1. Science, Society, and Public Policy Michael M. Crow Columbia University

2. -THE IMPORTANCE OF SCIENTIFIC AND TECHNOLOGICAL ADVANCE -THE SOCIAL SHAPING OF THE NATIONAL SCIENCE BASE -S&T POLICY: THE 1950’S MODEL - TRADING IN THE 1950’S MODEL

3. SCIENCE AS AN INSTRUMENT OF POLICY: Science is an instrument that can be used for a variety of social objectives, including: -Meeting Basic Human Needs -Making War - Improving the Quality of Life -Economic Growth and Development

4. SCIENCE, TECHNOLOGY, AND ECONOMIC GROWTH: Between 1870 and 1973, the U.S. economy had grown at an average rate of 3.4% annually. Between 1973 and 1993, the average rate of growth flattened to 2.3%.

5. GDP 3.4% long term rate 2.3% slow growth losses 19731993

6. Economic Growth: Importance Over the 20 years since 1973, the accumulated losses in goods and services due to slow growth have come to nearly $12 trillion, or $40,000 per person.

7. Economic Growth: Importance $12 trillion is more than enough to: –Have bought each of America’s landowners a new house; or, –Paid off all of our government, mortgage, and credit card debt; or, –Replaced all of our nation’s factories, including capital equipment, with new ones.

8. Economic Growth: Importance As the triangle grows over time, so does the cumulative damage. By the year 2013, assuming the post-1973 trend of growing just one-percent less than our historical average holds, the losses would be $35 trillion of lost production since 1973. This is a loss of over $100,000 per person.

9. Economic Growth: Importance Compounded over generations, a 1 or 2 percent reduction in the overall growth rate could be the difference between the standard of living merely doubling or increasing five-fold over a 100 year period.

10. Most economists agree that scientific and technical change accounts for as much as 50% of long-run economic growth. A large number of economists argue that, when we measure scientific and technical change properly, the figure is as high as 75%.

11. “ NATIONAL INNOVATION SYSTEMS” AND SCIENTIFIC/TECHNOLOGICAL ADVANCE National Innovation Systems: The Complex Network of Agents, Policies, and Institutions Supporting the Process of Scientific and Technical Advance in an Economy

12. The “Narrow” NIS Organizations and Institutions Directly Involved in Searching and Exploring Activities, e.g. Universities and Research Laboratories

13. The “Narrow” NIS Hybrid S&T Labs Public S&T Labs Scientific and Technological Societies Technology Sharing Regimes Technology Licensing Regimes Intellectual Property Regimes Mission Agencies Universities Private Firms

14. The “Broad” NIS Includes, In Addition To The Components Of The Narrow NIS, All Economic, Political, And Other Social Institutions Affecting Learning, Searching, And Exploring Activities, e.g. A Nation’s Financial System, Its Monetary Policies, And Internal Organization Of Private Firms

15. The “Broad” NIS Hybrid S&T Labs Public S&T Labs Scientific and Technological Societies Technology Sharing Regimes Technology Licensing Regimes Intellectual Property Regimes Mission Agencies Universities Private Firms Organization of Financial System User-Producer Relationships Internal Organization of Firms Industrial Organization Monetary Policies Natural Resources

16. National R&D Expenditures, By Performer: 1995 10% 9% 12% 71% 24% 67% 13% 49% 14% 3% 10% 2% 4% 7% 5% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% National R&D ($171.0 billion) Basic Research ($29.6 billion) Applied Research ($39.8 billion) Federal GovernmentIndustryAcademiaU&C FFRDCsOther

17. Sources of Academic R&D Funding, By Sector

18. The Complexity of the NIS Author Sector of the U.S. Papers Cited in Industry Patents % of Papers

19. Distribution of citations across U.S. performer sectors, by field: 1990-93 Field IndustryFederal FFRDC NonprofitOtherAcademia All fields 70.5 7.99.7 1.9 8.8 1.2 Clinical medicine 68.35.111.3 0.2 12.9 2.2 Biomedical research 72.76.99.4 0.7 9.7 0.7 Biology 79.22.813.4 0.2 3.2 1.3 Chemistry 78.512.8 4.3 2.7 1.3 0.2 Physics 63.120.9 5.1 9.3 1.5 0.1 Earth & space sciences 66.14.814.4 7.8 6.1 0.8 Mathematics 88.84.9 2.5 1.8 1.8 0.3 Engineering and 66.319.8 7.8 4.6 1.4 0.3 Technology Impact of University Research

20. Patterns of cross-sector citations, by citing sector Citing sectorIndustryFederal FFRDC NonprofitOtherAcademia United States, total 70.56.3 10.62.29.0 1.4 Academic institutions77.14.3 8.01.67.7 1.2 Industry46.936.1 8.12.75.2 1.0 1990-1993 articles United States, total 70.57.9 9.71.98.8 1.2 Academic institutions76.55.9 7.51.57.6 1.0 Industry47.835.7 7.81.85.9 0.9 1985-1988 articles Impact of University Research

21. Support for Academic R&D, 1935, and 1960-1990 (Millions of Current Dollars) $50 $646 $1,474 $2,335 $3,409 $6,077 $9,686 $16,000 24% 63% 73% 71% 68% 63% 67% 58% 0 2000 4000 6000 8000 10000 12000 14000 16000 19351960196519701975198019851990 Millions of Dollars 0% 10% 20% 30% 40% 50% 60% 70% 80% % Federally Supported Total Academic R&D% Federally Supported

22. The Complexity of the NIS Institutional Origin of Papers Cited in IBM Patents

24. The Components of the NIS Have Different Effects and Operate Differently Across Industries; For Example: University Science More Relevant to Some Industries than Others Different Extraindustry Sources of Technological Knowledge Across Different Industries Effectiveness of Patents Varies Across Industries

25. THE RELEVANCE OF UNIVERSITY SCIENCE TO INDUSTRIAL TECHNOLOGY Source: Rosenberg and Nelson (1994)

26. INDUSTRIES RATING UNIVERSITY RESEARCH AS “IMPORTANT” OR “VERY IMPORTANT” Fluid milk Dairy products except milk Canned specialties Logging and sawmills Semiconductors and related devices Pulp, paper, and paperboard mills Farm machinery and equipment Grain mill products Pesticides and agricultural chemicals Processed fruits and vegetables Engineering and scientific instruments Millwork, veneer, and plywood Synthetic rubber Drugs Animal Feed Source: Rosenberg and Nelson (1994)

27. THE RELEVANCE OF SCIENCE TO INDUSTRIAL TECHNOLOGY Source: Rosenberg and Nelson (1994)

28. EXTRAINDUSTRY SOURCES OF TECHNOLOGICAL KNOWLEDGE Source: Levin et al. (1987)

29. EFFECTIVENESS OF PATENT PROTECTION ACROSS INDUSTRIES WITH TEN OR MORE RESPONSES (MEAN SCORE ON SCALE OF 1-7) Source: Levin et al. (1987)

30. The Evolution of the American National Innovation System

31. The Evolution of the American National Innovation System: Four Periods Laissez-Faire (1790-1940) The War and Post-War Period (1940-1950) The Federalization Period (1950-1975) The Revisionist Period (1975-1990) Source: Crow (1994)

32. The Evolution of the American National Innovation System Laissez-Faire Period: 1790-1940 A Pre-Policy Period: Government Has No Distinct Science and Technology Policy or Mission The Key Institutions in the National Innovation System: Independent Corporate R&D Labs Government Does Establish Some R&D Labs to Support Weak Industries (i.e. Mining) Beginning of the Late 1800’s: Universities Emerge as the Home of Basic Science and Advanced Training

33. The Evolution of the American National Innovation System The War and Post-War Period 1940-1950 To Support the War Effort, the Government Establishes Many New R&D Institutions and a New, Expanded Role for Academic Science During the War, Large Scale Federal Investment, Federally Mandated R&D Objectives, Targeted Funding, and Industrial and Governmental Cooperation are the Norm By the end of the War, Hundreds of New R&D Labs had been established, and the potential of Large Scale R&D for meeting national objectives is demonstrated

34. The Evolution of the American National Innovation System The War and Post-War Period 1940-1950 Following the Dramatic Change in Science and Technology Policy During the War, Policy Makers Sensed the Potential of Science and Technology to Serve the National Interest

35. The Evolution of the American National Innovation System The War and Post-War Period 1940-1950 In 1944, President Roosevelt asked Vannevar Bush, the Director of the Wartime OSRD, to Look Ahead to the Role of Science in Peacetime. Bush’s Design, Presented in Science the Endless Frontier, Became the Foundation for U.S. Science Policy

36. LINEAR TECHNOLOGY DEVELOPMENT MODEL Pure Basic Research Directed Basic Research Intermediate Range Applied Research Applied Research Tech. Develop- ment Tech. Commer- cialization FUNDAMENTAL RESEARCH AND DISCOVERY FOCUSED RESEARCH AND PRELIMINARY DEVELOPMENT FOCUSED DEVELOP- MENT MARKET DRIVEN TECH. DEVELOP- MENT Increasing Role of UniversitiesIncreasing Role of Industry Increasing Role of Government

37. The Evolution of the American National Innovation System The Bush Design Was Built Around the Following Characteristics: Political Autonomy: Self Regulation by Scientists: Focus on science for science’s sake as well as problem solving Strong academic model of individual achievement General Accountability(linked to broad objectives of national well being) Single Major Basic Research Agency Limited resources for only the best scientists

38. The Evolution of the American National Innovation System: The Bush Design Political Autonomy Separation from Political Control Separate Governance

39. The Evolution of the American National Innovation System: The Bush Design Self-Regulation by Scientists Peer-Review

40. The Evolution of the American National Innovation System: The Bush Design Focus on Science for Science’s Sake As Well as Problem Solving Basic Science/Fundamental Discovery Applied Science

41. The Evolution of the American National Innovation System: The Bush Design Strong Academic Model of Individual Achievement Scientists as Individual Thinkers

42. The Evolution of the American National Innovation System: The Bush Design General Accountability (Linked to Broad Objectives of National Well-Bring) Success Measured by Overall National Achievement

43. The Evolution of the American National Innovation System: The Bush Design Single Major Basic Research Agency NSF in original design

44. The Evolution of the American National Innovation System: The Bush Design Limited Resources for Only the Best Scientists Small Budgets

45. The Evolution of the American National Innovation System Federalization Period: 1950-1975 By the end of the period, five types of institutions were important in the NIS: –Hundreds of Large Industrial Labs –Dozens of Large Federal Labs –Thousands of Small Technology Oriented Labs and Companies –Hundreds of Unconnected and Unplanned Federal Labs –Researchers at Universities

46. The Evolution of the American National Innovation System The Revisionist Period 1975-1990 Economic and Technological Position of the United States began to slip The Bush model prevailed: Research dollars concentrated on defense and on basic science However, pushed by local political demands, Congress did make some attempts to make to U.S. more competitive and to improve upon the Bush model

47. The Evolution of the American National Innovation System The Revisionist Period 1975-1990 Major Efforts to Change Science Policy Stevenson-Wydler Technology Act (1980) Bayh-Dole Act (1982) National Productivity and Innovation Act (1983) Federal Technology Transfer Act (1986)

48. The American NIS Today Today, the design parameters for basic science and the cultural design for basic science and technology remain essentially those suggested by Bush.

49. The American NIS Today The Bush design is in serious need of updating and improvement, and has been for some time. The rationale for updating is simply that Bush failed to build into the system the feedback and response mechanisms needed for a post-industrial democracy.

50. The American NIS Today In updating the Bush design, we must keep in mind that the NIS today is a complex web of institutions, actors structures, and relationships. We cannot completely overhaul it while it is in motion. We must be aware of the size and the complexity in the system before prescribing change

51. The American NIS Today: Examples of Size, Complexity Interactions between Public, Private, and Hybrid Science and Technology Labs Government Funding of Academic Basic Research, Applied Research, and Development Percentage of New Products and Processes Based on Recent Academic Research

52. Distribution of R&D Laboratory Type circa 1995-2005 Public Knowledge and Technology Products Private Knowledge and Technology Products Private Technology Labs Public Science Labs Private Science Labs Public S&T Labs Hybrid Science Labs Private S&T Labs Hybrid Technology Labs Hybrid S&T Labs Public Technology Labs

53. SUPPORT FOR ACADEMIC R&D, 1935, AND 1960- 1990 (MILLIONS OF CURRENT DOLLARS). Source: Rosenberg and Nelson (1994)

54. PERCENT OF FEDERAL RESEARCH FUNDS ORIGINATING WITHING PARTICULAR AGENCIES Source: Rosenberg and Nelson (1994)

55. FEDERAL AND NONFEDERAL R&D EXPENDITURES AT UNIVERSITIES AND COLLEGES, BY FIELD AND SOURCE OF FUNDS, 1989 Source: Rosenberg and Nelson (1994)

56. EXPENDITURES FOR BASIC RESEARCH, APPLIED RESEARCH, AND DEVELOPMENT, 1960-1990 (MILLIONS OF CURRENT DOLLARS) Source: Rosenberg and Nelson (1994)

57. UNIVERSITY-INDUSTRY RELATIONS Over the past two decades, there has been a significant increase in the fraction of academic research funded by industry and in the number and size of university-industry research centers. Some academics, while welcoming this trend, do not want industries to influence the research orientation of universities. Other academics both welcome industry funding and are eager to re- orient their research to make it more commercially relevant and rewarding. In the 1980s, industry leaders were enthusiastic about the ability of academics to contribute to technical advance in industry. Today, however, there is considerable skepticism in industry: a prevailing view is that academics should stick to basic research and training young scientists, and to stop thinking of themselves as the sources of new technology. Source: Rosenberg and Nelson (1994)

58. % OF NEW PRODUCTS AND PROCESSES BASED ON RECENT ACADEMIC RESEARCH, U.S., 1975-1985 Source: Rosenberg and Nelson (1994)

59. The American NIS Today: Updating the Bush Design

60. FREEMAN’S “THREE PHASES” OF SCIENCE POLICY Source: Crow (1994)

61. RECOMMENDATION I Political Autonomy Establishment of an institutional mechanism for forecasting our national science and technology needs

62. DESIGN PARAMETER I: POLITICAL AUTONOMY 1. CONGRESS should establish a National Science and Technology Forecasting Institute 2.OFFICE OF SCIENCE AND TECHNOLOGY POLICY would use the National Science and Technology Forecasting Institute to identify the specific S&T objective of a particular administration 3. MISSION AGENCIES would develop research agendas with regard to the S&T forecast 4. NSF’s research agenda and areas of focus would be mapped according to the S&T forecast

63. RECOMMENDATION II Self Regulation by Scientists Spending time and resources on educating the public about science and research Development of a formal science “court” for internal discipline and conflict resolution Broadening the criteria for peer review to include potential for social profit

64. DESIGN PARAMETER II: SELF REGULATION BY SCIENTISTS c Congress and the President would establish the U.S. Science Court. National Science Board would establish a greatly expanded public information and access program. Research Agencies would develop expanded criteria for project selection.

65. RECOMMENDATION III Focus on “Science for Science’s sake” as well as Problem Solving Eliminate references to “basic” and “applied” research projects without specific definitions of these projects Evaluate projects with regard to their purpose, realizing that the type of research conducted (basic, applied, and fundamental technology development) are functions of the missions of agencies Consider all projects and program areas as equal, regardless of scientific focus or technical objective, until they are evaluated

66. DESIGN PARAMETER III Focus on Science and Problem Solving 1. OMB would establish meaningful budget categories and improved nomenclature for defining research activity. Research would be classified by purpose and not by “function”. 2. OSTP would develop project and program classification nomenclature for research type and purpose for uniform use in all agencies. 3. Mission Agency research agendas would be contextually placed with regard to the S&T forecast. 4. NSF’s research agenda and areas of focus would be mapped according to the context setting forecast.

67. RECOMMENDATION IV Strong Academic Model of Individual Achievement Enhanced team funding mechanisms Expanded recognition mechanisms for team participation Evaluation of scientists on a group and disciplinary basis Including contributions to other fields or departments in the evaluation of particular fields or departments Heavy funding of “star” groups

68. DESIGN PARAMETER IV: STRONG ACADEMIC MODEL OF INDIVIDUAL ACHIEVEMENT 1. OMB would permit institution building among dispersed research enterprises such as universities. 2. Mission Agencies would concentrate funding on the best labs and roups. 3. NSF would concentrate funding by increasing grant size, develop more center type R&D efforts, and provide for enhanced linkage building between and among research groups at different institutions.

69. RECOMMENDATION V General Accountability Evaluating agency research programs based on their success or failure in attaining particular pre-defined goals and objectives Integrate these evaluations into funding and priority setting models

70. DESIGN PARAMETER V: GENERAL ACCOUNTABILITY 1. OSTP/Congress would set annual five and ten year objectives for National Science and Technology investment. 2. All Research Agencies would establish formal R&D evaluation capabilities at the agency and division levels.

71. RECOMMENDATION VI Single Basic Research Agency Define roles and functions of agencies with greater care Place National Science Foundation (NSF) in charge of building foundation knowledge and research tools for other programs of research Reorient NSF research agenda towards research foundation building needs

72. DESIGN PARAMETER VI: SINGLE BASIC SCIENCE AGENCY 1. Congress would require a linked science budget plan indicating who is doing what and how the NSF budget request is linked. 2. Research Agencies would re-think budget and planning models to define their roles as producers of foundation knowledge, basic knowledge, or specific solutions to problems.

73. RECOMMENDATION VII Limited Resources for only the Best Agencies should concentrate their resources in those fields of greatest importance to their individual missions Increase the size of average grants: more funding for fewer groups

74. DESIGN PARAMETER VII: LIMITED RESOURCES FOR ONLY THE BEST All Research Agencies would re-think allocation models so as to begin concentration of resources to the best research groups and labs. New allocation models would be based on: - Scientific Track Records - Institutional Infrastructure - Quality of science and support groups - Overall goal attainment