United Nations Capital Development Fund

MALAWI INTERGOVERNMENTAL FISCAL TRANSFERS STUDY

Dr. Jamie Boex
Mr. Randson Mwadiwa
Mr. Reckford Kampanje

Government of Malawi
United Nations Capital Development Fund /
United Nations Development Programme

Lilongwe and Atlanta, April 2001

EXECUTIVE SUMMARY

Objective of the Study

The principal objective of the study is to assist the Government in establishing an objective mechanism for the allocation of revenue from the central government to local assemblies. As such, the Final Report provides recommendations in specific areas of intergovernmental fiscal transfers. The study recommends:

a mechanism for determining the overall size of the transfer pool (as a share of the national budget);
options for the types of fiscal grants;
options for the formulas which are needed to accomplish the Government's objectives; and
how the grant system should be phased to provide financial support for local governments during this transition period.

General Design of Transfer System

Our preliminary recommendations with regard to the general design of the system of intergovernmental fiscal relations are in line with previous decentralization studies in Malawi. We recommend that the Government of Malawi introduce a vote in the annual budget to be known as "Transfers to Local Assemblies" under the control of the Ministry of Finance for the purpose of funding local government activities that have been devolved to the local government level. General funding to the local government level would be provided through the General Resource Fund (GRF). The GRF would replace the Resource Supplementary Grants, as well as other transfers currently in place to fund recurrent local government activity.

In addition, provisions within the sectoral votes would be made for several sectoral transfer funds, such as the Education Fund, the Health Fund, and possibly others. Each of these funds would be distributed on a formula basis, determined by the Local Government Finance Committee. Further, the central government should provide full cost-reimbursement if it relies on local governments to implement national policies.

Administratively, the Secretariat of the Local Government Finance Committee would determine transfers to each local government on a monthly basis. This determination would be based on the level of monthly revenue collections as well as the vertical sharing rules and the horizontal allocation formulas. Local assemblies would receive their monthly allocation as financial transfer with a predetermined disbursement schedule. An absolute commitment by the central government is needed to ensure that consistent and timely transfer payments will be made. After all, local governments will be given the responsibility for several key government activities, including primary education and primary health care. It would be irresponsible government policy to allow a disruption of the flow of resources to fund teachers, health care workers, and medicine.

While we recommend substantial fiscal discretion be accorded local governments, they would still be subject to the same financial management rules, accountability provisions and audits as central government agencies or parastatal enterprises. Local governments would be required to submit monthly reports and demonstrate on a quarterly basis that their budget expenditures conform to financial regulations and that local spending adheres to the conditions imposed by the sectoral grants. In order to assure fiscal accountability, local governments should be subject to disciplinary actions in case of financial mismanagement, and a court-appointed administrator might be placed in charge of a district's financial management in case of persistent failure to comply with key conditions of the transfer system. In addition, local governments might be required to demonstrate that they have sufficient administrative capacity before key sectoral responsibilities can be devolved.

The Role of Capital Development Funds

Community-focused, donor-supported capital development funds currently form a separate structure that operates similar to a system of intergovernmental transfers. Complete integration of the community-focused donor funds into the system of intergovernmental fiscal transfer (for now, at least) is neither administratively feasible, nor acceptable to the donor community. It would be desirable, however, if MASAF, DDF, and other donor funds could adopt a number of measures that would bring their practices in line with the new reality of locally elected assemblies:

First, it would be highly desirable if all development funds could distribute their funds among the local governments based on an allocation formula (the DDF already does this).
Second, the development funds should coordinate all their activities through the new local assemblies; allow the assemblies policy control over the selection of local development projects; and integrate the capital development planning into the annual district budget process.
Third, MASAF, DDF, and other funds could provide greater policy discretion to local governments by funding a wider array of development activities, including human capital development. In addition, donor should play an important role in the raising awareness in civil society, fiscal decentralization policy training for key officials and capacity building on the ground.

A number of long-term considerations should also be dealt with pertaining to capital development funds in Malawi. For instance, over time, it is advisable to make repayment of MASAF loans a responsibility of the local government level. The optimal (long-term sustainable) level of capital infrastructure funding ultimately will depend on the ability of local governments to repay capital development loans.

Vertical Allocation of Resources

The vertical allocation of resources refers to the allocation of resources between different levels of government; the main determinant of the vertical resource allocation is the manner in which expenditure responsibilities are assigned between different levels of government.

We recommend that the Government of Malawi determine funding rules for the GRF and the sectoral transfer funds that specify the transfer pools as a share of a broad-based national revenue base. The National Revenue Base should reflect the pool of public sector resources that is available at the discretion to the national government to fund public sector activities, either at the central or local government level. Thus revenue sources and budget items that are not at the discretion of government, such as the cost of servicing the national debt, should be excluded for the national revenue base. On the other hand, there is no reason to a priori exclude any discretionary budget items (even discretionary, targeted transfers from central to local government) from the national revenue base. The National Revenue Base (which is to be computed on a monthly basis in order to assign transfers to local governments) could be defined as actual tax collection, excluding the following deductions:

Debt service (interest payments)
Refunds and retentions
Pensions and gratuities

Due to the existence of notable macro-economic fluctuations in Malawi, we recommend that actual collections figures are used for the purpose of determining transfers, as opposed to relying on the nominal amount contained in the budget plan. The use of nominal budget figures would expose both the central government and local governments to destabilizing fluctuations in real resources when unexpected changes in the inflation rate or economic growth occur. In order to further enhance stability, we recommend that the funding rule be initially applied for a period of three years.

Based on historical expenditure patterns, preliminary recommendations for the size of the respective funds (as a percentage of the national revenue base) are:

Education Fund: 15 percent
Health Fund: 9 percent
General Resource Fund: 4 percent

It should be noted that the system of intergovernmental fiscal relations (including the transfer system) should be flexible and able to respond to changing needs over time. In order to accord local governments with greater policy discretion and revenue certainty over time, we suggest that the following issues are considered as part of the evolution of the system of intergovernmental relations in Malawi:

Decreased emphasis on sectoral funds;
Increased role for the General Resource Fund;
Increased emphasis on locally generated revenue sources, revenue sharing and ceded revenues;
Integration of transfers for recurrent expenditures and development funds (DDF and other donor-funded capital development funds).

Horizontal Allocation of Resources

The horizontal allocation of resources determines how the available pool of funds is distributed among eligible local government units. Section 5 contains extensive statistical analyses of current allocation patterns in Malawi. Several general principles should be considered in selecting variables to be used in allocating resources to local governments:

The factor should be statistically sound and regularly updated (for instance, census data should be used as opposed to variables occasionally produced)
The data source should be respected by all stakeholders, and the factor should be free from manipulation by local government officials (for instance, school enrollment and health care utilization figures can be easily manipulated by local governments; instead the National Statistical Office is seen as an independent source of data by all stakeholders)
The factor should reflect the need (demand) for a service (i.e., the number of potential clients for a government service), not necessarily the current supply. For instance, the use of infrastructure in an allocation formula will likely introduce inequity and inefficiency.
The formula should not rely on the "equality principle," otherwise:
- The system would be very inequitable (unfair): Mzuzu City would receive 15 times more resources per person than Lilongwe Rural District.
- The system would be inefficient, because resources would not be allocated where they are needed most (allocative inefficiency)
- Districts would have a reason to divide into inefficiently small districts (technical inefficiency)

Education Fund

The Ministry of Education has made significant progress in its preparedness for decentralization since December 2000. Indeed, Ministry officials have suggested that it is in principle ready for devolution of its primary education responsibilities to the local level. It has suggested that an appropriate approach would be to allocate 100 percent of the Education Fund to local assemblies in proportion to the number of school-aged children in each district. This variable is a demographic variable provided by the NSO, and thus cannot be manipulated by individual districts. This formula would result in a transfer of approximately MK 690 per school-aged child. We recommend the use of the number of school-aged children, as opposed to actual enrollment figures, due to the limited capacity of the central government to monitor enrollment figures reported by the local assemblies.

Two recommendations are made with regard to the devolution of this responsibility to local assemblies. First, it would be prudent to only devolve the responsibility for primary education to about half of the districts in 2001/02, notably those most capable of administering the transfer of responsibilities (presumably City Assemblies, Township Assemblies and LIAs). The remaining districts would assume the responsibility for primary education in the following budget year. Second, the formula for the allocation of the Education Fund could be phased in over time to assure a smooth transition, so that:

during the first year, districts receive the amount budgeted by MOEST;
during the second year, districts receive 50% based on previous year's allocation, and 50% based on formula;
during the third year, districts receive 100 % based on the formula.

Different types of capacity building and training will be needed within the education sector, including fiscal decentralization policy training at the ministrial level and training in budgeting and planning for DEOs at the district level.

Health Care Fund

Our observation with regard to the Ministry of Health's preparedness to devolve its responsibilities for primary health care is that while the Ministry has stated that it is "committed to decentralization," it is clearly felt that the previous timetable was "too fast" and that no devolution of health care activities should occur during the coming fiscal year. In addition, there was a consensus among the technical experts that the district health system should be devolved as a block, as opposed to devolving the responsibility for clinics first; instead, phasing the level of responsibility of assemblies over the health sector seems to be a better option (this contradicts the Ministry's current policy). However, the Ministry is unclear with regard to the administrative and managerial arrangement that should be put in place for the Ministry to continue monitoring health care delivery in a decentralized system. The Ministry of Health officials expressed a need for more training and sensitization on decentralization issues and desired improved coordination with the Decentralization Secretariat.

In the interim, the Ministry of Health is in the process of developing a formula in order to allocate resources for ORT among the districts. However, since these resources are flowing through the Ministry of Health to the District Health Officers, this remains a deconcentrated approach as opposed to a truly decentralized approach to health care delivery. The Ministry is planning to expand the use of formulas in future years to the allocation of personnel expenditures and the drug budget.

The formula currently proposed by the Ministry of Health relies largely on the out-patient utilization rate, existing infrastructure, and historical spending levels. There is a need to improve the proposed formula to assure that it is based on the need (demand) for health care services, and not on the current level that is supplied. We are further extremely concerned that in the absence of a strong central ability to monitor local government activities, this will cause inequities and inefficiencies in sectoral allocations; it is our understanding that the necessary data is currently not available within limits of acceptable quality. A simpler formula approach, based largely on demographic data, should be pursued as the necessary databases for more complex health care funding schemes are developed.

Formulating the General Resource Fund

The level of funding for the General Resource Fund should be based on the recurrent expenditure needs for local governments to engage in minor local government activities (for which no sectoral grants are given) plus 20 % overhead on all sectoral expenditures for the general operating cost of the local government level. This amount may be reduced by the aggregate level of locally generated resources that is available to local governments. We consider 4 percent of the National Revenue base to be a reasonable estimate of general local government needs to be funded through the GRF.

We recommend using the Resource Supplementary Grant (RSG) as a guide in determining the formula for the General Resource Fund (GRF). However, we suggest several modifications to the formula that is currently being used by the RSG. The table considers compares the recommended formula for the GRF with the current RSG formula:

Factor	RSG	GRF
"Equal Shares"	50	0
Population	30	80
Drought-stricken districts	13	0
Least-developed districts / Above-average poverty	7	20

In summary, we recommend that the General Resource Fund be allocated between the districts based on the following formula:

80 percent in proportion to population;
20 percent in proportion to above-average poverty

How did we move from the current RSG formula to our recommendation for the GRF? First, as noted earlier, the use of the equality principle is ill-advised, as it causes inefficiency and inequality into the transfer system. In line with recommendations made in the Accountability Report, we recommend the new formula combine the equality share (50 percent) and the population share (30 percent) and disburse 80 percent of the General Resource Fund to all local assemblies (cities, townships and rural districts) in accordance with population.

Second, the two remaining factors included in the RSG attempt to catch disadvantaged regions. However, the measures chosen seem rather arbitrary and ineffective. For instance, the formula targets drought-stricken regions, but does not tie the funding provided to these regions in any way to possible relief of the drought situation. Moreover, drought is only one of a number of natural emergencies that Malawi periodically faces; others include floods, crop disease and pests. In addition, response to emergencies is considered a national responsibility, and the National Disaster Fund can be used for this purpose. Additionally, the manner in which "least developed districts" are determined is not necessarily transparent. We recommend that in the new formula, 20 percent (equivalent to the 13+7 for drought-stricken and least developed districts) of the general resources should be allocated to rural districts with above-average levels of poverty. Only rural districts would be eligible for these funds, since urban areas and townships have significantly higher fiscal capacity.

As the decentralization process progresses, it may become necessary to consider additional sectoral funds. Sectors for which sectoral funds should be considered include environmental protection (including related activities, such as forestry), agricultural extension activities, and local streets and roads.

Implementation Concerns and Next Steps

In order to ensure clarity, transparency, and stability in the transition process, the GOM should move expeditiously in defining the overall timetable for the devolution of the various functions and the specific activities. In addition, a substantial number of implementation concerns remain that either directly or indirectly will affect the success of failure of the system of intergovernmental fiscal transfers that is envisioned. Successful implementation of the decentralization program will require addressing these concerns. The next steps in introducing a system of intergovernmental fiscal transfers in Malawi include:

Political approval of the policy recommendations contained in the report, and political endorsement and support from the highest government authority;
An expression of clear expectations and deadlines with regard to the devolution of primary education (pilot districts in 2001/02; all districts in 2002/02) and primary health care (pilot districts in 2002/03; all districts in 2003/04). It is crucial to keep the momentum going.
Continued efforts to raise awareness among central government officials, local assemblies, donors, other stakeholders, and the public at large about the next steps in the implementation of the decentralization reform;
Strengthen the Decentralization Secretariat for its role in implementing decentralization, including the inclusion of a number of sectoral line officers, assuring improved coordination between the Secretariat and the line ministries;
High-level fiscal decentralization policy training for key personnel at the ministerial level (including the Decentralization Secretariat, Finance, Health, and Education),
Capacity building activities and training for local government officials, in the areas of local fiscal policy, fiscal decentralization, and local government budgeting, planning and administration. This training should focus on local elected officials, local finance/planning/administrative officials, as well as local sectoral officials, including District Health Officers, District Education Officers, and their respective administrative staffs.
Finally, fiscal decentralization is an ongoing process. Over time, there should be an emphasis on increasing local revenue sources, moving away from sectoral funds, and periodic reviews of the transfer system to assure vertical and horizontal balance. Continued monitoring and evaluation of the decentralization process by a combination of Malawian and international fiscal policy experts is recommended.