How easily could the expected surplus in the world cereal markets be upset? Not easily, it appears, but several scenarios are considered. First minor variations, such as higher income and population growth rates, to the assumptions for the baseline simulation are considered. Then major variations are simulated. The minor variations represent likely alternatives to the baseline simulation while the major variations represent extreme cases. Even though the major variations are considered unlikely, they are included to deepen the understanding of the world food outlook. The effects of each alternative are simulated separately.

The effects of minor variations to the baseline assumptions can be evaluated reasonably well with the econometric model. Major variations are more difficult to evaluate because they are outside the range of historical experience and could well lead to policy as well as economic changes. The econometric model can still be useful in indicating the magnitude of the required adjustment however, but it cannot project the types of government policy responses which are likely.

The minor variations considered are for 10 per cent higher GDP growth, higher population growth as specified in the United Nations high variant alternative and a temporary doubling of energy and fertilizer prices. The major variations considered are for zero yield growth after 1990 and a high consumption alternative for developing economies. The zero yield growth alternative could be considered as being caused by severe environmental or biological constraints. It could also be viewed as the result of strict legislated environmental controls on fertilizers or agricultural chemicals.