[Senate Hearing 115-18]
[From the U.S. Government Publishing Office]





                                                         S. Hrg. 115-18
 
                AGENCY USE OF SCIENCE IN THE RULEMAKING
    PROCESS: PROPOSALS FOR IMPROVING TRANSPARENCY AND ACCOUNTABILITY

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

                                HEARING

                               before the

                            SUBCOMMITTEE ON
               REGULATORY AFFAIRS AND FEDERAL MANAGEMENT

                                 of the

                              COMMITTEE ON
                         HOMELAND SECURITY AND
                          GOVERNMENTAL AFFAIRS
                          UNITED STATES SENATE


                     ONE HUNDRED FIFTEENTH CONGRESS

                             FIRST SESSION

                               __________

                             MARCH 9, 2017

                               __________

                   Available via http://www.fdsys.gov

       Printed for the use of the Committee on Homeland Security
                        and Governmental Affairs
                        
                        
                        
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]



                           ________

                U.S. GOVERNMENT PUBLISHING OFFICE
                   
 25-313 PDF                 WASHINGTON : 2017       
____________________________________________________________________
 For sale by the Superintendent of Documents, U.S. Government Publishing Office,
Internet:bookstore.gpo.gov. Phone:toll free (866)512-1800;DC area (202)512-1800
  Fax:(202) 512-2104 Mail:Stop IDCC,Washington,DC 20402-001   

                        
                        
                        
                        
                        

        COMMITTEE ON HOMELAND SECURITY AND GOVERNMENTAL AFFAIRS

                    RON JOHNSON, Wisconsin, Chairman
JOHN McCAIN, Arizona                 CLAIRE McCASKILL, Missouri
ROB PORTMAN, Ohio                    THOMAS R. CARPER, Delaware
RAND PAUL, Kentucky                  JON TESTER, Montana
JAMES LANKFORD, Oklahoma             HEIDI HEITKAMP, North Dakota
MICHAEL B. ENZI, Wyoming             GARY C. PETERS, Michigan
JOHN HOEVEN, North Dakota            MAGGIE HASSAN, New Hampshire
STEVE DAINES, Montana                KAMALA D. HARRIS, California

                  Christopher R. Hixon, Staff Director
               Margaret E. Daum, Minority Staff Director
                     Laura W. Kilbride, Chief Clerk


       SUBCOMMITTEE ON REGULATORY AFFAIRS AND FEDERAL MANAGEMENT

                   JAMES LANKFORD, Oklahoma, Chairman
JOHN MCCAIN, Arizona                 HEIDI HEITKAMP, North Dakota
ROB PORTMAN, Ohio                    THOMAS R. CARPER, Delaware
MICHAEL B. ENZI, Wyoming             MAGGIE HASSAN, New Hampshire
STEVE DAINES, Montana                KAMALA D. HARRIS, California
                     John Cuaderess, Staff Director
                 James Mann, Professional Staff Member
                  Eric Bursch, Minority Staff Director
               Warren Flatau, Minority Legislative Fellow
     Katie Delacenserie, Subcommittee Clerk and Committee Archivist
     
                            C O N T E N T S

                                 ------                                
Opening statement:
                                                                   Page
    Senator Lankford.............................................     1
    Senator Heitkamp.............................................     8
    Senator Hassan...............................................     9
    Senator Harris...............................................    11
    Senator Daines...............................................    16
    Senator Carper...............................................    23
Prepared statement:
    Senator Lankford.............................................    35
    Senator Heitkamp.............................................    37

                               WITNESSES
                        Thursday, March 9, 2017

Susan Dudley, Director, Regulatory Studies Center, The George 
  Washington University..........................................     3
Andrew A. Rosenberg, Ph.D., Director, Center for Science and 
  Democracy, Union of Concerned Scientists.......................     5
Nancy Beck, Ph.D., Senior Director, Regulatory Science Policy, 
  American Chemistry Council.....................................     7

                     Alphabetical List of Witnesses

Beck, Nancy, Ph.D.:
    Testimony....................................................     7
    Prepared statement...........................................    56
Dudley, Susan:
    Testimony....................................................     3
    Prepared statement...........................................    38
Rosenberg, Andrew A., Ph.D.:
    Testimony....................................................     5
    Prepared statement...........................................    49

                                APPENDIX

Article submitted by Senator Harris..............................    74
Responses to post-hearing questions for the Record from:
    Ms. Dudley...................................................    77
    Mr. Rosenberg................................................    81
    Ms. Beck.....................................................    85


                      AGENCY USE OF SCIENCE IN THE



                   RULEMAKING PROCESS: PROPOSALS FOR



               IMPROVING TRANSPARENCY AND ACCOUNTABILITY

                              ----------                              


                        THURSDAY, MARCH 9, 2017

                                 U.S. Senate,      
                        Subcommittee on Regulatory,        
                      Affairs and Federal Management,      
                    of the Committee on Homeland Security  
                                  and Governmental Affairs,
                                                    Washington, DC.
    The Committee met, pursuant to notice, at 10:15 a.m., in 
room 342, Dirksen Senate Office Building, Hon. James Lankford, 
Chairman of the Subcommittee, presiding.
    Present: Senators Lankford, Daines, Heitkamp, Carper, 
Hassan, and Harris.

            OPENING STATEMENT OF SENATOR LANKFORD\1\

    Senator Lankford. Good morning. Welcome to today's 
Subcommittee hearing entitled Agency Use of Science in the 
Rulemaking Process: Proposals for Improving Transparency and 
Accountability.
---------------------------------------------------------------------------
    \1\ The prepared statement of Senator Lankford appears in the 
Appendix on page 35.
---------------------------------------------------------------------------
    Over 2 years ago, this Subcommittee began an in-depth 
review of the rulemaking process, tackling subjects such as 
retrospective review, agency use of guidance, and issues 
surrounding small business' concerns when it comes to improving 
the regulatory outcomes. This morning we will continue our 
regulatory work by examining how agencies use scientific 
information to inform their regulatory decision making.
    American people should be confident that when agencies 
regulate they are relying on up-to-date, accurate, and unbiased 
information. To put it simply, agencies should rely on the best 
available information and make decisions based on the weight of 
that information.
    When determining whether scientific information is the best 
available, agencies should consider things like whether 
information has been peer-reviewed by an independent third 
party, whether the conclusions are verifiable and reproducible, 
whether the information's use is consistent with its intended 
purpose, and whether the data is transparent and publicly 
available.
    This is not a new idea. Presidents from both parties have 
stressed the importance of relying on sound science to inform 
regulatory decisions. Executive Order (EO) 12866, which has 
been in place since 1993, and endorsed by every President 
since, directs agencies to base decisions on the best 
reasonably obtainable, scientific and technical information. 
Eight years ago, President Obama went even further by issuing a 
memorandum to agency heads guaranteeing scientific integrity by 
following a list of principles that included consideration of 
well-established scientific processes and urging transparency 
to the public. And in 2011, President Obama issued Executive 
Order 13563, where he directed each agency to ensure the 
objectivity of any scientific and technical information used to 
support regulatory actions.
    Yet, despite these clear directives, agencies continue to 
use questionable science to support their regulatory decisions, 
or we do not know the background on science. For example, in 
2015, when the Environmental Protection Agency (EPA) proposed a 
ban on chlorpyrifos, an insecticide that farmers have been 
using successfully for decades, the agency based the regulation 
on a study that was discredited by their own scientific 
advisory panel and by the U.S. Department of Agriculture 
(USDA).
    I understand agencies often face difficult choices and not 
all studies come to the same conclusion, but it is very 
concerning when agencies are not open about why they chose to 
use a study with such significant criticism. When agencies hide 
information from both Congress and the American people, it is 
our job to question motives and methods as a part of oversight.
    Transparency is not an unreasonable request. In fact, it 
will go a long way in enforcing better regulations and heading 
off lawsuits in the future. When agencies issue regulations 
that place legally binding requirements on the American people, 
the data and methods or models the agency used should be 
publicly available for independent, third-party review.
    When many of the most costly Clean Air Act (CAA) 
regulations are based on a single Harvard study from 1993, EPA 
should not be able to hide behind the excuse they can't release 
the study because they do not own it, Harvard does, despite the 
fact that Harvard receives well over half a billion dollars in 
Federal awards.
    Examples like these call into question whether agencies are 
using best information because it is unknown, not because there 
is just a false accusation. We just don't know, and the 
American people do deserve to be able to know when these things 
impact their lives in such a significant way.
    Each administration has their own priorities, but the 
principles supporting regulatory decisions should remain 
constant, regardless of who occupies the White House. If past 
administrations' attempts to encourage agencies to base their 
regulatory decisions on transparent, sound science have failed, 
Congress should consider establishing new legal requirements.
    I look forward to discussing steps Congress can take to 
implement basic, fundamental requirements that have been 
endorsed by both Democrat and Republican administrations for 
decades.
    Senator Heitkamp is running a little bit behind, though I 
will always tease her and say the lady is never late, but I 
will recognize her for an opening statement if she chooses to 
give that orally, or to be able to put that into the permanent 
record if she does not choose to.
    With that I would like to recognize our witnesses. Let me 
introduce them all and then we will do a swearing-in, and then 
we will take your testimony.
    The Hon. Susan Dudley is the Director of the Regulatory 
Studies Center and Distinguished Professor of Practice in the 
Trachtenberg School of Public Policy and Public Administration, 
George Washington (GW) University. Before joining the faculty 
at GW, she served as the Administrator of the Office of 
Information and Regulatory Affairs (OIRA), from 2007 to 2009. 
Thanks for being here.
    Dr. Andrew Rosenberg is the Director of the Center for 
Science and Democracy at the Union of Concerned Scientists. 
Before joining the Union of Concerned Scientists, he served as 
the Northeast Regional Administrator of the National Marine 
Fisheries Service (NMFS) at the National Oceanic and 
Atmospheric Administration (NOAA). Thanks again.
    Dr. Nancy Beck is a senior director of Regulatory and 
Science Policy at the American Chemistry Council (ACC). Prior 
to joining the ACC she served as a toxicologist and science 
policy analyst at the Office of Management and Budget (OMB's) 
Office of Information and Regulatory Affairs from 2002 to 2012.
    I do thank all of the witnesses for appearing.
    It is our custom, as a Subcommittee and as a committee, to 
swear in all witnesses, so if you would all please rise and 
raise your right hand.
    Do you swear the testimony you will give before this 
committee will be the truth, the whole truth, and nothing but 
the truth, so help you, God?
    Ms. Dudley. I do.
    Mr. Rosenberg. I do.
    Ms. Beck. I do.
    Senator Lankford. Thank you. You may be seated. Let the 
record reflect the witnesses have all answered in the 
affirmative.
    We do use a timing system here. That would be a 5-minute 
time clock that will start in front of you, countdown for your 
opening statements. Obviously all of your written statements 
will go into the permanent record already, and so we are 
welcome to be able to receive your oral testimony.
    Susan Dudley, you are up first.

  TESTIMONY OF SUSAN DUDLEY,\1\ DIRECTOR, REGULATORY STUDIES 
            CENTER, THE GEORGE WASHINGTON UNIVERSITY

    Ms. Dudley. Thank you very much, Chairman Lankford, and it 
is good to meet you, Senator Hassan.
---------------------------------------------------------------------------
    \1\ The prepared statement of Ms. Dudley appears in the Appendix on 
page 38.
---------------------------------------------------------------------------
    Effective regulatory policy that focuses resources on 
addressing real threats to public health and the environment 
depends on reliable scientific information and transparent 
policy choices, so your hearing is important.
    I do not think anyone wants science to be politicized, but 
no one is immune to the temptation to put a spin on science to 
advance a policy goal. Politicization can arise when political 
decisionmakers attempt to distort scientific findings, but also 
when scientists and others attempt to exert influence on policy 
decisions by how they present scientific information.
    My prepared testimony focuses on the second type of 
politicization. It identifies two problems that current 
regulatory institutions tend to aggravate. The first is when 
scientists, intentionally or not, insert but do not disclose 
their own policy preferences in the scientific advice they 
provide decisionmakers. I have called this ``hidden policy 
judgments'' and others have called it ``advocacy science'' or 
``normative science.'' The second problem is the tendency to 
camouflage controversial policy decisions as science. For this 
I adopt Wendy Wagner's colorful term, a ``science charade.''
    Institutional arrangements tend to aggravate both the 
hidden policy judgments and science charades, which threaten 
the credibility of science and harm regulatory policy. Many of 
those involved in regulatory decisions have incentives to hide 
policy preferences, such as how to deal with uncertainty 
inherent in assessments of risk, and to dismiss, and even 
denigrate dissenting views. Key policy choices, disguised as 
science, too often rest with technical staff, while 
policymakers are able to avoid responsibility by claiming that 
their hands were tied by the science.
    In thinking about reforms to improve how science is used in 
regulation, clarifying which aspects of the decision are 
matters of science and which are matters of policy is essential 
to avoid both hidden policy judgments and the science charade. 
This was a key finding of a 2009 Bipartisan Policy Center (BPC) 
report, as well as a 1983 National Research Council (NRC) book.
    In a forthcoming paper, Marcus Peacock and I offer 10 
recommendations. I boiled those down to eight in my written 
testimony, and will trim them down to five for the remaining 
few minutes that I have.
    First, we must recognize that science is a positive 
discipline that can inform, but not decide, appropriate policy. 
In drafting authorizing legislation, Congress should not 
delegate decisions to agencies on the pretense that science 
alone can make the normative decision of what policy ought to 
be.
    Second, science advisory panels can provide a necessary and 
valuable source of information and peer review for agencies, 
but greater efforts should be made to restrict their advice to 
matters of science and not ask them to recommend policy. 
Further, a 2012 Keystone Center report emphasized all potential 
panelists on those science advisory boards will have conscious 
and unconscious biases, so it is important to select panelists 
with diverse perspectives and expertise, who are willing to 
engage in open discussion and are open to considering other 
views.
    Third, risk assessments necessarily involve assumptions and 
judgments as well as pure scientific inputs, yet they often 
generate precise-sounding predictions that hide not only 
considerable uncertainty about actual risk but hidden 
judgments. Greater transparency regarding the assumptions and 
policy rationales for choosing one set of assumption or models 
over another would encourage more openness and constructive 
discussion about science and policy.
    Fourth, reproducibility is important and it requires 
sharing of underlying data. I understand and appreciate 
concerns about disclosure of personally identifiable 
information (PII) and confidential information, but many 
Federal agencies have successful guidelines allowing access to 
data containing PII, using tiered approaches commensurate with 
the sensitivity of the information.
    And finally, the scientific method depends on falsifiable 
hypotheses, data gathering, replication, dissent, and 
challenge--Reforms that intentionally engage rather than avoid 
competing views and that subject scientific predictions to ex-
post evaluation would go a long way toward improving underlying 
science and the decisions that depend on it.
    Legislation both Senators Lankford and Heitkamp have 
introduced, could be constructive. One bill that you introduced 
just yesterday would require advanced notices of proposed 
rulemaking for high-impact regulations, and one that you 
introduced together last session, would have required agencies 
to plan for ex-post review when proposing new rules, and that 
provides that feedback.
    So I will close there, and thank you again for inviting me.
    Senator Lankford. Thank you. Dr. Rosenberg.

 TESTIMONY OF ANDREW A. ROSENBERG, PH.D.,\1\ DIRECTOR, CENTER 
    FOR SCIENCE AND DEMOCRACY, UNION OF CONCERNED SCIENTISTS

    Mr. Rosenberg. Chairman Lankford, Senator Hassan, Members 
of the Subcommittee, and Ranking Member Heitkamp, thank you for 
the opportunity to testify today on the important role of 
science and rulemaking process.
---------------------------------------------------------------------------
    \1\ The prepared statement of Mr. Rosenberg appears in the Appendix 
on page 49.
---------------------------------------------------------------------------
    I am Andrew Rosenberg, Director of the Center for Science 
and Democracy at the Union of Concerned Scientists, and I have 
more than 25 years of experience as a scientist in government 
service, academia, and the private sector. I have served as a 
scientist and regulator in Democratic and Republican 
administrations, including as Deputy Director of NOAA 
Fisheries, as you noted. And in academia, I was the Dean of 
Life Sciences and Agriculture at the University of New 
Hampshire.
    Independent science plays a critical role in the policy 
decisions made by the Federal Government that impact Americans' 
health and safety, and the science process consists of 
continuous and incremental discoveries in multiple fields of 
study, accumulating a weight of evidence and building toward 
broad acceptance of facts within the scientific community.
    Weight of evidence refers to the cumulative body of 
scientific research and analysis that pertains to a particular 
subject. Weight refers not only to the number of studies but 
also their importance, robustness, and credibility in drawing 
scientific inference. Credibility relates to the design of the 
study, analytical methods, methods of inference, as well as 
provenance of the work with regard to potential conflicts of 
interest, peer reviews conducted, and comparisons with other 
relevant studies.
    A valid and credible scientific process consists of a 
rigorous examination of ideas, review and critique by 
technically qualified peers, open exchange of ideas among 
colleagues, and protection against manipulation of results by 
vested interests.
    I serve as a regular reviewer for scientific journals, as a 
member of editorial boards, and as an independent reviewer for 
national and international reports. In this capacity, I 
consider the framing of the study, the methods, the results, 
the researchers' interpretation in light of my knowledge of the 
field, and relevant scientific literature. I may not agree with 
the conclusions drawn by the researchers, but if the 
aforementioned components are well-executed then the study 
contributes to the body of evidence, in my view.
    Every study is subtly different and should be judged by 
experts in the field on its merits, and in my view, best 
available science is research that is conducted in accordance 
with well-established scientific practices, including well-
designed investigation, logical statistically rigorous 
analyses, clear documentation of data collection and analytical 
methods, careful peer review of results free from external 
influences seeking a particular policy outcome.
    I strongly believe that this process cannot be legislated 
without undermining innovation and unduly constraining the 
scientific information available to policymakers and to the 
public. If one were to legislate what should legally be 
considered best available science, it would prevent the 
innovation and flexibility that is inherent in the scientific 
process. This ability to innovate is essential for agencies as 
they address new discoveries, such as emerging public health 
threats.
    Public access to science that underlies regulatory 
decisions is important as part of our democracy and to ensure 
the rationale for decisions is clear, even if all do not agree 
with the final policy outcomes. On this point it is important 
to distinguish between data and science. The scientific 
information critical for an informed public concerns how 
studies are conducted, how information is interpreted, and 
inferences are drawn. This is not dissimilar from the 
information a peer reviewer like me considers in evaluating a 
study, albeit for the public in a non-technical form. Peer 
reviewers, in general, do not review raw data.
    Agency rulemaking must be informed by independent science 
advice that is free from political pressure. Twenty-four 
Federal agencies have developed scientific integrity policies 
that help ensure the independence of scientific advice. There 
is now legislation in both House and Senate that would codify 
scientific integrity policies into law, which I view as a 
positive step for protecting science-informed policymaking. 
Strengthening peer review policies in several agencies further 
safeguards government science.
    Best available science should be used to describe the 
weight of scientific evidence developed by a credible process 
for ensuring independence from political influence. Agency 
scientists supported by a commitment to independent science, 
scientific integrity policies, and appropriate transparency 
measures should be trusted to provide science advice to 
decisionmakers, and I agree it is advice. All Americans benefit 
when science is used to inform policy, and its integrity in the 
rulemaking process is imperative for a functional democracy and 
a safer, cleaner planet for all.
    Thank you Mr. Chairman, Ranking Member, and Members of the 
Committee. I appreciate the opportunity.
    Senator Lankford. Dr. Beck.

TESTIMONY OF NANCY BECK, PH.D.,\1\ SENIOR DIRECTOR, REGULATORY 
          SCIENCE POLICY, AMERICAN CHEMISTRY COUNCIL.

    Ms. Beck. Good morning, Chairman Lankford, Ranking Member 
Heitkamp, Senator Hassan, and Senator Harris. I am honored to 
be here today representing the American Chemistry Council. My 
name is Nancy Beck and I have spent over 15 years working at 
the intersection of science and policy.
---------------------------------------------------------------------------
    \1\ The prepared statement of Ms. Beck appears in the Appendix on 
page 56.
---------------------------------------------------------------------------
    The business of chemistry is a critical component for 
manufacturing safe, high-quality products. ACC member companies 
and the public rely on science to innovate, to advance product 
stewardship, and to prove the assessments of chemical risk. ACC 
members expect high-quality science and objective assessment 
processes to underpin regulatory decisions.
    Reliance on high-quality science is critical to ensuring 
public trust. Regrettably, that trust is eroding. In July 2016, 
almost 200 toxicologists signed an appeal for the integrity in 
science. They were expressing concerns that precautionary 
regulations and policies are being presented as objective 
science, when, in fact, they are not. Dr. Rosenberg recently 
stated, ``When science is sidelined from policy decisions, we 
all lose.'' The American Chemistry Council shares these 
concerns.
    Improving agency science should not be as challenging as it 
has been. As you have mentioned, Senator Lankford, government 
and non-governmental guidance already exists. In 2002, Federal 
agencies were directed through OMB's information quality 
guidelines to ensure the quality, objectivity, utility, and 
integrity of information which they disseminate to the public.
    Unfortunately, while most agencies have committed to 
meeting these standards, some agencies' scientific analyses 
falls short. Some examples of this are provided in the 
testimony ACC has provided in writing. These shortfalls are 
despite Federal guidance on risk assessment and peer review, 
and further guidance from organizations like the National 
Academy of Sciences (NAS), the Keystone Center, and the 
Bipartisan Policy Center.
    The bipartisan Lautenberg Chemical Safety Act (LCSA) now 
mandates that EPA apply high-quality, reliable, and relevant 
scientific information by using best available science and a 
weight of the evidence approach. To date, EPA appears to imply 
that business as usual is consistent with these standards. 
Rather than explicitly incorporating the best available science 
and the weight of the evidence standards into the draft 
framework rules it is developing to implement the act, the 
agency has suggested that it can simply rely on existing 
approaches and current practices. This is of great concern to 
the ACC.
    The peer review report from EPA's most recently released 
draft risk evaluation notes that EPA did not apply a weight of 
the evidence approach, nor did it use a systematic review 
process, which is a critical part of conducting a weight of the 
evidence evaluation. EPA simply chose the value that would 
provide the lowest allowable exposure, without transparently 
documenting the quality of the individual studies they 
considered.
    The good news here is that there are many potential 
solutions to improve agency science, and I am going to present 
just four.
    One, improving and clarifying scientific definitions. The 
intent of the scientific standard is to improve existing agency 
practices. Agencies should provide clear and specific 
definitions for terms like best available science and weight of 
the evidence. These definitions should address not only what 
agencies should consider when evaluating information, but also 
what information agencies should present in the evaluations, 
forcing the agencies to show their work and present their 
thought process in a clear and transparent manner.
    Two, there must be stronger oversight to ensure that 
existing guidance is actually followed. Existing guidance 
documents could be converted into agency checklists and agency 
staff could self-certify that their evaluations are consistent 
with these checklists. These checklists could also be used by 
stakeholders and peer reviewers.
    Third, as was mentioned before, the importance of improving 
peer review practices cannot be underestimated. Ensuring that 
peer review panels have the depth and breadth necessary to 
address scientific concerns while also ensuring that conflicts 
and biases are addressed is critical.
    And fourth, the government can play an important role in 
changing the incentives for grant funding such that decisions 
are not so heavily dependent on finding positive results, and 
the government standards for funding should ensure that 
research studies follow best scientific practices and are 
designed with the regulatory use in mind.
    In conclusion, ensuring that Federal decisionmaking is 
firmly based on the use of high-quality science can be achieved 
through common-sense reforms that will lead to a more efficient 
and effective regulatory process. ACC looks forward to working 
with Members of the Committee to enhance approaches to ensure 
that high-quality science is the foundation of regulatory 
decisionmaking.
    Thank you.
    Senator Lankford. Thank you. Before we move into our time 
of questions, as is our tradition on this committee, the 
Ranking Member and I will defer our questions at the end.
    Before we move to Senator Hassan, I want to recognize 
Senator Heitkamp for her opening statement.

             OPENING STATEMENT OF SENATOR HEITKAMP

    Senator Heitkamp. Just a quick apology to our panel. I had 
a markup in the Banking Committee where a bill of mine was 
being considered, and I wanted to make sure it got across the 
finish line. So I had a chance to review your testimony and 
really appreciate your attendance, and I ask that the Chairman 
include my opening comments for the record.\1\
---------------------------------------------------------------------------
    \1\ The prepared statement of Senator Heitkamp appears in the 
Appendix on page 37.
---------------------------------------------------------------------------
    Senator Lankford. Absolutely. Without objection. So did you 
get your bill across the finish line?
    Senator Heitkamp. Well, I got it out of committee.
    Senator Lankford. OK. Out of the gate is a start. 
[Laughter.]
    So that is great.
    Senator Hassan, you are recognized.

              OPENING STATEMENT OF SENATOR HASSAN

    Senator Hassan. Well, thank you, Mr. Chairman, and Ranking 
Member Heitkamp, and welcome to all of the witnesses today. 
Thank you all for being here.
    I wanted to start, Dr. Rosenberg, with a question to you. 
In your testimony you talk about the scientific community's 
view of what constitutes best available science, and you also 
state that generally accepted standards cannot clearly be 
legislated without jeopardizing the kind of scientific 
innovation that leads to some of our most important findings 
and discoveries.
    So I would ask you to take just a minute or so to elaborate 
a little bit more on what you see as the problem with 
legislating what should be considered the best available 
science.
    Mr. Rosenberg. Thank you, Senator. I think this is a really 
crucial point, of course, for the hearing, but in general as we 
think about the processes of bringing science into rulemaking.
    If you legislate best available science, I believe, from 
being an agency scientist and regulator, that is exactly what 
agencies will do and they will not deviate. And so even if 
there is clearly a better approach, even if there is new 
science that should be brought into the mix, agencies will say, 
``My mandate is to do it this way. I will do it that way until 
the mandate changes.''
    I know that there is some rhetoric that agencies do things 
differently from their mandates. From my experience, you spend 
an awful lot of time saying, ``OK, what is the law telling us 
we need to do? '' and you really do not have very much time to 
do other things. If that guidance is given in statute for best 
available science, it will become very rigid in agency 
practice, and the danger there is that, science does actually 
evolve, quite dramatically sometimes. You want people to be 
able to include new information, new types of data collection 
that lead to new inference in their results and bring that to 
bear on the best policies for the Nation.
    Senator Hassan. Thank you. Another question for you, Dr. 
Rosenberg. You highlighted the significance of understanding 
scientific uncertainties, but also the importance of acting 
once the weight of evidence is compelling enough to justify a 
reasonable policy solution. For example, while there may be a 
degree of uncertainty as to some aspects of climate change, 
there is widespread agreement that global climate change caused 
by human activities is occurring now and it is a growing threat 
to society, and as somebody whose home in New Hampshire just 
ran out of power because of sudden extreme winds this morning, 
like many Americans I am acutely aware of this growing sense of 
extreme weather.
    So why does science have uncertainty, why is it OK for 
scientific uncertainty to exist, and how does uncertainty play 
into the decisionmaking process.
    Mr. Rosenberg. Well, again, thank you for the question. 
Uncertainty exists not only in science, of course, but 
unfortunately in life, but certainly scientific evidence will 
always have some degree of uncertainty. I think the importance 
is to characterize that uncertainty but think of it in terms 
of, from a policy perspective, then, what risks are we 
incurring?
    And so uncertainty comes in many forms. As a scientist you 
think about uncertainty in really four different ways. You 
think about measurement error, which is your ability to 
actually measure what is happening in the world; process error, 
which is your understanding of the processes by which things 
move forward; implementation error, your ability if you 
implement something to actually do it accurately; and model 
error, your ability to characterize a process in a model which 
is inherently going to leave some aspects out.
    Well, that is helpful for scientists to think about those 
types of uncertainty and they mean something different, but to 
a broader public, and usually to policymakers, uncertainty 
means you do not really know. It does not necessarily mean 
that. I mean, all of those types of uncertainty are different.
    So if you go back to your climate change example, I would 
say it is not so much that there is a level of agreement among 
scientists as there is a very large weight of evidence, and if 
you look at that weight of evidence, it is quite compelling, 
any possible way you look at it. And sure, there may be people 
who say, ``Well, I can identify one piece of evidence that does 
not seem to fit,'' but the weight of evidence is entirely 
clear.
    Now you, as policymakers, are responsible for deciding how 
much risk we should take. That is not a science decision. I 
agree with Professor Dudley on that. We can say if you tell us 
something about risks that you believe society should incur, 
how the uncertainty that we have relates to those risks. Most 
people would say, if, in fact, climate is changing as rapidly 
as the evidence indicates, that the risks to many aspects of 
society are very high, so, therefore, you might want to take 
action with more uncertainty than, say, you would take in 
crossing the street.
    Senator Hassan. Right. Well, thank you, and that leads me 
to a question for the whole panel, a general one, and I have a 
little over a minute left. So I want to follow up on that, 
because Dr. Rosenberg just described the balance between 
uncertainty and risk to public. In New Hampshire, as you all 
know, like in many other States, we have a heroin and opioid 
epidemic that is devastating and it is a critical public health 
threat. And the threat, the scale of it is stunning, it is 
unprecedented, and we have been really working hard to get in 
place the prevention, treatment, recovery, and public safety 
measures needed to support those on the front lines of this 
epidemic.
    But sometimes those interventions and methods we are using 
are newer, and while we have evidence supporting our policies, 
we do not always have the kind of 100 percent certainty that we 
would like to have to approach this public health epidemic.
    So the real question is, does scientific uncertainty about 
something mean we should not take action at all when the risk 
is high, and what dangers do you think legislating a definition 
of best available science could pose to responding to this type 
of urgent crisis? Because I have only got like 10 seconds left, 
Dr. Dudley, I will start with you and then I will ask the 
others to submit a written answer to that.
    Ms. Dudley. Very briefly, no. Uncertainty does not mean act 
or do not act, but I agree with Dr. Rosenberg--it is a policy 
question. So scientists can tell you, ``Here is the 
uncertainty. Here is why we have this uncertainty,'' and it is 
policy officials' responsibility to say, ``Here is what action 
I would like to take,'' knowing that uncertainty.
    Senator Hassan. Thank you.
    Senator Lankford. Senator Harris.

              OPENING STATEMENT OF SENATOR HARRIS

    Senator Harris. Thank you. Thank you all for your work. I 
have to tell you my mother was a scientist, an endocrinologist, 
and, in fact, one of my first jobs ever was cleaning pipettes. 
I was awful. She fired me. [Laughter.]
    But I do appreciate your work and the importance of the 
work, and particularly the importance of politics not playing 
into your mission and the very important role you play in 
advancing us, in terms of human health and also what we need to 
do in terms of really thinking about what we can do in terms of 
being innovative as a country.
    So, Dr. Rosenberg, I have a question for you. This morning 
on CNBC, EPA Administrator Scott Pruitt stated that carbon 
dioxide is not a primary contributor to global warming. He 
said, and I quote, ``I think that measuring with precision 
human activity on the climate is something very challenging to 
do, and there is tremendous disagreement about the degree of 
impact, so no, I would not agree that it is a primary 
contributor to global warming that we see.'' And then he went 
on to say, ``We need to continue the debate and continue to 
review and continue our analysis.''
    Separately, according to an article in the New Republic on 
March 7, the EPA's Office of Science and Technology, which has 
historically been in charge, of developing clean water 
standards for the States, had on its website that their 
standards were guided by science. Since January 30 of this 
year, the reference on their website to science-based standards 
has disappeared.
    So my question to you is that we are obviously talking a 
lot here, in this hearing, about sound science policy, yet one 
of the most important agencies that should consider science has 
removed the reference to science from their website. So can you 
talk a little bit about what you believe the role of science 
will play in the EPA and any other Federal agencies, and your 
concerns, if the importance and significance of science is 
diminished?
    Mr. Rosenberg. Thank you, Senator, for the question. I was 
not good at cleaning pipettes either, so I became a fisheries 
scientist instead of a Senator, I guess. [Laughter.]
    The statements that you read are very concerning. I can 
only tell you what I believe the role of science should be. 
Clearly, the weight of evidence as--in my response to Senator 
Hassan--on not only the occurrence of global warming but also 
the human role in global warming, is very strong. In fact, it 
has been looked at by multiple studies, multiple investigators. 
All of the data sets indicate that you cannot really explain 
the data without attribution to greenhouse gas emissions, 
including CO2, but not restricted to CO2.
    So it is very concerning to say that he does not believe 
that CO2 is a primary contributor, and equally, that the 
uncertainty is so high that we cannot act, and I think this 
goes back to Senator Hassan's question directly. Whether it be 
an opioid crisis, global warming, or many other issues, you 
really have to think about what is the risk. So it seems to me 
that the uncertainty is relatively low about whether global 
warming is occurring. There certainly is higher uncertainty 
about localized effects, but the risk of not taking action is 
quite high, and there are elements of understanding that risk 
that come from the science.
    For an agency like the EPA, which is fundamentally focused 
on public health and safety protections, which are largely 
identified by scientific work, to, in any way, diminish the 
role of science in the agency is extremely concerning because 
it is public health and safety, and that is what they were set 
up by Congress to do, and those are fundamentally science and 
technical questions. How you address them, of course, is both a 
policy decision for you and for the Administration, depending 
on which policies we are talking about, but I hope we would all 
agree that without a very strong science basis and recognition 
of that science, then we would be in serious trouble with 
regard to public health and safety across the country.
    I was asked what I make of the change in the website for 
the EPA Office of Science. I was very concerned about it 
because it referenced economically viable solutions as opposed 
to public health, and that certainly worries me. In my 
experience as a regulator, there is an awful lot of discussion 
about what is economically viable for whom, which part of an 
industry and which part of the public, and so on, but public 
health threats need to be front and center and predominant, and 
that is a science and technical exercise, by and large.
    Senator Harris. And so we have a lot of students in the 
hearing room, which always makes us happy. Can you explain to 
me and them, and the Committee, what you believe this might 
means in terms of the next couple of years, and what the 
diminishing of science has the potential to do, in terms of 
policy for the next couple of years, and any long-term impact 
as it relates to the health and well-being of our citizens?
    Mr. Rosenberg. Well, first of all, again, I hope that the 
EPA continues to do what they have done for the Nation since it 
was created, and that is continue to protect public health and 
safety. And sure, I understand that people can criticize from 
different perspective, but fundamentally we should realize 
where we have come over the last 40 to 50 years, in terms of 
public health and safety. I live in the Boston area. Look at 
Boston Harbor or the Charles River or the air in Los Angeles, 
or whatever you would like, to decide whether that is 
important. So I hope the agency refocuses on using science and 
technical information.
    I believe that there are many actions that have been 
proposed or called out or mentioned by Administrator Pruitt and 
others in the Administration that would dramatically slow the 
ability to address public health and safety concerns, and I 
think that that is truly problematic, because while we have 
come a long way in 40 years, there certainly are many issues 
that continue to arise. So I hope students are thinking, first 
of all, it is really important to be involved in science and 
technology, and I have quoted you a number of times, Senator 
Hassan, from when we met, that, everyone here agrees that 
Science, Technology, Engineering and Math (STEM) education is 
really important. I hope we listen to the STEM-educated.
    And so I hope the students here think about STEM education, 
and also are thinking fundamentally about what history has 
shown us about public health and safety changes in the country. 
They may not remember L.A. when you could not drive down the 
streets, but it did occur, and we have learned a lot from the 
history of agencies like the EPA, and we cannot let up now.
    Senator Harris. Thank you. And, Mr. Chairman, I have asked 
unanimous consent that the article be entered into the 
record.\1\
---------------------------------------------------------------------------
    \1\ The article referenced by Senator Harris appears in the 
Appendix on page 74.
---------------------------------------------------------------------------
    Senator Lankford. Without objection.
    Senator Harris. Thank you.
    Senator Lankford. Before I recognize Ranking Member 
Heitkamp, I would encourage dialogue with us. We have a bill 
dealing with sound science, and it is not a partisan bill. That 
is everything, and for everyone, for either side of this. And 
so I would appreciate just the dialogue on that ongoing.
    So, Senator Heitkamp.
    Senator Heitkamp. Thank you, Mr. Chairman, and thank you to 
all of the students here. I know this seems like it might be 
really boring, and not consistent with some fireworks that you 
see in other committee hearings, but I think that this 
discussion is so important to the future of our country. How do 
we take fact-based analysis, science-based analysis? And you 
will never separate facts from judgment.
    So we are going to have different judgments about different 
facts, and we are going to have to analyze those. But the 
problem is that frequently people have different sets of facts, 
and different visions of science, some of which you heard 
today. The challenge that we have right now regarding CO2--we 
can have this discussion, but frequently, on things like that, 
the world moves around us. Right? The Chief Executive Officers 
(CEOs) of major Fortune 500 companies move around us, and say, 
in their judgment, they have looked at it, they need to do 
things differently in their own contexts. The public moves 
around us.
    And that is always the challenge we have with science, 
because frequently, in Washington, D.C., it is politicized. We 
just have to put that out there, that science is in the eyes of 
the beholder, what is, in fact, scientific fact. And one of the 
things that I resist that we are doing a little bit today is we 
are just looking at one area, whether it is climate or whether 
it is, in fact, the EPA.
    This discussion goes beyond that, and I want to engage 
first with Ms. Dudley, because we have had long conversations 
about the application of fact-based analysis to rulemaking, and 
there is a challenge here because people look at rulemaking, or 
these challenges, through the lens of what they believe to be 
fact.
    I want to just give you an example. Recently--this is 
actually true--a basketball player suggested that the world was 
flat. That was in a tweet. I do not know if you saw it. And the 
response that someone gave was, ``Well, we should respect his 
opinion.'' And I'm thinking, maybe not, because I thought 
Galileo dealt with this and was imprisoned. The Earth goes 
around the Sun. The Earth is not flat.
    But yet there seems to be an acceptance in this country for 
a wide variety of reasons which say let's respect this. And so 
there is a widespread kind of narrowing of the lanes. You can 
have a difference of opinion within a narrow lane, but we 
cannot be swimming outside the lanes so that we accept or have 
a reasonable dialogue about the Earth being flat. I am just 
sorry. I am not going to do it.
    And so the reason why I raised that is, can we, in fact, 
legislate that lane in a way that really gets us back to using 
science, but also recognizes that judgment comes with science? 
And I would like to start with Susan.
    Ms. Dudley. I think that is a great way to look at it. 
There are some things that clearly we know, based on science. 
The world is not flat. There are other things where there is 
more uncertainty. As Dr. Rosenberg said, there is a lot of 
uncertainty. It is inherent in everything, not just science but 
in everything. And being able to differentiate what is pure 
science, what is pure policy, and what is in that trans-science 
area in the middle that we need to translate from pure science 
to policy, that is the trickiest thing. And if we can be more 
explicit about that, if scientists can say, ``Here is where the 
science ends and here is the uncertainty around it,'' if we 
could be more explicit about that, I think we would solve a lot 
of the problems that you are talking about, where we are 
arguing about science when we really are not.
    I will just quote briefly from The Bipartisan Policy 
Center. They had a report in 2009, on improving use of science 
in regulatory policy, and they concluded that ``a tendency to 
frame regulatory issues as debates solely about science, 
regardless of the actual subject in dispute, is at the root of 
the stalemate and acrimony all too present in the regulatory 
system.''
    Senator Heitkamp. Yes. Dr. Rosenberg.
    Mr. Rosenberg. So I agree. I mean, this is an absolutely 
critical point. I believe, certainly, in the process of 
policymaking we should give different people their due, based 
on expertise. I certainly have an opinion on pretty much 
everything, my wife will tell you.
    Senator Heitkamp. And you would agree with me, the world is 
not flat. [Laughter.]
    Mr. Rosenberg. I am quite certain that the evidence is 
overwhelming.
    Senator Lankford. Have you ever been to western Oklahoma? 
[Laughter.]
    Mr. Rosenberg. No, sir.
    Senator Heitkamp. Or the Red River Valley in North Dakotas.
    Mr. Rosenberg. So I recognize there is some uncertainty---- 
[Laughter.]
    But I actually think that the scientific evidence and the 
inference that you draw from that evidence, and call science, 
should be done by people who are working through a scientific 
process, and, in fact, are scientists.
    By the same token, I certainly agree with Susan, that, 
everyone has an opinion and we need to try to be careful so 
that for me as a scientist to offer my opinion about, a policy 
matter and I am couching it clearly as this is my opinion about 
the policy matter, not that it is my science that dictates a 
particular policy matter. And so I often am in training for 
young scientists and students who will say it is important to 
distinguish--here is the evidence and here is what I think it 
means, in light of the body of scientific knowledge. Now if I 
want to say, ``Here is what I think you ought to do about it, 
Senator,'' that needs to be clearly separated.
    The difficulty in the current debate, of course, is that 
you have all kinds of people saying, ``Well, the science says 
this.'' ``No, it says that.'' Most of them seem to be non-
scientists or observers, and I think they should offer their 
opinion, whether the world is flat or not, but I think we 
should weight it and not consider it scientific evidence. That 
is why I continually refer to the scientific process in my 
testimony.
    And I do not think that science is the sole deciding factor 
for most policies. It never is. You have an incredibly 
difficult job of making those societal choices. Those societal 
choices that you deal with every day are not a scientific 
decision. They are not part of the science process. All we can 
do is try to inform you of what the threats are, problems are, 
the opportunities, and certainly the consequences of certain 
decisions, based on the evidence that we have collected or have 
in hand.
    Senator Heitkamp. Well, let us talk to Dr. Beck.
    Ms. Beck. Yes. It is a great question. To me, it comes down 
to transparency. I do not want you to trust me because I am a 
scientist. I want you to trust me because you have looked at my 
analysis, and you have been able to evaluate, if it rigorous, 
did it follow the scientific method, did anyone peer-review it? 
Right? If we can be transparent about the quality of the 
science and the uncertainties that come along with the science, 
then we can have a real dialogue about what the policy should 
be.
    So, to me, the solution is in clarity in transparency and 
the strength of the science.
    Senator Heitkamp. Yes. I mean, I just think it is a big 
circle, and, there is consensus that this is the unknown in the 
middle. These are the variables in the middle. But there are 
scientific facts that do, in fact, establish and are widely 
accepted, in spite of what some person may say. Unfortunately, 
too often, we argue--let us take climate. For so many people, 
climate is not in the center. Climate is on the outside. This 
is known. And, in part, because I think it got politicized 
before there was ever an evaluation of the science. You see 
what I am saying? Because the public, and, I think, opinion-
makers did not come to consensus on the science or on the 
variables before there was a lot of discussion about policy 
initiatives.
    And so it becomes a political issue as opposed to a 
scientific issue, and that is the challenge that we have, which 
is when does it cross the Rubicon? Can we achieve some kind of 
analysis? And Senator Lankford and I are struggling with this. 
How do we respect other opinions but also narrow the focus of 
what we consider scientific proof or sound science?
    And so with that, that is more rhetorical than anything 
else. I will allow Senator Daines to have a moment here.
    Senator Lankford. Yes. I recognize Senator Daines.

              OPENING STATEMENT OF SENATOR DAINES

    Senator Daines. It has been a great discussion. Thank you, 
Mr. Chairman and Ranking Member Heitkamp.
    I just would submit, for the record, that Exhibit A and B 
for the flat Earth thought would be probably in North Dakota 
and Oklahoma---- [Laughter.]
    And I will bring the contrarian opinion here, being from 
the Rockies of Montana.
    Well, thank you for testifying before the Committee today. 
I guess if you want to have a fun fact this morning, I am the 
only chemical engineer that serves in Congress, out of 535 of 
us here on the Hill. There are several engineers. I get to hold 
down the chemical engineering caucus of one. So I truly have a 
fundamental appreciation for the proper role of science in the 
Federal rulemaking process, because it should, above all, be 
objective, meaning two things above all. One, data should be 
able to stand up to public scrutiny, and, two, it can be 
replicable.
    When I worked at Procter and Gamble--I spent 28 years in 
the private sector before I came here to this new day job, 
where I wear a tie every day--our company used science to meet 
the needs of the mass consumer market, ultimately studying the 
needs of our consumers and rigorously developing a solution to 
meet that need. That is what made our business successful. It 
was the best type of sound and objective peer review you could 
ask for.
    As Ms. Dudley mentioned in her testimony, science can only 
tell us what is but not what ought to be, and for that the 
rulemaking process requires human judgment, which has run amok, 
in my opinion, as regulators continue to conflate science with 
their own personal judgments.
    I know the issue of climate came up. I will give you an 
example of that. I was in a hearing last year where we had Gina 
McCarthy, the Director of the EPA, come and testify. We were 
talking about the Clean Power Plan. And the Clean Power Plan, 
those regulations were going to affect the State of Montana 
more than any other State in the United States. It was going to 
have a devastating effect on our jobs--studies done by the 
University of Montana--our jobs, tax revenues, economic 
activity, double the increases in utility rates.
    So I asked Gina, I said, ``Cato ran the regulations through 
an algorithm called Magic, where you take the regulation, it 
then determines what the projected reduction in CO2 would be, 
and then projects what the impact would be on global 
temperature, which ultimately these regulations were means to 
an end, in terms of trying to reduce CO2 and reduce the impact 
on temperature.
    And when Cato ran those numbers, the answer came back, 0.02 
degrees centigrade, which is negligible. It is a rounding 
error. And I challenged Gina on that. I said, ``Gina, I 
understand what we want to do with these regulations and we can 
quantify the impact on people. We can also quantify the impact 
on the science on climate.'' And I was hoping she would refute 
my data, and tell me why I was wrong and why she had other 
studies.
    But in this limited peer review I had, from one senator to 
an EPA director, she did not refute the data. In fact, she 
said, ``We need to do this because it is almost a moral 
obligation to show the world, and show China, and India, and 
other places, how we are leading in that effort,'' to which I 
responded, pragmatically, ``Well, if you we cede the leadership 
of technology for coal-fired plants to the Chinese and to 
India, we will not have a better outcome. We will have a worse 
outcome, environmentally speaking, because we do a better job 
in the United States of protecting our environment, and sure, 
we could develop clean coal technology, innovate it, and affect 
the entire world.''
    So I think this is an example, again, where we need to get 
into the science and not be afraid to debate the science, 
debate the facts and the figures here, to ensure that we have 
appropriate regulations and outcomes.
    Ms. Dudley, in your written testimony, you mentioned that 
Congress too often cedes power to regulators, allowing them to 
set policy objectives. An example you offered was the 
difference between the Clean Air Act and the Safe Drinking 
Water Act (SDWA), and how Congress required agencies to 
consider economic costs in the Safe Drinking Water Act.
    So my question is, would you say that agencies are 
currently compliant with statutory guardrails like requirements 
to consider economic costs and benefits?
    Ms. Dudley. Every President, going all of the way back to 
Carter, but more explicitly Reagan and Clinton, set in place 
standards that required agencies to look at the benefits and 
the costs of new regulations, and try to maximize the net 
benefits--get more benefits than cost, whenever possible. But 
it is subject to statutory constraints. The Safe Drinking Water 
Act is explicit and says you should look at the benefits and 
costs. The Clean Air Act, especially the ambient air quality 
standards, is not. In fact, it has been interpreted as saying 
that you cannot consider the costs of achieving the standard.
    This gets back to what I was saying to you, Senator, 
Heitkamp, regarding the blurring of the lines. In the Clean Air 
Act we pretend that we are making our decision purely based on 
public health, because we are not allowed to consider costs. 
But nobody really thinks that is true because we do not set the 
standards at zero. In the Safe Drinking Water Act, there is 
much less acrimony, much less argument over whether the science 
is biased or not, because we can be clear, here is what the 
science says; now let us do some tradeoffs in setting the 
policy. And I think, to the extent statutes can look more like 
the Safe Drinking Water Act, that allow us to be honest about 
where the science ends and where the policymaking begins, that 
would improve the analysis and also the decisions.
    Senator Daines. Yes. If Congress were to have established 
these economic guardrails for agencies over past Congresses, do 
you have any estimate of how much cost it would have saved the 
economy, or how many jobs, perhaps, would have been created?
    Ms. Dudley. There are just so few good estimates of the 
real impacts of regulation, either benefits or costs, but I 
could just give you an example. An EPA rule--not the Clean 
Power Plan but another rule that affected electric utilities, 
the Mercury and Air Toxics Rule--the Supreme Court sent that 
back to EPA because that statute set standards that are 
appropriate.
    Senator Daines. Yes.
    Ms. Dudley. And the Court said, ``It's not appropriate to 
have $9 billion in costs to achieve hundreds of thousands of 
dollars in risk reduction.'' So, in that case, the Supreme 
Court said, when the statutory language says appropriate, it 
does mean this balancing of benefits and costs.
    Senator Daines. Yes.
    Ms. Dudley. But yes, if Congress were more explicit, I 
think that $9 billion you would have been able to save.
    Senator Daines. Yes. That is a pretty good number. Well, 
and I think you mentioned the Clean Power Plan. Certainly a 
little over a year ago the Supreme Court intervened, as well, 
on that.
    I want to talk about the redacting PII. I think that came 
up for public access.
    Oh, I am out of time, I see, Mr. Chairman. Are you OK?
    Senator Lankford. I will ask for unanimous consent to give 
you 12 more seconds. [Laughter.]
    That is fine.
    Senator Daines. I am having so much fun here, I am looking 
at----
    Senator Lankford. It is all right. Go ahead.
    Senator Daines [continuing]. The red light was on here, 
Senator.
    Regarding redacting PII for public access, Mr. Rosenberg 
spoke in his testimony that the public should not have access 
to scientific information used for regulations in order to 
maintain confidentiality. We do not use reams of paper anymore. 
We do use Excel and other forms. Dr. Beck, cannot sensitive 
information be redacted, since most data is electronic?
    Ms. Beck. Yes. I think sensitive information can be 
redacted, as well as protecting confidential business 
information, trade secret information. There are ways to share 
the data without sharing these important details.
    Senator Daines. So it seems that researchers, can data 
mine, they can slice-and-dice numbers to perhaps draw their 
preferred conclusions. And back to that important point of peer 
review and scrutiny, do you believe there is wisdom in numbers 
on this one, and public scrutiny of scientific information 
could perhaps foster better regulatory rules and outcomes?
    Ms. Beck. Reproducibility, to me, is critical. Right? So 
often times you need the underlying data to actually reproduce 
the analysis. We have had situations at the American Chemistry 
Council where we did not understand an analysis that was 
presented by EPA, and at first we worked with the journal 
researchers, to try to get that data. We worked with the agency 
to try to get that data. Eventually--and this took years--we 
were able to get the data through a data transfer agreement 
with the National Institutes of Health (NIH), and they 
protected whatever information needed to be protected, and we 
were able to conduct a re-analysis of that data. And that re-
analysis came out differently than the agency's analysis, so 
now we can actually have that important scientific dialogue to 
understand the data. So I think that access to information is 
really important.
    Senator Daines. Thank you.
    Senator Lankford. Thank you. So this whole conversation is 
the challenge that we have back and forth between, we want to 
get good information, whether it is science, or whether it is 
facts, or research, or whatever it may be, in any area, because 
the policymakers need that, and the agencies need that. So the 
conversation today is really all about how do we get that 
balance between getting good science, but understanding we are 
not asking the scientists to be the policymakers, and in a 
strange way, we ask the Federal Bureau of Investigation (FBI) 
to go research, to go get everything together, and then they 
give all of their information to the Department of Justice 
(DOJ), and the Department of Justice determines whether to 
prosecute. The two are kept separate, so you do not have the 
people doing all of the research and all of the gathering of 
information also making the final decision.
    In a strange way, science and policymakers are the same 
way. We ask folks in science, ``Do the research. Get the 
information.'' But we have to make some final decisions on this 
on how to be able to balance it, but we want to make sure that 
is done in the same in the agencies. So the challenge is, how 
do we pull this together?
    Professor Dudley, you made a comment that I would like to 
be able to highlight more. You did not talk about it much, 
during your oral testimony, but you did in your written 
testimony--about risk assessments, and getting multiple options 
and opportunities, knowing that science does not always agree 
on some things. So giving kind of a larger window view of that. 
Can you go into a little more depth on that?
    Ms. Dudley. Yes. I think that would be very important, and 
because the model----
    Senator Lankford. Describe what you mean by the risk 
assessments, and then take us through it.
    Not everyone in the room has read your written statement.
    Ms. Dudley. Well, you are missing something. [Laughter.]
    In developing, especially environmental and public health 
regulations, we go through a process where agencies assess the 
risk and then they try to manage the risk. And back in 1983, 
the National Academy of Sciences said, conceptually, we should 
think of those as two separate things-the risk assessment 
process and the risk management process, because the risk 
management process takes other factors into account--the 
economics, law, what other regulations are in effect, political 
goals--whereas the risk assessment is the scientific inputs. 
Again, risk assessment tells us what ``is'' versus what 
``should be.''
    But I think we have all talked about this. It is important 
to recognize that even in that risk assessment phase, it is not 
as straightforward as ``the Earth is flat or not.'' There is a 
lot of uncertainty. There are different models that predict 
different things, different data available, different 
assumptions used to apply those data to the situation that we 
are concerned about. And that is where a lot of these hidden 
policy judgments can get hidden.
    So one of my recommendations--and this is not my 
recommendation but it comes from other people in this field--is 
that it would be better if the scientists were to lay out, 
using this set of assumptions, we get this number, a different 
set of assumptions provides different numbers. Lay that out, 
and then the policymaker can look at those different estimates, 
and that is the policymaker's job. I mean, it is your job. It 
is the head of EPA's job. Knowing this is a range of estimates, 
what is the right policy?
    Senator Lankford. You have worked with us before, at OIRA, 
and you are familiar with the process. What would that look 
like, as far as an executive action or a codification or--how 
would you consider something like that fitting into how we do 
our government?
    Ms. Dudley. That is a great question. It is so much harder 
to put it into a statute.
    Senator Lankford. Correct.
    Ms. Dudley. So in 2007, we issued--I forget what it was 
called--the Office of Science and Technology Policy and OIRA 
issued a joint paper that Ms. Beck was very involved in, on 
risk analysis in general. Not just the regulatory process. And 
that laid out some things that really had been based on a set 
of guidelines issued in the Clinton Administration. Those have 
withstood the test of time. I wonder whether that would be 
something that might be worth fleshing out a little bit more.
    Senator Lankford. OK. That is interesting to be able to 
note.
    Dr. Rosenberg, you have mentioned a couple of times that 
you feel like it will especially squash science and innovation 
if best available science is codified at some point. So I want 
to delve into that a little bit more, of the why, because as I 
read through it, my competitive nature says if it is the best 
available science and it is new science coming up, it would 
drive them to be able to show the validity of it, multiple 
locations to be able to see how it has been used, and to show 
this science, this technique, this modeling is better than that 
modeling, and to be able to push them toward that.
    You seem to imply that if we do best available science that 
will actually push all innovation out and only old science, 
that has been well-tested and has, as you have mentioned, 
weight of evidence from being used over and over again, will be 
the one that is used. How do we hit a balance on that, because 
this is not theory anymore? This is already being codified by 
Congress. It is in the Toxic Substances Control Act (TSCA) Bill 
already, of a standard for best available science. It has been 
in Executive Orders (EO) for a very long time.
    Mr. Rosenberg. Thank you, Mr. Chairman. I think in TSCA and 
in other bills, there is some description of process, but if 
you codify, and, particularly, then becomes judicially 
reviewable, then people will retreat back to what is the most 
cautious interpretation in many cases. That is what I believe 
agencies will do. They will be careful about the information 
that they consider as part of that weight of evidence, and that 
means that you would be less willing to consider new 
information or information coming from other sources that might 
be critically important.
    I believe that you can certainly, through your oversight 
responsibilities, hold agencies to account that they have used 
a very effective process. My concern is, in legislation, then 
it becomes a much more rigid vehicle. So it is not that I do 
not think that there is no such thing as best available 
science. I described what I think the hallmarks of best 
available science are in my testimony.
    I do not think it is as rigid, and if I can refer to 
Professor Dudley's comments on risk assessment, as there is a 
good example. The view of how you do risk assessment in 
different fields has, in many cases, changed through time. You 
could legislate exactly how you should perform a risk 
assessment, including how do you paramaterize the models, if 
you like. Now you are getting down deeply into how do you 
actually conduct the science.
    On the other hand, there may be much better ways to 
approach that problem that emerge next year, the year after, or 
in a different field, and you want to make sure that people are 
able to do that and not say, ``That is not how we do it.''
    In TSCA, it seems to me, and you are obviously an expert on 
TSCA, but there is an opportunity to use a structured system, 
but it does not apply across many other fields. Even the term 
``replicable'' or ``reproducible''--what does that mean to me 
as a scientist who does field studies on marine biology?
    Senator Lankford. You have to wait for that same fish to 
swim by again, in the same spot. [Laughter.]
    Mr. Rosenberg. Yes. I would ask the fisherman to catch him 
again.
    Senator Lankford. Yes. We need to know--and I think the 
oversight portion of this feeds into what you just mentioned, 
that when people are looking at it, it is not just closest 
science, or the sciences in the cubicle next to you that was 
most convenient for the agency----
    Mr. Rosenberg. Absolutely.
    Senator Lankford [continuing]. Or it is someone that I 
graduated with, but there is an actual opening up to say, let 
us ask the broader question. Let us ask the entire science 
community. And going back to your comments earlier, about peer 
review--it seemed like you were very engaged in the issue of 
oversight on peer review process--to be able to ask the broader 
question, for us, as policymakers, to say, have we looked at a 
broader group of science? And just because you used them last 
year does not mean that this is the best available science this 
year. It could be someone besides your brother-in-law that has 
the best available science this year. Let us open this up.
    Mr. Rosenberg. I entirely agree that the mechanisms for 
ensuring that the science is not only best available but--and 
is independent, are critically important, and those are peer 
review, disclosure of conflicts of interest, ensuring that you 
are bringing in information from other fields, all of those 
things, that are done in understanding that all of this science 
is scrutinized through a public process as well.
    Senator Lankford. That is the ultimate peer review.
    Mr. Rosenberg. There are many aspects of peer review that 
occur in a regulatory process.
    Senator Lankford. But the question is, when you are going 
for best available science--and I am going to go to Senator 
Carper here in just a moment--but when you are going at best 
available science, our goal of that, and what I hope we can do, 
and what I hope we did in TSCA--TSCA was so well-received it 
was voice-voted here in the Senate, as putting in statute, for 
the first time, something that has been in executive action for 
a long time, to say let us look at how do we open this up to as 
many people in the science community as we can, and ask the 
EPA, in this setting, to be able to look specifically at who 
you have used, but is that the best method to use, and go 
research that out, and go ask the real questions of it, and 
then to be able to put that in that language, and also, as you 
have talked about before, the weight of evidence, which we will 
come back and talk about in a moment again. That becomes 
extremely important.
    What we are trying to figure out is, if this works in TSCA, 
does it work in other instances, to ask for the data, to ask 
for the method, to ask for the model, to do the transparency, 
to do, as you have just mentioned, for the public to be able to 
look at it and do the ultimate peer review, and to be able to 
evaluate and ask the hard questions that need to be asked, and 
know it is going to stand up, that the scientists are willing 
to be able to say, ``This is good science. We put our science 
out. Policymakers made their decisions based on that,'' but at 
the end of the day, people can look and review the science. And 
so if you disagree with a policymaker, you can also go back and 
ask them, ``Where did you get this thought to come from?'' and 
you can also go back and disagree with them, and argue through 
the facts as well.
    Mr. Rosenberg. All of those principles I agree with, but I 
would just add you also need to make it workable, because, of 
course, you cannot wait to make a decision about a toxic 
chemical, about a resource or anything else, until you have 
such broad engagement. And there always, of course, will be 
somebody with a contrarian view.
    Senator Lankford. Sure.
    Mr. Rosenberg. And so it does have to be a workable system 
that agencies can actually manage. Otherwise, they will not 
actually carry out their mandated results.
    Senator Lankford. I could not disagree more, but even in 
emergency situations, where we have seen agencies step in and 
say, ``This is so vitally important. We need to engage right 
now for public safety and health,'' there are mistakes even in 
that, and we need to allow for humanity to exist.
    One of the most obvious examples was recently, when 
agencies rushed in after the fertilizer plant explosion in West 
Texas, and it was immediately, we need to make these massive 
changes around the country, and then, 2 years later, the Bureau 
of Alcohol, Tobacco, Firearms and Explosives (ATF) finished 
their investigation and discovered that was not an accidental 
explosion. That was an intentionally set fire. Well, that 
changed everything on the premise, that for 2 years the 
agencies had run with a false premise.
    So I understand everyone makes mistakes in the process on 
that, but good oversight helps us in the process as well, and a 
good set of boundaries for it.
    We will come back to this. I want to recognize Senator 
Carper.

              OPENING STATEMENT OF SENATOR CARPER

    Senator Carper. Thanks. Thanks so much, Mr. Chairman. 
Senator Heitkamp, good morning. Nice to see you. To our 
witnesses, Susan, welcome back. You have been here before. You 
should probably pay us by the appearance. [Laughter.]
    Dr. Rosenberg, Dr. Beck. I have a specific question for Dr. 
Beck. Before I do, I applaud our Chair and Ranking Member for 
holding this hearing. We are grateful to our witnesses for 
coming.
    I have a question for Dr. Beck, but before I do that I 
would just like to ask Susan if each of you just take maybe 30 
or 45 seconds, since I did not get to hear your testimony, and 
just maybe give me what you thought would be a point that we 
should really walk away from here with. Give us a good nugget 
or two.
    Ms. Dudley. For me, I think it is that science is essential 
for informing regulatory policy, but it cannot decide 
regulatory policy, or policy, in general, and that there are 
other disciplines that are also important. And so often, when 
people accuse others of politicizing science, they really are 
talking about decisions that science can inform, but not 
decide. And if we could do a better job of being clear where 
the science ends and where the policy begins, I think it would 
open up to much more rational debate on the policy side, and we 
would not impugn science inappropriately.
    Senator Carper. All right. Good. Thank you. Dr. Rosenberg.
    Mr. Rosenberg. I would actually agree with that. I would 
also say that we should--when we are talking about concepts 
such as best available science, we should really focus on the 
process of developing that science. It is not a matter that 
everyone goes and redoes the analysis, that it is the science 
through a credible process, contributes to a weight of 
evidence.
    And then there is a judgment call on a policymaker's side--
is the weight of evidence enough to take action? But we should 
trust that there is a strong science process and continue to 
strengthen it with elements such as scientific integrity 
policies and transparency policies.
    Senator Carper. Good. Thank you. Dr. Beck same question.
    Ms. Beck. So I agree with Susan Dudley about--both of 
them--about the separation of science and policy, and I think 
this is why it is important that we be really transparent about 
the quality of science. And the confusion about whether or not 
to codify a definition is--I just do not understand it. I think 
that people should commit to using an approach--that uses a 
clear criteria to evaluate studies, to ensure they are peer-
reviewed. It does not put the finger on the scale for any 
particular study, but ensures that at that period in time, you 
are looking at all of the evidence and using the best evidence 
to define the science. And then you move it to the policy 
arena.
    Senator Carper. Would it be safe to say that the three of 
you agree that we should not be blinded by science?
    I think everybody nods yes. Thank you. Thomas Dolby would 
appreciate you saying that.
    Dr. Beck, my colleague has been talking about Toxic 
Substance Control Act and his compadre from Oklahoma, Jim 
Inhofe, following David Vitter from Louisiana, Tom Udall, 
myself, and a bunch of others, worked for years to enact the 
full legislation. And I think if they ever remake the filmstrip 
or civics video about how a bill becomes a law, I would 
nominate TSCA for the Oscar, and say--it took a long time, but, 
in the end, it was just really nicely done.
    The reason it was important that Congress act last year, 
was that TSCA had been essentially a broken law since 1991. I 
think some would say even before that, where an industry, I 
think in 1991, successfully sued to overturn the Bush 
Administration's proposed ban on asbestos. And they were able 
to do that because the old law required EPA, as you may recall, 
to choose the ``less burdensome'' regulation for industry, and 
industry argued that the EPA had not evaluated the costs and 
the benefits of all possible alternatives to a ban, even though 
EPA had spent, I guess, a decade or so writing the rule and had 
prepared thousands of pages of analysis.
    The EPA concluded that TSCA was unusable, and for decades 
it could not regulate the safety of chemicals it knew to be 
dangerous.
    Would you agree that the requirement that EPA select the 
``least burdensome'' regulation is a major part of why old TSCA 
was unusable and a major part of why we all worked so hard 
together to reform it, Ms. Beck?
    Ms. Beck. Yes, that language was very problematic.
    Senator Carper. If you want to say more that that language 
was very problematic, you can.
    Ms. Beck. Well, the language does not----
    Senator Carper. We do not charge you for testimony. We do 
not pay you either. [Laughter.]
    Ms. Beck. The language does not exist anymore, right? I 
think there was full agreement that the language confused the 
science and the policy. So now you have an unreasonable risk 
standard in the new Lautenberg Chemical Safety Act, which is 
based purely on the evaluation of the science. It looks at the 
hazards and the exposures under very specific exposure 
conditions, and then it makes a science determination. And then 
separate from that, there will be some risk management steps 
that consider costs and benefits, but there is no requirement 
on the agency to choose the least burdensome.
    Senator Carper. OK. When we wrote the new law over the last 
couple of years, we told EPA to--try to first figure out 
whether or not a chemical was dangerous, and then to consider 
the cost when the agency was deciding how to protect people 
from whatever risk the agency had identified. We did not tell 
EPA, though, that it had to study the costs and benefits of 
every possible regulation, and we did not tell EPA that it had 
to choose the cheapest option. We told EPA it had to protect 
the public against unsafe chemicals and consider costs when it 
did so.
    Let us just say, for a moment, that EPA proposes a 
regulation under TSCA that costs industry, we will say, a 
billion dollars. A lot of money. And industry tells EPA, during 
the required comment period, that there is an alternative that 
will protect people just as well, that costs only a million 
dollars. A million versus a billion. Is it not true that the 
Administrative Procedures Act (APA) requires EPA to consider 
and respond to industry's views on cost when they write the 
final rule? And that could be, again, you, Dr. Beck. Is it not 
true that the Administrative Procedures Act requires EPA to 
consider and respond to industry's views when EPA writes the 
final rule?
    Ms. Beck. So I am not an expert in the Administrative 
Procedures Act, but yes, the rulemaking process does require 
that the agency consider the public comments and provide 
responses to those comments. It does not require that the 
agency accept those comments but they have to consider them, 
discuss them, and explain how they were considered.
    Senator Carper. OK. Thanks.
    Ms. Beck. Yes.
    Senator Carper. Is it not also true that if EPA ignores 
that cheaper, equally protective measure in its final rule that 
industry could also sue and overturn the rule, under the 
Administrative Procedures Act? And if you do not know the 
answer to that question, maybe one of the other--your 
colleagues would. Dr. Rosenberg. Susan?
    Mr. Rosenberg. Well, it is probably more Susan's area than 
mine, but I would say that frequently happens with a claim that 
it is arbitrary and capricious, which is the language of the 
Administrative Procedures Act, and that was drummed into my 
head when I was a regulator, because that is in every challenge 
to a regulation--you have been arbitrary and capricious, which 
goes back to the role of science. So, yes, that is what APA 
does. It gives you that opportunity.
    Senator Carper. Yes. Susan? Same question.
    Ms. Dudley. Yes, that is right. Anybody could sue if the 
agency did not base their final regulation on the material in 
the docket, which would be comments from the public, the data 
they have, et cetera.
    Senator Carper. OK. Good. Who succeeded you at OIRA? Was it 
Cass Sunstein?
    [No audible response.]
    One of the things that President Obama asked him to do was 
to--I do not know if it had been done before, but I think they 
called it a look-back--to look back at regulations that had 
been adopted, in some cases many years ago, that may have 
served their purpose but did not anymore, and to see which ones 
should be saved, which ones modified, and which ones gotten rid 
of.
    I have read that that was a pretty successful endeavor. I 
believe it went on throughout the rest of the Obama 
Administration. Do you have any recollection of that?
    Ms. Dudley. You are right. It was actually something that 
every President since Carter has asked agencies to do.
    Senator Carper. Yes.
    Ms. Dudley. And yet they really do not do a good job of it, 
for several reasons. One is that the incentives are not there. 
It is much more interesting to look at the next problem to 
solve rather than looking back, and even regulated parties 
often are not interested, because--especially if there are 
investments they had to make, the last thing they want is for 
the agency to say, ``Oh, never mind. We should not have done 
that,'' because then their competitors will have an advantage.
    Senator Carper. Yes.
    Ms. Dudley. But it is also hard to do. So it is not----
    Senator Carper. Did you do it while you were running OIRA?
    Ms. Dudley. Yes, and so I was at OIRA for the last end of 
the Bush Administration----
    Senator Carper. Yes.
    Ms. Dudley [continuing]. So I was actually wrapping up 
things that had been begun before me, toward the end.
    Senator Carper. You were there on the clean-up?
    Ms. Dudley. I was there for the clean-up. I was finishing 
things.
    Senator Carper. Oh, that is good.
    Ms. Dudley. But one of the things that I think about 
retrospective review is that it is not just to see what does 
not work and to rescind and reduce costs. It is to find out 
whether our estimates of the risk reduction benefits were 
accurate. It is part of the scientific method that requires 
hypothesis testing, gathering data and testing your hypothesis.
    So I think it is very important. There was a bill that 
Senators Lankford and Heitkamp introduced last year that would 
have required agencies, when writing a new regulation, to plan 
for how they would review it, 5 or 10 years down the road. 
That, I think, would be huge for the science, the risk 
assessment that goes into regulation----
    Senator Carper. Right.
    Ms. Dudley [continuing]. Ex-ante as well as ex-post.
    Senator Carper. You are not just saying that because he is 
sitting here, are you?
    Ms. Dudley. No. Not because I am huge fan of both of yours, 
but no, that is not why.
    Senator Carper. Thank you. And the converse is true.
    Thanks so much for being here and for your help. Thanks. 
Thanks, Mr. Chairman.
    Senator Lankford. No, thank you, and can I finish up this 
conversation as well, when we were talking about finishing and 
closing and cleaning it up, and we will have about another 10 
minutes or so if you all are kind of wondering on time, unless 
we keep going, and then it goes until one. We will see.
    My question for you is this issue of retrospective review. 
How does that work in the science, and I am still thinking 
about the legislative side of this. Let us say you have a 
regulation that is a $100 million-plus regulation that comes 
out. Seven years later you go back and look at it. It is not 
accomplishing what you want. The science has moved, as Dr. 
Rosenberg, you have talked about often. The science continues 
to develop in research. You have better measurement tools and 
you figure out your hypothesis did not work. How do you do a 
retrospective review when there is also a scientific opinion 
that is sitting, that may, at that point, be 10 or 12 years 
old? Are you talking about a retrospective review that would 
also include re-evaluating the science again, or just looking 
at the numbers, or trying to evaluate that?
    Ms. Dudley. I actually think thats one of the under-
appreciated advantages of doing better retrospective review 
would be evaluating whether our scientific predictions were 
right. So it is not just valuable for the regulations that we 
issued 10 years ago, because it may be too late to change 
those, but it helps us the next time we issue a regulation, we 
will know which of our assumptions maybe were not right. So I 
think it will improve the science if we could do a better job 
of it.
    Senator Lankford. Right. And it is the same thing whether 
it is economics or whether it is science of any type, that you 
get a chance to note, this is what we estimated, this is what 
we thought would happen, this is the model, and then we look at 
it 5 years later and say, is the model proving to be correct, 
and sometimes you have not had enough time to be able to 
evaluate it, and sometimes you have.
    Ms. Dudley. It is just an essential part of the scientific 
method.
    Senator Lankford. OK. I want to get into this issue of 
weight of evidence. Dr. Beck, you and Dr. Rosenberg both 
brought this up significantly, both in written comments and in 
oral comments as well, by dealing with the weight of evidence. 
Dr. Beck, you mentioned several spots that you had some 
frustration already with the EPA not looking at weight of 
evidence. I would like to get a chance to talk about that and 
figure out how we address that, because that is one of the 
things in TSCA, specifically in statutory language, that this 
Congress said, ``no, we want EPA to look at weight of evidence 
in this role, and to be able to figure out where that comes 
from.''
    So, Dr. Beck, if you want to talk about how you see that 
being applied right now and then let us try to figure out how 
we actually deal with this.
    Ms. Beck. Yes. So in the Congressional Record associated 
with the TSCA rulemaking, there is a definition for weight of 
evidence, and we would like to see the agency simply adopt that 
definition. It basically would commit them to looking at the 
studies and evaluating their strengths and limitations using a 
systematic review process which is a standardized process for 
essentially evaluating evidence, providing sort of the recipe 
for how you are going to bake your cake before you bake it. 
Everybody knows the criteria that you are going to use, the 
plans that you are going to use for analysis. You put that up 
front. You are clear about how you are going to collect your 
literature and evaluate your literature and do the analysis. So 
once you have evaluated the strength of the evidence, right, 
the individual strength of each individual studies, based on 
the quality of those studies, you weight it all and you make 
your determination.
    So that is the definition to us that seems so 
straightforward and so consistent with the scientific process, 
we think it should be adopted into the regulations that EPA is 
writing, to implement TSCA.
    Senator Lankford. OK. Are there other definitions out there 
for weight of evidence that you would recommend, other than 
what was done in the Congressional Record?
    Ms. Beck. The term has been confused over the years. Some 
people use a strength of evidence approach. That is very 
different and we do not support a strength of evidence. That is 
how many studies are positive, how many are negative----
    Senator Lankford. Right.
    Ms. Beck [continuing]. And oftentimes people discount the 
negative studies. We only think the negative studies should be 
discounted if they are of bad quality. If they are of high 
quality they should be considered equally.
    So, to me, the cleanest approach is to adopt the one that 
the Congressional drafters had, and there was support for that 
definition in the House language as well as in the Senate 
language.
    Senator Lankford. OK. Dr. Rosenberg, any comments about 
just the weight of evidence?
    Mr. Rosenberg. Well, I would just say that the weight of 
evidence is more than just tallying up the number of studies, 
and I do not think that you can equally weight every study. And 
so I do not disagree that there should be clear guidance from 
the agency, in terms of--so that people can understand how the 
agency will interpret information going forward. But I would 
caution against saying, well, if it is a credible study then 
everything gets equal weight, because that is not how you would 
do it as a scientist, certainly. You would consider the 
uncertainty. You would consider a lot of factors around the 
study and weight things appropriately.
    Senator Lankford. Back to the range of options, Dr. Beck, 
you had mentioned one of your proposals, grant funding not 
based on positive outcome.
    Ms. Beck. Yes. So, in the scientific world there is this 
concept of publish or perish, and that if you do not have a 
strong publication record you are not going to get promotions 
in academia and you are not going to get tenured positions. And 
in the journal world, there is an incentive to only publish 
data that are positive. People do not like to publish negative 
studies. People have talked about creating journals of negative 
data so that people are not reproducing the same studies that 
fail over and over again because the public does not know that 
they fail, because nobody is publishing them.
    So if there was a way to, at least within the Federal 
Government, to encourage the publication and the presentation 
of negative data, that would be extremely helpful.
    We are aware of cases, for instance, at the National 
Institute of Environmental Health Sciences (NIEHS), where they 
have done some studies that are actually negative, and we have 
been talking to them, can you please publish these data? These 
are important studies. And at this point in time there are not 
incentives to them to release negative data as much as there is 
to release positive data.
    Senator Lankford. But when you talk about grant funding 
based on it, the grant funding should be, at the beginning of 
it they do the study, not at the end of it. Somewhat this 
sounds like----
    Ms. Beck [continuing]. Yes----
    Senator Lankford [continuing]. You want to make sure your 
study ends up positive at the end if you have got grant 
dollars, if you want to get more grant dollars.
    Ms. Beck. If you want to get more grant dollars. That is 
correct. You have a 5-year grant and you want to get your grant 
renewed, or you want your center renewed, you want your center 
to have positive findings or it is not likely to get renewed by 
a Federal agency.
    Senator Lankford. So is it the fear that the science either 
gets limited--there were, for lack of better terms, 10 things 
we learned not to do, one thing we learned to do, and we only 
published the one because you want to show positive, or is your 
fear that the science is going to make sure that they try to 
push it and steer it toward something positive, regardless?
    Ms. Beck. Well, it is actually both. The official term is 
publication bias, and that is that what exists in the 
literature are mostly positive findings. So then when people go 
to evaluate all of the evidence in totality, they are going to 
have a lot more positive studies than negative studies, and 
then what you cannot evaluate does not get to inform your 
regulation, right, so the positive studies are then informing 
the regulation and the negative studies, which are not 
published, are not known.
    Senator Lankford. OK. Dr. Rosenberg.
    Mr. Rosenberg. Just very briefly, while I agree that 
occurs, it is not restricted to grant-funded research. Of 
course, industry does the same thing. Clinical trials are the 
classic example of that. You do not put forward the clinical 
trials that did not show the positive effect. And so there is a 
problem of those incentives. It is all across industry and it 
is also in academia.
    Senator Lankford. OK. Both of you also mentioned the issue 
about peer review, and if there is a legislative--in fact, Dr. 
Rosenberg, I think you said if there is an area to address in 
legislation that peer review process or trying to improve that 
was one of the areas. But regardless of whether it is 
legislative or not, what would be your recommendations on 
trying to deal with the peer review process, or how we label 
this, this is a peer-reviewed study, and still determine the 
quality of that?
    Mr. Rosenberg. Well, I think I said that both scientific 
integrity and peer review were important areas for progress. I 
think the independence of peer review is critical, where you 
are actually getting the peer reviewers, and the breadth of 
that peer review. This is actually not the same as a peer 
review you do for academic journals, where, there is a limited 
number of reviewers and then you can move forward with 
publication. Obviously, in regulatory work, you have to have a 
more intensive process, and that goes to the selection of peer 
reviewers and so forth.
    I think, from a legislative perspective, it is important 
that agencies be clear about what their peer review process is, 
and how they select reviewers. We should be careful that 
reviewers are not representatives of different interests, so 
the idea that has been in some quarters in science advisory 
boards, that you should have representational members. That is 
a bad idea for peer review. You should be there because of your 
expertise in particular areas and you contribute to the review. 
There are questions in regulatory review about whether it 
should be anonymous review or not. The most important thing is 
that there is a clear and transparent, process for how we are 
going to conduct peer review. People can know that we did it--
--
    Senator Lankford. So let me press on that a little bit. 
When you are talking about representational review, do you have 
an issue, for instance, if you were doing--let us say we are 
testing safe drinking water, that EPA is both doing their own 
sample with their own agency folks. Industry that may be nearby 
that they know that they are suspect, should they be able to be 
at that site, be able to pull from that same--if you are going 
to pull 10 gallons of water out and be able to examine it, they 
pull a quart out of it as well, out of that same group, so they 
are able to run it through their lab? So you have multiple 
competing labs all checking it, because as you know well, there 
are some differences in all of the testing there. Do you have a 
problem with that type of peer review, where it is a group that 
could be affected by it, but they would have to be there onsite 
to be able to make sure their sample is consistent?
    Mr. Rosenberg. So I do not have any problem with having, as 
you describe, different entities involved in the sampling. What 
I would say is that if you are going to then apply peer review, 
which is independent, outside experts who look at the results, 
and look at the methodology, and do all of the things that you 
do in peer review--this is not, again, the raw data, but how it 
was done--that that be done not just for the agency analysis 
but it also be done then if the affected industry says, oh, 
well, we got different results----
    Senator Lankford. Show your data.
    Mr. Rosenberg [continuing]. It should have exactly the same 
peer review standards.
    Senator Lankford. Yes.
    Mr. Rosenberg. The difficulty I have had in regulatory 
settings is that I might go through, 20 public meetings on 
scientific analysis for, again, marine resources. I know you 
have many of those in Oklahoma. And someone else would stand up 
and say, ``Well, my results are different.'' You have 
absolutely no idea whether that has been reviewed or who has 
reviewed it. It is not meeting the same standard at all. So you 
have to compare like with like.
    Senator Lankford. Yes. You will be glad to know that 
Oklahoma has more fresh water shoreline than any other State in 
the union.
    Mr. Rosenberg. I am really happy to know that.
    Senator Lankford. We dig a lot of ponds. [Laughter.]
    So we do that. We know how to store our water.
    Let me ask for your quick response on that as well, and 
then I want to try to move on to a final question.
    Ms. Beck. Yes. Peer review is an area that I think could 
benefit from a lot of improvements. One of these issues 
relating to representative, equal review of all data is 
funding. I listened to a peer review panel where there was a 
very good study funded by industry, and when it came up one of 
the peer reviewers said, ``Wait. Was not that funded by 
industry? We are not going to consider that.'' It was just 
completely discounted, simply based on its funding, and there 
was no evaluation of the quality of the data.
    So, again, I think these peer-review panels, of course you 
need to address conflicts and biases, but you need the breadth 
and the depth and the expertise. If you are reviewing something 
about a manufacturing process and you have nobody that 
understands the manufacturing process, I do not think that peer 
review is going to end up being very good.
    In the FDA, they have ways of having experts that might 
have conflicts on panels to help educate the panels, but they 
do not vote. Senator Carper referred to the Oscar-winning TSCA 
rule. The TSCA legislation actually requires different 
representations on the review panel, and those representations 
include animal welfare, labor, industry. In putting that panel 
together for the industry representation, EPA has chosen two 
people from the pharmaceutical industry. These people do not 
understand the chemical manufacturing process and TSCA does not 
regulate pharmaceuticals.
    So I think you have to reach a point where you ensure that 
you have the right expertise and breadth to do a high-quality 
peer review.
    Senator Lankford. Yes, which makes it a challenge after the 
fact.
    Ms. Beck. Yes.
    Senator Lankford. The TSCA language on the science section, 
do any of the three of you have an issue with that, how it was 
written for that area, as much as you know the language? I am 
not going to ask you to know every word of it, but as much as 
you know. Dr. Beck.
    Ms. Beck. So I would just say that that language, it is 
consistent with what is in the information quality guidelines. 
It is very consistent with what is in the Safe Drinking Water 
Act. It should not be new to the agencies.
    Senator Lankford. No, it has been in existence for a long 
time----
    Ms. Beck. Yes.
    Senator Lankford [continuing]. As Executive Orders and 
other actions.
    Ms. Beck. It is not new. It is around, so it is actually 
nice to see it in there, and we hope that when EPA finalizes 
their framework rules under TSCA they will commit, to meeting 
those standards in their rules.
    Mr. Rosenberg. Chairman, I do not know the language in that 
little detail. I would just refer back to one comment that Dr. 
Beck made, about representation on a peer-review panel. That is 
exactly the problem of having people who are there to represent 
a particular interest. They are no longer peer reviewers. If 
they are there because they have particular expertise, I do not 
have a problem.
    Senator Lankford. Right.
    Mr. Rosenberg. But if you say, ``I am here as an industry 
reviewer. I am here to represent the interests of industry,'' 
or non-governmental organizations (NGO's), then I have a real 
problem. That is not a peer review to me.
    Senator Lankford. Right.
    Ms. Beck. I agree with you. It should be expertise.
    Senator Lankford. Right. So let me ask this question. Is 
there an issue with trying to get the data, the models, and the 
methods out, so when an agency concludes their work, you can 
ask the practical question, ``You came to this conclusion as an 
agency head. You based it on this report. Can we see--with 
private information excluded, obviously, and redacted--the 
data, the models, and the methods'' to be able to come to that? 
Susan Dudley.
    Ms. Dudley. I think that is essential, and I will just read 
from Science magazine which requires that now for their 
articles. The editors, when they made the change to make data 
available, said that ``when the greatest number of creative and 
insightful minds can find access and understand the essential 
features that led to the collection of a data set, the data 
reached their highest potential.''
    So I think that is an essential element of the transparency 
that you mentioned in your opening statement.
    Senator Lankford. OK. Dr. Rosenberg.
    Mr. Rosenberg. I do not agree that the raw data is needed 
and I do not think that is what Science requires, in most 
cases. At least whenever I publish there they do not ask for 
the raw data. I think the confidentiality is not just for 
public health records. It is also confidential business 
information, intellectual property. It gets very complicated 
quickly.
    Again, for most studies, you are not reviewing the raw data 
but you are doing exactly what Susan said. You are looking at 
how was the data collected, what is the provenance of the 
study, models, and so on. Does that mean that we would never 
consider results from a proprietary model, including the magic 
model that one of your committee members cited? For evaluating 
the Clean Power Plan we should not consider that because that 
model is proprietary. So there is a whole range of information, 
not just public health information.
    I think that to the extent that data can be released, fine, 
but I do not think that that should be the requirement, and if 
you cannot do it, for multiple reasons, therefore, you do not 
move forward.
    Senator Lankford. How do we have transparency without 
knowing the data, or without knowing the process of how it was 
put together? Because as you know well, the variables within 
data and the assumptions that underlie that can vary 
dramatically.
    Mr. Rosenberg. I have absolutely no difficulty with 
describing, in detail, how the data was collected. The data 
collection process is part of something I review in every study 
that I look at. The modeling process is something that I review 
in every study that I look at, because I tend to mostly review 
modeling studies. I do not review the raw data.
    Now, sure, some people may want to do alternative analyses 
of that raw data. It is almost inevitably going to be for 
regulatory matters, industry. Nobody else would have the 
capability of doing that. But again, then, interpreting those 
results, unless they are put back into exactly the same peer 
review process that the agency used, is almost impossible, and 
the difficulty then is you have gotten into a very difficult, 
very long process of review with almost no end, because anyone 
putting in a new process starts a new round of peer review and 
you just stay in peer review forever.
    So I think you need to be realistic about what the 
information will be. The information to review is the details 
of that process of how that data was collected, certainly, and 
all of the other elements of the study, as I indicated in my 
testimony.
    Senator Lankford. OK. Dr. Beck.
    Ms. Beck. I noted this is in my testimony. The public trust 
is eroding in science. There are articles in the Washington 
Post about how 50 percent of the science that is published is, 
maybe false and not true. And I am not saying every study is 
unreliable, but if you want to have trust and confidence in 
your regulatory decisions, you need to be as transparent as you 
can absolutely be about that underlying data, so that everyone 
is on the same page and can have a scientific dialogue. It 
helps to build trust and confidence.
    So proprietary information, private information, all that 
needs to be protected. There are ways to protect proprietary 
information in models while letting other people look at those 
models. I think there are legal ways data-sharing can happen, 
and to the extent that the scientific community can improve 
that, the public confidence in the science will be greatly 
improved and confidence in regulatory decisions will be 
improved.
    Senator Lankford. Yes, that is a balance, obviously. We 
want people to be able to do their research, to be able to 
protect their data. They have paid for it, they have worked 
through the process, unless it is the Federal taxpayer that 
paid for it, and that is a whole different set of issues.
    But the challenge that we face is trying to trust it and to 
say ``how can we get a chance to do a good evaluation,'' even 
things that are given the tag ``it was in a journal,'' ``it was 
published,'' or ``it was peer-reviewed.''
    One of my favorites studies, when we were going through and 
doing the preparation for all this, and the research, was 
finding Dr. Bohanan's tests from Harvard, the biologist there, 
where he wanted to find out how the scientific journals went, 
and so he created a completely bogus study, with bogus modeling 
and everything else, and presented it to 300 different 
scientific journals, and 157 of them accepted it. And you 
think, OK, there is a challenge here, even within scientific 
journals, on how they accept what is fact and what is, at that 
point, completely bogus information. So we have to be able to 
guard that, and that is part of gaining public trust.
    The one thing I would like to do is be able to maintain 
this conversation. We are seeking a way to be able to solve 
this legislatively. We do not want to overreact and what it has 
done and able to squash the future of science, or to be able to 
compel people to not do research or to be able to put 
proprietary information that should not be put in the public 
domain not in there.
    I would tell you I serve on the Intelligence Committee as 
well. We deal with a tremendous amount of redacted information 
and a tremendous amount of research that can be put aside with 
private information. So I think there are ways to be able to 
accomplish some of this without putting people's private 
information or proprietary information at risk, so we can still 
gain trust.
    So I would appreciate your feedback as we walk through this 
process. The legislative process, as I can assure you, is a 
messy process, and I would appreciate it if you would join us 
in the mess, and any ideas that you have.
    Before we adjourn I do want to announce that on April 6, 
the Subcommittee intends to hold a hearing regarding our 
continuing efforts to address potential problems and solutions 
associated with the Federal workforce.
    This concludes today's hearing. I do, again, want to thank 
your witnesses for being here today. I appreciate knowing your 
work. We did not get to even 10 percent of what you each 
submitted in your written work as well, and I do bemoan that, 
but we are out of time.
    The hearing record will remain open for 15 days until the 
close of business on March 24, for the submission of statements 
and questions for the record.
    I thank all of you again. This hearing is adjourned.
    [Whereupon, at 11:59 a.m., the Subcommittee was adjourned.]

                            A P P E N D I X

                              ----------      
                              
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]