0.2% per year.

The high economic growth branches (A1 and B1 families) of the SRES scenarios may represent upper bounds for future increases in labor productivity. Different assumptions about future rates of labor force participation do not appear to change this conclusion substantially. These scenarios essentially assume that, within the next few decades, most developing regions will experience an extended period of successful economic development analogous to the historical experience of Japan and the "Asian Tigers." An alternative view on future labor productivity growth is provided in a number of scenarios developed with the MiniCAM model, most notably A1v2-MiniCAM and the "transitional" scenario A2-A1-MiniCAM (see Box 4-6). In these scenarios global economic output is between 20% (A2-A1-MiniCAM) and 36% (A1v2-MiniCAM) lower than in the other scenarios of their respective scenario families.


Box 4-6: Possible Transitions between Scenario Families: The A2-A1-MiniCAM Scenario

The four scenario storylines have been stylized as global socio-economic developments that evolve in different directions, but globally and continuously. In reality, different regions may follow the developments pictured in the scenarios in different time periods. For example, the world may in reality develop according to one of the storylines and after some time move toward another. As an illustration of this, one scenario (A2-A1-MiniCAM) was elaborated by a member of the writing team. This scenario is described here even though some members of the writing team considered its inclusion undesirable and possibly confusing as it was submitted too late to have the team thoroughly discuss its consistency24 and to clarify its relationship to the four storylines. However, the point that in reality transitions between scenarios are possible is a valid one and the contrasting viewpoint is presented here for consideration of the reader, following IPCC practice.

The A2-A1-MiniCAM transition scenario explores a world in which the prerequisites for development, such as education, effective institutions, and high saving rates, take some time to develop, so that rapid development does not begin to occur until between 2020 and 2050 depending on the region. In this scenario, total GDP by 2100 is below the median of the historical scenario data base and, with the population 20% higher than the current median UN forecast, average per capita income is at the lower end of the historical data base range. In such a relatively poor world the economic structure is more sensitive to environmental change than in the marker scenario, and the institutional structure is less capable. Thus, the impacts of climate change are larger and the ability to adapt less than those in the A1 world. The primary driving forces for the A2-A1-MiniCAM transition scenario follow the logic of this story line, as detailed below.

Population is lower in the A2-A1-MiniCAM scenario variant than in the A2 marker, since its total completed fertility early in the 21st century is lower. This reflects the continued rapid historical declines and leads to slower population growth rates than in the population scenario adopted for the A2 scenario family. Total completed fertility declines slowly in the forecast period with a long-term asymptotic level of 2.25 for all regions, which results in a global population that is still growing by over 100 million per year in 2100. The values for migration and death rates used to generate the population trajectory follow those of the UN median forecast. The population scenario of A2-A1-MiniCAM is quite close to the UN medium population projection (see the discussion of the B2 scenario family below) until about 2060. Thereafter, however, A2-A1-MiniCAM's population scenario continues to grow linearly to about 12 billion, whereas the UN median projection stabilizes at about 10 billion by 2080 and slightly declines thereafter.

The regionally heterogeneous ("delayed") pattern of development of the precursors to rapid economic growth (e.g. education) means that some developing regions experience stagnant or very slowly growth in per capita incomes well into the 21st century. As regions begin rapid development they approach and follow the average OECD labor productivity pattern, so GDP growth rates accelerate post-2050 and average nearly 2.5% per year over the second half of the 21st century. GDP thus rises more slowly early in the 21st century in this scenario, reaching just over US$50 trillion by 2050. After 2050, GDP rises more rapidly to reach just under US$200 trillion in 2100.

Per capita final energy demands are limited by per capita income, and rise only as economic growth occurs. After growing slowly until 2050, per capita energy demands grow by more than 1.5% annually to reach 120 GJ per capita by 2100. With increases in the efficiency with which services are provided, this results in a global average level of energy services similar to that currently seen in Western Europe. Global per capita income and energy use by 2100 approach that of Western Europe in 1990. Therefore global energy intensities in 2100 approach those of Western Europe in 1990, a value lower than in other scenarios but in line with current observations.

Natural gas and oil dominate the primary energy system, and contribute slightly more than half of the primary energy. Non-fossil sources contribute about 30% of total primary energy, with coal providing the remainder. Total fossil energy carbon emissions reach 22 GtC by 2100. Sulfur controls are delayed in this implementation until economic growth takes off after 2050. With high levels of fossil fuel use, and relatively low rates of control, sulfur emissions are about ten million tons higher in 2100 than they are today.

The relatively slow growth in output and productivity in the economy in general is mirrored in the agricultural sector, with lower growth in agricultural productivity until post 2050. The large population complicates this problem, leading to large-scale expansion of agricultural lands and a resultant decrease in forested and unmanaged lands, especially in developing regions. Land-use changes do not offset any of these emissions, since the relatively high population, the rapid growth in income, and the growth in modern biomass result in essentially zero carbon emissions from land use and agriculture, rather than the substantial uptake seen in many other scenarios.


Other reports in this collection

IPCC Homepage