Page images

contain clauses reflecting the mutual interests in any changes and the nature of both parties' rights to terminate. The specifics of who does what and who is responsible for what would be expressed in the instrument that defines the relationship.

The need for the cooperative agreement mode was recognized while examining projects or programs in which grants and contracts were used interchangeably. It became apparent that the grant/contract confusion could be reduced only by segregating from both grants and contracts the relationships that were causing the confusion. Thus, the concept of a class of cooperative agreements is essential. Without it, there will be: (1) grants with all types of Federal involvement and, thus, a substantially reduced chance of getting a useful degree of Government-wide consistency in the assignment and understanding of respective responsibilities, and (2) a continuation of the pressure to move, under the rationale of the need for agency control, some of the cooperative agreement kinds of transactions into the procurement area.

tions which, in the judgment of the agencies, call for their use.

In our review of Federal grant-type programs we found that in most programs the agencies were using instruments consistent with our recommendations, that is, based on the need for agency involvement during performance. In a significant number of programs, however, the agencies were not using instruments consistent with our recommendations. This usually occurs because (1) the authorizing statutes require the use of a grant even though the agencies believe that they should be substantially involved with the recipient in performing the work being supported or stimulated, or (2) the agencies lack grant or cooperative agreement authority. These statute-based barriers to consistency discourage agency and executive branch efforts to achieve uniformity.

Statute-caused inconsistency is a serious problem in the newer agencies. HUD's funds are earmarked for grant programs or contract programs. In UMTA and LEAA programs, where the authorizing statutes require the use of grants, there is a great deal of agency and recipient uncertainty or concern as to roles and responsibilities. LEAA is criticized for doing too little and UMTA for doing too much. It is possible that many of these relationships are cooperative-agreement types of situations in which the instruments should be tailored to the requirements of each situation by negotiation between the parties. If the parties knew that the kind of instrument used established the parameters of relationship, a more careful delineation of respective responsibilities, a greater acceptance of respective roles, and more effective performance would result.

Enactment of legislation providing the general authority we have proposed should give the agencies ample authority to assign their projects or classes of projects to the appropriate relationship categories. This does not mean, however, that the same or similar projects need be treated in the same way by all agencies. The relationship category in which a project or class of projects falls will be determined by the mission of the agency and, therefore, by how the agency views the particular project or class of projects. A given project may


Enacting legislation to distinguish assistance from procurement by standardizing instruments to reflect types of relationships would be an important step in reducing the present confusion. A second step also is needed. Most agencies do not have authority to use these instruments in the relationship situations that call for them. At present there is no general authority for the use of cooperative agreements, although a few agencies, such as AEC and NSF, have ample authority for such an instrument, and USDA does use such an instrument in some situations. Individual programs have specific grant or grant-in-aid authority for their implementation, but the general grant authority of Public Law 85–934 is restricted, being available only for basic research in institutions of higher education or in nonprofit organizations whose primary purpose is research. The agencies need general authority to use grants, grants-in-aid, and cooperative agreements in relationship situa

[blocks in formation]

be procurement in one agency and assistance in another. But the way an agency classifies its transactions will become a meaningful public statement of how it views both the basic nature of its relationships with the non-Federal sector and its responsibilities. It is probable, therefore, that the adoption of the proposed distinctions would have beneficial effects in that comparisons of various ways of imple menting programs would be easier.

The discipline enforced on the agencies by adherence to relationship-based instrument definitions should increase Government productivity. Federal agency involvement would be less a function of personal or agency preference or habit than is now the case. Each agency would have to decide before making grant-type awards what degree of involvement would be required for each class of transaction. Thus, a basic research grant might have standard cost principles, audit requirements, one or two routine types of approval, and standard technical and financial reporting requirements. A cooperative agreement might contain some of the same requirements or standards as the basic research grant, but it would also specify in detail the roles of the recipient and the agency in the light of the needs of the particular kind of project and the particular recipient. It might specify how the agency was to be involved, for example, by approving contracts for certain sorts of procurements or grants for substantive parts of the work; by requiring consultation or mutual agreement on certain types of decisions; by specifying for ongoing programs how new standards or administrative requirements are to be developed and approved.

The grant or grant-in-aid will emphasize the responsibilities of the recipient and the Federal agency in terms of basic Federal requirements and standards. These arrangements will entail routine Federal/recipient interaction during performance. The cooperative agreement would be a more tailored and explicit statement of how the parties will work together during performance to insure success of the project or program. It will entail active and substantial Federal/recipient interaction.

[blocks in formation]

instruments appropriate to several types of relationships, eliminate some problems and ambiguities, and provide a basis for greater Government-wide consistency. But the stimulus to achieve maximum efficiency, consistency, simplicity, and effectiveness is likely to come only from a Government-wide assistance system spelling out the rationale for and specific guidance on methods, techniques, and requirements for assistance transactions and relationships.

Such a system would illuminate grant-type programs and the ways they are carried out so as to permit public scrutiny and encourage better understanding and needed improvements. The rapid growth of grant-type programs in the 1960's and early 1970's has created tremendous stresses on an already overburdened Federal administrative structure. There is excessive proliferation of requirements, undue complexity, serious lack of coordination, and inadequate management.

Federal grant-type assistance should "assist” non-Federal recipients. The lack of a recognized system to assign responsibilities for grant-type assistance programs has unfortunate effects. Where there is or is considered to be recipient inadequacy, the Federal administrator tends to try to remedy it by developing more and tighter Federal rules, procedures, and standards. He then feels more secure in the face of the scrutiny to which he may be

subjected. The recipient tends to see his part in this process as one of becoming a routine applier of Federal rules and regulations, with the program losing the flexibility necessary for optimum effectiveness. Without welldeveloped concepts of what Federal/nonFederal relationships should be, the dynamics of the process are likely to work against recipient initiative, responsibility, and growth, and against effective performance of assistance objectives. It is likely, for example, that many of the Federal and non-Federal participants in assistance projects do not recognize the "project management” needs of some projects. Assuring adequate contractor project management in a procurement context is difficult enough. We have yet to understand the need for, much less provide, guidance on assuring adequate project management in the different, supposedly cooperative, and admittedly more delicate, assistance relationship.

In emphasizing recipient responsibilities, we cannot lose sight of the Federal responsibility for assuring the effective expenditure of public funds. Assistance programs must strike a careful balance between utilizing and encouraging recipient capabilities and providing the standards and technical assistance, including management assistance, needed to assure effective performance. In view of the size of the "Federal Assistance System,” this can be done effectively only with Government-wide consid

[blocks in formation]

eration of programs and the way things are done. A comprehensive system of guidance is necessary if Federal and recipient staff effort is to be used effectively in assistance programs.

We found that most representatives of recipient interest groups, particularly those of State and local government, see the advantages of a more rational assistance system. They see a possibility that such a system might enhance the cooperative or partnership character of some relationships, which would reduce what they see as the tendency of Federal administrators to "impose" conditions on recipients without adequate recipient opportunity to comment or negotiate on the conditions. Most State and local recipients would like to participate to a larger extent in processes that now are often closed to them. They would welcome a structure which offered promise of open discussion and a clearer delineation of the respective roles of the Federal agencies and themselves. Of course, many of those involved in categorical programs, in and out of Government, are likely to be distrustful of any change that might affect the operation of their individual programs.

The creation of an assistance system would give the Congress a better overview that would assist it in making clearer its intentions when it authorizes new programs. It also would facilitate congressional review and oversight of how the agencies perceive and handle these programs.

GAO, in carrying out its audit and review responsibilities, makes judgments on where audit effort and attention are needed. For this reason, the question of accountability, which is closely related to responsibility, is of interest to GAO. If a basic Federal system in terms of which agency and recipient roles and responsibilities can be clarified is established, it should also permit a sharpening of the meaning of accountability. A standard framework should facilitate GAO's efforts to achieve effective accountability to the extent that it recognizes degrees and kinds of Federal and non-Federal responsibility, as well as means of associating degrees and kinds of responsibility with types of instruments or classes of transactions.

OMB has recognized the need to bring greater order to Federal assistance programs.

OMB studies in 1966 and 1969 3 were directed toward research at institutions of higher education. The current OMB Federal Assistance Review (FAR) Interagency Task Forces seek uniformity and simplicity in requirements used in transactions with several classes of performers, units of State and local governments, educational institutions, hospitals, and other nonprofits, including community-based nonprofit groups. FAR generally has not endeavored to distinguish assistance from procurement in any systematic way nor has it tried to distinguish between grants and contracts. It recognizes that the issuance of Government-wide guidance, to the extent that it requires distinguishing types of transactions, would face agency opposition and statutory barriers which would be difficult or impossible to overcome in the absence of explicit statutory authority to do so. Thus, FAR's effort to obtain interagency consensus on requirements has tended to result in establishing requirements which are least common denominators of present practice. While this effort will be helpful, what is needed is not only a simplification or standardization of present agency practices, but also a thorough examination of all kinds of assistance programs in an attempt to determine what a system of guidance for them should be. The feasibility of developing a comprehensive system should be assessed. There has been no concerted effort to examine Federal assistance programs, much less grant-type assistance programs, in systems terms.

A System of Guidance

Recommendation 2. Urge the Office of Federal Procurement Policy to undertake or sponsor a study of the feasibility of developing a system of guidance for Federal assistance programs and periodically inform Congress of the progress of this study.

Much of the attention currently devoted to the hundreds of assistance programs is concentrated on achieving individual program ob

• U.S. Bureau of the Budget, The Administration of Government Supported Research at Universities, 1966, and Report on the Project Concerning the Policies, Procedures, Terms and Conditions Used for Research Projects at Educational Institutione, June 20, 1969.

jectives. Much less effort has been devoted to generalizing from the methods used in assist ance programs. If assistance methods can be standardized and cataloged, it should be possible to take a long step in the direction of consistency and simplicity, and at the same time enhance program effectiveness, by establishing a system of guidance for generic aspects of the management of assistance programs.

The system that needs to be developed should cover all types of assistance relationships. The need is to: (1) identify the assistance universe comprehensively; (2) examine existing techniques and related considerations; (3) generalize to the extent possible from such data; and (4) explore the possibilities of developing new techniques. An analysis and evaluation of assistance techniques should consider, in addition to the usual grant-type transactions, loans, direct payments, and all forms of nonfinancial assistance. It also should consider subsidies which usually are not regarded as "assistance.” It also may be desirable to consider the applicability of assistance techniques to “revenue sharing.” It is a reasonable guess that questions of accountability, oversight, and the degree of active and passive Federal involvement in revenue sharing will become issues of national importance. A systematic review of all forms of Federal assistance and their operational methods and techniques could assist in decisions on how new forms of assistance should be structured to achieve desired ends.

The major purpose of identifying the assistance universe is to permit generalization from the types of Federal/recipient interaction which occur. It should be possible to identify all types of functional purposes and then by questionnaire and interview obtain from the agencies and from recipients in each kind of program in-depth data on types of involvement or interaction.

By involvement we mean the formal and informal ways in which the agency and the recipient interact before, during, and after performance of the recipient activity. Formal interactions include the use of a program or financial plan, the approval of specific milestone decisions, the approval of contracts or grants let under the prime recipient transaction, and so on. These are "active" types of

involvement. There are also formal "passive" kinds of involvement, such as the use of cost principles or other standards or requirements which must be followed.

The clauses in contracts, grants, and other instruments establish formal parameters of involvement or interaction. Some of these clauses require or lead to specific kinds of interaction. For this reason it is necessary to identify and analyze the statutory requirements for assistance programs and compare them in selected programs with the requirements added by the agencies. Beyond formal transactional requirements, there are formal agency and recipient requirements (established by agency or recipient policies and procedures) which lead to interaction.

There are also informal kinds of interaction (persuasion, use of political influence, interpersonal relationships) which should be identified to determine their significance, whether generalization from them is appropriate, and how they relate to the more formal kinds of interaction.

The data developed could be examined by matrix analysis. One coordinate could list the types of involvement or interaction which are identified, for example, budgetary control, approval of change orders, etc. The other coordinate could list types of function such as research, education, construction, demonstration, planning, etc. These program functions can be subdivided to accommodate various program purposes. For example, different kinds of research and different kinds of construction may call for different kinds of involvement or interaction. Finally, factors such as the nature of the recipient (unit of government, nonprofit organization, or profitmaking organization), degree of matching required, and the relationship of audit review to agency accountability may need to be included in the matrixes. An analysis of such a matrix should permit useful generalization with respect to assistance methods, techniques, standards, clauses, and relationships.

The use of matrixes to develop data for generalizing assistance techniques or methods must be supplemented by consideration of related factors which have a direct or indirect

Some OMB FAR data were put in such a form and might be helpful in this connection.

« PreviousContinue »