Balancing Flexibility and Accountability: Grant Program Design in
Education and Other Areas (Testimony, 02/11/98, GAO/T-GGD/HEHS-98-94).

Pursuant to a congressional request, GAO discussed balancing flexibility
and accountability in federal grant programs, focusing on education
programs.

GAO noted that: (1) flexible grants are an adaptable policy tool and are
used in fields from urban transit to community mental health; (2) they
are alike in that each addresses a national purpose but gives state or
local grantees the flexibility to adapt funded activities to fit the
state or local context; (3) however, there are vast differences among
them as well; (4) some offer flexibility within a narrow range, as do
many so-called categorical programs, while others offer choices so broad
that they come close to resembling revenue sharing; (5) these
differences reflect three critical design features on whether: (a) the
national objectives involved are performance-related or fiscal; (b) the
grant funds a distinct program or contributes to the stream of funds
supporting state and local activities; and (c) it supports a single
activity or diverse activities; (6) in combination, these features are
associated with differences in flexibility, accountability, and the
level of government that is accountable for performance; (7)
combinations that produce greater flexibility lodge accountability at
the state or local level and complicate the task of obtaining
program-wide measures of performance through grantee reporting; (8)
additional sources of information may be needed to support program
decisions at the national level; and (9) considering design features and
their implications can help policymakers ensure that accountability and
information needs are met, whatever type of design is selected.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  T-GGD/HEHS-98-94
     TITLE:  Balancing Flexibility and Accountability: Grant Program 
             Design in Education and Other Areas
      DATE:  02/11/98
   SUBJECT:  Accountability
             Education or training
             Block grants
             State-administered programs
             Intergovernmental fiscal relations
             Federal/state relations
             Reporting requirements
             Congressional oversight
             Grant administration

             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


Cover
================================================================ COVER


Before the Education Task Force
Committee on the Budget
U.S.  Senate

For Release on Delivery
Expected at 2:00 p.m.  EST
Wednesday,
February 11, 1998

BALANCING FLEXIBILITY AND
ACCOUNTABILITY - GRANT PROGRAM
DESIGN IN EDUCATION AND OTHER
AREAS

Statement of Susan S.  Westin, Associate Director
Advanced Studies and Evaluation Methodology
General Government Division

GAO/T-GGD/HEHS-98-94

GAO/GGD/HEHS-98-94T


(966708)


Abbreviations
=============================================================== ABBREV


BALANCING FLEXIBILITY AND
ACCOUNTABILITY:  GRANT PROGRAM
DESIGN IN EDUCATION AND OTHER
AREAS
====================================================== Chapter SUMMARY

GAO's testimony discusses balancing flexibility and accountability in
federal grant programs, with particular attention to programs in
education.  Today's statement previews the results of a comparative
study of flexible grant programs--in transportation, health, social
services, education, criminal justice and employment--to be completed
later this spring.  GAO's study covered all of the currently
operating 'block grants' plus other grant programs with similar
characteristics.  The testimony also reflects GAO's work on program
evaluation and information issues and on the accountability and
fiscal effects issues associated with block grants.  Finally, it
draws on GAO's many studies of education programs. 

Flexible grants are an adaptable policy tool and are used in fields
from urban transit to community mental health.  They are alike in
that each addresses a national purpose but gives state or local
grantees the flexibility to adapt funded activities to fit the state
or local context.  However, there are vast differences among them as
well.  Some offer flexibility within a narrow range, as do many
so-called 'categorical' programs, while others offer choice so broad
that they come close to resembling revenue sharing. 

These differences reflect three critical design features:  whether
the national objectives involved are performance-related or fiscal;
whether the grant funds a distinct 'program' or contributes to the
stream of funds supporting state and local activities; and whether it
supports a single activity or diverse activities.  In combination,
these features are associated with differences in flexibility,
accountability, and the level of government that is accountable for
performance.  Combinations that produce greater flexibility (such as
are typical of K-12 education programs) lodge accountability at the
state or local level and complicate the task of obtaining
program-wide measures of performance through grantee reporting. 
Additional sources of information may be needed to support program
decisions at the national level. 

Considering design features and their implications can help
policymakers ensure that accountability and information needs are
met, whatever type of design is selected.  The testimony closes with
a design framework in the form of a decision tree to assist in this
task.  GAO applies the framework to two programs:  the Title VI
Innovative Education grant program (with primarily fiscal objectives)
and the Safe and Drug Free Schools and Communities local grants (with
performance objectives). 


BALANCING FLEXIBILITY AND
ACCOUNTABILITY:  GRANT PROGRAM
DESIGN IN EDUCATION AND OTHER
AREAS
==================================================== Chapter STATEMENT

Mr.  Chairman and Members of the Education Task Force: 

I am pleased to be here today to discuss balancing flexibility and
accountability in federal grant programs, particularly in education. 
With block grant proposals on the horizon and the first performance
plans coming due under the Government Performance and Results Act,
this topic could hardly be more timely.  Our testimony responds to
the program design issues raised in previous hearings before this
task force and to your request that we provide a comparative
perspective on these issues and their implications for future
education programs. 

Flexible grant programs across the federal government take many
forms.  Three federal program design features appear critical:  (1)
objectives--whether the national objectives involved are
performance-related or fiscal, (2) nature of operations--whether the
grant operates as a program or a funding stream, and (3) diversity of
activities--whether the grant supports a single activity or diverse
activities.  In combination, these features are associated with
differences in flexibility, accountability, level of government that
is accountable for performance, availability of performance
information, and the need to draw from additional sources in order to
obtain the information needed to support program decisions at the
national level.  Considering design features and their implications
can help policymakers ensure that accountability and information
needs are met. 

This testimony previews the results of our comparative study of
flexible grant programs--in transportation, health, social services,
education, criminal justice and employment--to be completed later
this spring.  Our study covered all of the currently operating 'block
grants' plus other grant programs with similar characteristics.  The
testimony also draws from past GAO work on program evaluation and
information issues and on the accountability and fiscal effects
issues associated with block grants.  Finally, it draws on our
studies of education programs, many of which have been described in
testimony at earlier hearings. 

I will begin by considering the design of flexible programs
generally.  What are the key design features of flexible programs,
and how do they vary from program to program?  What issues of
flexibility, accountability and information do these features raise? 
This portion of the testimony will draw on examples from across
government, with particular attention to elementary and secondary
education.  Finally, I will present a framework for examining current
or proposed programs as they come before you for reauthorization or
authorization. 


   BACKGROUND
-------------------------------------------------- Chapter STATEMENT:1

Flexible grants are an adaptable policy tool and are used in fields
from urban transit to community mental health.  They are alike in
that each addresses a national purpose but gives state or local
grantees the flexibility to adapt funded activities to fit the state
or local context.  However, there are vast differences among them as
well.  Some offer flexibility within a narrow range, as do many
so-called 'categorical' programs, while others offer choice so broad
that they come close to resembling revenue sharing. 

These differences can be traced to a few key design
features--objectives, nature of operations, and diversity of
activities-- each representing an important policy choice.  Each
feature by itself has implications for accountability and
information. 


   KEY DESIGN FEATURES
-------------------------------------------------- Chapter STATEMENT:2


      OBJECTIVES: 
      PERFORMANCE-RELATED OR
      FISCAL
------------------------------------------------ Chapter STATEMENT:2.1

Our first feature concerns the nature of the national objectives to
be served through the federal grant program.  We are not speaking
here of such broad, ultimate national purposes as decreasing poverty,
but rather of the more immediate, direct, and concrete objectives to
be attained through the provision of grant funds.  Grant programs'
objectives can be characterized as either primarily
performance-related or primarily fiscal. 

  -- Performance-related objectives focus on services.  In our study,
     we found such objectives as:  expanding services, improving
     service quality, increasing coverage of targeted problems or
     populations, or achieving specified service outputs or outcomes. 
     For example, the central objective of the grants for Special
     Programs for the Aging--Nutrition Services is to provide
     nutritious meals to needy older Americans in such a way as to
     reduce social isolation.  Among grants for elementary and
     secondary education, national objectives like ensuring access to
     services, equity in educational opportunity, and targeting
     services to groups that are educationally disadvantaged have
     historically been important, reflecting civil rights concerns. 

  -- Fiscal or financial assistance objectives focus on providing
     dollars.  Typical fiscal objectives include increasing support
     for meritorious goods or under-funded services and targeting
     grant funding to needy jurisdictions.\1 For example, the
     objective of the Title VI Innovative Education grants is to
     provide funds to support local educational reform efforts. 


--------------------
\1 These and other fiscal objectives are discussed in Federal Grants: 
Design Improvements Could Help Federal Resources Go Further
(GAO/AIMD-97-7, Dec.  18, 1996). 


   NATURE OF OPERATION:  PROGRAM
   OR FUNDING STREAM
-------------------------------------------------- Chapter STATEMENT:3

A second critical feature concerns whether national objectives should
be achieved through a grant-specific operating program or simply
through adding to the stream of funds supporting ongoing state or
local programs.  An operating program is a "program" in the
common-sense meaning of the term.  (The Aging Nutrition Program again
is an example.) It has performance requirements and objectives and
carries out distinct program-wide functions through a distinct
delivery system in such a way that grant-funded activities, clients,
and products are clearly identifiable.  Criteria of successful
operation such as those outlined in our oral testimony last November
apply.  These include clear focus, stated objectives, monitoring and
reporting in terms of those objectives, and research-based
activities. 

Grants that operate as a funding stream are not federal "programs" in
this sense.  Here, the federal agency provides funds that are merged
with funds from state or local sources (and sometimes from other
federal sources as well) to support state or local activities
allowable under the flexible grant.  The grant is one funding source
among many.  The "programs" supported are state or local programs. 
For example, funding stream operation is typical in elementary and
secondary education, which is a state responsibility administered by
local education agencies.  Federal funding for kindergarten through
grade 12 (K-12) programs may go directly to local education agencies. 
Or, it may go to the states, either for provision of state-run
programs and services or to be passed through to local agencies under
state oversight and accountability. 


      ACTIVITIES:  SINGLE OR
      DIVERSE
------------------------------------------------ Chapter STATEMENT:3.1

The third critical feature concerns diversity of activities.  Some
flexible grants, as in the senior nutrition example, focus on a
single major activity or limited set of activities common to all
grantees.  At the other extreme, some allow unrestricted choice among
a wide variety of allowable activities.  For example, states with
Preventive Health and Health Services grants can choose among 60
allowable uses of funds. 


   IMPLICATIONS FOR FLEXIBILITY,
   ACCOUNTABILITY, AND INFORMATION
-------------------------------------------------- Chapter STATEMENT:4

These design features define the flexibility given to grantees,
accountability for performance, and the level of government at which
performance is managed.  Flexibility reflects the presence or absence
of national performance objectives.  Its range is most limited when
the grant funds a single major activity through a national program
and broadest when the grant functions as a funding stream and permits
choice among activities. 

Accountability for performance is established by the inclusion of
performance objectives--and provisions that implement them--in a
grant program.  When objectives are purely fiscal, accountability to
the federal agency focuses on fiscal matters.  For example, if the
national objective is to encourage states to provide more of a
nationally important service (like assisted housing), states may be
held accountable for using grant funds to supplement rather than to
replace or supplant their own spending on that service. 

Programmatic objectives and operation as a national 'program' give
the federal funding agency a role in managing performance under the
grant.  In programs with primarily fiscal objectives and those that
operate as a funding stream, the activities supported are managed at
the state or local level.  In the words of agency staff (quoting
state officials), "these aren't federal programs, they are state
programs that receive federal funds." The federal agency's role is
limited accordingly, and may involve little more than seeing that
applications for funding are properly submitted, compliance or audit
issues resolved, and money disbursed in timely fashion.  Where
grant-funded activities are managed at the local level as in the two
education programs we studied--Title VI Innovative Education and Safe
and Drug Free Schools and Communities--the state's role may be
similarly limited. 

Design features also affect the availability of performance
information.  Operation as a program simplifies the task of getting
uniform information about performance attributable to grant funds. 
It is possible to identify which activities were supported, the
amount of federal funds allocated to each, and, to various extents,
the results of federal support.  Operation as a funding stream
complicates this task.  When grant funds are part of a stream, it is
possible to identify which activities federal funds supported and the
amount allocated to each.  But once added to the overall budget for a
state or local activity, federal dollars lose their identity and
their results cannot be separated out--particularly when the federal
share is small.  Thus, the only program outcome measures available
are likely to be for the state or local service delivery program, not
the federal funding program. 

As the above discussion suggests, it is the combination of features,
rather than any single feature, that has implications for
flexibility, accountability and information. 


   DESIGN FEATURES IN COMBINATION
-------------------------------------------------- Chapter STATEMENT:5

Examining how the design features were used in programs across
government, we identified four major combinations or design types. 
We have summarized them in table 1, which shows design features,
examples, and summary comments associated with each type.  As the
last column indicates, state or local flexibility and control over
performance objectives and performance management increase as you
move down the table. 



                                     Table 1
                     
                       Grant design features in combination

                                Diversity of
Objectives      Nature          activities      Examples        Comments
--------------  --------------  --------------  --------------  ----------------
Performance-    Operating       Single major    Job Training    ï¿½Federal role
related         program         activity        Partnership     substantial
                                                Act Title II-   ï¿½State or local
                                                A and II-C      flexibility
                                                                narrowest
                                                Special         ï¿½Most likely to
                                                Programs for    include national
                                                the Aging -     service outcome
                                                Nutrition       objectives
                                                Services        ï¿½Most likely to
                                                                have program
                                                                performance
                                                                information at
                                                                the national
                                                                level

                Funding stream  Single major    Child Care and  ï¿½Federal/State
                                activity to     Development     balance an
                                diverse         Block Grant     issue
                                activities                      ï¿½Flexibility
                                                Preventive      varies with
                                                Health and      diversity of
                                                Health          activities
                                                Services Block  ï¿½Performance
                                                Grant           objectives and
                                                                measures may be
                                                                state or local
                                                                ï¿½Less likely to
                                                                have national
                                                                performance
                                                                information

                Operating       Diverse         Community       ï¿½State or local
                program         activities      Development     level dominant
                (project)                       Block Grant -   ï¿½Flexibility is
                                                Entitlement     broad
                                                                ï¿½Most likely to
                                                                have performance
                                                                and evaluation
                                                                information at
                                                                the project
                                                                level

                Funding stream  Single major    Title VI -      ï¿½Federal role
                                activity to     Innovative      confined to
                                diverse         Education       providing funds
                                activities      Program         ï¿½Flexibility
Fiscal                                          Strategies      varies with
                                                                diversity of
                                                Social          activities
                                                Services Block  ï¿½Broadest
                                                Grant           discretion to
                                                                grantee
                                                                ï¿½Least likely to
                                                                have performance
                                                                information
--------------------------------------------------------------------------------
I will review and illustrate the four combinations briefly, and then
move on to the accountability and information issues they raise. 

Our first type of grants pursue performance-related objectives
through a distinct operating structure (top row).  Grants of this
type are closest to the conventional notion of a "program." They
typically focus on a single major activity and include program-wide
performance objectives and sometimes even service outcome objectives. 
Because of this, these grants are most likely to be able to collect
uniform information about performance from grantees.  For example,
the national objectives of the Job Training Partnership Act are to
provide job training that leads to increases in employment and
earnings of youths and adults facing serious barriers to
participation in the work force.  To evaluate the results of the
program in achieving these objectives, recipients are required to
provide counts of activities provided, demographic characteristics of
individuals served, employment outcomes, and program costs.  No major
K-12 education grants are of this type. 

Our second type covers performance-related, funding stream grants
(second row)--grants that involve national performance objectives,
yet operate through state or local programs.  Most programs of this
type cover a state or local function or delivery system (such as
preventive health) involving various activities.  National
performance objectives typically concern system improvement or
capacity-building, ensuring access to services, service quality, and
targeting of activities to priority populations.  Several grants in
this group require state or local grantees to set their own
performance objectives of various kinds.  Provisions of the
Preventive Health and Health Services Block Grant, for example,
require each state to fund activities related to Healthy People 2000
objectives and to measure and report the progress of the state in
meeting the objectives selected. 

Our third type includes grants with fiscal objectives (third row)
that provide support for program-like--rather than ongoing--state or
local activities.  These activities often take the form of
projects--similar to operating programs in having clear boundaries,
but with a clear start and finish as well.  Grant provisions identify
allowable uses of funds and may include national criteria for
selecting activities, such as the benefits test that applies to
projects supported by Community Development--Entitlement block
grants.  Otherwise, performance objectives and measures are set at
the operating level.  Under the Drug Control and System Improvement
(Byrne Formula) Grant Program, for example, states are required to
set performance objectives for activities that are funded and to
evaluate the success of these activities in achieving those
objectives. 

Our fourth type concerns fiscal funding stream grants (bottom row). 
They allow a broad range of activities are the classic block grant
design of the early 1980s.  Consistent with their purpose, grants of
this design typically require only the information needed to
determine how much was spent on each activity and verify that funds
were used for allowable purposes and that any requirements related to
fiscal objectives (such as maintenance of effort) were met.  Some of
these programs make an effort to get service output information (such
as client counts), but even this can be difficult.  For example,
where actual counts of recipients served are not available, the
Social Service Block Grant accepts counts based on estimation
procedures which may vary in their statistical validity. 


   BALANCING FLEXIBILITY AND
   ACCOUNTABILITY AND IMPLICATIONS
   FOR INFORMATION
-------------------------------------------------- Chapter STATEMENT:6

These four design types present very different situations with
respect to grantee accountability--what grantees are held accountable
for and the level of government that is accountable for
performance--and the information needed to support it.  They also
differ with respect to the information needed to support program
decisions at the national level and prospects for getting this
information through grantee reporting, as opposed to other means. 


      GRANTEE ACCOUNTABILITY AND
      THE INFORMATION NEEDED TO
      SUPPORT IT
------------------------------------------------ Chapter STATEMENT:6.1

As a previous GAO report has noted, accountability is an elusive
concept whose meaning depends on the context.\2 At a minimum, all
grantees are accountable to the federal level for financial
management and for using funds to support allowable activities. 
Beyond that, what grant recipients are accountable for to the federal
level varies from grant to grant, reflecting (1) the type of
objective, and (2) if performance objectives are involved, whether
the federal level manages the program or merely adds to the stream of
funds supporting state or local programs.  I will now describe the
situation for each type of grant, with a focus on performance issues. 

Accountability for performance to the federal level is most extensive
in grants that include national performance objectives and operate as
distinct programs--grants with the most limited flexibility.  As
mentioned previously, programs of this type commonly collect and
report information in line with their performance objectives, which
may be concerned with program implementation, outputs, or (when
possible to measure) direct outcomes of services.  (Indirect or
ultimate outcomes are another matter, which we will discuss in the
next section.)

Objectives, information and reporting are similarly lined up in
programs with fiscal objectives that operate as a funding stream. 
But here, accountability focuses on fiscal matters.  The funding
agency is accountable for ensuring compliance with fiscal objectives. 
However, the activities funded are under state or local direction,
accountability for performance is to state or local authorities under
whatever arrangements they have put in place, and performance
information does not necessarily flow to the federal funding agency. 
If specific fiscal objectives (such as maintenance of effort) apply,
grantees are accountable for meeting them and for providing fiscal
information to verify that they have been met. 

The grants that combine federal performance objectives with operation
through state or local programs--a common pattern across K-12
education--present puzzling performance accountability issues,
particularly for service outcome objectives.  Activities supported
with federal funds and the information collected about performance
often differ from state to state.  (This difficulty affects
fiscal-objective operating programs as well.) While state or local
program outcomes in total may be measurable, the component
attributable to federal funding cannot be separated out.  Thus,
measuring performance at the level of the federal program may not be
feasible.  For accountability purposes, measuring overall performance
of the state or local program would not be appropriate, particularly
when the federal grant contributes only a small fraction of the cost. 

Reliance on delivering services through state and local programs or
projects thus raises other delicate issues of federalism as well. 
First, there is the question of whether achieving a given national
objective through existing state and local programs is feasible. 
This question is particularly relevant with respect to new service
outcome objectives, such as decreasing drug use among students. 
State and local programs designed with different objectives in mind
may have difficulty incorporating this new objective.  Or, conditions
that enable achievement of that outcome (such as solid knowledge of
how to produce it) may not be met. 

Assuming that operation through state or local programs is feasible,
how can national grant programs encourage achievement of national
performance objectives and encourage accountability for performance,
yet respect state and local authority, interests, and differences? 
We found several approaches to this dilemma among our programs.  Some
approaches strengthen accountability to the state or local agency
that receives federal funds.  (They mitigate the risk that existing
state or local oversight and management arrangements might be
insufficient to ensure strong performance.) For example, in grants
with the national objective of improving the quality of service
delivery, we found provisions that direct states or localities to

  -- set service delivery or quality standards and monitor whether
     standards are being met (as in the Child Care and Development
     Block Grant) or

  -- set performance criteria for schools funded under the grant and
     procedures to ensure that schools that don't meet these criteria
     are identified and remedial actions are initiated (as under
     Education's Title I program). 

States and localities are then accountable to the federal agency for
implementing these provisions. 

The Department of Education has been experimenting with a different
approach.  The Department grants temporary exemptions from certain
federal program requirements (waivers) to states or school districts
that demonstrate that the waiver will lead to educational
improvements.  These waivers are intended as a tool to expand the
flexibility available to local school districts in exchange for
increased accountability for student achievement.  The results of
this experiment are not yet in. 

One final example of an approach to serving national objectives
through state or local activities relies on the techniques embodied
in the Government Performance and Results Act--that is, requiring
states or localities to set--and to report to the federal agency in
terms of--performance objectives for the activities or projects they
choose to support with federal funds.  Provisions of the Safe and
Drug Free Schools and Communities Act, for examples, require states
and local education agencies to establish drug use and violence
prevention objectives, report the outcomes of state and local
programs, and assess their effectiveness toward meeting the
objectives. 

Under this 'results' approach, accountability for performance remains
at the level of the state or local agency doing the reporting, not
the federal or state agency to whom the report is directed.  The
federal or state agency receives the information, but does not use it
for program management. 

This information, however, can be useful in assessing the degree to
which national objectives for the program are being met, a subject to
which we now turn. 


--------------------
\2 Block Grants:  Issues in Designing Accountability Provisions
(GAO/AIMD-95-226, Sept.  1, 1995). 


      INFORMATION TO SUPPORT
      PROGRAM DECISIONS AT THE
      FEDERAL LEVEL
------------------------------------------------ Chapter STATEMENT:6.2

To make decisions about the programs they oversee, congressional
committees need evaluative information--information that tells them
whether, and in what important respects, a program is working well or
poorly as well as whether performance objectives are being met. 
Performance data collected from grantees is an important source.  But
they have limitations, and other sources are useful as well. 

Let me describe for you some limitations of performance data in the
flexible grant context.  First of all, our study indicates that,
beyond simple activity counts, few grant programs are able to obtain
these data program-wide.  Collecting reliable uniform data requires
conditions (such as uniformity of activities, objectives, measures)
that are unlikely to exist under many flexible program designs, and
even where overall performance can be measured, the amount
attributable to federal funding often cannot be separated out.  And
some programs have ultimate outcome goals, such as increasing highway
safety, which are measurable only through aggregate data. 

More importantly, performance data cannot answer the full range of
questions that are likely to arise during congressional oversight. 
To determine how, and in what respects, a program is working well or
poorly, we have found that Congress is likely to need not only
performance information but also

  -- Descriptive information sufficient to convey a sense of the
     variety of conditions under which the program operates and how
     federal funds are actually being used.  (Our earlier study found
     that general summaries of activities performed or clients served
     do not meet this need.) For flexible grants, information that
     shows how grant funds fit into the context of other programs is
     of particular interest. 

  -- Information about program implementation, including attention to
     whether feasibility or management problems are evident and
     whether the methods used to deliver services are of known or
     likely effectiveness. 

  -- Information concerning positive or negative side effects of the
     program. 

  -- Information that will help them determine whether this program's
     strategy is more effective in relation to its costs than others
     that serve the same purpose.\3

Some of this information may be available from federal agency staff,
particularly if the agency plays an active oversight or technical
assistance role. 

Formal evaluation studies are a second likely source of these four
kinds of information, and they can gather performance information
from a sample of sites under controlled conditions that could not be
obtained through grantee reporting.  We found evaluation studies in
programs of every type. 

Information on the effectiveness of service delivery methods comes
largely from research and demonstration studies.  Knowledge to
support effective practice is well established in some of the subject
areas covered in our sample of grants, and was incorporated into
program provisions (such as service standards) or in companion
technical assistance or knowledge dissemination programs. 

Finally, aggregate measures of social, environmental, educational or
health outcomes can be useful.  However, since no single program is
likely to have much of an impact on these measures they are most
useful for judging the combined results of programs that share a
common goal. 

To close this section, let me note that each of the sources mentioned
is more likely to be available when backed by statutory authorization
and budget resources than when it is not.  As we observed in our
earlier study, Congress is more likely to get the information it asks
for and pays for. 


--------------------
\3 Program Evaluation:  Improving the Flow of Information to the
Congress (GAO/PEMD-95-1, Jan.  30, 1995). 


   SUMMARY AND DESIGN FRAMEWORK
-------------------------------------------------- Chapter STATEMENT:7

In summary, the design of a flexible grant program involves choosing
among policy options that, in combination, establish the degree of
flexibility afforded to states or localities; the relevance of
performance objectives for grantee accountability; whether
accountability for performance rests at the federal, state, or local
level; and prospects for measuring performance through grantee
reporting.  Fortunately, performance measures are not the only source
that Congress can draw on in making program oversight decisions. 

We have developed an exhibit that depicts the grant design policy
choices discussed in this testimony and factors that might be
considered at each point in the form of a decision tree.  Each choice
has implications regarding the degree of flexibility provided states
or local entities, the type of performance information that can be
available, and the level at which this information can be used for
accountability purposes.  Our framework assumes that the decision to
establish a formula grant program (as opposed to another type of
program, such as a demonstration) has already been at least
tentatively made.\4

   Figure 1:  Grant Design
   Framework

   (See figure in printed
   edition.)



   (See figure in printed
   edition.)

To illustrate how the exhibit flows, let's consider how it would
apply to the Title VI Innovative Education Program Strategies grant
program.  The objectives of the grant, to support local education
reform and innovation, are primarily fiscal, putting us on the upper
decision path on our diagram.  Funds may be used to support local
projects (such as magnet schools), but Title VI's purpose does not
require that project-level performance objectives be set so we
continue to the step of designing provisions to match fiscal
objectives. 

Title VI has such provisions, stating that grant funds may not be
used to supplant funds from non-federal sources and the state must
maintain prior levels of fiscal effort.  To obtain information
required for accountability, the program requires local districts to
describe their intended uses of funds and how these will contribute
to the grant's objectives of supporting reform.  States, drawing on
district records, must report biennially on general uses of funds,
types of services furnished, and students served.  As these data are
of limited utility for program oversight, the Congress mandated
national evaluation reports on this program in 1986 and 1994.  The
1994 report provided information about federal share, the size of
state and local grants, how funds were used, the minimal performance
accountability requirements imposed by states, and the difficulty of
evaluating a program that provides supplemental resources for other
activities. 

To further illustrate, let's consider the Safe and Drug-Free Schools
and Communities grant, focusing on funding at the local level.  Grant
funds support activities that serve national performance objectives
of preventing violence in and around schools and the illegal use of
alcohol, tobacco and drugs.  The presence of these objectives puts us
on the lower, performance-oriented path of the diagram.  Funded
activities are not implemented through a national operating program,
but rather through state and local programs--reflecting at least the
hope that national objectives could be achieved through these
programs.  However--and here's an example of some of the federalism
issues we mentioned earlier--some national program provisions do
apply.  Local programs must be comprehensive and convey the message
that the illegal use of alcohol and other drugs is wrong and harmful. 
These national requirements notwithstanding, the local education
agencies are responsible for setting performance goals, deciding how
to pursue them, and reporting to the state in terms of those goals. 

Moving along the state and local path on our diagram, we come to the
question of whether drug and violence prevention programs function as
distinct projects or like a funding stream.  A recent evaluation
study--and here we jump ahead a little--suggests the latter. 
Examining what appeared to be comprehensive school-based drug
prevention programs, this study found so much variation within
districts in what was being done that local activities hardly met our
definition of a 'program.' As to the feasibility question on the
diagram, collecting performance data--beyond student counts--for drug
prevention programs has proven difficult.  Reporting requirements
make reference to local program outcomes, but states are simply asked
to provide whatever relevant data they can.  Reflecting these
limitations, provision has been made to gather data from other
sources, including state-level data from national surveys of youth
drug use, for program oversight.  Although the Department of
Education is required to produce a report on the national program
every three years, the lack of uniform information on program
activities and effectiveness may limit its usefulness.  An evaluation
study covering the period 1990-1995 provided insight into resource
adequacy, the extent to which activities reflect research findings,
implementation issues, student outcomes, and state and local
evaluations.  Further evaluation studies are planned. 


--------------------
\4 Other types of programs are discussed in the Program Evaluation
report previously cited. 


------------------------------------------------ Chapter STATEMENT:7.1

Mr.  Chairman, this concludes my illustration of the decision diagram
and also my prepared statement.  I appreciate the opportunity to
share this information with you and would be pleased to respond to
any questions you or members of the Task Force may have. 

RELATED GAO PRODUCTS

Federal Education Funding:  Multiple Programs and Lack of Data Raise
Efficiency and Effectiveness Concerns (GAO/T-HEHS-98-46, Nov.  6,
1997)

Managing for Results:  Analytic Challenges in Measuring Performance
(GAO/HEHS/GGD-97-138, May 30, 1997)

Federal Grants:  Design Improvements Could Help Federal Resources Go
Further (GAO-AIMD-97-7, Dec.  18, 1996)

Safe and Drug-Free Schools:  Balancing Accountability with State and
Local Flexibility (GAO/HEHS-98-3, Oct.  10, 1997)

Block Grants:  Issues in Designing Accountability Provisions
(GAO/AIMD-95-226, Sept 1, 1995)

Block Grants:  Characteristics, Experience, and Lessons Learned
(GAO/HEHS-95-74, Feb.  9, 1995)

Program Evaluation:  Improving the Flow of Information to the
Congress (GAO/PEMD-95-1, January 30, 1995)


*** End of document. ***