This group has addressed the issues of data on three levels. The first

is the creation of new and improved analytic techniques such as

regression methodologies.

The second is the exposition of existing

methods to make them more understandable and useful to those without highly specialized statistical training. The third is a concentrated application of these techniques to the costs, benefits and risks of medical activities, particularly innovations in surgery and applications to other areas of public policy as well.

Although there have been many technical contributions of importance and broad professional applicability, one major use has been the application of these methods to medical advances. Professor Mosteller and his colleagues have evaluated the benefits and risks of innovations in surgery and anesthesia. Surgical procedures whose risks and costs almost outweigh their benefits provided the most stringent test of the methods. One result has been to extend the technical ability to measure direct improvement in patient health and life expectancy and indirect improvement in the quality of the patient's life. Even with so much attention given to the quality of medical care, we do not know how to measure the relationship of medical expenditures to improvements in the quality of life. Their work has had direct use in designing clinical trials and in isolating the factors that will achieve more efficient resource allocation. Further, it is providing the basis for

- 41

improved public understanding of medical outcomes and costs--both when faced with individual decisions and when confronted with national

allocations.

This project was not designed initially to address public policy issues or medical decisions. The output, including publication of Costs, Risks, and Benefits of Surgery, in 1979, the first book that illustrates a systematic statistical attack on the problems of assessing medical and social policy in the surgical area, exemplifies the widespread utility and generalizability of fundamental work. In addition, three other books have appeared. F. Mosteller and J.W. Tukey's book, Data Analysis and Regression, Addision-Wesley, 1977, emphasized modern ways of looking at data both from confirmatory and exploratory points of view. It gave special attention to problems of regression and to interpretation of regression in both experimental and non-experimental settings. The research is being disseminated through graduate seminars and through the use of research results published in manuals, books, and journals. The final year of the grant will focus on the completion of the manuals to guide other researchers and will extend applications to non-medical policy areas. When completed, the grant will have produced the most comprehensive set of materials developed to date.

Funding Patterns

As a fraction of total Federal funding for basic research, support of the social sciences is relatively low: $132.6 million out of an

42

estimated total of over $4.5 billion in FY 1980. As illustrated in the attached figure, the pattern of support by discipline has tended toward the erratic. In terms of buying power, the overall level of effort is well below that of the early 1970's.

Although data show that basic research support for the social sciences grew recently, increasing by 31.4% (in current dollars) between FY 1977 and FY 1978, closer scrutiny is warranted. Of the $29.7 million increase for FY 1978, economics grew by 14.8% ($4.4M), anthropology by 13.5% ($4M), and sociology by 9% ($2.7M). The greatest share of the increase, 60.6% ($18M) went to that category of social science "not elsewhere classified," which includes linguistics, geography, history, law and research in education. This large fraction can be accounted for primarily by a $9.5M increase at DHEW (NIE received a $6.4M increase in basic research for FY 1978), a $7.4M increase at LEAA and an additional $2.7M at DOL. (NSF dropped by $2.4M in this category between FY 1977 and FY 1978.)

In subsequent years, the increase for

basic research in these areas diminished: the increase overall for FY 1979 was 4.9%; it is estimated as 1.5% for FY 1980, and is likely to be essentially level in the FY 1981 budget.

Even though overall growth diminished, selective emphases initiated in 77-78 continued. DHEW programs are estimated to have grown by 77.9% from FY 1977 to FY 1980, resulting in a doubling of the National Institute of Education programs and a 56% growth at the Alcohol, Drug

about $5M (a $3M drop from the FY 1978 high) and DOL has continued its program at a $2-3M level. The major portion of funds from LEAA, DOL, and NIE are not directed toward university-based, investigatorinitiated research in key disciplinary fields of social science. Thus, they do not overlap extensively the NSF research emphases.

Although basic research at universities retained a share of the growth during the 1977-1980 period, the average annual rate of increase was well under the cost-of-living. The increases were not distributed uniformly across fields. Three fields made stable gains: anthropology ($3.5M to $6.7M), sociology ($7.8M to $10.3M), and political science ($2M to $3.2M). Economics research averaged a 4.2% annual increase and the "not elsewhere classified" category averaged 4.9%. It should be noted that in each of these major fields (except sociology) the Foundation is unique in being the only agency that funds investigatorinitiated research. Major funding in economics is contributed by DOA, but this is strongly mission-oriented, non-competitive, and limited to intramural researchers and those at land grant institutions. Smithsonian Institution provides support for anthropology, but this is exclusively intramural. For political science, NSF provides essentially

100% of the available research funds.

The

In summary, NSF funding is vital if we are to maintain a strong scientific effort in the social and economic fields of research.

APPENDIX

SUMMARY

Allocations of FY 1981 Request

Physiology, Cellular, and Molecular Biology will be increased by 11.1 percent to a total of $74.8 million. Research in these areas in the broadest sense represents the forefront of modern (cell) biology and is moving into an era in which there are likely to be extensive and rapid commerical exploitation of the findings. Instrumentation needs for these fields continue to be great. In addition, there is a special need for advancing studies of cells, cell organelles, and their functions through techniques that employ immunochemical and recombinant DNA methods. Consistent with the explosively advancing developments, the research activity of this unit remains the largest in the Directorate. The primary concentrations are on genetic, biochemical, and physiological properties and functions of living organisms.

Environmental Biology includes the core of the activity most traditionally associated with biological research. We request an increase of about $3 million for these programs in FY 1981 to bring the total to $40.7 million. The program will stress research on (1) nutrient cycling, including the role of the terrestrial biomass in regulating the carbon dioxide content of the atmosphere, and the effects of acid rain on biological productivity, (2) the biology of tropics where biota of immense scientific and practical interest are abundant but fast disappearing, (3) long-term ecological research, and facilities for improved field research capability--essential components for research in all fields of environmental biology.