Second, analysts can elicit descriptions of the problem from sampled stakeholders, whose perspectives will vary widely. This technique has been repeatedly used in an attempt to achieve consensus among stakeholders on a definition of the forest policy problem in Costa Rica. Earlier efforts in this area had found that there were major definitional differences among stakeholders. These included factors such as high rural population growth, the absence of logging regulation to ensure sustained yields, the lack of a national policy to protect tropical forest resources, the absence of fiscal incentives to the forest based industry, a high US beef import quota that encouraged pasture expansion and burning of forests, and lack of a regulatory framework and lead enforcement department at the national level (Guess 1979).Problem definition was dominated then as now by the powerful beef-cattle industry, which in self-serving fashion defined tropical forest elimination as the inevitable effect of normal expansion of the agricultural frontier. This definition produced a non policy, or policy vacuum, on several fronts. It perpetuated major loss of tropical forest resources and associated jobs and income opportunities for the rural poor of Costa Rica. In eliciting problem representations, the analyst should avoid "positive hearing," in which "words are put into people's mouths that they would have said if they'd thought of them at the time" (Le Carre 1996, 277). Despite the assurance by Le Carre's character that "everybody does it anyway," the analyst must guard against the use of positive hearing in interviews (278).