search. A fellowship program in full operation that awarded 100 two-year fellowships per year would cost $10 million per year if the average annual cost were $50,000 per fellowship, including indirect costs. Five national centers could be maintained with small but strong core staffs for about $1 million per center per year. Funding for data acquisition and dissemination should remain, as in the fiscal 1991 budget, at about 20 percent of the funds allocated to human interactions research. On that basis, we recommend that funding for the data program be increased over the transition period to a level of $8-10 million.

The committee believes that this level of support would make possible the establishment of a balanced national research program on the human dimensions of global change and that the research community will be able to take on such a commitment over a three-year period if the funding is available. This level of funding would represent about 5 percent of the fiscal 1991 budget for the U.S. Global Change Research Program (USGCRP) or 4 percent of the proposed fiscal 1992 budget, in contrast to the 3 percent currently budgeted. In light of the National Research Council's conclusion that the human interactions science priority is ''the most critically underfunded in the FY 1991 budget for the USGCRP'' (National Research Council, 1990b:95), an increase of this magnitude over a short time period seems fully justified. Support for appropriate parts of the research program outlined here could come from an emerging Mitigation and Adaptation Research Strategies program as well as from the Global Change Research Program.