The different SRES scenario assumptions and results of the WorldScan model are based on the corresponding marker scenarios and their storylines of IPCC (2000). In fact, the scenarios for population, GDP, and energy use per energy carrier for the four IPCC regions have been directly taken from IPCC (2000). The base year data are taken from the GTAP4E data base (McDougall et al, 1998) (see model description). The WorldScan model is used to simulate the 12-region world economy based on the four SRES regions and to generate consistent trajectories for other economic parameters such as economic structure.
The main factors in WorldScan used to differentiate between the A and B scenarios and between the A1-B1 and A2-B2 scenarios are the investments, technology and consumer's preferences. In addition, WorldScan requires scenario assumptions on elasticities of production, trade and consumption, depreciation and capital flows. Finally, the WorldScan regions (12) need to be disaggregated to the IMAGE 2.2 regions (17).
The assumptions that were made for the development of the economic scenarios, amongst others for different parameters and exogenous variables, are listed below:
| 1. Changes made to the SRES input data for the four IPCC regions | |
| 2. Assumptions made in the WorldScan model | |
| Parameters | Exogenous variables |
| 3. Conversion from 12 WorldScan regions to 17 IMAGE regions | |
|
|
|
The data published in SRES for the four macro regions for population, GDP, energy per energy carrier (coal, oil, gas, modern biomass, and other non-fossil fuels) taken from IPCC (2000) are input to the WorldScan implementation of the SRES scenarios. The WorldScan model has been set up in such a way that it exactly reproduces these trajectories. However, some small changes have been made to the GDP trajectories from IPCC (2000):
The results for the four IPCC regions can be found under the IPCC-region view in the USS.
|
|
|
The Armington trade specification includes elasticities to describe the preference for consumption of domestically over internationally produced goods. WorldScan assumes a low short-term and higher long-term Armington elasticity, emphasizing the Law-of-one price in the long run. The Table below shows that the long-term elasticities in the globalization A1 and B1 scenarios are twice those of A2 and B2 (less globalization).
The production substitution elasticities and the consumption-share-income elasticity are are in line with CPB (1999). They are equal for all regions, sectors and scenarios and constant over time. This also holds for the depreciation rate, which equals 5% yr-1.
|
Armington trade elasticities (short to long-term) |
|||||
| Consumer capital, and Intermediate Goods | Oil, and other material goods | Coal, gas | Agriculture | Services, electricity, trade/transport | |
| A1, B1 | 3->6 | 4->16 | 2->2 | 4->10 | 2->5 |
| A2, B2 | 3->3 | 4->8 | 1->1 | 4->5 | 2->2.5 |
|
Production substitution elasticities |
Convergence of consumption |
||||
 vyValue added and all other inputs |
ex Energy and intermed. inputs |
xx Between intermed. inputs |
ee Between energy inputs |
c over indexed income levels compared to 1995 |
|
| A1, A2, B1, B2 | 0.4 | 0.8 | 0.8 | 2 | 1 |
|
|
|
The portfolio model relates capital flows between regions to total wealth multiplied by a logistic function with preferences for hosting countries and an indicator reflecting the degree of capital market integration. The capital markets will heavily integrate and that differences in nominal rates of return will strongly influence the investment flows in the A1 and B1 scenarios. In the less-globalizing A2 and B2 scenarios capital markets hardly integrate, and thus current preferences dominate the investment flows.
The indicator for the degree of capital-market integration is set at 10 (high) for A1 and B1, and 2 (low) for A2 and B2. This implies that in the A1 and B1 scenarios GNP will remain closer to GDP.
|
|
|
Risk premiums are a major determinant of the real interest rate, and indicate the degree of risk aversion of investors. The table below presents the risk premium factors as a percentage of the total interest rate. It can be seen that especially Japan and the European Union within the OECD, and Dynamic Asean countries have high values and tend to be aversive of risks.
The table below shows convergence of the risk-premium factor in A1 and B1 to 60% (equal to the global premium-factor in 1995). In A2 and B2 the risk-premium factor remains constant over time at the 1995 value. In A1 and B1 the Pacific OECD region can be seen to move up towards the global average. Thus, risk aversion will become more important in Pacific OECD, and thus slows down their economic growth compared to the other OECD regions.
|
Risk premium factor (as % of real interest rate in 1995) |
||||||
|
|
||||||
| USA | Japan | Pacific OECD | Western Europe | Eastern Europe | Former Soviet Union | |
| A1, B1 | 60 | 60 | 60 | 60 | 60 | 60 |
| A2, B2 | 30 | 84 | 45 | 68 | 46 | 90 |
| Middle East | Africa | Latin America | China | Dynamic Asian Economies | India and rest of the world | |
| A1, B1 | 60 | 60 | 60 | 60 | 60 | 60 |
| A2, B2 | 14 | 47 | -2 | 35 | 80 | 14 |
|
|
|
Expectations on the future combined with overall technological development determine the investments in new capacities to match the expected demand. This needs to be consistently matched with consumer's savings behavior. Saving rates depend heavily on unknown time preference rates, i.e. the extent to which consumers are willing to postpone consumption in favour of savings, either for future consumption or future generations. Savings decrease and comsumption increases with higher preference rates, because future consumption is preferred less over current consumption patterns.
Time preferences are calibrated to match base-year data and parameter assumptions, and they are region-specific. The relative position of the preference variable is presented for the base year in the table below. For example, the USA tends to consume more at the expense of savings and economic growth in the long run. In the A1 and B1 scenarios the time preferences converge by 2100. In the A2 and B2 scenarios the time preference for investments does not change over time.
|
Time preference for consumption* |
|||||
| USA | Japan | Pacific OECD | Western Europe | Eastern Europe | Former Soviet Union |
| ++ | 0 | + | + | + | 0 |
| Middle East | Africa | Latin America | China | Dynamic Asian Economies | India and rest of the world |
| - | + | + | -- | - | + |
|
* The relative time preferences are indicated by ++ (strong preference for short-term consumption), + (medium), 0, - (medium preference for savings), -- (strong preference for savings).
|
|||||
|
|
|
Consumption of goods is allocated to specific categories according to a Cobb-Douglas consumption function with time-dependent share-preference variables for different varieties of goods (see the table below). These share-preferences deviate from the current situation, as they refer to the long run and reflect the scneario narratives. As low-income regions start to develop, the consumption preferences will converge to the preferences prevailing in the OECD. The values for A1 and A2 are equal to the 1995 values in the USA. The long term preferences for energy-intensive goods (intermediate products, consumer durable goods, and capital) are assumed to be about 60% lower in the B scenarios than in the A scenarios.
|
Long-term consumption shares (%) |
|||||||
|
|
|||||||
| Intermediate goods | Consumer goods | Capital goods | Services goods | Electricity | Trade and Transport | Agriculture | |
| A1 , A2 | 4% | 11% | 8% | 51% | 1% | 25% | 1% |
| B1, B2 | 1% | 3% | 2% | 68% | 1% | 25% | 0% |
|
|
|
Technology is one of the main drivers of economic growth. Each sector is given technology growth such that it catches up to some pre-specified desired convergence level in agreement with the scenario narrative. In A1 and B1 there is a higher degree of convergence than in A2 and B2. In addition, a sector parameter is used to add sectoral shifts in the economies, which is important when differentiating between the A and B worlds.
The table below shows the relative technilogical growth for each WorldScan region within the SRES macro-regions. The technology leaders within specific macro-regions are indicated by "0". Thus, it can be seen that within the OECD, the USA is will experience less technological progress relative to Japan in the A1 scenario.
|
Relative technology indicator* within the SRES macro-regions |
|||||
| SRES region | WorldScan region | A1 | A2 | B1 | B2 |
| OECD | USA | - | + | 0 | ++ |
| Japan | 0 | 0 | 0 | 0 | |
| Pacific OECD | ++ | ++ | ++ | ++ | |
| European Union | ++ | ++ | ++ | ++ | |
| EFSU | Eastern Europe | 0 | 0 | 0 | 0 |
| Former Soviet Union | ++ | ++ | ++ | ++ | |
| ASIA | China | -- | -- | - | -- |
| Dynamic Asian Economies | 0 | 0 | 0 | 0 | |
| India and rest of the world | ++ | + | ++ | + | |
| ROW | Middle East | + | 0 | 0 | - |
| Africa | ++ | ++ | ++ | ++ | |
| Latin America | 0 | 0 | 0 | 0 | |
|
* Technology indicator denotes for each SRES macro-region the technological progress of a region relative to the technology leader; ++ = much faster progress, + faster, - slower, -- much slower.
|
|||||
|
|
|
WorldScan generates results for 12 WorldScan regions (see CPB, 1999). Disaggregation to the 17 IMAGE regions was necessary for Canada and Oceania (= WorldScan region Pacific OECD), Central and South America (WorldScan region Latin America) and the four African regions and Middle East (WorldScan regions Sub-Saharan Africa and North Africa and Middle East). The basic principle applied is assumed convergence in the GDP per capita levels in 2120 and 2140, for the A1 and B1, and A2 and B2 scenarios, respectively.
|
|
|