Niassa Agricultural Development Project (NADP) (2007) - IOE
Niassa Agricultural Development Project (NADP) (2007)
Completion evaluation
Project history and design
The design process started in 1991 before the peace agreement in 1992. Appraisal was done in 1993, and in April 1994, IFAD's Executive Board approved the Niassa Agricultural Development Project (NADP) with a budget of US$ 20.1 million financed by an IFAD loan of SDR 8.8 million (~US$ 12.4 million), co-financing of US$ 4.1 million from the OPEC Fund, and a contribution of the Government of Mozambique (GOM) corresponding to US$ 3.6 million. The budget was allocated for three components plus project management, including institutional development support: (i) Agricultural Support Services (US$ 7.2 million); (ii) Roads (US$ 8.2 million of which the Organisation of the Petroleum Exporting Countries (OPEC) contributed US$ 4.1 million for rehabilitation of a primary road); (iii) Community Initiatives (US$ 2.1 million). An amount of US$ 2.6 million was allocated for a Project Facilitation Unit (PFU), monitoring and evaluation and institutional development support.
In 1992, Mozambique was the poorest country in the world, emerging from a long period of (civil) war and the effects of oppressive colonial rule. And the situation in Niassa province, remotely located in the north-west corner, was far below national averages. At the time of design, security was a critical issue in many parts of Mozambique, and it was decided that NADP would cover only two of Niassa's 15 districts, i.e. Lichinga and Sanga which had a reasonable security situation. Niassa is Mozambique's biggest province accounting for 16 per cent of the area, but it is sparsely populated accounting for only five per cent of the population. Agricultural land is abundant. The population of Sanga and Lichinga districts is today about 140 000 accounting for less than one per cent of the national population. The Appraisal overestimated the population of the two districts, and thus the target group, by about 100 per cent. At the design stage (1993), it was assumed that 45 000 households (94 per cent of all households in the two districts) were farm-dependent and poor, thus constituting the initial target group. NADP's objective was to improve the levels of income, employment and food security of 45 000 poor family farm households. However, the 1997 population census recorded a total population in the two districts of 107 243 corresponding to about 25 000 households.
The design of the Agricultural Support Services Component included innovative features such as introduction of participatory and farmer-driven research and extension. To enhance outreach and ensure participation, 300 young progressive farmers (of whom 100 would be women) were to be appointed and facilitated as Village Extension Guides (VEGs) to assist their fellow farmers with adoption of improved crop husbandry methods. However, the VEG system was not introduced because the Training and Visit system at the time was the official extension policy. The design also included support for establishment of a Technology Support Fund (TSF) which was to be a revolving fund that would support artisans and farmers with manufacturing of simple equipment, and finance testing of labour-saving technologies. However, the TSF was not established and instead treadle mills/pumps were piloted. An intervention for Support to Rural Trade was included to provide technical assistance and training to individuals and groups involved in rural trade, and rehabilitate 115 village stores to be sold or leased to interested traders with the proceeds going into the District Development Fund. This sub-component was re-redesigned in 2000 and implemented quite differently. Finally, the design included support for establishment of pilot Savings and Credit Associations (SCAs). The target was to establish 95 SCAs benefiting 1 900 households. This sub-component was not implemented though some farmers' organisations did receive some advisory services from the extension system.
The road component had two main sub-components: (i) Primary road rehabilitation which included spot repairs and upgrading of 280 km primary road between Lichinga town and Cuamba in the south. At the time this was a dirt track but the only access to the province; the rehabilitation was to be financed by OPEC with US$ 4.1 million; and (ii) Construction and rehabilitation of feeder roads which would finance rehabilitation of 173 km of secondary and feeder roads as wells as maintenance of 300 km of secondary/feeder roads. In addition, the component would finance an environmental impact study and an economic analysis of an investment in rehabilitation of the railway link. Neither was done, however.
The Community Initiatives Component had two sub-components. (i) Village Infrastructure which would finance construction of 100 water points, rehabilitation of 16 health posts with medical equipment and furniture, and rehabilitation of 24 two-classroom primary schools including furniture and teaching materials. Rural promoters would be deployed to ensure community participation; and (ii) establishment of District Development Funds (DDFs) for funding micro-projects demanded by the local communities who would finance part of project costs. The DDF would be financed directly from project funds and indirectly from funds revolving from the support to rural traders.
Project coordination and management: At national level, the National Rural Development Institute (INDER) would coordinate and oversee implementation which would be delegated to the Provincial Planning Commission (PPC) of Niassa Province, later on the DPPF (Provincial Directorate for Planning and Finance). INDER would be facilitated through provision of computer equipment and travel budget. The PPC would be "strengthened by the establishment of a strong Project Facilitation Unit (PFU)" which would include a monitoring and evaluation unit.
Results and performance
It was highly relevant and in accordance with IFAD's mandate to provide support for Niassa province but it was also a major challenge as the capacity in the public and the private sector was extremely limited. At the time, NADP was the first major external support intervention in the province. The design significantly overestimated the population and target group while underestimating the challenges. Therefore, potential benefits were overestimated, in particular the benefits of the support for agriculture. Assumptions regarding how much NADP could increase yields and production were unrealistic given that input and produce markets hardly existed and considering the challenges of introducing improved crop husbandry in a traditional system of shifting cultivation (slash and burn). Though such a system to outsiders may be considered as a wasteful use of natural resources, it may for the farmers be the optimal choice given the land abundance in the area.
The agricultural support services component did improve the capacity of research and extension and for some years the outputs and services increased. However, as indicated above, the support was not implemented according to design. Output delivery was inefficient for the extension services and highly inefficient for the re-designed support for rural traders, which was poorly managed by the contracted government institute, Fund for Promotion of Small Industries (FFPI). The Lichinga Agricultural Research Station (EAL) was moderately efficient in delivering its outputs and services. However, the number of farmers covered by the services was significantly less than targeted and the impact in terms of increased yields and production is negligible. For the major crops, yields stagnated or declined during the support period. Thus, the component is assessed as unsuccessful in achieving its objectives (effectiveness).
The road component shows a highly mixed performance. Only half of the targeted primary road was rehabilitated with the OPEC funds, and the rehabilitated section needed further rehabilitation immediately after project works had been completed. The target for secondary and feeder roads was surpassed by 51 per cent, though there were numerous implementation problems due to the limited capacity and experience of contractors and implementing agencies. The rehabilitated and constructed secondary and feeder roads have provided very important benefits to a large part of the target group. Thus, in spite of efficiency problems, this sub-component was highly effective in achieving its objectives.
Under the Community Initiatives Component, the Village Infrastructure Sub-component has achieved 72 per cent more than the targeted water points, 58 per cent more than the targeted school rehabilitation and almost the target for health post rehabilitation. The sub-component is assessed as effective and moderately efficient. The support for water points stands out as a highly successful intervention, providing important benefits to about 43 000 people with a relatively modest investment. The District Development Fund (micro-project) Sub-component has no targets but is based on community demand and promoting community participation. It has funded various activities and investments, including water points, road maintenance, soccer equipment, electricity supply, and training of midwives. Implementation was constrained by lack of clear guidelines, and community participation in setting the priorities and implementing the micro-projects did not meet the ambitions of the design. However, some interventions did produce important benefits. The sub-component is assessed as moderately inefficient but moderately effective.
Economic Rate of Return: The Appraisal Report (AR) arrived at an estimated Economic Rate of Return (ERR) of 15.7 per cent by only including, on the benefit side, the incremental crop production, and on the cost side, the investment and recurrent costs of those components which are related to productive activities. Very large production increments were assumed. As actual production increments are nil or negligible, a repetition of this exercise at project completion results in a negative ERR. However, this disregards the important benefits obtained from the support for water points, secondary and feeder roads, and health and education services.
Impact on rural poverty
Overall impact: As it is impossible to attribute any increase in crop production and yields or any improved trade to the Agricultural Support Services Component, the component is assessed as unsuccessful in terms of impact on rural poverty. However, the component has contributed to improving the capacity of the research and extension systems. The support for primary roads is assessed as moderately unsuccessful while the support for secondary and feeder roads is found to be moderately successful. The Community Initiatives Component is assessed as moderately successful with the investment in water points standing out as being successful.
Targeting: The design did not include a differentiated targeting strategy. Spatially, it appears that NADP investments in social and economic infrastructure have been evenly distributed over the two target districts, which however are better off than most of the other districts in Niassa. No special priority has been given to the poorest areas within the target districts, e.g. the lakeside. Targeting within a community has largely been left to the village leaders who not always prioritise the poorest households when identifying the beneficiaries of a specific support intervention, e.g. the goat scheme.
Gender impact: The three components have had a highly differentiated gender impact. The Agricultural Support Services Component has had limited focus on gender issues and the outreach to women is found to be negligible. All of the 22 extension workers in the field are men who for cultural reasons seldom can work directly with female farmers. The Roads Component appears to have complied with government policy that women shall constitute minimum 25 per cent of those employed in road works, and many households including women have obtained employment and income from road construction. The Community Initiatives Component has provided important benefits for women, notably through the support for water points which has reduced women's labour burden and improved health standards. The support for adult literacy, midwives and other health and education services has also benefited many women.
Impact on Institutions, Policies and the Regulatory Framework: Partly as a consequence of NADP, the institutional capacity has been strengthened in the province and the two districts. The organisations, both public and private, involved in implementation have gained significant knowledge over the years, reflected in gradual (yet insufficient) improvements in implementation performance. NADP's participatory approach, working directly with districts, communities and Community based Organisations/Non-governmental Organisations (CBOs/NGOs), may together with other similar projects have influenced GOM's decentralisation policy and the design of other similar projects. However, there is no evidence to suggest that IFAD has used NADP as basis for policy dialogue with the government, with the view of actively influencing the decentralisation policy.
Innovation, replicability and scaling up: The key innovative design feature was introduction of participatory research & development as well as farmer-driven extension based on progressive farmers being appointed as VEGs. Unfortunately, this was not implemented because the official policy at the time was based on the T&V methodology. It remains unclear why this approach was rejected while GOM, by approving NADP, implicitly had given its endorsement. Other innovative design features, which were implemented though not fully according to the ambitions of the design, included the decentralisation of authority and promotion of community participation in setting priorities and implementing activities. These features are today being widely replicated. However, insufficient efforts were devoted to identifying and working with existing farmers' and community organizations.
Sustainability: Generally, activities which GOM considers as a core government responsibility are most likely to be sustained. Such include primary schools, health centres/posts, primary and secondary roads, and to a lesser extent feeder roads and water points. However, community participation may ensure the sustainability of water points and feeder roads. Fortunately, government resources, complemented by donor support, have increased considerably which improves the sustainability prospects for core public utilities and services. Activities, which are outside core government responsibility and which have dubious commercial viability, have limited likelihood of being sustained. Such include the rural trader scheme, which also is affected by mismanagement, the treadle mill scheme and the revolving goat scheme.
Project management and coordination
Frequent changes of project management and coordination structures have occurred since project start. The design envisaged that INDER would coordinate policy and planning with concerned ministries at the national level, but would delegate management responsibility to the provincial planning and finance directorate (DPPF) and the PFU in Lichinga as far as possible. Due to logistic reasons and problems of recruiting competent PFU staff to Lichinga, project management was mostly done from Maputo up to 2001 when the duty station of the Project Executive Director (in 2002 merged with the position of Project Coordinator) and the Financial Management staff was changed to Lichinga. One of several reasons why it took time to move the management of the project to Lichinga was that there were no banks in Lichinga that operated a foreign exchange account. Major issues throughout the project period include: (i) the limited capacity at provincial level to manage and coordinate the project; (ii) insufficient, irregular and unsystematic training of project staff and partners; (iii) frequent changes of project staff; (iv) the absence of an operational forum for coordinating activities and build consensus, also with other donors; (v) lack of internal guidelines and manuals for project implementation, such as a Project Implementation Manual (PIM); (vi) no clear mandate or terms of reference for the Steering Committee (decision making or advisory body?) and no beneficiary representation; and (vii) insufficient use of existing farmers' and community organizations.
NADP did not succeed in establishing an operational Monitoring and Evaluation system, despite considerable investments, an explicit requirement of the Loan Agreement, and recommendations in supervision reports and by the Mid-Term Review in 1999. A Monitoring and Evaluation Officer was in place only between April 2003 and February 2005, when his contract was terminated due to unsatisfactory performance. The consequence has been a lack of systematic data for indicators, ultimately leading to difficulties of monitoring, coordinating and planning project activities. Associated with this problem has been the absence of an Annual Work Plan and Budget following IFAD guidelines using Participatory Rural Appraisal and a proper LogFrame. These instruments were produced for the first time in 2003 with technical assistance support from IFAD.
Financial reporting was done according to the expenditure categories of the Loan Agreement and not according to components, interventions and outputs. Reportedly, the financial accounting system was modified around year 2000 so that it became possible to link expenditure in different categories to outputs and components. However, it has not been possible to obtain information on how much has been spent on the different components/sub-components/outputs over the project period. There is no updated asset registry specifying the economic value of items in the inventory but only a physical inventory with no values attached. Audit reports between 2000 and 2004 bring forward issues such as payments outside the objective of the project, procurement procedures in acquiring expensive IT equipment and services in 2003 for the monitoring system, erroneous booking of expenditure, and FFPI's non-compliance with the contract. A special audit of the support to rural traders through FFPI is currently being conducted and it is recommended that IFAD and GOM follow up on this.
Performance of partners
IFAD: In spite of the fact that IFAD has provided relevant support in line with its mandate, the overall performance of IFAD is assessed as moderately unsuccessful. The design was based on several wrong assumptions, in particular an overestimate of the target group by 100 per cent, which may partly be explained by the fluid and uncertain situation during the design period. However, as normalcy returned, IFAD should have revisited the design. It was too late to wait for the Mid-Term Review in 1999. IFAD allowed the project to become effective and go ahead in spite of the fact that an important condition for loan effectiveness was not fulfilled, namely the requirement for establishing effective arrangements for monitoring and evaluation. IFAD appears to have been distant, in particular in the early period. It was only very late in the implementation period that IFAD started to provide direct support to the PFU, e.g. advisory support on how to prepare participatory work plans and budgets. The re-designed rural traders' scheme appears to have been approved without any in-depth scrutiny. Finally, there is no evidence to suggest that IFAD has made active use of the NADP as a basis for engaging in policy dialogue.
At the end of the implementation period, IFAD established country presence and there is no doubt that NADP would have benefited from having an IFAD representative in Maputo throughout the entire project period. This would have helped to identify various constraints faced by the implementing institutions, and to ensure a more timely response to deviations from project design. An additional benefit could have been a more active participation in donor coordination and in policy dialogue processes.
Cooperating Institutions: In the period 1995 to 1998, the World Bank served as Cooperating Institution (CI), supervising project implementation. This was considered a practical arrangement because the Bank's missions also supervised and visited the World Bank financed Rural Rehabilitation Project. In January 1999, United Nations Office for Project Services (UNOPS) took over as the CI, in full agreement with government and the World Bank whose project had ended. Both CIs have provided timely and frequent supervision missions, however for different reasons their performance is assessed as moderately unsuccessful.
Though there is noting to indicate that the Bank's supervision has not satisfied the formal requirements, the Bank's efforts (and flexibility) to facilitate progress under very difficult circumstances were insufficient. The Bank was strict on enforcing the requirements and conditions, defined in the Loan Agreement as well as in its own operational policies but did not arrange for the capacity development support that national partners needed to deal with these requirements. Applying standard procurement rules assumes a certain capacity within government implementing agencies as well as a private sector with some capacity and competition.
Compared with the Bank, UNOPS tended to adopt a more pragmatic approach, shifting focus from procedures to outcomes and achievements. UNOPS was more flexible on project management, speeding up the process of loan disbursement and procurement. However, though UNOPS' supervision reports were more detailed than the Bank's, they failed to establish a transparent picture of physical outputs, and the expenditure on each output/sub-component/component. Stricter guidance should have been provided on how to report on physical progress and financial expenditure, setting out the actions and responsibilities of the PFU and implementation partners.
Government: Several government institutions have been involved in the implementation: (i) project management and coordination has been the responsibility of INDER (later DNPDR) through the PFU; (ii) provincial directorates (agriculture, public works, education and health); (iii) district administrations (micro-projects); (iv) EAL (agricultural research); and (v) FFPI (rural traders' scheme). The performance of most provincial directorates and the district administrations is assessed as moderately successful while the performance of the overall project management and coordination (INDER/DNPDR/PFU) is assessed as unsuccessful. EAL has delivered many of the services and outputs expected, however with limited impact, and its performance is assessed as moderately successful. FFPI has mismanaged the support (US$ 600 000) for rural traders and its performance is assessed as highly unsuccessful.
NGOs: A number of NGOs have been involved in the implementation. For water points, the international NGO Water Aid was contracted, which in turn contracted national NGOs. Generally, these NGO's have performed well and their overall performance is rated as successful.
Co-financiers: The OPEC Fund played a passive role as a co-financier but did efficiently what it was expected to do, e.g. quick release of funds. The overall performance is assessed as moderately successful.
Overall assessment of NADP
The overall rating of NADP is "moderately unsuccessful" which as explained above hides over big differences. The two major budget items, i.e. the support for agriculture and the primary road, are rated between "unsuccessful" and "moderately unsuccessful" which pulls down the project average. However, even when the components and sub-components are not weighted with their respective estimated share of total expenditure, the overall rating is "moderately unsuccessful". The support for secondary & feeder roads and community initiatives is rated as "moderately successful", with the support for water points achieving a rating of "successful". Overall partner performance is rated "moderately unsuccessful" with NGOs achieving a rating of "successful" while at the other end, FFPI's performance is rated "highly unsuccessful". IFAD and cooperating institutions obtain a rating of moderately unsuccessful. "Project Performance" (relevance, effectiveness, efficiency) scores closer to "moderately successful" than "moderately unsuccessful" thanks to high ratings of relevance.
In spite of a modest overall rating, the Evaluation Mission agrees with those Mozambican observers who find that NAPD has been an important project. At the time, NAPD was the first major intervention in Niassa and it had a pioneering role, paving the way for others to follow. Though important, it is difficult to quantify the benefits of this pioneering role. NADP, together with other support, has made an important contribution to the development of the capacity in Niassa's public and private sectors from a very negligible base. And, NADP has positively changed the daily life of many households through its support for water points, roads, and rehabilitation of education and health facilities.
Insights and recommendations
When project design takes place in situations of extreme uncertainty, e.g. post-conflict and post-emergency, an inception review should be implemented to assess the validity of assumptions and the design, and introduce the needed design modifications early in the implementation period. The review would be done jointly with the government and have the mandate to recommend the design changes required. Preferably it should be done when the project management team is on board and has had some time (say six months) to assess the situation and the project design.
It is unusually difficult to forecast the project context when designing a project at the end of or just after a conflict (or emergency), where the situation is fluid, volatile and uncertain. When "normalcy" returns, the reality may change dramatically within a short period of time (one year), making some of the project assumptions and part of project design irrelevant. The time from project approval to Mid-Term Review (MTR) is normally four to six years, depending on the elapsed time from approval to effectiveness, and this is too long to wait for adjusting design to the changed realities. NADP was designed during 1991-1993, i.e. just around the peace agreement in 2002. Serious implementation only started around 1996/97 and the MTR was implemented in 1999. The NADP design would most probably have been different in a number of aspects if it had been designed during 1995/96.
The Government of Mozambique should assess if there is a need for strengthening its internal systems for appraising and approving donor-assisted projects. It was unfortunate that some of the innovative design features, notably the VEG system, were rejected following the approval. This may be an indication that the government did not fully take ownership of the design process and/or that the government's internal appraisal/approval system did not sufficiently analyse the details of the proposal. Since then, the internal systems may have improved but it is suggested that GOM assesses whether the current systems are such that the experiences of NADP are avoided.
Projects in remote areas need to be managed and coordinated locally and the feasibility and problems of doing so need careful assessment at the design stage. During NADP's first five years, implementation was managed and coordinated from Maputo which negatively influenced performance. This was contrary to the design but considered necessary for a number of unforeseen practical reasons. It is, therefore, important to assess the obstacles already at the design stage and include measures to reduce the obstacles in the design (special employment conditions, communication facilities etc.).
In remote areas with limited human and institutional capital, considerable investments in capacity development are required during the early implementation phase, and repeat-training is needed to bring new staff on board. These needs are significantly higher in multi-sector projects with many implementation partners. Though NADP had an adequate budget for training it was not utilised, and too little was done initially to define implementation rules and procedures and to train the many partners. In post-conflict and post-emergency situations, there is a natural urge to deliver something on the ground quickly, and there may be a tendency to jump into implementation without first establishing the foundation for doing so. Therefore, in such situations the effort to develop capacity and systems need to be intensive but brief. Subsequently, when implementation had started, NADP should have done more to organise repeat-training in order to bring new staff on board. Such investments may be considered as "delivery costs" and for a multi-sector project in a remote area working with numerous institutions with limited capacity, these costs are considerable but necessary for achievement of impact.
Since 1994, IFAD has obtained more grant resources for supporting the portfolio and the partner countries. In a future project similar to NADP, it would be relevant to use grants for providing intensive support for capacity development and for close guidance and implementation support, in particular during the initial stage.
Major interventions in areas of core government responsibility have performed better than small scattered interventions in the border area of the public sector domain. The investments in secondary and feeder roads and in water points, schools and health posts have made an important difference in the districts and have reasonable sustainability prospects. Performance of these interventions was considerably better than the performance of small scattered activities such as goat re-stocking, rural traders' network, treadle pumps etc. which tended to become appendices without the required management support. While there are provincial directorates which have mandate, responsibility, staff and budget for roads, and health and education facilities and services, there is no provincial directorate whose core responsibility (with related budget) is to promote goat re-stocking, treadle pumps, or village shops. Furthermore, government institutions and staff are generally not specialised in developing commercially viable businesses. When the management capacity is limited, there is a need for focus and a less diversified support menu, even under a participatory approach.
Furthermore, the NADP-supported interventions of a public goods nature have better prospects of sustainability than the NADP-assisted commercial activities because the commercial viability of the latter is uncertain. However, government needs to take follow-up actions to ensure the sustainability and benefits of some of the public goods interventions, including:
- In the roads component, there are some unfinished works including construction of an aquaduct on the N'toto – Choule road, two bridges, and completion of Unango – Muembe road. These works need to be funded and completed in 2006. Reportedly, the budget for these investments is included in the provincial Annual Social and Economic Plan (PES) for 2006 but it needs to be ensured that it is spent for this purpose.
- A decision regarding future ownership and management of the equipment financed by NADP for road construction and maintenance needs to be taken before project closure.
- The support for mobile midwives has had a positive impact and is much appreciated. However, to sustain the services, the maternity kits provided to the midwives need to be replaced periodically.
Targeting strategies and selection criteria are required in design and implementation. All farm households were included in the target group leaving it to community and district leaders to select the beneficiaries of various support interventions. Occasionally, beneficiaries were not poor, and the needs of the most vulnerable households were neglected.
Design and implementation should clearly address gender aspects, strategies and resources. In doing so it is important to take into account cultural and traditional constraints and practices. For example, it may prove difficult for male extension workers to reach female farmers since interaction with the opposite gender often is seen as inappropriate, particularly for married women. Targets and objectives should also be defined, allowing for gender disaggregated M&E indicators, and progress reporting on gender strategies and targets. For example, abandoning the design strategy of appointing 100 female VEGs should have been an issue for discussion.
An effective approach to raising agricultural productivity, production and incomes in sparsely populated remote areas with land abundance has not been identified. This is a negative lesson learned and very frustrating since a major effort was invested in this objective. It remains for discussion why the efforts had limited impact and what could have been done differently to ensure the achievement of this important objective.
If the explanation is that farmers do not respond to recommendations on improved crop husbandry because they cannot access the required inputs and tools and/or because there is no market for their incremental production, then this would indicate that investments in research and extension are premature and need to await the development of market infrastructure and services. For the government, which is trying to prioritise a limited resource envelope, which is insufficient to support all areas of Mozambique with research and extension, this would indicate that priority should be given to commercial agricultural areas with a certain level of market development, while road infrastructure and livelihood investments (e.g. water points) initially would be more relevant for remote areas like Niassa. However, an alternative explanation could be that farmers have no demand for changing their traditional extensive shifting cultivation system as long as land is abundant. If this is the case, one may question the role of research and extension, at least the approach.
The capacity of District Administrations has improved significantly over the project period, and if NADP were to be implemented today, District Administrations could serve as implementing agencies for many of the NADP supported activities, while the provincial government would have the overall implementation responsibility. A District Administration of 2006 is completely different from a District Administration of 1994. With some support, most district administrations would today be capable of implementing many of the NADP activities. As compared to the current management from a provincial directorate in Lichinga town, this would bring management closer to the beneficiaries, with a number of consequential benefits.
Finally, in a possible future area-based multi-sector project, IFAD and GOM may consider to assign the overall implementation responsibility to the provincial government, instead of a national line agency, but with the Ministry of Finance or Ministry of Planning and Development providing general oversight and serving as the official national counterpart.