Accompanying Report of the National Performance Review Office of the Vice President Washington, DC September 1993
The Agency for International Development (AID) has increasingly relied on personal service contractors (PSCs) to manage its assistance activities. Restricting staffing resources has contributed to deterioration of AID's ability to manage itself and its programs by shifting the duties of AID project officers away from hands-on project management into process management (e.g., preparing proposals and managing contracts).
The agency is being forced to shed, through attrition, much of its program implementation expertise, with a resultant lack of continuity in program administration. In a downward spiral, AID managers are increasingly preoccupied with contract management and lose touch with the reality of field work, further diminishing their effectiveness in overseeing the performance of contractors. Although the direct-hire workforce has declined, the increase in PSC levels has more than made up for the direct-hire decrease.
AID has about 1,700 Foreign Service employees, approximately 1,100 overseas and 600 in Washington; nearly 1,000 Foreign Service Nationals overseas; and about 1,500 Civil Service employees in Washington. Besides its direct-hire employees, AID will employ 7,000 to 8,000 personal service contractors (PSCs) in fiscal year 1993. About one percent of PSCs served in Washington in 1992 to support central programs and the rapid expansion of AID program activities into the former Soviet Union; the balance served abroad. Normally, authority to employ PSCs is granted only for contractors working overseas. About 10 percent of the PSCs are U.S. citizens.
The President's Commission on the Management of AID Programs (the Ferris Commission) noted AID's heavy reliance on contractors to carry out increasing amounts of project design and service delivery. The commission concluded that "a major factor which makes human resource management most difficult in AID is that external parties determine the annual operating expense funds and career personnel ceilings AID must adhere to without reference to the program to be carried out."(1) This refers to the separate operating expense appropriation from Congress and the full-time equivalent (FTE) ceilings imposed on agencies by the Office of Management and Budget (OMB).(2) Conflicts have existed between AID management's vision of the agency's overseas structure and the visions of Congress and OMB.
Declining levels of operating expense funds and FTEs have changed the role and makeup of AID's overseas workforce. In a constrained resource environment, the decision to turn to PSCs is a logical one for program managers. PSCs provide a flexible alternative to directhire employees, and are especially useful for quickly changing the workforce skill balance to react to evolving requirements. PSCs, doing much the same work that a direct-hire employee does, can be charged to program funds, rather than more tightly controlled operating expense funds. Direct-hire employees cannot have their salaries charged to program accounts. Consequently, among program managers, there is a movement away from the implementation of development activities to the management and monitoring of contracts.
According to the General Accounting Office (GAO), "Although overseas U.S. direct-hire employees spent 32 percent of their work years on project management, their responsibilities increasingly involved managing and monitoring contractors, rather than the technical aspects of project implementation. Accordingly, missions and overseas offices place great reliance on foreign national direct-hires and PSC to manage day-to-day project implementation."(3)
In fiscal year 1990, more than two-thirds of AID's overseas work years (totaling approximately 8,000 years) were expended by American and foreign national PSCs, who are not separately identified and reported to Congress. Foreign national PSCs made up 60 percent of the overseas workforce in 1990, followed by 15 percent for American direct hires, 12 percent for foreign national direct hires, and 13 percent for American PSCs.(4)
The reductions in direct-hire employee staffing have not been evenly distributed. According to the Ferris Commission report:
The continuing reductions in direct-hire levels required by [OMB] have hit the field missions the hardest. During the period FY '85-FY '89, U.S. direct-hire staff decreased by 2.6 percent while directhire staff overseas decreased by 14.5 percent. This occurred in the face of [Inspector General and General Accounting Office] criticisms that AID did not have sufficient project management in the field.(5)
AID determined those specific staffing reductions according to its own internal priorities.
The personnel system has not responded to the need for staff with training or experience in contract management.(6) GAO indicated that AID has not adjusted its culture, training, and personnel practices to the increased use of contractors to deliver services, noting that large numbers of AID employees consider contract management to be a low priority, despite the fact that the bulk of project officers' responsibilities increasingly revolve around contract development and management. GAO also noted problems with supervision of overseas contractors and data collection concerning the number and nature of personal service contracts.(7)
Currently, AID budgets salaries for direct-hire staff out of a central agency pool and assigns FTE levels to each operational unit. Budgeting salaries out of a central fund does not create incentives for managers to factor payroll costs into their planning or to consider potential savings when hiring. Giving line managers increased control over staffing resources is an important reinvention principle.(8) Managing the total AID workforce--direct hires as well as PSCs--through the budgets of line managers is the only way to effectively control the size of the workforce.
Relaxing FTE restrictions may not be the only action required to realign the AID workforce. A critical related issue is that Congress appropriates funds separately for operating expenses, the account from which direct-hire employees are paid. Reductions in AID's operating expense budget have recently kept the agency from using its full allocation of FTEs. Contractors can be paid out of program money because their work is generally tied directly to project implementation.
The proposal to grant line managers more direct control over staffing resources is written with the presumption that AID is already concentrating its overseas staffing resources by phasing out some programs and consolidating some services in regional offices. If these changes were not already underway, a further recommendation would involve using program money to fund direct-hire positions where legitimate direct program oversight responsibility exists.
The following actions should not be initiated until AID has produced a thorough budget and policy analysis and has put in place the reporting and monitoring systems to ensure accomplishment of its mission.
This should include (to the extent feasible given governmentwide classification constraints) the salary and benefits portions of the operating expenses. Both actions must be taken concurrently to fully implement the proposal. To the extent possible, pilot efforts should be initiated to enable line managers to manage staffing tied to resources rather than to work years. Managers should be rewarded for using staff more creatively to reduce costs.
2. AID should put systems in place to accurately monitor employee work years for both direct-hire employees and PSCs.
Staffing analysis should include the total AID workforce, not just direct-hire employees. AID should thoroughly understand the real number of work years required to carry out its programs and be prepared to report that information to OMB and Congress.
3. AID should undertake a full review of the impact of its policies and practices concerning use of PSCs.
Such a review should address the trade-offs between accountability and the need for staffing flexibility, and whether it is not more beneficial to define accountability as assessment of project and program impact rather than contract micromanagement.
Providing line managers more control over their total budget resources will create incentives for more efficient operation. Because of the administration's commitment to reduce the federal workforce, the combined totals of PSCs and direct-hire staff must be monitored carefully. Also, the multiple pay plans and retirement systems, as well as the high proportion of total operating expenses represented by salaries and benefits, require very close monitoring to avoid exceeding the amount budgeted.
The fiscal implications cannot be determined. Experience has shown, however, that giving line managers control of personnel resources often yields opportunities to reprogram salary money into staff development or other administrative support areas. By considering both the direct-hire and PSC workforce, reductions should be made in the total workforce size. The precise fiscal impact of the recommendation, however, cannot be estimated.
The Agency for International Development (AID) must significantly improve its information management capability to adjust to reduced funding levels and a field presence restructured in response to the end of the Cold War. Reductions in the level of operating expense funds between fiscal years 1993 and 1994 make it difficult, however, for AID to reengineer its processes. Without capital funding for information systems development, AID will be hindered in meeting its reporting requirements and empowering managers with useful project information. AID must invest in order to build its capacity to accomplish its mission.
Information Management Deficiencies. AID has been widely criticized for deficiencies in its information management. The agency's systems are considered antiquated and unintegrated and consequently do not support management decision making. The accounting system cannot comply with some federal accounting standards. Layers of redundancy exist in the current system structures. Management improvements proposed by various studies conducted over the last several years regularly involve the application of new information technology as integral parts of the corrective actions.
Deficiencies in both financial accounting and program information processing have combined to exacerbate AID's resource management problems. Financial management systems and operations in the agency have been reported by AID as high-risk material weaknesses under the Federal Managers' Financial Integrity Act since 1988.(1) The General Accounting Office (GAO) has criticized AID for its lack of effective funds control. GAO noted the proliferation of unintegrated systems, identifying more than 100. One core criticism has been that the financial accounting and control system and the mission accounting and control systems are not in conformance with government-wide accounting standards.(2) GAO recommended that AID develop a comprehensive business plan, centralize information resources management (IRM) leadership, incorporate IRM planning into agency strategic planning and resource allocation processes, develop system standards, educate agency decision makers on information initiatives, and complete initiatives to identify information management requirements.(3)
Unintegrated Reporting. The Joint OMB-AID SWAT Team, which comprised analysts from both agencies, reported that:
AID cannot assemble an "official" portrait of AID's large, diverse field activity portfolio because the agency's information systems lack essential data, are not coordinated, and do not collect information in a consistent manner.
The report continued:
AID uses many automated and manual systems to monitor field activity. AID's systems are not integrated, operate on older proprietary computer technologies, and are often duplicate and overlapping. This leads to inconsistent, inaccurate and incomplete reporting that managers frequently do not trust.(4)
During the National Performance Review (NPR) interview process, criticism was frequently expressed regarding the decentralized management structure that had developed at AID and an apparent failure by senior management to exert centralized policy control over automation initiatives. Recent responses to a worldwide AID request for comments indicated a strong field concern about the number and obvious duplication of headquarters reporting requirements.(5)
Consolidating Data. Corporate data sharing (centralized data shared by the entire agency) would serve to consolidate reporting and make higher quality and more comprehensive information available faster to both headquarters and field staff. Until AID has the telecommunications capability to provide connectivity with all of the field missions, the agency automation systems will be operating in a mixed mode, with some information coming in electronically and some by incompatible diskette requiring re-entry in Washington--a process that substantially increases the likelihood for data error.(6)
AID Information Systems Plan. AID has begun to respond to the criticisms. The agency has completed a comprehensive information systems plan (ISP) geared toward integrating the basic financial, procurement, personnel, and program management functions of the agency by sharing data electronically across office boundaries throughout the agency. A comprehensive business process reengineering initiative is under way as part of the ISP process. This ISP has been favorably reviewed by General Services Administration (GSA). The GSA team briefed AID officials, OMB financial and technical analysts, and NPR staff in June 1993 regarding the outcome of its review. The ISP was seen as technically sound but ambitious, given the lack of resources available to carry it out.
AID has actively pursued implementation of the Army Corps of Engineers Financial Management System (CEFMS) to replace AID's current accounting system. CEFMS was selected by AID over off-theshelf software because AID personnel believed it was: (1) from an agency with project management requirements similar to AID, and hence about 70 percent compatible without modification; and (2) available without licensing fees. AID will have to modify either type of software to suit its specialized requirements. AID contends that commercial software will require much higher levels of costly modification for full implementation. The system is currently undergoing extensive testing at AID. The new system is based on establishing centralized corporate data sharing and relating resources to program measures. AID has provided briefings on its findings to the departments of State, Agriculture, and Commerce, and the United States Information Agency.
Financial Constraints. Upgrading information management (along with other capital-intensive innovations) at AID has been seriously impeded by rapid expansion of the agency into the countries of the former Soviet Union without program reductions elsewhere. It has also been affected by constraints on the operational expenses account. These two situations have strained AID's ability to put its automation infrastructure in place. AID has requested funding for automation initiatives every year for three years and has received only partial funding. AID has limited authority, with OMB approval, to reprogram some project money into the operational expenses account. Such a transfer has been authorized by OMB and could be expanded with cost analysis by AID.
Funding Structures. Increasingly common in state and local governments, innovation funds allow organizations to invest in capital intensive information management hardware, training, or other infrastructure required to fundamentally reengineer work processes. As in the private sector, reengineering business processes depends on both the application of the right amount of capital investment at the right time and a measurable return on investment. A good example for AID would be development of a paperless travel management system, which would cost money to implement but would yield substantial savings in reduced staff-time requirements. Opportunities like this exist throughout the agency. One or more of the following approaches could provide a source of innovation funds for AID, if legislative and funding issues can be resolved:
Repayment to the Fund. Money advanced to pay for an innovation should be considered as a capital investment loan that must be repaid. A request should be accompanied by a comprehensive rate of return analysis that demonstrates whether investment in capital information technology will reduce future operating costs and yield a positive return.
The lenders, in this case Congress and OMB, need to have assurance that the resources will not simply fall into the general operating account and dissipate. Funds made available for capital projects must not be diluted. Objectives must be clear and progress measurable. A clear repayment structure must be agreed upon at the beginning, reflecting the useful life of the innovation, and any policy-related conditions must be laid out unambiguously. Ideally, the repayment should originate with the end users themselves based on actual use of the innovation.
In AID's case, given the current financial constraints under which the agency is operating, a separate appropriation or an emergency supplemental seems best. The specifics of any funding arrangement must be agreed upon by the parties, but several general principles apply. To begin realizing savings, sufficient funding authority must be established to substantially implement the innovation or capital project. The investment capital fund should be established in a separate appropriation, which would be offset against some portion of AID's current information management funding levels and would be accounted for over several fiscal years. AID should request authority from OMB and Congress to replenish the fund from savings in operating expenses and program accounts achieved by eliminating obsolete systems and other management improvements. Payments against the innovation loan should be subtracted directly from AID's appropriation. Savings realized by the innovation should be available for reprogramming by the AID administrator.
2. AID should target sound projects from the Information Systems Plan for funding.
The agency's focus should be on: (1) agencywide data sharing through improved connectivity, (2) the financial management system, and (3) the automated program management system. These are the areas that have either been subjected to the majority of the criticisms or have the greatest likelihood of success. This would demonstrate the utility of the new technology, build faith in outside observers regarding the agency's ability to correct internal problems, and improve the management of the agency.
3. AID should conduct a thorough cost-benefit analysis of any proposed capital expenditure.
Congress and OMB must be satisfied that a proposed capital expenditure will produce a positive benefit justifying its cost. Accomplishing a fundamental change in the way that an agency does business, using unproven technology with borrowed funds, is a proposition that requires a high level of confidence from the participants.
4. AID should obtain customer input during business process reengineering.
AID has a good but tenuous start on analyzing its business processes. Key to the success of this endeavor is the need to bring those who will be using the systems into the design process. In AID's case these customers are all over the globe. This critical process cannot be cut short, and will require a commitment of resources from the innovation fund to ensure adequate participation of all customers, particularly overseas missions. The systems implemented should serve both the user at the country mission level and the managers in Washington. There must be incentives for each user to support the system. Focus should be on the three areas outlined in the second recommendation.
5. Implementation of the Army Corps of Engineers Financial Management System (CEFMS) should be expedited.
This system has been made available to AID as a solution to its project-oriented financial management systems requirements. It is viewed as an alternative to more costly private sector off-the-shelf systems. Tests are underway to determine modifications necessary to meet AID and core financial system requirements. OMB has recommended to GSA that a waiver under OMB circular A-127 be granted, to permit AID to acquire CEFMS, if AID meets several specific OMB-imposed conditions.
The recently completed information systems plan is a large step forward for AID, but more than a good plan will be required to gain the confidence of the agency's critics. The recommendations will result in quicker and more comprehensive implementation of AID's information systems. The financial decision making on these substantial capital expenditures will be made more transparent, and the return on these investments will be better understood and documented. With competent business process reengineering, work will be performed more efficiently and be better understood by both management and those at the implementation level.
A thorough reengineering of management systems has the potential for creating substantial operational savings; however, these savings cannot be estimated at this time.
The President's fiscal year 1994 international affairs budget totals about $21.9 billion, of which $6.8 billion is requested for the Agency for International Development (AID).(1) Much of AID's budget is spent on programs and projects that range in complexity from a single training workshop to developing financial systems in a foreign country's government. AID funds approximately 2,000 projects annually in more than 100 countries.
Through the years, AID has shifted its focus from the funding of discrete projects to an emphasis on comprehensive programs and increased use of what AID calls non-project assistance, which promotes broad-based economic and policy reforms in developing countries. This portion of the report will focus on what AID traditionally has called the project management system. However, it also has wide application to AID's programs and non-project activities.
Although AID headquarters in Washington, D.C., plays an active role in project management, AID's field offices or missions have primary responsibility. A mission is based in the foreign country that receives AID assistance. AID refers to these foreign countries as host countries. Generally, an AID mission funds a development project to respond to a specific problem within the host country, although funds sometimes are earmarked by Congress for a particular country, sector, or activity.
Since a major portion of the U.S. foreign assistance budget is specifically targeted to the goal of alleviating poverty in developing countries, the participation of the poor in these countries in AID projects is critical for project success. A successful project is defined as a sustainable one that not only achieves its objectives, but also is continued by the intended beneficiary after AID funding terminates. Therefore, project sustainability is largely contingent on beneficiary support.
Congress also recognized the important role that project beneficiaries play in the development process. In 1966, Congress mandated through Title IX of the Foreign Assistance Act that priority "be placed on assuring maximum participation in the task of economic development by the people of the developing countries."(2) The Foreign Assistance Act of 1973 advocated "self-help efforts . . . and the involvement of the people in the development process through democratic participation."(3) In 1975, subsequent legislation provided that "greatest emphasis . . . be placed on countries and activities that effectively involve the poor in development."(4)
Although AID understands the linkage between customer involvement and project sustainability, the current project management system can prevent AID from fulfilling its congressional mandate in this area.(5) As a result, many AID projects are developed that either do not adequately serve AID's customers or lack the necessary customer commitment. Therefore, such projects may not continue once AID funding terminates. Based on an extensive review of the literature and interviews with AID staff and other development specialists, several problems in the way that AID's project management process is structured have been identified. These problems seriously constrain the ability of AID to respond to the needs of the poor in developing countries.
Recent reports have highlighted numerous problems with AID structures and procedures that hinder the agency's ability to achieve results. The project management cycle generally involves seven phases: problem identification, project design and negotiation, obligation of funds, implementation, evaluation, project redesign, and closure. Each phase requires an array of internal reviews, streams of paperwork, and complicated procedures, resulting in a rigidly crafted blueprint design of the development project.
The detailed blueprint of the project and the requirements of the system do not allow for the necessary flexibility and may inhibit the beneficiary input required for project sustainability. A 1988 internal evaluation of 62 AID health projects indicated that more than half either failed before the project ended or were unlikely to be continued after U.S. funds ceased.(6)
This high failure rate has been attributed to AID's top-down, blueprint project design and management approach, which impedes the ability of project field staff to make the necessary mid-course corrections during project implementation.(7) It also has fostered an environment that discourages risk-taking and the use of creative problem-solving strategies that may deviate from the project design.
Although changes may be warranted to respond to the uncertainties inherent in all phases of the project cycle, the standardized, rigid structure of the project cycle specified in AID's operational project handbooks does not allow sufficient flexibility. AID has streamlined some of its project requirements to allow for greater flexibility in certain regions of the world and plans to adapt these simplified procedures agencywide in early fiscal 1994.
The number of distinct steps necessary for project approval involving field, headquarters, and Congress results in a burdensome, duplicative, and time-consuming system. Frequently, a project can take up to three years before it begins. By the time a project gets off the ground, the assumptions built into the project design and the initial project participants may have changed. Continuity is lost. The implementation team attempts to carry out a project that was sometimes planned by another team three years earlier. Adding to the problem is AID's overseas contracting and procurement system, which often impedes project results by contributing to logistical delays, higher costs, and customer dissatisfaction.(8)
The incentives at AID tend to reward work in one part of the project management cycle (design and obligation of funds) at the expense of others (e.g., start-up, implementation, evaluation, results). People are rewarded for designing new projects, not necessarily for implementing them.
The General Accounting Office (GAO) recently suggested in a draft report that the push to obligate earmarked funds drives AID to fund million-dollar projects to which the host country is indifferent, as in the case of judicial reform initiatives in Central America.(9) In the area of implementation, more focus has been placed on audit trails, processes, and resource accounting than on how or if the project is achieving results. Evaluations focus greater attention on inputs (number of activities) than outputs (project impact).(10)
The mission director in each host country has ultimate responsibility and accountability for a specific AID project in his or her jurisdiction. Ensuring that the project works on a day-to-day basis, however, involves many people working on separate pieces. The AID project officer often is assigned to a project that was designed, budgeted, and authorized by others and is rarely present throughout the entire life cycle.
One reason for this is that AID projects are generally long-term, while Foreign Service personnel have shorter rotation assignments. It should be noted that this is addressed under a separate AID recommendation which allows for longer stays in host countries. The current structure fragments accountability. Project development involves the whole range of mission staff, from legal advisors to contracting officers, to ensure compliance with AID's rules and regulations. It appears that everyone is involved, yet nobody is really in charge to ensure overall project results. Noted management specialist Michael Hammer summarized the problem:
Classical business structures that specialize work and fragment processes are self-perpetuating because they stifle innovation and creativity in an organization . . . are unresponsive to large changes. . . . [I]nflexibility, unresponsiveness, the absence of customer focus, an obsession with activity rather than result, bureaucratic paralysis, lack of innovation, high overhead--these are the legacies of one hundred years of American industrial leadership.(11)
Compounding the problem, AID relies heavily on different outside contractors to handle separate phases of the same project. A project could be designed by one contractor, implemented by another, and redesigned by still another contractor. Contractors are evidently relied upon to provide skills not available within AID and to compensate for reduced staffing levels. Many staff do not have the training required for AID's new program directions of environment and democracy, including management of judicial reform projects in Latin America.(12)
As a result, AID sometimes pays contractors at U.S. rates for work that indigenous organizations might be able to do at the local rate for a fraction of the cost. Indigenous organizations are those that have been formed by the local population within the host country. Although U.S. contractors are needed in countries where technical expertise is lacking, there is a definite advantage in using the growing reservoir of host country professionals and indigenous organizations to assist with development projects. AID's existing policies, which create incentives that encourage U.S. contractors to form partnerships with indigenous organizations, are among the steps that AID has taken to help indigenous organizations strengthen their institutional capabilities.
Many procedures and incentives, however, preclude the full involvement of intended beneficiaries such as indigenous organizations in developing countries. Even though many AID staff expressed interest in working directly with indigenous organizations, they cited a number of requirements in AID's operational project handbooks that are confusing and onerous and act as barriers to beneficiary participation. Some of the requirements include audit, financial statements, personnel systems, and an existing track record. Many indigenous organizations may not have access to such basics as typewriters and accountants that are standard for U.S. organizations but are luxuries in developing countries.
Experience has shown that projects that support and do not supplant community efforts within developing countries are more likely to achieve results. Many innovative and successful models have been structured to build upon the efforts of people within developing countries. The following four projects have been included in this report to illustrate how their processes, incentives, and structures have been designed to foster an environment that focuses the organization on results and the customer. AID continues to rethink traditional approaches and processes to further its reinvention endeavors and has incorporated some of these innovations into existing projects. It should be noted that the following four projects represent only a few of the many innovations occurring within the development field.(13)
The Grameen Bank of Bangladesh. The Grameen Bank is an example of a development model in which credit has been used as an entry point for the poor and now serves as a catalyst for change.(14) It is not directly funded by AID, although AID has recently provided $2 million to the Grameen Trust, which provides services and technical assistance to other microenterprise programs around the world. Basically, the Grameen Bank provides small loans to poor people who own less than half an acre of cultivable land or assets not exceeding the value of an acre of land. Loan repayment has been excellent. Grameen has helped its borrowers accumulate capital and generate employment opportunities.
The Grameen Bank was developed using careful experimentation to determine what did and did not work. It continues to evolve using this method. Grameen recognized early that there was no prescribed formula for development. Had the Grameen been designed under a rigid, blueprint approach, it no doubt would have failed in the project design phase. It needed the experimentation of a social learning process in the implementation phase for innovation, risk-taking, and flexibility to adjust to local conditions. It continues to learn from its mistakes and to respond to feedback drawn from all levels of the organization. People are encouraged to bring up problems and are rewarded for developing creative solutions for solving them. It has been very effective at balancing the administrative demands of the organization with the participatory demands of the local community.
Grameen's success is also attributed to its incentive structure. Borrowers organize themselves into small groups. The members support each other and apply pressure when needed to repay loans. This pressure substitutes for collateral. Incentives are organized in a way that creates an interdependency between bank workers and bank beneficiaries. Within the group, mutual accountability ensures repayment and project sustainability.
Grameen has received worldwide attention for being a development project that works very well. It has been successfully replicated in other countries when it has been adapted to fit their unique cultural and economic conditions. It is important to note that there is no Grameen blueprint. The Grameen Bank has worked in Bangladesh only after many years of experimentation.
Canada's International Development Research Centre (IDRC). The IDRC's mission revolves around the belief that the problems of developing countries cannot be solved by importing outside solutions. The IDRC promotes development by providing funds to developing country researchers who are working within their own country on solutions to their nation's problems. It also helps in the dissemination of research results so that the research findings can be widely applied. This approach offers research opportunities that build up scientific capacity and thereby reduces the scientific brain drain in developing countries. It relies on a board of directors who are all from developing countries. The Canadian government has given the IDRC complete latitude to accomplish its mission by keeping it free from micromanagement restrictions and structuring it in a way that insulates it from Canadian politics. This project offers a variety of lessons for development organizations on how people within developing countries can be supported to solve their problems.
The Philippine National Irrigation Administration (NIA). Local participatory approaches to development are often viewed as not particularly relevant to organizations such as AID, which mainly work with host country governments at the ministry level and rarely with rural villagers. NIA, however, is an example of a large government agency in the developing world that implemented programs in ways that empowered local communities to take an active role in development activities. In fact, AID has promoted water users associations throughout Asia, based on NIA experience. AID has also worked with other donors to establish an international irrigation institute based on this model. It also provided funds in 1987 to NIA to support participatory approaches to irrigation systems.(15)
The NIA sponsored small-scale communal irrigation systems throughout the Philippines that involved local farmers from project initiation through implementation and evaluation phases. The NIA's program is one innovative approach to development that transformed itself from the conventional non-participatory approach where agency personnel viewed themselves as the implementors while the local farmers were considered passive beneficiaries, to a highly participatory one where farmers eventually took over the operations of the project and continued to sustain it through local fees.
Restructured incentives--such as partial recovery of collection fees and investment costs, amortization payments, clear points of responsibility for the results, and new performance measures based on financial viability where income exceeds expenses--encouraged greater participation and helped to curtail unnecessary spending, while promoting a sense of collective responsibility and ownership of the irrigation facilities.(16)
Since much of the AID funding is channeled through government agencies, this project offers a variety of relevant lessons about field approaches, organizational design, and management processes needed to transform the agency's role in developing countries from rowing to steering.(17) Although AID has incorporated many of these approaches in Asia, where AID sponsors irrigation and natural resource management, these methods could be expanded to other sectors and countries.
Inter-American Foundation (IAF). The IAF exists to support the efforts of indigenous organizations working on behalf of the poor in Latin America. Unlike AID, the IAF has no field presence. It accomplishes its mission through grants awarded directly to indigenous organizations. The IAF provides grants for projects that are initiated, designed, and implemented by indigenous people within host countries, and not by IAF staff. In-country support teams made up of indigenous people are responsible for monitoring and evaluating project performance. Although the IAF has been criticized for becoming too bogged down with regulations perceived as hampering innovation, its strength has traditionally been attributed to its unbureaucratic internal structure and decentralized decision making processes, which have helped the organization foster risk-taking and beneficiary participation and the freedom to experiment and learn from mistakes.
The importance of timely, accurate, and relevant information moving vertically and horizontally throughout an entire organization in all phases of the project process cannot be overestimated. One way that AID obtains feedback on its projects is through the evaluation process. It appears, however, that evaluation findings are not always acted upon or taken into account in the planning and budget allocation processes. Sometimes funding continues even after evaluations conclude that projects are producing no benefits and the host country has not demonstrated a commitment to the project.
Several external barriers prevent AID from withdrawing funds from unsuccessful projects. These include:
AID's 1992 In-Country Presence Report concludes that because the programs are broad and staff is increasingly stretched thin, AID missions may be working in isolation from other donors, with the result that missions can be unaware of donor activities within the same sector or country.(18) This could result in a very skewed approach to development, with some regions inundated with too much funding, while others are totally ignored.
Efforts should focus on:
These missions should be given waivers on rules and regulations that are deemed impediments to achieving project results. The four models presented in this report provide a number of innovative approaches that could be adapted for AID projects.
3. AID should structure the reward and incentive system in project and program management to ensure that performance and accountability are linked to accomplishing project results and that innovation is encouraged.
Project success should be the primary factor in evaluation. Rewards and incentives should be designed to create project success based on interdependency between key players. AID staff and project beneficiaries should be rewarded on project success based on clear performance indicators including mutual accountability, cooperation, and results. For example, performance appraisal systems should include beneficiary satisfaction and quality of beneficiary participation as performance indicators to measure AID's staff and contractor performance at the field level.
Likewise, host countries should be required to demonstrate sufficient commitment toward a project before it begins. Future project funding should take into account host country performance on existing projects. Where appropriate, future contract and grant awards should take into account a contractor's past performance and its ability to establish collaborative linkages to indigenous organizations.
Incentives should also be included that reward innovative approaches to development at the field level. The projects listed in this report provide a number of ways that organizations can encourage innovation. These models should be analyzed and adapted for AID use in the field and headquarters where appropriate.
4. AID should establish systems for continuing critical review of all existing projects to ensure that they are achieving desired outcomes.
As part of this review, each mission should address the following:
Reviews should determine whether projects need to be terminated or provided additional funding. AID should identify and remove existing internal and external barriers that have prevented the agency from terminating unsuccessful projects. Mechanisms are needed to ensure that resources continue to flow to those activities that achieve results.
5. AID should strengthen the project and program evaluation process and integrate it into the planning, budget, and project allocation processes.
AID should develop mechanisms that incorporate evaluation findings into all stages of the decision making process. Evaluations should be used to help guide managers with planning, budget, and project allocation decisions. Evaluations should focus on results. Funding for projects that are not producing any benefits should be discontinued.
Impact evaluations should be more than lessons learned. They need to be taken seriously and their recommendations should be acted upon. It needs to be recognized that this may require strengthened analytical capabilities at AID. Increased focus on results in other assessment and accountability mechanisms, such as audits, inspections, and other reviews, would assist managers in planning, budgeting, and project allocations.(19)
6. AID should improve donor coordination, both at headquarters and in the field.
AID mission directors should designate an officer responsible for coordinating the efforts of the AID mission and other donors. Where appropriate, all AID mission staff should be encouraged to coordinate policies and leverage resources through such means as extensive consultations, joint ventures, and partnerships with other donors.
These recommendations support the reinvention efforts outlined by the AID administrator. Overall, they are intended to create a highperformance, results-driven agency. At AID, the internal structures, external barriers, and burdensome procedures that have accumulated over the years make it more difficult to achieve program results.
Reengineering the project management system under a Grameen-type social learning model of development could provide a more effective process, supporting creativity and flexibility while focusing on the customer to achieve project results. Incentives that tie performance to results will increase productivity and ensure accountability to beneficiaries.
The system and processes will be adapted to bring in the customer at each stage of the development process. By focusing on the customers (in this case the project beneficiaries), projects will be designed to adapt to local conditions. Bringing indigenous groups more directly into the project cycle will ensure that projects better reflect their needs. As a result, greater beneficiary ownership will be achieved to ensure sustainability and accountability. AID will leverage resources in collaboration with other donors and NGOs to support host country activities.
Projects deemed unsuccessful in the evaluation process should be phased out. Reallocated resources will flow to projects that are tied to new priorities and are actually achieving results. Although these recommendations target the project cycle, they can also be applied to AID programs and non-project activities that may encounter the same types of problems experienced in the project cycle.
The specific savings associated with reengineering the project and program management system cannot be determined. Reduction of overhead administrative costs, increased reliance on indigenous populations, stringent reviews, and termination of unsuccessful projects could result in substantial savings. The process of reengineering is long and arduous, and an increased focus on strategic planning and evaluation may result in up-front costs. This is not expected, however, to create new spending demands. On the contrary, the removal of existing internal and external barriers should reduce costly administrative overhead. This should offset any added costs incurred as the agency reengineers its management of projects and programs.
In his confirmation hearing on April 29th, the administrator of the Agency for International Development (AID), J. Brian Atwood, declared that he did not want reinvention efforts to be limited to a few pilot projects. Instead, he announced that the entire agency would be offered up as a reinvention laboratory. Since then, Administrator Atwood has taken a number of steps to turn rhetoric into reality. Comprehensive plans are underway to transform AID into a high performance, results-driven organization that can respond effectively and efficiently to global challenges.
There is widespread recognition that AID must change profoundly if it is to be relevant in the post-Cold War era. This agreement on the need for fundamental change encompasses AID career staff, other U.S. government agencies, Congress, and the public. It is an important asset that supports AID reinvention efforts. But it also presents a complicated challenge because there are diverse views on what kind of change is appropriate. The recently completed report by the Deputy Secretary of State has started to resolve this debate. That report will be the basis for establishing a broad consensus on the appropriate role and function of AID.
AID reinvention efforts encompass a wide number of activities. A Quality Council representing a cross-section of AID employees has been established to coordinate the reinvention plans. This Council meets weekly to review reinvention efforts and to the greatest extent possible ensure that all AID employees are given the opportunity to get involved in the reinvention process. To date, reinvention efforts include:
Vision for the Future
Within the next five years, AID will become a leaner, more focused organization with a clearly defined mission. To a significant extent, the success of AID's reinvention efforts will depend on its ability to secure cooperation and support from other key executive branch agencies (such as the State Department and the Office of Management and Budget) and Congress. In addition, a focus on results will mean that the current programming, monitoring, and auditing systems overwhelmingly geared to tracking inputs will have to be reexamined.
Change in Budget Authority by Fiscal Year (Dollars in Millions)
Recommendation
1994 1995 1996 1997 1998 1999 Total Change in FTEs ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ AID01: cbe cbe cbe cbe cbe cbe cbe cbe AID02: cbe cbe cbe cbe cbe cbe cbe cbe AID03: n/a n/a n/a n/a n/a n/a n/a n/a AID04: cbe cbe cbe cbe cbe cbe cbe cbe AID05: n/a n/a n/a n/a n/a n/a n/a n/a AID06: cbe cbe cbe cbe cbe cbe cbe cbe AID07: cbe cbe cbe cbe cbe cbe cbe cbe ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Total 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0
Change in Outlays by Fiscal Year
(Dollars in Millions)
Recommendation
1994 1995 1996 1997 1998 1999 Total ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ AID01: cbe cbe cbe cbe cbe cbe cbe
AID02: cbe cbe cbe cbe cbe cbe cbe
AID03: n/a n/a n/a n/a n/a n/a n/a
AID04: cbe cbe cbe cbe cbe cbe cbe
AID05: n/a n/a n/a n/a n/a n/a n/a
AID06: cbe cbe cbe cbe cbe cbe cbe
AID07: cbe cbe cbe cbe cbe cbe cbe ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Total 0.0 0.0 0.0 0.0 0.0 0.0 0.0
cbe = Cannot be estimated (due to data limitations or uncertainties about implementation timelines).
n/a = Not applicable (recommendation improves efficiency or redirects resources but does not directly reduce outlays).
Governmental Systems
Changing Internal Culture Abbr. ************************* **** Creating Quality Leadership and Management QUAL Streamlining Management Control SMC Transforming Organizational Structures ORG Improving Customer Service ICS
Mission-Driven, Results-Oriented Budgeting BGT
Improving Financial Management FM
Reinventing Human Resource Management HRM
Reinventing Federal Procurement PROC
Reinventing Support Services SUP
Reengineering Through Information Technology IT
Rethinking Program Design DES
Strengthening the Partnership in
Intergovernmental Service Delivery FSL
Reinventing Environmental Management ENV
Improving Regulatory Systems REG
Agencies and Departments Abbr. ************************ ***** Agency for International Development AID Department of Agriculture USDA Department of Commerce DOC Department of Defense DOD Department of Education ED Department of Energy DOE Environmental Protection Agency EPA Executive Office of the President EOP Federal Emergency Management Agency FEMA General Services Administration GSA Department of Health and Human Services HHS Department of Housing and Urban Development HUD Intelligence Community INTEL Department of the Interior DOI Department of Justice DOJ Department of Labor DOL National Aeronautics and Space Administration NASA
National Science Foundation/Office
of Science and Technology Policy NSF
Office of Personnel Management OPM
Small Business Administration SBA
Department of State/ U.S. Information Agency DOS
Department of Transportation DOT
Department of the Treasury/
Resolution Trust Corporation TRE
Department of Veterans Affairs DVA