development moves in industrial or postindustrial directions is also uncertain. Much room exists for research and for pilot experiments with policy options as ways to reduce the uncertainty.
These and other human systems will determine the extent to which the Western experience with energy efficiency will proceed further or be repeated in other countries. The future will depend on the ways these systems interact in each country and on the ways national and local policies intervene in them.
Intensification of the greenhouse effect is likely to alter rainfall patterns on a regional scale. As a rule, regions that receive increased rainfall are likely to benefit; decreased rainfall is the more serious concern. The history of the human consequences of severe drought can be instructive about the variety of human consequences of, and responses to, unmitigated climatic change.
The human role in causing drought in the Sahel region of sub-Saharan Africa is a matter of controversy. Throughout the modern history of drought-famine association in the region, there has been a tendency to interpret extreme events as indicators of trends and to attribute the presumed trends to human mismanagement of the local environment. In fact, Sahelian droughts have been recurrent events. The droughts of the 1970s and 1980s were preceded by several others in this century, one of which, in 1910-1915, resulted in intense famine with high mortality. The controversy over the human role in causing Sahelian drought revived with the drought of 1968-1974. The prevailing view was that desertification was an anthropogenic process reflecting deforestation, overgrazing, overfarming, burning, and mismanaged irrigation resulting in salinization of soil and water.
Lack of good data is a major obstacle to understanding the causes of Sahelian drought. Although some evidence supports the orthodox view, some recent research using remote sensing, field measurements, and intensive investigations of small areas has called that view into question. Observable ecological changes are less significant than had been supposed and correlate better with rainfall records than with land management (Mortimore, 1989).
The consequences of Sahelian droughts in this century have depended on the ability of indigenous systems of livelihood to