[House Hearing, 107 Congress]
[From the U.S. Government Publishing Office]
H.R. 5215, CONFIDENTIAL INFORMATION PROTECTION AND STATISTICAL
EFFICIENCY ACT OF 2002
=======================================================================
HEARING
before the
SUBCOMMITTEE ON GOVERNMENT EFFICIENCY,
FINANCIAL MANAGEMENT AND
INTERGOVERNMENTAL RELATIONS
of the
COMMITTEE ON GOVERNMENT REFORM
HOUSE OF REPRESENTATIVES
ONE HUNDRED SEVENTH CONGRESS
SECOND SESSION
ON
H.R. 5215
TO PROTECT THE CONFIDENTIALITY OF INFORMATION ACQUIRED FROM THE PUBLIC
FOR STATISTICAL PURPOSES, AND TO PERMIT THE EXCHANGE OF BUSINESS DATA
AMONG DESIGNATED STATISTICAL AGENCIES FOR STATISTICAL PURPOSES ONLY
__________
SEPTEMBER 17, 2002
__________
Serial No. 107-227
__________
Printed for the use of the Committee on Government Reform
Available via the World Wide Web: http://www.gpo.gov/congress/house
http://www.house.gov/reform
______
88-326 U.S. GOVERNMENT PRINTING OFFICE
WASHINGTON : 2003
____________________________________________________________________________
For Sale by the Superintendent of Documents, U.S. Government Printing Office
Internet: bookstore.gpr.gov Phone: toll free (866) 512-1800; (202) 512�091800
Fax: (202) 512�092250 Mail: Stop SSOP, Washington, DC 20402�090001
COMMITTEE ON GOVERNMENT REFORM
DAN BURTON, Indiana, Chairman
BENJAMIN A. GILMAN, New York HENRY A. WAXMAN, California
CONSTANCE A. MORELLA, Maryland TOM LANTOS, California
CHRISTOPHER SHAYS, Connecticut MAJOR R. OWENS, New York
ILEANA ROS-LEHTINEN, Florida EDOLPHUS TOWNS, New York
JOHN M. McHUGH, New York PAUL E. KANJORSKI, Pennsylvania
STEPHEN HORN, California PATSY T. MINK, Hawaii
JOHN L. MICA, Florida CAROLYN B. MALONEY, New York
THOMAS M. DAVIS, Virginia ELEANOR HOLMES NORTON, Washington,
MARK E. SOUDER, Indiana DC
STEVEN C. LaTOURETTE, Ohio ELIJAH E. CUMMINGS, Maryland
BOB BARR, Georgia DENNIS J. KUCINICH, Ohio
DAN MILLER, Florida ROD R. BLAGOJEVICH, Illinois
DOUG OSE, California DANNY K. DAVIS, Illinois
RON LEWIS, Kentucky JOHN F. TIERNEY, Massachusetts
JO ANN DAVIS, Virginia JIM TURNER, Texas
TODD RUSSELL PLATTS, Pennsylvania THOMAS H. ALLEN, Maine
DAVE WELDON, Florida JANICE D. SCHAKOWSKY, Illinois
CHRIS CANNON, Utah WM. LACY CLAY, Missouri
ADAM H. PUTNAM, Florida DIANE E. WATSON, California
C.L. ``BUTCH'' OTTER, Idaho STEPHEN F. LYNCH, Massachusetts
EDWARD L. SCHROCK, Virginia ------
JOHN J. DUNCAN, Jr., Tennessee BERNARD SANDERS, Vermont
JOHN SULLIVAN, Oklahoma (Independent)
Kevin Binger, Staff Director
Daniel R. Moll, Deputy Staff Director
James C. Wilson, Chief Counsel
Robert A. Briggs, Chief Clerk
Phil Schiliro, Minority Staff Director
Subcommittee on Government Efficiency, Financial Management and
Intergovernmental Relations
STEPHEN HORN, California, Chairman
RON LEWIS, Kentucky JANICE D. SCHAKOWSKY, Illinois
DOUG OSE, California MAJOR R. OWENS, New York
ADAM H. PUTNAM, Florida PAUL E. KANJORSKI, Pennsylvania
JOHN SULLIVAN, Oklahoma CAROLYN B. MALONEY, New York
Ex Officio
DAN BURTON, Indiana HENRY A. WAXMAN, California
Bonnie Heald, Staff Director
Dan Daly, Counsel
Chris Barkley, Clerk
David McMillen, Minority Professional Staff Member
C O N T E N T S
----------
Page
Hearing held on September 17, 2002............................... 1
Text of H.R. 5215................................................ 5
Statement of:
Haver, Maurine, Ph.D., Chair, Statistics Committee, National
Association for Business Economics; William D. Nordhaus,
Ph.D., Sterling professor of Economics, Department of
Economics, Yale University; Ralph Rector, Ph.D., research
fellow and project manager, Center for Data Analysis, the
Heritage Foundation........................................ 54
Kroszner, Randall S., member, Council of Economic Advisers,
Executive Office of the President; Kathleen B. Cooper,
Under Secretary for Economic Affairs, Department of
Commerce; and Kathleen P. Utgoff, Commissioner, Bureau of
Labor Statistics, Department of Labor...................... 33
Sawyer, Hon. Tom, a Representative in Congress from the State
of Ohio.................................................... 24
Letters, statements, etc., submitted for the record by:
Haver, Maurine, Ph.D., Chair, Statistics Committee, National
Association for Business Economics, prepared statement of.. 56
Horn, Hon. Stephen Horn, a Representative in Congress from
the State of California, prepared statement of............. 2
Kroszner, Randall S., member, Council of Economic Advisers,
Executive Office of the President, prepared statement of... 36
Maloney, Hon. Carolyn B., a Representative in Congress from
the State of New York, prepared statement of............... 83
Nordhaus, William D., Ph.D., Sterling professor of Economics,
Department of Economics, Yale University, prepared
statement of............................................... 62
Rector, Ralph, Ph.D., research fellow and project manager,
Center for Data Analysis, the Heritage Foundation, prepared
statement of............................................... 70
Sawyer, Hon. Tom, a Representative in Congress from the State
of Ohio, prepared statement of............................. 26
Utgoff, Kathleen P., Commissioner, Bureau of Labor
Statistics, Department of Labor, prepared statement of..... 45
H.R. 5215, CONFIDENTIAL INFORMATION PROTECTION AND STATISTICAL
EFFICIENCY ACT OF 2002
----------
TUESDAY, SEPTEMBER 17, 2002
House of Representatives,
Subcommittee on Government Efficiency, Financial
Management and Intergovernmental Relations,
Committee on Government Reform,
Washington, DC.
The subcommittee met, pursuant to notice, at 1:32 p.m., in
room 2154, Rayburn House Office Building, Hon. Stephen Horn
(chairman of the subcommittee) presiding.
Present: Representative Horn.
Staff present: Bonnie Heald, staff director; Henry Wray,
senior counsel; Dan Daly, counsel; Chris Barkley, clerk; David
McMillen, minority professional staff member; Jean Gosa,
minority clerk; and Earley Green, minority assistant clerk.
Mr. Horn. A quorum being present, the subcommittee on
Government Efficiency, Financial Management, and
Intergovernmental Relations will come to order.
Today, the subcommittee will consider a bill which I
introduced on behalf of myself and Representatives Tom Sawyer
of Ohio and Carolyn Maloney of New York. The bill is H.R. 5125,
the Confidential Information Protection and Statistical
Efficiency Act of 2002. H.R. 5215 has primary objectives. One
objective is to enable the Federal Government's three principle
statistical agencies, the Bureau of the Census, the Bureau of
Labor Statistics, and the Bureau of Economic Analysis, to share
the business data they collect. Such data sharing would
substantially enhance the accuracy of economic statistics by
resolving serious inconsistencies that now exist. It would also
reduce reporting burdens on businesses that must now supply
data separately to the individual agencies.
The bill's second and equally important objective is to
ensure that the confidential data that citizens and businesses
provide to Federal agencies for statistical purposes are
subject to uniform and rigorous statutory protections against
their unauthorized use. Currently, confidentiality protections
vary among agencies and are often not based in law. This bill
would raise confidentiality standards for all Federal
statistical agencies to the highest standard that now exists.
The administration strongly supports H.R. 5215. This bill
is similar to another bill I had introduced in the 106th
Congress, H.R. 2885, the Statistical Efficiency Act of 1999.
That bill received strong bipartisan support, and passed the
House under suspension of the rules. H.R. 5215 differs from its
predecessor by narrowing the data sharing provisions and
broadening the confidentiality protections.
H.R. 5215 is a bipartisan, common-sense bill that we should
enact this year. Therefore, the subcommittee will hold a markup
on the bill immediately following the hearing.
The Heritage Foundation has raised a concern that the
confidentiality provisions in H.R. 5215 could be misconstrued
to prevent the release of some data that is now available to
the non-governmental researchers. This information is released
in a form that does not directly or indirectly reveal the
identity of the data provider. That is not the intent of H.R.
5215. With the support of the administration, I will offer an
amendment at the markup to clarify the language of the bill in
this regard. My amendment also will strengthen the bills
oversight provisions.
[The prepared statement of Hon. Stephen Horn and the text
of H.R. 5215 follow:]
[GRAPHIC] [TIFF OMITTED] T8326.001
[GRAPHIC] [TIFF OMITTED] T8326.002
[GRAPHIC] [TIFF OMITTED] T8326.038
[GRAPHIC] [TIFF OMITTED] T8326.039
[GRAPHIC] [TIFF OMITTED] T8326.040
[GRAPHIC] [TIFF OMITTED] T8326.041
[GRAPHIC] [TIFF OMITTED] T8326.042
[GRAPHIC] [TIFF OMITTED] T8326.043
[GRAPHIC] [TIFF OMITTED] T8326.044
[GRAPHIC] [TIFF OMITTED] T8326.045
[GRAPHIC] [TIFF OMITTED] T8326.046
[GRAPHIC] [TIFF OMITTED] T8326.047
[GRAPHIC] [TIFF OMITTED] T8326.048
[GRAPHIC] [TIFF OMITTED] T8326.049
[GRAPHIC] [TIFF OMITTED] T8326.050
[GRAPHIC] [TIFF OMITTED] T8326.051
[GRAPHIC] [TIFF OMITTED] T8326.052
[GRAPHIC] [TIFF OMITTED] T8326.053
[GRAPHIC] [TIFF OMITTED] T8326.054
[GRAPHIC] [TIFF OMITTED] T8326.055
[GRAPHIC] [TIFF OMITTED] T8326.056
Mr. Horn. I welcome all of our excellent witnesses, and
look forward to their testimony. I am particularly pleased that
our colleague, Representative Tom Sawyer from Ohio, he is a
very distinguished legislator and we are sorry to see him leave
after this Congress. He has been an outstanding leader for many
years in many efforts to improve in this issue, to improve the
Federal statistical activities. And if Mr. Sawyer would like to
come forward and make his statement, and if you wish to, after
your statement, come to the dais. And you are welcome to answer
questions and all the rest of it. So, Tom, we are looking
forward to it.
STATEMENT OF HON. TOM SAWYER, A REPRESENTATIVE IN CONGRESS FROM
THE STATE OF OHIO
Mr. Sawyer. Thank you very much, Mr. Chairman. And thank
you for those very kind words. It has been a pleasure to work
with you over all the years that we have worked here together,
and on no topic more than the one that brings us here together
today.
You are right, I have spent a lot of my time in Congress in
working on ways to improve Federal statistical systems so that
the policies that we struggle over and the formulas that we
agonize over have some meaningful life when they go forward in
the real world and distribute dollars and lead to changed
policies.
I am really pleased that you are moving forward on this
legislation today. As you point out, this measure has been
several years in the making. It builds on your approach to
provide limited data sharing among agencies for the efficiency
of the work of those agencies and my bill to strengthen the
confidentiality of Government statistics. Together, these two
approaches, I believe, will reduce statistical errors in many
important arenas of Federal endeavor. Both parts of this bill
are worthwhile, but I will focus my remarks on confidentiality
today.
In that regard, the bill provides a clear and consistent
standard for the use of confidential statistical information,
and prohibits its use for any non-statistical purpose. It
ensures that data gathered under a pledge of confidentiality
are used only for statistical purposes, and imposes penalties
for the willful disclosure of confidential information. It
would replace the current patchwork of rules, and extend those
protections to all individually identifiable data collected for
statistical purpose, and in that way, encourage greater public
cooperation with Government surveys and improve the quality of
Federal statistics.
This measure, Mr. Chairman, I think is both timely and
necessary. Confidence in Government data-gathering is fragile
at best. In the 2000 Census, more than a quarter of the
occupied housing units in this country did not respond to the
mailed questionnaire. And while that was an improvement over
the previous Census, it is an enormously difficult problem to
overcome. There are many reasons for this. One is the growing
unease about confidentiality that has grown with the
unprecedented surge in the immigrant population. While there
are no known cases of Federal agencies misusing such personal
information in exactly this way, the risk is real. The
temptation to diminish civil liberties in the name of national
security leaves ordinary people and businesses feeling
vulnerable to disclosure.
We saw that risk in the post-Pearl Harbor anti-Japanese
tide. At that time, the Census Bureau worked closely with the
War Department to help locate Japanese Americans. They didn't
turn over actual records or violate the less stringent
standards of the time, but the appearance of complicity
acknowledged only 2 years ago shook the confidence of many
immigrant communities, especially Asian American communities.
Now, after September 11, the firewalls between individually
identifiable information and aggregate statistics appear to be
at risk.
I am pleased that the House Homeland Security bill made
clear that it would not alter existing law on information
collected for statistical purposes. However, in too many cases,
existing law does not ensure that such personal information
will remain confidential. More than 70 Federal agencies or
statistical units collect such data, but only 12, as you
suggest, Mr. Chairman, are covered by regulations to protect
personally identifiable information from disclosure, and only a
handful of those have the stronger protection of law.
Some of these uncovered units collect information on highly
sensitive topics ranging from health care and substance abuse
and mental health. It involves millions of dollars of sensitive
data, and deserves the most stringent of protections from
disclosure.
While agency policy may have once been enough, real public
trust requires that information be shielded by the force of
law. Statutory protection under H.R. 5215 would prevent any
regulatory or law enforcement misuse of these data. This
recommendation was first made under the Privacy Act of 1944.
However, that act has several loopholes that allow for the
disclosure of personally identifiable information without the
informed consent of those who supplied the information.
There are 12 categories of such exemptions, and the act
fails to distinguish between data collected for research
purposes and that collected for administrative purposes, and so
offers minimal protection from improper disclosure.
The commission at that time that arose from the Privacy Act
recommended that no record or information collected for
statistical purpose be used in identifiable form to make any
decision or take any action directly affecting the person to
whom the record pertains. H.R. 5215 embodies the commission's
recommendation in that regard.
In summary, these improvements, Mr. Chairman, are long
overdue. They are needed to protect the public and ensure
continued public participation in essential governmental
research. Informed public policy relies on it. Thank you, Mr.
Chairman, for the chance to be here.
Mr. Horn. Thank you. And if you wish to come up here, why,
without objection we are delighted to have you with us.
Mr. Sawyer. Thank you, Mr. Chairman.
[The prepared statement of Hon. Tom Sawyer follows:]
[GRAPHIC] [TIFF OMITTED] T8326.003
[GRAPHIC] [TIFF OMITTED] T8326.004
[GRAPHIC] [TIFF OMITTED] T8326.005
[GRAPHIC] [TIFF OMITTED] T8326.006
[GRAPHIC] [TIFF OMITTED] T8326.007
[GRAPHIC] [TIFF OMITTED] T8326.008
[GRAPHIC] [TIFF OMITTED] T8326.009
Mr. Horn. We will now go to the panel two, and that is the
Honorable Randall S. Kroszner, member, Council of Economic
Advisers, Executive Office of the President; the Honorable
Kathleen B. Cooper, Under Secretary for Economic Affairs,
Department of Commerce; the Honorable Kathleen P. Utgoff,
Commissioner of the Bureau of Labor Statistics, Department of
Labor. And since it is a big table, and three more--we will
throw in panel three, if you would like to come up to the table
and just grab one of the chairs if we are missing them. And
Maurine Haver, doctorate, Chair, statistics committee,
representing the National Association for Business Economics;
William D. Nordhaus, Sterling professor of economics,
Department of Economics at Yale; Dr. Ralph Rector is the
research fellow and project manager, Center for Data Analysis,
the Heritage Foundation.
And we have to swear witnesses. If you will stand and raise
your right hand. And if you have anybody supporting you there,
get them behind you, too, so we don't have to do it halfway
through the panel.
[Witnesses sworn.]
Mr. Horn. And the clerk will note that all six affirmed.
And we will start then with what it is on the agenda, and that
is the Honorable Randall Kroszner, member, Council of Economic
Advisers. Those were people that were authorized by law under
President Truman; and the old humor that the President had, he
said he was tired of somebody saying on this bit and that bit
and so forth, I want some people that can give me some good
economic data and not just tell it on the right, he is saying,
and then the left hand. And this is the Council of Economic
Advisers.
STATEMENTS OF RANDALL S. KROSZNER, MEMBER, COUNCIL OF ECONOMIC
ADVISERS, EXECUTIVE OFFICE OF THE PRESIDENT; KATHLEEN B.
COOPER, UNDER SECRETARY FOR ECONOMIC AFFAIRS, DEPARTMENT OF
COMMERCE; AND KATHLEEN P. UTGOFF, COMMISSIONER, BUREAU OF LABOR
STATISTICS, DEPARTMENT OF LABOR
Mr. Kroszner. Thank you very much, Mr. Chairman, and
members of the subcommittee. And I am very, very pleased to be
speaking before you today on what I consider an extremely
important initiative that has very big implications, but one
that has no budget implications. And so what I am going to
argue is that this is a very, very valuable piece of
legislation that does not take any additional costs on the--for
the public sector, and actually can reduce burdens on the
private sector. And of course, since I am on leave from the
University of Chicago, our main theme there is, ``there ain't
no such thing as a free lunch.''
So, why hasn't this happened before? Well, members of the
staffs of all of the agencies that have been involved as well
as the leaders of those can tell you that this was not a free
lunch. It required a lot of work to make sure that we could get
the language correct, to make sure that we had the appropriate
ideas in there, appropriate scope. And we got just fabulous
support and cooperation among the different agencies and with
Capitol Hill.
So I am extremely pleased to be able to speak to you about
this very important issue.
As we well know, Federal statistics in the United States
are among the best, if not the best, in the world. But that
should not make us complacent. We can still improve them. The
U.S. economy is an extremely dynamic one, and one in which we
must respond to changes, constant changes. And that requires
constant improvement in our statistics. As we well know,
Government statistics play an important role not only for
Government decisions, whether they be Social Security
decisions, budgetary decisions, monetary policy decisions,
small errors can have very large effects when we do our budget
projections. And this is something that is extremely important
to the Council of Economic Advisers in providing information to
the President to be able to provide accurate forecasts.
The private sector, of course, relies very heavily on
statistics for their own budget and planning purposes, and, of
course, academics around the world rely on U.S. statistics to
be able to do an appropriate analysis of both the effects of
policy as well as more fundamental research. And so what we
need to do is to improve those statistics in a way that will
help to reduce some of the data problems that we have seen
recently.
We have had very large revisions of GDP. We have different
estimates of productivity, one of the most important aspects of
our economy going forward. Depending on which measure you use
from the income and project accounts versus other accounts, the
difference can be 35 basis points. Now, that is about a third
of a percent. That doesn't seem like very much, but that is a
third of a percent difference in growth every year going out
into the future. That makes an enormous difference for our
well-being over time, it makes an enormous difference to our
budget projections. There is an approximate effect of about
$200 billion for every point 1 percent difference in GDP growth
over a 10-year budget horizon. So we are talking real money
here.
Also, when we think about the classification of firms, a
very limited study was done a few years ago looking at the
differences between the Census and the Bureau of Labor
statistics on how they classified firms. They found about a 30
percent difference in which industries firms are classified
into. This has led to very big differences in the estimates for
the size of particular sectors of the economy, the chemical
sector, high-tech sector. And in particular, in the new
sectors, in emerging sectors it becomes very difficult to try
to classify these firms.
And that is how these data anomalies and these areas creep
in. Even though I think all the agencies do a superb job, just
trying to look over something as large as the U.S. economy is a
very difficult thing, given the very limited budgets they have.
By being able to talk with each other and say, well, we have
classified this firm this way, but you've classified it that
way, let's try to understand what would be the best
classification, that is a very simple, straightforward thing
that is basically impossible to do now. And if we can allow for
that, we can improve our Government statistics and improve the
numbers that we get out of that process and be much more
accurate.
What this act would allow us to do is do that through two
means, through formal data sharing through a variety of
memoranda of understanding among the different agencies, but
also by having consistent high confidentiality protections
amongst all of the agencies. And this is very valuable for
allowing the data sharing to occur, because there has to be the
same level of protection, and we want to make sure that it is a
high level of protection if the data is going to move from one
agency to another. And so that becomes an extremely important
part of the legislation itself.
And what this will also do is reduce the burdens on the
private sector. If the agencies can coordinate with each other,
eventually they may be able to reduce duplicative surveys, they
can, in general, just reduce the burdens on the private sector.
And so I consider this something that is a triple win.
First, by improving these statistics directly through improved
business lists, we get better data from what is provided to the
statistical agencies. But by boosting confidence and by
lowering the burdens on the private sector, we are likely to
get much more accurate data from the private sector so we also
have more efficient Government, we have lower burdens on the
private sector, we have no budget cost. I consider this a
triple win.
I think this is something that should have bipartisan
support going forward for everyone's agenda because it is
something that I think clearly improves the Government,
improves the private sector, has no budget costs, and I see
very little--actually, absolutely no downside. Thank you very
much, Mr. Chairman.
Mr. Horn. Thank you. We appreciate your fine succinct
presentation, and we will wait and go with your colleagues and
then we will start the question period.
[The prepared statement of Mr. Kroszner follows:]
[GRAPHIC] [TIFF OMITTED] T8326.010
[GRAPHIC] [TIFF OMITTED] T8326.011
[GRAPHIC] [TIFF OMITTED] T8326.012
[GRAPHIC] [TIFF OMITTED] T8326.013
Mr. Horn. We now have the Honorable Kathleen Cooper, Under
Secretary for Economic Affairs at the Department of Commerce.
Ms. Cooper. Thank you very much, Chairman Horn, Congressman
Sawyer. I am very pleased to be here with you today to discuss
H.R. 5215. I need to get the--better volume. Is that better?
All right. I hope it is not too loud. But I also want to thank
you, Congressman Horn--I mean, Chairman Horn and Congressman
Sawyer, but certainly Chairman Horn, for your leadership on
this issue for a good number of years. As Under Secretary for
Economic Affairs of the Department of Commerce, I have the
privilege of overseeing the fine work of two of the jewels of
the Federal statistical system, the Bureau of Economic Analysis
and the Census Bureau.
And our economic statistics from these Bureaus are
calculated by experts and professionals who produce the most
accurate numbers possible with the tools available to them. The
President has made enhancing our economic data a priority and
wants to give the Bureaus the tools they need to measure the
twenty-first century economy. He appreciates that better
information is fundamental to better public and private
decisionmaking. With the President's budget request for the
Census Bureau and BEA, you will see a range of economic data
sooner. The release of international trade data, for example,
will be available 20 days earlier. We plan to implement an
annual measure of investment in information technology and
quarterly measures of the services sector, the Census Bureau's
first new economic indicator in 50 years.
Unfortunately, while these agencies are striving to improve
economic statistics, the fiscal year 2003 funding level
approved by Senate appropriators is significantly below the
President's request, and we simply will not be able to
undertake these important improvements with the Senate's flat
funding level.
But today, I would like to discuss one way to improve our
Federal statistics at next to no cost. If enacted, this
legislation will help us improve the measurement of
inventories, one of the most volatile components of GDP. We
will develop more efficient samples, reduce the reporting
burden, improve regional and State data, and reduce revisions.
And I would like to share with you some real-life examples of
gaps in our Federal statistical system the data-sharing bill
would close.
Most of BEA's data comes from elsewhere, the Census Bureau
and the Bureau of Labor Statistics being the main building
blocks for BEA products. The Census Bureau, BLS, and BEA
already work hand in hand. BEA is the Census Bureau's most
important customer, and the two agencies are in contact every
day. The staffs at BLS and the Census Bureau meet routinely
with their counterparts at BEA. Throughout the year, managers
collaborate and ensure that our statistical infrastructure is
efficient and productive. However, H.R. 5215 would allow BEA,
the Census Bureau, and BLS to work even more efficiently
together, to share knowledge, and to borrow strengths. The most
important result would be, as my colleague mentioned,
consistent classification of businesses by the Census Bureau
and BLS. BEA would be the first to stand and cheer such an
accomplishment. The Census Bureau and BLS place one third of
the businesses in different boxes, and the BEA has to sort out
the resulting data.
In determining real output, for example, BEA looks at
shipments from Census and prices for BLS, and must untangle the
classification confusion. With the data-sharing bill, the
statistical agencies could cross-validate their company lists
and determine the most appropriate classification. By comparing
corporate financial reports with BLS and Census surveys, BEA
can improve estimates of profits and of wages and salaries.
This would help to reconcile the near $100 billion statistical
discrepancy between gross domestic income and gross domestic
product.
At the end of July, we saw large revisions to profits that
indicated that corporate profits had peaked in 1997 rather than
in 2000, and that other incomes were weaker than earlier
estimated. If BEA were able to access the firm level data from
the Census Bureau, if you compare that information with
corporate return data from the IRS and publicly released
financial reports and in this way BEA could better capture the
impact of corporate activity where there is a difference
between tax and financial accounting methods.
I would like to see an improvement in the accuracy of State
personal income as well. For the year 2000, estimates of growth
and payrolls for Delaware vary from a BLS estimate of 6 percent
to Census Bureau's 14 percent estimate. Likewise, for Virginia,
there are sizable differences. Even for New York where the
differences in growth are smaller, the difference in dollars in
over $7 billion in wages or $380 million in State and local
taxes for the State.
Amid the dynamic economy, how can our statistical agencies
keep track of businesses that come and go? The Census Bureau
and BLS have different sources of information that provide
insight into companies births and deaths. Combining the two
measures should give us better information than from either
source alone.
Other legislation under congressional consideration would
allow the agencies limited access to IRS data. Today, Census
and BEA have access to difference set office tax data. Enhanced
access to IRS material would allow BEA to make its measure of
corporate profits and other business income all the more
accurate.
As you will hear and have heard in part already from my
colleagues, my fellow panelists, this bill also builds on the
agency's unmatched record of confidentiality. It provides
equally stringent protection for all data, and avoids any
perception of inappropriate use.
Finally, Mr. Chairman, I would like to alert you to a
concern that I have about the most basic step in our data
gathering, voluntary participation in our surveys. In April,
reporting by large semi-conductor companies in the Census
Bureau's monthly survey--in the Census Bureau's monthly survey
of manufacturing activities dropped to the point where the
Bureau had to discontinue publishing data on semi-conductors.
As a result, the Census Bureau could no longer produce
bellwether sales and inventory data for this very important
industry.
The Census Bureau and the semi-conductor industry have
agreed to a test to determine whether the industry can provide
the desired data. The results will not be known for 5 months,
and in the meantime, gaps in our data persist. We will continue
to work with businesses to find efficient means for them to
report. Ultimately, Congress and the administration must
encourage participation that yields information vital to
informed decisionmaking.
I thank you, Mr. Chairman. I appreciate your efforts to
improve the quality and the efficiency of our Nation's
statistics while protecting its confidentiality. Thank you.
Mr. Horn. Thank you. We are delighted to have you with us.
And we now will have the last part on panel one, and that
is the Honorable Kathleen Utgoff, Commissioner of the Bureau of
Labor Statistics.
During the Eisenhower administration, I was assistant to
the Secretary of Labor, and the first thing he said to me was:
Nobody around here fools with the commissioner of Bureau of
Labor Statistics. And all of us were told he is untouchable.
Now, you are untouchable. So.
Ms. Utgoff. Thank you for that piece of history, Mr.
Chairman. Good afternoon and--good afternoon, Mr. Chairman,
Congressman Sawyer. I appreciate the opportunity to testify in
support of H.R. 5215. Having come on board as BLS commissioner
just last month, I am particularly pleased that my first chance
to appear before Congress is an opportunity to support
something that is so important to the BLS. There are three main
reasons we believe the proposal currently before the
subcommittee is a good one, and you will hear the other
panelists give similar reasons.
First, all data users, researchers, analysts, policymakers,
private citizens, Government agencies, corporate decisionmakers
will benefit from a higher quality economic statistics. Second,
the major Federal statistical agencies will be able to operate
more efficiently. Third, the protections of confidential
statistical information will be enhanced.
This carefully crafted bill is designed to meet all these
aims. It has come about through an extraordinary level of
cooperation among representatives from the Council of Economic
Advisers, the Bureau of the Census, the Bureau of Economic
Analysis, the BLS, the Office of Management and Budget, and the
Internal Revenue Service, coupled with outstanding
congressional collaboration. The bill is a testament to hard
work and a shared belief in better Government.
Because BLS has an extensive network of cooperative
arrangements with the State's statistical agencies to produce
State and subcommittee--sublabor market estimates, I would like
to note that the intended benefits from data sharing will also
extend to the States. In addition to being large producers of
their own economic data, State governments are large consumers
of federally produced data, and therefore have an interest in
Federal program improvement.
The enhanced data sharing that would be permitted under
Title 2 will improve the ability of BLS, BEA, Census, and the
States to track rapidly changing trends in the U.S. economy. It
will facilitate joint projects among the agencies to improve
data quality and to reduce the reporting burden and costs of
programs. In particular, both the Federal and States'
statistical agencies will be better able to classify business
establishments in appropriate industries, they will be able to
resolve data anomalies and correct reporting errors more
quickly and more efficiently.
Reconciling discrepancies between the BLS and the Census
Bureau business registers is one critical example of why
enhanced data sharing matters. The tangible benefits will
include improved employment, unemployment, and income measures,
better survey sampling frames, improved payroll data for
forecasting State government revenues, and a better foundation
for economic development plans. More accurate business
classification will ultimately allow for the production of more
accurate industry statistics, a vital part of our national and
State economic intelligence picture.
Another important area of potential improvement to BLS data
series is the measurement of multifactor productivity. These
data series track the contributions of capital, technology, and
labor to output. Productivity is widely regarded by analysts as
a key ingredient of economic performance and the standards of
living. By linking BLS work force and occupational data to
Census Bureau production inputs and outputs data, better
measures of productivity can be developed. These better
measures allow more complete research into understanding the
factors that cause productivity change particularly at the firm
level.
Having access to data the Census Bureau collects on firm
revenues by specific product lines would allow BLS to improve
sampling and reduce respondent burden for the producer price
index, or the PPI. The PPI is one of the Nation's most watched
economic indicators. It measures price pressures at the
wholesale level of the economy.
Turning now from the data-sharing provisions of the bill to
the confidentiality provisions, I would like to emphasize the
importance of Title I to Bureau of Labor Statistics. The BLS
has long needed the explicit statutory confidentiality
protection that this legislation offers for the data it
collects. The fact that BLS lacks the specific comprehensive
protections already in place with the Census Bureau, to cite
one example, is a historical omission that needs correcting.
BLS has been successful in protecting the confidentiality of
the data it collects through an amalgam of statutes,
precedents, rules, and practices. A clear example--a clear,
explicit, and comprehensive statutory assurance of
confidentiality is essential to maintain and improve our
response rates by increasing respondent confidence in improving
our ability to protect their data.
In fact, it is not an overstatement to say that
confidentiality is the lifeblood for Government statistical
agencies like BLS that depend upon the voluntary participation
of survey respondents. We rely on individual citizens in their
private households and on companies of all sizes in all
industries and across all States to entrust us with their vital
economic data. The voluntary transmittal of citizens to the
Government of private information on, for example, job status
or earnings, or by a company of its core business information
such as employment wages and revenues is a remarkable example
of public/private partnership. Without the trust of our survey
respondents, the BLS surveys and the important statistics they
gen-
erate would be in jeopardy. H.R. 5215 will strengthen that
trust.
Thank you, Mr. Chairman. I would be happy to answer any
questions that you have.
Mr. Horn. Thank you.
[The prepared statement of Ms. Utgoff follows:]
[GRAPHIC] [TIFF OMITTED] T8326.014
[GRAPHIC] [TIFF OMITTED] T8326.015
[GRAPHIC] [TIFF OMITTED] T8326.016
[GRAPHIC] [TIFF OMITTED] T8326.017
Mr. Horn. We will ask a few questions before we get to the
next panel, and then if you could all stay around, why, we
would appreciate it. And that way we would have a better look
at it.
Could you elaborate on how the bill will reduce data
disparities. For example, how will it solve the problem of
inconsistent standard industrial classification codes? Let's
just go right down the way.
Mr. Kroszner. Thank you very much, Mr. Chairman. As I had
mentioned in my earlier remarks, a limited survey or limited
study that had been done in 1994 trying to compare the
classifications of firms by BLS versus the Census found about a
30 percent difference in the classifications. And so that's a
very, very concrete illustration of how, by working together,
they can try to work out some of the inconsistencies in the
classification of firms.
Now, part of this is because firms change in a dynamic
economy and it is very difficult to keep up as rapidly with
them. So the Census may have done their classification in 1995.
That firm today may still have the same name, but it may be
doing something completely different that may be in a new area.
Also, it may be doing something that wasn't even part of the
classification codes from 5 to 10 years ago. And that is why it
is very important to do this. And also, I want to emphasize
that I think the agencies do a superb job of this, but it is
just hard to keep track of such a vibrant entrepreneurial
economy that we have.
And so I think it is extremely important to allow for this,
and we can reduce the anomalies and inconsistencies by allowing
the data to be shared, the different agencies to talk to each
other to keep up with our very dynamic economy.
Mr. Horn. Can you give us an idea of how much duplicate
data collection now occurs among the three agencies?
Ms. Cooper. How much? I'm sorry?
Mr. Horn. Just give us an idea of how much duplicate data
collection now occurs among the three prime agencies here.
Ms. Cooper. Well, I think it's hard to give a general
statement. My fellow panelist, Mr. Kroszner, mentioned with
regard to classification of business firms. I think we also
have to think about what will happen over time among the three
agencies. I think it is just a very difficult question to
answer; but we think that over time, we will reduce the
duplication, and that clearly is a plus for the reporters, for
businesses out there in this economy. It will lower the burden
on them. And with the comparisons that we could make of cross
agencies, that has to be a very real plus. But it is one of
those numbers that is very difficult to figure out.
Mr. Horn. Well, this next question is about the same as the
last one. But can you estimate the time and resources that this
bill will give the agencies by reducing duplicate information
collection?
Ms. Utgoff. Again, I don't think we could give you a
numerical answer to that question. But there are many
instances. For instance, the International Price Program. If we
are able to use the Census data, we will be able to reduce the
number of people who have to respond to the survey, and we will
be able to reduce the information that we get from each
respondent.
Ms. Cooper. I might just add there, from the Census
Bureau's point of view, that if we were able to reduce one
survey across our broad set of businesses that we tend to
survey, that, in and of itself, would save us some $2 million.
So I think we can begin to see that there are some real savings
out there as we go forward, but it is just very difficult to
come up with a precise number.
Mr. Horn. Now I am going to yield 10 minutes, to start
with, with the gentleman from Ohio on any questions you want to
ask.
Mr. Sawyer. Thank you very much, Mr. Chairman.
Ms. Cooper, it's my understanding that the Census Bureau is
currently conducting research on matching data from household
surveys with data from business surveys and censuses. The idea
is to use them to model changes in labor force composition,
future pension demands, and a variety of other important
economic conditions. It's my understanding that under the terms
of this legislation, that those data could not be shared with
researchers at BLS or BEA, despite the fact of the substantial
expertise in those agencies and the work that they might be
able to do to understand those data. Can you explain why the
legislation has been changed to exclude those data? Or am I
incorrect in my understanding of it?
Ms. Cooper. That this legislation has been changed to
exclude it?
Mr. Sawyer. I believe so.
Ms. Cooper. I honestly don't know the answer to that one.
Ms. Utgoff. There were questions--this bill deals only with
the information for firms. It does not deal with individual
household response. In the previous efforts to have this
legislation passed, the household response became very
controversial, so it was not included in this round of the
bill.
Mr. Sawyer. I would simply hope that as we gain experience
with this kind of information sharing, that we could find a way
to enable that kind of sharing. It is important. I couldn't
agree with you more about the sensitivity of personal
individual household information. But where it can be used to
shape policy in important ways, I think it would valuable to
do.
It leads me to my second question: Does the administration
have a plan of action with regard to how, in future, this kind
of legislation might be expanded to include sharing household
data in order to improve our understanding of the changing
nature of poverty, access to early childhood education, pension
coverage, and a myriad of other kinds of social statistics that
shape an awful lot of the debate that characterize our work
here?
Mr. Kroszner. There is no specific plan right now. I think
what we want to do is gain experience with the data sharing
with the business data to avoid the concerns and controversies
that had been raised with the individual data. And I think as
we have the experience with that, both the researchers outside
of the Government and internally, we will be able to see how
best to shape the memoranda of understanding to make sure that
no data--no confidentiality agreements are violated.
And so I think it is an important foundation and first
step. Much like we had some experience with sharing some
international data back in 1990, I think that provides the
basis for how well the agencies can work with that data to show
that this will be functioning very well. And then perhaps in
the future we can take it another step, but there is no
particular plan right now for that next step to be taken.
Mr. Sawyer. Go ahead.
Ms. Cooper. And could I just add, Congressman Sawyer, going
back again to your question before and tying it to this one. I
agree that the work that is being done that ties the labor
market and other information together, the Census Bureau, is
innovative and can be very helpful longer term, but as long as
we--at this stage we can only do that in the confines of the
Census Bureau. But as my colleague has indicated, once we can
demonstrate that there will not be concerns longer term--I am
not sure how long that takes, but we are all hopeful that it
will be sooner rather than later--then perhaps this could be
considered later, much later.
Mr. Sawyer. I appreciate the concerns that you have for
confidentiality, which is what led me to my half of this
legislation. I don't want those concerns, however, to stand in
the way of serious innovation that can come about as a result
of better sharing.
Let me ask each of you to respond to the fact that the
administration sought and was granted a provision under the
PATRIOT Act that provides the Attorney General access to
individually identifiable survey records held by the National
Center for Education Statistics. If the administration sought
access to similar records held by the Census Bureau or BEA or
even the Council of Economic Advisers, would you support or
oppose access to survey records for law enforcement purposes?
Each of you, please.
Mr. Kroszner. Well, we have no records at CEA. So, for our
data, we are happy to share what we have with anyone.
Certainly, that's one issue that has arisen about maintaining
confidentiality versus collecting data for law enforcement
purposes, and there can sometimes be a tension there.
Mr. Sawyer. They are very much in tension, and I appreciate
that. I shouldn't interrupt. Go ahead.
Mr. Kroszner. No, no. And so we have tried to ensure that
nothing that we have done in this legislation would, in any
way, inhibit the ability of the Department of Justice to mete
out justice to wrongdoers. But I don't think we at CEA have a
particular view on that broader question.
Mr. Sawyer. Others?
Ms. Cooper. I don't think I have more to say on that,
either. I think that is something that has to be settled
elsewhere, and that is a very real tension and concern.
Ms. Utgoff. I agree with the other panelists on this point.
Mr. Sawyer. Let me touch on something that the chairman
touched on. This is kind of off the track. Nearly 20 years ago
when I was a mayor in Akron, Ohio, we had gone from an--about
80 years where our signature industry was in the tire and
rubber industry. It was 1984, and we hadn't made a passenger
car tire in Akron since 1979. And the truth is, we were trying
to chart a new future for ourselves as a community with some
sense of realism.
We had done a great deal of work in expanding the product
applications of synthetic materials and other kinds of
polymeric applications in a wide range of different kind of
product fields. And, but before we committed ourselves to that
kind of work and to bringing together resources, as a city we
decided we wanted to get a sense of the current state of play
in what we were loosely referring to as the polymer industry,
and discovered to our frustration that it was very difficult to
track that because standard industrial codes simply did not
reflect the way in which the industry had shifted its field,
not just individual companies--B F Goodrich, by the middle of
the 1980's, was no longer a tire company, they were a chemical
and aerospace company. But that's different from when an entire
industrial field shifts its ground.
Can you talk to me just briefly about the how the ability
to share data will help you track not the changing character of
companies, but the changing character of large-scale enterprise
in the United States?
Mr. Kroszner. By being able to--actually, in some sense, by
being forced to resolve anomalies between the different
agencies when they've classify one firm one way and another
firm another way, that forces the agencies to address exactly
that issue much more head-on than they otherwise would have to,
because suddenly they now have the same firm classified in two
different ways, and neither of those classifications is
appropriate for that firm anymore. And so by talking to each
other, they'll say, well, maybe our classifications aren't
appropriate. We have to try to build on our standard industrial
classifications to take into account this dynamism.
And so I think that is one of the ways in which could help
to have the agencies speak with each other, because they can
then use their expertise together to say there is something
wrong here, we've both misclassified this firm, it should be
something new, and we have to innovate to come up with a new
classification for it.
Mr. Sawyer. Thank you, Mr. Chairman.
Mr. Horn. Thank you. About 5 years ago, probably at maybe
the same relevant question of this, I wanted, when I had all
those brains looking at me, and I knew you had the answer. And
that is, how does the OMB have one set of assumptions for their
accuracy of Federal statistics, and the CBO on Capitol Hill use
different assumptions? And is there any way we can get both of
those wonderful, powerful operations that they can agree on a
base? And how can we handle that? I realize that might have
nothing to do with what you are saying. But I just want that
now that I have got a few bright economists.
Mr. Kroszner. Well, I think it is a very important issue,
because it is sharing of a different type of sense of data and
different types of assumptions across different groups. I think
this is certainly an area of--in which reasonable people can
differ about looking out, let's see, 10 years hence about what
economic growth will be.
Actually, the long-term economic forecasts which we
developed through the process of CEA and OMB and Treasury, much
of the economics is actually quite similar to what we find in
the CBO forecasts for--especially for the long-range growth
assumptions? There is some differences on how the business
cycle will move over the short-term and then differences on
views on how you turn GDP into revenues. And so I think
increasing the dialog between our groups would be very--would
certainly be very valuable, but I don't unfortunately have sort
of the magic bullet that can make us all agree. And if you look
at private sector forecasters, they often have very different
assumptions that they make, and in some sense that's
appropriate that people have different views, and those should
be taken into account. I am not sure that we want to have just
one single view. I think we are relatively close; we're not too
far apart, but I think further dialog would certainly be
helpful.
Mr. Horn. Ms. Cooper.
Ms. Cooper. I might just add that indeed they certainly
have been fairly close in the last year, 2 years. They
operate--the two organizations operate under different rules in
terms of what the CBO's goal is and what the OMB's goal is. And
the timing of their forecasts clearly is slightly different,
not largely different. So I am impressed that they are as close
to one another as they are, especially after they have a little
bit of time to readjust.
But I would agree with my colleague that it's good to have
not just one view going forward. It would be surprising if we
had terribly different points of view on potential GDP growth
and we actually don't. It's trying to wrestle at this point
more than anything else with what level of revenues come with
each dollar of GDP.
Mr. Horn. Ms. Commissioner.
Ms. Utgoff. I don't have anything to add to that. I agree
with what the panelists have said.
Mr. Horn. Well, maybe we will get something out of the next
panel. Because up here--and, you know, various presidents have
said, well, gee, I've got OMB; at that point it is what the
President wants to have done. And then up here, we'd like to
having something because, we are trying on spending and we are
trying on not spending. And so everything we do, going through
the appropriations, we have got to show them that this is going
to be in either 5 years or whatever. And that tightens up
things quite a bit around here. And so I still think there
ought to be some way that they get together and they say, look,
this is what is here and this is what we ought to know, and
deal with it.
So, then in other administrations, why, they say, gee, we
like what CBO did and so forth. And it's all back and forth.
It's like a tennis game and little people return running every
day over the, with getting the ball halfway over the net. So,
anyhow, that's one little gripe I have. And I don't think it
will ever be solved, but it would be nice if we had sort of a
treaty of Versailles here for economics and, you know, maybe
the mirrors in the palace of Versailles would do it with all
the things they can get it through the sun, through the windows
and everything.
Mr. Sawyer. That was as remarkable of a mixed metaphor as I
think I've heard all day. I just thought maybe we could think
of these guys as line judges in your game of tennis.
Mr. Horn. My colleague here and I have both been dealing
with the European parliament since we got here, and he's very
good with that group. And so I will leave it right there.
OK. Is there anything that we should have asked you that we
didn't ask you? Well, that's to make a good conscience
somewhere.
But we will get the next panel up. And then if we could get
all--if you can stay, we can get the questions with all of you
there.
We have Maurine Haver, William D. Nordhaus and Ralph
Rector. They were previously sworn. Let's take Dr. Haver, start
with her. She is the chair of statistics committee of the
National Association for Business Economics.
STATEMENTS OF MAURINE HAVER, PH.D., CHAIR, STATISTICS
COMMITTEE, NATIONAL ASSOCIATION FOR BUSINESS ECONOMICS; WILLIAM
D. NORDHAUS, PH.D., STERLING PROFESSOR OF ECONOMICS, DEPARTMENT
OF ECONOMICS, YALE UNIVERSITY; RALPH RECTOR, PH.D., RESEARCH
FELLOW AND PROJECT MANAGER, CENTER FOR DATA ANALYSIS, THE
HERITAGE FOUNDATION
Ms. Haver. I am pleased to testify today in support of H.R.
5215. I am speaking in my capacity of chair of the statistics
committee of the National Association for Business Economics,
on behalf of our 3,000 members. The members of our association
have a keen interest in the quality of economic statistics
produced by the BEA, Census and the BLS. We use these
statistics daily in our work to help our companies and clients
make informed business decisions that have real dollar
consequences.
We believe that passage of this legislation will protect
the confidentiality of our companies' proprietary information
supplied to the Government. It will minimize the burden imposed
on our companies by the statistical agencies because duplicate
surveys can be eliminated. And, finally, it will improve the
quality of our national economic information, especially
statistics at the industry level, because of more complete and
consistent source data.
The business community and financial markets derive
significant benefits from the collection and dissemination of
economic data. Complaints of respondent burden are often
misinterpreted. Most companies recognize the value of
Government statistics and actively use statistics produced by
these agencies for the basis of many operation and planning
decisions. Companies need industry and national statistics to
have highest quality and are willing to do their part as long
as confidentiality of their proprietary information can be
assured and data collection is done efficiently.
This bill is important because it extends confidentiality
protection for respondents to all Federal agencies that collect
data for statistical purposes under a pledge of
confidentiality, and it prohibits the use of those data for any
other purpose. It also specifically prohibits disclosure of
information under the Freedom of Information Act. This uniform
set of confidentiality protections will do a great deal toward
reducing concern about reporting to national Government
agencies.
I believe the prior witnesses have devoted sufficient time
to talking about the improvements that data sharing will make
in our statistics. Let me just say that current statutory
barriers to the sharing of business data do result in duplicate
surveys that not only increase respondent burden, but also
introduce classification errors that reduce accuracy. This bill
will address both of these obstacles.
In summary, we believe this bill will encourage business
participation in Government surveys and will improve the
quality of the statistics available to business and
policymakers.
We strongly urge passage of H.R. 5215. I would like to
thank you, Mr. Chairman, for your invitation to participate in
this hearing today, and also for your efforts ever since 1995,
when I was President of the NABE. You were there to help us
work toward a more efficient Federal statistical system, and we
greatly appreciate that. I would be happy to answer your
questions.
[The prepared statement of Ms. Haver follows:]
[GRAPHIC] [TIFF OMITTED] T8326.018
[GRAPHIC] [TIFF OMITTED] T8326.019
[GRAPHIC] [TIFF OMITTED] T8326.020
Mr. Horn. Thank you, and we will pursue the questions with
the two our presenters here.
William D. Nordhaus, Dr. Nordhaus, Sterling Professor of
Economics, Department of Economics at Yale.
Mr. Nordhaus. Thank you very much, Mr. Chairman. I would
like to--I have some prepared remarks which I would like to
submit for the record.
Mr. Horn. That is automatically put in the record.
Mr. Nordhaus. I will summarize those briefly. I am
delighted to have the opportunity to discuss the proposal for
the sharing of statistical data in H.R. 5215. I think it is an
important and useful bill and urge its passage.
I am involved with a number of different groups that are
involved with economic statistics. I won't go over all of
those, but they do include one that is closely related to the
statistical agencies that I am the chair of the advisory
committee of the Bureau of Economic Analysis. That committee
has--actually saw an early draft of the data sharing proposal a
couple of times in its oversight meeting and discussed those.
I emphasize that the remarks I am making are only my own
and don't implicate any other organizations who like to have
their own views. But I did do a nonscientific sample of a
number of statisticians and economists and they are highly
enthusiastic of this bill.
Good economic statistics, as you know, are important
because they are critical inputs into the decisions of public
and private decisionmakers, the Congress in its budgetary
decisions, companies on their investments, State and local
governments on their infrastructures, and the private sector
and households on their financial decisions. All of these
issues involve and require good statistical information.
Earlier this year, the Commerce Department conducted a
brainstorming session of leading academic and business
economists to consider improvements in the national economic
accounts. And then earlier this year, the Joint Economic
Committee held some hearings where it inquired into some
different proposals for improving the Federal statistical
system.
I appended at the end of my testimony a summary list of the
recommendations that I made to the JEC. One of the major
recommendations I made was that the Congress should move ahead
expeditiously with improved data sharing among statistical
agencies.
Now, the Federal--I would just say one word about source
data. This is one of the less romantic parts of the statistical
system that people don't really know much about. We see
published in the newspaper every day the data on the GDP or the
inflation rate or the balance of trade.
But these are really just the visible tips of the
statistical icebergs, and below the surface lies vast volumes
of source data from all corners of the economy, and they are
collected by the Census Bureau, the Bureau of Economic
Analysis, the Bureau of Labor statistics, the IRS, the Federal
Reserve and many other Federal agencies, and the quality of our
economic statistics depends crucially on accurate, timely and
comprehensive source data.
Now, this bill--there are many ways to improve source data,
but this bill proposes one that is extremely economical and
useful, and that is solving the difficulties that arise from
the decentralized nature of our Federal statistical system.
The current system has many advantages, but one big
disadvantage is that agencies have a variety of statistical
data that they cannot share. In a sense the Government has
imported data in its left pocket, but that data cannot be moved
to the right pocket, and that restriction just make no sense
and should be lifted.
There are many examples of how data sharing will improve
the quality of Federal economic statistics. I will concentrate
on the national income and product accounts, which is the
system I know best. I will just mention four briefly.
One is that early estimates of our gross domestic product
are subject to large revisions because the source data are
sparse and often based on voluntary reporting.
With data sharing, BEA will be able to use statistical
techniques to correct the data for omissions to get more
accurate early estimates of quarterly GDP.
A second problem is annual revisions which come in the
middle of every year. And these are often large because many
monthly Census surveys are based on voluntarily supplied data.
The July 2002 revisions were particularly large because some of
the data that came in during this year showed revisions that
were far different from the preliminary data.
By working with individual company data and comparing them
with publicly available data, BEA can identify discrepancies
earlier and thereby reduce annual revision.
A third example is the statistical discrepancy between the
product and income side of our accounts. This is currently
running at $166 billion in the last quarter, which is more than
1\1/2\ percent of total GDP.
The source of the statistical discrepancy is still a
mystery. But by comparing IRS, Census, and public data, BEA may
be able to sharpen its estimates of different sources,
particularly of income, and narrow that discrepancy.
A final example, and one that has been very much in the
news, involves data on corporate profits. These are one of the
most important and hard to measure of the statistics. And
accurate measures of total corporate profits are produced only
with a 3-year delay, and this is because it takes that amount
of time to gather all of the different tax returns and tabulate
them completely.
Inaccurate profit data may well have contributed to the
boom and bust cycle of stock prices in the last few years. I
believe that by triangulating data from tax returns, quarterly
financial reports and publicly available financial statements,
BEA can develop statistics on corporate profits more accurately
and in a more timely fashion, and this, of course, can help
investors gauge the true movement of profits in an era when
financial reports are not always reliable.
As Dr. Kroszner noted, these statistical innovations can
improve the quality of Federal statistics with little, no or
even negative cost.
So, in summary, I think the proposal for data sharing
contained in H.R. 5215 is a small but important step toward
improving the efficiency and the use of Federal statistical
resources, and I support strongly its enactment.
[The prepared statement of Mr. Nordhaus follows:]
[GRAPHIC] [TIFF OMITTED] T8326.021
[GRAPHIC] [TIFF OMITTED] T8326.022
[GRAPHIC] [TIFF OMITTED] T8326.023
[GRAPHIC] [TIFF OMITTED] T8326.024
[GRAPHIC] [TIFF OMITTED] T8326.025
[GRAPHIC] [TIFF OMITTED] T8326.026
Mr. Horn. Dr. Ralph Rector, is the research fellow and
project manager at the Center for Data Analysis at the Heritage
Foundation. Welcome.
Mr. Rector. Thank you. Chairman Horn, thank you for
inviting me to testify today. My name is Ralph Rector. I am the
Project Manager of the Heritage Foundation's Center for Data
Analysis. It should be noted that the following testimony is my
own view, not necessarily that of the Heritage Foundation or
any other organization.
Today, I would like to discuss three standards I believe
should guide any proposal to improve America's statistical
system. These standards are, first, protection of individual
identity for the respondents who provide data; second,
production of useful, timely information for data users; and
third, independent evaluations of data for decisionmakers. I
think of these as the three I's of statistical policy: Identity
protection, information value, and independent evaluation.
The sections concerning statistical efficiency contained in
title II of H.R. 5215 are examples of measures that can enhance
information value by improving the accuracy and timeliness of
economic data.
My testimony will focus primarily on the issues related to
title I, Standard 1, identity protection. Those who provide
data to statistical agencies should not have to worry that the
data they provide to the Government will be used against them.
In addition, statistical agencies must protect the identity of
individuals who provide data that may eventually be released to
the public. Provisions for protecting individual identities can
be found in plans such as H.R. 5215, which clearly distinguish
between statistical and nonstatistical data.
Also, these prohibit the release of data in a form that
could reasonably be expected, either directly or indirectly, to
yield the identity of the respondent.
Standard 2, information value. Although necessary,
procedures that protect confidentiality also tend to reduce the
amount and the value of data that can be released. It is not,
however, necessary to adopt such extreme forms of data
suppression as those found in H.R. 5215.
As currently written, this bill states that agencies cannot
disclose data in identifiable form. The bill further defines
identifiable form to mean representation of information that
permits information about a respondent to be reasonably
inferred through either direct or indirect means.
This method of protecting confidentiality precludes the
disclosure of all individual level information that respondents
would provide, despite the use of safeguards that protect the
identify of the respondents. The problem with H.R. 5215 arises
because it does not clearly distinguish between the identity of
the individual respondent and the information they provide.
Denying researchers access to all individual level data
would drastically reduce the value of publicly available
information and would undermine the quality of important
research conducted in the United States.
Standard 3, independent evaluation. Although valuable, it
is not enough for Government statisticians to approach data
availability solely in terms of the amount of data they
provide. In addition, the data should be sufficient so that
researchers outside the Government can respond effectively to
evaluate the proposals, either to validate them or challenge
them.
My written testimony includes two examples that may help
clarify why the distinction between the amount and the form of
data accessibility is so important to nongovernment researchers
who provide public policy analysis.
To implement three statistical standards described in my
testimony, I believe Congress should, with regard to identity
protection and information value, provide guidelines similar to
H.R. 5215. However, the guidelines should be modified to
clearly indicate that confidentiality applies to the identity
of the respondent. The current version of H.R. 5215 is not
sufficiently clear in this respect.
With regard to independent evaluation, Congress should
require that whenever possible Federal agencies provide data to
independent researchers in a form that permits them to conduct
complete and independent evaluations.
Thank you.
[The prepared statement of Mr. Rector follows:]
[GRAPHIC] [TIFF OMITTED] T8326.027
[GRAPHIC] [TIFF OMITTED] T8326.028
[GRAPHIC] [TIFF OMITTED] T8326.029
[GRAPHIC] [TIFF OMITTED] T8326.030
[GRAPHIC] [TIFF OMITTED] T8326.031
[GRAPHIC] [TIFF OMITTED] T8326.032
[GRAPHIC] [TIFF OMITTED] T8326.033
[GRAPHIC] [TIFF OMITTED] T8326.034
Mr. Horn. Thank you. Let's go to a few questions here.
Where in the bill would you like, Dr. Rector, to solve that
problem?
Mr. Rector. I believe that the problem exists in the
definition of identifiable form, and I have provided specific
line numbers to the staff so that they can see exactly what the
language would, I think--it would need to be in order to very
clearly distinguish between the individual respondent, their
identity and the information that they provided.
Mr. Horn. And my staff has provided you a copy of the
amendments I intend to offer at today's markup on H.R. 5215.
Does this amendment satisfy your concerns with the original
language of the bill?
Mr. Rector. I have read the substitute amendment to H.R.
5215 provided by your staff, and I do believe that the new
version does correct this problem.
Mr. Horn. So that solves that problem, good. Something is
happening today anyhow.
Let me ask you a few questions that my colleague didn't
have a chance to do it, Dr. Sawyer, who was needed elsewhere.
And he and I both worry about the Attorney General, who has
sought access to survey information provided by individuals.
What would be the effect on economic statistics of information
collected from businesses for statistical purposes if it was
used in legal proceedings by the Government against the
businesses and the executives that provided that information?
How do you feel about that one way or the other?
Ms. Haver. I am only speaking for myself now, because we
certainly haven't queried our members. But I would not be
supportive of the Attorney General having access to information
that is provided to the Government for statistical purposes.
Mr. Horn. Dr. Nordhaus.
Mr. Nordhaus. My own view again, it depends a little bit on
the context. But it seems to me that whether it is a person or
a business responding to a survey for statistical purposes, I
think it would be very chilling for--to get good responses to
that if there were the possibility of it being used in legal
proceedings. I think that is particularly applicable for
voluntary surveys. For mandatory then the person would have
some--would have a tug of war between which of the two
provisions were more--he was more fearful of.
But in the case of voluntary surveys, people would say,
well, it is voluntary, why should I fill it in if there is any
chance that this would be used against me in a legal
proceeding. So I think that is a serious, serious concern.
Mr. Horn. Dr. Nordhaus, the Federal Economic Statistics
Advisory Committee is a unique committee in that it is charged
with addressing the interaction among these three statistical
agencies. Would you please comment on the role that the
committee might play in advising those agencies in carrying out
the intent of this legislation?
Mr. Nordhaus. The FESAC, as it is called, is an agency or
was a group that was set up a couple of years ago. The Bureau
of Labor Statistics was the lead agency on that, although the
Census Bureau and the Bureau of Economic Analysis were also
participants. I was--I have met with that as--in terms of my
advisory capacity of the BEA.
I think all of these advisory committees can play a very
useful role in terms of reviewing proposals like this. As I
think I mentioned in any oral remarks, the BEA advisory
committee actually discussed the proposal that Dr. Kroszner
brought before it earlier this year. I think it was a very
useful discussion among the different people, because people
from different points of view, from business, from the research
community and from Government, all had something to say.
So I think this is a useful forum, these are useful forums
for discussions of those kinds of issues. It has been useful in
the past, and I hope it can be again in the future.
Mr. Horn. Mr. Sawyer's question is to Dr. Nordhaus. While
this bill provides a mechanism to bring the collective talents
of these three agencies to bear to improve economic statistics,
it does not provide a mechanism to draw on the expertise in
universities and private research organizations.
Do you have thoughts on how we can take that next step?
Mr. Nordhaus. Well, I--there are two separate issues here.
One is the actual production of statistics. I regard producing
the GDP and the CBI as production. I mean it is a very
intellectual and high level production, but it is like
producing cars in the sense you have got to roll them out every
month.
There is a separate question, which is the research that
lies behind those statistics, and there is where a very useful
role can be played. Some of the agencies, particularly the
Census, has taken the lead on this, have basically taken
researchers as employees of those statistical agencies to help
do research on behavior, on the behavior of particular series
or relationships, and I think those have been very useful in
bringing academic research to bear on the questions.
But for the most part, actually I think the research staffs
of those agencies are very high level, and I think the data
sharing will go a long way in improving some of the easily
fixed problems with the statistics.
Mr. Horn. Dr. Haver, do you agree with him?
Ms. Haver. Yes, I do. I think that certainly the production
of the statistics is something that requires compromises that
sometimes our academic colleagues would prefer not to see. We
occasionally find the theory is wonderful, but then there is
the application of that theory, and that sometimes becomes much
more difficult. But I think that the academic community has a
lot to contribute and is doing so, for example, through
organizations like the BEA Advisory Committee which Professor
Nordhaus chairs and other organizations like these.
Mr. Horn. Dr. Rector, would you agree with both of those
colleagues?
Mr. Rector. I would, but I would like to followup on a
comment just made about the compromises. As I had indicated in
my testimony, I think that there is a tradeoff between
protecting individual identity and information value.
The protection of identity tends to reduce the quality and
the value--the amount of information that can be released. I
believe that the agencies' data disclosure review boards, the
boards that make many of the compromises that were just
described, they do not work closely enough with data users to
produce compromises that are needed, and I would encourage
Government statisticians to work with users more closely as
they go through their normal effort to review these tradeoffs,
make the decisions, and to involve data users earlier on in
that process.
Mr. Horn. Well, I think that makes sense. As a user of
Government data, what suggestions would you give to the Federal
statistical agencies as they implement the provisions of this
bill?
Mr. Rector. Well, again my focus is more on the title I
aspects, and those are very wide ranging. They don't just deal
with the business data. I think that there are many data bases
that the Government produces, not only that have to do with
statistical reports and the activities of the statistical
agencies, but there are many data bases that are produced in
conjunction with reports that Congress has mandated for policy
evaluation, policy review.
It is difficult, sometimes impossible, for independent
researchers, whether it is in a think tank, or whether it is in
the Academy, to gain access to these data bases. I would
encourage agencies, again not just the statistical agencies but
all Federal agencies that collect data, to require as part of
that, particularly the program evaluation studies, to include a
mechanism for the timely release of these data bases to
independent researchers.
Mr. Horn. I learned about 30 years ago that it was very
difficult to get faculties involved with the politicians, now
that I am one, and the problem is time, and we need it in 24
hours. They need the sabbatical every 7th year, and they will
think about it. So that is a little problem that we have got
there.
But what do you feel should be done by the Government side,
although you touched some of it, what can be done with these
data?
Mr. Rector. Well, specifically with regard to policy
evaluations, because so many of those are mandated by Congress,
I would like to see that data dissemination actually be built
into the requirement when a report is released, that the data
be made available at that point in time for independent review,
peer review, evaluation by other researchers, that just be part
of the grant process.
Mr. Horn. What do you see as the greatest hurdle in trying
to improve response rates from private sector companies?
Ms. Haver. I think it is a very big problem actually of--
certainly as corporations are having more and more financial
difficulties or are not performing as well as they might like,
they are looking at all corners of their business and trying to
make sure that every person in that company is doing something
that enhances profitability.
So clearly filling out Government forms isn't on that list,
or isn't high on that list. So it really is important that the
data that is released by the Government really accurately
reflect the industry, if it happens to be industry data, that
companies might actually find useful.
I think we will go a long way toward improving response
rates if in fact companies do believe that the information
reported on their industry is accurate.
We had a reference earlier today to the semiconductor
reporting problem. That was very simply that the companies did
not think that those statistics accurately reflected their
business. And so I think it is very important that we improve
those statistics. And, as was explained earlier, part of
improving our industry data is simply getting establishments
put into the right classification. And if our--if the study
that was done with 1994 data is accurate today, and I think it
is probably worse rather than better, then we have to believe
that only 70 percent of single establishment companies are
classified correctly.
And, therefore, you know, we have a lot of mistakes going
into that information, which explains why companies often don't
think they truly reflect the reality that is out there.
Mr. Horn. Dr. Haver, you indicate that data users in
industry are strong supporters of H.R. 5215. Do you know what
industry data suppliers think of the bill?
Ms. Haver. Well, after I was asked to testify today I
called up a variety of representatives of data suppliers, the
National Association of Manufacturers, the U.S. Chamber of
Commerce, the NFIB, the Semiconductor Industry Association, and
so on, to discuss this bill with them.
They were quick to point out that they really couldn't
speak on behalf of their members because they had not asked
their members their opinions on this legislation. However, they
did say to me that they could see absolutely nothing in this
bill that would cause their members, in their personal opinion,
to not support the legislation.
In other words, it seemed to them, in one person's words, a
no-brainer. This was legislation that improved assurances of
confidentiality for companies and at the same time provided our
fundamental general statistical agencies with the capabilities
of sharing information to really improve our data and to make
it more relevant.
So although I can't say that there is not some company out
there that might have a problem with this legislation, I have
to say, among the numerous organizations I did talk to, I did
not hear anyone express that view.
Mr. Horn. For all three of you, do you see any downside to
the bill from the viewpoint of industry? We have heard Dr.
Haver.
Ms. Haver. No.
Mr. Horn. Dr. Nordhaus.
Mr. Nordhaus. No. I think it is a plus. I think that we can
actually do more with the data that we have now with data
sharing. So I think it is a plus.
Mr. Rector. I am unaware of any downside.
Mr. Horn. OK. What do you see as the greatest benefits of
data sharing provisions for users of Government data like
yourselves?
Ms. Haver. More accurate, relevant data. I think that we
have discussed this really at some length today, so I won't go
through some of the points again. But there are many cases
where the ability to share information among our agencies will
at least give us the possibility of improving our statistics,
and there are many situations today where we know that somebody
is wrong.
You know, in 1997, information technology firms had a
certain level of employment, but if you compare the numbers of
the Bureau of Labor statistics and the Census Bureau for a
year--that was after all an economic census year--there is a
rather sizable discrepancy of--I don't quite remember the
number, but between 10 and 15 percent.
So these are the kind of anomalies that hopefully the
agencies can start investigating. They can try to figure out
why is it that BLS has more workers in this industry. So I
think what we are going to have will be data that are much more
consistent across our three agencies and hopefully, as I said
before, a better reflection of the truth.
Mr. Horn. As a user of Government data, what suggestions
would you give to the Federal statistical agencies as they
implement the provisions of this bill?
Ms. Haver. Well, I would say, first of all, they should
focus their energies where they are going to have the greatest
payoff, and one clear example is the business list. I would
like to see one business list, but I do know that Census
perhaps has its way of using its list and BLS has yet another
way of using its own. So at a minimum I hope we can get to the
discrepancies. But what I would like to see would be simply one
list.
I think also, and it may be very difficult for our agencies
right now given their budget constraints, at least the first
markups on the appropriations do not look very good for these
agencies. But I would like to see them investigate ways of
reducing duplicate surveys so that we can really go to
companies and say we are not going to ask this of you more than
once, that efforts are really being undertaken in the
statistical system to reduce the burden.
I think that would be a big selling point to companies when
a survey arrives to be filled out or when they receive a letter
asking them to participate in a survey.
Mr. Horn. Thank you, and I think that completes the
presentation, unless there is something any of you want to put
on.
And Mrs. Maloney, who has been an excellent ranking person,
she would like to submit a statement for the record. And
without objection, that will be in at this point.
And now I want to thank the staff of the subcommittee,
Bonnie Heald, the staff director back there against the wall.
And the gentleman doing all of the work here in many ways is
the senior counsel, Henry Wray, to my left, your right, and
counsel Dan Daly. Where is Dan Daly? Come on, don't be shy. Put
your hand up there, fellows. And Chris Barkley, majority clerk,
he is over there.
You know, when you get next to a wall, why, there is a
problem there. And then minority staff member, David McMillen,
professional staff, back here. He gives us a lot of advice. And
Jean Gosa, minority clerk, and she is with the staff of the
minority. And our court reporters, and we are delighted to have
them, and that is, Desirae Jura, and Mark Stuart. Thank you
very much. We appreciate all of the work you have done over the
years.
So, with that, we thank you all and wish you well.
[Whereupon, at 3:10 p.m., the subcommittee was adjourned.]
[The prepared statement of Hon. Carolyn B. Maloney
follows:]
[GRAPHIC] [TIFF OMITTED] T8326.035
[GRAPHIC] [TIFF OMITTED] T8326.036
[GRAPHIC] [TIFF OMITTED] T8326.037