[House Hearing, 111 Congress]
[From the U.S. Government Publishing Office]


 
                        INVESTING IN HIGH-RISK,
                          HIGH-REWARD RESEARCH

=======================================================================

                                HEARING

                               BEFORE THE

                      SUBCOMMITTEE ON RESEARCH AND
                           SCIENCE EDUCATION

                  COMMITTEE ON SCIENCE AND TECHNOLOGY
                        HOUSE OF REPRESENTATIVES

                     ONE HUNDRED ELEVENTH CONGRESS

                             FIRST SESSION

                               __________

                            OCTOBER 8, 2009

                               __________

                           Serial No. 111-55

                               __________

     Printed for the use of the Committee on Science and Technology


     Available via the World Wide Web: http://www.science.house.gov


                  U.S. GOVERNMENT PRINTING OFFICE
52-484                    WASHINGTON : 2010
-----------------------------------------------------------------------
For Sale by the Superintendent of Documents, U.S. Government Printing Office
Internet: bookstore.gpo.gov  Phone: toll free (866) 512-1800; (202) 512ï¿½091800  
Fax: (202) 512ï¿½092104 Mail: Stop IDCC, Washington, DC 20402ï¿½090001



                                 ______

                  COMMITTEE ON SCIENCE AND TECHNOLOGY

                   HON. BART GORDON, Tennessee, Chair
JERRY F. COSTELLO, Illinois          RALPH M. HALL, Texas
EDDIE BERNICE JOHNSON, Texas         F. JAMES SENSENBRENNER JR., 
LYNN C. WOOLSEY, California              Wisconsin
DAVID WU, Oregon                     LAMAR S. SMITH, Texas
BRIAN BAIRD, Washington              DANA ROHRABACHER, California
BRAD MILLER, North Carolina          ROSCOE G. BARTLETT, Maryland
DANIEL LIPINSKI, Illinois            VERNON J. EHLERS, Michigan
GABRIELLE GIFFORDS, Arizona          FRANK D. LUCAS, Oklahoma
DONNA F. EDWARDS, Maryland           JUDY BIGGERT, Illinois
MARCIA L. FUDGE, Ohio                W. TODD AKIN, Missouri
BEN R. LUJAN, New Mexico             RANDY NEUGEBAUER, Texas
PAUL D. TONKO, New York              BOB INGLIS, South Carolina
PARKER GRIFFITH, Alabama             MICHAEL T. MCCAUL, Texas
STEVEN R. ROTHMAN, New Jersey        MARIO DIAZ-BALART, Florida
JIM MATHESON, Utah                   BRIAN P. BILBRAY, California
LINCOLN DAVIS, Tennessee             ADRIAN SMITH, Nebraska
BEN CHANDLER, Kentucky               PAUL C. BROUN, Georgia
RUSS CARNAHAN, Missouri              PETE OLSON, Texas
BARON P. HILL, Indiana
HARRY E. MITCHELL, Arizona
CHARLES A. WILSON, Ohio
KATHLEEN DAHLKEMPER, Pennsylvania
ALAN GRAYSON, Florida
SUZANNE M. KOSMAS, Florida
GARY C. PETERS, Michigan
VACANCY
                                 ------                                

             Subcommittee on Research and Science Education

                 HON. DANIEL LIPINSKI, Illinois, Chair
EDDIE BERNICE JOHNSON, Texas         VERNON J. EHLERS, Michigan
BRIAN BAIRD, Washington              RANDY NEUGEBAUER, Texas
MARCIA L. FUDGE, Ohio                BOB INGLIS, South Carolina
PAUL D. TONKO, New York              BRIAN P. BILBRAY, California
PARKER GRIFFITH, Alabama                 
RUSS CARNAHAN, Missouri                  
BART GORDON, Tennessee               RALPH M. HALL, Texas
               DAHLIA SOKOLOV Subcommittee Staff Director
            MARCY GALLO Democratic Professional Staff Member
           MELE WILLIAMS Republican Professional Staff Member
                    BESS CAUGHRAN Research Assistant


                            C O N T E N T S

                            October 8, 2009

                                                                   Page
Witness List.....................................................     2

Hearing Charter..................................................     3

                           Opening Statements

Statement by Representative Daniel Lipinski, Chairman, 
  Subcommittee on Research and Science Education, Committee on 
  Science and Technology, U.S. House of Representatives..........     9
    Written Statement............................................    10

Statement by Representative Vernon J. Ehlers, Ranking Minority 
  Member, Subcommittee on Research and Science Education, 
  Committee on Science and Technology, U.S. House of 
  Representatives................................................    10
    Written Statement............................................    12

                               Witnesses:

Dr. Neal F. Lane, Malcolm Gillis University Professor and Senior 
  Fellow, James A. Baker III Institute for Public Policy, Rice 
  University
    Oral Statement...............................................    13
    Written Statement............................................    14
    Biography....................................................    20

Dr. James P. Collins, Assistant Director, Directorate for 
  Biological Sciences, National Science Foundation
    Oral Statement...............................................    21
    Written Statement............................................    23
    Biography....................................................    33

Dr. Richard D. McCullough, Vice President for Research; Professor 
  of Chemistry, Carnegie Mellon University
    Oral Statement...............................................    33
    Written Statement............................................    35
    Biography....................................................    39

Dr. Gerald M. Rubin, Vice President and Director, Janelia Farm 
  Research Campus, Howard Hughes Medical Institute
    Oral Statement...............................................    39
    Written Statement............................................    41
    Biography....................................................    47

Discussion.......................................................    47

             Appendix 1: Answers to Post-Hearing Questions

Dr. Neal F. Lane, Malcolm Gillis University Professor and Senior 
  Fellow, James A. Baker III Institute for Public Policy, Rice 
  University.....................................................    62

Dr. James P. Collins, Assistant Director, Directorate for 
  Biological Sciences, National Science Foundation...............    64

Dr. Richard D. McCullough, Vice President for Research; Professor 
  of Chemistry, Carnegie Mellon University.......................    66

Dr. Gerald M. Rubin, Vice President and Director, Janelia Farm 
  Research Campus, Howard Hughes Medical Institute...............    67

             Appendix 2: Additional Material for the Record

Statement of Professor Franklin M. Orr, Jr., Stanford University, 
  representing the David and Lucile Packard Foundation...........    70


              INVESTING IN HIGH-RISK, HIGH-REWARD RESEARCH

                              ----------                              


                       THURSDAY, OCTOBER 8, 2009

                  House of Representatives,
    Subcommittee on Research and Science Education,
                       Committee on Science and Technology,
                                                    Washington, DC.

    The Subcommittee met, pursuant to call, at 1:07 p.m., in 
Room 2318 of the Rayburn House Office Building, Hon. Daniel 
Lipinski [Chairman of the Subcommittee] presiding.



                            HEARING CHARTER

             SUBCOMMITTEE ON RESEARCH AND SCIENCE EDUCATION

                  COMMITTEE ON SCIENCE AND TECHNOLOGY

                     U.S. HOUSE OF REPRESENTATIVES

                        Investing in High-Risk,

                          High-Reward Research

                       thursday, october 8, 2009
                          1:00 p.m.-3:00 p.m.
                   2318 rayburn house office building

1. Purpose

    The purpose of this hearing is to examine mechanisms for funding 
high-risk, potentially high-reward research, and the appropriate role 
of the Federal Government in supporting such research.

2. Witnesses:

  Dr. James P. Collins, Assistant Director for Biological 
Sciences, National Science Foundation.

  Dr. Neal F. Lane, Malcolm Gillis University Professor and 
Senior Fellow, James A. Baker III Institute for Public Policy, Rice 
University. Dr. Lane was a member of the American Academy of Arts & 
Sciences committee that published the report, ARISE: Advancing Research 
in Science and Engineering.

  Dr. Richard D. McCullough, Professor of Chemistry and Vice 
President of Research, Carnegie Mellon University.

  Dr. Gerald M. Rubin, Vice President and Director, Janelia 
Farm Research Campus, Howard Hughes Medical Institute.

3. Overarching Questions:

  What is high-risk, high-payoff research? How does it differ 
from the research traditionally funded by federal science agencies? 
What metrics should be used to evaluate the success of any approach to 
funding high-risk research?

  Relative to the total funding for basic science and 
engineering research from all sources, is the current level of support 
for high-risk research appropriate? If funding for high-risk research 
were to be increased as recommended in several recent reports, what 
should be the responsibility of the Federal Government in achieving 
that increase, and how does that responsibility differ from that of 
private sector research organizations and funding sources as well as 
research universities?

  How can federal science agencies such as the National Science 
Foundation (NSF) increase their support for high-risk research? In 
particular, what are the pros and cons of establishing targeted 
programs or set-asides for high-risk research versus changing how 
proposals are reviewed and selected across an agency's research 
portfolio? What are the biggest challenges or risks associated with 
each of these approaches?

4. Background

What is high-risk, high-reward research?
    The terms `high-risk, high-reward' (or `high-risk, high-payoff') 
and `transformative' research are often used interchangeably. The 
National Science Board has proposed the following definition for 
transformative research:

         Transformative research is defined as research driven by ideas 
        that have the potential to radically change our understanding 
        of an important existing scientific or engineering concept or 
        leading to the creation of a new paradigm or field of science 
        or engineering. Such research is also characterized by its 
        challenge to current understanding or its pathway to new 
        frontiers.

    The Board, mindful of NSF's unique role in funding basic research 
across the disciplines, says nothing in its definition about research 
leading to new technologies or solutions to societal challenges. 
Federal mission agencies, on the other hand, use a mission inspired 
definition for high-risk, high-reward research, or some comparable 
term. For example, a few years ago NIH created the Pioneer Awards for 
this purpose.

         The term ``pioneering'' is used to describe highly innovative 
        approaches that have the potential to produce an unusually high 
        impact on a broad area of biomedical or behavioral research.

    A handful of philanthropic organizations also invest in high-risk 
research. One such organization, the Keck Foundation, makes a 
distinction between ``high-risk'' and ``transformative'' as follows:

         ``High-risk'' comprises a number of factors, including 
        questions that push the edge of the field, present 
        unconventional approaches to intractable problems, or challenge 
        the prevailing paradigm. ``Transformative'' may mean creation 
        of a new field of research, development of new instrumentation 
        enabling observations not previously possible, or discovery of 
        knowledge that challenges prevailing perspectives.

    What is common to all definitions of high-risk, high-reward, or 
transformative (or pioneering) research is a tolerance for failure that 
departs from the overwhelming tendency, within the federal system at 
least, to fund research for which there is already a proof of concept 
or preliminary data, and for which the likelihood of achieving the 
stated aims is pretty high. In other words, scientists and engineers 
are not encouraged by the current federal funding system to propose 
their wildest (but scientifically sound) ideas; rather, they believe 
their only chance at getting funded is to propose something that they 
already know will work.\1\ The resulting incremental advances in 
science and engineering are a necessary, but not sufficient element of 
the science and technology enterprise. In many if not most cases, great 
breakthroughs and paradigm shifts in S&T were the result of scientists 
and engineers stumbling upon some unexpected result or suddenly 
imagining some new application and then having the funding and/or 
flexibility to alter their research plans accordingly.
---------------------------------------------------------------------------
    \1\ One historically successful federal model for funding high-risk 
research is DARPA, credited with funding early development of the 
Internet, not to mention countless advanced military technologies. In 
2007, the S&T Committee applied the DARPA model to the Department of 
Energy (DOE) by creating ARPA-E. ARPA-E invests in technologies that 
will promise true transformations in how we use or produce energy--what 
DOE describes on their web site as high-risk, high-payoff concepts. 
While there may be elements of DARPA and ARPA-E that are broadly 
applicable to all models for funding high-risk research, the ARPA model 
is driven by a need for mission-specific technologies, making it 
inappropriate for replication in basic science agencies.

The call for a greater federal role in funding high-risk research
    In 2006, the National Academies Committee on Prospering in the 
Global Economy of the 21st Century released the report, Rising Above 
the Gathering Storm, that became both the impetus and intellectual 
foundation for the 2007 America COMPETES Act. In addition to the many 
recommendations regarding K-12 STEM education, funding for basic 
research in the non biomedical sciences, and creation of an ARPA-E that 
were implemented as part of the COMPETES Act, the Academies Committee 
recommended that at least eight percent of the budgets of federal 
research agencies should be set aside for discretionary funding managed 
by technical program managers in those agencies to catalyze high-risk, 
high-payoff research. They provided no further details on how that 
might be done and chose eight percent because it was a compromise 
between committee members who thought five percent was sufficient and 
those who argued for 10 percent.
    In 2004, the National Science Board convened a task force on 
transformative research to make recommendations on how the National 
Science Foundation (NSF) could encourage more funding of high-risk, 
potentially high-reward research. In the resulting 2007 report,\2\ the 
Board recommended that NSF develop a distinct, Foundation-wide 
Transformative Research Initiative ``distinguishable by its potential 
impact on prevailing paradigms and by the potential to create new 
fields of science, to develop new technologies, and to open new 
frontiers.'' Beyond defining transformative research and stating that 
the NSF Director's leadership is essential its success, the Board did 
not go into any details on how such an initiative should be carried 
out, nor did it recommend a specific percentage of the NSF budget for 
investment in transformative research.
---------------------------------------------------------------------------
    \2\ http://www.nsf.gov/pubs/2007/nsb0732/nsb0732.pdf
---------------------------------------------------------------------------
    Perhaps in recognition of the absence of details in these reports, 
the American Academy of Arts and Sciences launched a new study in 2007 
to develop specific recommendations for how federal agencies, 
universities and private foundations can encourage more high-risk, 
high-reward research, even absent significant growth in overall 
research budgets. The Academy assembled a distinguished committee of 
Nobel Laureates, (former) agency and National Lab directors, university 
presidents, private research organization directors and other notables 
for this purpose. The Committee also addressed support for early-career 
faculty, which shares some challenges in common with support for high-
risk research. The resulting report, Advancing Research in Science and 
Engineering: Investing in Early-Career Scientists and High-Risk, High-
Reward Research (ARISE),\3\ was completed in 2008.
---------------------------------------------------------------------------
    \3\ http://www.amacad.org/AriseFolder/

Role of Charitable Organizations and Universities
    According to NSF, non-federal, non-business entities provided $23 
billion in funding for R&D in the United States in FY 2006, out of a 
total of $340 billion from all sources. This ``other'' category is 
pretty broad, including state and local governments, nonprofit 
organizations (e.g., charitable foundations), and universities. Funding 
for academic R&D in FY 2006 totaled $48 billion.\4\ Institutional 
(university) funds accounted for $9.1 billion, or 19 percent of that 
total. A different category of ``other'' sources of funds for academic 
R&D, including nonprofit organizations and gifts from private 
individuals, accounted for $3.2 billion, or seven percent of all 
academic R&D in FY 2006.\5\
---------------------------------------------------------------------------
    \4\ Institutional funds encompass: 1) institutionally financed 
organized research expenditures, and 2) unreimbursed indirect costs and 
related sponsored research.
    \5\ From 2008 Science and Engineering Indicators: http://
www.nsf.gov/statistics/seind08/?org=NSF
---------------------------------------------------------------------------
    There are many charitable foundations of varying size that fund 
what they consider to be high-risk, high-reward research, sometimes at 
universities and sometimes in their own, privately run research labs. 
The Keck Foundation, for example, funds academic research projects 
across all disciplines that might not be funded otherwise. Keck's 
evaluation criteria are: 1) is this idea scientifically sound?; 2) if 
anyone can pull it off, can this particular individual/team?; and 3) 
does this individual/team have the tools at their disposal to carry out 
this research? In other words, Keck takes a chance on people with 
strong track records and access to first class research facilities. The 
Howard Hughes Medical Institute (HHMI) similarly takes a chance on the 
reputation of individual scientists, but HHMI investigators become HHMI 
employees, freeing them from the constant pursuit of federal support, 
or they join HHMI's own world class research campus, severing ties with 
their home institutions altogether. Some foundations make lump-sum 
grants to universities and rely on the leadership within the university 
to run an internal competition for the best ideas.
    Institutions also support their own faculty, in particular by 
providing start-up funds to newly recruited faculty. In the case of 
young investigators just starting out, the new faculty need money to 
build their labs and gather preliminary data before they can apply for 
federal funding with a reasonable chance of success. But universities 
may also offer generous packages to well established scientists 
recruited from other universities. In general, institutional funding 
may provide more flexibility for faculty wanting to pursue high-risk 
ideas than do standard federal research grants.

Challenges and Approaches to Investing in High-Risk Research
    There is little doubt that flat research budgets and low proposal 
success rates across agencies such as NSF and NIH have contributed to 
more conservative funding decisions on the part of peer review panels. 
When budgets are constrained and success rates low, a single critical 
review by a peer may be sufficient to scuttle a proposal. Human nature 
surely plays a role as well. As an expert in the same field as the 
applicant, the critical reviewer may have his or her own career 
invested in the paradigm being challenged by the applicant. The peer-
review system is, on balance, strong, functional and successful, but it 
is not perfect.
    In general, there are two approaches to funding more high-risk 
research, described in detail in the ARISE report: creation of targeted 
programs or grant mechanisms, or systemic reform of the current peer-
review process.
    In the case of targeted programs or grant mechanisms, the agencies, 
or Congress, must decide how much of the total research dollars to set 
aside for this purpose. The National Science Foundation has such a 
mechanism already, one that they have had in place for a number of 
years. It was called Small Grants for Exploratory Research (SGER) and 
just this year (partially to satisfy a requirement in the COMPETES Act) 
was split into two programs: Exploratory Grants for Early Research 
(EAGER), and RAPID grants for urgent response research, typically after 
a natural disaster.
    EAGER grants are reviewed only internally at NSF and may be up to 
$300,000 and for up to two years in duration. Program officers were 
allowed to use up to five percent of their program budget for the 
former SGER awards. In FY 2008, a total of 389 SGER grants were awarded 
across all directorates, accounting for only 0.6 percent of NSF 
research obligations.\6\ The directorate that made the most use of SGER 
grants was Computer and Information Sciences and Engineering (CISE), at 
1.9 percent. Similarly, NIH has its Pioneer Awards, but they account 
for about 0.01 percent of NIH's total budget and have a dismal success 
rate that discourages many potential applicants.
---------------------------------------------------------------------------
    \6\ For a directorate by directorate breakdown, see Appendix 8 of 
the NSB's 2008 Merit Review Report: http://www.nsf.gov/nsb/
publications/2009/
nsb0943-merit-review-2008.pdf
---------------------------------------------------------------------------
    The ARISE Committee also makes a number of recommendations for 
strengthening the entire system to support more high-risk research, 
from changing the make-up of review panels to altering the charge to 
those panels. Finally, the ARISE Committee recommends greater 
investment in agency program officers to strengthen program leadership 
and facilitate the injection of new ideas into agency and community 
deliberations.
    In the FY 2010 budget request, NSF announced a new Foundation-wide 
transformative research initiative in which each research division will 
set aside a minimum of $2 million ($92 million Foundation-wide) to 
explore methodologies that help support transformative research.

Metrics for Success
    The ARISE Committee also took on the question of how to measure the 
success of any new policy or program to support high-risk research. 
They recommended evaluating programs in two phases. The first phase 
involves determining whether the new program or policy was successful 
in attracting high-risk research proposals and in funding proposals 
that would normally be rejected under the traditional peer-review 
system. The second phase should occur no sooner than 10 years after the 
initiation, according to the Committee, and would involve evaluation of 
scientific outcomes.
    Evaluating the effectiveness or impact of any basic research 
program is a difficult, perhaps impossible task, thereby making them 
easy targets during the zero-sum game appropriations battles. Policies 
or programs for high-risk research, therefore, could face even greater 
uncertainty in the federal budget process. For that reason, some argue 
that charitable organizations and universities are better positioned to 
ensure long-term support for high-risk research.

5. Questions for Witnesses

James Collins, NSF

        1.  Please describe the National Science Foundation's (NSF) 
        proposed transformative research initiative. What definition is 
        NSF using for `potentially transformative research?' What 
        guidance has been provided to research divisions regarding 
        implementation of this initiative and how was that guidance 
        developed? To what extent does this initiative entail targeted 
        programs and grant mechanisms versus modifying the standard 
        grant review process across the Foundation? To what extent does 
        it overlap with initiatives to support young investigators? How 
        will NSF evaluate the impact of its transformative research 
        initiative?

        2.  How in particular is your directorate, Biological Sciences, 
        planning to implement and evaluate the transformative research 
        initiative?

        3.  What is the role of the program officer in identifying and 
        funding potentially transformative research? What guidance is 
        provided to program officers regarding their role? To what 
        extent does that guidance vary across disciplines/divisions? 
        What has been the impact of flat agency operations budgets on 
        program officers' ability to identify and support potentially 
        transformative research proposals?

        4.  Is there a unique role for NSF versus the university and 
        the private sector in investing in potentially transformative 
        research? How can NSF's models for support of potentially 
        transformative research complement or facilitate university as 
        well as private sector, including philanthropic support for 
        such research?

Neal Lane, Rice University

        1.  What were the key findings and recommendations in the 2008 
        American Academy of Arts and Sciences report, ``Advancing 
        Research in Science and Engineering (ARISE): Investing in 
        Early-Career Scientists and High-Risk, High-Reward Research.'' 
        In particular, what were the key findings and recommendations 
        with respect to support for high-risk, high-reward research, 
        especially in non-biomedical disciplines?

        2.  What are the pros and cons of establishing targeted 
        programs or set-asides for high-risk research versus changing 
        how proposals are reviewed and selected across a federal 
        science agency? What are the biggest challenges or risks 
        associated with each of these approaches? What metrics should 
        be used to evaluate the success of any approach to funding 
        high-risk research?

        3.  What are the appropriate roles and responsibilities of the 
        various funders, including the federal science agencies, the 
        private sector and universities themselves, in supporting high-
        risk research? How can federal investments in high-risk 
        research be used to leverage private sector and university 
        investments, and vice-versa?

Richard McCullough, Carnegie Mellon University

        1.  What percentage of science and engineering research funding 
        at your institution comes from the Federal Government? The 
        private sector? The university itself? How do the proposal 
        selection methods and criteria vary across the funding sources?

        2.  Which of the funding sources described previously provides 
        the most flexibility to your faculty to pursue high-risk, high-
        reward (or `transformative') research? Do all of your science 
        and engineering faculty have equal access to those sources (or 
        types of sources) of funding given meritorious proposals?

        3.  Given the total funding for academic science and 
        engineering research from all sources, is the ratio of funding 
        for high-risk research appropriate? If the ratio were to be 
        increased as recommended in several recent reports, what should 
        be the responsibility of the Federal Government in achieving 
        that increase, and how does that responsibility differ from 
        that of the university itself and the private sector?

        4.  Do you have any specific recommendations for how federal 
        science agencies such as the National Science Foundation could 
        increase their support for high-risk research? In particular, 
        what are the pros and cons of establishing targeted programs or 
        set-asides for high-risk research versus changing how proposals 
        are reviewed and selected across a federal science agency? What 
        are the biggest challenges or risks associated with each of 
        these approaches? What metrics should be used to evaluate the 
        success of any approach to funding high-risk research?

Gerald Rubin, Howard Hughes Medical Institute

        1.  What is Howard Hughes Medical Institute's model for funding 
        high-risk, high-payoff research? What are the benefits of this 
        model? What are the challenges? Is this a model that could or 
        should be duplicated by federal funding agencies or federally 
        funded research and development centers such as the Department 
        of Energy National Labs or the National Institutes of Health?

        2.  Given the total funding for basic science and engineering 
        research from all sources, is the ratio of funding for high-
        risk research appropriate? If the ratio were to be increased as 
        recommended in several recent reports, what should be the 
        responsibility of the Federal Government in achieving that 
        increase, and how does that responsibility differ from that of 
        private sector research organizations and funding sources such 
        as HHMI?

        3.  Do you have any specific recommendations for how federal 
        science agencies such as the National Science Foundation could 
        increase their support for high-risk research? In particular, 
        what are the pros and cons of establishing targeted programs or 
        set-asides for high-risk research versus changing how proposals 
        are reviewed and selected across a federal science agency? What 
        are the biggest challenges or risks associated with each of 
        these approaches? What metrics should be used to evaluate the 
        success of any approach to funding high-risk research?
    Chairman Lipinski. Good afternoon and welcome to this 
Research and Science Education Subcommittee hearing on high-
risk, high-reward research.
    Before I start, it is important to make clear that high-
risk, high-reward research is also known by many other names 
including high-risk, high-payoff, transformative, pioneering, 
and even high-risk, transformative research. There is neither a 
distinct definition for each of those terms nor a common 
definition for all of them. We chose high-risk, high-reward 
because it is the term used by the ARISE Committee, that is, 
the Advancing Research in Science and Engineering Committee 
whose report we will be discussing today.
    Three years ago in the now famous Rising Above the 
Gathering Storm report, a distinguished National Academies 
committee recommended that each Federal research agency set 
aside eight percent of its budget for high-risk, high-payoff 
research. Not long after that, the National Science Board 
recommended that the National Science Foundation establish a 
transformative research initiative.
    Both of those reports reflected a growing consensus in the 
research community that the peer-review system has become too 
conservative in its funding decisions and that even the 
brightest and most creative scientists are not bothering to 
submit more ambitious proposals. But both reports were also 
short on details. That same year we were working on the America 
COMPETES Act, a bill that essentially took every recommendation 
of the Gathering Storm report within the Science and Technology 
Committee's jurisdiction and translated it into law--that is, 
every recommendation except the one that set aside eight 
percent at every research agency for high-risk research. We all 
agreed there was an unmet need, and the Senate even made a 
commendable attempt to implement that recommendation in their 
bill, but during conference we all agreed to put off 
implementing this recommendation until we could better answer 
these questions. First, what exactly is high-risk research? 
Second, why eight percent? Third, why a set-aside as opposed to 
reforming the peer-review system? Fourth, does this really make 
sense for every federal research agency? And fifth, does this 
make sense for Federal agencies at all?
    As we look ahead to our 2010 reauthorization of the America 
COMPETES Act and how we can address high-risk research in that 
bill, we turn to the distinguished panelists before us today 
and to their many expert colleagues in the community to help us 
answer these questions.
    My colleagues and I up here on the dais also have a 
political challenge. Whatever your choice of words or 
definitions, high-risk research means more failures in the 
short-term, and funding for failures is not easy to justify in 
an era of ballooning deficits. It is hard enough to secure 
sustainable funding increases for basic research, and it is all 
too easy to cut science in appropriations battles. As the 
Energy and Water Appropriations Subcommittee Chair once said in 
response to concerns about cuts to the DOE Office of Science, 
floods kill people. So often times, funding does not seem to be 
at the highest end of priorities.
    Therefore, I worry even more about the risks of creating a 
discretionary pot of funding that a priori assumes a large 
failure rate. I say that to remind us all of the political 
context that surrounds our discussions this afternoon, and I 
certainly welcome any thoughts Dr. Lane may have on that topic 
given his many years of experience in Washington.
    I want to thank all the witnesses for being here today, and 
I look forward to your testimony.
    The Chair now recognizes Dr. Ehlers for an opening 
statement.
    [The prepared statement of Chairman Lipinski follows:]

             Prepared Statement of Chairman Daniel Lipinski

    Good afternoon and welcome to this Research and Science Education 
Subcommittee hearing on high-risk, high-reward research. Before I 
start, it is important to make clear that `high-risk, high-reward' 
research is also known by many other names including `high-risk, high-
payoff,' `transformative,' `pioneering,' and even `high-risk, 
transformative' research. There is neither a distinct definition for 
each of those terms nor a common definition for all of them. We chose 
`high-risk, high-reward' because it is the term used by the ARISE 
Committee--that is, the Advancing Research in Science and Engineering 
Committee--whose report we will be discussing today.
    Three years ago in the now famous Rising Above the Gathering Storm 
report, a distinguished National Academies committee recommended that 
each federal research agency set aside eight percent of its budget for 
`high-risk, high-payoff' (their term of choice) research. Not long 
after that, the National Science Board recommended that the National 
Science Foundation establish a `transformative' research initiative.
    Both of those reports reflected a growing consensus in the research 
community that the peer-review system has become too conservative in 
its funding decisions and that even the brightest and most creative 
scientists and engineers are not bothering to submit more ambitious 
proposals. But both reports were also short on details. That same year 
we were working on the America COMPETES Act, a bill that essentially 
took every recommendation of the Gathering Storm report within the 
Science and Technology Committee's jurisdiction and translated it into 
law. That is, every recommendation except the one to set aside eight 
percent at every research agency for high-risk research. We all agreed 
there was an unmet need, and the Senate even made a commendable attempt 
to implement that recommendation in their bill, but during conference 
we all agreed to put off implementing this recommendation until we 
could better answer these questions:

        1.  What exactly is high-risk research?

        2.  Why eight percent?

        3.  Why a set-aside as opposed to reforming the peer-review 
        system?

        4.  Does this really make sense for every federal research 
        agency?

        5.  Does this make sense for federal agencies at all?

    As we look ahead to our 2010 reauthorization of the America 
COMPETES Act and how we can address high-risk research in that bill, we 
turn to the distinguished panelists before us today and to their many 
expert colleagues in the community to help us answer these questions.
    My colleagues and I up here on the dais also have a political 
challenge. Whatever your choice of words or definitions, high-risk 
research means more failures in the short-term, and ``funding for 
failures'' is not easy to justify in an era of ballooning deficits. It 
is hard enough to secure sustainable funding increases for basic 
research, and it is all-too-easy to cut science in appropriations 
battles. [As the Energy and Water Appropriations Subcommittee Chair 
once said in response to concerns about cuts to the DOE Office of 
Science, ``floods kill people.'']
    Therefore, I worry even more about the risks of creating a 
discretionary pot of funding that a priori assumes a large failure 
rate. I say that to remind us all of the political context that 
surrounds our discussions this afternoon, and I certainly welcome any 
thoughts Dr. Lane may have on that topic given his many years of 
experience in Washington.
    I thank the witnesses for being here this afternoon and I look 
forward to your testimony.

    Mr. Ehlers. Thank you, Mr. Chairman, and I agree with your 
assessment of the situation and your viewpoint of it. Excellent 
panel that we have today.
    We will be hearing from a panel of experts testifying on 
the Federal Government funding, both risky and potentially 
rewarding research, and I will get back to that in just a few 
minutes. Tight agency research budgets and intense competition 
have established an environment of cautious, incremental 
research proposals from many scientists seeking Federal 
support. Ironically, many of the greatest challenges faced by 
our nation in health care, energy and national security may not 
be addressed in a timely manner because there are limited 
opportunities for promising ideas to be heard and funded simply 
because they are outside of the box.
    Though the National Science Board, the National Academies, 
the American Academy and others have identified the need to 
address transformative research in our basic Federal research 
portfolio, it is necessary to examine how to best facilitate 
the introduction and proliferation of this type of research. 
Transformative discoveries have emerged from federally funded 
research in spite of a lack of dedicated programs for this 
purpose. Learning how we might adapt our Federal funding system 
to elicit more ground-breaking discoveries is a worthy goal, I 
look forward to hearing the insights on this topic from our 
witnesses today. I also thank the Chairman for instigating this 
particular hearing, and I think it is badly needed.
    One other dimension I would like to give to this, and I 
addressed a panel of manufacturers yesterday, and we discussed 
research. They were concerned about the ``Valley of Death'' and 
so forth. And I pointed out to them that they should be 
supporting basic research for the simple reason that they are 
dependent on it, and they don't have the resources to do basic 
research because of the high-risk factor. We may be able to 
borrow money to make the world's next best widget, but it is 
very difficult to borrow money when you say I don't know what I 
am going to find, but I think if I do research, I will probably 
find something good and I will probably be able to make some 
money and I will probably be able to pay back the loan. But you 
don't get a loan that way, and that is the point I was simply 
trying to make.
    They should be lobbying us to do the appropriate basic 
research at the Federal level because the Federal Government 
has to take the high-risk opportunities. We can afford to miss 
a major discovery now and then, but we should be trying 
constantly to really bridge the gap and get rid of the ``Valley 
of Death.'' Do the basic research where we are taking risks 
where we won't even know what the result might be, but if we do 
100 such projects, we are likely to hit pay dirt in anywhere 
from two to ten which will more than pay for all the research.
    And one of my favorite examples which I give to lay people 
is the laser, which is today ubiquitous, truly ubiquitous. And 
yet, it was started by some research, some in Russia, some in 
the U.S. Charlie Townes is a good friend of mine, did some very 
good research indicated and had some good success. He built a 
laser and then went ahead--I don't know how much federal money 
he got, but I would guess given the value of the dollar at that 
time, it was probably not much more than a million dollars, 10 
million at the very most. Now, you add together what the laser 
industry revenues are today, and it is, multi-billions of 
dollars. Not only that, it is, as I said, a ubiquitous device 
which is extremely helpful in many areas of life, many areas of 
manufacturing, and we just take it for granted. Without Charlie 
Townes and some federal money, we might not have discovered it 
for another 10 or 20 years. That time difference alone and the 
revenue that was generated during that 10 to 20 years is more 
than enough to cover the NSF budget for a number of years.
    So I think it is self-evident that the Federal Government 
has a very serious responsibility in supporting fundamental 
basic research because that is the springboard for innovation, 
creativity, and that is the springboard in turn for 
manufacturing and economic growth.
    I yield back.
    [The prepared statement of Mr. Ehlers follows:]

         Prepared Statement of Representative Vernon J. Ehlers

    Today our subcommittee will hear from a panel of experts testifying 
on how the Federal Government should fund both risky and potentially 
rewarding research. Tight agency research budgets and intense 
competition have established an environment of cautious, incremental 
research proposals from many scientists seeking federal support. 
Ironically, many of the greatest challenges faced by our nation in 
health care, energy and national security may not be addressed in a 
timely manner because there are limited opportunities for promising 
ideas to be heard and funded simply because they are ``outside of the 
box.''
    Though the National Science Board, the National Academies, the 
American Academy and others have identified the need to address 
transformative research in our federal basic research portfolio, it is 
necessary to examine how to best facilitate the introduction and 
proliferation of this type of research. Transformative discoveries have 
emerged from federally funded research in spite of a lack of dedicated 
programs for this purpose. Learning how we might adapt our federal 
funding system to elicit more ground-breaking discoveries is a worthy 
goal, and I look forward to hearing the insights on this topic from our 
witnesses today.

    Chairman Lipinski. Thank you, Dr. Ehlers. As always, I 
appreciate your perspective that you bring here.
    Now, if there are any Members who wish to submit additional 
opening statements, your statement will be added to the record 
at this point.
    At this time I would like to introduce our witnesses. 
First, Dr. Neal Lane who is the Malcolm Gillis University 
Professor at Rice University. He also holds appointments as 
Senior Fellow of the James A. Baker III Institute for Public 
Policy. Dr. Lane was a member of the American Academy of Arts 
and Sciences Committee that published the report, ARISE: 
Advancing Research in Science and Engineering.
    Dr. James Collins is the Assistant Director for Biological 
Sciences at the National Science Foundation.
    Dr. Richard McCullough is a Professor of Chemistry and Vice 
President of Research at Carnegie Mellon University.
    And finally, Dr. Gerald M. Rubin is Vice President of the 
Howard Hughes Medical Institute and Director of the Janelia 
Farm Research Campus.
    Okay. How do you pronounce that?
    Dr. Rubin. The way you did.
    Chairman Lipinski. Very good. I should have stopped myself 
right there then.
    As our witnesses should know, you will each have five 
minutes for your spoken testimony. Your written testimony will 
be included in the record for the hearing. When you all have 
completed your testimony, we will begin with questions. Each 
Member will have five minutes to question the panel. We will 
start with Dr. Lane, so Dr. Lane?

   STATEMENT OF DR. NEAL F. LANE, MALCOLM GILLIS UNIVERSITY 
 PROFESSOR AND SENIOR FELLOW, JAMES A. BAKER III INSTITUTE FOR 
                 PUBLIC POLICY, RICE UNIVERSITY

    Dr. Lane. Thank you very much, Chairman Lipinski and 
Ranking Member Ehlers, Members of the Committee. I greatly 
appreciate the opportunity to testify today on behalf of the 
American Academy of Arts and Sciences on High-Risk, High-
Rewards Research. I commend the Subcommittee for holding a 
hearing on this topic, which I believe is vital to the future 
progress of American science and technology and innovation.
    Last year, as you noted, Mr. Chairman, the American Academy 
released the report called ARISE which stands for Advancing 
Research in Science and Engineering, and my comments today are 
drawn largely from the findings and recommendations of that 
report.
    I was privileged to serve on the ARISE Committee, which was 
chaired by Nobel laureate and former Howard Hughes Medical 
Institute President Thomas Cech.
    Many studies have focused on the need to increase the level 
of funding for science and technology research, and I am not 
here to argue against that proposition. But the Academy 
Committee took on a different question. Regardless of the 
levels of overall Federal funding, what are the things that all 
stakeholders--government, universities, foundations--must do to 
ensure the most efficient and effective use of those Federal 
funds?
    In considering this question, the Committee identified two 
issues that we felt were central to the vitality of America's 
research enterprise. One was the support of early-career 
investigators, and the second was the need to encourage high-
risk, high-reward, sometimes called transformative, research. 
The two issues are, of course, related, since many fresh, new 
ideas come from researchers who are in the early stages of 
their career.
    The ARISE report recommended several policy actions that we 
believe will strengthen the opportunities for early career 
investigators, and those recommendations are detailed in my 
written statement.
    With regard to high-risk, high-reward research, our 
Committee concluded that across virtually all government 
agencies and departments that fund science and engineering 
research, short-term, low-risk and measurable results tend to 
dominate the funding decisions. While the current system does a 
very good job of identifying meritorious proposals using peer 
review, some potentially path-breaking research is not being 
funded because it just looks too risky. So there is a lot of 
truth to the often-heard advice, ``Don't put it in your grant 
proposal unless you know it will work.''
    Some agencies recognize the problem and are taking steps to 
address it. Indeed, NSF and its National Science Board 
highlighted the issue some time ago and were encouraged by the 
statements of the Director, Arden Bement, and the inclusion in 
the President's 2010 NSF budget request of $92 million 
specifically to foster transformative research.
    We urge this Subcommittee and the Congress to build on 
these commitments by the Administration, and specifically, the 
Academy's ARISE report recommends that NSF and the other 
agencies that support science and engineering research take the 
following steps. First to establish and strengthen policies, 
programs and targeted funding mechanisms designed to foster 
potentially transformative research. Two, provide high-risk, 
high-reward research programs with sufficient support to allow 
funding of a significant portion of applicants. Three, 
establish appropriate metrics with which to evaluate the 
success of targeted research programs. Four, adopt funding 
mechanisms and policies that nurture transformative research in 
all award programs, not just those targeted at high-risk, high-
reward, research. Five, strengthen application and review 
processes. High-risk research proposals face an even greater 
challenge in a stressed peer-review system that is not equipped 
to appreciate them. Six, strengthen investments in the career 
development of agency program officers who are indispensable to 
the vitality and productivity of the entire research 
enterprise.
    Let me just comment on the last recommendation, supporting 
program officers at the agencies. As a former NSF Director--and 
actually, back in the late '70s, I was NSF Physics Division 
Director as a rotator--I experienced firsthand the commitment, 
the quality, the dedication of these program officers, and I 
have seen their workload increase and their support decrease. 
They must have the resources to be in touch with their 
communities. They must be viewed as leaders in the science and 
engineering research community, make the site visits that are 
necessary so that not only are they staying current in their 
field, but the field itself recognizes and has credibility in 
their decision making.
    A key recommendation of the ARISE report was the creation 
of targeted grant programs, specifically aimed at high-risk, 
high-reward research. There are several advantages to creation 
of targeted grant programs and some challenges which we could 
go into in the question session. The Committee also considered 
institutions other than Federal agencies that are vital to the 
enterprise, particularly universities and foundations, and has 
offered recommendations helpful to them.
    Final point, the Academy is going to have a look now at a 
second phase of this project with the impact of these modes of 
Federal funding on the universities themselves. What is it 
doing about the curriculum? What is it doing to the nature of 
the faculty culture? Are there some lessons to be learned from 
that? If the Subcommittee has interest in that issue, has 
concerns that you would like us to address, we would be pleased 
to hear from you.
    Thank you, Mr. Chairman.
    [The prepared statement of Dr. Lane follows:]

                   Prepared Statement of Neal F. Lane

    Chairman Lipinski, Ranking Member Ehlers, and Members of the 
Committee: I am Neal Lane, the Malcolm Gillis University Professor at 
Rice University. I also hold appointments as a Senior Fellow of the 
James A. Baker III Institute for Public Policy, where I am engaged in 
matters of science and technology policy, and in the Department of 
Physics and Astronomy. Prior to returning to Rice University, I served 
in the Federal Government during the Clinton Administration as 
Assistant to the President for Science and Technology and Director of 
the White House Office of Science and Technology Policy, from August 
1998 to January 2001, and as Director of the National Science 
Foundation (NSF) and member (ex officio) of the National Science Board, 
from October 1993 to August 1998.
    I am also proud to be a Fellow of the American Academy of Arts and 
Sciences and to serve on its Council. I co-chair with Charles Vest the 
Advisory Committee for the American Academy's Initiative for Science, 
Engineering, and Technology. Last year, as part of the Initiative, the 
Academy released a report, ARISE: Advancing Research In Science and 
Engineering. I am pleased to appear today on behalf of the American 
Academy to discuss the findings of the ARISE report as they apply to 
the issue of federal funding for high-risk, high-reward research.
    The American Academy of Arts & Sciences was founded in 1780 by John 
Adams and other scholar-patriots to encourage dialogue among leaders of 
science, the arts, business and public affairs. Today, the Academy is 
an independent policy research institute, engaged in the study of 
complex problems vital to our nation's future. Through its projects and 
studies, and publications like the ARISE report, the Academy pursues 
practical policy responses to pressing national and global problems. On 
behalf of the Academy I wish to thank the Subcommittee for inviting me 
to summarize briefly this report's timely findings and recommendations. 
They are, we believe, vital to the future progress and prosperity of 
the Nation.
    I would also like to acknowledge the distinguished Fellows of the 
Academy who served on the committee that developed the ARISE report. 
The group was chaired by Nobel laureate and former Howard Hughes 
Medical Institute President Thomas Cech. Committee members included 
some of the Nation's preeminent scientists and policy leaders from 
government, academia, and industry. In particular, I want to mention 
University of Maryland President C. D. (Dan) Mote, Jr. Before the 
hearing date was changed, President Mote rearranged his schedule in 
order to testify before the Subcommittee on behalf of the Academy, an 
indication of his strong commitment to the issues raised in the ARISE 
report. He was a valuable member of the committee and is a leader on 
competitiveness and science and technology research issues at his own 
university and nationally.
    Many studies have focused on the need to increase the level of 
federal funding for science and technology research in order to sustain 
America's competitive advantage. The Academy committee that generated 
the ARISE report began its deliberations with a different question: 
Regardless of the levels of overall federal research funding, what are 
the things that all stakeholders--government, universities and 
foundations--must do to ensure the most efficient and effective use of 
those federal research funds?
    In considering this question, the committee identified two issues 
central to the vitality of America's research enterprise: 1) the 
support of early-career investigators; and 2) the encouragement of 
high-risk, high-reward research.

Early-Career Faculty

    Before turning to the Committee's interest in high-risk, high-
reward research, permit me to briefly summarize key points from the 
ARISE report related to new tenure-track faculty, those most talented 
individuals who will lead our science and technology enterprise into 
the future. The two issues are, of course, related since many of the 
most novel ideas come from early-career researchers.
    In recent years, many early-career faculty have faced greater 
obstacles in launching and sustaining their careers than their senior 
colleagues. Many, probably most, early-career investigators spend 
excessive amounts of time constantly preparing and submitting multiple 
grant proposals for awards, and when they succeed, new awards often are 
inadequate in size and too short in duration. New researchers must 
sustain an intense pursuit of funding, diverting time from their 
research and teaching during the formative years of their research and 
teaching careers.
    Data from the National Science Foundation and the National 
Institutes of Health confirm worrisome trends, shared, we suspect, 
across all fields of physical sciences and engineering. In general, 
early-career investigators must compete harder, succeed less often, and 
start careers later than did older, established investigators, most of 
whom also confront intense competition for limited resources.
    While NSF and NIH have helpful trend data on early-career faculty, 
most mission agencies lack comparable data and analyses; they do not 
track demographic data on their applicant and investigator populations. 
The enterprise as a whole lacks an analytical capability to produce a 
systemic view across all agencies and fields of research.
    Early-career investigators typically have had to wait too long to 
receive their first grant. The average age of first-time NIH awardees 
has risen steadily and in 2007 stood at 42.6. In many cases, tenure-
track faculty will be facing an up-or-out tenure decision before they 
have received their first competitive grant and had time to demonstrate 
their research capability. In such cases, the university loses a 
promising faculty member and the investment it has made with a start-up 
package.
    Of new investigators who applied for NIH awards in 2007, 20.6 
percent succeed compared with 23.8 percent of established researchers, 
according to data reported by the Institutes.
    In 1980, about 33 percent of NIH individual investigator awards 
went to first-time investigators; by 2006 less than 25 percent of 
awards went to early-career investigators.
    One-half of new NSF investigators never again receive NSF funding 
after their initial awards.
    Meanwhile, NSF and NIH data confirm that the investigator 
population across the sciences and engineering is graying even as non-
tenure track ranks continue to grow.
    In light of these trends, high frustration levels and low morale 
felt by many new tenure-track researchers are being communicated to 
promising undergraduate and graduate students as they make their own 
career decisions. Discouraging our brightest students from pursuing 
research careers is an ineffective strategy for assuring our nation's 
science and technological leadership in the future.
    Despite these worrisome findings, there is some good news. The 
Obama Administration, and NSF and NIH in particular, recognize the 
importance of these issues and are taking steps to address them. There 
is evidence that mission agencies are also becoming aware of the 
particular challenges facing early-career investigators. But more must 
be done.
    Recognizing that the Subcommittee's jurisdiction does not extend to 
all of the Federal Government's science and technology research-funding 
agencies, the American Academy encourages this subcommittee and the 
Congress to support initiatives designed to strengthen incentives and 
opportunities for early-career investigators. Specifically we ask you 
to:

        1.  Monitor closely actions taken to address the needs of 
        early-career researchers across the sciences and engineering 
        disciplines;

        2.  Encourage all agencies to establish targeted programs for 
        early-career faculty;

        3.  Encourage all agencies to establish new research programs 
        only if they have sufficient fiscal support to fund a 
        reasonable percentage of applicants. Grant programs that fund a 
        very small percentage of applications are inefficient uses of 
        money, time, and effort;

        4.  Encourage agencies to give special attention to proposals 
        of early-career investigators during competitive merit review 
        and to adopt career-stage-appropriate expectations for grant 
        funding;

        5.  Encourage agencies to create seed funding programs for 
        early-career investigators to enable them to explore new ideas 
        for which no results have yet been achieved;

        6.  Encourage agencies to remove barriers affecting those who 
        serve their families as primary caregivers, for example, by 
        providing grant extensions or other appropriate support 
        mechanisms, and, finally;

        7.  Encourage agencies to collect and analyze demographic data 
        on applicants and principal investigators government-wide and 
        in a uniform format to establish a comprehensive federal 
        database on how agencies support research. The current 
        nonstandardized tracking among funding agencies hinders efforts 
        to analyze funding trends. Since NSF has an excellent track 
        record of collecting and analyzing data relevant to the future 
        of the Nation's science, engineering and technology enterprise, 
        its example could be helpful to other agencies that do not have 
        such a tradition.

High-Risk, High-Reward Research

    Turning now to high-risk, high-reward research, the ARISE report 
highlights several important themes that I believe merit consideration 
by the Subcommittee.
    Most research scientists and engineers achieve their goals by 
persistent, step-by-step work built on the discoveries and advances of 
others. This is, and must remain, the vital foundation of our research 
enterprise. Important breakthroughs do result from incremental 
research.
    Science also progresses from bold innovation in methods, 
instruments, and computer software. Curiosity-based or intuition-based 
boldness can require even greater leaps into the unpredictable unknown. 
Most such efforts will fail, but the few pioneers who are successful 
can profoundly influence the direction of science by challenging 
accepted paradigms. Such research can generate deep changes in 
concepts, create new subfields of science or bring together different 
fields to make discoveries and advances that would otherwise be 
impossible. This research can also allow the entire community to extend 
its reach by creating revolutionary technologies, new products, new 
markets and industries and high quality jobs. Thus, high-risk, high-
reward research is needed to maintain the U.S. position of leadership 
in science and technology and to ensure the Nation's future economic 
competitiveness. The ARISE report cites several examples of such 
transformative payoffs, including the transistor, quantum mechanics, 
and angiogenesis. The report recommends that every agency set aside a 
certain portion of its research budget for high-risk research.
    For most of its history, the NSF has received far more proposals 
that have been judged by the competitive peer-review system to merit 
funding than the agency has sufficient funds to award. It is up to the 
program officers to make the final judgments as to which proposals 
receive awards and the large majority that do not. Other research 
agencies are in a similar position. When funds are this tight, all 
components of the system--researchers writing the proposals, experts 
reviewing the proposals, and program officers making the final 
decisions--naturally tend to become more risk averse. They tend to give 
highest priority to projects likely produce incremental success in the 
near-term. Short-term, low-risk and measurable results dominate 
competitive review and program management systems and decisions. The 
ARISE Committee summed it up in these words:

         ``As the resulting constant hunt for dollars fosters 
        conservative thinking, it also impedes the pace of research. 
        The thought, `Don't put it in your grant proposal unless you 
        know it will work,' too often guides senior and junior faculty 
        alike as they compete in an intense national grant-writing 
        mill.''

    It is important to emphasize that the system continues to fund 
excellent research, that it does help prepare the next generation of 
scientists and engineers, and that virtually all proposed research 
projects are challenging and are judged to advance scientific and 
technical understanding. But, some potentially path-breaking research 
is not being funded because it just looks too risky.
    The American Academy and the ARISE Committee are encouraged by 
several promising recent developments designed to counter the 
prevailing incentive system.
    In 2007, Congress created the Department of Energy (DOE) ARPA-E 
program as part of the America COMPETES Act. ARPA-E is modeled after 
DARPA with the goal of enhancing the economic and energy security of 
the United States through research into transformative energy 
technologies. DOE is currently evaluating the first round of 
applications, and successful proposals will be funded by the American 
Recovery and Reinvestment Act (ARRA).
    Similarly, this year will see the first grants awarded under the 
NIH Transformative R01 program (TR01), a targeted high-risk, high-
reward initiative designed as a result of strategic planning to fund 
ground-breaking research opportunities. The proposed FY 2010 budget 
expands funding for this program to $70 million, double the 2009 
funding level.
    The economic stimulus program enacted by Congress will support 
promising high-risk research at other agencies as well. NSF Director 
Arden Bement has pledged to give increased priority to new principal 
investigators and high-risk, high-return research in allocating ARRA 
funds. Building on the momentum provided by stimulus funding, the 
proposed NSF FY 2010 budget sets aside $92 million specifically to 
foster transformative research.
    The Academy commends the Congress, the National Science Board and 
NSF for their early recognition of the need to nurture high-risk 
research and their recent actions to address this need. The Foundation 
has taken important first steps to expand opportunities for new and 
established researchers alike to pursue high-risk opportunities. For 
example, NSF program officers now have the flexibility to award up to 
two years of funding for potentially transformative research through 
the EAGER program (EArly-concept Grants for Exploratory Research). This 
mechanism should be used more frequently across the NSF grant programs 
and at other funding agencies as well. Clearly, each agency must stand 
behind the program officers making these difficult decisions, since 
many of the truly bold, high-risk ideas will not bear fruit. If the 
agencies' expectations are too high, the entire effort will fail.
    President Obama's Innovation Strategy aims to restore American 
leadership in fundamental research. In outlining this strategy in a 
September 21st speech in Troy, New York, the President stressed the 
importance of valuing and promoting ``the risk takers who have always 
been at the center of our success'' and pledged ``more support for 
high-risk, high-return research, for multi-disciplinary research, and 
for scientists and engineers at the beginning of their careers.''
    Looking to the future, the Obama Administration has emphasized the 
need to build on these commitments to encourage potentially 
transformative research. In an August 4 memorandum from Office of 
Science and Technology Policy Director John Holdren and Office of 
Management and Budget Director Peter Orszag, executive departments and 
agencies were asked to prioritize high-risk, high-reward research in 
preparing FY 2011 budget requests, stating ``Agencies should pursue 
transformational solutions to the Nation's practical challenges, and 
budget submissions should therefore explain how agencies will provide 
support for long-term, visionary thinkers proposing high-risk, high-
payoff research.'' The directive also asked agencies to create metrics 
to evaluate the success of programs designed to promote high-risk 
research.
    To these ends, the Academy respectfully asks this subcommittee and 
the Congress to encourage all of the science and engineering research 
agencies to:

        1.  Establish and strengthen policies, programs, and targeted 
        funding mechanisms designed to foster potentially 
        transformative research:

                  Applications should be relatively short and 
                focused on the qualifications of the researcher, an 
                explanation of the potentially transformative nature of 
                the research, and an explanation of why the researcher 
                believes the proposed approach could succeed.

                  The proposal and the review process should 
                place a premium on innovation.

                  Fast-track seed money to evaluate a novel 
                idea should be made available.

                  Agencies should be open to providing longer 
                funding periods for those proposals that require it.

                  A possible model for sustained funding is the 
                NSF Industry/University Cooperative Research Centers 
                program--an initial five-year grant that, if moving 
                forward appropriately, can be renewed for an additional 
                five-year period at a reduced level of funding.

           Because federal research agencies are highly diverse in 
        their missions, needs, and programs, funding mechanisms that 
        support potentially transformative research will and should 
        vary across departments and agencies. Such diversity is a 
        national asset and the foundation of the research enterprise. 
        Therefore a final recommendation is:

                  Convene interagency meetings to share 
                information on how departments and agencies design, 
                organize, implement, and evaluate their investments in 
                potentially transformative research.

        2.  Nurture high-risk, high-reward research programs that have 
        a critical mass.

        3.  Establish metrics with which to evaluate the success of 
        targeted research programs:

                  Short-term metrics: Are proposals of higher 
                quality compared to those submitted to standard grant 
                programs? Does the funding rate discourage future 
                applicants?

                  Long-term metrics: Wait ten years to evaluate 
                scientific outcomes--fruits of transformative research 
                are not apparent in the short-term.

        4.  Adopt funding mechanisms and policies that nurture 
        transformative research in all award programs, not just those 
        targeted at high-risk, high-reward research:

                  Charge reviewers to identify new ideas, 
                innovation, and creativity. Consider alternative ways 
                to select and mentor reviewers.

                  Give program administrators in all agencies 
                the flexibility to provide extra resources or time to 
                research unexpected but promising developments, 
                potentially using the NSF EAGER grants as a model.

                  Recognize in grant-reporting requirements the 
                value of fortuitous findings not related to the main 
                objective of the research proposal.

                  For grant renewals or new grants on the same 
                topic, restrict the number of submitted publications 
                and require a self-assessment of each cited 
                publication's impact.

        5.  Strengthen application and review processes. High-risk 
        research proposals face even greater challenges in a stressed 
        peer-review system not equipped to appreciate them:

                  Require recipients of multiple grants from an 
                agency to serve as reviewers.

                  Achieve greater continuity in reviewers.

                  Require applicants to address the following 
                question about their proposed research: ``If this 
                works, what long-term scientific difference will it 
                make?'' Evaluate proposals based on this criterion.

                  Establish interdisciplinary review panels to 
                consider high-risk research proposals across programs 
                and fields.

                  Evaluate renewals for first awards for high-
                risk, high-reward research on the basis of project 
                execution and potential scientific impact, not on 
                deliverables. Resist fine-grain assessments of whether 
                a project ``worked''; expect some hypotheses to fail.

        6.  Strengthen investments in the career development of agency 
        program officers who are indispensable to the vitality and 
        productivity of the entire research enterprise. They should be 
        encouraged and expected to engage with the professional 
        communities they fund. This requires an adequate administrative 
        budget, which should not come at the expense of the research 
        budget:

                  Program officers should be leaders not only 
                within their agencies but within their external 
                scientific communities as well.

                  Program officers should be able, indeed 
                encouraged, to attend professional meetings and to 
                visit institutions and laboratories funded by programs 
                for which they are responsible.

                  Many university faculty members serve as 
                temporary program officers at NSF, or ``rotators,'' 
                while on leave from their university. They provide 
                essential service and leadership for NSF's research 
                programs. Consideration should be given to providing 
                this flexibility to other agencies as well.

    As a former Director of the National Science Foundation, I wish to 
affirm and commend the dedication and the quality of its program 
officers. They are the core of the NSF. The encouragement and support 
they receive directly determines how well NSF performs its important 
work. They must be able to travel to professional meetings, make site 
visits to universities, and in other ways become more active and 
visible leaders in their fields. Just as the program officers need to 
stay current on the latest developments in science and engineering 
research, the research community needs to know and respect these 
professionals, who have such large responsibilities for the quality of 
U.S. science and engineering. I urge Congress, through its oversight 
and appropriations roles, to provide the resources the NSF requests for 
support of the agency's staff.
    The Committee will note that the ARISE report recommends both the 
creation of targeted grant programs specifically aimed at high-risk, 
high-reward research and the promotion of such research within all 
existing funding programs. There are several advantages to the creation 
of targeted grant programs, and a few attendant challenges. High-risk, 
high-reward research involves unique objectives, time-frames and 
evaluation metrics, and targeted programs permit these research 
proposals to be evaluated separately from standard proposals. It may 
also be faster and easier to implement a new targeted program than to 
re-tool standard funding processes to accommodate the particular needs 
of high-risk, high-reward proposals.
    Challenges associated with targeted funding programs include the 
potential for extremely low funding rates that could discourage future 
applicants. A further challenge is that funding agencies must be 
prepared to follow unexpected research directions arising from high-
risk, high-reward research. Finally, in evaluating the merits of high-
risk, high-reward research programs, it must be kept in mind that the 
fruits of transformative research are often not apparent for at least 
ten years. Near-term evaluation of these programs must be based on 
different metrics, for example, whether the quality of proposals 
differs from those received through standard grant programs.
    The ARISE Committee was also concerned with the role of other 
institutions, particularly universities, in supporting high-risk 
research. Institutions of higher education--especially medical 
schools--have tended to enlarge their faculty in times of expanding 
federal investment by shifting the salary burden to faculty. For the 
federal funding agencies, this salary support reduces the number of 
projects that can be funded. For the faculty member, this requirement 
fosters conservative, risk-averse thinking as the path to sustained 
funding. When funding tightens, faculty, especially early-career 
faculty, after years of training often simply leave the field.
    Two final ARISE recommendations directly address the role of 
universities in supporting early-career scientists and high-risk, high-
reward research. These recommendations aim to mitigate concerns over 
the effects that boom and bust funding cycles have on tenure, training, 
and capital investment on campuses:

        1.  Universities should accept greater institutional 
        responsibility for the salaries of faculty members.

        2.  In building new facilities and programs, universities 
        should shoulder a larger share of the financial cost.

    Thus, university resources are needed to buffer the scientific 
enterprise from the ups and downs of federal funding. If funding 
campaigns for construction were expected to assume some significant 
portion of the research expenses, it would lead universities to limit 
excessive building programs based on unrealistic expectations about the 
expansion of the research enterprise. Some universities are now 
beginning to recognize the wisdom of setting aside money from building 
campaigns for research and equipment. Universities could go even 
further and underwrite the creation and maintenance of centers 
specifically devoted to potentially transformative research. In times 
of economic downturn and shrinking endowments, the government and 
universities should consider ways to provide general support for 
science and engineering research that protect against disruptive boom 
and bust funding cycles.
    The Academy is initiating a second phase of ARISE to study how the 
distribution of federal funds affects the administration, faculty, 
students, and the academic mission of the university. The Academy would 
be grateful to this subcommittee for its input as we develop this phase 
of the ARISE study.
    I look forward to your questions about all aspects of the ARISE 
report. Thank you, once again, for this opportunity.

                       Biography for Neal F. Lane

    Dr. Neal Lane is the Malcolm Gillis University Professor at Rice 
University in Houston, Texas. He also holds appointments as Senior 
Fellow of the James A. Baker III Institute for Public Policy, where he 
is engaged in matters of science and technology policy, and in the 
Department of Physics and Astronomy.
    Prior to returning to Rice University, Dr. Lane served in the 
Federal Government during the Clinton Administration as Assistant to 
the President for Science and Technology and Director of the White 
House Office of Science and Technology Policy, from August 1998 to 
January 2001, and as Director of the National Science Foundation (NSF) 
and member (ex officio) of the National Science Board, from October 
1993 to August 1998.
    Before becoming the NSF Director, Dr. Lane was Provost and 
Professor of Physics at Rice University in Houston, Texas, a position 
he had held since 1986. He first came to Rice in 1966, when he joined 
the Department of Physics as an assistant professor. In 1972, he became 
Professor of Physics and Space Physics and Astronomy. He left Rice from 
mid-1984 to 1986 to serve as Chancellor of the University of Colorado 
at Colorado Springs. In addition, from 1979 to 1980, while on leave 
from Rice, he worked at the NSF as Director of the Division of Physics.
    Widely regarded as a distinguished scientist and educator, Dr. 
Lane's many writings and presentations include topics in theoretical 
atomic and molecular physics and science and technology policy. Early 
in his career he received the W. Alton Jones Graduate Fellowship and 
held an NSF Doctoral Fellowship (University of Oklahoma), an NSF Post-
Doctoral Fellowship (while in residence at Queen's University, Belfast, 
Northern Ireland) and an Alfred P. Sloan Foundation Fellowship (at Rice 
University and on research leave at Oxford University). He earned Phi 
Beta Kappa honors in 1960 and was inducted into Sigma Xi National 
Research Society in 1964, serving as its national president in 1993. He 
served as Visiting Fellow at the Joint Institute for Laboratory 
Astrophysics in 1965-66 and 1975-76. While a Professor at Rice, he was 
two-time recipient of the University's George R. Brown Prize for 
Superior Teaching.
    Through his work with scientific and professional organizations and 
his participation on review and advisory committees for federal and 
State agencies, Dr. Lane has contributed to public service throughout 
his career. He is a fellow of the American Physical Society, the 
American Academy of Arts and Sciences (member of its governing 
council), the American Association for Advancement of Science, the 
Association for Women in Science and a member of the American 
Association of Physics Teachers. He serves on several boards and 
advisory committees.
    Dr. Lane has received numerous prizes, awards, including the AAAS 
Philip Hauge Abelson Award, AAAS William D. Carey Award, American 
Society of Mechanical Engineers President's Award, American Chemical 
Society Public Service Award, American Astronomical Society/American 
Mathematical Society/American Physical Society Public Service Award, 
NASA Distinguished Service Award, Council of Science Societies 
Presidents Support of Science Award, Distinguished Alumni Award of the 
University of Oklahoma, and over a dozen honorary degrees. In 2009, Dr. 
Lane received the National Academy of Sciences Public Welfare Medal, 
the American Institute of Physics K.T. Compton Medal for Leadership in 
Physics, and the Association of Rice Alumni Gold Medal for service to 
Rice University.
    Born in Oklahoma City in 1938, Dr. Lane earned his B.S., M.S., and 
Ph.D. (1964) degrees in physics from the University of Oklahoma. His 
thesis advisor was Chun C. Lin (currently at the University of 
Wisconsin-Madison). He is married to Joni Sue (Williams) Lane and has 
two children, Christy Saydjari and John Lane, and four grandchildren, 
Allia and Alex Saydjari, and Matthew and Jessica Lane.

    Chairman Lipinski. Thank you, Dr. Lane. The Chair now 
recognizes Dr. Collins.

    STATEMENT OF DR. JAMES P. COLLINS, ASSISTANT DIRECTOR, 
     DIRECTORATE FOR BIOLOGICAL SCIENCES, NATIONAL SCIENCE 
                           FOUNDATION

    Dr. Collins. Congressman Lipinski, Congressman Ehlers and 
Committee Members, thank you for the opportunity to testify for 
the Research and Science Education Subcommittee.
    My name is James P. Collins. I have served as Assistant 
Director for Biological Sciences at the National Science 
Foundation. I am on leave from Arizona State University where I 
am Virginia M. Ullman Professor in the School of Life Sciences, 
and in this capacity I maintain a research lab. So I understand 
this intersection, this deep and important intersection between 
basic research and the university environment and the sort of 
policy issues that are at work here in terms of the Federal 
Government.
    Today we will discuss high-risk, high-reward research in 
the context of transformative science at the National Science 
Foundation. The U.S. National Science Foundation is first and 
foremost an innovation agency. NSF has a long history of 
success in supporting research with far-reaching impacts on the 
U.S. economy and the well-being of all Americans. Since 1950, 
this success has relied on. first, the close partnership with 
America's colleges and universities; second, on a merit-review 
system based in the scientific community; and third, on a 
continuously refreshed cadre of program officers who are 
stewards of the Nation's investment in basic scientific 
research and education.
    Unlike industry, whose typically shorter-term goals and 
proprietary results are aimed at the marketplace, NSF 
investments are both short- and long-term, and most 
importantly, its results are public. It is sometimes mistakenly 
assumed that research investments that are scientifically 
successful in the short run can produce similar short-term 
economic gains, and that this outcome is the only valid measure 
of success. In fact, the transformative impacts of the 
knowledge and technologies that result from successful 
scientific investments on subsequent scientific research, the 
economy, and society are often realized only many years later.
    Research in the history, philosophy and social studies of 
science teaches us that attempts to predict the individual 
ideas or projects that will be transformative are imprecise at 
best. The process of scientific discovery is a community 
endeavor. This endeavor takes time and is by design cumulative, 
skeptical and critical of new results. Transformative 
discoveries happen because of these qualities.
    NSF seeks to advance the transformative science by 
encouraging high-risk, high-reward research in the context of 
the structures, programs, and policies of an innovation agency. 
Transformative science is supported by institutions designed to 
foster such research, and NSF is just such an institution.
    NSF's merit-review process, which is based in the 
scientific community, is a form of what is now called ``crowd 
sourcing.'' It uses the collaborative wisdom of the crowd to 
identify the best research. NSF merit review, done by convening 
groups of experts, creates a special role for NSF in the 
evolution of the values in the American scientific community. 
As we discuss transformative research and investments in risky 
but potentially high-reward research with NSF panels, 
researchers incorporate these ideas in their evaluations and 
promote them in their own scientific venues.
    Establishing and sustaining interactions among NSF 
reviewers, program officers, applicants and awardees has shaped 
the culture of American science and is at the heart of the 
process of discovery in U.S. science. NSF program officers are 
stewards of the Nation's investment in research and science 
education. In addition to merit review, they manage awards, 
they mentor post-doctoral fellows and early career scientists, 
they facilitate national and international connections within 
and across fields and engage in outreach to promote broader 
participation in education for knowledge economy.
    But as the research enterprise accelerates and becomes 
interdisciplinary, the demands of proposal and award management 
are becoming overwhelming. Time for just thinking about a 
problem, interacting with researchers, and imagining creative 
new ways to find and fund the best research is decreasing. As 
NSF experiments with new methods of review and funding directed 
at enabling transformative science, program officers will 
experience even greater demands on their time and attention in 
order to manage these innovative processes.
    Like other Federal funding agencies, NSF seeks to describe 
itself in terms of its research awards. However, in searching 
for more meaningful assessments, NSF is exploring new methods 
and measures--to understand the transformative contributions of 
new scientific knowledge to economic and social outcomes, to 
inform future investments, and to convey this information to 
policy-makers and the public.
    For nearly 60 years, NSF has been forward-looking in its 
management of the Nation's scientific enterprise. Our challenge 
for the future is to sustain a culture of creativity and 
innovation that pervades NSF and guides our decisions. NSF must 
continue to innovate, even in the midst of excellence.
    Once again, Mr. Chairman, thank you for giving me the 
opportunity to testify on this important subject. I would be 
pleased to answer any questions that you have.
    [The prepared statement of Dr. Collins follows:]

                 Prepared Statement of James P. Collins

INVESTING IN HIGH-RISK, HIGH-REWARD RESEARCH

    Chairman Lipinski, Ranking Member Ehlers, and distinguished Members 
of the Subcommittee on Research and Science Education, thank you for 
inviting me to participate in this hearing on ``Investing in High-Risk, 
High-Reward Research.''
    Mr. Chairman, as you know, the U.S. National Science Foundation 
(NSF) is first and foremost an innovation agency that has a long 
history of success in supporting research with far-reaching impacts on 
the U.S. economy and the well-being of Americans. Since 1950 this 
success has relied on a close partnership with America's colleges and 
universities, which are the principal locus of the research NSF funds. 
NSF research grants are made for the short- or long-term and its 
results are public, unlike industry which usually has shorter-term 
goals aimed at the market place and proprietary results. An NSF 
hallmark is its continuing effort to advance transformative science by 
encouraging high-risk/high-reward research in the context of the 
structures, programs, and policies needed to function as innovation 
agency.
    Scientific discovery is a social process, a community endeavor that 
takes time, and is by design cumulative, skeptical, and critical of new 
results. Transformative discoveries happen because of these qualities 
(not in spite of them). Moving from an ``aha moment'' to value creation 
in a knowledge economy is a complex process involving interactions 
among people, social structures, and institutional practices and 
cultures. Research in history, philosophy, and social studies of 
science teaches us that attempts to predict which individual ideas or 
projects that are likely to be ``transformative'' are challenging and 
imprecise at best.
    The challenge for agencies like NSF that fund research done by 
other organizations is to create and sustain a culture of innovation in 
which the flow of information among its members creates an 
institutional culture and framework that stimulates, reinforces, and 
rewards creativity, and pervades the agency and guides its decision-
making process.

Creating and sustaining innovation

    NSF's decisions are based on the advice of its constituents though 
merit review, which is a form of what is now called ``crowd sourcing'' 
or a way to leverage group collaboration toward the goal of identifying 
the best research. Most merit review at NSF is done by convening groups 
of scientific experts, which creates a special institutional role for 
NSF in the evolution of values in the American scientific community: in 
this case valuing the importance of potentially transformative ideas 
and investment in potentially high-reward research that has risks.
    As we share and discuss transformative science with reviewers, 
panelists, and advisory committees, they incorporate that idea in their 
own evaluations and promote it in other scientific venues. Interactions 
among NSF reviewers, program officers, applicants for research and 
education funding and awardees have shaped and are shaping the culture 
of American science. Establishing and sustaining this three-way 
relationship is a signal contribution of NSF and at the heart of the 
process of discovery in U.S. science.
    The recent Netflix million-dollar prize competition is a compelling 
example of the successful use of crowd sourcing for technological 
discovery while also contributing to a culture of innovation. Netflix 
offered $1 million to anyone who could improve their algorithm for 
matching movies with customers. The incentive was hugely successful. Of 
the many creative submissions, two proposed the same promising and 
highly transformative approach. These two submissions were 20 minutes 
apart so that under the rules of the contest, the first submission won. 
However, as described in a recent New York Times (September 22, 2009) 
report by Steve Lohr:

         ``. . . the scientists and engineers on the second-place team, 
        and the employers who gave many of them the time and freedom to 
        compete in the contest, were hardly despairing.

         Arnab Gupta, Chief Executive of Opera Solutions . . . took a 
        small group of his leading researchers off other work for two 
        years. `We've already had a $10 million dollar payoff from what 
        we've learned,' Mr. Gupta said. `So for us, the $1 million 
        dollar prize was secondary, almost trivial.' ''

    By any measure, the outcomes of NSF's investments in frontier 
research in science, engineering, and science education are impressive. 
NSF's tradition of merit review that enables new ideas to be tested and 
funded has served the Nation well. The hallmarks of NSF merit review 
are:

          Review criteria that identify those ideas that will 
        make a difference both in terms of intellectual merit and 
        broader impacts;

          A selection process that combines evaluation by 
        independent expert merit reviewers with the professional 
        scientific experience and judgment of NSF program officers;

          Management of the merit review process by a 
        combination of permanent program officers, who provide 
        institutional memory and experience, and visiting scientist 
        program officers who contribute recent research expertise.

    In the May 12, 2008 issue of The New Yorker, James Surowiecki (The 
open secret of success; http://www.newyorker.com/talk/financial/2008/
05/12/080512ta- talk-surowiecki) writes about 
innovation at the Toyota car corporation, which has two elements. 
First, Toyota turns principles, such as eliminate waste, have parts 
arrive when needed, fix problems as soon as they arise, into practice 
better than its competitors. And second, Toyota defines ``innovation as 
an incremental process, in which the goal is not to make huge, sudden 
leaps but, rather, to make things better on a daily basis . . .. 
Instead of trying to throw long touchdown passes, as it were, Toyota 
moves down the field by means of short and steady gains.'' This leads 
Surowiecki to conclude: ``And so it [this process] rejects the idea 
that innovation is the province of an elect few; instead it's taken to 
be an everyday task for which everyone is responsible.'' Said 
differently, innovation succeeds in practice when it is 
``institutionalized,'' when it is central to the institution's culture, 
and when the institution itself is structured to create and sustain 
innovative thinking.
    Multiple lines of evidence support the conclusion that discovering 
the very best science to fund is a social process. The results are 
context dependent, which means that is crucial to create and sustain an 
institutional culture that is open to transformative ideas since hoped 
for discoveries are often resisted because ideas are premature. 
Discoveries are prized because they are often challenged and tough to 
achieve. Path breaking is hard work, and the decision to follow someone 
down a new road is not always the obvious thing to do. Making that 
decision requires experience and often wisdom.
    NSF's Program Officers are at the center of this decision-making 
process; they are the keystone of the agency's culture of innovation.

The NSF Program Officer's role in fostering transformative research

    NSF relies on the expertise and experience of its permanent and 
visiting scientist program officers for funding recommendations. After 
reading proposals, listening to visiting panel reviewers and gleaning 
advice from external referees it is the program officer who recommends 
action on a proposal. It is her or his responsibility to integrate all 
of the information and make a final recommendation based on an 
understanding of all of the sources. For this reason program officers 
play a central role in identifying potentially transformative research.
    Stewardship and scholarship responsibilities of program officers go 
beyond merit review responsibilities. These science administrators look 
for the extraordinary in the proposals they review to create an award 
portfolio of emerging ideas and outcomes. Beyond the ideas in 
proposals, new areas for support emerge from a broad and constant set 
of interactions with the scientific community. As stewards of the 
Nation's investment in research and science education they determine 
enabling levels and durations of funding, mentor postdoctoral fellows 
and early career scientists, facilitate national and international 
connections within and across fields, and engage in outreach to promote 
broader participation and the education of a new generation of 
scientists as well as the general workforce.
    A culture of creativity at NSF is encouraged by regular exercises 
in which program officers identify and present exciting and emerging 
areas for future investment within and across directorates. ``Blue Sky 
projects'' not limited by disciplinary boundaries are encouraged. Such 
exercises help program officers to incorporate risky, transformative, 
and/or interdisciplinary research and education projects as essential 
parts of their award portfolios.
    As NSF experiments with and develops new methods of review and 
funding directed at enabling transformative science, program officers 
will experience even greater demands on their time and attention in 
order to manage these innovative processes and their anticipated 
additional workload. The Subcommittee asked for an assessment of ``the 
impact of flat agency operations budgets on Program Officers' ability 
to identify and support potentially transformative research 
proposals.'' As the research enterprise accelerates and becomes more 
interdisciplinary, the day to day obligations of proposal and award 
process management are significantly increasing. Time needed for ``just 
thinking'' about a problem, interacting with researchers, and imagining 
creative new ways to find the best research to fund is decreasing. 
Fostering program officer creativity requires investment of time and 
money. Sufficient personnel and infrastructure support, as requested in 
the President's 2010 Budget, is needed to ensure that NSF remains a 
21st century innovation agency.

Supporting institutional creativity through practices and policies

    Identifying proposals during the review process that will produce 
transformative results before the research is conducted and before the 
scientific community can assimilate the findings is challenging and 
imprecise at best. However, the Foundation can and does identify 
proposals that contain potentially transformative research ideas or 
concepts, and as discussed already is shaping the institution in ways 
that facilitate the identification of transformative research. 
Specifically, NSF has:

          Modified the intellectual merit review criterion to 
        include potentially transformative concepts;

          Established an operational definition of 
        transformative research;

          Provided training to new program officers on the 
        importance of supporting potentially transformative as part of 
        a balanced awards portfolio.

    Modifying the Intellectual Merit Review criterion. As a result of 
discussions with the National Science Board and within NSF, a simple 
but important addition to the NSF Intellectual Merit review criterion 
was adopted to emphasize to the scientific community and to NSF staff 
members the importance of potentially transformative research. On 
September 24, 2007, NSF's Director issued Important Notice No. 130 on 
transformative research; important notices are sent to presidents of 
universities and colleges and heads of other NSF awardee organizations. 
The notice stated that effective October 1, 2007, the NSF Grant 
Proposal Guide, as well as new funding opportunities issued after that 
date, would incorporate the following revised Intellectual Merit 
Criterion--the new wording is underlined:

         What is the intellectual merit of the proposed activity?

         How important is the proposed activity to advancing knowledge 
        and understanding within its own field or across different 
        fields? How well qualified is the proposer (individual or team) 
        to conduct the project? (If appropriate, the reviewer will 
        comment on the quality of prior work.) To what extent does the 
        proposed activity suggest and explore creative, original, or 
        potentially transformative concepts? How well conceived and 
        organized is the proposed activity? Is there sufficient access 
        to resources?

    All proposals received after January 5, 2008, have been reviewed 
using this revised criterion. Program officers instruct reviewers to 
pay special attention to those proposals that may include potentially 
transformative research.
    Defining potentially transformative research. The National Science 
Board (NSB) defined transformative research as ``research driven by 
ideas that have the potential to radically change our understanding of 
an important existing scientific or engineering concept or leading to 
the creation of a new paradigm or field of science or engineering. Such 
research also is characterized by its challenge to current 
understanding or its pathway to new frontiers.'' To make the NSB 
definition operational within the context of NSF's funding programs, 
the NSF uses the following definition, which builds on the NSB 
definition with explanatory text and examples:

         Transformative research involves ideas, discoveries, or tools 
        that radically change our understanding of an important 
        existing scientific or engineering concept or educational 
        practice or leads to the creation of a new paradigm or field of 
        science, engineering, or education. Such research challenges 
        current understanding or provides pathways to new frontiers.

         Transformative research results often do not fit within 
        established models or theories and may initially be unexpected 
        or difficult to interpret; their transformative nature and 
        utility might not be recognized until years later. 
        Characteristics of transformative research are that it:

                a.  Challenges conventional wisdom,

                b.  Leads to unexpected insights that enable new 
                techniques or methodologies, or

                c.  Redefines the boundaries of science, engineering, 
                or education.

    NSF Senior Managers, such as Division Directors, discuss concerns 
about the conservative aspects of peer review with every panel in order 
to raise consciousness about the importance of risk-taking and 
creativity in research. Panels are asked to flag high-risk/high-reward/
transformative research.
    Training new program officers. To ensure that program officers 
understand NSF's commitment to supporting high-risk/high-reward/
transformative research, the Foundation developed a training 
presentation for new program officers. Senior NSF staff members are 
advisors at each training session. New program officers receive the 
May, 2007, NSB Report, ``Enhancing Support of Transformative Research 
at the National Science Foundation;'' the Foundation's guiding 
principles in support of transformative research; the Foundation's 
working definition of transformative research, including examples; and 
a set of Frequently Asked Questions (and answers) related to 
potentially transformative research. Finally, NSF's Annual Report to 
Employees in 2007 provided guidance to all NSF staff members about the 
critical importance of identifying and supporting potentially 
transformative research.
    While we cannot predict which research investments will invariably 
produce transformative results, we can create institutional structures 
and cultures, such as those discussed already, that provide a context 
for recognizing and supporting projects that have the greatest chance 
of leading to fundamentally new discoveries. Collectively, these 
institutional mechanisms constitute the process of discovery for 
potentially transformative research. But if NSF is to be America's 
premier ``innovation agency,'' the institution itself must always be 
looking for novel mechanisms to discover the best research to fund. 
Here are some ways NSF is exploring this exciting frontier.

New approaches for identifying potentially transformative research

    NSF is experimenting with novel mechanisms for developing, 
reviewing, and funding exploratory and especially creative research. 
All are new ways to foster NSF's process of discovery.
    In January, 2009, NSF announced a new foundation-wide funding 
mechanism modeled on the Small Grants for Exploratory Research Program. 
EAGER (Early-concept Grants for Exploratory Research) awards support 
the initial stages of untested, but potentially transformative research 
ideas or approaches. The work may be considered especially ``high-
risk--high-payoff'' in the sense that it involves radically different 
approaches, applies new expertise, or engages novel disciplinary or 
interdisciplinary perspectives.
    Appendix I has a summary of seven targeted NSF programs that 
support potentially transformative research.
    At the Subcommittee's request, activities in the Directorate for 
Biological Sciences will be reviewed to illustrate how NSF has many 
features of an innovation agency, and is actively developing 
structures, programs, and policies needed to function as such an 
institution.
    Biology research and education today increasingly differ from how 
they were done 10, even five years ago. Frontiers are often at 
disciplinary ``edges:'' the intersection of biology and computer and 
information sciences, engineering, geosciences, mathematics, physical 
sciences, and social sciences. To the extent that it ever did, biology 
no longer stops at disciplinary margins, but is reflected in 
interdisciplinary areas such as bioengineering, biogeochemistry, 
biomathematics, chemical biology, and evolutionary psychology. The 
Directorate for Biological Sciences is responding to this reality 
through:

          Joint CAREER panels involving the Directorate for 
        Biological Sciences and the Directorate for Math and Physical 
        Sciences, which have for six years successfully reviewed 
        proposals from young investigators that integrate innovative 
        research and education at the interface of biology and physics.

          A shared program officer between the Directorate for 
        Biological Sciences and the Directorate for Math and Physical 
        Sciences who is charged with identifying and reviewing 
        proposals in the emerging interdisciplinary area of chemical 
        biology. The success of this activity led us to expand this 
        model with the Geosciences Directorate.

          An Integrated Global Systems Science activity will 
        bring together program officers and professional science 
        support staff members from the Directorate for Biological 
        Sciences and the Directorate for Geosciences in an effort to 
        identify and support the best interdisciplinary research needed 
        to address the global challenges we face as a planet.

          The recently released report ``Transitions and 
        Tipping Points in Complex Environmental Systems'' from NSF's 
        Advisory Committee for Environmental Research and Education 
        warns that ``The global footprint of humans is such that we are 
        stressing natural and social systems beyond their capacities. 
        We must address these complex environmental challenges, and 
        mitigate global-scale environmental change--or accept likely 
        all-pervasive disruptions.'' This challenge requires both 
        interdisciplinary research at the interface of natural and 
        human systems and improved environmental literacy that will 
        enable policy-makers both in the U.S. and around the globe to 
        make the informed decisions that will enable us to live 
        sustainably on Earth. A three-year-old Memorandum of 
        Understanding among the Directorates for Biological Sciences, 
        Geosciences, and Social, Behavioral and Economic Sciences to 
        establish Coupled Natural and Human Systems (CNH) as an ongoing 
        cross-directorate program is a successful example of cross-
        directorate thinking put into action.

          The Directorate for Biological Sciences is exploring 
        the idea of ``Fellowships at the Interface,'' which will 
        provide training and experience at the interface of biology and 
        other scientific disciplines and education. Consideration also 
        is being given to expanding this program (with an additional 
        investment) to include experience for mid-career scientists at 
        the interface of biology and education.

    About 18 months ago Malcolm Gladwell argued in an article in The 
New Yorker that ideas are easy to come by; implementing them is hard. 
Ideas, Gladwell argued, are not precious, but everywhere. He concluded, 
therefore, ``maybe the extraordinary process that we thought necessary 
for invention--genius, obsession, serendipity, epiphany--wasn't 
necessary at all.'' The trick, he felt, was getting together a group of 
thoughtful, creative people all thinking about how to solve a problem: 
(``In the Air;'' http://www.newyorker.com/reporting/2008/05/12/
080512fa-fact-gladwell/?yrail).
    The Directorate for Biological Sciences is using three methods to 
take advantage of this line of reasoning.

          The ``Sandpit'' is an experiment in real time, 
        interactive peer review to explore novel solutions to existing 
        problems or indentify new areas of research. The Directorate 
        for Biological Sciences, with participation and support from 
        the Directorates for Math and Physical Sciences, Engineering, 
        Social, Behavioral and Economic Sciences, and Computer and 
        Information Sciences and Engineering, sponsored its first 
        sandpit in the area of synthetic biology in conjunction with 
        the United Kingdom's Engineering and Physical Sciences Research 
        Council (EPSRC) in April, 2009. This sandpit produced five 
        interdisciplinary, multi-investigator projects with support 
        from NSF and EPSRC.

          The Directorates for Biological Sciences, 
        Engineering, and Social, Behavioral and Economic Sciences also 
        funded an EAGER proposal that focuses on developing a 
        ``prediction market'' for synthetic biology. A prediction 
        market is a social networking method used to predict the most 
        likely outcome of an event like a presidential election or next 
        quarter's sales for a business. The principal investigator for 
        this award will use the method to assess where the most 
        creative research investments can be made to advance the area 
        of synthetic biology.

          Synthesis Centers promote the process of collecting 
        and connecting disparate data, concepts, or theories to 
        generate new knowledge or understanding. Beyond its necessity 
        for innovation in basic science, synthesis increasingly 
        contributes to novel and effective solutions for pressing 
        problems, and to the emergence of new ideas or fields of 
        inquiry that would not otherwise exist. Biology Directorate-
        funded synthesis Centers in conjunction with other NSF 
        Directorates and federal agencies emphasize interdisciplinary 
        research and education in critical areas of the biological, 
        computer, and social sciences. Current centers include: the 
        National Center for Ecological Analysis and Synthesis, the 
        National Evolutionary Synthesis Center, the National Institute 
        for Mathematical and Biological Sciences, and the iPlant 
        Collaborative. These centers advance our understanding by 
        interdisciplinary activities as well as by ``getting together a 
        group of thoughtful, creative people all thinking about how to 
        solve a problem.''

    Modern cyberinfrastructure can greatly facilitate these ways of 
identifying the likely places for a commitment to supporting high-risk/
high-reward/transformative research. The social networking manifest in 
models like crowd sourcing or prediction markets is based on arguments 
that there is great value in a collective effort focused on uncovering 
the best sort of research to fund--the so-called ``wisdom of the 
crowd'' argument. However, as noted already, NSF's merit review system 
is at its root a wisdom-of-the-crowd model. The new extensions of this 
fundamental model rely on modern computer and information sciences to 
integrate tens, hundreds, or even, as in the case of the Netflix Prize 
also discussed earlier, thousands of researchers focused on solving a 
common problem. These sorts of social networking models are 
potentially, in an analogy with Clayton Christian's The Innovator's 
Dilemma, a ``disruptive technology'' when it comes to discovery related 
to research and education.
    But every presumptive innovation carries with it an implicit 
challenge: How would one know that a novel idea, invention, or method 
really made a difference? How can we assess any effort at creativity?

The assessment challenge

    NSF tends to describe itself in terms of its awards, just as other 
federal basic research funding agencies. One form of assessment, then, 
is a review of the narrative summarizing the kinds of research the 
agency funds.
    NSF tracks research outcomes in the form of highlights, which are 
short descriptions of research and educational outcomes composed by 
program officers using material provided by principal investigators. 
Just as for research proposals, merit review can be applied to 
institutions, and NSF also uses this method. NSF relies on the judgment 
of external experts to maintain high standards of program management, 
to provide advice for continuous improvement of NSF performance, and to 
ensure openness to the research and education community served by the 
Foundation.
    Every NSF program is evaluated by a Committee of Visitors (CoV) 
every three years. Each CoV submits a detailed report to the 
appropriate NSF Advisory Committee, which itself is composed of members 
drawn from the communities NSF supports. All CoV reports are available 
(http://www.nsf.gov/od/oia/activities/cov/covs.jsp). CoV reviews 
provide NSF with assessments of the quality and integrity of program 
operations and program-level technical and managerial matters 
pertaining to proposal decisions. Each CoV comments on how the results 
generated by awardees contribute to NSF's mission and strategic outcome 
goals, including an assessment of the division/program's investments in 
high-risk/high-reward/transformative research projects.
    The Advisory Committee for GPRA (Government Performance and Results 
Act) Performance Assessment (AC/GPA) is charged with determining 
whether NSF has demonstrated ``significant achievement'' under its 
strategic outcome goals. This Foundation-wide Advisory Committee has 22 
members from outside of NSF drawn from academia, industry, and 
government. AC/GPA reports to NSF's Director. In its annual evaluation, 
the committee focuses on program highlights, reports from CoVs, and 
issues such as transformative research, broadening participation, and 
societal benefit. The most recent report notes:

         It is the unanimous judgment of the 2008 Advisory Committee 
        for GPRA Performance Assessment (AC/GPA) that the National 
        Science Foundation successfully met its performance objectives 
        by demonstrating significant achievement for each of the 
        following three long-term, qualitative, strategic outcome goals 
        in its 2006-2011 Strategic Plan:

                  DISCOVERY: Fostering research that will 
                advance the frontiers of knowledge, emphasizing areas 
                of greatest opportunity and potential benefit and 
                establishing the Nation as a global leader in 
                fundamental and transformation science and engineering.

                  LEARNING: Cultivating a world-class, broadly 
                inclusive science and engineering workforce, and expand 
                the scientific literacy of all citizens.

                  RESEARCH INFRASTRUCTURE: Building the 
                Nation's research capability through critical 
                investments in advanced instrumentation, facilities, 
                cyberinfrastructure and experimental tools.

    However, the AC/GPA also took issue with the practice of evaluating 
NSF's performance using only highlights because they were limiting in 
several ways:

          Highlights are annually scoped and cannot address 
        long-term outcomes or societal impacts.

          Highlights are written about individual awards or 
        projects, not fields or communities. The relevance of an 
        individual project or result cannot be understood in isolation.

          Highlights do not capture ``people outcomes,'' which 
        are central to NSF's vision.

          Highlights are anecdotal, both in subject matter and 
        in the non-systematic nature of their collection.

    At any given time, these assessment mechanisms provide a 
contemporary case history of how research results from NSF awards 
relate to the agency's mission and strategic goals. However, the 
longer-term ``transformative'' impacts of the knowledge and 
technologies that result from these successful scientific investments--
on subsequent scientific research, the economy, and society--are often 
realized years later. For funding agencies like NSF, identifying 
proposals during the review process that will produce transformative 
results before the research is conducted and before the scientific 
community can assimilate the findings is challenging at best.
    Mistakenly, it is sometimes assumed that research discoveries can 
be quickly brought to market and this rate can serve as an assessment 
metric. But it is intrinsic to the research enterprise that investments 
that are scientifically successful in the short-term cannot guarantee 
similar short-term economic gains. Dr. Julia Lane of NSF noted recently 
that:

         ``. . . [A] focus on economic value alone may also understate 
        the true returns of investments in science. Indeed, one strand 
        of research is attempting to develop a public value mapping of 
        outcomes: outcomes that are public, nonsubstitutable, and 
        oriented to future generations and that capture dimensions such 
        as competitiveness, equity, safety, security, infrastructure 
        and environment.'' [Assessing the Impact of Science Funding. 
        2009. Science 324, 1273-1275]

    The 2008 AC/GPA recommended that NSF ``consider ways to convey the 
long view of NSF investments in science and engineering'' and ``track 
future outcomes from people trained and supported by the Foundation.'' 
However, the absence of computer information systems designed to manage 
information rather than to simply process reviews, awards, or reports 
is a serious impediment to understanding how NSF awards connect to 
leading edge science and long-term outcomes. What is needed is a 
program information management system that connects the agency's award 
portfolios with one another, with other federal research agencies, with 
the scientific community, and with the public. Such a system would 
enable a reciprocal interaction (another form of crowd sourcing) among 
all of these elements.
    The NSTC's Science of Science Policy Interagency Group has 
identified this lack as a major issue in its recent Roadmap (http://
www.ostp.gov/galleries/NSTC%20Reports/
39924-PDF%20Proof.pdf). In particular, there are currently 
no data infrastructure that identifies the universe of individuals 
funded by federal science agencies (PIs, co-PIs, graduate and 
undergraduate students, lab technicians, science administrators, etc.) 
and that systematically couples science funding with the outcomes 
generated by those individuals. In searching for prototypes for the 
development of more meaningful assessment methods, NSF has begun to 
look within--to the Directorate for Social, Behavioral and Economic 
Sciences (SBE), where Program Officers and a research community think 
about these things--for the methods and measures needed to understand 
the transformative contributions of new scientific knowledge to 
economic and social outcomes, to inform future investments, and to 
convey this information in a manner that is understandable to policy-
makers and the public. SBE has programs such as Science of Science and 
Innovation Policy (SciSIP) and Science, Technology and Society (STS) 
that are funding work on next generation science assessment. Also, 
SBE's Science Resource Statistics Division, the Nation's resource for 
science statistics, is dedicated to continual improvement through 
ongoing workshops and consultations.

The role of NSF, universities, and the private sector in supporting 
                    potentially transformative research

    As noted earlier, NSF has a long history of success in supporting 
research with far-reaching impacts on the U.S. economy and the well-
being of Americans. Since 1950 this success has relied on a close 
integration with America's colleges and universities, which are the 
principal locus of the research NSF funds; unlike other federal 
agencies, NSF has no intramural labs or research staff. Significantly, 
NSF research grants are made for the short- or long-term, and results 
are not classified, but readily published in the open literature. In 
contrast, industry usually has shorter-term goals aimed at the market 
place, and results are often proprietary and therefore not readily 
shared.
    In the October, 2008, issue of Computerworld, Gary Anthes wrote: 
``By most measures, the U.S. is in a decade-long decline in global 
technological competitiveness. The reasons are many and complex, but 
central among them is the country's retreat from long-term basic 
research in science and technology, coupled with a surge in R&D by 
countries such as China.'' He went on to note that ``the kind of pure 
research that led to the invention of the transistor and the Internet 
has steadily declined as companies bow to the pressure for quarterly 
and annual results.'' He emphasized how many companies now support 
development, as opposed to the kind of basic research done at colleges 
and universities with NSF support. And there is also an increasing 
trend on industry's part to take even the basic research that it does 
offshore. Thomas Friedman recently noted to his dismay that ``America's 
premier solar equipment maker, Applied Materials, is about to open the 
world's largest privately funded research facility--in Xian, China.'' 
(The New Sputnik. New York Times, Sunday, September 27, 2009: p. wk 
12.)
    If federal agencies such as NSF were to adopt shorter-term 
perspectives exclusively as a way to meet new national needs, we risk 
an eventual intellectual and technological vacuum. Anthes feels this is 
already happening: ``The refocus from long-term research to shorter-
term development in industry--and Bell Labs is by no means the only 
example--has been mirrored by a similar trend among the Washington 
agencies that fund science and technology, such as the Departments of 
Defense and Energy, the National Institutes of Health and the National 
Science Foundation. Federal funding for R&D has not declined overall--
it has, in fact, increased. But since the early 1990s, funding has been 
more and more focused on the short-term needs of government.'' He 
reports no evidence in support of this claim, but the point deserves 
reflection.
    The U.S. must continue to support transformative research with 
potential long-term benefits. In a science and technology-based world 
that will underlie knowledge-based economies to divert our focus from 
the frontier is to disadvantage us in many ways. Sometimes it just 
takes unfettered time to make discoveries at the leading edges of 
knowledge: it is just this freedom that is the essential quality of the 
R&D that NSF as an innovation agency supports in partnership with 
America's institutions of higher learning. The NSF activities in 
Appendix II are examples of the productive intersection between basic 
research supported by a federal agency and the private sector and 
universities.
    For nearly 60 years NSF has been forward looking in terms of how 
the agency manages the scientific enterprise. Merit review fosters the 
``process of discovery,'' that is the means by which researchers can 
identify and answer leading/transformative/grand challenge questions. 
At the heart of the task of being a manager or administrator of the 
scientific enterprise there should be an abiding interest in the best 
ways to identify leading/transformative/grand challenge research 
opportunities. As new modes of science management emerge, especially 
those facilitated by modern information management systems, science 
administrators at the frontier will increasingly experiment with these 
new methods.
    Mr. Chairman, as I noted at the start of my testimony, NSF has many 
features of an innovation agency, and these features will continue to 
evolve in ways that will ensure NSF's place as first and foremost an 
innovation agency dedicated to funding the world's best research and 
education.
    I appreciate the opportunity to appear before the Subcommittee to 
speak to you on this important topic. I would be pleased to answer any 
questions that you may have.

Appendix I

                   NSF-targeted activities supporting

                  potentially transformative research

    In the Directorate for Engineering (ENG), the Office of Emerging 
Frontiers of Research and Innovation (EFRI) was conceived specifically 
to support high-risk, high-reward research. Beginning with its first 
awards in 2007, EFRI has funded investigations in areas where new 
concepts, new collaborations, and new approaches are essential to 
address grand engineering challenges or national needs. For example, 
EFRI researchers are investigating the topic of autonomously 
reconfigurable systems, which can respond to even unanticipated changes 
of circumstance. Teams are conducting unprecedented research to forge a 
theoretical framework for embedding autonomous reconfigurability into 
any type of complex system, including air traffic, wireless 
communication networks, and urban transportation networks. One team is 
creating a group of robots that can sense variables in their 
surroundings and self-assemble into a structure best suited for that 
particular environment. Engineering this new capability into human-made 
systems could transform infrastructure reliability and disaster 
response.
    Since its inception, the Engineering Research Center (ERC) program 
has supported high-risk, transformative research and the development of 
the Nation's leaders in innovation. The 2009 solicitation focuses 
explicitly on new mechanisms to link discovery to technological 
innovation in order to concurrently advance technologies and produce 
engineers who can lead U.S. innovation in a globally competitive 
economy. Two examples of transformative results from ERC-supported 
research include the portable defibrillator and early warning systems 
for tornadoes and other low-ground storm systems.
    In the Directorate for Computer & Information Science & Engineering 
(CISE), the focus in 2010 for transformative research will include the 
Expeditions in Computing Program. Expeditions are large multi-
disciplinary awards targeted to compelling, transformative research 
agendas that promise disruptive innovations in computing and 
information science and engineering. Funded at levels of up to $10M, 
Expeditions represent some of the largest single investments currently 
made by CISE.
    The NSF-wide Cyber-enabled Discovery and Innovation (CDI) program 
is another example of NSF's support for potentially transformative 
research. CDI recognizes that ``computational thinking'' (i.e., 
computational methods, concepts, models, algorithms and tools) will 
transform how all science and engineering will be conducted in the 21st 
Century. Computational abstractions, as much as high-speed computers 
and high-bandwidth networks will enable scientists and engineers to 
make new discoveries by changing the very questions they can ask. Above 
and beyond the usual NSF requirements, CDI uniquely requires that 
research projects advance two or more disciplines as well as 
innovations in or innovative uses of computational thinking.
    The NSF Office of Cyberinfrastructure (OCI) will focus investments 
on the Strategic Technologies for Cyberinfrastructure (STCI) Program 
whose primary purpose is to support work leading to the development 
and/or demonstration of innovative cyberinfrastructure services for 
science and engineering research and education that fill gaps left by 
more targeted funding opportunities. In addition, STCI considers highly 
innovative cyberinfrastructure education, outreach and training 
proposals that lie outside the scope of targeted solicitations.
    The Directorate for Social, Behavioral and Economic Sciences (SBE) 
is working to catalyze transformative science in three major ways. 
First, its largest funding opportunities are for multi-disciplinary 
research projects, thus encouraging the transformations that are 
possible when disciplinary silos are shattered. Second, SBE has alerted 
its scientists that it is interested in funding complexity science 
projects. Complexity science lies at the edge of normal science and is 
especially promising terrain for transformative insights. Third, SBE is 
working with its communities to identify and create major 
infrastructure--particularly new databases and new tools for 
assembling, analyzing and managing data--that will enable next 
generation analyses of social, behavioral and economic phenomena. SBE 
has chosen to do all this by integrating these transformative 
mechanisms into its regular standing scientific programs rather than by 
creating separate activities. This is because they want to ensure that 
the appreciation and norms for reviewing and supporting potentially 
transformative science are visible to and integrated into the entire 
community, rather than separated from normal scientific review and 
discussion.
    The NSF Plant Genome Research Program (PGRP) within the Directorate 
for Biological Sciences (BIO) began in February 1998 as part of the 
National Plant Genome Initiative (NPGI), which is managed across 
federal agencies by an Interagency Working Group on Plant Genomes. The 
long-term goal of the NPGI is to develop and apply basic plant genome 
knowledge to a comprehensive understanding of economically important 
plants and plant processes. Connecting basic research to plant 
performance in the field accelerates basic discovery and innovation, 
which enables improved management of agriculture, natural resources, 
and the environment. To date the PGRP has contributed to the genome 
sequences and tools for studying both model and crop plants, including 
Arabidopsis, maize (corn), soybean, potato, tomato and Medicago. 
Training and outreach is built into all PGRP projects. PGRP-supported 
tools such as Targeted Induced Local lesions IN Genomes (TILLING) are 
now used in research and commercial settings for a wide range of plants 
and animals. TILLING technology has led to a spin-off company that is 
now part of Arcadia Biosciences. Since agricultural challenges do not 
stop at national borders, the PGRP, in coordination with USDA and 
USAID, expanded its efforts in 2004 to include Developing Country 
Collaborations for Plant Genome Research. In 2009, the NSF in 
partnership with the Bill & Melinda Gates Foundation (BMGF) established 
a new program called Basic Research to Enable Agricultural Development 
(BREAD). With equal support from NSF and BMGF (a total of $48 million 
over five years), BREAD will fund basic research to develop innovative 
solutions to the agricultural problems faced by small farmers in 
developing countries. This exciting new partnership will enable NSF to 
leverage basic research advances made through the NPGI with BMGF 
funding for implementation to international partners. The Plant Genome 
Research Program has developed tools and resources that not only have 
transformed our understanding of plant structure and function, but that 
now are enabling us to tackle pressing needs for new plant-based 
materials, new energy sources, and plants that adapt to environmental 
stresses resulting from a changing climate.

Appendix II

                   Examples of NSF activities at the

                 intersection of federally funded basic

                    research and the private sector

                            and universities

    NSF-funded Centers are designed from the outset with built-in 
flexibility so that investigators can pursue innovative ideas within 
the context of a defined program of research. Examples are legion, and 
include the Mosaic web browser developed at NSF's National Center for 
Supercomputing Applications at the University of Illinois. NSF's 
creation of two Centers for the Environmental Implications of 
Nanotechnology (CEIN) in 2008 exemplify innovative networks that are 
connected to other research organizations, industry, and government 
agencies to strengthen our nation's commitment to understanding the 
potential environmental hazards of nanomaterials and to provide basic 
information leading to the safe environmentally responsible design of 
future nanomaterials.
    The Industry/University Cooperative Research Centers (I/UCRC) 
program develops long-term partnerships among industry, academe, and 
government. Each I/UCRC contributes to the Nation's research 
infrastructure, enhances the intellectual capacity of the STEM 
workforce by integrating research with education, and encourages and 
fosters international cooperation and collaborative projects. For 
example, the NSF Industry/University Collaborative Research Center (I/
UCRC) known as the Berkeley Sensor and Actuator Center conducts 
industry-relevant, interdisciplinary research on micro- and nano-scale 
sensors, moving mechanical elements, microfluidics, materials, and 
processes that take advantage of progress made in integrated-circuit, 
bio, and polymer technologies. This I/UCRC has developed and 
demonstrated a hand-held device that allows verified diagnostic assays 
for several infectious diseases currently presenting significant 
threats to public health, including dengue, malaria, and HIV. The 
device uses a dramatically simplified testing protocol that makes it 
suitable for use by moderately-trained personnel in a point-of-care or 
home setting. The center has also created many spin-off ventures 
including companies in the areas of wireless sensor networks for 
intelligent buildings; MEMS mirror arrays for adaptive optics; and 
optical flow sensors for industrial, commercial, and medical 
applications.
    The objective of the NSF Small Business Innovation Research (SBIR) 
program is to increase the incentive and opportunity for small firms to 
undertake cutting-edge research that would have a high potential 
economic payoff if successful. For example, in 1985, Andrew Viterbi and 
six colleagues formed ``QUALity COMMunications.'' In 1987-1988 NSF SBIR 
provided $265,000 (Phase I 8660104 and Phase II 8801254) for single 
chip implementation of the Viterbi decoder algorithm. Qualcomm 
introduced CDMA (code division multiple access) which replaced TDMA 
(time division multiple access) as a cellular communications standard 
in 1989. This advance led to high-speed data transmission via wireless 
and satellite. Now the $78B company holds more than 10,100 U.S. 
patents, licensed to more than 165 companies. Another example--Machine 
Intelligence Corp. was supported by SBIR Phase I and Phase II awards to 
develop desktop computer software that could alphabetize words, a feat 
that previously had been accomplished only on supercomputers. When 
Machine Intelligence went bankrupt, principal investigator Gary Hendrix 
founded Symantec and continued the project. The line of research 
resulted in the first personal computer software that understood 
English, marketed as ``Q&A Software.'' Q&A quickly became an extremely 
successful commercial product and remains a widespread commercial 
application of natural language processing. Symantec research supported 
by NSF SBIR eventually led to six other commercial products and 
contributed to 20 others. Now, Symantec is a leading anti-virus and PC-
utilities Software Company valued at $12B with more than 17,500 
employees worldwide.
    NSF launched the Integrative Graduate Education and Traineeship 
Program (IGERT) in 1997 to encourage innovative models for graduate 
education at colleges and universities across the Nation that would 
catalyze a cultural change in graduate education--for students, faculty 
and institutions. IGERT was designed to challenge narrow disciplinary 
structures, to facilitate greater diversity in student participation 
and preparation, and to contribute to the development of a diverse, 
globally-engaged science and engineering workforce. The result has been 
a cadre of imaginative and creative young researchers. For example, an 
NSF-funded IGERT award to the Scripps Institute of Oceanography (NSF 
#0333444) supported a doctoral student who successfully modeled the 
extinction of the Caribbean monk seal and demonstrated the magnitude of 
the impact of over-fishing on Caribbean coral reefs. This research 
developed improved ecological models, which may influence environmental 
policy and ultimately lead to the preservation of species and 
ecosystems for future generations.

                     Biography for James P. Collins

    Dr. James Collins received his B.S. from Manhattan College in 1969 
and his Ph.D. from the University of Michigan in 1975. He then moved to 
Arizona State University where he is currently Virginia M. Ullman 
Professor of Natural History and the Environment in the School of Life 
Sciences. From 1989 to 2002 he was Chairman of the Zoology, then 
Biology Department. At the National Science Foundation (NSF), Dr. 
Collins was Director of the Population Biology and Physiological 
Ecology program from 1985 to 1986, and Assistant Director for 
Biological Sciences from 2005 to 2009. NSF is the U.S. Government's 
only agency dedicated to supporting basic research and education in all 
fields of science and engineering at all levels. Collins oversaw a 
research and education portfolio that spanned molecular and cellular 
biosciences to global change as well as biological infrastructure. He 
coordinated collaborations between NSF and other federal agencies 
though the President's National Science and Technology Council where he 
chaired the Biotechnology Subcommittee and co-chaired the Interagency 
Working Group on Plant Genomics. He was also NSF's liaison to NIH.
    Dr. Collins's research has centered on the causes of intraspecific 
variation. Amphibians are model organisms for field and laboratory 
studies of the ecological and evolutionary forces shaping this 
variation and its affect on population dynamics. A recent research 
focus is host-pathogen biology and its relationship to population 
dynamics and species extinctions. The role of pathogens in the global 
decline of amphibians is the model system for this research.
    The intellectual and institutional factors that have shaped 
Ecology's development as a science are also a focus of Dr. Collins's 
research, as is the emerging research area of ecological ethics. 
Federal, State, and private institutions have supported his research.
    Dr. Collins teaches graduate and undergraduate courses in ecology, 
evolutionary biology, statistics, introductory biology, evolutionary 
ecology, and professional values in science; he has directed 33 
graduate students to completion of doctoral or Masters degrees. Collins 
was Founding Director of ASU's Undergraduate Biology Enrichment 
Program, and served as Co-Director of ASU's Undergraduate Mentoring in 
Environmental Biology and Minority Access to Research Careers programs.
    Honors include the Pettingill Lecture in Natural History at the 
University of Michigan Biological Station; the Thomas Hall Lecture at 
Washington University, St. Louis; and serving as Kaeser Visiting 
Scholar at the University of Wisconsin-Madison. ASU's College of 
Liberal Arts and Sciences awarded him its Distinguished Faculty Award. 
He is a Fellow of the American Association for the Advancement of 
Science and a Fellow of the Association for Women in Science.
    Dr. Collins has served on the editorial board of Ecology and 
Ecological Monographs as well as Evolution. He is the author of over 
100 peer reviewed papers and book chapters, co-editor of three special 
journal issues, and co-author with Dr. Martha Crump of Extinction in 
Our Times. Global Amphibian Decline (Oxford University Press, 2009).

    Chairman Lipinski. Thank you, Dr. Collins. Dr. McCullough.

  STATEMENT OF DR. RICHARD D. MCCULLOUGH, VICE PRESIDENT FOR 
  RESEARCH; PROFESSOR OF CHEMISTRY, CARNEGIE MELLON UNIVERSITY

    Dr. McCullough. Thank you very much, Chairman Lipinski, 
Ranking Member Congressman Ehlers, Members of the Committee, 
ladies and gentlemen. It is a distinct honor to be before you 
today testifying about investing in high-risk, high-reward 
research.
    I am the Vice President at Carnegie Mellon, as has been 
noted, and I am also an active researcher. I am still a funded 
researcher, and I have a lab of many graduate students. I also 
have taken NSF funding to generate innovation in my lab which 
has turned into a company that employs 70 people. So I 
understand high-risk research from the beginning all the way 
through the cycle, so I think I can speak with it in a special 
way.
    Today I really want to give you a report from the front 
lines, in the trenches, perspective on high-risk, high-reward 
research, and I want to sort of let you know that when a 
researcher comes up with an idea, that researcher is faced with 
how to fund the idea. Often they see it as a high-risk 
proposal, but they don't necessarily understand whether it is a 
high-reward or transformative idea. So often they turn to very 
few sources they have to fund this idea. One needs to get 
preliminary results in order to go to the standard funding 
agencies like NSF or NIH to get your proposal funded, but there 
are no funds to fund the graduate students or post-docs who 
actually do the work, so that you can show that you have 
preliminary results, so that you can actually have these 
proposals funded. Whereas there are programs at the NSF and NIH 
to fund these programs, and NIH has actually done quite a good 
job recently of funding these high-risk proposals, except there 
is very little money to go after, and whether that is real or 
perceived, it is mostly at least perceived. At our university, 
we only have one such high-risk, high-reward proposal that has 
been funded in the whole entire university, and that is at the 
NSF and it is for $66,000.
    So most people go to the standard proposal mechanism to get 
their research funded. So it is the chicken and the egg problem 
in that they don't have the funds to get started, and so they 
have to find funds either within their existing grants or 
within the university or sometimes with foundations to be able 
to get these projects going so that they can get the reward of 
a grant at the NSF or the NIH, so that they can move this 
project forward.
    We have many examples which I have included in my 
testimony, and I won't walk through all of those today. But 
some of the most striking examples of high-risk, high-reward 
research are where a computer scientist, who is an expert in 
data mining, comes together with a person who does brain 
imaging, and the two of them can combine high-output brain 
imaging scans and data mining together to be able to tell what 
people are thinking. Now, you may be scared of that in some 
respect, but it is really important for brain research from an 
injury, for our soldiers, and things like this that we can 
understand how people can heal. This proposal wasn't funded two 
times by the NSF because all the panel people said that it was 
way too high-risk. It hadn't been proven. And I have many other 
examples like that.
    So the system itself has some failings. It is not the 
program officers that are at fault. They are just faced with a 
very difficult challenge of lots of proposals that they have to 
fund, including often determining whether someone is going to 
get tenure or not or whether they are going to remove the 
funding from an established researcher.
    So I have a few recommendations that I will make to you 
today. One is, I think high-risk, high-reward research has to 
be funded. I think it has to be a set-aside so there is not 
competition within the agencies to direct those funds into 
normal proposals, which I think are low-risk and not 
necessarily innovative.
    I think there needs to be a process, guidance by reviewers 
with language written in these high-risk programs that says 
what they expect these proposals to actually be, that maybe 
preliminary results are not needed. We can't depend completely 
on the program officers because often these areas are so 
interdisciplinary and so broad that it is impossible to expect 
one person to be able to make that decision alone. So I think 
special panels with special guidance are important.
    I would recommend funding a milestone proof-of-concept 
where you can apply for high-risk, high-reward research where 
you get $100,000, $200,000 to work on that project, and then if 
you show you can get proof of concept, then you can get further 
payments down the line so that these turn into not just high-
risk with no reward but high-risk so they can be put into the 
system in a normal way.
    Another recommendation that I would make is that there 
should be a mechanism for basic research that becomes 
transformative. Often great discoveries are made that we don't 
know are going to happen until they happen. But when they 
become transformative, there ought to be mechanisms for 
accelerator funds from these agencies that a person can go to 
and say, I have made this amazing discovery. Please help me 
take this and make it transformative. And right now there are 
no mechanisms to do that except to work off the back of these 
proposals.
    So thank you very much for the opportunity to tell you what 
is going on in the front lines.
    [The prepared statement of Dr. McCullough follows:]

              Prepared Statement of Richard D. McCullough

    Chairman Lipinski, Ranking Member Congressman Ehlers, Members of 
the Committee, and ladies and gentlemen. It is a distinct honor to 
testify before the Committee on Investing in High-Risk/High-Reward 
Research.
    My name is Rick McCullough and I am the Vice President for Research 
and a Professor of Chemistry at Carnegie Mellon University. In addition 
to my administrative job, I remain active in doing research. I am also 
a co-founder of Plextronics, Inc., a Pittsburgh-based, high-tech start-
up company with over 70 employees that produces printable, green solar 
technologies and printable inks for lighting and display applications. 
So I have had a variety of experiences with high-risk/high-reward 
research.
    Today, I want to give you a ``frontline/in the trenches'' 
perspective on high-risk/high-reward research. As you know, there are a 
number of excellent reports on high-risk/high-reward or transformative 
research providing an enormous amount of motivating background 
information. These include: the 2007 National Academy Report, ``Rising 
Above the Gathering Storm: Energizing and Employing America for a 
Brighter Economic Future,'' the 2007 National Science Board Report, 
``Enhancing Support of Transformative Research at the NSF,'' and the 
more recent 2009 American Academy of Arts and Sciences (AAAS) Report, 
``ARISE: Advancing Research in Science and Engineering.''
    The United States' leadership in science and technology is at risk. 
This is particularly troublesome when one considers how vital 
innovation is to the US economy and our ability to be competitive as a 
nation. While increased resources for basic research are absolutely 
vital to our ability to remain leaders in science and technology, it is 
also important to consider if the process for obtaining funds for high-
risk/high-reward research is broken. Consider what happens when a 
researcher has a new idea. First, this will require funding to pursue 
the research needed to test that idea. The faculty member can pursue 
basic research funding or a high-risk/high-reward funding.
    Where would a faculty member turn for research funding? Like most 
Tier 1 research universities, Carnegie Mellon receives most of its 
research funding from the Federal Government. Carnegie Mellon's 
percentage of federal science and engineering funding is around 82 
percent, with 13 percent coming from private sources and five percent 
coming from the university. So a faculty member generally thinks of 
federal agencies such as the NIH, NSF, DOE, NASA, or the DOD as sources 
of funding for their new ideas. However, the researcher is faced with 
an extremely competitive grant climate and must maximize the odds of 
receiving funding for the project. What faculty members know or feel is 
that hit rates on NSF proposals have dropped 13 percent over the last 
four years at Carnegie Mellon and NIH hit rates have dropped 18 percent 
over the last three years. Great progress has been made by Congress to 
increase research funding and we are most grateful, however there is a 
lag to realize this new funding. To maximize the probability of getting 
your grant funded (in a regular program or one of the very small high-
risk programs), one of the most important factors is the ability to 
demonstrate proof of concept and/or present preliminary results that 
show the feasibility of the proposed approach. In order to get 
preliminary results, the faculty must either have funded graduate 
students or postdoctoral researchers that actually perform the work. 
Faculty members can sometimes find overlap between the high-risk 
research idea and projects funded by other grants. However, if the idea 
is truly transformational, then probability of success in obtaining 
funding is a problem. That is, you need results to get funded and you 
need funding to get results. I would be shocked if the NIH or the NSF 
had programs where the idea is truly new and is high-risk/high-reward, 
if that proposal would be funded without preliminary results. I could 
be wrong, but I assure you that the number of high-risk funding 
opportunities without preliminary results is diminutive.
    Nevertheless, the NIH is working hard to create new programs such 
as the NIH Director's Pioneer Award, New Innovator Award, and the 
Transformative RO1, all of which accounted for the awarding of $348M to 
115 grantees. This is a tremendous start. However, when a faculty 
member or a brand new researcher is setting out on a new strategic area 
of research he or she may find it difficult to obtain the rare (18 in 
2009) Director's Award. I hope for an increase in the number of 
pioneers for the future. I recommend that Congress explore directing 
additional funding toward Pioneer Awards that stimulate high-risk 
research projects.
    If you go to the NSF, the situation is worse. In my opinion, the 
system is broken. The NSF has had the Small Grants for Exploratory 
Research (SGER) program that evolved to the Early-concept Grants for 
Exploratory Research (EAGER). These grants began, as I recall, as one-
time $50K grants that were rarely funded. I can tell you about a grant 
that I submitted with two other top researchers that would create a 
completely new way to make plastic superconductors that was not funded; 
it was probably too risky and we did not have proof of concept. 
Nevertheless, the program has expanded where $2M/division has been 
allocated for transformative research. This is a start, but I believe 
that the system of evaluation and funding of high-risk/high-reward 
research at the NSF needs to be improved. My colleagues at Carnegie 
Mellon have related to me that it is often easier to get resources for 
high-risk research by getting preliminary results at a very slow pace 
and then using the normal grant mechanisms to fund transformational 
research. This is the way I look for funding for high-risk research as 
well. From of the perspective of these faculty members, high-risk/high-
reward research funding is virtually unavailable from traditional 
federal sources.
    Reading the National Science Board's 2007 report entitled 
``Enhancing Support of Transformative Research at the NSF,'' one can 
find that many of the needed improvements to the program are 
recommended in that report. I find that report echoes many of the 
recommendations I would make to you today.
    For example, I agree with the NSB report that our first challenge 
is clearly defining transformative or high-risk/high-reward research 
and how to distinguish it from the definition of basic research. It is 
important to note two caveats to defining high-risk/high-reward 
research: 1. scientists and engineers are often not that good at 
marketing and sales and many will rarely think of their ideas initially 
as high-reward or transformative and 2. many scientific discoveries 
occur in basic science and are even accidental and then become 
transformative.
    In addition, in the EAGER program at NSF leaves funding of high-
risk/high-reward proposals to program directors. This presents multiple 
challenges in the evaluation process, such as: 1. program officers 
often do not have the expertise to determine what is high-risk/high-
reward research; 2. program officers do not often have the expertise to 
judge the proposals which can be broad and highly interdisciplinary in 
scope; and 3. the monies that are set-asides are usually at the 
discretion of the program officers who are faced with the pressure of 
not having enough resources to fund highly rated proposals. For 
example, a program officer who is faced with funding a mid-career 
scientific leader, or funding the last attempt by a junior faculty 
member who is up for tenure, would find it extremely difficult to 
divert funds for high-risk/high-reward research. In addition, highly 
interdisciplinary research that is seeking high-risk research funding 
will find itself in one discipline with a program officer from that one 
discipline. In theory, such program officers can collaborate to fund 
the proposal across disciplines by going to other program officers and 
asking if they are interested in jointly funding the proposal. However, 
collective funding across divisions is probably a difficult process. 
This is not to be critical of the NSF program managers. They have a 
very difficult task because the reality is that they do not have enough 
resources to fund all the great proposals that they receive and they 
face ever-changing reporting requirements and short-term 
accountability.
    Consequently, high-risk/high-reward proposal programs are not 
viable options in cases such as these. As an example, Carnegie Mellon 
has one $66,000 EAGER grant from the NSF and zero NIH Director's 
Pioneer Awards, zero New Innovator Awards, and zero Transformative RO1 
grants.
    Alternatively, a researcher might hope to get funding for a high-
risk/high-reward proposal via the normal NSF or NIH process; however 
these proposals are not a good fit for that process either. Typical 
panels that review the basic research proposals clearly do not reward 
high-risk/high-reward proposals with funding. Panels generally (not 
all) reward incremental research where preliminary results are 
absolutely critical to funding. Panels are often the ``white blood 
cells'' of high-risk/high-reward research, since these proposals are 
easy targets and the reason for elimination from competition. As one 
advisory board member to one of the divisions of the NSF said, the 
system is set up to reward low-risk research. One program manager's 
response was, if he is expected to report in one year how this research 
has contributed to our country, how can he take a chance on high-risk 
research? I will give you multiple anecdotes on proposals in the 
regular process that get killed for being high-risk/high-reward 
proposals.
    I do believe that one solution might be to create special panels 
led by hand-picked committee chairs that would review proposals for 
their potential as transformational or high-reward. New guidance by the 
NSF could instruct special panels and/or outside reviewers that 
preliminary results are not necessary so that researchers (new and old) 
moving into new areas of high-risk research can have a chance at 
funding. I would also suggest a system where seed funding can be 
provided and, after proof of success, additional funds can be released. 
For example, funding might be provided for two years and with success 
of converting the high-risk research into proof of concept results, an 
additional release of funds could occur.
    Faculty members can also turn to foundations for the support of 
high-risk/high-reward research. Examples where Carnegie Mellon has had 
success in this regard would include the Keck Foundation, the Heinz 
Endowments, the R.K. Mellon Foundation, the Gordon and Betty Moore 
Foundation, the John D. and Catherine T. MacArthur Foundation, and the 
Doris Duke Foundation. However, the opportunities for funding from 
these foundations are highly limited to a few faculty members within 
the university. In the same vein, private support is limited to a few 
selected centers or individuals. An example would be private support 
for programs such as the Ray and Stephanie Lane Center for 
Computational Biology.
    In addition, one strange aspect to high-risk/high-reward research 
is that many great discoveries are accidental. As the late Carnegie 
Mellon Nobel Prize winner Herb Simon used to say, to do world-class 
research, one should look for surprises and explain them. This is how 
the material C60 was discovered. The late Nobel Prize 
winner, Rick Smalley of Rice was shooting high powered lasers at 
graphite and off came buckyballs or C60. It was later found 
that when C60 is combined with certain conducting polymers 
(that we discovered), one can make an ink that can be printed to form a 
plastic solar cell that absorbs light from the sun and makes energy. 
The transformational discovery of C60 may end up 
transforming energy production by making solar incredibly inexpensive.

Examples of High-Risk/High-Reward Projects at Carnegie Mellon

Reading Minds with Computers
    In the early 2000's two of our top professors (one in psychology 
and one in computer science) wrote two NSF proposals to seek funding 
for research that applies machine learning to fMRI (functional magnetic 
resonance imaging) brain image analysis. The idea is that using high-
speed/data mining of brain scans, it might be possible to understand 
human thoughts. The use in medical brain research and therapy such as 
the treatment of traumatic brain injury, as only one example, would be 
profound. The first proposal received weak reviews and was not funded. 
The reviews said that while the impact of the proposed work would be 
very high, the techniques were unproven and the work was too high-risk. 
A year or so later, a second proposal was submitted, this time with 
compelling preliminary results showing that the researchers could train 
machine learning programs to decode various cognitive states of a 
person from their brain image data (e.g., whether they were reading a 
sentence or viewing a picture). Again, the reviews said this was 
unproven technology and the proposed research was too high-risk, in 
comparison with other proposals. It was headed for a rejection, but a 
wise NSF program manager used his discretion to bump it up into the 
barely fundable category, and the NSF provided small grant so that we 
could start the work. The Provost's office at Carnegie Mellon provide 
funds when the NSF funds ran out and eventually we were able to get 
some funding from the Keck Foundation. This work has been a huge 
success and has been featured recently on 60 Minutes. The one of the 
success stories of a high-risk/high-reward project.

Using the Power of Ubiquitous Sensors and Computers as Safety Sensors
    We have a team of top professors in Civil Engineering and 
Electrical and Computer Engineering that have created hardware sensors 
and software that can be used anywhere at anytime to monitor buildings, 
roads, bridges, water infrastructure, etc. This group recently 
submitted a proposal, whose reviews were generally quite complimentary, 
and described by many of the reviewers as a clear example of a high-
risk, high-reward endeavor. However, they were also criticized for not 
presenting sufficient results to back up the proposed approach as being 
feasible. My office at Carnegie Mellon is currently funding the project 
and supporting one student. However, the project is at risk of not 
continuing.

Using Free Human Work on the Internet to Digitize Books
    We have a project by an award winning computer science professor 
that proposes to use computer programs to digitize books. When people 
open accounts on gmail, Yahoo, etc., or buy tickets on-line they have 
to translate a distorted word to be able to open said account or buy 
tickets. These distorted words called CAPTCHAs prevent computers from 
opening the accounts, because computers cannot read the distorted 
words. However, humans can translate the distorted words with ease. It 
turns out that distorted words are a problem when books are digitized. 
A person makes a copy of the book and at the edge, some of the words 
are distorted and therefore cannot be read by a computer. The 
professor's idea was to use the same distorted words from book 
digitization as the words that need to be translated for book 
digitization. Therefore, free human work to translate the distorted 
words to open accounts gets sent back and help to digitize books. The 
NSF declined to fund this work. The work was funded internally and led 
to ReCAPTCHA and a spin-out company from Carnegie Mellon that was 
recently sold to Google.

Others
    We had a project that uses machines to interpret biomedical 
research data and the computer can teach itself what to look for in 
cancer diagnostics. We have proven that machines can do this work 
better than humans can. This project was funded by the Scaife 
Foundation, then Keck, and by private sources, but was always reviewed 
by the NIH as high-risk/high-reward research and was never initially 
funded. Another similar project uses high-power computer science to 
attack massive data sets related to cancer diagnostics. The professor 
told me that he wrote a proposal to the NSF that was funded and is 
funding the high-risk project at a 10 percent level from that grant. 
His initial grant focusing on this approach was rejected as being too 
high-risk.
    Our work at the university in Green Chemistry has had a very 
difficult time securing federal funding. One of our professors has 
created revolutionary new catalysts that activate non-toxic hydrogen 
peroxide to create systems that, in a green way, can be used to clean 
up toxic rivers, bleach pulp in the paper bleaching process, allow very 
little water to be used in laundry wash cycles, etc. He has not been 
able to secure NSF funding.
    We have multiple areas of futuristic research at Carnegie Mellon, 
such as Claytronics (the ability to make programmable matter) that have 
struggled mightily to receive any funding. These are just a few 
examples of high-risk/high-reward research just at Carnegie Mellon, so 
you can imagine what high-risk/high-reward research that is being not 
(and not funded) at other top universities.
    In closing, I want to again express appreciation for the support 
Congress has shown in restoring growth to federal research funding. In 
combination with the innovation funding provided in the American 
Recovery and Reinvestment Act, this support reflects the critical role 
that American higher education must play in restoring economic 
competitiveness and growth. The comments I have shared with you today 
reflect my belief that this full potential can only be realized by 
recognizing the critical importance of supporting high-risk/high-reward 
research. I believe that actions to increase support for those programs 
that do fund high-risk research and efforts to infuse a focus on 
breakthrough research into existing program review processes can bring 
the full return we must realize from this renewed investment in 
American research.

References

Rising Above the Gathering Storm: Energizing and Employing America for 
        a Brighter Economic Future, pdf of book found at http://
        books.nap.edu/catalog.php?record-id=11463#toc

2007 National Science Board Report, ``Enhancing Support of 
        Transformative Research at the NSF,'' http://www.nsf.gov/nsb/
        publications/landing/nsb0732.jsp

2009 AAAS Report, ``ARISE: Advances Research in Science and 
        Engineering,'' http://www.amacad.org/arisefolder/
        ariseReport.pdf

                  Biography for Richard D. McCullough

    Richard McCullough was appointed Vice President for Research at 
Carnegie Mellon University in July 2007. In this new senior leadership 
position, McCullough will nurture interdisciplinary research 
initiatives and oversee sponsored research, technology 
commercialization and a number of cross-college research centers. Prior 
to this position he served as the Dean of the Mellon College of Science 
at Carnegie Mellon University. He came to Carnegie Mellon in 1990 as an 
Assistant Professor and quickly rose through the tenure ranks, being 
promoted to Associate Professor in 1995 and Professor in 1998. In 1998 
he assumed the role of Department Head of Chemistry. He was appointed 
Dean of the Mellon College of Science in 2001.
    McCullough is internationally known as the world's expert in the 
area of printable electronics and is well known for his discovery of 
regioregular polythiophenes--a material that led to plastic solar cells 
and plastic transistors. His research focuses on the design and 
understanding of the structure-property relationships in conducting 
materials and nanoelectronics.
    In addition to his position at Carnegie Mellon, McCullough is also 
the chief scientist and founder of Plextronics, Inc., the world leader 
in developing active layer technology for printed electronics devices, 
such as organic light-emitting diode displays, polymer solar cells and 
plastic circuitry. Since its inception in 2002, the Pittsburgh Company 
has grown to more than 70 employees and received numerous honors, among 
them being named the 2008 Going Green Top 100 Company and a Wall Street 
Journal Technology Award runner-up.
    He was a postdoctoral fellow at Columbia University and holds a 
Ph.D. from the Johns Hopkins University and a B.S. from the University 
of Texas at Dallas.

    Chairman Lipinski. Thank you, Dr. McCullough. Dr. Rubin.

STATEMENT OF DR. GERALD M. RUBIN, VICE PRESIDENT AND DIRECTOR, 
 JANELIA FARM RESEARCH CAMPUS, HOWARD HUGHES MEDICAL INSTITUTE

    Dr. Rubin. Thank you for the opportunity to speak before 
you today. As a science philanthrope whose explicit goal is the 
discovery of new knowledge, the Howard Hughes Medical 
Institute, or HHMI, seeks to use its investment of intellectual 
and financial capital to see growth and change to foster fresh 
thinking.
    HHMI's biomedical research philosophy can be summarized in 
three words: people, not projects. By appointing scientists as 
HHMI investigators rather than awarding research grants, the 
Institute provides long-term flexible funding that enables its 
researchers to pursue their scientific interests wherever they 
lead. These two flagship research programs, the HHMI 
Investigator Program, currently employs 346 researchers 
selected through rigorous national competitions who direct 
institute laboratories on the campuses of 72 universities and 
other research organizations throughout the United States.
    HHMI scientists include mathematicians, physicists, 
engineers, physicians, chemists and classically trained 
molecular and cellular biologists.
    The success of HHMI's people-not-projects philosophy can be 
seen in the high productivity and breakthrough insights 
generated by HHMI investigators. For example, HHMI 
investigators have been awarded Nobel Prizes in nine of the 
past 11 years, and 14 investigators have all received the Nobel 
Prize. Just earlier this week, HHMI investigator Jack Szostack 
was awarded the Nobel Prize in physiology or medicine, and HHMI 
researcher Thomas Steitz was awarded the Nobel Prize in 
chemistry.
    With freedom and flexibility come high expectations for 
intellectual output. HHMI demands creativity and innovation. 
Investigators are expected to work at the frontiers of their 
chosen field, to ask fundamental questions, and to take risks.
    Although the Institute already had the highly successful 
HHMI investigator program, the scientific leadership continue 
to explore new ways to support the research of some of this 
nation's most creative scientists. These discussions led to the 
establishment of HHMI's Janelia Farm Research Campus in 
Northern Virginia, which opened in 2006.
    The blueprint for Janelia Farm grew out of an 
acknowledgment by HHMI leadership that while most biomedical 
problems are handled well in the university setting, there are 
some that are better addressed in a place where small groups of 
researchers with different skills can work together without the 
barriers typically encountered at a university.
    I have described Janelia Farm more fully in my written 
testimony, but in the interest of time, I would like to 
conclude by giving my personal perspective on Federal support 
for innovative research.
    The central question I have been asked to address is what 
is the best mechanism that Federal funding agencies can use to 
support high-risk, high-reward research? With regard to 
funding, my own personal belief, backed up by HHMI's nearly 30-
year experiment, is that in the long run, high-reward research 
comes from focusing on people, not projects. In today's funding 
environment, researchers are compelled to define and advance 
the goals, methods, and likely outcomes of the research project 
in a detailed grant application. While this funding model is 
appropriate for some types of biomedical research, it has two 
major limitations. First, proposals for higher risk projects, 
even those that may have enormous impacts if successful, have 
traditionally fared poorly. Second, the ability to move quickly 
to take advantage of unforeseen targets of opportunity is 
fairly constrained. How can a scientist capitalize on a flash 
of insight if he or she must first write a grant proposal and 
then wait a year, even if the grant application is successful 
for the funding to test the idea? Federal funding agencies need 
to do a better job of providing research support under terms 
that permit rapid changes in research direction, encourage 
taking on challenging research problems, even if the chance of 
short-term success is low.
    I think these changes could bring more innovation per 
dollar spent. In 2003 I was asked to join a task force convened 
by Dr. Elias Zerhouni, NIH Director at that time. The group was 
charged with recommending new ways to find high-risk, high-
impact research. The primary recommendation of our panel was to 
establish a new set of awards to researchers based on their 
track record of innovation. In fact, the journal Science 
covered our panel's recommendations in a news story headlined, 
``NIH to Award People, Not Projects.'' That headline nicely 
summed up our recommendations, but in practice, the NIH came up 
short in carrying out this initiative.
    Take the NIH Director's Pioneer Awards, for example, which 
aimed specifically at stimulating highly innovative research 
and supporting promising new investigators. Our task force 
recommended that the NIH award 10 percent of its R01 grants, 
which would equate to roughly 700 grants per year, on a people-
not-projects basis. In 2004, the first year the awards were 
made, the NIH selected only nine Pioneer Award recipients from 
among approximately 1,000 nominations.
    It is somewhat more encouraging to see that this year the 
NIH has awarded a total of 115 grants for high-risk, high-
reward research through its Pioneer Awards, New Innovator 
Awards, and the NIH Director's Transformative R01 Awards. The 
total number of these awards, however, still falls far short of 
our 2003 recommendation.
    Even with these new awards, the NIH is still heavily 
weighted toward project-oriented research with 98 percent of 
grants going to projects. As I stated earlier, I strongly 
believe that giving money to scientists of exceptional and 
demonstrated creativity and allowing them to follow their 
instincts is a better way to promote innovation. In my opinion, 
even a modest shift in the Federal research portfolio, going 
from perhaps 98 percent to 90 percent project-oriented could 
make a big difference in producing innovative and potentially 
transformative research results.
    I would like to end with a quotation from the Nobel Prize 
winner, Max Perutz, who directed the Medical Research Council 
Laboratory of Molecular Biology in England for more than 20 
years. ``Creativity in science, as in the arts, cannot be 
organized. It arises spontaneously from individual talent. 
Well-run laboratories can foster it, but hierarchical 
organization, inflexible, bureaucratic rules, and mounds of 
futile paperwork can kill it. Discoveries cannot be planned; 
they pop up, like Puck, in unexpected corners.``
    Thank you, Mr. Chairman.
    [The prepared statement of Dr. Rubin follows:]

                 Prepared Statement of Gerald M. Rubin

Mr. Chairman and Members of the Committee:

    Thank you for the opportunity to speak before you today. I am 
Gerald Rubin, a Vice President at the Howard Hughes Medical Institute 
(HHMI) and Director of the Janelia Farm Research Campus in Ashburn, 
Virginia. I am honored to testify before the committee as it begins to 
examine the mechanisms for funding high-risk, high-reward research, and 
the appropriate role of the Federal Government in supporting such 
research in the United States.
    My testimony will cover three broad areas: HHMI's approach to 
biomedical research; HHMI's motivation for creating a new kind of 
research center at Janelia Farm; and a summary statement that reflects 
my perspective on how the Federal Government could improve its support 
of high-risk, high-reward research.

The Howard Hughes Medical Institute Invests in People, Not Projects

    Nearly 25 years ago, as the HHMI Trustees prepared to sell the 
Hughes Aircraft Company to General Motors Corp., in order to establish 
the first permanent endowment for the Howard Hughes Medical Institute, 
The New York Times issued an emphatic challenge to the leadership of 
the newly reorganized entity. In an editorial that was published on 
June 15, 1985, the newspaper urged the Institute to avoid the 
temptation to plug gaps in federal spending and instead to ``be more 
venturesome and fund high-risk research, and by methods as different as 
possible from the Government's.''
    As a science philanthropy whose explicit goal is the discovery of 
new knowledge, HHMI seeks to use its investments of intellectual and 
financial capital to seed growth and change, to foster fresh thinking.
    HHMI's biomedical research philosophy can be summarized in three 
words: people, not projects. By appointing scientists as HHMI 
investigators--rather than awarding research grants--the Institute 
provides long-term, flexible funding that enables its researchers to 
pursue their scientific interests wherever they lead.
    The Institute takes the ``long view,'' preferring to nurture the 
creativity and intellectual daring of scientists who are willing to set 
aside conventional wisdom or the ``easy'' question for a fundamental 
problem that may take many years to solve. Among the distinguishing 
characteristics of HHMI's scientists are qualities such as creativity, 
a high tolerance for risk-taking, and a commitment to discovery, 
productivity, and perseverance.
    HHMI's unique research model is an imaginative and powerful 
alternative to project-based research support or funding biomedical 
research through grants. The Institute's flagship research program, the 
HHMI Investigator Program, currently employs 346 researchers who direct 
Institute laboratories on the campuses of 72 universities and other 
research organizations throughout the United States. HHMI scientists 
represent a wide range of biomedical research disciplines--from 
chemistry, neuroscience, and bioinformatics to structural biology, 
immunology, and clinical genetics. They include mathematicians, 
physicists, engineers, physicians, chemists, and classically trained 
molecular and cellular biologists.
    The success of HHMI's ``people, not projects'' philosophy can be 
seen in the high productivity and breakthrough insights generated by 
HHMI investigators. In recent years, HHMI researchers have made many 
major research advances, including:

          Identifying a new drug that is now approved by the 
        FDA to treat patients whose chronic myeloid leukemia failed to 
        respond to standard treatment with Gleevec

          New microscopes and imaging techniques that let 
        researchers visualize cells and proteins with unprecedented 
        resolution

          A non-invasive test for genetic mutations associated 
        with colon cancer

          Gene microarrays and ``protein chips,'' enabling 
        researchers to simultaneously measure the function of thousands 
        of genes or proteins.

    HHMI investigators have been awarded Nobel Prizes in eight of the 
last 10 years, and 12 investigators overall have received the Nobel 
Prize. Currently, there are 131 HHMI investigators who are members of 
the National Academy of Sciences. Election to the Academy--one of the 
highest honors a scientist can receive--is based on distinguished and 
continuing achievement in original research. HHMI investigators 
presently compose about six percent of the Academy's 2,100 current 
members (this does not include foreign associates).
    Since the early 1990s, investigators have been selected through 
rigorous national competitions. The Institute solicits applications 
directly from scientists at medical schools and other research 
institutions in the United States, with the aim of identifying those 
who have the potential to make significant contributions to science. 
HHMI employs an open application process to ensure that it is selecting 
its researchers from a broad and deep pool of scientific talent.
    After they have been selected, HHMI investigators continue to be 
based at their home institutions, typically leading a research group of 
10-25 students, postdoctoral associates and technicians, but they 
become Institute employees and are supported by HHMI field staff 
throughout the country.
    With freedom and flexibility come high expectations for 
intellectual output. HHMI demands creativity and innovation. 
Investigators are expected to work at the frontiers of their chosen 
field, to ask fundamental questions, and to take risks. HHMI prizes 
impact over publication volume in its merit-based renewal of 
investigator appointments and recognizes that some areas of research 
will proceed more slowly than others.
    In reviewing its scientists, HHMI expects not only that its 
investigators be talented and productive scientists, but also that they 
demonstrate some combination of the following attributes to an extent 
that clearly distinguishes them from other highly competent researchers 
in their field:

          They identify and pursue significant biological 
        questions in a rigorous and deep manner.

          They push their chosen research field into new areas 
        of inquiry, being consistently at its forefront.

          They develop new tools and methods that enable 
        creative experimental approaches to biological questions, 
        bringing to bear, when necessary, concepts or techniques from 
        other disciplines.

          They forge links between basic biology and medicine.

          They demonstrate great promise of future original and 
        innovative contributions.

    HHMI's annual research budget, though substantial, is dwarfed by 
the Nation's investment in research through the National Institutes of 
Health and the National Science Foundation. Yet in holding fast to a 
distinctive model for supporting scientific research, HHMI uniquely 
serves science, creating a culture of inquiry that encourages the free 
and unfettered pursuit of knowledge.

Examples of HHMI's Approach to Science

    HHMI scientists work avidly and passionately toward tomorrow's 
discoveries. Sometimes inventing wholly new areas of study, HHMI 
researchers are pioneers in such areas as neuroscience, genomics, and 
computational biology. The examples below are just a few of many that 
illustrate HHMI's approach to science.

Richard Axel and Linda Buck

    The olfactory mechanics that make possible the exquisite ability to 
discern smells from the most subtle to the blatant have been the 
subject of study for HHMI investigators Richard Axel and Linda B. Buck 
for much of their research careers. Axel and Buck, who joined HHMI in 
1984 and 1994, respectively, were awarded the 2004 Nobel Prize in 
Physiology or Medicine for their discoveries of ``odorant receptors and 
the organization of the olfactory system.''
    The process of smelling an odor begins with odorant receptors that 
are located on the surface of nerve cells inside the nose. Researchers 
now know that when an odorant receptor detects an odor molecule, it 
triggers a nerve signal that travels to a way station in the brain 
called the olfactory bulb. Signals from the olfactory bulb, in turn, 
travel to the brain's olfactory cortex. Information from the olfactory 
cortex is then sent to many regions of the brain, ultimately leading to 
the perceptions of odors and their emotional and physiological effects.
    The trail to the Nobel began many years earlier as an attempt to 
understand how the brain creates an internal representation of the 
external sensory world. Little was known about the mechanics of smell 
before Axel and Buck published their seminal discovery of odorant 
receptors.
    In 1991, Axel and Buck (who was working on her second postdoctoral 
fellowship in Axel's lab), were three years into their search for 
odorant receptors. Approaching the problem with her training in 
immunology, Buck had been trying to identify rearranged genes in the 
mammalian nervous system. She was intrigued by the possibility that 
gene rearrangement or gene conversion might be involved in the 
generation of a varied set of odorant receptors or regulate their 
expression, as with antigen receptors in the immune system. Buck became 
obsessed with finding the odorant receptors and stayed on in Axel's lab 
to look for them.
    Buck and Axel, who is at Columbia University, initially adopted an 
``unbiased approach'' with regard to the structure of odorant 
receptors, choosing to focus on two assumptions: that the receptor 
proteins would be selectively expressed by olfactory sensory neurons 
and, given the structural diversity of odorants, there would be a 
family of related, but varied, odorant receptors that would be encoded 
by a family of related genes.
    Their efforts produced nothing at first. The tide turned when, 
using scattered evidence from other labs, Buck decided to narrow her 
search to G protein-coupled receptors (GPCRs), many of which were known 
to be involved in cell signaling. Making use of the recently developed 
gene amplification technology called PCR, or polymerase chain reaction, 
Buck decided to conduct an exhaustive search for GPCRs in the olfactory 
epithelium by taking a novel approach.
    Further analysis of the PCR products narrowed the search to one 
candidate. Buck cloned this PCR product, sequenced five of the clones, 
and found precisely what she had been looking for. When Buck finally 
found the genes in 1991, she could not believe her search was over. 
Furthermore, none of the genes she found had ever been seen before. 
They were all different, but all related to each other.

Roderick MacKinnon

    Roderick MacKinnon of The Rockefeller University joined the Howard 
Hughes Medical Institute in 1997 as a self-taught structural biologist. 
Already an accomplished scientist, MacKinnon considered his HHMI 
appointment a special opportunity to take an entirely new research 
direction in order to further his work.
    Prior to coming to Rockefeller, MacKinnon was a successful 
scientist at Harvard Medical School, where he ran a laboratory that 
studied ion channels, tiny doughnut-shaped pores that penetrate the 
membrane that surrounds living cells. They permit ions--charged atoms 
of potassium, sodium, chloride, and calcium--to flow across cell 
membranes, thereby generating electrical signals. Ion channels are 
fundamental to health and to the normal function of the human body; 
their impulses create the sparks of the brain and nervous system, 
allowing us to walk, talk, fall in love, and, for example, cast a 
fishing line with accuracy.
    Building on decades of clever observations by their predecessors, 
MacKinnon and others had been inching toward a deeper understanding of 
how the pores performed their feats of exquisite discrimination among 
ions and responsiveness to minute changes in their environment--
enabling the cell membrane to suddenly become permeable, but only to 
highly specific types of ions.
    But though the genes behind the channel proteins had been cloned, 
which gave scientists new traction on the problem, channel aficionados 
were still struggling.
    Trained as a physician, MacKinnon decided to teach himself the 
rudiments of x-ray crystallography because he wanted to find a way to 
solve a specific problem: defining the structure and mechanism of the 
channel that controls the flow of potassium into the cell. To devote 
himself to this pursuit, he moved his laboratory from Harvard to 
Rockefeller University, where he was named an HHMI investigator shortly 
after joining the faculty. His creativity, ability to approach his 
research from a new perspective, and single-minded pursuit of a 
significant scientific problem exemplify many of the attributes HHMI 
seeks in its investigators.
    In April 1998, the journal Science published two elegant articles 
by MacKinnon. In the first article, he defined the ``inverted teepee'' 
structure of the potassium channel in a strain of bacteria and in the 
second he confirmed that the human potassium channel was structurally 
similar. MacKinnon continues to generate new insights that illuminate 
the structure and function of ion channels. These insights are critical 
to understanding new approaches for treating human diseases as varied 
as hypertension and epilepsy. Like many other HHMI investigators, 
MacKinnon has focused on fundamental biological questions that have 
significant implications for the understanding and treatment of human 
disease.
    Five years after those Science articles were published, MacKinnon 
received the ultimate vindication of his out-of-the-box creativity and 
persistence in the face of high-risk: He shared the 2003 Nobel Prize 
for Chemistry with Johns Hopkins researcher Peter C. Agre who 
discovered water channels in cells.

Huda Zoghbi

    Using some of the most advanced techniques in genetics and cell 
biology, HHMI investigator Huda Zoghbi and her collaborators unraveled 
the genetic underpinnings of a number of devastating neurological 
disorders, including Rett syndrome and spinocerebellar ataxia type 1. 
Their discoveries may one day lead to better methods for treating these 
diseases and provide new ways of thinking about more common 
neurological disorders, including autism, mental retardation, and 
Parkinson's disease.
    Zoghbi's interest in Rett syndrome began long before she 
established her own research laboratory at Baylor College of Medicine. 
While in the second year of medical residency, Zoghbi encountered a 
very puzzling patient. The girl had been a perfectly healthy child, 
playing and singing and otherwise acting like a typical toddler. Around 
the age of two, she stopped making eye contact, shied away from social 
interactions, ceased to communicate, and started obsessively wringing 
her hands. The girl made a huge impression on Zoghbi, who set out to 
determine what could have caused this sudden neurological 
deterioration.
    Sixteen years after she saw that first patient, Zoghbi and her 
collaborators identified MECP2, the gene responsible for Rett syndrome. 
Children afflicted with this rare neurodevelopmental disorder develop 
normally for about six to 18 months and then start to regress, losing 
the ability to speak, walk, and use their hands to hold, lift, or even 
point at things. MECP2, it turns out, encodes a protein whose activity 
is critical for the normal functioning of mature neurons in the brain; 
it is produced when nerve cells are forming connections as a child 
interacts with the world. The disease occurs primarily in females, 
because boys who inherit an inactive form of MECP2--which lies on the X 
chromosome--usually die shortly after birth. Girls survive because, 
with two X chromosomes, they stand a good chance of inheriting a 
healthy copy of the gene.
    For the first 15 years of her career, Zoghbi spent 20 percent of 
her time seeing patients with childhood neurological disorders. Driven 
by a desire to improve the clinical outcome of her patients, she became 
convinced that more basic research was needed.
    Zoghbi and her colleagues have also identified the mutation 
responsible for spinocerebellar ataxia type 1 (SCA1), a 
neurodegenerative disorder that renders its victims unable to walk or 
talk clearly, or eventually to even swallow or breathe. The culprit is 
a sort of genetic stutter that increases the size of the SCA1 gene. The 
normal gene harbors a stretch of nucleotides in which the sequence CAG 
is repeated about 30 times. In individuals with the disease, the tract 
expands to include 40 to 100 iterations. As a result, the product of 
the mutant gene--a protein called ataxin-1--grows large and sticky, 
forming clumps throughout the cell. These ataxin-1 aggregates overwhelm 
the molecular machinery that cells use to recycle damaged proteins and 
eventually disable the neurons involved in controlling movement. Using 
mice and flies that produce the mutant protein, Zoghbi is now searching 
for compounds that enhance the clearance of ataxin-1 tangles. Such 
drugs could slow the progression of the disease or prevent it 
altogether.

Creating a New Scientific Culture at Janelia Farm

    Although the Institute already had the highly successful HHMI 
Investigator Program, the scientific leadership continued to explore 
new ways to support the research of some of this nation's most creative 
scientists. The genesis of the Janelia Farm Research Campus occurred in 
1999 in a series of informal conversations at HHMI about ways to expand 
the boundaries of biomedical research.
    The blueprint for Janelia Farm grew out of an acknowledgment by 
HHMI leadership that while most biomedical problems are handled well in 
a university setting, there are some that are better addressed in a 
place where small groups of researchers with different skills can work 
together without the barriers typically encountered at a university. 
Development of new tools to facilitate biological discovery, for 
example, can require diverse expertise. But at universities, scientists 
from different fields are often compartmentalized, and demands placed 
on researchers by their departments may restrict collaboration outside 
those walls. To avoid these constraints, HHMI decided to bring together 
researchers from disparate disciplines in a free-standing campus.
    Scientists at the Janelia Farm Research Campus, which opened in 
2006, are working in two synergistic areas: discovering the basic rules 
and mechanisms of the brain's information-processing system, and 
developing biological and computational technologies for creating and 
interpreting biological images. These two areas were chosen because 
they are truly large, unsolved problems in biology and because there is 
a very good chance that they will not be solved by one laboratory or by 
scientists in one discipline.
    In planning Janelia Farm, HHMI carefully studied the structure and 
scientific culture of other important research models at both academic 
and for-profit biomedical laboratories, including the Medical Research 
Council Laboratory of Molecular Biology (MRC LMB) in England and the 
former AT&T Bell Laboratories in the United States. The MRC LMB and 
AT&T's Bell Labs are generally considered to have been the most 
successful research institutions in biology and electronics, 
respectively.
    Though the MRC LMB and Bell Labs were different in many ways, they 
did have several things in common. Both institutions kept research 
groups small, and principal investigators worked at the lab bench. The 
single sponsor provided all funding--applying for outside grants was 
not allowed--and good support services and infrastructure were in 
place. Notably, both institutions evaluated their own people rather 
than rely on expert opinions from outsiders. HHMI decided to 
incorporate these core concepts into Janelia Farm.
    Researchers at Janelia Farm are freed from most of the 
administrative, grant writing, and teaching duties that consume time at 
a university. Traditional academic environments are suitable for a 
large proportion of research projects, but they can be too conservative 
and restrictive, stifling the kinds of creative, long-term projects 
that can lead to true breakthroughs. This is true, in part, because the 
reliance on external funding sources forces scientists to define their 
research programs in advance when they apply for grants.
    By setting the course of the research plan up front, scientists are 
restricted in their ability to pursue questions and opportunities that 
arise during their studies. The bulk of the scientific community is 
limited to projects that can be funded by peer-review committees, which 
tend to be very conservative. These grants have to be reviewed every 
three to five years, making it very difficult for people to take on 
high-risk, high-reward projects.
    It is important to remember that we think of Janelia Farm as an 
experiment. We don't have all the answers. We have a working 
hypothesis. We formulated the hypothesis by studying previously 
successful research institutions and analyzing what made them 
successful. We may not get it exactly right at first, but we'll adapt. 
We will revise the hypothesis, like any good scientist would do.
    Ultimately, we believe the success of our approach might be 
measured by a ``deletion test.'' Twenty years from now, would the 
scientific landscape look substantially different if Janelia Farm's 
contributions were to be deleted? Of course, since Janelia Farm is only 
three years old, we do not know the answer yet.

Summary Statement and Perspective on Federal Support for Scientific 
                    Research

    The central question that I have been asked to address is what is 
the best mechanism that federal funding agencies can use to support 
high-risk, high-reward research. I have outlined HHMI's approaches, 
which focus on people, not projects. It is worth noting here that 
although there are numerous organizational cultures in which scientific 
research is conducted, from HHMI's perspective, no single culture has 
emerged as ``the best.''
    But with regard to funding, my own personal bias, backed up by 
HHMI's nearly 30-year ``experiment,'' is that in the long run, high-
reward research comes from focusing on people, not projects. And I 
believe that federal funding agencies, such as the National Institutes 
of Health and the National Science Foundation, should allocate a 
greater portion of their research portfolios to supporting truly 
innovative scientists (identified as such by their track record) and 
not make funding decisions based on the projects those researchers 
propose to study.
    In today's funding environment, researchers are compelled to define 
in advance the goals, methods and likely outcomes of their research 
project in a detailed grant application. While this ``funding model'' 
may be appropriate for some types of biomedical research, it has two 
major limitations. First, proposals for higher-risk projects--even 
those that may have enormous impact if successful--have traditionally 
fared poorly. Second, the ability to move quickly to take advantage of 
unforeseen targets of opportunity is severely constrained.
    As I like to say, how can a scientist capitalize on a flash of 
insight that occurs at 3 A.M., if he or she must first write a grant 
proposal and then wait a year--even if their grant application is 
successful--for funding to test the idea? Federal funding agencies need 
to do a better job of providing research support under terms that 
permit rapid changes in research direction and encourage taking on 
challenging research problems, even if the chance of short-term success 
is low.
    I think these changes will bring ``more innovation per dollar 
spent'' without adding more money into the research budgets of these 
agencies. In 2003, I was asked to join a task force convened by Dr. 
Elias Zerhouni, NIH Director at that time. The group was charged with 
recommending new ways to fund high-risk, high-impact research. Our 
panel made three main recommendations, but I will focus on just one of 
those: establishing a new set of awards to researchers based on their 
track record. In fact, the journal Science covered our panel's 
recommendations in a news story headlined, ``NIH to Award People, Not 
Projects.'' That headline nicely summed up our recommendations. But in 
practice, the NIH came up short in carrying out this initiative.
    Take the NIH Director's Pioneer Awards, for example, which were 
aimed specifically at stimulating highly innovative research and 
supporting promising new investigators. Our task force recommended that 
the NIH award 10 percent of its R01 grants--which would equate to 
roughly 700 grants--on a ``people, not project'' basis. In 2004, the 
first year the awards were made, the NIH selected only nine Pioneer 
Award recipients from among approximately 1,000 nominations.
    It is somewhat more encouraging to see that this year NIH has 
awarded a total of 115 grants for high-risk, high-reward research 
through its Pioneer Awards, New Innovator Awards, and the NIH 
Director's Transformative R01 Awards. The total number of these types 
of awards, however, still falls far short of our 2003 recommendations.
    Even with these new awards, the NIH research budget is still 
heavily weighted toward project-oriented research, with 98 percent of 
grants going to projects. As I stated earlier, I strongly believe that 
giving money to scientists of exceptional and demonstrated creativity 
is a better way to promote innovation. In my opinion, even a modest 
shift in the federal research funding portfolio--going from 98 percent 
to 90 percent project-oriented--could make a big difference in 
producing innovative and potentially transformative research results.
    I would like to end with a quotation from the Nobel Prize winner 
Max Perutz, who directed the Medical Research Council Laboratory of 
Molecular Biology in England for more than 20 years: ``. . . 
(C)reativity in science, as in the arts, cannot be organized. It arises 
spontaneously from individual talent. Well-run laboratories can foster 
it, but hierarchical organization, inflexible, bureaucratic rules, and 
mounds of futile paperwork can kill it. Discoveries cannot be planned; 
they pop up, like Puck, in unexpected corners.''
    Thank you, Mr. Chairman. I would be pleased to answer any questions 
that the Committee might have.

                     Biography for Gerald M. Rubin

    A Vice President of the Howard Hughes Medical Institute (HHMI) 
since 2000, Gerald M. Rubin was named in 2003 the first Director of 
HHMI's Janelia Farm Research Campus. At Janelia, Rubin directs 
scientific programs designed to speed the development and application 
of new tools for transforming the study of biology and medicine. A 
760,000 square-foot biomedical research complex in Ashburn, Virginia, 
which opened in the summer of 2006, Janelia will eventually accommodate 
a research staff more than 300. It houses laboratories and provides 
short-term housing for visiting researchers, along with a conference 
center.
    Rubin served as HHMI's Vice President for biomedical research from 
2000 to 2002, when he was appointed Vice President and Director of 
Planning for Janelia Farm. Before moving to HHMI headquarters, Rubin 
was an HHMI investigator at the University of California, Berkeley, 
where he was the John D. MacArthur Professor of Genetics in the 
Department of Molecular and Cellular Biology. An internationally 
recognized geneticist, Rubin led the publicly funded effort to sequence 
the fruit fly Drosophila melanogaster genome. In addition, his 
laboratory has worked to determine the function of fruit fly genes that 
have homology to human genes and, more recently, to develop genetic 
tools to help probe the structure and function of the fruit fly brain.
    Rubin received his Bachelor's degree from the Massachusetts 
Institute of Technology and earned his Ph.D. in molecular biology from 
the University of Cambridge in England. He did postdoctoral work at the 
Stanford University School of Medicine before joining Harvard Medical 
School in 1977 as an assistant professor of biological chemistry. In 
1980 he joined the Carnegie Institution of Washington as a staff member 
in the department of embryology, and three years later moved to UC-
Berkeley. Rubin is a member of the National Academy of Sciences, the 
Institute of Medicine, and the American Academy of Arts and Sciences.

                               Discussion

    Chairman Lipinski. Thank you, Dr. Rubin. I thank all our 
witnesses for their testimony, and at this point, we are going 
to begin our first round of questions, and the Chair will 
recognize Mr. Tonko.
    Mr. Tonko. Thank you, Mr. Chair. Dr. Lane, you are probably 
as familiar as anyone in this room with the annual budget 
process and the challenges that accompany that. In these very 
difficult budget times, what advice can you give us as to 
defending the dedication of a pot of money in those tough times 
when there is a high probability of failure?
    Dr. Lane. Mr. Tonko, I thank you and appreciate the 
question. The ARISE report didn't answer that specific 
question, so I will give you my best take on it based in part 
on the discussions of the Academy Committee but also my own 
experience.
    I think there is a problem with terminology here. My sense 
is that even the highest-risk ideas proposed to our agencies by 
researchers who have strong reputations and track records are 
not likely to fail. It is quite possible that the particular 
goal that was put forward in the research proposal might not 
pan out. But in the meantime, young people are educated, often 
technologies are developed if this is an experimental process.
    So I think even though the ARISE report recommended in 
terms of the merit-review system focusing differently on high-
risk, high-reward research from the normal grant program, I 
think in both cases one should look at the potential for 
transforming the field, a major breakthrough, and at the same 
time, other outcomes, other aspects of the research that are 
almost assured to come out. I don't think we have to tell the 
public and your colleagues here on the Hill that we are going 
to spend more money on research that has a high probability of 
failing. I think the research that we invest in here is likely 
to seed on many dimensions, even though it might not actually 
reach the conclusion that the researcher hopes for or create 
the device that the researcher wants.
    An example might be the LIGO project, Laser Interferometer 
Gravitational Wave Observatory at Caltech. Significant funding 
hasn't yet seen a gravitational wave as far as I know, but out 
of that have come spin-off companies, new technologies, lots of 
educated young people. So I think we need to talk about what is 
a better way to articulate what we are doing here. I think 
there is a good story to be told, and maybe we haven't worked 
hard enough on doing that.
    Mr. Tonko. Might there be a way for some of these projects 
to prove themselves before they become a target? Is there----
    Dr. Lane. I think the issue is--one of the issues is one of 
funding. Sometimes it is rather expensive to try this idea. You 
can't really do it on the cheap. So we proposed seed funding, 
for example, to allow an investigator with a wild idea, let us 
say, to explore, enough data, enough experience that he can 
convince the peer reviewers that this has a higher probability 
of success. I think that is what you can do with seed funding. 
That is what universities could do more of if they had the 
resources to devote in that way.
    Mr. Tonko. Dr. McCullough, you have talked of the failure 
of some of these proposals or at least the process that relates 
to these proposals at NSF. But it is colleagues from the 
research universities that are oftentimes the reviewers serving 
on these panels. Is there any sort of input or encouragement 
that could be provided to these colleagues because they are 
scoring these given situations? And you know, again, they are 
impacted by these tough times, economic times. Is there some 
sort of encouragement or training that can be provided to the 
colleagues?
    Dr. McCullough. That is one of the reasons I think it would 
be good to have a set-aside and direct instructions from the 
agencies that we want to fund things that are a big more high-
risk, high-reward if you will, proposals with sort of specific 
guidelines. I think the way it happens now is high-risk, high-
reward research is often sort of a check box that means that 
the proposal should not be funded because it has not been 
proven yet. There is no proof of concept for those proposals. 
And so I think having sort of a mechanism by which the panels, 
chair people, are guided, outside reviewers are guided in a 
special program and not leaving it up to the unfortunate 
program officers who are really struggling to try to find the 
proposals that are very, very highly rated already.
    Mr. Tonko. So in a sense it is a cultural change that we 
need to incorporate into the review process? Is it just perhaps 
requiring that a certain bid of high-risk projects be looked, 
at maybe taking the top threshold of those projects?
    Dr. McCullough. In my opinion, my personal opinion, yes, 
because if you look at the testimony that I have written, you 
will see examples of five or six projects that we have at the 
University that are just unbelievably spectacular . . . that we 
really struggle to try to fund them internally, through 
foundations, providing those seed funds. And as I was saying 
earlier in my spoken testimony, I think providing funds as you 
suggested where, as Dr. Lane suggested, we can provide seed 
funding so people can get the projects off the ground. And then 
if they don't show proof of concept, then they can't maybe get 
follow-on funds.
    So it is less of a risk as you were indicating if we had 
some sort of limit in the initial funds, but then the ability 
to open a gate for them to bring these things and deliver on 
the high-reward transformative nature of the research programs.
    Mr. Tonko. Mr. Chair, if I might just ask one more 
question. The relationship with venture capitalists--I am often 
told that venture capitalists will walk away from some of these 
high-risk situations, but is there a way, is there a threshold 
of involvement from the public sector that might be 
incorporated with a venture capitalist funding that might 
maximize what we could do here?
    Dr. McCullough. I think that one of the things we are doing 
at Carnegie Mellon--I am directing something called an 
innovation ecosystem, and what we are trying to do is create 
funds within the University so that once these projects get to 
the research phase, that we can find funding mechanisms to 
delivering--now I am using proof of concept in a different way 
now--but proof of concept for commercial situations. And so to 
create pre-commercial research and take it to the proof of 
concept for commercial reasons, there are funds like that that 
are not available, that we are trying to raise those privately, 
work with venture capital funds, angels, et cetera, to try to 
keep things from approaching the ``Valley of Death'' and these 
concepts so that we can have higher probability for these 
research ideas to end up creating jobs, and help us to be 
leaders in technology innovation. I think that is an absolutely 
critical thing because as a person who has a start-up company, 
I have lived through this, and these are the kinds of funds 
that we are trying to do within our University and work with 
other agencies and foundations to try to create.
    Mr. Tonko. Do any of the other three of our witnesses have 
anything they want to say on that partnership that can be had 
with venture capitalists in the high-risk area? Yes, Dr. 
Collins.
    Dr. Collins. Well, let me make two points, Mr. Tonko. 
First, with respect to the notion of failure of a project, I 
would build on Dr. Lane's remarks, and actually something that 
you said, in that it would be a mistake to focus on an 
individual project and think about it solely as succeeding or 
failing, and that's the reason why the United States invests in 
the basic research enterprise.
    For me, the United States invests in a basic research 
enterprise in order to sustain what is an innovation ecosystem 
as far as the United States is concerned. Some projects will 
succeed, some projects will fail. But the point is that you 
have individuals constantly trying to think about where the 
next steps are with respect to the whole process of discovery, 
and I think a great example of this we saw recently with the 
Netflix competition, which had to do with a better algorithm 
for picking movies as far as the Netflix Corporation was 
concerned. Interesting thing about the analysis of that that 
appeared in The New York Times was that there was a prize for 
the winner, and they were of course quite happy about that. But 
when the second-place individual was interviewed, that company, 
they said we gained as much by participating in this 
competition, even though we did not win, because the very fact 
of trying to think through these problems affected the culture 
of our institution.
    It seems to me that is what you can do with Federal 
research dollars, is you can give our institutions the freedom 
to take those kinds of risks. That is the reason we do it. We 
fund institutions in order that we can do this kind of work. So 
your phrase was exactly right, it is a cultural issue. It is a 
cultural issue within our universities and within our research 
institutes, and it is a cultural issue within a funding agency 
like the National Science Foundation. You are absolutely right 
again in that what we do is we direct the panels, we direct the 
program officers, we direct our senior managers to discuss with 
the faculty members, with the reviewers, what it means to think 
about transformative research and risky research, and we ask 
them to take that into consideration when they look at these 
proposals coming in. It seems to me that is where you put your 
finger on the issue and fostering the combination of the 
culture and institutions that are open to these kinds of 
issues.
    Mr. Tonko. Thank you.
    Dr. Lane. May I just comment, Dr. Tonko, to Dr. Collins. I 
think it is important I think to make the point that high-risk, 
high-payoff research is not always just basic research. It can 
certainly be research that has some toward directed hoped-for 
goal, very much on the practical side. I mean, the transistor 
came because there was an effort at Bell Labs to find a 
replacement for a vacuum tube triode. Well, the rest is 
history. So there is great opportunity, I think, for agencies 
like the NSF to work with the private sector, largely through 
the partnership with the universities as it does with its 
Engineering Research Centers, Industry/University Cooperative 
Research Centers, the SBIR grants that are made directly to 
industry. One might have a look at those mechanisms, some of 
which have longer-term time horizons on them to see whether 
there is not an opportunity to do more to address the high-
risk, high-payoff goals that we are talking about today through 
some of those mechanisms.
    Mr. Tonko. Thank you.
    Chairman Lipinski. Thank you, Mr. Tonko, for your 
questions. I know it did go on a while, but I think those were 
all very interesting questions, and we are getting a little off 
the--a little far afield, to some extent, talking about 
bringing private funding, but I think it is a critical question 
and because this is something that I am interested in a broader 
sense of how we best do these public/private partnerships 
essentially in research and development. I would like to hear 
what was said there also.
    So with that, I will now recognize Dr. Ehlers for five 
minutes or however long. He always has plenty of good 
questions, so I will let him go ahead.
    Mr. Ehlers. I better be a little careful. My dad was a 
preacher, so we could be here quite a while.
    Chairman Lipinski. Okay, not that long.
    Mr. Ehlers. Not that long. Okay. I was interested in the 
comments about people, not projects, and I would like to pursue 
that with each of you. I happen personally to think that is in 
general a good idea because I recall in my days in university 
life, it wasn't too hard to pick out the really outstanding and 
bright young people. But I also learned that they don't 
necessarily make the best investigators. There is a big 
difference between thinking of an idea and carrying it out. And 
so I guess I am asking for comments from all of you because I 
think the concept is good. I wonder how it can be executed well 
in practice.
    I will just give you one example that could be a problem. 
There are others. A junior researcher just starting out doesn't 
have much of a track record. He may appear to be very bright, 
but you are really not sure. How do you evaluate that? I happen 
to know one who is really quite bright and has done very well 
but had a terrible time getting his first NSF grant because he 
wasn't well-known, even to the extent that an idea he had which 
he had discussed with another researcher at another university, 
that person picked up on it and submitted the proposal and got 
it funded, whereas the young person who thought of the idea 
submitted the proposal which didn't get funded.
    So I would just like an open discussion from all of you, 
and Dr. Rubin, I will let you start because you mentioned the 
people, not projects first. But how does this work out in 
practice?
    Dr. Rubin. I think you raised a couple of good issues here, 
so I would say at the beginning, I don't think you could 
convert the entire research enterprise to a people not 
projects, and I certainly wouldn't suggest that. As I said in 
my verbal remarks, I am talking of going from two percent to 
ten percent of the grants awarded, recognizing people based on 
their track record for innovation and success.
    So this does create a problem for people who are just 
starting, but our experience at the Howard Hughes Medical 
Institute, we have no trouble identifying people who had an 
independent, say, faculty position at a university for four or 
five years. At that point it is pretty clear who the 
innovative, creative individuals are, and we believe that past 
performance is a better indictor of future success than any 
written proposal will ever be, and that people who have an 
ability and desire to be innovative and are willing to go in 
uncharted territory, that is a personality trait which carries 
over. And you can identify such people, and at least in our 
experience, we do much better by saying to someone, we are 
going to give you generous funding for your research for five 
years. You better do something with it because in five years we 
are going to look over what you did, and if we don't like it, 
we are not going to give you any more money. But we are not 
going to tell you what to do with the money. We are betting on 
you as an individual, and we are going to win or lose our bet.
    I would say that our experiment that we have done could 
make a very good case that betting on those--placing your bets 
in that way gives you a higher rate of success than placing 
your bets by reading a stack of research proposals, because in 
research proposals, you often reward the people who are very 
articulate and can write very good research proposals, rather 
than the people who have the good ideas or are going to be able 
to execute if you give them the money.
    So it is just an alternative, and I think a portfolio, a 
diversified way of funding research is always better than 
putting all your eggs in one basket. I just think that we are 
out of balance now in the way we fund research in this country.
    Mr. Ehlers. Let me just get back to you on the one issue 
there. I don't think this would work at the NSF which does in 
fact review stacks of proposals because they really--I don't 
think either the program officers or the panel of reviewers 
generally don't know the applicant that well. I take it that 
you, at HHMI, is that right?
    Dr. Rubin. Yeah.
    Mr. Ehlers. Really get to know these people well and bet on 
them because you have investigated them and worked with them or 
talked to them enough that you are quite convinced that they 
really are above the pack.
    Dr. Rubin. Well, I think the way they send in applications 
initially to be appointed, and--they are reviewed in a way not 
dissimilar from typical peer review. I would say--let me give 
you an example just from my NSF colleagues.
    NSF has an award every year called the Waterman Award which 
is supposed to give money to the most creative, best scientists 
under the age of 35 or something like this. And I am sure they 
have 20 or 30 very good nominees for that. I would think the 
NSF--a good use of the NSF budget would be to say these 20 
people who are nominated for this award are all outstanding. 
Let us just give them each $10 million to fund their--whatever 
they want to do in science for the next five years instead of 
reviewing a lot of little grants. I think they would get more 
output in research dollars. I mean, I am making these numbers 
off the top, but they already have a mechanism in place to do 
that. It would require very little extra person power to 
implement a policy like that.
    Mr. Ehlers. Are you basically suggesting the McArthur 
approach?
    Dr. Rubin. Something along that ilk, I think to a certain 
amount. I mean, this approach has limitations, but I do think 
that the most creative innovative ideas come from when you give 
innovative, creative people some money and you don't try to 
tell them what they should do about it because by definition, 
an innovative, transformative idea--if someone can write down 
in their proposal and submit it to you, it is not an 
innovative, transformative idea, almost by definition, by my 
definition of those terms.
    Mr. Ehlers. Okay. I want to get back to that in a minute 
but first Dr. McCullough, you have been smiling broadly.
    Dr. McCullough. Well, I mean, there is certainly room for 
these sorts of programs where HHMI and Waterman Award winners, 
who are the rarified group of people at the top who get funded, 
probably are amazing. We have a McArthur genius at our place, 
and he is just out of this world, you know, an idea all of the 
time. But you know, for those of us who went to community 
college and you know, peaked later in life and you know, some 
of the examples of colleagues at my university, one in 
particular I think of, who came from Poland and after he became 
a professor really, although it is not necessarily the rule, 
but is now the most-cited chemist in the world and he is often 
mentioned as a Nobel Prize candidate.
    And I know that Dr. Rubin is not suggesting that we change 
the whole program to bet on horses. In the DoD world they often 
do this in terms of funding. They will find someone who they 
like and they can work with, and they will bet on that 
professor and they will fund that professor. And they get often 
locked into the system. I think that there are pitfalls with 
that approach, and one has to be worried about funding people.
    You know, there is also the aspect of beyond the proposal. 
You know, you meet people at conferences, they hear you talk, 
there is access, and especially for young researchers, I think 
it is sometimes very difficult to see who is going to be the 
greatest innovator. So I think that there is a role. I am not 
disagreeing with Dr. Rubin, but I also believe that it is 
important for those who peak later in life to--it doesn't mean 
that they are not going to be innovative. I do agree with the 
ARISE report that the most creative ideas often come from 
brand-new faculty members who are really thinking out of the 
box. So I think that is an area of concern.
    Mr. Ehlers. Yeah, I am still waiting for myself to reach my 
peak. I have to admire the creativity of some of my colleagues, 
particularly when they have done something wrong and they are 
explaining it to the press. I think that is a different sort of 
creativity.
    Dr. Collins, would you like to offer a few comments from 
the NSF perspective?
    Dr. Collins. I would. Thank you, Congressman. A couple of 
thoughts. I would agree with you that there is a difference 
between just being able to think up an idea and be able to 
carry out that idea. It suggests that at times you do want to 
be taking chances as far as individuals are concerned, and I 
will come back to your point also about young investigators, 
and we can tie those two things together across.
    For example, the program managers at the National Science 
Foundation, where, as I indicated in my oral testimony and 
opened up in the written testimony, these individuals not only 
manage awards but they work and mentor post-docs, they 
facilitate connections, they engage in this outreach. They 
really do work with the individuals. We summarize that under 
this single word of ``stewardship.'' It is the notion that 
these program officers are not only just processing paper, but 
they really are deeply engaged in the scientific process itself 
in an ongoing conversation. And this is the heart of the 
enterprise, and it is what we need to continue to foster in 
terms of something that is being challenged right now with the 
terrific workload that is coming as far as NSF is concerned, 
and I might mention NIH as well.
    So you do want to evaluate individuals and indeed that is 
what happens when we do take chances on young people. So for 
example, the career program which you know and funds on average 
about 450 individuals a year--this year, because of the ARRA 
funding that was able to go up to 700 individuals. And those 
then enter into a relationship with our program officers, where 
they work with these young investigators as well as they work 
with senior investigators in terms of stewardship.
    Now, the senior investigators also, and young investigators 
can also call out their program officer and say I do have this 
great idea. I have this great idea at 3:00 in the morning, and 
we now have the eager mechanism where that individual can get 
up to $300,000 for a couple years to begin to pursue that idea 
and get the preliminary data that is needed in order to move 
that idea along to a proposal.
    Furthermore, program officers once again have the 
prerogative for a successful program or for a program that is 
moving along. They can call out that investigator and discuss a 
creativity extension where there is minimal application needed 
in order to continue that funding for another couple of years 
or three years.
    So there are a variety of mechanisms that are in the hands 
of the program officers, and it goes back to my point earlier 
on having a culture of an institution that is willing to make 
these kinds of engagements and these kinds of investments over 
time, so that you can use past performance as an indicator, but 
you can also work with young people in order to take the risks 
that are needed to begin to build up the infrastructure within 
the country. I was taken by this this week with the Nobel 
Prizes. NSF molecular and cellular biology had funded two of 
the individuals who received Nobel Prizes this week, and they 
funded them early in their careers. It means that the program 
officers 20 years ago, 25 years ago, were prescient enough to 
look at that group of applicants and say, this is someone to 
fund. Twenty-five, thirty years later, you see the fruits of 
that labor and you also see it as a result of our relationship 
between the funding agencies in the United States, where the 
NSF got these individual started, and in one instance is still 
funding that individual. But the individual receives support 
from other kinds of funding agencies.
    So it speaks to the need for diversity of institutions, not 
only diversity of approaches, that give you really breadth of 
support for an innovation enterprise as far as the country is 
concerned.
    Mr. Ehlers. Okay. I think we will have to cut this off 
here. I have used too much time already. We may get back to you 
later, Neal.
    Just a comment. I forget who mentioned accelerator funds. I 
think that is a great idea, Dr. McCullough. When you find 
something, and I have been in that situation where you are 
doing some research and you find something really great, and 
you want to pursue that, it is nice to have a mechanism to do 
that.
    The other comment I want to make is I don't like the term 
transformative research. Obviously high-risk doesn't go too 
well, it doesn't survive the Proxmire Golden Fleece award 
requirements where, you know, you talk to the public about 
doing high-risk research, that doesn't really mesh. But 
transformative doesn't mean too much to the public, either. I 
might suggest you come up with something like NASCAR research 
because the point of NASCAR is you take some very high risks in 
hopes of winning. That is exactly what you are doing here. So 
you are taking the NASCAR approach. Let us fund the stuff that 
we think is really going to payoff, and we know there is some 
risk attached to it, just as there is to driving a car at 200 
miles an hour. But the rewards can be great. And the public can 
clearly identify with that one.
    With that, I yield back. Thank you, Mr. Chairman.
    Chairman Lipinski. Thank you. Dr. Ehlers, I am thinking 
about that NASCAR award and thinking about if somehow we could 
have crashes and other ways to somehow interest the general 
public while this research is being done, then maybe that could 
work. Colorful cars and other things like that. We have to keep 
people interested in the race while it is going on, not just 
the final.
    Mr. Ehlers. Perhaps we can ask them to contribute to down 
payments for the first trip to a dark hole. That is a pretty 
safe bet actually.
    Chairman Lipinski. The Chair now recognizes himself for 
five minutes, and I have a lot of questions. I am going to try 
to limit it a little bit, but if we can try to limit our 
responses a little might be good. I am glad that Dr. Ehlers 
went down the road of the--Dr. Rubin had talked about the 
people, not projects, and we talked a little bit about that. I 
know a couple years ago Secretary Chu, at that time at Berkeley 
Lab, that was one thing that he really--a couple years ago I 
had lunch with him out there, and the one thing that he 
impressed upon me that that is something that he really thought 
was a good way to go.
    I sit here as a former political scientist, maybe I still 
am a political scientist, but I wonder, does this work better--
no one may have any strong opinions on this or thoughts, but I 
sit here thinking, does this work better for some disciplines 
than others in some different scientific areas, disciplines? 
Would this not work as well, do you not see as much 
possibilities or does this question not really pertain? Does 
anyone have any thoughts on that? Dr. Lane.
    Dr. Lane. I don't have any wisdom on it, I have thought. It 
connects a little bit with my earlier observation that I think 
high-risk--you already said in your opening statement, Mr. 
Chairman--there is still a remaining question about what is 
this that we are talking about, high-risk, high-reward 
research, and I wanted to make the point that it doesn't just 
have to be basic research, it can also be research that is 
being done with some particular outcome in mind. I think that 
if we go field by field, and my colleagues can help me in many 
of these fields, but my sense is this question is apt in all 
fields that I know anything about for a couple of reasons. One 
reason is that breakthroughs often occur as total surprises so 
the research that was being done maybe strikes us as somewhat 
routine, somewhat dull, whoever is funding it, and suddenly 
there is a funny blip on the screen, there is a number that is 
unexpected, there is some surprise that comes out of the 
research. And the ability to pursue those surprises, the 
flexibility, the freedom of the investigator to pursue such a 
surprise, it is extremely important. First all, it has to be a 
kind of person who is so curious and so driven that he or she 
wants to do that, but then the environment, the funding, the 
organization, the institution has to be willing to go with 
that. Bell Labs is a really good example, so you know, not too 
long ago they had seven Nobel Prizes to their alumni. It must 
be ten or a dozen or something like that now. There was an 
environment in which funding was available. People were being 
bet on, and when they found surprises, they were able to pursue 
them, including the observation of background blackbody 
radiation from the origin, from the first big bang in the 
universe.
    So I would think that, in any field that one could think 
of, there is an opportunity for these kind of breakthroughs. 
And therefore, the question is at least apt whether one is 
properly addressing the issue of high-risk, high-reward 
research.
    Chairman Lipinski. Anyone else? Dr. Rubin.
    Dr. Rubin. Well, I agree that in any field you need a range 
of research projects. So in, say, biology, one of our most 
successful projects recently was the Human Genome Project, 
which was a very well-defined project with a clear goal and 
could be measured. I wouldn't call that--it was only high-
reward, it wasn't high-risk. On the other hand, you have other 
projects which are higher risk, or more unpredictable is a word 
I would prefer. But I just want to emphasize what Dr. Lane said 
about flexibility in being able to alter the goals that you are 
working on to pursue a new--take advantage of some unexpected 
result. I don't think that, the way the funding agencies work 
now. A lot of it is just the problem of peer review, are very 
good at rewarding people for not pursuing what they originally 
said they were going to do but to take advantage of something 
much more interesting or important that came up as an 
unexpected result within that.
    So anything that we can do to change the culture to support 
individuals having more freedom, I think would be a good thing.
    Chairman Lipinski. Dr. Collins.
    Dr. Collins. Mr. Chairman, I would agree with Dr. Lane. I 
don't think this is a discipline-limited issue, and in fact, I 
wouldn't even limit it to the sciences. I think there are high-
risk activities in the humanities and the arts as well where 
individuals take chances and they take risk, and sometimes it 
works and sometimes it doesn't. Breakthroughs indeed do come as 
a result of these surprises that are present. They are inherent 
in the research enterprise itself. After all, that is why we 
call it research and we don't call it demonstration. It is 
something that we are involved in as a process of discovery.
    Chairman Lipinski. One thing I would go back to. I am not 
sure who had made the comment. I don't know if it was Dr. 
McCullough. My experience, and again we are talking social 
sciences, political sciences, was you don't propose anything 
that you don't already know the answer for. And so that sort of 
goes counter to the research that is sort of wide. It is still 
possible to discover other things, but that narrows it down 
much, much more. And I tend to think we are not going to stay 
here all afternoon and the next few weeks talking here, so I am 
not going to open this up, but the culture issues are certainly 
also very important. But please don't start talking about that 
because we will be here forever. But was there something else 
you wanted to----
    Dr. Collins. I would agree with you. I think the culture is 
important, and that is why on this issue of surprise and 
breakthroughs, program officers have the flexibility to work 
with the investigator at that point. That is why we give 
grants, we don't do contracts. And it is that flexibility that 
is inherent in the process as far as NSF is concerned.
    Chairman Lipinski. One thing, shifting gears here, my 
colleague, Bob Inglis, and I had sponsored the H Prize. We had 
legislation for the H Prize. We eventually worked, got that 
into the energy bill that we passed a couple years ago. The 
idea of prizes, I just wanted to throw that out there. That is 
one way to avoid the political issues associated with high rate 
of failures. You know, you put out a prize worth something 
specific. The H Prize was for advances in use of hydrogen for 
transportation. But you also have examples of prizes promoting 
basic research. For example $1 million in Millennium Prizes 
offered by Clay Mathematics Institute.
    So what extent do you think that prizes can motivate, 
transform into research? Dr. McCullough.
    Dr. McCullough. I think that they do play a major role. If 
I can point to the DARPA Urban Grand Challenge as a situation 
where they asked for autonomous vehicles to drive within an 
urban situation. That was a challenge, and universities came 
together to accomplish this. Of course, I bring it up because 
we won at Carnegie Mellon, and there was $1 million prize. 
Carnegie Mellon and private groups and companies invested in 
this, and much technology came out of these things to create 
these autonomous vehicles. There is nothing like a good 
competition to get people's juices flowing and actually create 
something and bring teams of people together that maybe 
normally don't work.
    So I think it is a very interesting idea, and I think it 
certainly plays a role like many other funding mechanisms, but 
I think it is a very interesting one.
    Chairman Lipinski. Anyone else?
    Dr. Lane. I would just add to that, I completely agree that 
the prizes can be enormously stimulating and have the 
advantages you just described, but you sort of have to know 
what the goal is so you can decide who won, and if it is a 
high-reward research that is going to show its worth in 20 
years or 15 years, then it is a somewhat different category. So 
I think there are several ways to stimulate people to be 
thinking, to be taking risks, with the possibility of great 
payoffs. But there are many dimensions to that issue, and a 
prize certainly would be a very important one.
    Chairman Lipinski. Thank you. The Chair will now 
recognize--Mr. Carnahan, do you have questions? Mr. Carnahan 
for five minutes.
    Mr. Carnahan. Thank you, Mr. Chairman. And again, I thank 
the panel. I apologize I had to come in late. I was in some 
other meetings, but I had a chance to look at some of the 
written testimony and wanted to particularly ask Dr. Collins. 
You cited an article by Gary Anthes who stated, ``The kind of 
pure research that led to the invention of the transistor and 
the Internet has steadily declined as companies bow to the 
pressure for quarterly and annual results.'' Well, during this 
year's energy and water appropriations, I, along with several 
other colleagues, passed appropriations to fund Secretary Chu's 
request for energy innovation hubs. Three were funded out of 
the eight requested. These hubs as you know are modeled after 
the research labs, involved in the Manhattan Project Labs, 
Lincoln Labs at MIT and AT&T/Bell Labs that developed the 
transistor. Each hub envisioned would embrace within these 
topical areas the goals of both understanding and use without 
erecting barriers between basic and applied research. It will 
seek advances in highly promising areas of energy, science, and 
technology and will result in many solutions being deployed 
into the marketplace.
    Today's hearing is focused on suggestions moving forward 
for funding high-risk, high-reward research. Would you qualify 
these energy innovation hubs? It is models that other agencies 
should employ? And are these good models for government 
agencies to fund but not necessarily institutions like NSF? Dr. 
Collins?
    Dr. Collins. Continuing on from really the last 
conversation and the last question, the real key, it seems to 
me as far as innovation is concerned, and especially within the 
structure of the Federal agencies is to have a diversity of 
different kinds of approaches across the different agencies. In 
fact, to go back to the point about competition, that in and of 
itself can be affected. But even within an institution, to have 
different ways in which one would go about this whole process 
of discovery, whether you are using individual applications, 
whether you are looking at groups, whether you are using 
centers at some times, the Science and Technology Centers, for 
example, as far as the NSF is concerned.
    The thing you have to be careful about is this intersection 
between basic and applied. You have to know what you're going 
for, especially if you are going to use something like a prize, 
and the danger here in terms of basic research is to have a 
metric that is too short. So basic research sometimes takes 
quite a bit of time. Manhattan Project can be a pretty good 
example for some kinds of things. It is a pretty 
straightforward engineering solution that you're looking for. 
But if the path forward were perfectly clear, we wouldn't call 
it basic research. We wouldn't need to have that time that's 
needed in order to discover the fundamental issues that are at 
work in order to then come up with the application. So it's 
really this mix of things that is needed both within and in 
between institutions in order to stay ahead.
    Mr. Carnahan. And let me ask Dr. Lane if you would comment 
as well.
    Dr. Lane. Thank you very much, Mr. Carnahan, for the 
question. The first thing I should say is this particular issue 
was not addressed in the ARISE report, so I don't want to 
answer with the thought that I am reflecting on--however Steve 
Chu, or Secretary Chu, was a member of our Committee, so he 
probably had it in his mind at that time.
    Just adding to what Dr. Collins said about diversity, the 
Department of Energy of course has a rich experience in not 
only funding university research, high-quality university 
research, but also laboratories, national laboratories. And the 
advantage of a laboratory, whether it is one of these new hubs 
or the existing national labs, is you have a cadre of talented 
people there who can fairly quickly, if needed, work together 
in different ways to address a major national need, like 
energy, for example. And I think--I don't know what Secretary 
Chu has specifically in mind here, but they look a little bit 
like little Bell Laboratories, having that ability to focus on 
the quality of people, qualify an innovative idea or move 
rapidly, not be judged too much on short-term timelines and 
such matters. And so that is a very thoughtful concept. Then, 
it seems to me, it is appropriate for the Department of Energy. 
It is not the way academic research works in universities. It 
is not so easy to quickly put together large teams of 
researchers around this single goal. So I think the diversity 
issue Dr. Collins spoke to is the right way to think about it, 
and I personally am very pleased to hear your support for 
Secretary Chu's ideas. But it was not something we addressed in 
the ARISE report.
    Mr. Carnahan. Any others? Comments?
    Dr. McCullough.I would just give you one ancillary effect 
of these innovation hubs which is really very positive. Having 
these very large programs, what happens is groups of people 
start teaming together across universities and companies and 
national labs and forming groups that would have never be 
formed in any other situation, and often there is a great 
benefit for these teams to be formed and trying to chase after 
this sort of money that's around the innovation hubs. But there 
are great things that come out of that because people who would 
not normally come together find each other and start to 
collaborate and find other sources of going after funds.
    So these kinds of programs and the diversity of these 
programs are really important to get people's attention, to 
bring them together. So I think there's a great effect beyond 
what you will see just out of the program. You will see other 
groups that will be formed that will not be funded, and great 
things will come from them as well.
    Mr. Carnahan. Thank you, Doctor. Thank you very much, Mr. 
Chairman.
    Chairman Lipinski. Thank you, Mr. Carnahan. And before we 
close, I recognize Dr. Ehlers for closing comments.
    Mr. Ehlers. Thank you, Mr. Chairman. I really want to thank 
you, Mr. Chairman, for an excellent panel, good cross-section 
of people to deal with this topic. And I have learned a great 
deal here, so I thank you for coming and thank you for your 
comments and your answers.
    One other comment, when I talked about the NASCAR award, I 
was not suggesting, although maybe it had seemed that way to 
you--one thing I have become very sensitive to as a scientist 
in Congress and that is the scientific community should be very 
careful about how they say things, and I recall the time I had 
to dash to the floor because one of my colleagues got up and 
offered an amendment to cut the budget of the National Science 
Foundation because they were going to fund gain theory research 
and ATMs. And that is what was in the bill. That is what came 
up. And I dashed down there just to point out that gain theory 
was not what they thought it was. It was a very important part 
of theoretical physics and also the ATM that they were 
ridiculing, my colleagues said, the banks use ATMs. Let them 
pay for the research. And I said, I am sorry but ATM stands for 
a-synchronous transfer mode, and we need some research on that 
so you can make the internet better. So NASCAR is not a bad 
idea. I am not really going to go to bat for it or publish it 
but my point is simply transformative sounds like gobbledygook. 
High-risk does pass through a Senator Proxmire test. So see if 
you can come up with a better term. And I won't patent the 
NASCAR returns. So thank you very much.
    Chairman Lipinski. Thank you, Dr. Ehlers. I think it is 
just the high-reward part that we could work on some other 
names for it. I want to thank the witnesses for testifying 
today, and certainly this was a very interesting topic, not one 
that I think has a really wide appeal here. But it is 
critically important, and how we best do this. We know how to 
keep the United States at the forefront of technology. We have 
to be doing this research and have to be doing this research 
that is whatever you want to call it, that we get the high 
rewards. I think that sounds very critical. And as we move into 
early next year, this Subcommittee will be working on writing 
the NSF reauthorization. This is something that we will be 
looking closely at, and as the Full Committee works on America 
COMPETES through next year also, this will certainly be a part 
of what we are working on doing.
    So again, I want to thank all the witnesses for their 
testimony. The record will remain open for two weeks for 
additional statements from the Members and for answers to any 
follow-up questions the Committee may ask of the witnesses. 
With that, the witnesses are excused, and the hearing is now 
adjourned.
    [Whereupon, at 2:29 p.m., the Subcommittee was adjourned.]

                              Appendix 1:

                              ----------                              


                   Answers to Post-Hearing Questions


Responses by Neal F. Lane, Malcolm Gillis University Professor and 
        Senior Fellow, James A. Baker III Institute for Public Policy, 
        Rice University

Questions submitted by Representative Vernon J. Ehlers

Q1.  Do you have any recommendations on how to modify the peer review 
process--without changing all the things about it that currently work 
well--that would help reveal which investigators are truly attempting 
high-reward research? In other words, is there a better way for grant 
committees to ``get to know'' investigators?

A1. This is a critically important question and I offer the following 
observations in response, based largely on the deliberations and 
recommendations of the Academy's ARISE Committee. My comments focus on 
three primary elements of the process: the quality of outside 
reviewers; the criteria agencies use to evaluate potentially 
transformative research proposals; and the resources made available to 
agencies' professional program staff. Each of these factors has a 
decisive impact of agencies' ability to recognize and provide 
appropriate support for worthy applicants pursuing potentially 
transformative research.

Quality of Outside Reviewers

    Strong reviewers are a lynchpin of a successful, high-integrity 
peer review system. The best possible reviewers are attracted to 
participate if they perceive the process to be well-managed.
    The peer review systems operated by the National Science Foundation 
and the National Institutes of Health have long been considered to be 
the gold standard of competitive research award systems. However, as 
noted in the ARISE report, the rapid increase in applications to these 
agencies has placed serious strains on their peer review systems. As 
the workloads of reviewers, program officers, and staff grew, NIH's own 
analysis documented an erosion of quality resulting from a loss of 
continuity in panel membership and rapid turnover of program officers 
and reviewers. Similarly, more than one-third of NSF reviewers reported 
``great'' or ``somewhat'' decreased attention to each proposal.
    A high-quality, well-organized review process attracts gifted 
reviewers. Potential reviewers will be discouraged from participating 
if they have reason to believe that their time will be wasted--whether 
from personal experience or based on the experiences of colleagues.
    We recommended the following steps to strengthen the application 
and review processes:

          Require recipients of multiple grants from an agency 
        to serve as reviewers.

          Achieve greater continuity in reviewers.

          Establish interdisciplinary review panels to consider 
        high-risk research proposals across programs and fields.

          Consider alternative ways to select and mentor 
        reviewers.

          Consider dividing applications of more senior 
        researchers from new investigators and form separate review 
        panels with separate quotas.

Evaluation Criteria

    The ARISE Committee suggested steps for promoting the prospects of 
investigators pursuing high-risk, high-reward research, and also noted 
a number of ways in which funding agencies sometimes inadvertently 
discourage such research. Recognizing the inherent uncertainties and 
additional time required for such research, the Committee offered the 
following recommendations for adjustments to peer review systems 
specifically to foster high-risk, high-reward research:

          Applications should be relatively short and focused 
        on the qualifications of the researcher, an explanation of the 
        potentially transformative nature of the research, and an 
        explanation of why the researcher believes the proposed 
        approach could succeed.

          The proposal and the review process should place a 
        premium on innovation and reviewers should be charged to 
        identify new ideas, innovation, and creativity. Require 
        applicants to address the following question about their 
        proposed research: ``If this works, what long-term scientific 
        difference will it make?'' Evaluate proposals based on this 
        criterion.

          Agencies should not reject proposals solely on the 
        grounds that the proposed work is ``overly ambitious.''

          Fast-track seed money to evaluate a novel idea should 
        be made available.

          Agencies should be open to providing longer funding 
        periods for those proposals that require it.

          Recognize in grant-reporting requirements the value 
        of fortuitous findings not related to the main objective of the 
        research proposal and give program administrators the 
        flexibility and expectation to provide extra resources or time 
        to research unexpected but promising developments.

          For grant renewals or new grants on the same topic, 
        restrict the number of submitted publications and require a 
        self-assessment of each cited publication's impact.

          Evaluate renewals for first awards for high-risk, 
        high-reward research on the basis of project execution and 
        potential scientific impact, not on deliverables. Resist fine-
        grain assessments of whether a project ``worked''; expect some 
        hypotheses to fail.

Support for Professional Staff

    The ARISE Committee focused particular attention on the 
indispensable role of program officers in creating and maintaining the 
vitality and productivity of the research enterprise. Program officers 
manage millions of taxpayer dollars; their careers and opportunities 
for participation and leadership in their professional communities must 
be strengthened. The entire research system will greatly benefit if 
program officers are given greater opportunities to exercise leadership 
within the professional communities they fund and for whom they are 
responsible.
    If agencies and departments improve the professional opportunities 
of their research program officers, several benefits will follow. 
Program leadership will be strengthened, and career satisfaction will 
be improved. New ideas will be injected into agency and community 
deliberations. Researchers and program managers will be challenged in 
creative, timely, and innovative ways. Mutual understanding and 
communication will be strengthened. Counterproductive misperceptions 
will be identified more quickly. The return on investment of taxpayer 
dollars will be enhanced.
    Just as the program officers need to stay current on the latest 
developments in science and engineering research, the research 
community needs to know and respect these professionals, who have such 
large responsibilities for the quality of U.S. science and engineering.
    The ARISE Committee recommended the following steps to strengthen 
the system:

          Administrative budgets should keep pace with research 
        budgets.

          Program officers should be leaders not only within 
        their agencies but within their external scientific communities 
        as well.

          Program officers should be encouraged to attend 
        professional meetings and to visit institutions and 
        laboratories funded by programs for which they are responsible, 
        and agencies should make the resources available for them to do 
        so.

          Many university faculty members serve at NSF as 
        temporary program officers, or ``rotators,'' while on leave 
        from their university. They provide essential service and 
        leadership for NSF's research programs. This practice should be 
        encouraged and program funds should be allocated for this 
        purpose at other agencies as well.

    Again, I deeply appreciate the opportunity to bring these issues to 
the attention of the Committee.
                   Answers to Post-Hearing Questions
Responses by James P. Collins, Assistant Director, Directorate for 
        Biological Sciences, National Science Foundation

Questions submitted by Representative Vernon J. Ehlers

Q1.  Do you have any recommendations on how to modify the peer review 
process--without changing all the things about it that currently work 
well--that would help reveal which investigators are truly attempting 
high-reward research? In other words, is there a better way for grant 
committees to ``get to know'' investigators?

A1. In Dr. Collins' written testimony submitted for the record, there 
is a section on ``New approaches for identifying potentially 
transformative research'' that addresses these questions.
    Briefly, NSF is experimenting with novel mechanisms for developing, 
reviewing, and funding exploratory and especially creative research. 
All are new ways to foster NSF's process of discovery and thus ``reveal 
which investigators are truly attempting high-reward research.''
    About 18 months ago Malcolm Gladwell argued in an article in The 
New Yorker that ideas are easy to come by; implementing them is hard. 
Ideas, Gladwell argued, are not precious, but everywhere. He concluded, 
therefore, ``maybe the extraordinary process that we thought necessary 
for invention--genius, obsession, serendipity, epiphany--wasn't 
necessary at all.'' The trick, he felt, was getting together a group of 
thoughtful, creative people all thinking about how to solve a problem: 
(``In the Air;'' http://www.newyorker.com/reporting/2008/05/12/
080512fa-fact-gladwell/?yrail).
    NSF's is using three methods to take advantage of this line of 
reasoning.

          The ``Sandpit'' is an experiment in real time, 
        interactive peer review to explore novel solutions to existing 
        problems or identify new areas of research. The Directorate for 
        Biological Sciences, with participation and support from the 
        Directorates for Math and Physical Sciences, Engineering, 
        Social, Behavioral and Economic Sciences, and Computer and 
        Information Sciences and Engineering, sponsored its first 
        sandpit in the area of synthetic biology in conjunction with 
        the United Kingdom's Engineering and Physical Sciences Research 
        Council (EPSRC) in April, 2009. This sandpit produced five 
        interdisciplinary, multi-investigator projects with support 
        from NSF and EPSRC.

          The Directorates for Biological Sciences, 
        Engineering, and Social, Behavioral and Economic Sciences also 
        funded an EAGER proposal that focuses on developing a 
        ``prediction market'' for synthetic biology. A prediction 
        market is a social networking method used to predict the most 
        likely outcome of an event like a presidential election or next 
        quarter's sales for a business. The principal investigator for 
        this award will use the method to assess where the most 
        creative research investments can be made to advance the area 
        of synthetic biology.

          Synthesis Centers promote the process of collecting 
        and connecting disparate data, concepts, or theories to 
        generate new knowledge or understanding. Beyond its necessity 
        for innovation in basic science, synthesis increasingly 
        contributes to novel and effective solutions for pressing 
        problems, and to the emergence of new ideas or fields of 
        inquiry that would not otherwise exist. Biology Directorate-
        funded Synthesis Centers in conjunction with other NSF 
        Directorates and federal agencies emphasize interdisciplinary 
        research and education in critical areas of the biological, 
        computer, and social sciences. Current centers include: the 
        National Center for Ecological Analysis and Synthesis, the 
        National Evolutionary Synthesis Center, the National Institute 
        for Mathematical and Biological Sciences, and the iPlant 
        Collaborative. These centers advance our understanding by 
        interdisciplinary activities as well as by ``getting together a 
        group of thoughtful, creative people all thinking about how to 
        solve a problem.''

    Modern cyberinfrastructure can greatly facilitate these ways of 
identifying the likely places for a commitment to supporting high-risk/
high-reward/transformative research. The social networking manifest in 
models like crowd sourcing or prediction markets is based on arguments 
that there is great value in a collective effort focused on uncovering 
the best sort of research to fund--the so-called ``wisdom of the 
crowd'' argument. However, as noted elsewhere in Dr. Collins' written 
testimony, NSF's merit review system is at its root a wisdom-of-the-
crowd model. The new extensions of this fundamental model rely on 
modern computer and information sciences to integrate tens, hundreds, 
or even thousands of researchers focused on solving a common problem.
    These sorts of social networking models are potentially, in an 
analogy with Clayton Christian's The Innovator's Dilemma, a 
``disruptive technology'' when it comes to discovery related to 
research and education. In relation to the question posed by the 
Subcommittee, these mechanisms are ways ``to modify the peer review 
process--without changing all the things about it that currently work 
well.''

                   Answers to Post-Hearing Questions

Responses by Richard D. McCullough, Vice President for Research; 
        Professor of Chemistry, Carnegie Mellon University

Questions submitted by Representative Vernon J. Ehlers

Q1.  Do you have any recommendations on how to modify the peer review 
process--without changing all the things about it that currently work 
well--that would help reveal which investigators are truly attempting 
high-reward research? In other words, is there a better way for grant 
committees to ``get to know'' investigators?

A1. The key to increasing program manager and review committee 
engagement with researchers is to expand the tools that facilitate 
quality interactions. These tools include early investigator awards 
that expand the exposure of young faculty members and encourage more 
expansive and higher-risk research activities early in careers. In 
addition, the expansion of seed and challenge grants that could provide 
$100,000 for early stage exploratory projects in high-risk, high-reward 
areas could provide a platform for both stimulating faculty engagement 
and expanding access to program managers. These seed grants would be 
designed to advance concepts to a stage where they may be applicable 
for traditional program competitions. Further, the funding could be 
stage-gated with $100K/year for two years then $200K/year for two years 
if the science becomes innovation.

                   Answers to Post-Hearing Questions

Responses by Dr. Gerald M. Rubin, Vice President and Director, Janelia 
        Farm Research Campus, Howard Hughes Medical Institute

Questions submitted by Representative Vernon J. Ehlers

Q1.  Do you have any recommendations on how to modify the peer review 
process--without changing all the things about it that currently work 
well--that would help reveal which investigators are truly attempting 
high-reward research? In other words, is there a better way for grant 
committees to ``get to know'' investigators?

A1. One way to modify the peer review process is to change the criteria 
for reviewing grants to place more emphasis on creativity and 
originality and that would be more tolerant of the chance of failure, 
especially in cases where the reward for success is high. As many types 
of research are needed to advance knowledge, it might be best to have a 
separate category of grants that specifically emphasized so-called 
high-risk/high-reward research. The NIH is trying this approach, but, 
in my opinion, at too small a scale. As part of the review process for 
the Pioneer Award, NIH does conduct personal interviews of the top 
applicants.

                              Appendix 2:

                              ----------                              


                   Additional Material for the Record


              Statement of Professor Franklin M. Orr, Jr.,
                   Stanford University, representing
                the David and Lucile Packard Foundation

    The David and Lucile Packard Foundation appreciates the invitation 
to share our views on high-risk research with the Committee. This 
response to questions from the Committee staff is presented by Franklin 
M. Orr, Jr., trustee of the Foundation from 1999 to 2008, who has long 
been involved with the two programs discussed below.

1.  Why does the Packard Foundation fund basic research? How does, or 
should the Foundation's role differ from that of the Federal 
Government?

    David Packard was co-founder of the Hewlett-Packard Company. The 
success of the Hewlett-Packard Company has been built on technology, 
derived in large measure from research and development in university 
laboratories. Because the endowment of the David and Lucile Packard 
Foundation would not have been possible without the success of HP and 
the research performed by university-educated engineers and scientists 
employed by this company, the Foundation has a long-standing interest 
in strengthening both university-based research and graduate education. 
In 1988, the Foundation established the Packard Fellowships for Science 
and Engineering to allow the Nation's most promising young professors 
to pursue their science and engineering research with few funding 
restrictions and limited paperwork requirements. The goal of the 
program is to encourage talented young faculty to build research groups 
that make career-long contributions by training talented graduate 
students and by conducting research that will be the basis for future 
scientific and economic progress. The Foundation also supports the 
Monterey Bay Aquarium Research Institute (MBARI). These two programs 
receive support of approximately $50 million per year.
    MBARI's mission is to conduct advanced research and education in 
ocean science and technology, and to do so through the development of 
better instruments, systems, and methods for scientific research in the 
deep waters of the ocean. MBARI emphasizes the peer relationship 
between engineers and scientists as a basic principle of its operation. 
MBARI has been a leader in oceanographic science and in the development 
of remotely-operated vehicles (ROVs), autonomous underwater vehicles 
(AUVs), ocean observatories, and in situ chemical and biological 
sensors for research. Both the science and engineering have been high-
risk in many respects. Some developments pioneered at MBARI have taken 
five or more years to field, but have enabled ground-breaking 
discoveries on important issues such as ocean acidification, nitrogen 
uptake, harmful algal blooms, and invasive species. Once proven, these 
new tools have been transferred from MBARI to other oceanographic 
institutions, NOAA labs, and commercial vendors such as Satlantic and 
Battelle. Such developments would be difficult, if not impossible, to 
undertake with traditional, short-term government grants, and would 
also be unattractive to industries concerned with near-term profits.
    The Packard Fellowships are designed to identify 16-20 of the most 
promising early career faculty members, chosen from a field nominated 
by 50 of the Nation's leading research universities. The intent of the 
fellowships is not high-risk research directly. Instead, the aim is to 
provide some very talented and creative young scientists and engineers 
with the opportunity to pursue their research interests with 
substantial unrestricted research funds that can be used flexibly over 
the course of the five-year fellowship. The result of that arrangement 
has been a rich flow of innovative research across a very wide range of 
disciplines. It is often high-risk, in the opinion of the Packard 
Fellows, in the sense that they report that the questions they 
investigated often could not have been supported, at least at the 
outset, because these problems or approaches were not yet at the stage 
where they could attract federal research support. Many Fellows also 
report that they feel that the flexibility of the fellowship confers on 
them an obligation to use the funds in ways that open new areas for 
their research groups or develop research areas, experimental 
approaches, or theoretical attacks that require effort over a period 
that is long compared to shorter federal funding cycles. In other 
words, they feel that they should take advantage of the fellowship 
funds to do something that would be difficult or impossible to do in 
the context of the traditional funding mechanisms.

2.  What is the Foundation's model for funding high-risk, high-payoff 
research? What are the benefits of this model? What are the challenges? 
Is this a model that could or should be duplicated by federal funding 
agencies or federally funded research and development centers such as 
the Department of Energy National Labs or the National Institutes of 
Health?

    In contrast to the current federal system of reviewing research 
proposals, MBARI reviews the researchers. Scientists and engineers who 
are highly productive, show exceptional creativity, remain relevant to 
the Institute's strategic plan, and show good citizenship receive 
steady support. Formation of interdisciplinary teams that can focus on 
a topic for an extended period is encouraged. Over its 20 years of 
existence, MBARI has attracted scientists and engineers who dare to 
push the limits of what is possible in a culture that rewards risk 
taking.
    In accord with David Packard's wishes, MBARI limits the fraction of 
its total funding that comes from federal sources. The intent of this 
approach is to preserve the independence of the Institute and its 
ability to investigate problems that are not constrained by 
programmatic objectives and the inevitable increases and decreases in 
support typical of changeable funding cycles. In addition, this 
approach complements the federal portfolio by allowing sustained effort 
on research challenges that may require extended periods of 
development. This is a model that cannot be transferred, as stated, to 
the federal research establishment, although the autonomy of the 
Defense Advanced Research Projects Agency may approach a similar degree 
of independence in the short run.
    The intent of the Packard Fellowship Program is to identify and 
provide support for unusually creative young faculty researchers early 
in their careers (nominees must be in the first three years of their 
academic careers). The Foundation seeks to support innovative 
individual research that involves the Fellows, their students, and 
junior colleagues, rather than extensions or components of large-scale, 
ongoing research programs.
    Fellows are selected in two stages. Each of 50 invited research 
universities nominates two candidates. The competition within the 
universities is tough enough to produce good candidates. An independent 
advisory panel, whose research expertise spans a wide range of 
scientific and engineering specialties, then reviews the applications 
of the 100 candidates and the recommendations they solicit from mentors 
and leaders in their fields. The Foundation awards up to 20 fellowships 
each year based on the recommendations of the review panel. The number 
of nominations was selected to be large enough to provide a very good 
pool of candidates, but small enough to make the review process 
manageable and the probability of success reasonable.
    Again, the Foundation's process focuses on selecting individuals, 
rather than their research projects. Fellows are encouraged to take 
risks and to change their research plans in the course of their 
fellowships if they judge that it makes sense to do so. Given their 
talents and creativity, of course, it is hard to prove that the Fellows 
would not have pursued high-risk research absent the Packard 
Fellowship. Similarly, their rapid advancement and acknowledged 
leadership across a wide range of scientific and engineering 
disciplines cannot be attributed to the fellowship alone. As noted 
above, the Fellows report that this approach gives them a highly valued 
opportunity to pursue high-risk research. In effect, this approach 
replaces a review process based on a detailed review of specific 
research proposals with a process that attempts to evaluate the 
creative potential of the investigators. Either process inevitably has 
its own challenges and imperfections.
    It is clear that there are good candidates in the pool of nominees 
each year who do not receive fellowships and that there are good young 
faculty at institutions not on the list of institutions invited to 
nominate. The amount of support available from the Foundation is small 
($14-17.5 million per year) compared to the size of the national 
research enterprise. NSF CAREER awards do provide support for young 
investigators, and there are other young investigator awards that do so 
as well. There is likely to be room for additional support that could 
be applied productively. In addition, a program that provided support 
for high-risk research with potential for significant breakthroughs 
would augment these young-investigator programs. It should be noted 
that a federal program that focuses on high-risk research will require 
modification of the traditional peer review process. Reviewers will 
have to be conditioned to accept the risk inherent in an attempt to 
support research that is risky but has high-potential rewards if it is 
successful. In the current review process, creative ideas for research 
may be rejected for funding because reviewers are not convinced (by 
detailed proof of concept experimental results, for example) that the 
investigators can achieve the goal. One way to deal with that problem 
would be to create a tiered program in which shorter-term projects with 
lower funding levels (e.g., at the amount required to support a post-
doctoral fellow for long enough to do the proof of concept experiment) 
are considered in addition to longer-term projects for which the 
pathway is established even though there are many hurdles to overcome.

3.  Given the total funding for basic science and engineering research 
from all sources, is the ratio of funding for high-risk research 
appropriate? If the ratio were to be increased as recommended in 
several recent reports, what should be the responsibility of the 
Federal Government in achieving that increase, and how does that 
responsibility differ from that of private sector research 
organizations and funding sources such as the Packard Foundation?

    These questions are important ones that reflect choices about the 
portfolio of research programs and funding levels. The Foundation has 
not attempted to study these questions and can offer no detailed 
informed judgment concerning the appropriate balance. It seems 
reasonable to argue, however, that there should be some fraction of 
research funding that should support new ideas that involve risk. Our 
commitment to independent funding of excellent research that can attack 
areas that are long-term and therefore involve risk is reflected in the 
international reputations earned by MBARI and by a large majority of 
the past and present Packard Fellows. It is our intent to continue to 
fund those efforts as the endowment of the Foundation allows (the 
decline in the value of the Foundation's endowment this last year led 
to a reduction from 20 to 16 fellowships this year). We believe that 
our support of these scientists and engineers has demonstrated the 
value of supporting those who undertake high-risk research. Given the 
magnitude of the federal research enterprise compared to that funded by 
foundations, however, significant expansion of funding of high-risk 
endeavors will have to come from growth in federal support for such 
research.

4.  Do you have any specific recommendations for how federal science 
agencies such as the National Science Foundation could increase their 
support for high-risk research? In particular, what are the pros and 
cons of establishing targeted programs or set-asides for high-risk 
research versus changing how proposals are reviewed and selected across 
a federal science agency? What are the biggest challenges or risks 
associated with each of these approaches? What metrics should be used 
to evaluate the success of any approach to funding high-risk research?

    David Packard believed in finding excellent scientists and 
engineers, providing resources, and then trusting the investigators to 
use the funding wisely. This is a thread connecting the Foundation's 
support of unmanned research vehicles and the science they can carry 
out in the ocean with our support of early career faculty members. By 
affording creative scientists the independence to pursue their 
curiosity, with an understanding that ``failures'' are part of the 
research enterprise, creative results can follow. To the extent that 
programs focused on high-risk studies can achieve a similar end, they 
serve the Nation's interest. At the same time, it should be noted that 
not all of the research needed by the Nation is or should be high-risk 
in nature, and funding should also reflect the reality that the 
contributions of science and engineering to human well-being arise from 
solid, reliable understanding of nature, won only in part through 
breakthrough studies.
    The notion of risk implies that some fraction of the work that is 
done will fail to produce the originally intended result, though it may 
have unanticipated value that becomes apparent later. Any program that 
emphasizes high-risk work must tolerate redirection as the work 
proceeds and the possibility that some work will be unsuccessful. 
Because the time scales for high-risk research are likely to be long, 
any attempt to measure effectiveness will have to reflect that time 
scale. Simple metrics based on counts of papers and published and 
citations of them will likely not be useful in the short-term. It seems 
likely that the real value of such programs will be much more apparent 
in retrospect, when those areas that have developed rich and productive 
sets of ideas and results can be identified more readily. Thus, taking 
the long view of measuring program effectiveness will be essential.
    The Foundation is grateful for the opportunity to provide its views 
to the Committee.
