NEX in the Context of China

As in every country, NEX in China partakes of the unique characteristics of the national bureaucratic structure and culture. China remains, despite the pace of reform, a highly centralized and hierarchical political system. Governmental and semi-governmental units regularly supplement partial budgetary support with secondary income generation through aggressive cost recovery activities, and tertiary income generation through investment in businesses (restaurants, taxi companies, etc.) with little or no relation to the unit's purpose, but as a means of supplementing inadequate budgetary allocations. China offers a particularly difficult institutional milieu for outsiders, and a difficult environment for international donors accustomed to a high degree of transparency in programme activities and accounting.

NEX has rapidly become the dominant modality of execution. Out of 128 active projects, 98 UNDP projects with total budgets of $148.3 million are executed by the Government, and practically all projects approved under the Third Country Programme (1991-1995) are nationally executed. The third country program resources at the start were: $189.9 million IPF funds; $18.5 million Government cost sharing and $4.0 third party cost sharing. However during the course of the present program the IPF was cut back to $145.6 million.

The Government's Perspective. The leaders of the China International Centre for Economic and Technical Exchanges (CICETE), the government's designated "counterpart" and "focal point" for the UNDP Programme since 1985, stress that NEX provides the most efficient and effective execution modality, especially in a huge country such as China where international personnel and institutions are at a decided disadvantage regarding linguistic ability and knowledge of the complex situation. They emphasize that the recent decisions of the General Assembly and the Governing Council provide that all UNDP development programmes are to be "nationally owned, directed and controlled." Furthermore, they stress that under these new policies, "establishment of priorities, achievement of development objectives, and sustainability of results are the responsibility of the government." It is the stated view of CICETE that the UNDP field office "continues to perform much of the programme, financial and administrative functions that the policy of NEX prescribes for national entities like CICETE."

Third Country Programme. NEX is the modality of choice for all projects and programmes of the Third Country Programme (CP3), and is therefore applied to the full range of UNDP programming. Nineteen programmes, including over 100 projects, are grouped under five areas of concentration:

• Rural economic development and production;
• Energy, transport and communications;
• Industrial productivity and profitability;
• Social development and quality of life; and
• Economic and public administration reform and management.

The Third Country Programme also stresses several important multisectoral themes such as public sector management, environment, development of poor areas, women in development; and TCDC.

Institutional Arrangements. At the core of the management of the country program is CICETE, a monopoly, self-financing parastatal agency under the auspices of the Ministry of Foreign Trade and Economic Cooperation (MOFTEC). For all nationally executed projects, CICETE serves as the executing agency, controlling the budgets and financial reporting on behalf of a "recipient" ministry for a given programme. Within that ministry's programme, each component project provides a package of resources (study tours, vehicles, equipment, and technical assistance) to a "project authority", typically a semi-autonomous research institute affiliated with the ministry.

Under the Third Country Programme, the role of CICETE has expanded rapidly as the level of nationally executed projects increased to 95%, and the UNDP field office switched from the system of providing direct payments to providing quarterly advances for all projects to CICETE. UNDP has arranged that programmes funded from sources other than IPF are handled by agencies other than CICETE, thus providing some diversification in the execution of UNDP-administered programmes. CICETE handles a very large volume of national and international personnel recruitment, commodity procurement, quarterly advances and payments, financial accounting, and substantive monitoring, for projects implemented by numerous national project authorities and international agencies, including FAO, UNIDO, ILO, IMF and others.

Projects are formulated by CICETE in collaboration with the UNDP country office, after the prioritization exercise which is carried out by CICETE within the overall directives of the State Planning Organization; provided, of course, that concordance exists between the country programme of UNDP and the priorities already set by the Government. Once the project is formulated and the "recipient" agency is designated by CICETE, the project is carried out according to work plans and CICETE maintains its supervisory role and assumes the responsibility of providing the required inputs.

The quarterly advances from UNDP to CICETE are made according to the work plans and justifications presented by CICETE. The financial relationship between "recipient" units and CICETE is direct, without the involvement of UNDP. CICETE also assumes full responsibility for the outputs of each project. In many ways, this arrangement with CICETE resembles the traditional "agency execution" modality, the major difference being that CICETE must apply for advances and account for expenses quarterly. CICETE staff express resentment for being given less autonomy than the UN agencies.

Financing of CICETE. In exchange for receiving the package of resources, the recipient ministry and/or project authority pays CICETE a "management fee" from its own resources, consisting of three percent (3%) of the costs financed by UNDP funds, and one percent (1%) of Government cost sharing funds, to be paid in full by the recipient unit at the inception of a project. As CICETE receives little direct budgetary support from the central government, it must support itself through these fees and other means of income generation. In addition, CICETE has also recently requested that recipient units provide additional funds so that it can meet its financial obligations. Within the present setup, it appears to be normal that CICETE also keeps a portion of the amounts allocated for fellowship stipends and DSA allowances, and requests partial repayment in local currency for equipment purchased with hard currency under project equipment budgets.

Impact on Projects. Institutional relations in China are different from most UNDP country programmes. CICETE views itself as the "counterpart" to UNDP, acting on UNDP's behalf within China. It also views itself as the representative of the Chinese government regarding matters pertaining to UNDP programmes. Finally, as an "executing agency" it views itself in the role of a UN specialized agency, responsible for directing the implementation of projects through various national ministries and research institutes.

In China it appears difficult to make the distinction between "executing" agency and "implementing" agency since the same Chinese character represents both English terms. CICETE often acts partially in both modes. CICETE's practical role varies according to the strength, capacity, and influence of the government implementing agency. With "weak" ministries, CICETE effectively runs all aspects of the project, making key management, strategic, personnel, and financial decisions. With "strong" ministries, CICETE cedes overall technical management to the ministry, keeping at least some of the procurement and consultant contracting and payment for itself. Where there is a UN specialized agency implementing the project, international procurement and recruitment may be divided between the agency and CICETE.

Impact on Organizational Relations. Although the participation of UN specialized agencies is very greatly reduced with the rapid move to a complete NEX mode, they still play important roles in project and programme formulation, implementation and monitoring. UN specialized agencies serve as partial implementing agencies in 12 out of the 98 nationally executed projects, implementing $20 million, or about 15% of the total Third Country Programme IPF. The TSS-2 committed budgets amount to $0.9 million and AOS funds amount to $1.1 million.

In most countries the introduction of NEX involves an increased involvement of UNDP country office staff in the national programme. In China, as CICETE takes up more responsibility and demands more autonomy, the UNDP programme officers find themselves increasingly out of the loop and frustrated that they are unable to perform effectively the oversight functions required of them.

Financial Accountability. The preparation of financial accounts is the responsibility of CICETE. The country office reviews and approves the quarterly requests for advances, and the quarterly rendering of the expenditures. The rate of compliance and the flow of information from CICETE to the country office appear to be good, at least in a formalistic sense, since without proper quarterly financial information, advances cannot be made. Accounts are rendered fully and on time, although it is not easy for UNDP programme officers to match up expenditures with specific project activities. Project accounts are audited annually by the government auditing authorities.

Substantive Accountability. The responsibility for substantive accountability, in the first instance, rests with CICETE as the executing agency. The ultimate responsibility, however, remains with UNDP. CICETE itself lacks the required technical expertise to carry out the substantive monitoring and evaluation functions. For substantive monitoring it relies on the expertise of its recipient agencies and, in some instances, on participating international agencies. The UNDP country office takes the lead in preparing TPRs and evaluations in which CICETE and "recipient agencies" participate. In the context of China several factors should be considered in evaluating substantive accountability.

• Firstly, the sheer number of required project evaluations is extremely large.
• Secondly, neither UNDP nor CICETE have the resources to carry out all the mandatory evaluations.
• Finally, there is a prevailing feeling - especially on the Government side - that evaluations are costly and often unnecessary. CICETE's view is that if the project is going well, according to all accounts, there is no need to mount an investigation.

The end result is that the ratio between the number of actual evaluations and TPR meetings to the number planned is about 1:3. Both the country office and CICETE rely more on informal interactions between UNDP officers and designated officers of CICETE and the interactions between the latter and the project personnel. As part of its extensive mid-term review process, UNDP carried out six major sector evaluations.

Financing. As explained above, the major financing source during CP3 continues to be the IPF, although it now stands at a level which is 23% less than originally planned. On the other hand, actual cost sharing is now 10% higher than the planned amount, reaching $26.4 million. The bulk of the cost sharing comes from Government sources. Third party contributions continue to be negligible.

Programme and Project Performance under NEX. It is generally agreed that the performance of the programmes and projects in China is quite good. The recipient project authorities aggressively make very good use of the resources provided to them through the UNDP-supported programme, and CICETE provides effective overall coordination for this huge programme. Furthermore, although the Chinese are insistent upon their prerogatives in NEX, they are, for recent historical reasons, anxious to maximize contacts with the world at large, and technology transfer from abroad. One does not find in China the insistence found elsewhere that outside technical assistance is not needed or welcome. Certainly, a good number of projects are quite effective in carrying out their specified activities and moving in a timely manner toward achieving the project objectives, despite strong differences in the manner of implementation:

• In the Forestry Programme, strong leadership is provided by the Ministry of Forestry, with support from CICETE and FAO. One of its projects is rapidly putting in place a GIS and remote sensing-based system for mapping and monitoring the forest inventory on a national scale.
• For the troubled Machine Tool Programme, CICETE and UNIDO have divided the components between them, with the project authorities taking a very subsidiary role. This programme has faced a major crisis, as one of its international subcontractors went bankrupt without delivering all of the services for which it had been paid. This component has been frozen by UNDP while CICETE, UNDP, and UNIDO decide how best to utilize the remaining project resources.
• The Tax Reform Project at the Ministry of Finance was designed by the IMF which serves as a cooperating and cost sharing agency. IMF experts are working closely with their Chinese counterparts to redesign the taxation system, including new corporate taxes and a value added tax.

Cost Effectiveness. The Chinese responsible for project execution and implementation are very cost conscious, partly from a strong sense of ownership and partly from the participating institutions' need to partially finance their own operations. Cost effectiveness depends on the success of the projects. To the extent that the programmes and projects accomplish their objectives while keeping costs down, there appears to be a strong element of cost effectiveness. However, some of the cost cutting, while in the interest of the participating institutions is not in the interest of the UNDP or of the participating individuals. For instance, trainees sent abroad with inadequate funds reflect negatively upon UNDP and also inhibit the abilities of some trainees to get the most out of their experience. Delaying procurements so that CICETE can group them into larger orders, can delay the progress of a project activity and cost many times more than the savings on the item. Due to the mechanism for financing CICETE's operating costs, there is perhaps more squeezing of some expenditures than is good for project performance. However CICETE is financed, it is clear that a very large programme is being effectively implemented in China at a relatively low management cost.

Capacity Building. CICETE has developed an impressive capacity over the years for handling a large volume of UNDP projects, including procurement, contracting, recruitment, and financial reporting. As the number of nationally executed projects has increased rapidly, at a time when the expanding economy also provides more attractive opportunities for CICETE staff, CICETE is under a resource strain. This strain is exacerbated by the shift in some of the UNDP programming from technology transfer to poverty alleviation, impacting negatively on CICETE's ability to collect management fees from its recipient agencies. Although CICETE's capacity has grown over the years, there is considerable room for improvement if it is to continue to handle the huge volume resulting from 100% NEX of IPF-funded projects. The danger at present is that CICETE's capability to meet growing responsibilities, is actually deteriorating as it struggles to retain staff in the face of growing private sector opportunities.

Promotion of Self-Reliance. The prevailing opinion expressed by CICETE and project personnel is that NEX projects in China enhance self-reliance. However for China, self-reliance is less important than the sense of ownership, since four and a half decades of isolation and revolution, have made self-reliance a national virtue.

Sustainability. While the sustainability quotient appears quite high in China, as anywhere else, it would be simplistic to assume that there is an automatic and direct functional relationship between NEX and sustainability. The UNDP-assisted activities are closely derived from the government's programmes, they are integrated into the normal performance of duties of the staff of the recipient units, they are viewed as their own projects (ownership), and they are for the most part effectively managed and implemented. Whether the results of NEX projects will be sustained or not will depend upon how useful the government finds such results, what budgetary provisions it is able to make, and the degree of political will that exists in counterpart agencies. Moreover sustainability is also a function of how well the project is designed, how efficiently it is carried out, how many quality outputs are at hand, and whether the project is executed by a national entity or by a UN specialized agency.

CICETE has reached a point of financial crisis, whereby its sources of income are inadequate to support its operating costs. Its primary source of income, management fees from recipient agencies, has decreased for three reasons.

• The shift in the UNDP portfolio to poverty alleviation means that CICETE's new recipient units have little or no resources and no ability to pay CICETE a management fee.
• CICETE collects its management fee from recipient units upfront, at the time the project is signed. With the reduction of the IPF, few new projects are being signed.
• Some of CICETE's tertiary income generating schemes have done poorly, draining rather than supplementing resources.

At present, UNDP and CICETE are working together to reorganize the institutional arrangements for NEX. Under this proposal, the government ministry, MOFTEC, will take up the overall responsibility of programme execution. CICETE would be renamed as the China Aid Support Organization (CASO) and reorganized to focus its considerable institutional talents on providing the necessary project support services to UNDP, on the one hand, and to the government agencies, on the other. It is hoped that this new arrangement will remove the somewhat conflicting multiple roles of CICETE as both government decision-making institution and a project service organization that can benefit financially from such decisions. It is expected that in the future CASO will seek contracts with other aid donors. Similarly, it is expected that some future projects will be executed by ministries other than MOFTEC, and that some projects executed by MOFTEC will be implemented by institutions other than CASO.

Additionally, the proposed new arrangements provide for more transparent mechanisms for CASO's financing. MOFTEC will provide a management fee to CASO, paid from the IPF; and CASO will continue to collect a fee from recipient agencies when appropriate. CASO will refrain from operating tertiary income generation enterprises unrelated to its responsibilities. Also, UNDP will provide capacity building support to CICETE in order to improve its management systems, especially in the areas of procurement, subcontracting, and personnel recruitment.

A glimpse of the future. Within the context of this overall evaluation of the worldwide implementation of NEX, the Chinese case might appear as so unique that it provides little illumination of other countries' experiences. In some respects, however, China offers a picture of NEX that is very relevant. With its years of broad NEX experience, China provides a glimpse of the future of NEX in many countries, once they complete the transition from agency execution to full national execution. The Chinese window on the future portends two developments for NEX in other countries:

• The much discussed phenomenon of "country office execution" is a transitional issue. The dependency of government agencies on the constant advice and assistance of UNDP programme officers and financial staff will give way to a more assertive autonomy that will challenge the oversight responsibilities of country office staff. Several years of increased involvement in project implementation may be followed by an increasing marginalization of UNDP programme officers once their proteges learn what they know.

• National programme execution must be adequately financed one way or another, and the longer the issue is ignored and the greater the volume of nationally executed projects, the worse it will get. NEX may be less expensive than agency execution, but it is not free. Many country programmes, newly embarking on NEX, are utilizing temporary measures of financing government NEX support units of one sort or another, hoping that with time these expenses will be picked up more or less by government. The China case suggests that the UNDP country offices and the governments should agree early on realistic, and adequate means of permanently financing programme support and coordination activities.

Although Ethiopia had significant experience with NEX (and Government Execution) for many years, most of the government personnel from those times have been replaced by the victors of the civil war. NEX is thus being introduced into a relatively new government apparatus, one that remains, in many areas, understaffed and in the throes of reorganization and decentralization. Due to the political situation, the Country Programme in Ethiopia is more than a year behind schedule, at the time of writing.

The IPF for Ethiopia at the time of the approval of the current Country Programme was $146 million, the highest in Africa. This includes almost $24 million carried forward from the fourth cycle IPF. The total revised IPF is $107 million. The percentage of the annual IPF budget allocated under the NEX modality has increased steadily since 1992, from 19% to 91% in 1995 and a projected 100% for 1996.

The Government's Perspective. Government representatives in Ethiopia regard the idea that donor-funded projects should be executed by themselves as perfectly natural and correct. They express great confidence in their abilities and in their prerogatives to manage their own affairs, and this confidence has been increased through full participation in, and leadership of, the process of programme formulation. The sense of ownership of the Country Programme, and the commitment to the principle of NEX, is so strong that returning toward the previous system of agency execution would be impossible in Ethiopia.

It is difficult to consider the application of NEX in Ethiopia without also considering the simultaneous introduction of the Programme Approach. Both government and UNDP programme personnel stress that the real challenge has been the introduction of the Programme Approach. The idea of NEX follows logically from the Programme Approach, especially to the extent that the programmes are truly those of the Ethiopian government. As government officials and UNDP programme personnel see it: "If it is the government's programme, who else should be the executing agency?"

Fifth Cycle Country Programme. The Country Programme is comprised of six large nationally executed programmes that have been slowly and painstakingly developed and adopted by the government through the UNDP-funded programme formulation process. The UNDP-supported part of each programme is a large (between $10 and $25 million each), fully integrated, multi-disciplinary, decentralized, operation, involving over half a dozen implementing agencies. These programmes are more complex and more unified than is usually the case. The fully integrated Ethiopian programmes stand in contrast to the collection of semi-discrete, closely related projects or components that comprise programmes in many countries. The six Ethiopian programmes are in the areas of: Agriculture, Natural Resources, Human Resources Development and Utilization, Disaster Prevention and Mitigation, Management of Economic and Technical Change, and Economic Recovery and Reconstruction.

An objective of all the programmes is capacity building, and especially capacity building in support of the decentralization of decision-making and service provision. The formulation of the programmes is the work of multi-disciplinary Programme Formulation Teams (PFTs), each led by a Programme Formulation Coordinator (PFC). These teams include national experts with the appropriate ministries, national and international consultants, and the UNDP programme officers.

A lead ministry is designated as the focal institution and executing agency, with other institutions, national and international, governmental and non-governmental, serving as implementing agencies for specific outputs and activities. Each programme is led by a National Programme Coordinator who works with a National Programme Management Secretariat (PMS) comprising the programme coordinators from each of the participating implementing agencies. It is planned that each of these National Programme Secretariats will hire an administrative staff including a logistics/procurement officer, a finance officer, a national advisor, and a secretary. In the case where a single ministry is the executing agency for more than one programme, a single enlarged NPMS serves both interrelated programmes.

Impact on Programmes. One important impact of NEX (and the new programme approach) has been the lengthy process of programme formulation and the consequent postponement of implementation. Programme Officers and National Programme Coordinators express confidence that the implementation will go well. They stress that through full participation in preparing the Programme Support Document (PSD) and especially the Programme Support Implementation Arrangement Document (PSIA) all members of the PMS and their respective agencies are fully on board and proud of the programme they have designed. Implementation should follow naturally from the detailed PSIA they have prepared, and the national programme and subprogramme coordinators know the details by heart. However, past the half way point in the ongoing Country Programme, only three of the six programmes had been fully signed and were ready to begin implementation.

Two factors threaten to further delay implementation:

• One is the slowness of government to approve the hiring of the administrative staff for the PMSs. UNDP has rightly informed the government that they are not prepared to advance funds to programmes that do not have a fully staffed PMS in place.
• The other is the current reshuffling of government personnel. Should a significant number of national programme and subprogramme coordinators be reassigned or terminated, much of the momentum and experience gained from the long formulation exercise will be lost.

Only one of the programmes, the first phase of the Management of Economic and Technical Change (METC) programme has been operational for any significant time. METC has just completed its first year with considerable success, despite the fact that none of the administrative staff for its NPMS are yet in place. A recent review reports that the programme completed over 70% of its planned activities for the year. Accomplishments include the establishment of the underpinnings for a new national accounts system to provide improved modelling of short-term and medium-term economic performance, the development of a national programme in support of private sector development, and capacity building for economic management at the regional level.

Another project that has been functioning is entitled "Upgrading National Capacity for Technical Cooperation, Programme Planning, Co-ordination, and Management during the Fifth Country Programme." This $1.1 million project aims to strengthen the Ministry of External Economic Cooperation's (MEEC) capacity to respond to and coordinate the NEX initiative. Progress is reported to be quite good: an excellent NEX manual has been prepared, training, both overseas and national, has been delivered, and the understanding of NEX has expanded within the government. The project has effectively involved government personnel, international consultants, and national experts. However, the administrative and financial details of these programmes/

components have been handled directly by the UNDP country office.

Impact on Organizational Relations. As expected, the role of the specialized agencies is reduced under NEX in Ethiopia. Under the previous Country Programme, 30% of which was Government Executed, it is estimated by government that as much as 40% of project budgets went for international experts, provided largely through UN specialized agencies. By contrast, the utilization of international experts provided by specialized agencies under the present Country Programme is quite selective, with less than 15% of the programme implementation budgets allocated for international experts.

The specialized agencies played a significant, but supporting role in the programme formulation process. Some have complained that they did not receive "their share" of the programme implementation activity under the NEX modality. The country office stresses that the responsibility is with the specialized agencies to convince the government of the value of their services. Government spokespersons feel that the programme formulation process has judiciously utilized agency-provided international expertise whenever such expertise was deemed unavailable locally. They are prepared to use more international expertise whenever and wherever they discover that they have overestimated the national capacity. However, it is reported that there is a tendency in this government against foreign expertise in general and specialized agencies in particular.

Financial Accountability. To date, the country office handles all financial management and accountability. Regarding operations such as procurement; recruiting; contracting and payment of consultants; financial payments; and accounts; the Country Programme in Ethiopia is "Country Office Executed". "We are the finance department and personnel department for the government," is a common view in the country office. This responsibility will overwhelm the country office once full implementation of this huge programme begins in earnest. As implementation gets underway and the National Programme Management Secretariats are staffed with financial and procurement officers, these responsibilities must be taken up by government as fully and quickly as possible. To the extent that the country office continues to provide such services, it will charge a management service fee of 3%. One Government department, upon receiving UNDP's letter regarding this service fee, responded positively that the government agrees to pay UNDP the service fee for international procurement provided that the items are delivered on time.

Substantive Accountability. The core of the programme monitoring mechanism is the Programme Review Committee (PRC). The PRC is chaired by the Ministry of External Economic Cooperation (MEEC) and is composed of senior technical officers of the implementing agencies, the national programme coordinator, regional programme coordinators, representatives of participating specialized agencies, and donors. It is planned that the PRC will meet twice a year. The National Programme Coordinator will prepare a progress report for each review meeting. UNDP participates in what is essentially a government process. Experience to date with the METC Phase I Programme indicates that the process is effective. In addition to the semi-annual PRCs, there is provision for evaluations in which UNDP, government, and appropriate specialized agencies will participate.

Financing. The revised IPF of approximately $107 million had been allocated, as of December 31, 1994 to six programme areas. Of the six planned programme areas, three were fully approved with signed PSIAs and detailed output budgets. Two programmes had completed Project Support Documents, and were operating on a preliminary phase basis while preparation of the PSIA was in hand. One programme was in the early stages of formulation. Given the delays in formulation, the Country Programme has more, much more, resources than it is likely to deliver.

Programme Performance under NEX. It is too early to judge the implementation performance of the programmes, with the exception of the METC programme and the Component to Upgrade the National Capacity for Technical Cooperation, mentioned above due to the fact that implementation has barely begun. The performance of the Programme Formulation Teams, however, has been very strong. The Programme Support Documents set out the objectives, strategies, activities, outputs and inputs required for the programme. The PSIAs, now signed by UNDP and Government, provide a clear, broadly agreed-upon, fully understood, implementation programme. Workplans, budgets disaggregated by subprogrammes, components, and implementing agencies provide a clear road map to all participants. The members of the Programme Formulation Teams are rightly proud of their accomplishment to date.

Cost Effectiveness. There are evident cost savings in the greater utilization of national talent and apparent efficiencies in the programme approach, especially when the programmes are large, as in Ethiopia. Cost effectiveness will depend on the extent to which the programmes fulfill their objectives by the end of the Fifth Cycle. The long period of formulation combined with what is already shaping up to be a slow start-up for implementation (with the delays in approval of administrative personnel), may make it difficult for the programmes to achieve all of their objectives. The long formulation period is, in Ethiopia as elsewhere, the price that is paid for ownership. It is premature to judge whether ownership translates into effective, and ultimately cost effective, implementation. If the long process of formulation and implementation translates into having a significant portion of Fifth Cycle resources not utilized, then the potential cost effectiveness of NEX will be lost. However, the delays are attributable as much to national political conditions and the complexity of the Ethiopian utilization of the programme approach as to NEX itself.

To the extent that one important objective of the UNDP support is capacity building, the long formulation process contributes to this as effectively as the actual implementation. However, there is the danger that NEX in Ethiopia may become more of the end product of the country programme than the means to achieving, not only capacity building, but poverty alleviation, improved opportunities for women, environmental protection, and so on. Given the size of the UNDP-supported programmes, the proposed management structures are lean by any standard. Should they be staffed up quickly and prove reasonably effective, they will be cost effective. There is, however, evidence that these resources may be provided on a "too little, too late" basis, a problem compounded by the staff shortages and reshuffling in the government. The most expensive component of technical assistance projects, international expertise, has been greatly reduced between the Fourth and Fifth Cycles. There appears to be, at least in theory, an adequate planned utilization of outside expertise, carefully selected and tailored to support programme objectives.

Individual. UNDP Programme Officers and National Programme Coordinators alike report a significant increase in the abilities of all participants in the programme formulation process to conceptualize, negotiate, operationalize, and budget a large and complex development programme. Interviewing National Programme Coordinators, one cannot help but be impressed by their grasp of both the broad concepts and the small details of their programmes. To the extent that these individuals have the opportunity to participate in the implementation of the programmes they helped design, it is likely that their professional capabilities will be greatly enhanced.

Central Government. The government clearly lacks the capacity to carry out these programmes without significant training and other supporting inputs from UNDP. The country office has contributed much to strengthening the governmental capacity by preparing (with government input) a thorough and readable Guidelines for National Execution of UNDP-Assisted Programmes in Ethiopia, and related manuals. The combination of this manual, training on NEX procedures, the detailed PSIAs, and the staffing of the PMSs should significantly enhance the capacity of government units to carry out these UNDP-supported programmes and similar development activities. Given the likelihood of continuing personnel changes, UNDP is prepared to provide training in all aspects of NEX repeatedly, and build that training capacity into government. The larger problem in building government capacity is that many of the most capable civil servants are leaving government for the NGO and private sectors.

Regional and Zonal Institutions. A major thrust of the current government's agenda and of the programmes assisted by UNDP is the "regionalization" of authority and service delivery. If lack of capacity is a problem in the central ministries, it is a much larger problem out in the regions and zones. The management structures of the UNDP-assisted programmes include the appropriate representatives of the regions and zones, and in a few instances regional personnel were involved (much to their surprise and appreciation, it is reported) in the programme formulation process. In addition to this participation in programme formulation and implementation management, many of the activities of the programmes are directed at building regional and zonal capacity through training, workshops, and so on. The capacity building and involvement of regional, zonal, and community organizations will likely prove to be the largest challenge facing substantive implementation of the Country Programme.

The fear is expressed by some that the overlaying of decentralized management and operations on top of the complex multi-disciplinary, multi-institution programmes may prove to be more than the structure can bear. Again, people stress that although they are confident regarding NEX, the combination of the complexity of the UNDP-assisted programmes with the uncertainty of the political programme of decentralization poses an enormous challenge to the Fifth Country Programme in Ethiopia.

Sustainability. The Country Programme in Ethiopia is certainly designed to maximize sustainability. The programmes are thoroughly embedded in government with a minimum of UNDP-financed management and administrative resources, given their size. The government's sense of ownership appears strong. As stated above, the Programme Management Secretariats' staffing is lean. Although one does not expect these extra positions to be absorbed into government, many of the functions performed could reasonably be absorbed over time. The salaries anticipated for these contracted administrative staff are at local market rates. Government may discover that such contracting of selected functions is cost effective.

There is the classic problem of motivating underpaid civil servants. Some of the active government personnel will probably receive incentives of one kind or another through participation in the UNDP-assisted programmes. A modest proposal from the Ministry of External Economic Cooperation (MEEC) is presently under discussion at high governmental levels. And a high-level UNDP mission is presently studying the broader options for reform of civil service salaries and conditions of service.

The outputs of the programmes are designed to increase institutional capacity in a sustainable manner. New national accounts databases, trained regional authorities, a framework for supporting private sector development and entrepreneurship training, and others have the makings of providing a continued flow of benefits to beneficiaries after programme funding ends.

Functions. The operations side of the country office is, to date, carrying out the functions relating to financial, procurement, recruitment, and logistical aspects of the Country Programme. Although it hopes to devolve most of these activities onto the programme management secretariats once they are staffed and trained, it anticipates that it will continue to carry out a number of functions for the government, on a management contract basis. As stated above, the almost total country office involvement in recruiting, procurement, payment, and accounting must cease with the beginning of broad implementation, or the system will collapse of its own weight. The programme side of the country office stresses that it is the government that has taken the lead in the programme formulation process. The most important level of execution responsibilities is with government, albeit with considerable deft guidance, leading from behind, and diplomacy on the part of the programme officers. More than in the past, the programme officers find themselves in the challenging but appropriate and rewarding role of facilitators.

Staffing. To the extent that Operations recovers 3% for its services it may be able to use the funds to hire the necessary consultants and support staff. On the Programme side, this role of facilitators requires experienced and talented programme officers. The programme officers in Ethiopia are perhaps more experienced than those found in many African UNDP offices. Yet they admit that the formulation process was a learning process that taxed and stretched their abilities. Under NEX, the role of the programme officer becomes more of a high level facilitator and less that of a traffic cop for paper trails with specialized agencies. The implication is that NEX requires not additional staff, but certainly more experienced staff, than agency execution.

The implementation of the NEX modality combined with the Programme Approach is a huge and challenging endeavor, especially in view of the political situation in Ethiopia -a relatively new post-war regime that is embarking on a strong decentralization programme after centuries of rule from the center. Notwithstanding the difficulties, the country office programme staff and the government officials involved in the Fifth Country Programme are all extremely enthusiastic regarding NEX, and proud to play a role in its introduction. The process of programme formulation, and the resultant Programme Support Documents and Programme Support Implementation Arrangement documents are of unusually high quality with especially strong ownership on the part of the participating government personnel.

Country Programme delivery stands at an important crossroads. If the government does not quickly and decisively staff up the Programme Management Secretariats, reach agreements on DSAs and related issues, and approve the Guidelines for National Execution, then implementation will not begin and there will, for all intents and purposes, be little implemented under the Fifth Country Programme. UNDP programme officers are hopeful that these impediments to implementation will be solved. Others are very doubtful. The capacity of government to manage such huge programmes remains fragile. The reluctance of government to accept that National Programme Coordinators should be full time, and the reluctance to authorize the approved and necessary support staffing of the PMSs, indicate that many in government do not acknowledge what is needed. There remains a large and on-going job to be done in bringing higher level decision-makers, on one hand, and regional authorities, on the other, to the point of understanding and appreciating NEX.

The UNDP Fifth Cycle Programme began shortly after the new Government took control in 1990. The new Government needed extensive

support from IFIs in order to put its programme into implementation, as well as UNDP's support and dialogue as a "neutral agency." Peru's relations with IFIs were quite strained in the past. These now had to be normalized as speedily as possible. UNDP's resources and its logistical capacity to mobilize bi-lateral as well as multi-lateral funding sources for the Government were of primary importance.

To use UNDP also as an instrument to attract high level expatriate nationals or local private sector experts into the public sector, was crucial especially since the established bureaucracy was not quite attuned to the requirements of stabilization and structural adjustment policies which the Government had to implement as fast as possible. Thus, during the Fifth Country Programme UNDP had to assume the role of a privileged partner of the Government to facilitate the consolidation of Peru's financial economic and political rehabilitation. The Government itself was fully conscious that the state had to take charge of the recovery process with a modern, efficient, and scaled-down civil service apparatus.

The NEX modality could be used not only to formulate and implement emergency programs, but also to mobilize financial resources and deploy human resources which were abundant in the private sector but extremely scarce in the public sector. The modality was not a novel approach in Peru. NEX arrangements were used during the Fourth Cycle, hence a certain degree of know-how had already been accumulated. But the modality had partly become a budget support mechanism. The Fifth Cycle aimed to avoid this distortion and to contribute to the development objectives of the Government.

The programme is concentrated in the following areas: a) state reform (54.6%), b) employment creation and social development (16.2%), c) productive sectors (18.4% and d) environment (8.5%). The common denominator in all projects is institution building and support to administrative structure.

NEX in Peru has come to be virtually the only modality of execution. As of December 31, 1994, only 7 out 126 active projects were not being

executed/implemented nationally, and 96% of the total funds were allocated to NEX projects. During the Fourth Cycle only 57% were allocated to NEX projects. Total planned IPF for the Fourth and Fifth Cycles has not changed much (a little over $13 million). However, while during the Fourth Cycle cost sharing funds were $22.5 million, during the Fifth Cycle these funds will amount to $133.9 million.

The Government's Perspective. The Government of Peru chose to adopt NEX not so much for the "ownership" of the projects or for their cost-effectiveness but rather as a means to a remedy flawed bureaucracy. The Fifth Cycle coincided with the present Government's drastic measures aimed at breaking its isolation from the world financial markets; attracting foreign private investment which, due to reasons explained above, were reduced to next to nil; and returning to a true market economy. Reforms for economic stabilization and structural adjustment had to be realized within a democratic framework while complying with the requirements of IFIs. In addition, Peru needed an organization that could act as a catalyst in securing bi-lateral assistance. Hence the fortunate confluence of the Government's economic-political needs, and one of the attributes of NEX, namely responding speedily to the needs and incorporating national capacity into the public sector-at the expense of creating a parallel bureaucracy-led the Government to embrace NEX. Moreover, the limited but positive experience with NEX during the Fourth Cycle also had a demonstration effect.

The Government viewed, and continues to view, NEX as operationally useful, as long as its implementing agencies do not assume the responsibility of financial management. Capacity building in that area does not seem to be of importance. The speed with which the projects are implemented, additional human resources are incorporated into the existing cadre, and, to a lesser extent, technical capacity building, appear to be the attractive aspects of NEX.

Application of NEX. As was briefly stated above, IPF and cost sharing funds-bilateral and national-are used in institution building and respond to the requirements of economic reform. Most NEX projects are implemented directly by Government agencies. Beyond that there is a wide variety of implementation arrangements within NEX. Subcontracting national/international consulting firms and consultants appear to be the most prevalent method. Financial administration rests with the UNDP Country Office. UN specialized agencies are not involved as implementing agencies, nor are they sub-contracted by the national implementing entities. The UN agencies are also excluded from getting involved in project formulation/preparation, since the prevailing impression is that they are extremely slow in responding to the needs which are very fluid and changeable.

One important aspect of NEX is resource mobilization. While the IPF resources have been shrinking, cost sharing has been increasing. The originally estimated amount is now tripled. Recognizing the operational utility of NEX, the Government of Peru has cost-shared the projects in some cases. Neither IPF resources nor cost sharing by multilateral and bi-lateral entities enter into the national budget nor are they subject to the established budgetary procedure and control of the Government.

Impact on Projects. NEX has shifted the substantive control of the projects towards the Government. To a certain extent the articulation of development objectives through this modality has also shifted towards the Government. It would be erroneous to conclude, however, that NEX was instrumental in this process in a direct manner. Certainly, the UNDP Country Office and the Government have entered into a fruitful dialogue since the beginning of the Fifth Cycle. The Government appears to be quite clear with regard to the operational advantages of NEX: first of all, NEX proved to be a speedy mechanism. Secondly, through NEX, cost sharing through third parties became much more viable. In other words, given the existing circumstances, NEX was an appropriate and more practical modality.

It is difficult to ascertain to what extent the NEX modality ensures the sustainability of the project results. The budgetary provisions, once the project is terminated, will be the determining factor. Compared to a number of other countries Peru has already built capacity. Whether the termination of a NEX project will lead to retaining this capacity within the public sector will depend on a series of conditions which are exogenous to the modality.

Impact on Organizational Relations. The involvement of the UN specialized agencies in the NEX modality is virtually nil. The first and primary reason is the nature of the majority of the projects. Building institutions and, as needed, creating a parallel civil service cadre, obviously cannot be carried out by the specialized agencies, but will have to be done with national consultants, complemented and supplemented from time to time with international consultants. The only feasible involvement of specialized agencies in NEX projects in Peru may be at the formulation and design level. However, the agencies are perceived to be slow and bureaucratic and not able to respond to the needs of the Government and/or the UNDP Country Office. Moreover, the process of obtaining funds for technical support services (TSS-1 and TSS-2) are viewed by the agencies as cumbersome and not worth pursuing.

The Government's relationship with the UNDP Country Office is rather close. NEX project management, from the financial standpoint, is still basically the responsibility of the Country Office. For reasons that seem to be endemic in Latin America, implementing agencies shy away from assuming financial management. The reasons can be found in cumbersome budgetary procedures in the public sector, the slowness with which the public sector can disburse funds, etc. Hence, the UNDP Country Office has had no choice but to take over the financial management for the sake of speedy implementation of the projects. In some cases the Country Office also provides additional support, such as contracting consultants, assisting implementing agencies by supervising their contract bids, etc., which can be interpreted as the Country Office undertaking part of the implementation.

Financial Accountability. The preparation of financial accounts is the responsibility of the project administration. The role of the UNDP appears to be at two separate but related levels. The accounts prepared by the project are first checked, controlled and revised, if need be, by the UNDP Programme Officers. The Finance Unit, in turn, makes the payments and keeps the books in order. Therefore, the Country Office carries out a cashier function and has the ultimate responsibility for financial accountability, although the project designs foresee the administration as having the major responsibility in this area.

The administration of the projects becomes essentially obliged to delegate the financial administration to the UNDP. The reason lies not in the unavailability of technical capacity in the counterpart, but rather in the bureaucratic ways of the Government that do not permit project administrations to carry out the project requirements with the necessary speed. The rate of compliance and the flow of information from the projects to UNDP is not as speedy as it should be, but no serious non-compliance has been detected. Project accounts are audited at the UNDP by an international firm selected jointly by the Government and UNDP. In some isolated instances

there are some delays in the flow of information between the Country Office and Headquarters.

Substantive Accountability. Substantive accountability is an issue which is yet to be solved. The country's single executing agency, is far from having the required human resources to carry out the monitoring functions that an executing agency should. Substantive accountability falls, therefore, on the UNDP Country Office. The Office prepares the TPRs and mid-term evaluations as the Programme Officers deem appropriate, which may or may not coincide with the requirements of the project document. The primary argument is that the Programme Officers are monitoring project implementation on an ongoing basis and that the formal evaluation requirements are not necessarily the healthiest way of maintaining substantive accountability. TPRs and mid-term evaluations have, in fact, been carried out in a number of cases. The Programme Officers argue almost unanimously that NEX per se does not necessarily increase or decrease substantive accountability.

Financing. As was noted in the preceding paragraphs, IPF funds have declined during the Fifth Cycle and cost sharing funds have increased

substantially. The World Bank, IDB, bi-lateral donors, and the Government of Peru have been involved in cost sharing agreements with the UNDP Country Office for a number of projects. In fact, 68% of total cost sharing originates from Government's funds. The UNDP Country Office appears to be an effective catalyst in articulating and matching available funds to the needs of the Government.

Cost Effectiveness. The prevailing opinion in Peru is that NEX is cost effective simply because the UNDP management fee is 3.5 % (only on cost-shared funds and not on IPF funds) as opposed to the usual 13 % charged by UN specialized agencies. However, it must be kept in mind that a number of hidden costs, such as the increased workload of the Programme Officers, the time and effort spent on training the administrative personnel of the projects, which cannot be easily quantified, are not included in the 3.5% of the administrative costs. The experience of Peru does not permit one to state categorically that NEX is cost effective. The UNDP Country Office appears to cover all its administrative costs out of this charge of 3.5 %. Nonetheless, the cost sharing volume being so high, there may be economies of scale which may or may not be achieved in other countries.

Capacity Building, Nationally and of Government. Since the projects include human resource development components, one can deduce a priori that there is capacity building. However, the particular circumstances of Peru do not permit one to come to that conclusion. The present Government has been pursuing a policy of downsizing the public sector. As a result many project personnel, to a lesser or greater extent, are assuming responsibilities and functions of public sector employees. The capacity building resulting from NEX must be questioned, since clearly designated counterparts do not exist in many of the projects.

Promotion of Self-Reliance. The nationals' psychological reaction to NEX projects is that it does enhance self-reliance. In the final analysis, self-reliance is a matter of perception given the present socio-economic circumstances of the country. In this sense NEX projects in Peru do promote self-reliance.

Sustainability. The prevailing opinion, both at the UNDP Country Office and at the implementing agencies, is that sustainability is a function of a number of circumstances not related to the modality of execution. Whether the results of the NEX projects will be sustained or not, will depend upon budgetary provisions, political will, and the perception of the Government of the utility of results. There are indications, however, that a number of project results are likely to be sustainable, since budgetary provision has already been made to carry out the activity conducive to maintaining the output that NEX projects have produced or are likely to produce in the future. The Peru experience fortifies one's opinion that directly linking NEX to sustainability would be erroneous.

Under NEX, the Country Offices need a certain amount of flexibility, autonomy, and re-organization. One most commonly emphasized virtue of the NEX modality is the speed and agility with which Country Office responds to the needs of the public sector, in particular, and of the country, in general. To this effect it is important that the Country Office maintains its autonomy. A second organizational issue is the function of the Programme Officers. Unlike in the tri-partite relationship under the UN agency execution, the Programme Officers are vested with both financial and substantive responsibilities. In more cases than not, due to the desire for transparency and the time requirements of proper financial management, substantive accountability may tend to suffer. If NEX is to be successful Programme Officers should be able to draw on the support of independent monitors. This would permit the Programme Officers and the management of the Country Office to properly assume the substantive accountability that is imposed upon them.

In Peru the strong comparative advantage of the UNDP Country Office lies in providing technical assistance and support to the Government in reorganizing the public sector and revitalizing the productive sectors. The traditional agency execution was not a viable alternative, given the Government's need of agility and flexibility. In turn, agility and flexibility were, and still are, the primary virtues of the NEX modality.

The conditions in Peru have also created special relations between the Government and the UNDP Country Office. The trust built between the Government and the UNDP Country Office has been a determining factor in the mobilization of resources. NEX has turned out to be an extremely useful means in this sense. This trust has institutional as well as human components. In addition to the greater autonomy of the UNDP Country Office the continuity of human relations versus frequent rotation of Resident Representatives and their deputies is a matter of concern. Although, for certain reasons, one may favour the Country Office assignment being of a longer term than is currently the case, there might be equally strong arguments not to lengthen tours of duty, in order to sustain an effective relationship between the government and the UNDP.

The experiences from the People's Republic of China, Ethiopia and Peru illustrate the broad diversity of environments and approaches to NEX. In each country context NEX offers a unique opportunity to address real and felt needs. While each country is taking a different approach in implementing NEX, the common denominator is participation in the ownership of a process by which each country's human and organizational capacities for development may be strengthened. In the next chapter this report will review the major imperatives arising from the findings and concerns addressed in chapters three and four for the future of the NEX modality and operations.