Case Study Mozambique

Municipal Development Programme

Donor grants: Strengthening autonomy or dependency?

Maputo, 30 December 2010

Contributor: Marc de Tollenaere
Peer-review: Mali, Bangladesh

Video Intro
Have a look at the video intro to get an overview of the case study and its contents.

Fact sheet

1. Context

1. Mozambique has a two-pronged approach to decentralisation. On the one hand the central state deconcentrates competences and resources to the so-called Local State Bodies. These are made up of an intermediary level (10 provinces) and a local level (128 districts – varying between 3000 and 250000 people). The leaders of these Local State Bodies, respectively governors and administrators, are appointed by the President. Provincial and district administrations are staffed by civil servants. Since 2009 the provinces have elected assemblies who essentially have to approve the annual plan and budget of each province and oversee the provincial government’s performance. At district level there are no elected bodies but there is a district development council composed of people that are representative for the local population (businessmen, traditional leaders, religious leaders, civil society leaders,…). This council can advice the local government on the allocation of budgets and development grants. Provincial and district governments have very limited autonomy and are essentially local executive branches for decisions taken at central level. In 2010, provinces managed about 15% of the State Budget and districts ca. 10%. In 2011, this will be 13% and 12,5 respectively. This is a considerable chunk of the budget, but practically all of it is for the payment of salaries, so there is very limited discretionary power on budget allocation. The investment budget is still managed for 85% at central level.
2. Besides this deconcentration, Mozambique introduced in 1998 “autarquias” or municipalities. These municipalities have legally mandated autonomy over local economic development, sanita-tion, waste collection, secondary roads, etc. They are also vested with political and limited revenue generating authority. The municipalities have directly elected mayors and elected municipal assemblies. There are currently 43 municipalities. About 13 of these have clear urban characteristics while the others rather rural service centers (the smallest has 10000 inhabitants, the largest 1.2 million). About 30% of the Mozambican population lives in the 43 municipalities. Besides gen-erating own revenues from legally defined fees and taxes, the municipalities receive two block grants from central level. One for recurrent costs and one for investments. The total of transfers is not more than 1% of the annual State Budget.
3. Mozambique has no clear decentralisation strategy or vision. There are no clear objectives or trajectories in terms of administrative, fiscal or political decentralisation over the next decade(s). There are obvious frictions and tensions between the deconcentration to (dependent) Local State Bodies and the devolution to (autonomous) municipalities because of territorial and mandatory overlaps. The legal framework is extremely complex as the legislation on local governments is not always aligned and harmonized with sector legislation (for example on land, water, natural re-sources, etc.). Moreover there is still a “capacity gap” between legislation and reality. Planning and management capacities are low and institutions weak (performance often depends from the effort of one or more individuals).
4. SDC is supporting both Local State Bodies and municipalities, but for this case study we will only concentrate on our support to municipalities. After the creation of the first batch of 33 municipalities in 1998, SDC was one of the first donors to develop a project. 5 of the poorest municipalities were selected. The project was directly implemented by SDC. Only limited funds were transferred to the municipalities. Project planning and financial system were parallel to the national system. This programme (called PADEM) was operational from 2000 to 2007.
5. In 2007, SDC decided to join forces with the Danish and Austrian Cooperation who had also gained experience with municipal development. Besides increased harmonization the new pro-gramme also intended further alignment with national systems and to reduce the transaction costs of municipalities receiving external support. The programme started in 2008 and will run until June 2011. The formulation of a successor phase is ongoing.
6. The theory of change of the current programme is that through technical and financial support the 13 municipalities will be able to sustainably improve the living conditions of their citizens.
7. The reason why we want to share this case study is that our support to municipal development both in substance and modality has been and continues very dynamic and has by itself been a learning curve. The reason why we opted for donor grants as central theme is that that particular element has been the biggest innovation in how we deliver support to municipalities.

2. Activities and approaches

1. The support strategy/approach was characterized by:
1. A highly participatory formulation process. An individual needs assessment was conducted in each of the municipalities (and in close collaboration with the local actors). Various seminars took place where each municipality was represented by 3 or 4 elected officials. Within broad parameters (basically poverty reduction and improvement of the urban environment) and based on the needs assessment each municipality identified specific priorities for a 3-year period. All these priorities were then bundled in 5 components (see outcomes in fact sheet).
2. As a result of this participatory formulation the programme became very open and demand-driven. The idea was that a tailor-made approach would work best given the enormous differ-ences between the municipalities (smallest 15000, largest 500000 inhabitants).
3. A more systemic approach. Until 2007 we only worked with municipalities. Now it was decided to involve also key ministries (Finance, Public Works, Environment and State Administration – only the latter two would involve actively, the others quickly lost interest) and to involve the National Association of Municipalities. All participating institutions (20 in total) formed the governing body of the programme (Steering Committee). The Ministry of Environment is the political host of the programme as well as the host of the technical assistance team and the programme coordination.
4. Increased alignment. The programme follows the official budget and planning cycle and the externally funded activities must be integrated in the annual municipal plan and budget that is approved by the municipal assembly. The programme uses the national procurement procedures. Yet the funding is handled through a separate account and still requires separate re-porting.
5. Increased harmonization. The 3 donors agreed to join their funds in one common mechanism managed by the Danish Embassy. The donors agreed not to have the common fund managed by a central ministry as it was considered essential that the primary entry point had to be the municipalities (previous experiences of using a central ministry as entry point for funding of local governments did not work well). The Ministry of Finance indicated that they were not yet ready to receive the funds as earmarked budget support and the national Association is a very weak institution. So it was by exclusion or default, rather than by preference, that the funds are still managed by a donor.
   
2. The grant system is primarily output-based. Each institution (13 municipalities, 2 ministries and 1 National Association) receives an annual allocation. The formula for the municipalities is simplified. We distinguish three categories: 2 big municipalities (ca. 400000CHF/year), 8 medium sized municipalities (ca. 300000CHF/year) and 3 small municipalities (ca. 200000CHF/year). The municipalities need to submit an annual plan and budget to the Steering Committee. Approval by the Steering Committee is a precondition to get access to funding. After that each activity needs a technical and financial green light of the programme coordination unit (in which the TA is integrated) before funds are released. Moreover each beneficiary has to submit monthly reports. Problems with the reporting can also delay further transfers.
3. In 2010 an additional performance grant was introduced. Those municipalities that had spent over 75% of their budget in 2009 and had up-to-date reporting received a substantial performance bonus (ca. 400000CHF). 5 municipalities qualified in 2010.
4. This grant-based system is considered an intermediary step between project funding and budget support. The theory of change was that a more aligned form of support would be more effective and would also contribute to more structural changes rather than ad hoc investments.

3. Experiences

1. The support strategy has certainly been affected by the complex management structure that was put in place. That running a programme with 20 institutions in a not always very conducive environment would be difficult was anticipated. Yet, decision-making was heavier and slower than expected as a result of:
1. Disregard by some municipalities of the main objectives of the programme when presenting annual plans. Some activities (for ex. the construction of a sports stadium) had nothing to do with the programme objectives. These gave rise to lengthy discussions in the Steering Committee, the need for amendments and thus a delay in the disbursement of the grants.
2. Delays in the recruitment of a programme coordinator.
3. The inclusion of central ministries and the National Association turned the Steering Committee into a more political body where other agendas rather than the programme would regularly get the upper hand.
   
2. Progress has been made in terms of alignment with national systems and procedures, yet a donor managed common fund entails limitations. The grants are strictly programme based and thus require separate reporting. The municipalities have complained that the various verification mechanisms before grants are released have slowed down execution rates, yet the manifold corrections that had to be made by the technical and financial staff of the programme were not to bring requests in line with parallel bureaucratic conditions. is an indicator that municipalities still had to improve
3. Municipalities have used the grants in various different ways. Some used the funds for a limited number of investments and others for a high variety of expenditures, including small operational costs.
4. The accountability on the use of the grants is still highly focused on the donor-beneficiary relation. The programme has triggered the availability of more data (for example on own revenues and central transfers) but the information is still insufficiently disseminated.
5. The performance based grant initiated in 2010 was controversial. It was intended as an incentive for well managed municipalities, but it actually gave an advantage to those municipalities that used their grants for a few big investments (rather than slower and more difficult capacity development). It is unlikely that the experiment will be repeated.
6. The focus is almost exclusively on short term visible or tangible gains. During a recent Steering Committee meeting one mayor explained that if given the option between investment or capacity building, he would always go for the investment. If the programme wanted to improve management aspects that should be imposed rather than optional. The demand-driven philosophy of the programme simply did not work, least of all when it came to fiscal and financial management aspects.

4. Lessons learnt

1. From our (donor) perspective we certainly learned that the current more harmonized approach is better than running competing and individual projects. Yet this comes at a cost of increased complexity as regards the management of the programme and it is a dynamic process to improve dealing with that complexity.
2. A 3-year phase has proven to be short to set up this new approach to funding municipal development. A 5-year phase would have been more adequate. Both donor and partner need to get used to the new modality and this takes time. In particular the meetings of the Steering Committee where all involved partners are represented started off on a difficult an tense foot but have now turning into platforms where some of the structural problems related to the Mozambican decentralization process are discussed (for example on the boundaries of municipal autonomy frequently transgressed by central institutions). The Steering Committee meetings have evolved in peer reviews where performance is compared (budget execution rates, tax collection, reporting,…) and commented with all present. This has created a dynamic where local governments are increasingly interested in learning from each other and get stimulated as they see how some peers manage to perform better.
3. Particularly challenging was the management of the fiduciary risk: grants are prone to a higher fiduciary risk than a project approach, yet it is a much better way to foster development. The balance between advantages and disadvantages has been difficult to handle from a donor point of view: when do fiduciary risks outweigh the advantages of the grant approach?
4. The formula to allocate grants received insufficient attention. The mayors tend to favour an equal distribution of the grants, independent from population or budget. That is a political rather than a technocratic position. We agreed on 3 categories that generally speaking give an advantage to the smaller municipalities. For the large municipalities our grants do not represent more than 5% of their annual budget. In some smaller municipalities our grants represent up to 60% of their annual budget. The argument used in favour of that bias is that the smaller municipalities have more difficulties raising own revenues (true) and that they have higher poverty rates (not possible to confirm as there are no reliable comparative data). The counterarguments are that the smaller municipalities have less absorptive capacity (demonstrated by lower execution rates) and that the investments can not be followed by the available recurrent budget to secure proper operation and maintenance. This can make our investments less sustainable and increase dependency. If the external contribution to larger municipalities is only a small percentage, that may also reduce the leverage of the programme to induce change.
5. Municipalities do not like to raise their own revenues. The 13 municipalities we support raise between 5% and 50% of their annual budget from local fees and taxes. Recent analytical work demonstrated that municipalities on average effectively raise not more than a quarter of their potential fiscal income. Therefore there is no doubt that our grants reduce the pressure to raise own reve-nues. A bolder approach will be required in the future to address the link between own revenues and external support.

5. Conclusions

1. Municipalities have improved service delivery and infrastructure but were insufficiently interested to use the grants for institutional development and capacity building. This may compromise some of the short-term gains in the medium to long term.
2. Long-term engagement is important. We have been able to improve the way we provide grants through successive programme phases. Longer-term engagement gives the possibility to learn provided one does not shy away from it. It is hard and intensive work. Adapting the support strategy is a very necessary dynamic process. Some time down the line, the donor funding should be channelled through an earmarked contribution to the central state budget and than on to the municipalities as additional central transfers, but that would also require changes to the current legislation as well as the establishment of the necessary safeguards at local level.
3. It is essential to get reliable baseline data on municipal finances, human resources and services. In the Mozambique case it took us nearly 10 years to get that and it continues difficult to update. Financial data for each municipality are very important to inform for example the formula for grant allocation.
4. SDC needs to secure qualified staff. Without qualified staff we would not have been able to gradually improve our approach.
5. It is very difficult to find the right balance between incentives and conditionalities and also this needs time to develop. We are actually on a path where we gradually increase incentives (more money, less programme control) as well as conditionalities (more focused funding rather than the demand-driven approach).
6. There is a need to accept that a programme like this has besides many technical dimensions also a strong political dimension and that therefore compromise rather than the technical optimum rules.