Most Important Federal Education Programs Need Fundamental Overhaul
Testimony prepared for delivery to the Subcommittee
on Early Childhood, Youth and Families, Committee on Education and
the Workforce, U.S. House of Representatives
May 11, 2000
Chester E. Finn, Jr.
John M. Olin Fellow
Mr. Chairman and members of the Subcommittee, thank you for the
opportunity to talk with you today about a subject that's close
to my heart, important to the nation, and in urgent need of your
The programs and units that you are examining for reauthorization–the
National Center for Education Statistics (NCES), The National Assessment
of Educational Progress (NAEP), the National Assessment Governing
Board (NAGB)–are the irreducible core of the federal role in education.
Indeed, when the very first federal education department was formed
during the Civil War, it was an information agency, pure and simple.
In those days, Uncle Sam didn't administer complex, formula-based
funding programs, competitive grant programs, sprawling student
aid programs and innumerable school improvement programs. But the
Congress determined in 1867 that it was a legitimate federal responsibility
to "collect…such statistics and facts as shall show the condition
and progress of education in the several States and territories".
The original Department of Education was also charged by Congress
with "diffusing such information respecting the organization
and management of schools and school systems and methods of teaching
as shall aid the people of the United States in the establishment
and maintenance of efficient school systems…."
To all intents and purposes, this was the origin of the federal
role in education, and I submit that it remains the most important
thing that Washington does in this field (although civil rights
enforcement is a close second).
When I worked at the Education Department from 1985 to 1988, our
top priority for OERI–shared by Secretary Bennett and the rest of
the Reagan administration–was to strengthen the Education Department's
data-gathering and assessment work. We believed that the most powerful
lever for improving American education was ample timely and accurate
information about it, especially about the performance of its schools
To that end, we sharply increased the NCES budget, which had languished
for years, launched a number of new data-gathering efforts and strengthened
We appointed the "Alexander-James" panel on the future
of NAEP, which came back with pathbreaking recommendations for NAEP's
overhaul and expansion (especially the addition of state-specific
achievement data) and for the creation of what became NAGB.
We submitted legislation to enact the Alexander-James recommendations
and worked with the Congress to bring this about (in 1988).
After leaving the Department, I had the privilege of serving two
terms as a NAGB member (they were four-year terms in those days)
and of chairing the Board for its first two years.
My interest in and enthusiasm for these activities remain keen.
The past decade has, I think, amply demonstrated the value of reliable
education statistics and assessment information. Without them, I
think we would not have an education reform effort worthy of the
name. We surely wouldn't know how well it was working!
Unfortunately, the past decade has also shown how vulnerable these
activities are to all manner of interference, manipulation, political
agendas, incompetence and simple mischief. It turns out that they
are nowhere near to being adequately immunized against Washington's
three great plagues:
* The pressing political agendas and evanescent policy passions
of elected officials (in both executive and legislative branches)
and their appointees and aides.
* The depredations and incursions of self-serving interest groups
and lobbyists (of which no field has more than education).
* Plain old bureaucratic bungling and incompetence.
One could trot out numerous examples to illustrate these problems.
Maris Vinovskis has done a heroic job of documenting many of them,
both in his recent testimony to this Subcommittee and in his extensive
historical and evaluative writings on OERI, NAEP, NCES, NAGB, etc.
I will mention just two fairly recent (and linked) episodes: the
Education Department's willingness to allow Vice President Gore
to use the release of NAEP data to draw attention to himself and
his administration, thus turning what should have been a neutral
"information event" into a political one; and the still-murky
circumstances under which former Commissioner of Education Statistics
Pascal Forgione was eased out of his post. (Though much sand has
been thrown in the nation's eyes about this episode, it's my impression
that the chief reason Forgione was not retained by the White House
was his forthrightness in stating that the Vice President had erred,
i.e. his vigilance in defending the integrity of the data that his
agency was charged with collecting.)
What's important about these two episodes is how clearly they show
the inadequacy of present arrangements to insulate these vital government
functions from politics. The National Assessment Governing Board
can set a thousand policies to govern the release of NAEP data,
but the White House and Education Department are perfectly free
to ignore NAGB and do what they like when reporting NAEP results.
So much for an independent governing board! And despite the Congress's
effort to turn the post of Commissioner of Education Statistics
into one insulated from politics, Forgione's departure demonstrates
that this has not worked well.
More examples could be supplied. During my time on NAGB, a month
seldom passed without someone at the Education Department second-guessing
a NAGB policy decision or dragging their bureaucratic feet on its
implementation. But we don’t need lots of examples. The essential
point is that the present arrangement cannot shield the education
statistics and assessment activities from politics. Which means
it is unable to vouchsafe the integrity, timeliness, accuracy and
credibility of the resulting data.
This is not, for the most part, a matter of malign intentions,
but of bona fide differences in priorities combined with a flawed
structural arrangement–and the absence from the Education Department
of the kind of organizational culture that successfully shields,
say, the work of the Bureau of Labor Statistics from politics.
Matters would have been worse these past seven years if the incumbent
Education Secretary, Dick Riley, were not himself an honorable man
(and NAGB alumnus) with respect for honest data. He has sheltered
NAEP and NAGB on many occasions from inappropriate interference;
has generally respected NAGB's policy role; and has delegated to
that Board a number of important responsibilities vis-à-vis NAEP.
But even the estimable Dick Riley has made some ill-considered
appointments to NAGB. He evidently went along with Forgione's non-retention.
He apparently assented to the Vice President's abuse of the NAEP
The central point, however, is that the integrity of federal education
data–and the professional independence of those who gather, analyze
and disseminate it–ought not hinge on who happens to sit in the
Secretary's office. They shouldn't depend on individuals. They should
be cemented into the basic structure such that it doesn't matter
who occupies which office.
Data aren't much good if you can't trust them. Numbers are only
as helpful as the credibility attached to them. No matter how hotly
contested an education policy or program issue may be, all factions
should be able to agree on the facts. And it has to be the federal
government’s solemn obligation, in gathering and disseminating all
those numbers, to vouchsafe their integrity.
That means building a firewall between NCES, NAEP and NAGB on the
one hand, and Washington's innumerable political arsonists on the
In embarking on the NCES/NAEP/NAGB reauthorization process, I submit,
that should be the Congress's top priority: the creation of a structure
that assures the long-term independence and integrity of the data.
I recognize that not everybody thinks this is a good idea. In 1994,
the Congress reduced NAEP's independence and NAGB's authority–and
likely would have obliterated the very idea of an independent governing
board were it not for your own watchfulness, Mr. Chairman.
Today, too, forces are at work trying to shrink the federal role
in statistics and assessment. One of the tools they are employing
is to reduce the independent authority of NAGB. Both commercial
interests and political agendas are involved.
At the same time, other forces are pressing in the opposite direction,
seeking, for example, to beef up NAEP, to require participation
in it, even to turn NAEP into a kind of national "high stakes
test" by which states would be rewarded and punished for their
This is not the place to settle the desirability of those policy
ideas. (I’m personally uneasy about high-stakes uses of NAEP results.)
Nor am I going to venture into the minefield known as voluntary
national testing. My principal point today is that neither expansionists
nor shrinkers, neither detractors nor boosters, neither Republicans
nor Democrats, neither home schoolers nor governors, should be able
to treat federal education statistics as their own playpen. There
needs to be a structure in place that, frankly, makes it difficult
to do anything that threatens the integrity of the data.
This is not easy. It's famously hard to build independent structures
into the federal government, particularly if one also seeks to avoid
calcification. How, for example, to draw the line such that a program
remains adaptable to changing circumstances in the real world, and
responsive to evolving needs in the society, yet without being pushed
by special interests and tugged by election year politics.
It's a classic governmental dilemma–actually a constitutional dilemma–and
it's rarely handled well. Indeed, only a few such structures within
the federal government work well–and I can't name any that work
perfectly. But think of the government agencies and organizations
that connote independence and integrity, and it becomes clear that
this task, while authentically difficult, is not impossible. Think
of the Federal Reserve System. The Smithsonian Institution. The
Food and Drug Administration. The aforementioned Bureau of Labor
Statistics. The research protocols of the National Institutes of
Health. Even the Census Bureau, though enmeshed in a fractious debate
about which census methods to use and what questions to ask, has
not, to my knowledge, had anyone question the integrity of its data.
Can this not be done for education data, too?
Let me not be heard to say that reorganization is a silver bullet.
The success stories that I've mentioned are the result not just
of structural insulation but also of habit, culture, leadership
But some restructuring is also called for. And I believe it must
begin with a decision to extricate the statistics and assessment
functions from the Department of Education. I don't believe they
can be adequately insulated and immunized so long as they are subject
to its inherently political imperatives and organizational dictates.
I append to my testimony a statement of seven "principles"
for reauthorizing (and restructuring) the federal role in education
research, statistics, assessment and program evaluation. An earlier
version was circulated last week. It has been endorsed by a number
of former Education Department officials and other experts on education
research and statistics. Some of the signers have misgivings about
"agency proliferation". But nobody has a better solution
to the problem before us.
Note, too, that the administration’s bill would make sweeping structural
changes. It would abolish OERI as we know it, replacing it with
a research institute and NCES, evidently co-equal to one another
but both existing within the organizational framework of the Education
The administration’s bill would also make certain changes to NAEP
and NAGB, including the much-needed restoration of NAGB terms from
three years to four. This is important both for practical reasons
(the stuff that NAGB grapples with is genuinely complex and it takes
a year or two for a new member to master it) and to thicken the
political insulation a bit. (Let me note, though, that several valuable
NAGB members will end their terms in September as a result of the
three-year limit that’s in force today.)
But the administration bill doesn’t build a true firewall. It doesn't
even begin to. It leaves the research enterprise, the statistics
agency and NAEP at the beck and call of a dozen assistant secretaries
and at the mercy of thousands of bureaucrats, not to mention interference
from the White House and Capitol Hill. It leaves program evaluation
in the Department's policy unit, which suffers from an ineradicable
conflict of interest in this area. And it leaves NAGB dependent
on the goodwill of whomever occupies the Secretary's office–those
delegations of authority that currently give NAGB much of its freedom
can be undone at any time–and leaves in place today's awkward ambiguity
as to who really runs NAEP.
I judge that the necessary insulation cannot be supplied for these
functions so long as they remain within the Department of Education.
The way to shield NCES and NAEP and NAGB (and research and program
evaluation) from the arsonists is to extricate them from that organization
and set them free, by creating a new and fully independent Education
This would parallel actions that several states have taken to separate
the agencies that operate their schools and education programs from
those that monitor the performance of those schools and programs.
It’s what every corporation does that engages an independent auditor
to check the books of the company treasurer.
It’s what Congress did when it created a General Accounting Office
to keep watch over the performance of executive branch agencies.
This is how we should be thinking about federal education statistics
and assessment. It's folly to suppose that the units responsible
for running programs and institutions should also be expected to
report honestly on the performance of those same programs and institutions.
There's an inherent conflict of interest. They need to be separated,
just as an auditor is autonomous from the treasurer of the company
whose financial statements he is examining. If he weren't fully
independent--the shareholders' way of double-checking on the performance
of the company's management–he would not be trusted and his work
would have little value.
At this early stage of the Congress's deliberations about NCES,
NAEP and NAGB, it’s premature to venture too far into the weeds
of structural alternatives. At this stage, the principles matter
more than the particulars. (In time, of course, the details will
be crucial.) A number of different structures can be imagined that
might carry out the essential principles–and we must acknowledge
that none of them will get the job done unless structural changes
are joined by the requisite advances in habit, culture, leadership
Principles for Reauthorizing
OERI, NAEP and NAGB
(May 4, 2000; Revised May 10, 2000)
We have come to believe that the federal government's handling
of education research, statistics, assessment and program evaluation
needs a fundamental overhaul and complete reorganization. At a time
when the need for scientifically rigorous and objective education
research, statistics, assessment and evaluation is greater than
ever, the federal government has failed to live up to that challenge
despite having spent billions of dollars over the past quarter century.
At several recent House and Senate hearings, expert witnesses have
almost uniformly detailed the weak and disorganized state of federal
education research, program evaluation, and statistical efforts
in the Office of Educational Research and Improvement (OERI) and
the Planning and Evaluation Service (PES) in the U.S. Department
of Education today. Among the problems identified at these hearings
are: a) Despite an increase in funding in OERI since the mid 1980’s,
the funding has been spent on activities other than high quality
research, b) Congressional mandates on how OERI must spent its research
funds continue to hamper the ability of the agency to operate efficiently
and effectively, c) Federal involvement in educational research,
development, statistics, and evaluation has often suffered from
unstable and weak intellectual leadership, d) The agency lacks distinguished
and innovative researchers, e) Large-scale, systematic development
is largely absent at the Department of Education, f) Neither PES
nor OERI are providing a sufficient number of scientifically sound
and educationally relevant program evaluations, g) The Department
of Education has not always delivered timely and objective evaluations
of its major programs, h) The overall quality of the research, development,
and program evaluations needs improvement, I) Politics continue
to intrude in the operations of OERI and PES, and J) the National
Assessment of Educational Progress also suffers from its lack of
insulation and from ambiguity as to who has policy control.
It is time to reorganize the current structure and operations of
OERI, PES, and NAGB in order to build upon the best parts of the
current agencies while placing more emphasis on the independence,
integrity, and excellence of the units that deal with educational
research, statistics, assessment, and evaluation. More specifically
we suggest seven principles to guide Congress as it commences the
reauthorization of these agencies:
Education research, statistics, assessment and evaluation ought
not be subordinated to political needs, program concerns, reform
enthusiasms or policy preferences. One promising possibility is
to separate these functions from the Education Department and
lodge them in a new, independent agency, perhaps named the "Education
The new agency should be dedicated to the acquisition and dissemination
of timely, accurate information; the canons of scientific inquiry;
and the pursuit of truth, without fear or favor. The Education
Audit Agency should not be a program-operating agency, "improvement"
enterprise or reform operation. Its sole stock in trade is trustworthy
information by which the nation can appraise the performance of
its children and its educational institutions, evaluate efforts
to strengthen that performance, and develop knowledge by which
future such efforts may be more successful.
Though located within the executive branch, the new agency should
be structured to ensure maximum freedom from political manipulation
and interest group influence. Several structures should be considered;
the optimal model might resemble the Federal Reserve Board.
The Education Audit Agency should conduct all of the bona fide
research (but not the "improvement") functions of the
Office of Educational Research and Improvement (OERI), as well
as those of the National Center for Education Statistics (NCES)
and the National Assessment of Educational Progress (NAEP). It
should also have the capacity to provide objective, high quality
program evaluations. In conducting such evaluations, the Education
Audit Agency shall serve the interests of truth, timeliness and
rigor. (This does not preclude the Education Department from also
having a policy analysis unit, but it is expected that Congress
will ordinarily assign future program evaluations to the Audit
Agency, not to the Department.)
Affiliated with, but independent of, the Education Audit Agency,
the National Assessment Governing Board (NAGB) should be placed
in charge of all policies pertaining to NAEP. NAGB should itself
be strengthened through longer terms for its members, its own
nominating capacity, and other measures designed to ensure its
full independence and integrity. (The Audit Agency will administer
NAEP on behalf of NAGB.)
In its conduct of education research, the Education Audit Agency
should strive for scientific rigor, including, to the maximum
degree possible, randomized field trials.
Congress should consider overseeing the Education Audit Agency—and
ensuring the quality, integrity and independence of its work—through
a special joint committee (akin to the Joint Committee on Taxation).
William J. Bennett, Empower America
(former U.S. Secretary of Education)
Chester E. Finn, Jr., Manhattan Institute
(former Assistant Secretary for Research and Improvement)
William D. Hansen, Education Finance Council
(former Assistant Secretary for Management and Budget)
Tom Loveless, The Brookings Institution
Bruno V. Manno, The Annie E. Casey Foundation
(former Assistant Secretary of Education)
Diane Ravitch, New York University
(former Assistant Secretary for Research and Improvement)
Nina Shokraii Rees, The Heritage Foundation
Maris Vinovskis, University of Michigan
(former research adviser to the Office of Education Research and