always be new frontiers to open up when current areas are exhausted. And a history of nationalistic feelings—only now changing—encourages forest clearing: people have felt that Brazil must develop all of its resources to assume its rightful place as a world power. The incentives for policy makers to promote conservation have been weak: many of the costs of destroying tropical moist forests (for example, the resultant loss of biodiversity) are unknown or borne by those residing in other parts of the world, and consequently were ignored until recently by Brazilian policy makers. Under prevailing institutional arrangements, the incentives for individuals and corporations favored "mining" the forests, not conserving them. Although the proximal causes of the destruction of ancient forests are not difficult to pinpoint, the story becomes much more complex as we shift our attention to underlying causes that must be understood in order to control or redirect the behavior in question. The story is even more complex when one moves beyond Brazilian Amazonia and considers that the underlying causes and their relationships can be quite different from one country to another.
What do these stories tell us? They support the general notion that global environmental change is driven by trends in global production and consumption. But to recognize this is not necessarily useful for understanding specific types and rates of change or the complex set of factors influencing production and consumption. Single-factor explanations of the anthropogenic sources of global environmental change are apt to be misleading at best. It follows as well that simple, one-dimensional policies, such as a carbon tax or a uniform law of the atmosphere, cannot by themselves control global environmental change.
The stories also suggest the need to build stronger links between the natural sciences and the social sciences in efforts to understand global environmental changes and to devise public policies to respond to them in an effective manner. To project such changes, natural scientists must also project human behavior: what actions might affect the environment, how people might respond to environmental changes, and how people might use information in making decisions about their relationship to the environment. The quality of the environmental analysis is limited by the quality of the behavioral analysis that it includes. Erroneous assumptions about the future course of human behavior can lead analysts