[House Hearing, 115 Congress]
[From the U.S. Government Publishing Office]


CBO OVERSIGHT: THE ROLE OF BEHAVIORAL MODELING IN SCORING AND BASELINE 
                              CONSTRUCTION

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

                                HEARING

                               BEFORE THE

                        COMMITTEE ON THE BUDGET
                        HOUSE OF REPRESENTATIVES

                     ONE HUNDRED FIFTEENTH CONGRESS

                             SECOND SESSION

                               __________

          HEARING HELD IN WASHINGTON, D.C., FEBRUARY 27, 2018

                               __________

                           Serial No. 115-10

                               __________

           Printed for the use of the Committee on the Budget
           
           
[GRAPHIC NOT AVAILABLE IN TIFF FORMAT]           
           


                       Available on the Internet:                   
                            www.govinfo.gov
                            
                              __________
                               

                    U.S. GOVERNMENT PUBLISHING OFFICE                    
30-540                       WASHINGTON : 2018                     
          
-----------------------------------------------------------------------------------
For sale by the Superintendent of Documents, U.S. Government Publishing Office, 
http://bookstore.gpo.gov. For more information, contact the GPO Customer Contact Center, 
U.S. Government Publishing Office. Phone 202-512-1800, or 866-512-1800 (toll-free). 
E-mail, gpo@custhelp.com.                           
                            
                            
                            
                            
                            
                            
                            
                            
                            
                        COMMITTEE ON THE BUDGET

                    STEVE WOMACK, Arkansas, Chairman
TODD ROKITA, Indiana, Vice Chairman  JOHN A. YARMUTH, Kentucky,
DIANE BLACK, Tennessee                 Ranking Minority Member
MARIO DIAZ-BALART, Florida           BARBARA LEE, California
TOM COLE, Oklahoma                   MICHELLE LUJAN GRISHAM, New Mexico
TOM McCLINTOCK, California           SETH MOULTON, Massachusetts
ROB WOODALL, Georgia                 HAKEEM S. JEFFRIES, New York
MARK SANFORD, South Carolina         BRIAN HIGGINS, New York
DAVE BRAT, Virginia                  SUZAN K. DelBENE, Washington
GLENN GROTHMAN, Wisconsin            DEBBIE WASSERMAN SCHULTZ, Florida
GARY J. PALMER, Alabama              BRENDAN F. BOYLE, Pennsylvania
BRUCE WESTERMAN, Arkansas            RO KHANNA, California
JAMES B. RENACCI, Ohio               PRAMILA JAYAPAL, Washington,
BILL JOHNSON, Ohio                     Vice Ranking Minority Member
JASON SMITH, Missouri                SALUD O. CARBAJAL, California
JASON LEWIS, Minnesota               SHEILA JACKSON LEE, Texas
JACK BERGMAN, Michigan               JANICE D. SCHAKOWSKY, Illinois
JOHN J. FASO, New York
LLOYD SMUCKER, Pennsylvania
MATT GAETZ, Florida
JODEY C. ARRINGTON, Texas
A. DREW FERGUSON IV, Georgia

                           Professional Staff

                       Dan Keniry, Staff Director
                  Ellen Balis, Minority Staff Director
                                
                                CONTENTS

                                                                   Page
Hearing held in Washington, D.C., February 27, 2018..............     1
    Hon. Steve Womack, Chairman, Committee on the Budget.........     1
        Prepared statement of....................................     3
    Hon. John A. Yarmuth, Ranking Member, Committee on the Budget     5
        Prepared statement of....................................     6
    Keith Hall, Director, Congressional Budget Office............     8
        Prepared statement of....................................    12
    Hon. Steve Womack, Chairman, Committee on the Budget, 
      questions submitted for the record.........................    50
    Answers to questions submitted for the record................    51

 
CBO OVERSIGHT: THE ROLE OF BEHAVIORAL MODELING IN SCORING AND BASELINE 
                              CONSTRUCTION

                              ----------                              


                       TUESDAY, FEBRUARY 27, 2018

                          House of Representatives,
                                   Committee on the Budget,
                                                   Washington, D.C.
    The Committee met, pursuant to call, at 10:00 a.m., in Room 
1334, Longworth House Office Building, Hon. Steven Womack 
[Chairman of the Committee] presiding.
    Present: Representatives Womack, Renacci, Johnson, Faso, 
Smucker, Lewis, Ferguson, Smith, Sanford, Bergman, Woodall, 
Arrington, Grothman, Yarmuth, Jeffries, DelBene, Schakowsky, 
Jayapal, and Khanna.
    Chairman Womack. This hearing will come to order. Welcome 
to the Committee on the Budget's hearing on the Congressional 
Budget Office's Role of Behavioral Modeling in Scoring and 
Baseline Construction. This is the third of five hearings that 
the Committee plans to hold this spring on oversight of the 
Congressional Budget Office.
    Today's hearing, as we said, marks the third in our five-
part series. As I have mentioned before, we are conducting 
these hearings to learn more about how CBO carries out its 
nonpartisan mandate and to consider potential areas for 
improvement.
    Since its formation in 1974, CBO has been a vital support 
agency to Congress providing nonpartisan budgetary analysis and 
directly assisting the House and Senate Budget Committees. 
While CBO was established by the Congressional Budget 
Impoundment Act of 1974, the agency has not undergone a 
comprehensive review since then--more than 40 years ago. The 
demands on the agency have undoubtedly changed.
    CBO's mission of supporting the Congressional budget 
process remains the same, but, as we have already learned, 
expectations of the agency have evolved over time. So, like our 
last two hearings, we are here to make sure the agency has 
everything it needs to effectively and efficiently fulfill its 
mission in the 21st century.
    We open the oversight series with CBO director Dr. Keith 
Hall to broadly discuss the agency's organizational and 
operational structure including its staffing, assumptions 
processes, and work products.
    In our last hearing we began to explore some of the more 
technical aspects of how CBO actually crafts the impartial work 
products Congress relies on to make informed legislative 
decisions. And today, we will focus on how CBO conducts its 
modeling and scores legislative proposals including the 
assumptions made and processes used by the agencies analysts.
    CBO uses models as a tool for analysts to estimate the cost 
of legislation, project baseline impacts, and produce long-term 
economic analysis. In order to estimate the effects of 
legislative proposals, CBO must make assumptions about the 
likely responses to new policies. There is no doubt a 
challenge, and I know we have a lot to learn today, from the 
experts. With that, I want to introduce our witnesses each 
coming with knowledge and understanding of CBO's current 
modeling framework.
    At the witness stand today is the CBO director Dr. Keith 
Hall. A little later on Dr. Jessica Banthin, Deputy Assistant 
Director for health retirement long-term analysis. Dr. Banthin 
is here to share her expertise on how the agency practically 
applies its health insurance simulation model, as well as other 
models, in the course of cost estimating. And we will be joined 
by Dr. Jeffrey Kling Associate Director for Economic Analysis. 
He is here in his capacity as the Associate Director 
responsible for maintaining and enhancing the quality and 
transparency of CBO's work.
    Before we start our discussions today, I want to remind 
everyone that our goal here is to learn and identify potential 
areas for improvement at CBO. CBO is vitally important to the 
congressional budgeting process. So, these discussions are not 
meant to be one sided or accusatory. Throughout this hearing 
series, CBO has welcomed the opportunity to discuss its work 
and undergo oversight.
    However, it is certainly fair to acknowledge that CBO's 
scoring of legislative proposals has led to those scores being 
questioned--particularly related to healthcare policy. But, 
when it comes to recent efforts to repeal and replace 
Obamacare, the failure of Congress to do so is certainly not 
CBO's fault. However, it is a legitimate reaction of Congress 
to investigate how CBO, as a congressional support agency, 
comes to any of its conclusions and scores legislative 
proposals, including those related to healthcare policy.
    Much of the interest in and debate surrounding CBO has been 
centered on transparency, and these hearings have already 
enhanced communication between the agency and Congress. They 
have also started to clarify CBO's role in the congressional 
budgeting process and begun to improve understanding of how CBO 
does its work. Without question, there is genuine interest on 
both sides of the aisle to fix the broken budget process, and I 
want to remind my colleagues that CBO is a critical piece to 
ensure Congress has a successful in fully functioning budget 
process.
    Today, I look forward to hearing from our witnesses and 
engaging in more productive conversations with CBO. And with 
that I would like to yield to the Ranking Member, Mr. Yarmuth 
from the Commonwealth of Kentucky.
    [The prepared statement of Chairman Womack follows:]
[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]
    
    Mr. Yarmuth. Thank you very much, Mr. Chairman. I would 
like to join you in welcoming our three CBO witnesses today: 
Director Hall, Dr. Banthin, and Dr. Kling. We appreciate you 
coming before the Committee to testify and answer questions 
about how CBO incorporates behavioral responses into its 
analysis.
    I anticipate that a lot of today's discussion will center 
on healthcare, and I am glad the Committee is looking into it. 
I am not necessarily being critical of the timing of this 
hearing, but I do think it is important to point out that it 
would have made more sense to hold this hearing before voting 
on the majority's bill to repeal the Affordable Care Act and 
gut Medicaid or before voting on their tax bill which repealed 
a key component of the ACA or before voting on a budget 
resolution that cuts $1 trillion from Medicaid and ends the 
Medicare guarantee. Nevertheless, I am pleased that you were 
here Dr. Hall, Dr. Banthin, and Dr. Kling.
    A year ago, the President said nobody knew healthcare could 
be so complicated. I am quite certain that all three of you 
have long known just how complicated it is. That is why 
Congress relies on your objective nonpartisan expertise to 
research model and estimate the effects of large-scale changes 
to our nation's healthcare system. When you consider that our 
system represents fully one-sixth of the U.S. economy, there is 
absolutely no way we can responsibly legislate on behalf of the 
American people without knowing the entire impact of our 
healthcare proposals.
    During the debate on the Affordable Care Act, long ago as 
that was, I called healthcare policy a giant Rubik's Cube. You 
may think you are changing one small part of one side but that 
change unexpectedly impacts a different side or many different 
sides. Healthcare has a lot more sides then a Rubik's Cube, 
which is why we have to rely on impartial analysis and input 
from unbiased experts, or we will find ourselves neck deep in 
unintended consequences.
    And when it comes to access to healthcare, these unintended 
consequences can translate into severe consequences for 
American families. For example, after rigorous modeling, CBO 
ultimately estimated the House-passed bill to repeal the ACA 
would reduce the number of people with health insurance by 23 
million in 2026. Now it is still unclear to me if that was the 
goal of my Republican colleagues or an unintended consequence, 
but either way, it is critical information for us to have. And 
we must have it before we vote.
    That said, Director Hall, I am interested in hearing more 
about this estimate as well as about how CBO is updating the 
baseline to reflect actions taken by Republicans in Congress 
and the administration to deter ACA enrollment, destabilize the 
insurance markets, and prevent the law from being implemented 
as designed. I am also interested in hearing about how CBO is 
re-estimating proposals in the President's 2019 budget request.
    My Democratic colleagues and I have grave concerns about 
many of his policies including massive cuts to healthcare, 
antipoverty programs, and investments in economic growth.
    Again, I thank our witnesses for coming, and I look forward 
to their testimony. And I yield back.
    [The prepared statement of Mr. Yarmuth follows:]
 [GRAPHICS NOT AVAILABLE IN TIFF FORMAT]
 
    Chairman Womack. Thank you, Mr. Yarmuth. In the interest of 
time if any other members have opening statements I would like 
to ask for unanimous consent that members submit them for the 
record. Without objection.
    Chairman Womack. We would now like to recognize the 
Director of the Congressional Budget Office back to the chamber 
here, Dr. Keith Hall. Dr. Hall, thank you for your time again 
today. Committee has received your written statement it will be 
made part of the formal hearing record. You will have 10 
minutes to deliver your oral remarks, and that time begins 
right now. Sir, you are recognized.

 STATEMENT OF KEITH HALL, DIRECTOR, CONGRESSIONAL BUDGET OFFICE

    Mr. Hall. Chairman Womack, Ranking Member Yarmuth, and 
Members of the Committee, thank you for inviting me to testify 
this morning about how CBO does its analysis. I would like to 
make seven points.
    First, our goal is to provide budget estimates in the 
middle of the distribution of possible outcomes. All budget 
estimates are uncertain, but the budget process relies on 
estimates of specific dollar amounts. So those are what we 
produce in our baseline budget projections and in cost 
estimates for legislative proposals. To the extent feasible, we 
describe the uncertainty associated with those estimates, and 
we regularly compare our estimates with actual outcomes when 
available and to improve our estimating methods.
    Second, our estimates are derived from data and research. 
Baseline projections and analysis of policy proposals rely on 
various types of information depending on the program involved 
in the specifics of any particular proposal. When applicable, 
we use historical information about spending, revenues, and 
other factors affecting budgetary outcomes.
    We draw upon research from the professional community and 
utilize information about outcomes in analogous circumstances. 
We consult with congressional staff, staff members at relevant 
Federal agencies, and other experts with diverse views 
including experts from state governments, industry think tanks, 
and universities.
    And we receive input from our panel of economic advisors 
and our panel of health advisors whose members represent a 
variety of perspectives. We write about the data and research 
that informed our analysis in our description of the basis of 
an estimate.
    Third, our estimates incorporate behavioral reactions to 
proposed policy changes. Estimating how individuals, firms, 
government entities would react to a policy is a fundamental 
part of analyzing its effects. For example, our estimates 
account for changes in the production of various crops that 
would result from new farm policies, changes in people's 
likelihood of claiming government benefits if policies 
pertaining to those benefits are altered, and changes in the 
quantity of health services that would be provided if 
Medicare's payment rates to certain providers were adjusted.
    Fourth, for major legislation and when practicable, our 
cost estimates reflect additional behavior changes that would 
affect total output in the economy. Those macroeconomic 
changes, including changes in the labor supply or private 
investment, are incorporated into certain cost estimates using 
what is sometimes termed dynamic scoring.
    Because such macroeconomic analyses are complex and time-
consuming they are produced for only a small number of 
proposals. Usually when the gross budgetary effect is at least 
one quarter of 1 percent of gross domestic product in any year 
over the next 10 years, and when there is sufficient time to 
conduct the analysis.
    Fifth, our analyses cover a broad range of topics. At the 
end of 2017 we had 40 people working on issues related to 
Medicare, Medicaid, the Affordable Care Act, prescription 
drugs, and other health-related topics; 29 working on education 
immigration, income security, labor, and retirement issues; 23 
on national security issues; 17 on tax issues; 16 on energy and 
natural resource issues; 15 on macroeconomics; nine on the 
overall budget outlook; and eight on finance and housing 
issues.
    Those people have specialized training to work on those 
topics and develop analyses specific to issues at hand. The 
analysis involved regularly used hundreds of models and other 
estimating techniques.
    For example, in its analysis of our proposal to increase 
the counseling people received before obtaining a mortgage, CBO 
used evidence about how such counseling reduces the volume of 
loans and default rates among borrowers to estimate how the 
proposal would affect the cost of loan guarantees made by the 
Federal Housing Administration.
    As another example, the agency estimated the effects of the 
ACA on labor supply in its economic projections. Mainly by 
calculating the effects of the law and marginal average tax 
rates and drawing upon research about changes in the labor 
supply resulting from changes in tax rates.
    Six, their estimates are produced by a team of people not 
by models. Although our analysis often use models in preparing 
cost estimates, they also use information obtained from experts 
data and research to each other in which models or other inputs 
to use: how to distill the proposed changes in law, into 
inputs, into those models, and how to combine the results of 
the models with their other available information to produce a 
final estimate. The general process for preparing cost 
estimates is followed in our analysis of major proposals that 
would affect health insurance coverage for people under the age 
of 65.
    For such proposals an especially large number of analysts 
and modeling efforts are usually involved because of the 
complexity of health insurance decisions. In those analyses CBO 
focuses on estimating the effects on coverage, premiums, and 
Federal spending, and the staff on the Joint Committee on 
Taxation estimates the tax related budgetary effects.
    The analyses have three main steps. First, develop an 
analytical strategy. We first review the proposal and identify 
the key effects that it would have. With an estimate issue 
surrounding its implementation, such as how eligibility for a 
subsidy would be verified, assessing the probable timing of the 
effects.
    As part of that process, we consult with outside experts 
and review existing evidence. Second, we model the effects of 
the proposal. We use several models including our health 
insurance simulation model, models of Medicaid, and JCT's 
individual tax model to analyze the proposals effects on health 
insurance coverage in the Federal budget.
    We translate the features that proposal into changes 
relative to current law and the price and generosity of health 
plans and other in factors that would affect the decisions of 
all parties involved--states, employers, insurers, individuals, 
and others--and use those changes as inputs into modeling the 
proposal's effects on health insurance coverage and premiums. 
We then use the results from those analyses as building blocks 
to project the proposals budgetary effects including effects on 
the cost of the Medicaid program and on receipts of individual 
income taxes.
    Third, we review and write about the estimate. At several 
points, we thoroughly review the projections for objectivity 
and analytical soundness. That rigorous process involves 
multiple people at different levels in CBO and JCT. When an 
estimate of the proposals total budgetary effect is nearly 
complete we write up their results along with a detailed 
explanation of how we arrived at them for the Congress. We 
followed these steps to prepare our November 2017 estimate of 
the effects of repealing the ACA's mandates for individual to 
have health insurance.
    In developing an analytical strategy, we examine data and 
research that might shed light on the effects, we discussed the 
issues at a meeting of CBO's panel of health advisors and 
consulted other experts. In modeling the effects of the 
proposal, we estimated the number of people who would no longer 
enroll in health insurance as a result of the repeal, the 
effects and premiums, and the speed at which those changes 
would occur. And we projected how those changes in coverage and 
premiums would affect tax revenues, Medicaid spending, and 
other spending.
    In writing about the estimate, we also indicated that we 
were undertaking considerable work to revise our methodology 
and that the preliminary results indicated that the estimated 
effects on health insurance coverage would be smaller than 
previously reported. In addition to other technical and 
economic changes, our next baseline projections will 
incorporate estimates of those effects because they repeal the 
penalty for not having insurance is now part of current law.
    Seventh, we strive to make our analysis transparent, and we 
recently reallocated resources to make it still more so. We 
have released new publications this year describing our 
processes for producing economic forecast, budget baselines, 
and cost estimates. Key staff are making presentations to 
congressional staff about those processes.
    In the coming months, some of our specific efforts to 
explain how our models have contributed to our estimates will 
include the following: exploring ways to make more supporting 
documentation for the methods used in baseline projections and 
costly estimates publicly available; publishing detailed 
information about key aspects of our updated model for 
stimulating health insurance coverage, including computer code, 
in how an analyst used the model in preparing estimates; 
developing a version of our model for projecting spending on 
discretionary programs to facilitate replication of roughly 40 
percent of CBO's formal cost estimates.
    Releasing technical documentation and computer code 
explaining how key parts of our long-term budget model work and 
how they contribute to our analyses; providing information 
online that enables users to examine how a large variety of 
changes in baseline economic projections can affect projections 
of the Federal budget; publishing revised estimates of how 
certain changes to laws governing medical malpractice would 
affect medical spending; explaining the reasons behind the 
revisions to the methodology used; documenting the model used 
to project how those changes to laws would affect medical cost; 
and making computer code for that model available.
    Posting a tool for examining the cost of different military 
force structures on our website; and providing computer code 
that generates results discussed in our working paper about our 
model of the maximum sustainable output of the economy.
    More broadly, we plan to increase the public documentation 
of our modeling efforts by publishing more slide decks, working 
papers, appendices, supplemental data, related spreadsheets, 
and other technical material. We look forward to getting 
feedback on the usefulness of these transparency efforts.
    I appreciate the opportunity to explain how we do our work, 
and I am happy to answer your questions. And thank you for your 
support and guidance as we carry out our mission to provide 
information to the Congress as you grapple with the difficult 
issues facing the Nation.
    [The prepared statement of Keith Hall follows:]
[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]
    
    Chairman Womack. With 3 seconds to go, that is very good, 
Dr. Hall.
    Mr. Hall. I have had some practice lately.
    Chairman Womack. I am impressed. We appreciate your 
testimony this morning, and as we begin the Q-and-A, I would 
like to welcome two additional witnesses to the table. They are 
Jessica Banthin, Deputy Assistant Director for health 
retirement long-term analysis at the CBO, and Jeffery Kling, 
Associate Director for economic analysis, at the CBO as well.
    Chairman Womack. All three of our witnesses are available 
to answer the questions.
    It should be noted that the Ranking Member and I have 
conferred prior to the meeting, and we have both agreed to 
defer our opening questions to the end out of respect for the 
colleagues of ours on both sides of the aisle who have full 
schedules and have other things going during the course of this 
particular hearing. So, with that, I am going to recognize 
first up my good friend Mr. Renacci from Ohio. The floor is 
yours.
    Mr. Renacci. Thank you, Mr. Chairman, for holding the 
hearing, and I want to thank Director Hall and his associates 
for the participation work at the CBO. You were tasked with the 
incredibly difficult job of forecasting outcomes of complex 
legislation. Oftentimes having to use various assumptions to 
predict the implications of policies that have never been 
tested before.
    I want to touch upon a process the CBO uses to make 
assumptions on human behavior in response to the enactment of 
new policy. Last fall, S&P rating analysis estimated that 
repealing the individual mandate penalty for uninsured 
individuals will lead to 3 to 5 million Americans no longer 
having healthcare coverage. Compare this to CBO's analysis that 
estimated that 13 million Americans would no longer have 
coverage in a report issued in November 2017. This is quite a 
dramatic difference in analysis and is based on assumptions 
made by each organization.
    Dr. Hall can you first please explain the difference in 
assumptions made by the CBO and S&P which led to such 
discrepancies, and why you think your assumptions are 
dramatically different from institutions like the S&P?
    Mr. Hall. Well, let me start with just a broad 
characterization on the mandate. You know, our previous 
estimates of the effects of the mandate were in some respects 
anticipating. They were sort of projecting the likely impact.
    So what has happened with the ACA is we have had a 
significant increase in coverage. So the question is kind of an 
empirical question. That big increase in coverage, how much of 
it was due to the mandate? How much was it due to just 
publicity? How much of it was due to other things?
    So we are trying to parse out what the effects of the 
mandate were from the effects of other things, and we are 
beginning to get some data now, but we do not have very much 
data. And that has been our big challenge. It is because we 
have a little bit of data.
    We probably do not have enough data to come up with a solid 
statistical estimate on those numbers. That is where we have 
this sort of notion that we are probably a little high. We were 
probably a bit high in our mandate, but we were not very 
confident, we now know how much to lower it over time. Dr. 
Banthin can probably fill in a little bit more if that is okay?
    Ms. Banthin. Good morning, Mr. Renacci. So, we published a 
slide deck in September 2017 talking about some of the issues 
Dr. Hall just mentioned. Both the financial effects of the 
penalty associated with the mandate as well as some of the 
nonfinancial effects.
    For example, people's desire to comply with the law, 
possible changes in social norms regarding insurance coverage. 
Those nonfinancial effects are much harder to tease out. We are 
in the process of revising our methodology, and we will be 
discussing that in our forthcoming baseline report.
    Mr. Renacci. It is an important question because, as you 
can imagine, the number is always used, especially when it 
comes from CBO. Well, you are going to take 13 million 
Americans off of healthcare coverage where another organization 
said three to five. So, it is important that we figure out 
where your 13 million came from. Especially, when many times 
the other side uses those numbers against the legislation.
    So, I think it is important. And I am not condemning you. I 
am just trying to make sure we understand how these differences 
come about because clearly when it is for me it has been real 
key. As a businessman for almost 30 years, I rely on 
information, and if your information is different from other 
organizations then, again, we have to figure out what the 
differences are. So I am hoping that in the future that can 
become more open.
    I want to shift gears to assistance that is provided to the 
CBO through the CBO's panel of health advisors. Dr. Hall, can 
you talk about the composition of the panel of healthcare 
advisors and what role they play in assisting the CBO in 
producing its reports?
    Mr. Hall. Sure. Well, the idea of the panel is to have a 
diversity of views and have people on there who can offer us 
advice on our work on healthcare. In particular, they meet once 
a year and we talk about topics of interest. So this is part of 
our way of sort of magnifying our expertise. We have a number 
of experts on the topics but getting some outside experts on 
there as well is really helpful.
    We most recently use them on exactly the issue that you 
mentioned, on the mandate. We were thinking about changing our 
methodology, and we made it a major topic at our health panel 
meeting and got input from our health panel members. And they 
agreed that they thought our estimate was a little too high. 
But it did wind up being that helpful for us in actually 
figuring out how to change our methodology and how much to 
lower the estimate.
    Mr. Renacci. Thank you. I yield back, Mr. Chairman.
    Chairman Womack. Mr. Jeffries, New York.
    Mr. Jeffries. Thank you, Mr. Chairman, and I thank the 
distinguished Ranking Member as well. Director Hall, the 
Congressional Budget Act created the CBO to give Congress 
direct access to expertise is that correct?
    Mr. Hall. Yes.
    Mr. Jeffries. And that was done in 1974, is that right?
    Mr. Hall. That is right.
    Mr. Jeffries. And prior to the 1974 Act, Congress largely 
relied on the executive branch for most of its budgetary 
information. Is that right?
    Mr. Hall. Yes.
    Mr. Jeffries. And is it fair to say that relying on the 
executive branch to make congressional decisions largely put 
Congress at an informational disadvantage since the executive 
branch estimates were designed to enhance support for White 
House and administration policy?
    Mr. Hall. Actually, no, that was one of the reasons that 
Congress decided to create us.
    Mr. Jeffries. Now, for the last 40-plus years CBO has 
provided high quality information to Congress under, I believe, 
nine different directors appointed by the House Speaker and the 
Senate President pro tempore. Is that right?
    Mr. Hall. That is correct.
    Mr. Jeffries. And the CBO appoints staff based on their 
qualifications to do their job without any deference to 
ideology or political party. Is that right?
    Mr. Hall. That is correct.
    Mr. Jeffries. Now, the Trump administration has been very 
critical recently of the Congressional Budget Office. Is that 
right?
    Mr. Hall. I have heard some, yes.
    Mr. Jeffries. So, for example, the OMB Director, Mick 
Mulvaney, asked rhetorically, ``Has the day of the CBO come and 
gone?'' Are you familiar with that statement?
    Mr. Hall. I am.
    Mr. Jeffries. The Trump administration attacks have sought 
to discredit the CBO's analysis of Republican efforts with 
respect to healthcare. Is that right?
    Mr. Hall. Yes.
    Mr. Jeffries. So, in July of 2017, the White House tweeted 
a video claiming that the CBO inaccurately estimates healthcare 
covers. Is that right?
    Mr. Hall. Yes.
    Mr. Jeffries. Now, before unceremoniously resigning last 
September, HHS Secretary Tom Price claimed, ``The CBO was wrong 
when they analyzed Obamacare's effect on cost and coverage.'' 
Are you familiar with that statement?
    Mr. Hall. I am.
    Mr. Jeffries. But as House Budget Chair, prior to his 
ascension, Tom Price recommended you for your current position. 
Is that right?
    Mr. Hall. That is correct.
    Mr. Jeffries. And in fact, when Tom Price recommended you 
he praised you for your vast understanding of economic 
policies. Is that right?
    Mr. Hall. I believe so.
    Mr. Jeffries. And you were appointed as Director of the CBO 
by John Boehner, the Republican Speaker of the House of 
Representatives. Is that true?
    Mr. Hall. Yes.
    Mr. Jeffries. Now the CBO's estimates, with respect to the 
ACA, have largely been accurate. Is that fair to say?
    Mr. Hall. I think so. I think that is right. We are not 
perfect, but I believe we have done pretty well.
    Mr. Jeffries. So, for instance, in July of 2012, the CBO 
estimated that 30 million people would be uninsured in 2016. Is 
that right?
    Mr. Hall. Yes.
    Mr. Jeffries. And the actual number turned out to be a 
close approximation about 27 million. Is that right?
    Mr. Hall. That is correct.
    Mr. Jeffries. And the CBO estimated that 7 million people 
would be enrolled in the Obamacare marketplaces and receiving 
subsidies in 2014. Is that right?
    Mr. Hall. I think that is right.
    Mr. Jeffries. I think the actual number turned out to be 
about 5 million. Is that correct?
    Mr. Hall. I think that is right.
    Mr. Jeffries. Now, were those estimates not more accurate 
than the estimates from many other entities including the RAND 
Corporation. Is that true?
    Mr. Hall. I think generally we have done very well.
    Mr. Jeffries. It was more accurate than an estimate from an 
estimate from the Lewin Group. Is that correct?
    Mr. Hall. Yes.
    Mr. Jeffries. It was more accurate than estimates from the 
Department of Health and Human Services Centers for Medicare 
and Medicaid. Is that right?
    Mr. Hall. That one is about the same.
    Mr. Jeffries. Okay. So, you have exceeded the estimations 
of almost every other entity that has analyzed the future of 
the Affordable Care Act. Is that correct?
    Mr. Hall. I believe we have done pretty well, yes.
    Mr. Jeffries. Now, is it fair to say that recent attacks on 
the CBO are part of a pattern by the Trump administration of 
undermining expertise in our democratic institutions?
    Mr. Hall. I would not want to comment on that, but we are 
certainly not above criticism and critiques on our work.
    Mr. Jeffries. Okay. I appreciate that. Republicans have 
attacked the FBI, attacked Special Counsel Bob Mueller, who has 
had a distinguished career and is a life-long Republican, 
attacked the intelligence community, attacked the Article III 
judicial branch as embedded in the United States Constitution, 
attacked the free press, so, I would just suggest Mr. Director, 
that you are in good company, and, as Jay-Z from Bedford-
Stuyvesant, would say, ``Brush your shoulders off.'' I yield 
back.
    Chairman Womack. Mr. Johnson from Ohio.
    Mr. Johnson. I thought this was a hearing on CBO, but I 
guess I missed something, Mr. Chairman. And I thank you for 
your diligence and having this series of hearings and helping 
us understand how we can, first of all, reauthorize and then 
reform the budget process and CBO's role and responsibility in 
that. We have had some discussions, Mr. Director, with you and 
your staff.
    So, I have got some sort of basic questions here. Since CBO 
was founded on the principles of objectivity and transparency, 
why or how would relying on outside data impact CBO's core 
duties and functions in serving the effort to provide Congress 
with scoring?
    Mr. Hall. Well, we do in fact rely on outside data a fair 
amount. We look for data almost anywhere where we can find it. 
With keeping our eyes open, of course, of the data source and 
the level of objectivity and that sort of thing.
    Mr. Johnson. Okay, well why is there so much tension then 
between CBO's access to private data and data transparency.
    Mr. Hall. Well, lots of times we buy private data from 
companies, and, even if we do not buy the data they are kind 
enough to let us use it, we are careful not to disclose it 
freely because they are in the process of selling the data to 
other people. And if they notice I have it, and we let it go 
freely then that is a problem for them. So, we have to worry 
about confidentiality.
    Mr. Johnson. Do you get pushback from data sources that say 
we do not want to release this data?
    Mr. Hall. Oh, absolutely.
    Mr. Johnson. How do you get past that hurdle when you think 
you need that data to do an accurate score?
    Mr. Hall. Well, we do not always have to disclose the 
details of the data. We do try to detail where we got the data, 
on the nature of the data. We just feel like we often cannot 
actually release the data itself.
    Mr. Johnson. Okay, how does CBO verify information provided 
by the various agencies through its own research? How do you 
verify that what they are telling you is accurate?
    Mr. Hall. Well, we do rely on information from the 
different agencies. We do make an effort to look at other data 
as well to see if it is consistent. Though, actually, lots of 
times where agencies will perform an analysis and, rather than 
take the analysis on, we will ask for the data and do our own 
analysis. So, we do make an effort to make sure that the data 
we are getting from agencies is accurate, and, I say, by 
comparing it to other sources of data sometimes.
    Mr. Johnson. Okay. All right. Moving on to another subject 
I know that CBO makes assumptions, has to from time to time in 
doing your scoring work. So, how does CBO make assumptions when 
the existing evidence and literature are unclear on the effects 
and the impacts of a particular legislative proposal?
    Mr. Hall. Well, the first thing we do is we try to keep in 
mind that there is uncertainty in anything that we do and when 
we make an estimate we are trying to be in the middle of the 
range. We actually, literally sometimes, will sit down and come 
up with our best estimate, and then go through the exercise of 
saying, ``Well, what are the chances the real number is higher? 
What are the chances it is lower? Are we in the middle of 
that?'' So, we really try to make an effort to deal how we can 
with the uncertainty.
    And, you know, sometimes we can look at data that is sort 
of related, maybe not exact data on the issue. Maybe how 
individuals make choices on different types of healthcare. How 
that informs their choices, for example, on Medicaid or 
something like that.
    Mr. Johnson. Okay. All right. Shifting gears just a little 
bit, and I think we talked about this the last time you were 
before our Committee here, about the workload and the 
challenges of how many different scores you have on your plate 
or scoring requests at any particular time. So how does a CBO 
analysts manage the challenges of making assumptions with 
limited time or access to information? You are trying to get 
through this process quickly. How do you manage that?
    Mr. Hall. Well, we do what we can to support them. 
Actually, we have supervisors who help them and help them work 
things out. We will host some additional resources sometimes 
when data is harder to come by. But that is a continuing 
challenge for us, the turnaround times. Sometimes they are very 
quick. One of the things managers spent a lot of time on, are 
sort of urging committees to let us look at legislation early 
so we can start putting together the data that we are going to 
need once the legislation becomes more formal.
    Mr. Johnson. Got you. Okay, Mr. Chairman, I yield back.
    Chairman Womack. Ms. DelBene from Washington, you are 
recognized.
    Ms. DelBene. Thank you, Mr. Chairman, and thanks to all our 
witnesses for being with us today. Since the Trump 
administration took office in 2017 there have been a number of 
significant policy changes to the Affordable Care Act, 
including the elimination of the individual mandate.
    A study released by the Urban Institute just yesterday 
found that the elimination of the individual mandate penalties 
and the other policy changes: such as the withdrawal of cost-
sharing reduction payments and diminished Federal investments 
in advertising and enrollment assistance during 2017 that 
affected the 2018 open enrollment period, that those together 
will lead to an additional 6.4 million people uninsured in 2019 
compared with prior law.
    So, the advertising budget alone for open enrollment was 
cut by 90 percent. How will CBO account for these types of 
actions in its upcoming baseline projections?
    Mr. Hall. Sure. Do you want to go?
    Ms. Banthin. So we are watching closely all information 
regarding open enrollment advertising and all actions of the 
government that influence that, and we will be incorporating 
that into our baseline. We are also following and analyzing 
recently proposed regulations regarding association health 
plans and short-term plans. So we will speak to that when we 
publish our forthcoming baseline.
    Ms. DelBene. So, but how do you incorporate that? You 
brought up short-term health insurance plans proposed rule to 
expand short-term health insurance plan that the administration 
has released. So when you look at that rule, changes and 
advertising budget, how does that affect coverage and premiums 
in the marketplace? How are you going to address that 
specifically? Those are two specific examples but how are you 
going to approach those?
    Ms. Banthin. So, we look at all of the changes that we 
believe will affect enrollment in the marketplace, and we 
discuss what impact we think they will have, we seek evidence 
and information that may help us analyze those effects, maybe 
through historical evidence.
    We talk to outside experts, and we compile all that 
information and then use our model to assess the impact on 
enrollment. When a regulation is proposed but not yet final, we 
follow a budget rule where we incorporate just 50 percent of 
the effect until it is final and then we incorporate the full 
effect.
    Ms. DelBene. Is that a standard that you used before, the 
50 percent of the effect?
    Ms. Banthin. Yes.
    Ms. DelBene. Okay. How did that come about?
    Mr. Hall. It is just a tough way of dealing with a proposed 
rule that may not become a final rule. So, there is some 
probability it will become final, and so if you do not take it 
into account at all that is a zero chance. We do not want take 
100 percent because they do not all turn into final rules. So, 
we have we generally viewed this as a 50/50 chance that the 
rule will become final and incorporate that in our baseline. 
And then, of course, once it becomes final then we incorporate 
the rest of it.
    Ms. DelBene. I understand. Okay, thank you very much; I 
yield back, Mr. Chairman.
    Chairman Womack. Next we go to Mr. Faso of New York.
    Mr. Faso. Thank you, Mr. Chairman, and I thank you and the 
Ranking Member for letting us go ahead of you. That is very 
indulgent. So, I appreciate.
    Chairman Womack. Trying to be respectful of your time.
    Mr. Faso. Well, thank you sir. On the issue of the 
individual mandate that has been routed about throughout the 
Nation over the last number of years: is it correct that when 
CBO did its forecast it counted as those people who would be 
losing insurance those who were currently paying the penalty to 
avoid buying insurance?
    Ms. Banthin. So, we publish estimates that represent 
average annual coverage numbers, and we realize that in the 
real-world people may be uncovered for half a year. So, in real 
life some of the people who pay the penalty may pay a prorated 
penalty due to certain number of months uninsured, but they 
have been covered for the other months of the year. So, our 
estimates try to account for that.
    Mr. Faso. So, when you did that how many people were paying 
the penalty in the most recent year?
    Ms. Banthin. I do not know the number off the top of my 
head.
    Mr. Faso. So, in other words when you did your analysis of 
loss of insurance, you did not actually count all of the people 
who were paying the penalty as losing insurance?
    Ms. Banthin. That is correct.
    Mr. Faso. If the fine or penalty is repealed, you counted 
only a portion of them?
    Ms. Banthin. That is true, because we are presenting 
average annual statistics. So, for example, the real data from 
the IRS may show lots of people paying a small portion of the 
penalty. Let's say a whole lot of people paid half the penalty. 
We would then cut that number in half.
    Mr. Faso. I think this was an example where frankly what 
came from CBO was a bit of a muddle. It was not really 
understood, I think, by either side in this political debate 
and certainly not in the media and among the general public. 
And that might be an area where you might seek more clarity--
just a point.
    The other thing, I am aware that the budget office in 
Canada publishes models that citizens can look at. Netherlands 
invites international and domestic experts to come in and 
critique their CBO equivalent office. Have you considered doing 
such a thing, Director Hall, as well? In terms of trying to 
modernize and improve the forecasting abilities of your office.
    Mr. Hall. Yes, actually if you sort of look at our 
transparency plan we have a few proposals that are doing just 
that.
    Mr. Faso. So, what specifically are you going to be doing?
    Mr. Hall. Yeah, well we have what we are calling our 
waterfall table. Which is looking at discretionary spending by 
category, and we are going to put that up so you can sort of 
see how our projection of discretionary spending is going to 
change from year to year. We are putting up a model on a 
military force structure that is sort of similar to that.
    I think we are also going to work towards, what I call, a 
rule of thumb interactive model. Which will be based on our 
budget outlook forecast where people can then change some 
things like productivity or labor force participation and see 
the effects on the budget once you sort of vary those things. 
So, we have a few things like those that are designed to be 
interactive.
    Mr. Faso. And what about in terms of outside experts to 
come in and help review and critique performance?
    Mr. Hall. Well, we certainly bring them in all the time to 
look at our models. Especially when we are developing new 
models, we go to our experts, and we certainly use our panels 
to talk about our performance and get feedback from them on how 
we are doing in various aspects, especially on the economic 
forecast.
    Mr. Faso. And is that something where you report to the 
committees in the House and Senate in terms of that process.
    Mr. Hall. I do not think we officially report, but we 
certainly could do that if there is interest.
    Mr. Faso. I think it might be useful to at least advise the 
committees as to who you have brought in, what the discussion 
was, and what decisions or determinations you made as a result 
of those evaluations that outside parties assist you with.
    Mr. Kling. So, there is a process for that, you know, panel 
economic advisors, for example. Where at the meeting there is a 
number of the Committee staff who are attending the panel 
meeting itself. So, they can hear in real time what the panel 
members are telling us, what the discussion is about, the 
economic forecast, and then there is also a follow-up meeting 
with Budget Committee staff where they give us information 
about their takeaways from the meeting. And we talk about what 
we think we learned. So, that process is happening at a staff 
level at this time.
    Mr. Faso. Great. Thank you, Mr. Chairman, I yield back.
    Chairman Womack. Ms. Jayapal, Washington.
    Ms. Jayapal. Thank you, Mr. Chairman, and thank you all so 
much for being with us. I wanted to go to the healthcare law 
and how you have incorporated changes that were made by the 
Republicans last year into this analysis, including rules that 
were just applied.
    So a new analysis by the Urban Institute indicates that 
continuing unraveling of the ACA will increase premiums by 
double-digit percentages in most states next year, an estimated 
18 percent on average. And that includes the elimination of the 
individual mandate along with withdrawal of cost-sharing 
reduction payments and slashing of budgets to get the word out 
about the ACA and encourage enrollment.
    On top of this, pulling Americans out of the ACA and into 
short-term limited duration policies would raise the uninsured 
in this country to 36.9 million next year. How are you 
incorporating these effects of the changes in both policy and 
law into your baseline estimates?
    Mr. Hall. Sure. This is something we routinely do with our 
baseline. We have a specific healthcare baseline. So, we will 
do exactly this. We take on board the changes in law and et 
cetera. The rules are a little bit trickier, as we sort of 
discussed you know, because rules are not always finalized. And 
it is sometimes hard to see that, but we will be doing our own 
analysis with this. We will probably look at the Urban work, 
but we will do our own independent work to see what the effects 
will likely to be.
    Ms. Jayapal. And, Director Hall, is part of the reason that 
these rules are difficult to analyze, or even some of the early 
projections around the ACA, when you assumed for example that 
states would welcome Medicaid expansion because it would help 
their populations in their states. That would be a logical 
assumption to make, I believe. I do not know if you agree with 
that or not. You certainly made it in your initial estimates.
    Mr. Hall. That is right. That was the issue the Supreme 
Court overruled, the part that made it a requirement for states 
to expand.
    Ms. Jayapal. So then you revised your budget estimates and 
you actually had to take into account the fact that certain 
states would not be expanding Medicaid. So, there would be 
fewer people in those states, but some of your projections 
underestimated the deep need across the country for health 
insurance. So, actually more people are enrolled in Medicaid 
and are pushing the projections up a little. Is that correct?
    Mr. Hall. Well, that is right; we did underestimate the 
Medicaid numbers a little bit.
    Ms. Jayapal. So, you know, what I am trying to understand 
is how much you look at behavior, the finality of a rule. So, 
for example, let us move to immigration in the budget and the 
estimates in the budget. DACA repeal and deportation of about 
800,000 DACA recipients would reduce economic growth by an 
estimated $280 billion and cost more than $460 billion in 
economic output over the next decade.
    Is the update to the baseline that CBO is currently working 
on taking into account this economic impact if the President 
and the Republicans refused to extend DACA permanent solution 
for DACA recipients? How do you take something like that into 
consideration?
    Mr. Hall. Well, this is part of our annual update of the 
baseline where we, in fact, do try to take those on board. It 
is a little tricky for us when it happens off our schedule. So, 
we try not to make too many adjustments off our regular 
schedule for updating the baseline, but the changes that happen 
in time for us to put them into the baseline we definitely will 
take them into account.
    Ms. Jayapal. So what would you do in a case like that. 
Would you estimate that DACA recipients are not going to have a 
permanent solution? The deadline is March 5th. So, obviously 
the Supreme Court has ruled, and so that may, you know, provide 
us with a little bit more time. But the reality is at some 
point this year, those recipients are going to be deportable if 
there is not legislation passed. So, how do you take something 
like that into consideration?
    Mr. Hall. I do not know how we are going to do it exactly 
this time. Obviously, we will be finishing our baseline over 
the next month or so, and so we will have to have to look at 
these things and decide.
    Ms. Jayapal. Let me ask you about another extended 
complication or result of any proposals to deport or to limit 
legal migration to this country. At the end of 2012 the Social 
Security earnings suspense file--one of my favorite accounts 
because nobody knows anything about it, but I have studied it 
for years--contained $1.2 trillion in uncredited earnings.
    So these are earnings from undocumented immigrants using 
Social Security numbers that are not legitimate Social Security 
numbers. They are still paying taxes into the Federal budget 
that supports Social Security for all Americans.
    If you were to deport undocumented immigrants, or if you 
were to limit legal migration, how do you estimate the enormous 
impact that that would have on the economics of our budget? And 
I see that I am out of time. So can I have 5 seconds for him to 
answer?
    Chairman Womack. I will give you 10.
    Ms. Jayapal. Oh, you are wonderful. Thank you.
    Mr. Hall. Well one of the best things maybe to look at is 
we fairly recently did an estimate of the DACA, and we took a 
lot of these things right into account explicitly in their 
write-up of the estimate. It will give you an idea of how we 
think about this, and how we may deal with this if it comes up.
    Ms. Jayapal. I will work with you afterwards just to see 
exactly that.
    Mr. Hall. Yeah, we would be happy to follow up.
    Ms. Jayapal. Thank you very much. I yield back.
    Chairman Womack. Mr. Smucker, Pennsylvania.
    Mr. Smucker. Thank you, Mr. Chairman. Director Hall, my 
congressional district, 16th district in Pennsylvania, leads 
the State in agricultural production. Local agriculture sector 
supports more than 96,000 jobs, and farmers producers in my 
district rely in many cases on a specific provisions within the 
farm bill to, during tough times, keep their operations open.
    And today it is tough times. Net farm income is that at 12-
year low. Farmers in my area are in crisis and in some 
situations on the brink of bankruptcy. So, it is critical that 
Congress gets this upcoming farm bill reauthorization right, 
and, as you know, CBO's cost estimate is an integral part of 
that reauthorization process.
    So I would like to understand a little bit about the 
assumptions that you make in regard to the farm bill. And 
particularly my question is, with ag being one of the most 
volatile markets that that we have, how does CBO formulate its 
models to anticipate potential catastrophic risks, whether it 
is weather-related or in other areas? And how does it 
anticipate how that applies to farming practices?
    Mr. Hall. I will give a sort of a general answer and then 
pass it on. But this is one of the reasons why we have experts 
in so many different areas. We have experts in ag. We have some 
ag economists on board who spent a lot of time with these 
issues, and you talk with ag experts at the Department of 
Agriculture and elsewhere.
    So they have an understanding of these particular issues 
for agriculture. So, of course, our effort is to be as 
professional and use our expertise as much as possible and have 
these models accurately represent what would happen.
    Mr. Kling. So you can see on our website, with every 
baseline in the spring there is a set of tables, 20 or so, that 
show for different kinds of crops what we think the yields are 
going to be and the prices are going to be on average. And 
then, the key thing that, I think, you are interested in, is 
that sometimes it is not the average. Sometimes it is higher. 
Sometimes it is lower.
    So we have developed a set of modeling tools to look at 
history and see how much volatility has there been, and if the 
future is like the past then we can use that as a way to figure 
out what are the chances that prices are going to be below some 
threshold that might trigger some kind of payment. So, we are 
trying to keep track of those things through the model.
    Mr. Smucker. Do you have different models for different 
titles of the farm bill?
    Mr. Kling. Yes.
    Mr. Smucker. Do you work with any of the other Federal 
agencies such as the USDA or private organizations to help 
better assess the uncertainty in the ag markets?
    Mr. Kling. Yeah. So, we get a lot of information from them, 
and we also have an annual meeting where we have a bunch of 
people who study the agricultural sector--both private 
organizations, thinks tanks, universities, people from the 
executive branch--come in and talk about what is our outlook 
for the market and what are the key sources of uncertainty. And 
we try to take that information all on board.
    Mr. Smucker. Thank you. I have a little time so I am going 
to switch gears. My fellow member from New York has already 
asked some questions in regards to other agencies in other 
countries internationally that we could potentially learn from. 
I know, as a business owner, some of my best ideas were when we 
share best practices with a similar organization. So I would 
like to hear just a little bit more about whether you have 
spent any significant amount of time looking at operations in 
other areas.
    Mr. Hall. Sure. Well, first of all, we get lots of visits 
actually from Parliamentary Budget Offices of other countries 
and members to come and talk with us. And we have some 
discussions back and forth that way. The most formal thing we 
do is we participate in the OECD: Organization for Economic 
Cooperation and Development. They have actually a group of 
parliamentary budget or budget offices. They get together once 
a year and talk about common problems and such. So we do that. 
And then, there have been some pretty good examples, you know, 
Canada and the U.K. where.
    Mr. Smucker. Okay, and I am almost out of time, but that is 
one of the most interesting ideas in information the Committee 
has provided for us about online tools that Canada has made 
available so that legislators, or even members of the public 
who might be interested, could actually look at specific 
impacts of various policy proposals on their own time as they 
are developing those policies. So, and I know there may not be 
time to answer this, but are you looking to improve what we 
have in regard to that?
    Mr. Hall. Yeah, absolutely. And I say we have a pilot on a 
number of things, and we are looking for feedback. When we try 
something, whether you find it helpful, that is great. Then we 
will put more resources into doing that kind of thing. One of 
our challenges is there is so many different ways of being 
transparent, and it takes resources to do all of them. We have 
to sort of make up, what I call business decisions, about which 
ones you find most useful that we should spend our time on and 
which ones probably are not worth our time.
    Chairman Womack. Mr. Khanna, California
    Mr. Khanna. Thank you, Mr. Chairman. Director Hall, I just 
wanted to go over your background so people understand some of 
your qualifications. Tyler Cowen, who I disagree with but is 
one of the most brilliant economists in the world, said when 
you were appointed that you were an outstanding pick. Is that 
correct?
    Mr. Hall. I hope so. I am not sure I heard that.
    Mr. Khanna. And you served President George W. Bush as a 
White House Chief Economist in his administration, correct?
    Mr. Hall. That is correct.
    Mr. Khanna. And you were appointed by Speaker Boehner, 
correct?
    Mr. Hall. Yes.
    Mr. Khanna. And before you were appointed, you wrote op-eds 
criticizing the increase in the minimum wage and 
philosophically criticizing the Affordable Care Act. Correct?
    Mr. Hall. I would not use the word criticizing, but I was 
pointing out some effects that I thought were not being taken 
into account.
    Mr. Khanna. It would be fair on the spectrum of economists 
to say, before you took this role, you were more conservative 
in the conservative economists, correct?
    Mr. Hall. Yes.
    Mr. Khanna. So, how do you feel when you hear the President 
or the Speaker attack the credibility of the CBO? I mean, do 
you do you think President George W. Bush would ever have done 
that?
    Mr. Hall. No, I found the CBO to be a just a really great 
place with very competent people who do really great objective 
work. So, I would I would disagree with those criticisms.
    Mr. Khanna. I mean, you served as the Economic Advisor to 
President George W. Bush. Do you think he or his administration 
ever contemplated those kinds of attacks on the CBO?
    Mr. Hall. I do not know.
    Mr. Khanna. Do you think that the attacks and the rhetoric 
against the CBO is undermining an essential part of our 
democracy in a way that for 40 years we have never seen before?
    Mr. Hall. I can say I am not happy about it. But on the 
positive side, I have been very pleased with the people who 
have come to our defense. Very capable people who really 
understand what we do. So, in that respect it is not been all 
negative for us.
    Mr. Khanna. I mean, because if you do come with a slightly 
conservative perspective, as there have been people who have 
come with a liberal perspective, would you say that you can 
vouch that the analysis of the CBO does not have a liberal 
bias?
    Mr. Hall. Well, I think that is correct, and let me be 
clear that we are not above criticism. If people have 
complaints, if they think we are doing something wrong and want 
to give us feedback that we can take on board, that is actually 
helpful. So, I do not disagree that we should not be 
criticized.
    Mr. Khanna. But, sir, do you think there is any reason to 
doubt because of bias or ideological bias when you come out 
with estimates of how many people would lose insurance or gain 
insurance? Is there any reason to think that there is 
ideological bias behind that?
    Mr. Hall. No, we work we work pretty hard to keep bias out 
of it.
    Mr. Khanna. And what would you be your recommendation, 
then, to those of us in Congress in terms of our rhetoric in 
attacking the CBO?
    Mr. Hall. Well, I would I would hope that that people who 
are unhappy or unclear about what we have done would let us 
talk to them. You know, we have done a lot of outreach. I think 
we probably need to do more outreach to where you can explain 
what we have done and why we have done things. It would 
probably at least get some of the mystery out of what we have 
done. So, people can see that there is, in fact, real analysis 
behind our work.
    Mr. Khanna. Well, those are the only questions. I just want 
to make this comment that, you know, I feel for you. Because I 
think I disagree with a lot of your work before you were 
appointed in this position. I mean, I disagreed with the op-ed 
you wrote on minimum wage, the op-ed expressing skepticism of 
the Affordable Care Act.
    But I think it is outrageous for people to question your 
integrity or the CBO's integrity when you have conservative 
economists questioning, invalidating your credibility--some of 
the leading economists--when you probably understand economics 
more than 95 percent of the people sitting in this body. To 
question the integrity, your integrity, or the integrity of the 
body I think is an ultimate disservice to our democracy. And I 
really appreciate your withstanding those political attacks and 
continuing to do your job.
    Chairman Womack. Let's go now to Minnesota, and the 
gentleman from Minnesota is recognized, Jason Lewis.
    Mr. Lewis. Thank you, Mr. Chairman. I have to say, I do get 
somewhat of a kick out of my colleagues on the other side 
bemoaning attacks while immediately attacking anyone who 
disagrees with them. But be that as it may, I am not going to 
attack anybody today.
    I do want to talk a little bit about your estimates on 
baseline, though, on baseline budgeting and some of the more 
volatile variables that you have to assume there. For instance, 
as I understand it, when you develop a baseline budget, you 
assume that temporary provisions will continue, or the 
discretionary appropriations passed each year will continue, as 
is emergency funds will continue; trust fund depletion will not 
be a factor in spending; mandatory programs continue unabated.
    What has been your experience when doing the baseline? What 
have been the most volatile variables that you think, in an 
honest, objective way, you have probably been wrong on?
    Mr. Hall. Well, let me first go to the first part of your 
question. All those rules that you are talking about are rules 
that are set by either law or the Budget Committee or with 
consultation with the Budget Committee. So, we are actually not 
really making those decisions about how to do those. Those 
decisions are coming from the Budget Committees.
    Mr. Lewis. And we need to evaluate whether those decisions 
are prudent. And so, I am asking what variables have been the 
most problematic in assuming a baseline budget, say, for 10 
years?
    Mr. Kling. I would say the most problematic variable has 
actually come from the economic forecast. It has been our 
forecast on interest rates. Our projections of net interest 
have been farther off than any other budget category on a 
percentage basis, and we are not at all alone in that we 
forecast that interest rates would return to closer to the 
historical levels. Like just about every private sector 
forecaster, we have been very close to the blue chip consensus 
on our interest rate forecast, but nevertheless, everybody has 
been wrong, and it has been bad for our baseline budget because 
our net interest payments have been----
    Mr. Lewis. Emergency spending; you assume that continues? 
Has that been an issue?
    Mr. Kling. So we do project forward emergency spending, and 
so that is a rule that, you know, Congress changes that from 
time to time. Sometimes it is higher; sometimes it is lower. 
That has not been what I would call the biggest category.
    Mr. Lewis. Let's talk a little bit about dynamic scoring. 
Again, our friends on the other side of the aisle think you do 
too much of it; some of us on this side think you do too little 
of it. When we are factoring that into the costs estimates, do 
you have a rule? How much do you use, and what is the rule?
    Mr. Hall. Sure. Well, the biggest thing is the legislation 
needs to be big enough to have a dynamic effect to actually 
affect that the GDP or the labor supply or something like that. 
So, the rule we use is it has got to be greater than one 
quarter of 1 percent of GDP, so that puts it at, right now, 
probably $40 billion or something like that.
    And then, the second thing is, to be honest, whether or not 
Congress has the patience to wait for us to take a couple extra 
weeks to do the dynamic aspect of legislation. That has not 
always happened. Because it does take some time and effort to 
do that. Let me just say, just generally, about the dynamic: 
dynamic analysis is always part of our baseline. It is always 
part of our analysis of the President's budget.
    So dynamic elements just have always been part of what 
worked for a long time, and a big part of what we do. Dynamic 
scoring is just now introducing it to individual pieces of 
legislation.
    Mr. Kling. One important thing to add there is that we are 
following the guidelines that are set forth in the budget 
resolution, and so like the quarter percent of GDP, it is not 
something where we have made a decision about that. It is that 
you collectively have made that decision, and we are 
implementing it.
    Mr. Lewis. True. But, I mean, calculating GDP is also a 
little bit like calculating interest rates. If you could 
predict either one, we would probably all be doing something 
else for a living. I mean, I understand how difficult that is.
    I do want to take a microcosm of the dynamic effect, 
however. In my home state of Minnesota the Department of Health 
just released another report that 116,000 more Minnesotans lost 
their health insurance in the last two years. I happen to think 
the exchange from the ACA is not working too well. That is 
349,000 now without health insurance.
    So let me just ask you a very simple question, Director 
Hall. And that is, in your opinion, would more or fewer people 
buy health insurance if, indeed, the price of, say, a 
catastrophic policy, what my parents used to call major medical 
only--without the mandates, without the wellness benefits 
required from the ACA--if that price were to drop 
precipitously, would more or less people buy health insurance?
    Mr. Hall. Well, more, yes. On the catastrophic? Yes.
    Mr. Lewis. Yes. I mean, so those effects of the ACA had an 
effect on fewer people buying health insurance?
    Mr. Hall. Right. Now, one of the challenges we have got, 
right? Is how do you define coverage. Is it substantial enough 
to be coverage, or is it too small?
    Mr. Lewis. My time is up. I thank you for your indulgence 
and the Chairman's indulgence.
    Chairman Womack. Mr. Ferguson from Georgia.
    Mr. Ferguson. Thank you, Mr. Chairman. And, Dr. Hall, I 
want to thank you again for spending time in our office and 
having some really good conversations. As I sit here, and I 
kind of think about it--I listen to the comments from my 
colleague from California, and talk about, you know, the 
attacks on the CBO office--you all probably feel a lot like we 
do sitting over here. You are involved in a process that has 
too many acceptable pathways to failure, because that is the 
way that it has been designed, and all you are doing is you are 
operating within the framework that we tell you to. Okay?
    So, you know, what I am more interested in is more of the 
processes, and I have been trying to focus on the accuracy of 
the CBO scores going out into further years. And I think 
before, when we had the discussion, we talked about how the 
fact that you have got an accuracy rate--I think you said 2 or 
3 percent at the 6-year mark. Okay? From what you scored on day 
one, going to 6 years out, I think you are at about 3 percent, 
right?
    Mr. Hall. That is on the outlays. That is right.
    Mr. Ferguson. Yes, on the outlays. So, what I want to look 
at and talk about a little bit today is to get your thoughts 
about, well, how do we get better numbers for years 7, 8, 9, 
and 10? Okay? Because what we are doing is we are making 
decisions on 10-year numbers, and we have absolutely no idea 
what the accuracy of that 10-year number is. I mean, I think we 
feel like a lot of times, and what we are learning, is that we 
are, you know, closing the door, cutting the lights out, and 
throwing a dart at those out years.
    So, when we say we are going to make this budget 
assumption, and it is going to score for this many, whether it 
is positive or negative, what kind of tools do you need? What 
do we need to think about? What conversations do we need to 
have about understanding what the accuracy of year 7 through 10 
is?
    If we cannot do that, do we need to clarify in the CBO 
scoring that the accuracy is really good in years one and two; 
it is this number in years three and four; and it goes out. And 
so, by the end--by the time you get to years 9 and 10, you 
know, you can say, ``Well, we are dealing with about a 50 
percent accuracy rate.'' So, I will let you all take it and 
roll with that.
    Mr. Kling. So, we have been doing 10-year projections 
starting in 1996. So, 1996, we could look ahead to 2007, and in 
1997, we could look ahead to 2008. Those are the big years--the 
recession, for example--and so a lot of our 10-year projections 
have been at this time where the economy was in a place where I 
think people really did not expect it to be there, and we 
certainly did not expect that in the mid-1990s. And so, using 
that very short history, those 11 observations where we have 
the 10-year projections, to get a sense of what we expect our 
accuracy going forward is not really very helpful, because 
those conditions were very unusual.
    It is the case that over a longer period of time, if we 
look at our 4-year, 5-year, 6-year projections, that they tend 
to be fairly stable, and so we do not have any reason to think 
that years 7, 8, or 9 particularly worse than 4, 5, and 6 in 
that regard.
    Mr. Ferguson. So, I mean, that is an interesting point that 
you can make that, but you cannot prove that those outlying 
years are just as stable. And just as you point out, it does 
not matter what the reason is; we were way the heck off in 
those 3 years that you just described. The point I am trying to 
make here is that we put so much value on that--either side--
whether you like the score, or you hate the score, you go, and 
you put a lot of value on that 10-year number. Okay? And the 
point is we need to know what the probability of that number 
being right is, and you just pointed out a great example.
    You had no idea in 1996 and 1997 what 2007 and 2008 were 
going to look like. Okay? So, I just think we need to be more 
honest with ourselves. Can we and do we need to say, ``We can 
give you really good numbers for these 5 years, but beyond 
that, there are too many variables, and we just need to make 
decisions based on that on that uncertainty in the outlying 
years?'' And you know, we are putting that 10-year window in 
there because we need that length of time to be able to 
balance, to get to balance.
    So, my point is that I want you all to think about how we 
communicate the accuracy of those outlying years so that we can 
make better informed decisions as a Committee.
    Chairman Womack. Mr. Smith from Missouri.
    Mr. Smith. Thank you, Mr. Chairman. Director, I want to 
make sure these numbers are correct. But in 2013, CBO projected 
24 million people would be enrolled in the healthcare 
exchanges, is that correct?
    Mr. Hall. Yes.
    Mr. Smith. And in fact, how many in 2017 were enrolled?
    Mr. Hall. Do you have this?
    Mr. Smith. More like 10. So, that is a big miscalculation 
from 24 million to around 10 million. Do you know why there was 
such a miscalculation there?
    Ms. Banthin. As we stated earlier, we did pretty well in 
forecasting 2014 and 2015 subsidies. We overestimated subsidies 
in 2016, and we have since revised down our estimates. Some of 
the reasons that we overestimated subsidies in 2016 have to do 
with----
    Mr. Smith. We are talking 2017.
    Ms. Banthin. Right.
    Mr. Smith. So, you are only talking about 2016.
    Ms. Banthin. Okay, so, 2017, but we are referring to our 
earlier estimate, say, from 2013. We did not anticipate the 
fact that employers would continue to offer as much coverage as 
they have. They are continuing to offer----
    Mr. Smith. Why did you not anticipate that number?
    Ms. Banthin. We assumed employers would either stop 
offering or new firms might decide not to offer coverage so 
that their employees could take advantage of subsidies in the 
marketplace.
    Mr. Smith. But why did you assume that?
    Ms. Banthin. Because the subsidies were available to low-
income employees.
    Mr. Smith. And so, why did people not take advantage of 
that?
    Ms. Banthin. We believe that, given the uncertainty 
surrounding the marketplace and also the nature of the coverage 
in the marketplace, employers made the decision to continue to 
offer typical employer policies to their workers.
    Mr. Smith. So, did you not consider that as a factor in 
making these evaluations early on in 2013?
    Ms. Banthin. We have learned from that experience, and----
    Mr. Smith. But that was a reasonable expectation, and I am 
wondering why, when this report was done in 2013, that that 
reasonable outcome was not part of the factor?
    Mr. Hall. So, in 2012 we published a report about employer 
response, and at the time there was a great deal of 
uncertainty, and there was some evidence that it could be very 
small, as it turned out to be. But there was also some evidence 
that pointed to it might be very, very large. And so, we were 
trying to be in the middle of the distribution of those 
potential outcomes, and we had a kind of moderate employer 
response.
    It did turn out to be smaller, but at the time in 2012, 
looking at the debate at that point, there was also reason to 
be concerned that it could have been very large, and so that 
was one of those cases where we did----
    Mr. Smith. So, you erred on the side of extremely large.
    Mr. Hall.--the kind of analysis that we talked about 
before, which was trying to----
    Mr. Smith. I have got a minute-45. I want to ask some more 
questions.
    Mr. Hall. Sure.
    Mr. Smith. Director, when Mr. Lewis asked you a question, 
you made a statement that caught my interest, and you said, 
``The definition of coverage is what matters,'' and maybe that 
is why you did not go into the definition of how many people 
would have ``coverage'' under catastrophic plans.
    In regards to evaluating the healthcare bill that was 
passed out of the House and we discussed, would coverage have 
been defined under CBO if we would have put in the language of 
the bill that anyone with some kind of healthcare coverage--if 
we defined it, then would you have considered HSA involvement 
or catastrophic plans as coverage? If we would have defined it 
in the law?
    Ms. Banthin. Yes.
    Mr. Hall. Yes.
    Mr. Smith. And so, if we would have defined it in the law, 
under your review and expectations we probably would have had 
an increase of people having access to healthcare rather than 
the decrease because of the individual mandate.
    Ms. Banthin. I am not sure that would have been the 
outcome----
    Mr. Smith. But you have never figured that.
    Ms. Banthin. Our definition of coverage is very broad.
    Mr. Smith. Have you ever figured that? So, what is the 
definition of coverage that you use?
    Ms. Banthin. It is just insurance that covers serious 
illness and a range of services.
    Mr. Smith. So a catastrophic plan is not considered 
coverage?
    Ms. Banthin. Yes, it is.
    Mr. Smith. So what was the number that you felt, under the 
healthcare bill that was passed, would increase coverage under 
a catastrophic plan?
    Ms. Banthin. Catastrophic plans under the proposal would 
not----
    Mr. Smith. The healthcare bill that you evaluated.
    Ms. Banthin. The EHB would still have been in effect, so 
those plans----
    Mr. Smith. So it would not have been considered coverage?
    Ms. Banthin. They would not have been available for sale.
    Mr. Smith. So it would not have been considered coverage?
    Ms. Banthin. There would be no policies like that to 
purchase.
    Mr. Smith. But there were policies within the healthcare 
bill that you evaluated and gave a CBO score that we passed out 
of the house. And you did not consider coverage for that?
    Ms. Banthin. We did not have to make that determination. 
People were either covered through current law or not.
    Mr. Smith. Okay. I wish I had more time. But thank you.
    Mr. Hall. Well, let me just say, the short answer is if it 
had included catastrophic coverage, that would have counted. We 
are just saying that the legislation we evaluated did not have 
policies with just catastrophic coverage. And we can follow it 
up but----
    Mr. Smith. Yeah, we will follow it up, because it did.
    Chairman Womack. Welcome back, Ms. Schakowsky. You are 
recognized for 5 minutes.
    Ms. Schakowsky. Thank you, Mr. Chairman. In the March 13, 
2017 score for American Healthcare Act, the CBO estimated the 
effects of defunding Planned Parenthood. ``To the extent that 
there would be reduction in access to care under the 
legislation, they would affect services that help women avert 
pregnancies. The people most likely to experience reduced 
access to care would probably reside in areas without other 
healthcare clinics or medical practitioners who serve low-
income populations. CBO projects that about 15 percent of those 
people would lose access to coverage.'' So, on what basis did 
the CBO estimate the 15 percent of those people would lose 
coverage? Dr. Kling?
    Mr. Kling. So, that estimate is based on CBO's analysis of 
data on the locations and capacity of providers that specialize 
in offering family planning services, as compiled by the Health 
Resources and Services Administration.
    So some areas of the country would lack other providers who 
serve low-income populations; other areas of the country would 
have those providers, but they would have limited capacity to 
serve all of Planned Parenthood's clients, and a change in 
providers would affect access in instances where those 
providers do not offer the same breadth of family planning 
services as those offered by providers who specialize in family 
planning services, such long-acting and reversible 
contraceptives, same-day appointments, on-site refills, and 
confidentiality for minors.
    Ms. Schakowsky. Thank you. CBO also looked at the effect 
that loss of access to reproductive health care would have on 
births. ``CBO estimates that the additional births stemming 
from the reduced access under the legislation would add to 
Federal spending for Medicaid. In addition, some of those 
children would themselves qualify for Medicaid and possibly for 
other Federal programs.
    ``By CBO's estimate, in the 1-year period in which Federal 
funds for Planned Parenthood would be prohibited under the 
legislation, the number of births in the Medicaid program would 
increase by several thousand, increasing direct spending from 
Medicaid by to $21 million in 2017 and by $77 million over the 
2017 to 2026 period.'' And on what basis did CBO determine that 
kind of reduction would mean more births?
    Mr. Kling. So, CBO used data on the family planning 
services available from different types of health care 
providers and relied on research identifying the relative 
effectiveness of those services in preventing an unintended 
pregnancy. We projected that less effective services would be 
available under the proposal on average, resulting in 
additional births that would increase Medicaid spending.
    Ms. Schakowsky. Thank you. Did you want to say something?
    Mr. Hall. No.
    Ms. Schakowsky. So, the former Chairman of this Committee, 
Diane Black, contested the CBO estimate during our budget mark-
up last year. She said, ``It is unreasonable for CBO to assume 
that women are incapable of finding another provider. Women are 
smarter than this.'' So, when calculating the estimate, did the 
CBO assume that women are not smart enough to find another 
provider, or was its estimate based on some evidence that the 
closures of health providers in medically underserved areas 
could reduce access to birth control and thus lead to more 
births?
    Mr. Kling. It was based on our analysis of the locations 
and the capacity of providers that we talked about in the 
answer to your first question.
    Ms. Schakowsky. Thank you. One source of uncertainty in CBO 
estimates is that your office cannot always anticipate how a 
law will be implemented, but I understand that the CBO does 
adjust its baseline and cost estimate as it receives new data 
and observes changes in implementation. The Trump 
administration cut the outreach budget for the insurance 
marketplace under the Affordable Care Act by 9 percent. Is that 
the sort of change that the CBO would look at?
    Mr. Hall. Yes, it would, absolutely. We will be taking a 
look at that. We have not made any conclusions yet.
    Ms. Schakowsky. The administration wants to let states take 
Medicaid healthcare benefits away from families who cannot 
provide employment. Is that the sort of change in 
implementation CBO would consider as it updates its model?
    Mr. Hall. Yes.
    Ms. Schakowsky. And after years of decline, the uninsured 
rate went up--is that the last year? Okay. Is that the sort of 
new data the CBO would incorporate in its baseline future 
estimate?
    Ms. Banthin. I can answer that. We take account of all 
survey results that you are referring to, the Gallup poll. 
However, we place a little bit more weight on the National 
Health Interview Survey because of its larger sample and much 
higher response rate. That was released last week.
    Ms. Schakowsky. And who does that survey?
    Ms. Banthin. The Center for Health Statistics, CDC, HHS. It 
is a large, federally sponsored survey. That showed a tiny 
uptick that was not statistically significant in the number of 
uninsured.
    Mr. Hall. But that only covered the first three quarters of 
2017, so we have some more data to get.
    Ms. Schakowsky. Good. Thank you. I yield back.
    Chairman Womack. Governor Sanford, South Carolina.
    Mr. Sanford. Thank you, Chairman, and I thank each of you 
all for being here. I come from the coast of South Carolina, 
and there the tides gently come in, and they gently go out each 
day. When I look at your forecasts, here is my concern. Is it 
optimistic? Are we lulling ourselves to sleep based on 
forecasts that are relatively benign in what may prove not to 
be the case?
    And, specifically, I want to get at the notion of asset 
values and the behavior that comes with that in terms of 
consumer spending and other. I will go back to what Mr. Smith 
was just getting at. If you look at your projections in 2012 on 
the ACA, it was not margin of error. It was just plain error.
    And I am not faulting you for the different things you 
thought, but, I mean, the numbers; he focused in on a single 
year. We are talking from sort of a high point of 67 percent 
projected to, as the trend ran out of the window, down to 27 
percent are projected. I mean, again, not margin of error, but 
just outright error. And I will go back to what you said, Dr. 
Kling, just a moment ago, which was, in fact, interest rates 
were really the larger area of error in forecasts.
    So my question is this: if that is the case, and we were 
off on our interest rate numbers, given interest rate as a 
driver of the Federal budget going forward over the next 10 
years, we could put ourselves into one heck of a financial 
conundrum. And so, I look at some of the wealth numbers; we are 
in uncharted waters with regard to Federal and central banks 
with zero interest rate policy around the world. We have a 
global debt number that has run up to the highest ever.
    I pulled the Shiller PE ratio. Our numbers are now higher 
than on Black Tuesday in terms of the Shiller PE index. I look 
at household net worth to disposable income; we have never 
before in the history of our country been at as high a level. I 
look at household net worth to GDP; again, same phenomenon. We 
had blips, both prior to the tech bubble and the crash of 
roughly 2008, but never before at the levels that we are seeing 
right now.
    So maybe we are living in a bubble, and we do not even know 
it, and if that was the case, what does that do in terms of 
your projections? Because there is a wealth effect in the way 
that people spend money that would be real and have, I think, 
very grave consequences, not just in terms of the Federal 
budget but in terms of, ultimately, people's wealth in this 
country. Give me a little bit in terms of the degree to which 
you are incorporating that into your budget forecast.
    Mr. Hall. Sure. Actually, the risk, I think, is a little 
difficult to talk about, difficult to summarize for folks. You 
know, when we do our economic forecaster budget forecasts, we 
give our best estimate. We try to give an idea of what the 
ranges are, and then we can say things like when the debt gets 
up to really high levels, the risk is higher, but it is hard to 
communicate exactly how high.
    Mr. Sanford. I am not faulting you, but I am just pointing 
out that there is a substantial contingent liability built into 
whether our budget--in fact, if the CBO budget is, from my 
standpoint, generous, then the administration budget is wildly 
optimistic. Would that be fair?
    Mr. Hall. Well, we have not looked at the President's 
budget yet, so we know their take on that.
    Mr. Sanford. I understand the political distance, but, I 
mean, if yours is optimistic, then that would be wildly 
optimistic by comparative standard, correct?
    Mr. Hall. Yeah, we are usually pretty close to private 
forecasters. We are pretty close to the consensus, because we 
are all looking at the same data.
    Mr. Sanford. I have got 45 seconds. Dig in just a little 
bit on the thought you were getting at, though, which is what 
Rumsfeld talked about, the known unknown, or the unknown that 
we do not know about. Where are we? I mean, to what degree is 
that built or not built into what you are forecasting?
    Mr. Kling. So, one of the things that you might be 
interested in is in our long-term budget outlook we have 
analysis of budget uncertainty, where we look at some different 
factors of ``Well, if productivity growth and economy is higher 
or lower than we project, then what are the effects on the 
budget? What are the effects of higher or lower healthcare 
spending?''
    And so if you take a number of factors into account, you 
can see, like, the path of debt is either much higher than our 
central forecast, or it seems to be relatively flat. So, it 
would be unlikely for it to turn out that debt would be a lot 
lower than current levels. We project that it is going to be 
somewhat higher; it could be considerably higher, or it could 
remain around the same. But that cone is definitely mostly up, 
and so that, I think, is something that would be of interest to 
you.
    Mr. Sanford. Thank you.
    Chairman Womack. General Bergman, Michigan.
    Mr. Bergman. Thank you, Mr. Chairman, and thank you all for 
being here, because this is always an educational experience, I 
think, for all of us who are listening. Sometimes I think it, 
may not be impossible, but if you are talking, you are not 
listening. So, I am always appreciative of the opportunity to 
sit here and listen to people who are deep in the details of 
what you do on a daily basis.
    We are all, in our own way, deep in the details of what we 
do, but unless we share and collaborate across any arbitrary 
lines, artificial lines, then none of us are really doing our 
job on behalf of the American people.
    My district, the first district of Michigan, and a couple 
of important details. Michigan has 14 districts, congressional 
districts. My district has 46 percent of the landmass. Okay? 
Why is that important? I get a chance to drive a lot of 
beautiful places, but I also get a chance to meet a lot of 
people.
    For over a hundred years, the Upper Peninsula of Michigan 
has been on its economic back because of closure of mines, 
decreases in logging, all sorts of different things that go. 
So, we, just in Michigan alone, as the Upper Peninsula, feel 
left out of the decisions made in Lansing at times, our state 
capital. So there is a certain sensitivity, but a certain sense 
of realism, about what it means to be able to live your life, 
raise your family in the Upper Peninsula of Michigan.
    So we know that on the bell curve we use, the UP fall out 
on the outside the statistical norm. Okay? When you look at 
assumptions, and you look at then validation of those 
assumptions, can you say they are in that statistical norm of 
the 80 percent across the board, for the most part?
    Mr. Hall. Yeah, that is a good question, because so much of 
what we deal with is not based on statistical analysis. You 
know, if we had a nice statistical model, we could make an 
estimate, and we would have some idea of the uncertainty. So 
much of the uncertainty is unknown. These are the unknown 
unknowns. And so, we are----
    Mr. Bergman. But if you take a list of--because I know time 
goes fast. When you are listing or creating your assumptions, 
that assumption, if you put it right back into a bell curve, 
does it fit in the middle, or do you ever look at ``Where does 
my assumption fit into a norm?'' Or is it just kind of pie-in-
the-sky? How do you develop your list of assumptions?
    But also, more importantly, do you share those lists of 
assumptions going in with folks like us who know our districts 
like the back of our hand? And when we look at something, I can 
say that there is no fit in my district for that assumption, 
because it is not a valid assumption. That is why I am curious 
as to how you do that.
    Ms. Banthin. So, one example is that we do--in thinking 
about marketplaces and how well those marketplaces are 
functioning, we consider insurer participation, and we realize 
that in more rural areas there is less competition among 
insurers. We try to work that uncertainty and to discuss it in 
our reports and work it into our estimates. That is just one 
example I can provide.
    Mr. Bergman. Okay. Does the CBO communicate to Congress if 
you have a low level of confidence in any of the, you know, 
underlying assumptions, you know, of a cost estimate? Do you 
let us know ``Hey, I feel really good about this,'' or ``It is 
a little bit shaky''?
    Mr. Hall. We try to make that as clear as we can in the 
write-up of the cost estimate. For example, with the ACA repeal 
proposals from last year, we had a section called uncertainty. 
So we discussed the uncertainty in some detail, because we 
thought there was a high level of uncertainty.
    Mr. Bergman. Okay. And also, does the CBO ever collaborate 
with the private sector or foreign budget agencies when 
attempting to develop cost estimates in any particular area 
that you are looking at? Do you look at outside, you know, 
models and others that you say, ``Hey, maybe we can apply 
this,'' or ``This does not work for us because--'' what is your 
process for doing that?
    Mr. Hall. Well, absolutely. It is one of the reasons why we 
have expertise in-house. They spend a lot of time not only 
developing their own models, but looking to see other models 
that are available, whether it is the private sector or 
somewhere else, to see if that captures something that we are 
capturing. We will occasionally use outside models if it is 
something that we do not have on board.
    Mr. Bergman. Thank you. Mr. Chairman, I see my time is out. 
Thank you very much.
    Chairman Womack. Mr. Woodall, Georgia.
    Mr. Woodall. Thank you, Mr. Chairman. Thank you for having 
this hearing and thank you all for being here. I wanted to 
follow up on something that Dr. Banthin said, that CBO was 
surprised by how many employers went above and beyond to 
protect their employees from the uncertainty of the 
marketplace.
    Continuing to provide that insurance might not have been 
there in their immediate financial interest, but certainly, in 
the long-term financial interest, taking care of your employees 
is good for business. How will that new understanding of how 
employers operate be reflected in the baseline going forward?
    Ms. Banthin. Well, so we talk about this in our report on 
the process we follow in developing a cost estimate for any 
major proposal, but it is true for our baseline as well. We try 
to consider the timing of responses by affected parties. So 
employers are one of those important groups.
    Mr. Woodall. I might not be asking the right question, and 
so I apologize for interrupting. But I guess what I am saying 
is I thought what I heard you say is you expected employers to 
dump their employees. They did not, and they did not because 
they were protecting their employees from uncertainty.
    Ms. Banthin. Exactly.
    Mr. Woodall. So you were surprised by the effort that 
employers went to protect their employees. My question is not 
the timing of that protection but how does the baseline going 
forward reflect that employers are protecting their employees 
more than you expected?
    I am reading again about bonuses coming out from the tax 
bill. I do not think that was predicted behavior by JCT, but it 
is, in fact, behavior that is happening. So, how do we 
incorporate this new and better understanding of how employers 
are trying harder than you expected to take care of their 
employees?
    Mr. Hall. Let me just say we will in fact take that into 
account in our model going forward and we have. This take on 
board that companies are going to be more likely to continue 
offering coverage. We can and do. That is one of those 
behavioral assumptions which we will change now, because what 
happened was different than what we expected.
    Mr. Woodall. Changing, Director Hall, just with respect to 
healthcare coverage, or change it with respect to dental 
coverage and child care and salary assumptions? Is it a 
behavioral change, that we thought employers were out to 
feather their own nest, but it turns out that to run a good 
business you have to take care of your employees?
    And so, it is across the spectrum, and we will see it 
reflected in almost every employer-related bill going forward, 
or have you restricted that surprise to just healthcare? And 
employers are doing this unexpected behavior only with 
healthcare, but they are going to do exactly what we expected 
them to do with child care and transportation assistance and 
salaries and everything else.
    Mr. Hall. Well, I mean, certainly, we have changed it. We 
will change it with healthcare. Yeah, I think we will probably 
think about other things, see that this experience helps inform 
our assumptions of behavior, behavioral assumptions in other 
things like child care and some of those things.
    Mr. Woodall. I see the headline over and over again; 23 
million Americans expected to lose healthcare. It is not your 
fault; the media gets to report what the media wants to report. 
I have not seen--and thank goodness you all were here today--I 
had not seen that observation that employers take better care 
of their people to protect them from uncertainty than models 
had previously assumed. I understand that was reported, but it 
was not highlighted.
    Does CBO go out of its way to highlight for us? You know, 
folks have been here for 10, 20 years; they expect the CBO to 
do what the CBO does. When CBO uncovers new behavioral 
information, is that highlighted for folks somewhere? ``These 
are the changes we are assuming going forward from 2016; these 
are the changes we are assuming from 2017.''
    Mr. Hall. Well, we certainly do that in our budget outlook. 
We actually have sections on there about how we are changing 
things. We do it also in what we call an annual analysis of 
actuals, where we look at all the budget categories once a 
year; see how we are doing, how we did; see if we need to 
adjust our forecast, adjust our models. We have done that in 
private so far.
    One of the things that we are proposing to do is start 
releasing that. It would give you information on areas where we 
think our forecasting was not as accurate as it could be and 
where we have made adjustments. It would give you a feel for 
when we are making those changes.
    Mr. Woodall. We cannot sort this next one out in the 26 
seconds that are remaining, but I noticed your reluctance to 
respond to Mr. Sanford's ``Are the President's expectations 
wildly optimistic for the economy?'' When I go back to 1972 and 
the debate around the Act of 1974, the goal was to have CBO be 
the advocate for Article I. I know that Article II will not 
have any problem talking about how we have done our math wrong, 
and I wonder what the wisdom is of us not that being equally 
strong in defense of our work here. A different issue for a 
different hearing. Mr. Chairman, I thank you.
    Chairman Womack. All right, thank you. The dais is now 
clear of the members seeking to ask questions. I am going to 
now yield to the Ranking Member, Mr. Yarmuth.
    Mr. Yarmuth. Thank you, Mr. Chairman, and thanks to our 
witnesses for their responses. Just segueing off of Mr. 
Woodall's question, I am curious; maybe when you have analyzed 
or were analyzing the tax bill that recently was enacted and 
trying to model behavior, did you have a sense of how much--or 
the Joint Tax--have a sense of what the behavior was going to 
be with corporations? Because Mr. Woodall mentioned the 
bonuses.
    And Morgan Stanley just did an estimate, according to a New 
York Times editorial, that 43 percent of corporate tax savings 
would go to buybacks and dividends; nearly 90 percent would 
help pay for mergers and acquisitions. Just 17 percent would be 
used for capital investment, and even a smaller share, 13 
percent, would go toward bonuses and raises. Do you know 
whether those numbers correspond to expectations that maybe 
Joint Tax used or that you would use in your analysis of the 
tax bill?
    Mr. Hall. Yeah, I do not know offhand. I do know that is 
going to be an important part of our budget update, our 
analysis of the tax bill, how exactly we view that. We will try 
to detail it once we sort of iron it all out. I kind of do not 
want to give away pieces of what we are still working on.
    Mr. Kling. The outlook that we are preparing for you to 
publish this spring does have the advantage of this additional 
information that has come out in the last few months that the 
staff of the Joint Committee on Taxation did not have when they 
were doing their estimates in December. And so, we do have more 
information, and our estimates will take that on board.
    Mr. Yarmuth. Great. I appreciate that. Thanks very much. We 
have talked a lot about healthcare today; I do not really want 
to talk too much about it anymore, so I want to move on to a 
different subject that may or may not enter the national 
dialogue pretty soon. That is the question of infrastructure. 
And, you know, the President has a plan that relies on 
basically a 20 percent investment by the Federal Government in 
projects, matched by an 80 percent investment by either state, 
local, or private investors.
    And I assume that different types of infrastructure would 
yield different returns to the economy, would have different 
types of impact, and I am curious as to what factors might 
determine the impact of an infrastructure project. I just read 
something that I had never really thought of but makes good 
sense, that there are two different types of infrastructure 
spending; one is maintenance, and one is essentially new 
construction. And I am just curious as to how you would 
evaluate all of those different factors.
    Mr. Hall. Shall we just say, just in general, it is pretty 
clear that specific infrastructure projects vary quite a lot in 
their return, but it is really hard to characterize types of 
infrastructure that way, because there is so much variation in 
individual things. With respect to something like 
transportation, you know, we have done some work on that; you 
know, for example, shifting resources from rural areas to 
congested urban areas. That does increase the productivity of 
Federal spending.
    I think what you are probably referring to is repairing and 
rehabilitating facilities like highways. Probably has a higher 
return then adding new highways, that sort of analysis we have 
found.
    Mr. Yarmuth. Yes, that is interesting. And you mentioned 
rural and urban differences and how they interact. I was just 
reading something the other day that talked about--and I do not 
want to pit rural America against urban America--but that there 
is a possibility, depending on how you allocate resources, that 
you would be spending a disproportionate amount of money--
disproportionate to the population--that might end up in a less 
than optimum return.
    If you were spending 40 percent of the budget on rural 
areas as areas which have either a much smaller percentage of 
the population or a much smaller portion of GDP, is that 
something that should be considered as part of any 
infrastructure plan?
    Mr. Hall. I think our point really kind of goes more toward 
the idea that there are urban areas that are very congested in 
terms of highways, and spending more in those areas has a 
bigger payoff than urban areas. It is not because they are 
rural areas; it is because they may not be as congested as 
urban areas.
    Mr. Yarmuth. Got you. This is a question that is obviously 
timely, and I throw it out there because I have no idea. Has 
the CBO ever done any economic impact of gun violence? Did the 
CBO score the assault weapons ban in 1994? Are you aware?
    Mr. Hall. We will have to back to you. I do not know.
    Mr. Yarmuth. If there was assault weapons ban legislation 
introduced, what would CBO consider doing a score of assault 
weapons ban?
    Mr. Hall. Well, I suppose if it came out of a Committee to 
the floor, we would have been required to do it, to do an 
estimate of that. So, we could check.
    Mr. Yarmuth. But as far as you know, you do not have any 
sense now of what factors you would consider if you were 
actually considering doing a score of that. It would be new 
territory.
    Mr. Kling. Unfortunately, you have the wrong people the 
table for that. We will have to get back to you.
    Mr. Yarmuth. That is good. Going back to healthcare for 
just a second, I just want to get something into the record, 
when we were considering the Affordable Care Act in 2009 and 
2010, CBO did score that legislation.
    And I raise this question because, just during the last 
debate, a lot of our friends on the other side of the aisle 
tried to make a parallel that, yes, we used reconciliation to 
pass the Affordable Care Act in 2010, and they used 
reconciliation to pass the tax bill. But, of course, the 
difference was that the ACA had a positive score at that point 
over both the initial 10-year window and the second 10-year 
window. Is that not correct?
    Mr. Hall. Yeah, that is right.
    Mr. Yarmuth. And the tax bill obviously had a proposed 
deficit of $1.5 trillion. So, it is a considerable difference 
in the budgetary impact of those two. One other question about 
healthcare. One of the things that we know we have to do, and 
we continue to talk about it and work on it, is reducing costs. 
And it seems to me that CBO has an incredible amount of 
knowledge and expertise and resources that it has devoted to 
analyzing certain cost impacts of certain policies.
    And I guess my question is, is there any way Congress could 
avail itself of that expertise without regard to an individual 
proposal? Because it seems to me that you all have probably a 
better handle on what actually might drive cost down in the 
healthcare delivery system than anybody else in this body. And 
is there any way to get, essentially, a generic analysis of 
what possible steps we might consider? Essentially, asking you 
to propose solutions.
    Ms. Banthin. Well, so we do not make policy 
recommendations, but----
    Mr. Hall. But if you throw some at us, we can tell you tell 
you what we think.
    Ms. Banthin. Or rank them somehow.
    Mr. Yarmuth. Okay, that is fair enough. One other thing, 
just in the minute I have left. And I have talked about this 
many times before, and that is the pace of change in society 
and what a difficult time that gives you or anybody trying to 
make projections of more than a year or two.
    I read an interview recently with the managing director of 
Mercedes-Benz. And he talked about a wide range of things, but 
one of the things he talked about which he says could happen 
essentially any day now is that the impact of artificial 
intelligence, and specifically IBM's Watson, on careers like 
basic legal profession, accounting, radiologists and so forth, 
that essentially IBM's Watson has proven itself to be more 
accurate than humans in many of these areas and that people 
could start losing jobs in these areas almost immediately. Is 
that something that you think about or have any way of 
incorporating into projections?
    Mr. Hall. No, that is really hard to project that sort of 
thing. The sort of thing that we can and do is do these sorts 
of sensitivity analysis. What if productivity was much higher 
than we expect at the moment? And get some idea of what effect 
that would have--or much lower. So, we get some idea of the 
possible range of outcomes. But it is really hard to obviously 
predict something like that.
    Mr. Yarmuth. Absolutely. Well, thank you very much for your 
work and for your participation today. I yield back.
    Chairman Womack. Mr. Arrington, are you ready?
    Mr. Arrington. I am, Mr. Chairman.
    Chairman Womack. Let's go to Texas.
    Mr. Arrington. Thank you, Mr. Chairman. And I was Chairing 
a Subcommittee roundtable, and I apologize for being late. And 
Mr. Hall and your colleagues, I appreciate your time and input 
here. I have got several questions but let me start with just 
an observation stemming from the Affordable Care Act 
implementation and your analysis around that.
    The Obamacare, as we refer to it sometimes, expanded 
Medicaid beyond the 100 percent or 120 percent poverty. In some 
cases, because the Federal Government put that carrot out that 
they would pay 100 percent of it, it expanded upwards of 400 
percent poverty in some states. When it did that, it captured 
able-bodied people who are able to work and were not working 
and are not working but receive government assistance.
    And so, in that sort of vein of trying to get on behalf of 
the taxpayers and recipients of welfare, because we think it is 
good for everyone to have an honest, hard day's work and the 
value of that, along with we have a shortage of labor, along 
with, as fiduciaries of taxpayers, this is a really important 
thing for the American people.
    Your analysis, though, suggested, I think, early on--and 
rightfully so, but maybe off the mark--that when President 
Obama and the Congress at that time passed the Affordable Care 
Act, that work would actually decline as a result based on the 
dynamic I just mentioned. And then you came back, I think, at a 
later time--I think maybe the first report was 2010, and then 
in 2014--revised those numbers.
    So, now we are entering into this opportunity for the 
American people, whether it is with food stamps or Medicaid or 
whatever the government assistance program, compassionately and 
respectfully and responsibly, though, asking people who are 
able to work to work if they receive these benefits. Tell me 
about your behavioral modeling in that instance and why it 
changed, and have you perfected it and----
    Mr. Hall. Sure. Taking into account the labor supply 
effective policies is an important behavioral component, I 
think, of what we do, and we do try to do that. I will give 
you, for example, I think it was 2016 we did an analysis of not 
a repeal and replace but just a full repeal of the ACA, and we 
did our dynamic portion of that, where we actually estimated 
the effect on the labor supply of eliminating the ACA. So you 
get this sort of labor supply impact.
    We did that, and then we followed up with a piece, a 
research paper on how we came up with our labor supply effects. 
You know, what research we used et cetera. In fact, I think I 
referenced it in my statement as well. That is the sort of 
thing which may help inform your thinking, but that is 
certainly something we would want to take into account.
    Mr. Arrington. Were your findings that less people would 
work as a result of expanding Medicaid?
    Mr. Hall. Yes.
    Mr. Kling. The Medicaid effect is relatively small, so our 
analysis of the effects of the Affordable Care Act on labor 
supply mostly stems from the taxes in the act and from the 
phase-out of the premium tax credits, and very little from 
Medicaid itself.
    Mr. Arrington. But expanding Medicaid and raising taxes in 
that combination or some--and maybe there are other factors--
caused less people to participate in the workforce. Is that yes 
or no?
    Mr. Kling. Yes.
    Mr. Arrington. Were the numbers higher than you predicted 
in terms of people not participating in the workforce as a 
result of this Federal policy and legislation enacted or less?
    Mr. Kling. So it is pretty hard to know exactly what those 
effects have been, because we only observe what happened, and 
we did not observe what would have happened if the law had not 
been enacted.
    Mr. Arrington. My only point is you came out with ``Here is 
what will happen: fewer people will be in the workforce; less 
people will be incentivized to work as a result of the factors 
we just discussed.'' Were you off the mark because there were 
more people that did not participate and pulled back from the 
workforce, or were there less when you revised in----
    Mr. Kling. The research that has come out since our 
original estimate, effects have been smaller in the literature, 
and that has caused us to revise down a little bit but not very 
much.
    Mr. Arrington. Okay. I am out of time. I yield back, Mr. 
Chairman.
    Chairman Womack. Mr. Grothman, Wisconsin.
    Mr. Grothman. Thank you much, Mr. Chairman. In your view, 
what are the greatest challenges that you guys face in modeling 
for cost estimates that require extensive behavioral 
assumptions? Can you give me some examples of what you go 
through and [inaudible]?
    Mr. Hall. Sure. Well, we are literally always fighting the 
clock, first of all. You know, lots of times the Congress moves 
at paces that are just much faster than we can move, so we do 
our best to sort of try to, you know, put out models and 
understand some of the behavioral issues ahead of time before 
that. And we kind of do the best we can with our current 
resources, frankly. We work flat-out on things.
    Mr. Grothman. Okay. Could you republish your confidence 
intervals for cost estimates produced using behavioral 
modeling?
    Mr. Hall. Sometimes it might be possible. Most of the time 
we can try to give you some idea of the uncertainty. For us to 
produce actual ranges like that, it is kind of a decision, I 
would think, from the Budget Committee whether you want to see 
that or not. I think there are some issues with that, because 
you all want a point estimate, but when they see a range, they 
can pick the number that they like, so they will pick the low 
one or the high one. So, that can cause some confusion, but 
that is sort of above our paygrade as to whether you want us to 
produce those.
    Mr. Grothman. Can you walk us through the review process at 
to how you arrive at those figures?
    Mr. Hall. Sure, sure. Well, just generally, you know, we 
have analysts who spend a lot of time on a piece, and they go 
through; they look at prior estimates to see what informed 
things. They do a lot of talking with folks to see how things 
will be implemented. We have sort of a process like that, but 
then there is a supervisor involved helping with that, helping 
with that, with the assumptions, with the modeling, and et 
cetera.
    And then we have review at various levels. It is not just 
within one division. The work, especially for the important 
pieces, the review comes up to review at different levels 
within the organization. And then everything goes out of the 
director's office, so we get a chance to look at everything 
that eventually comes out where we are.
    Mr. Grothman. You are frequently at a rush for time. How 
does that affect your outcomes, do you think?
    Mr. Hall. It probably adds to the uncertainty. It increases 
the chance that we make a mistake, to be honest. You know, I 
would say it is like anything else: If you have more time, you 
have a better chance to be more accurate and more comfortable 
with things. That is always a balance for us to feel 
comfortable enough with an estimate that we are ready to let it 
go.
    Mr. Grothman. Okay. I will give you another question that 
always concerns me. Obviously, when you are under of the clock, 
you to a certain extent have to rely on information provided to 
you by the agencies, and sometimes the agencies may have an 
agenda. How often do you review the agency's assumptions, or do 
you do independent checking on what you are getting from them?
    Mr. Hall. Well, we generally try to do that, in fact. We 
really do try to do that. We try to get data rather than 
analysis from agencies. If we have experience with an agency, 
and we found that some of the information we are getting is 
unreliable, we will go to a different source other than the 
agency sometimes, to be honest. That is actually one of the 
delicate things for us, because we find people within agencies 
that have the information and we think are reliable, and they 
do not always want to be identified.
    Mr. Grothman. Thank you. I will [inaudible].
    Chairman Womack. All right, thank you. We have cleared the 
dais again, which is a good thing, and I will get to my 
questions. Dr. Hall, Dr. Banthin, Dr. Kling, are you all 
baseball fans?
    Mr. Hall. Yes.
    Ms. Banthin. Of course.
    Chairman Womack. Follow a little bit of baseball? You know, 
we are 30 days away from opening day. You guys are at the 
plate; you are the hitters. Life; the economy; climate; all 
kinds of conditions are the pitcher. And if you follow 
baseball, you know that the pitcher does not just throw 
fastballs.
    If the CBO was just up there to hit fastballs--fastballs 
meaning facts, easy-to-apply data in your calculations--life 
would be really easy for you. But the fact is there are a lot 
of curve balls out there. What are the breaking pitches that 
come at you that are the hardest to hit?
    Mr. Hall. It is the things that are very new, where we have 
no experience, where we have not done an estimate that is at 
all like it before, where there is not necessarily much data, 
there is not much research on a topic. That becomes very, very 
hard for us. And then also in cases that are really complex.
    Sometimes we even get estimates where, in terms of 
behavioral responses, one portion of the estimate pushes 
behavior one way; another pushes the estimate, pushes it 
exactly the opposite direction, so we wind up having this 
judgment as to which one is pushing harder. Those are the sort 
of things that are difficult. But it is the lack of 
information, lack of basis for making our assumptions that is 
the most difficult for us.
    Chairman Womack. Hitters like to find the sweet spot. Dr. 
Banthin, is it hard to find the sweet spot in your 
calculations?
    Ms. Banthin. Sometimes it is. When we are estimating a 
major health proposal that changes the entire healthcare 
system, that is when we have to think very carefully, review 
all available evidence, and then go beyond that and reach out 
to a range of experts. That will take us a lot of time, because 
if a new world is going to be created, and there is no real 
evidence we then have to talk to key stakeholders--insurance 
companies; state commissioners--and we cannot just talk to one 
or two. We have to talk to three, four, five, or six, and 
really think carefully about how these stakeholders will 
respond to the new policy.
    Chairman Womack. Yeah. If you follow baseball long enough, 
you have hitters that just cannot hit a curve ball. They just 
cannot. Dr. Kling, is it important for CBO to be capable of 
adjusting on the fly to the unpredictability of the whole 
plethora of things that you have to consider when you are 
performing calculations for Congress?
    Mr. Kling. I mean, it is absolutely important to try to pay 
attention to all the things that are happening. And so, as Dr. 
Banthin mentioned, in a case that is very uncertain, one of the 
ways we try to cope with that is to reach out to as many 
different experts with a variety of perspectives in order to 
take advantage of the wisdom of crowds in that circumstance.
    Chairman Womack. Dr. Hall, are there any areas in your 
business that you just lack the adequate tools with which to be 
able to adjust to the changing environment where you have to 
deal every day on the basis? And the fact that you are on the 
clock, having to do it was some kind of speed and a level of 
accuracy that can be redeeming to the agency and at the same 
time provide all of us actionable information that is accurate 
enough that we can make well-informed decisions.
    Mr. Hall. I would say the biggest challenge probably is how 
nimble we are, because when we get new topics we did not 
anticipate, we can acquire some expertise, we can acquire 
models and et cetera, but sometimes it takes time. And so, it 
is this sort of nimbleness that is probably our biggest 
challenge, because we have a really broad range of expertise 
throughout the place, but it is not so broad that that we have 
experts in everything that we could possibly get asked about, 
unfortunately.
    Chairman Womack. Are the needs of the CBO reflected in your 
budget request?
    Mr. Hall. Yes. And I have to say, ever since I have been in 
the job, we have been asking for more resources for healthcare 
and dynamic scoring, and we are hoping to get some more. And I 
think we actually have a plan going forward where we would like 
to over the next 2 or 3 years add about 20 people, just 
generally, so we can deal better with the transparency that 
that you all--seems like would want to have an increased level 
transparency from us, and a quicker turnaround.
    We have real issues with peak load issues. Right? Because 
we cover a really broad waterfront of areas, when get a lot of 
action in one area, we do not always have a lot of people, so 
we are trying to figure out how to be more deeper or have a 
deeper bench going forward. So, we are asking for some more 
resources for that.
    Chairman Womack. I want to follow up on a question that Mr. 
Grothman asked just a minute ago, and that is the input 
information you get from other bureaucratic agencies. It has 
been asserted more than once that these agencies can provide 
input to you that can be way out of bounds, terribly 
inaccurate, a swag, if you will.
    How often do you go back and verify, and is there a 
failsafe way to be able to take information from one agency 
that is critical for you to make a calculation and know within 
a certain amount of reason that that particular piece of 
information is really nothing but garbage to you? Garbage in; 
garbage out. So, how often do you do it, and are there 
mechanisms that you have in place or need to have in place to 
ensure that we can troubleshoot--maybe not necessarily second-
guess, but it sounds like that is where I am going with this--
these agencies?
    Mr. Hall. Well, one of the things we do is when we get 
legislation, the very first thing we do is we will ask the 
Committee staff work in legislation, ``Who should we talk to? 
What experts do you think we should be talking to? What data do 
you think we should look at?'' So, we start from the same place 
that Committee staff are starting from, and then we go out on 
our independent way to sort of look at other data sources and 
other things. So, we try not to just rely on agency data; we 
try to do our best to get other data and confirm it and be sure 
that we touch all the bases.
    Ms. Banthin. Yeah. In healthcare, we have enough expertise 
that I do not believe we fall into any trap like that. We get 
data from the executive branch agencies: Medicare claims data; 
Medicaid claims data; Household Survey data. And so, we analyze 
those data ourselves independently.
    Chairman Womack. So, quantifiable data I understand, but if 
it is just seeking information from an agency, does it ever 
happen where the analyst says, ``That does not look right. 
There is no way that that is accurate,'' and go back and make 
them show their work as to why that data is being provided? Is 
there ever a circumstance like that?
    Mr. Hall. There are, actually, and there are probably 
agencies who historically have not been very forthcoming with 
us we have struggled with. I would have to ask around for some 
other examples, but that certainly is true that we do have that 
happen sometimes.
    Chairman Womack. You have described the process as a little 
bit of art and a little bit of science. You know, that sounds 
like a recipe at home where, based on experience, you get all 
of those components just right because you have done this over 
and over again, and you know exactly whether it needs a pinch 
of that or a dash of something else. How capable are you of 
making sure that the outcome that we are getting is based on 
accurate measurements and less of a particular bias that one or 
others may have toward the given subject?
    Mr. Hall. We work in that under almost every level. When we 
hire people, we hire them based on their expertise, and we 
actually vet them, to the degree we can, about their ability to 
do objective work and not bring their personal views in. And we 
train them very carefully about what we want them to do, how we 
want them to approach issues, how we want them trained. Then we 
have a review process of how people are doing.
    Are they doing a good job? Are they being accurate? Are 
they following the right procedures to be objective? Are they 
touching all the bases? Products that come up; they get 
reviewed at a number of different levels as well. We just 
really try everything we can to be sure, because everybody has 
opinions, and we need to have people who can take their 
personal views out and work just purely on their professional 
views, and I think CBO does a really good job of that. I have 
been in a lot of places.
    Chairman Womack. I want to thank you, Dr. Hall, and the two 
other witnesses that are with us today--Dr. Kling, Dr. 
Banthin--for your testimony today. I want to advise members 
that they can submit written questions to be answered later in 
writing. Those questions and your answers will be made part of 
the formal hearing record. Any members who wish to submit 
questions or any extraneous material for the record may do so 
within 7 days. And with that, we wish you a very pleasant rest 
of the day, and this committee hearing stands adjourned.
    [Whereupon, at 12:09 p.m., the Committee was adjourned.]
    [GRAPHICS NOT AVAILABLE IN TIFF FORMAT]