Report No. 54692-MZ MOZAMBIQUE PUBLIC EXPENDITURE REVIEW FOR THE WATER SECTOR MAY, 2010 ACRONYMS AND ABBREVIATIONS AfDB African Development Bank INGC National Institute of Disaster Management Water and Sanitation Infrastructure AIAS JMP Joint Monitoring Programme Management Agency ARA-Centro Water Basin Agency for the Central Region MCC Millennium Challenge Corporation ARA-Sul Water Basin Agency for the Southern Region MdF/MoF Ministry of Finance (Ministerio das Finanças) Regional Water Basin Agencies ARAs MoPH Ministry of Public Works and Housing (Administraçoes Regionais de Aguas) On-budget and On-CUT non-project external ASAS MDG Millennium Development Goals funding Review of the Economic and Social Plan Ministry of Planning and Development BdPES MPD (Balanço do Plano Economico Social) (Ministerio da Planificaçao e Desenvolvimento) BER Budget Execution Report MTFF Medium-Term Fiscal Framework CGE State Final Accounts (Conta Geral do Estado) NGO Non-Governmental Organization CPI Consumer Price Index NRW Non-Revenue Water Council for the Regulation of Water Supplyl CRA (Conselho de Regulação do Abastecimento de ODA Overseas Development Assistance Água) Single Treasury Account (Conta Única do Overseas Development Assistance Mozambique CUT ODAMOZ Tesouro) Database DAS Provincial Directorates of Water OE State Budget (Orçamento do Estado) Poverty Reduction Action Plan (Plano de DMF Delegated Management Framework PARPA Acçao para a Reducçao da Pobreza Absoluta) Economic and Social Plan (Plano Economico e DNA National Water Directorate PES Social) Budget Execution Report (Relatorio de EBIT Earnings Before Interest and Tax REO/BER Execuçao Orçámental) EC European Commission SBS Sector Budget Support National Financial Administrative System European Development Fund (Fundo Europeu FED Sistafe (Sistema Administracao Financiera do Estado) de Desenvolvimento) (E-Sistafe is the electronic versión of Sistafe) Water Assets and Investment Fund (Fundo de FIPAG Investimentos e Patriônio de Abastecimento de SWAp Sector Wide Approach Água) Supreme Audit Institution (Tribunal GBS General Budget Support TA Administrativo) GDP Gross Domestic Product VAT Value Added Tax Inspectorate General for Finance (Inspecçao IGF WB World Bank Geral das Finanças) National Institute of Statistics (Instituto INE WRM Water Resources Management Nacional de Estatistica de Moçambique) WSS Water Supply and Sanitation i Table of Contents WAY FORWARD: SUMMARY OF RECOMMENDATIONS ......................................................................................... 1 EXECUTIVE SUMMARY .................................................................................................................................................. 9 1. OVERVIEW OF THE WATER SECTOR........................................................................................................................ 9 2. REVIEW OF WATER SECTOR SPENDING ................................................................................................................ 10 3. ASSESSMENT OF FUNDING FLOWS ........................................................................................................................ 11 4. EFFICIENCY OF FUNDING ...................................................................................................................................... 12 4.1 ALLOCATIVE EFFICIENCY ..................................................................................................................................... 12 4.2 BUDGET EFFICIENCY ............................................................................................................................................ 13 4.3 EFFICIENCY OF INVESTMENTS .............................................................................................................................. 14 5. FINANCING FUTURE WATER INVESTMENTS .......................................................................................................... 14 1. INTRODUCTION .................................................................................................................................................... 16 2. METHODOLOGY ................................................................................................................................................... 17 3. SECTOR BACKGROUND ...................................................................................................................................... 19 4. INSTITUTIONAL CONTEXT AND LEGAL FRAMEWORK ............................................................................. 21 4.1 INSTITUTIONAL CONTEXT AND LEGAL FRAMEWORK ........................................................................................... 21 4.2 BUDGET CYCLE ..................................................................................................................................................... 26 5. WATER SUPPLY AND SANITATION SECTOR PERFORMANCE ................................................................... 28 5.1 ACCESS TO IMPROVED WATER SOURCES.............................................................................................................. 28 5.2 PERFORMANCE OF WATER SUPPLY SECTOR ......................................................................................................... 34 5.2.1 Rural water supply sector performance ...................................................................................................... 34 5.2.2 Urban water and sanitation performance .................................................................................................... 37 5.2.2.1 Performance of the Urban Water Sector using International Benchmarks ....................................... 37 5.2.2.2 Sustainability of Urban Water Utilities............................................................................................. 38 5.3 SANITATION PERFORMANCE ................................................................................................................................. 43 6. WHAT IS BEING SPENT ON WATER SUPPLY AND SANITATION? ................................................................ 46 6.1 INTRODUCTION ..................................................................................................................................................... 46 6.2 TOTAL FLOW OF FUNDS IN THE SECTOR ............................................................................................................... 49 6.3 FLOWS BY INSTITUTION ........................................................................................................................................ 52 6.4 EXPENDITURE BY SUB-SECTOR ............................................................................................................................. 52 6.5 ALLOCATION ACROSS EXPENSE CATEGORIES ...................................................................................................... 54 6.6 SPATIAL ALLOCATION .......................................................................................................................................... 56 6.7 WATER AGENCIES‘ OWN REVENUES .................................................................................................................... 58 6.8 DONOR FUNDING IN THE SECTOR ........................................................................................................................... 59 7. HOW EFFICIENT IS SPEND ..................................................................................................................................... 61 7.1 ALLOCATIVE EFFICIENCY ..................................................................................................................................... 61 7.2.1 Budget Efficiency....................................................................................................................................... 63 7.2.1.1 Budget Procedures ............................................................................................................................ 67 7.2.1.2 Investment Planning ......................................................................................................................... 68 7.2.1.3 Procurement ...................................................................................................................................... 70 7.2.1.4 Audit Processes................................................................................................................................ 71 7.2.1.5 The Value Added Tax ...................................................................................................................... 72 7.2.1.6 Donor Fragmentation ...................................................................................................................... 74 7.2.1.7 Project Fragmentation ....................................................................................................................... 76 7.2.2 Efficiency of Investments ........................................................................................................................... 77 7.2.2.1 Efficiency in Rural Water Supply and Sanitation ............................................................................ 78 ii 7.2.2.2 Efficiency of Investments in Urban Areas ...................................................................................... 79 REFERENCES .................................................................................................................................................................. 84 BASIC DATA .................................................................................................................................................................... 86 DATA ANNEX ................................................................................................................................................................. 87 List of Tables Table 4.1 Organizations and their Roles and Responsibilities in the Water Sector ........................................................... 25 Table 5.1 Use of improved water source by income quintile ............................................................................................. 33 Table 5.2 Distribution of time to the nearest drinking water source by income quintile, 2008 ......................................... 34 Table 5.3 Use of Improved Water Sources in Rural Areas between 1990-2006 ................................................................ 34 Table 5.4 Annual Production of Water Points between 2003 and 2008 ............................................................................. 35 Table 5.5 Use of Improved Water Sources in Urban Areas between 1990-2006............................................................... 37 Table 5.6 Operating Cost Coverage Ratios for FIPAG, 2000-2008 ................................................................................... 38 Table 6.1 Annual Required Investment Needs to Meet MDGs in Water Supply and Sanitation ....................................... 50 Table 6.2 Own Revenue at Select Public Water Entities, 2000-2008 (MZN millions) ...................................................... 58 Table 6.3 Bilateral Aid data from various sources (US$ million) ...................................................................................... 60 Table 6.4 Execution of projects from non-OECD donor countries .................................................................................... 60 Table 7.1 Performance Budget Execution Ratios show wide variations on an annual basis.............................................. 64 Table 7.2 Accumulated Arrears in VAT and Counterpart Government Contributions up to end-2006 ............................. 72 Table 7.3 Number of Donors and Institutions benefiting from donor funding to the water and sanitation sector (2003- 2008) .................................................................................................................................................................................. 75 Table 7.4 Unit Costs Comparison of a 43m borehole in Mozambique and South Africa .................................................. 79 List of Figures Figure 3.1. Mozambique: Real GDP Growth Rates and Rainfall (1988-2002).................................................................. 20 Figure 4.1 Flow of Funds in Water Supply and Sanitation Sub-Sector.............................................................................. 22 Figure 4.2 Delegated Management Framework ................................................................................................................. 23 Figure 4.3 Flow of Funds in WRM Sub-Sector ................................................................................................................. 24 Figure 5.1 Use of Improved Water Sources in Mozambique, various years ...................................................................... 28 Figure 5.2 Use of Improved Water Sources shows large geographical variation............................................................... 30 Figure 5.3 Change in Use of Improved Water Sources between 2000 and 2008 by Province ........................................... 30 Figure 5.4 Use of Improved Water Supply Services by Consumption, Quintiles in 2008 ................................................. 32 Figure 5.5 Distribution of water points in a province in Mozambique, 2004 .................................................................... 36 Figure 5.6 Operating Cost Coverage Ratio for 13 utilities in 2008 ................................................................................... 39 Figure 5.7 Non-Revenue Water (%) in the utilities under FIPAG in 2009 ........................................................................ 40 Figure 5.8 Staff Ratio per 1000 connections in 13 utilities in Mozambique in 2004 and 2008 ......................................... 41 Figure 5.9 Tariff Levels in 13 utilities in Mozambique in 2000 and 2008 ........................................................................ 42 Figure 5.10 Access to Improved Sanitation 2000/01 and 2008 .......................................................................................... 43 Figure 5.11 Sanitation facilities in urban and rural areas ................................................................................................... 44 Figure 5.12 Households with access to basic sanitation by type of service by region in 2008 .......................................... 44 Figure 5.13 Access to Sanitation by Income Quintile 2008 ............................................................................................... 45 Figure 5.14 Percentage of households with no access to sanitation service between 2002/03 and 2008 ........................... 45 Figure 6.1 Millennium Challenge Account & Other Donor Commitments in ODAmoz Database, 2007 (US$millions) . 47 Figure 6.2 Increase in Actual Nominal Water Expenditure between 2002 and 2008 (MZN millions) .............................. 49 Figure 6.3 Evolution of Expenditure as % of GDP for Selected Priority Sectors .............................................................. 50 Figure 6.4 Annual Real Change in Expenditure for Selected Priority Sectors ................................................................... 51 Figure 6.5 Total Nominal Expenditure in the Sector by Institution (‗000 MZN) .............................................................. 52 Figure 6.6 Total Nominal Expenditure by Sub-sector (‗000 MZN) ................................................................................... 53 Figure 6.7 Composition of Nominal Public Expenditure in the Water and Sanitation Sector ........................................... 55 Figure 6.8 Donor Commitments to Water Subsectors, 2005-2007 (US$).......................................................................... 56 Figure 6.9 Urban and Rural per Capita Water Expenditure, 2007 (in MZN) ..................................................................... 57 iii Figure 6.10 Nominal External Funding by Recipient Institutions (‗000 MZN) ................................................................. 59 Figure 7.1 MDG Targets for Water and Sanitation and Sector Expenditure in Various Years .......................................... 61 Figure 7.2 Donor funding – cumulative actual commitments and cumulative actual disbursements (In MlnUS$) ........... 65 Figure 7.3 Actual Disbursements on a Quarterly Basis in US$ million ............................................................................. 66 Figure 7.4 ARA-Sul‘s Allocations of Domestic Funding (‗000 MZN) ............................................................................. 68 Figure 7.5 Planning and Control Instruments in Mozambique‘s Water Sector .................................................................. 70 Figure 7.6 Number of water sector projects by financial size as measured in US$ (as of May 2009) ............................... 74 Figure 7.7 Water Spending in Real MZN million .............................................................................................................. 77 Figure 7.8 Hidden costs in Urban Water Utilities (excluding Maputo) against Best Practice in Mozambique ................. 82 Figure 7.9 Hidden costs in Maputo against Best Practice in Mozambique (2006) ............................................................ 82 iv WAY FORWARD: SUMMARY OF RECOMMENDATIONS A. Introduction Reform is definitely under way in the water sector. Mozambique‘s goals and macro strategies are laid out in the Poverty Reduction Action Plan (PARPA), which is the main reference for water supply and sanitation policies. A water policy was established in 2007, which sets national targets for water supply and sanitation coverage, specifically to achieve the Millennium Development Goals (MDG). The sector also has a National Strategy for the Management of Water Resources. In addition, the different water subsectors have generated a large set of additional plans and strategies, including Visao 2025, Plano Quinquenal do Governo, the Economic and Social Plan (Plano Economico e Social, PES), the Medium-Term Fiscal Framework (MTFF) with specific proposals for each institution/agency in the sector, and the Economic and Social Plan for Rural Agriculture (Programa de Agua Rural, PESA-rural). These planning documents at the provincial and local levels require strong coordination to ensure that the sector uses an integrated approach that balances the interests of the different sub-sectors with that of the overall sector. Yet, currently the sector still lacks a strong coordination mechanism. An increasing share of the country’s capital budget is allocated to fund water infrastructure, especially water supply infrastructure. Between 2002 and 2008, actual expenditure in the water sector (excluding the own-generated funding) grew from MZN 636 million in 2002 to MZN 2.3 billion – an increase of 255 percent. Donors supply the bulk of capital funding to the sector, and have committed more funds to the water sector than ever before. Donor coordination has been institutionalized with a process of joint reviews, but this enhanced donor coordination has not yet resulted in a harmonized approach in the allocation, procurement, disbursement and reporting on donor funds. Water Expenditure as % of GDP Use of Improved Water Sources The increased funding has produced many outputs. Major water works, pipelines, reservoirs, and new connections have been constructed. In the case of Maputo, for instance, the number of piped water connections increased by 40 percent between 2003 and 2008. The number of rural water access points increased by 45 percent over the same period. 1 But the translation from outputs into outcomes, as measured by an increased in access to improved services, has been slow. Preliminary data from household surveys conducted in 2008 and preliminary data from the 2007 Census show that between 1997 and 2008, an additional 3 million people used improved water sources – of which 1.7 million live in urban areas and the remainder in rural areas. Yet, at the same time, the percentage of the population which uses improved water sources has only increased marginally in recent years. This is partially the result of rapid urbanization, which is making it difficult to keep pace with urban population growth. The gap between increased funding and outcome improvements can be largely explained by a lack of efficiency. In urban utilities, inefficiencies in the provision of services results in dependence on government subsidies and backlogs in proper maintenance, which crowd out funding for new priorities in the water sector. This is especially important, because rapid urbanization creates additional pressure to meet MDG targets. In the rural areas, the high breakdown rate of rural water points also indicates inefficiencies in the delivery of these services. The overall effect is slow progress in achieving the MDG targets. The sector deals with a number of issues that explain why the goal of improving access to water services is moving more slowly than expected. Some of these issues are largely outside the control of the water sector, but they do influence progress in the sector. Other issues are within the direct control of the water sector and need to be addressed to ensure that the sector can move more quickly to achieve its goals. B. Sector-specific issues: Sector-specific issues play a major role in explaining the sector’s performance in translating funds into actual outcomes. The sector currently requires significant subsidies for rehabilitation, crowding out investments for expansion and as such delaying improvements in access to water services. Improving the efficiency with which resources are used would do much to improve the absorptive capacity of the sector, increase the efficiency of public expenditure, and ensure improvements in access to and use of services. To that effect, the following measures are suggested: (i) establish a strong sector coordination mechanism; (ii) improve sector investment planning; (iii) improve the capacity in the sector to conduct procurement and disbursement; (iv) put a sharper focus on including incentives in funding allocation; (v) promote efficiency gains that lower the operating cost and/or capital costs; and (vi) promote sustainability in tariff setting while safeguarding the affordability of access. With many stakeholders operating in the sector, the sector will need to address a set of issues that with each of the different stakeholders undertaking actions to move the sector forward. B1. Strong Sector Coordination A large number of different agencies are operating in the sector. The National Water Directorate (DNA) as the lead agency is surrounded by a set of agencies that are each undertaking specific tasks with different levels of autonomy. These agencies include: (i) the Regional Water Basin Agencies (Administraçoes Regionais de Aguas, ARAs), which have responsibility for water resources management; (ii) the Water Assets and Investment Fund (Fundo de Investimentos e 2 Patriônio de Abastecimento de Água, FIPAG) which has responsibility for urban water through a delegated management framework; (iii) the Council for the Regulation of Water Supply Services (Conselho de Regulação do Abastecimento de Água, CRA) in charge of regulation in the water and sanitation sector; and (iv) more recently, the Water and Sanitation Infrastructure Management Agency (AIAS), which is to operate water supply and sanitation systems in urban areas outside the Delegated Management Framework (DMF) through provincial water and sanitation boards to be set up by the individual provincial governments. These various agencies are matched with a large set of strategies that exist or are being developed for Mozambique‘s water sector with each agency preparing its own strategy. Although DNA is the lead agency, it does not yet lead the sector in overall policy setting, sectoral planning, and monitoring. Policy setting and coordination of sector planning at the overall sector level will also enable more explicit trade-offs between the various sub- sectors. Decentralization of the sector will make this need for coordination in sector planning and budgeting even more important due to the large increase in the number of stakeholders in the sector. To enable DNA to function effectively as the lead sector agency entrusted with coordinating overall sector planning and budgeting, the agency will likely require reorganization. B2. Improve Sector Investment Planning The sector is in need of an overall investment strategy that enables priority setting not only for each sub-sector or agency separately, but for all sub-sectors combined in accordance with national and sector objectives. o The government should prepare a long-term1 investment strategy for the water sector that includes all sub-sectors and sets up a procedure to prioritize investments while ensuring their cost effectiveness. This way, the sector can build a robust pipeline of projects that provides the best value for money and will guide donor-funding priorities. o To enable the definition of a long-term investment strategy, the sector should have access to a uniform mechanism for screening investments. To that effect, it is necessary to build capacity in the sector to ensure that staff can adequately and transparently screen investments both at the agency level and at the overall sector level. o The overall sector plan (covering all sub-sectors) should be directly linked to the budget. The Medium-Term Fiscal Framework (MTFF) is the direct point of intervention, but to date, the annual deviations from the plans show that as a tool, the multi-year resource envelope does not really function very well. This is partly because the level of detail in the MTFF is so minimal that it does not allow for strategic decision-making. o The costs of maintenance of water infrastructure should be considered as an integral part of overall investment decisions. The medium-term costs play little role in the formulation of the annual budget. This is an issue in a sector where service delivery requires not only investments but also subsequent operation and maintenance of these assets over time. The lack of attention to the operation and maintenance cost consequences of investments results in scarce resources being used to pay for these costs – either through government subsidies or through a lack of 1 Long-term refers to an investment plan that covers the next 5 -10 years. 3 maintenance and subsequent reduction in the economic life of infrastructure, or a combination of both. The result is that most funding in such a setup will cater to existing consumers and ultimately lead to increased delays in access and use of services for those not yet served. B.3 Improve the predictability of timeliness and amount of funds to be released The lack of predictability of funds is a serious obstacle and a major cause of delays in implementation and additional costs. o The different sector agencies should strengthen their procurement, disbursement, and auditing functions in such a way that: (i) improves the link between procurement and disbursement planning and budget preparation; and (ii) provides training to their staff to ensure that they are proficient in procurement. B.4 A sharper focus on including incentives in the allocation of funding In the design of infrastructure programs, the Government should put more focus on how to improve the efficiency of its investment programs through: (i) performance-based transfers; and (ii) better poverty targeting. Such mechanisms can only be contemplated if there is a monitoring system that can measure the performance of the sector (including access, quality, and sustainability of services). o Accountability and achieving efficiency in sector performance requires well-functioning systems for monitoring and evaluation. This means building reporting systems that measure the efficacy and efficiency of such programs in achieving measurable outcomes in terms of access, quality, and sustainability of services. Evaluations could provide valuable information on what works and what does not and provide information for the design of future programs. CRA has initiated this process for a set of utilities, but it is important to link the utility data with that of census and household survey data, while similar initiatives have to be undertaken for other subsectors. o Introduce, where possible, more performance- based incentives in the transfer of funds. A very large part of the sector‘s funding is being transferred to urban utilities while the performance of these utilities still hold ample room for improvements which can reduce the cost of the service and subsequently reduce the need for large tariff increases. It would be useful to provide these funds to these utilities in a way that links investments, especially expansion investments, to elements of their performance that are under the utilities‘ control. An example would be the provision of expansion investments only when certain levels of performance are achieved or the use of outputs-based disbursement schemes. In rural service provision, a least subsidy approach could be used. B.5 The efficiency of investments can be significantly improved upon The volume of subsidies that are still provided to the sector can be used as a proxy for investment efficiency. The high breakdown rate of rural water supply infrastructure requires the sector to use a significant amount of the funding for rehabilitating the infrastructure that is not working anymore. As an important part of the funding is hence assigned for rehabilitation, expansion of services to areas and populations not yet served gets delayed. Performance data of urban utilities show that 4 even though operating subsidies are declining, the hidden subsidies to urban water authorities (in the form of financial and operational inefficiencies) are still very large. The following measures could be taken to improve the efficiency of investments: o Technology Choice. Different technology choices can make a significant difference in total investment requirements2. The cost of supplying water and sanitation services varies widely by the level of service provided, especially in rural areas because of lower population densities and transport costs. In view of such large cost differentials, and the fact that high service levels are much more likely to be used by wealthier consumers, there is a rationale for providing a minimum service level to consumers whereby households themselves finance higher levels of services. o In rural water supply, improve planning of the location of water points. Water points are often built in clusters, which results in some villages having multiple water points while others having little or no access. Further, the location of water points is often correlated to the presence of roads. The logistical challenges of drilling boreholes in rural areas with no proximity to a road add to inequalities in rural access, so that while many new water points may be built in a given area, the gap in access remains. In order to better match investments with needs, improved planning of new rural water points is needed. o Public investment funding to urban utilities should be linked to performance improvements. The current financing system should be made contingent on improvements in the financial and operational performance of water utilities. Utilities that are not capable of covering basic operation and maintenance costs should be provided with support to improve their operational performance (possibly through a capacity building program). Where a utility‘s revenues exceed its operation and maintenance costs, the utility would gain access to funds for rehabilitation. Only once a utility has proved its ability to cover basic operations and maintenance costs and funds for depreciation (to fund rehabilitation of existing infrastructure) will it graduate to gain access for expansion investment. As is shown in the hidden cost analysis, there is ample room to improve the performance of utilities by further reducing over- employment and increasing collection and billing efficiencies3. o The sheer size of the water utility of Maputo will require that the sector pay attention to this utility because improvements in Maputo are likely to affect the overall urban sector performance significantly. This utility accounts for a significant part of all water sold by urban utilities, and has much room to improve its performance as the size of its hidden subsidies shows that it is inefficient in its use of resources. 2 The Africa Infrastructure Country Diagnostic Study estimates that a minimum (or zero) level scenario to meet the MDGs would cost an annual $333 million, while the most expensive scenario would cost an annual $523 million. These estimates only include water supply and sanitation investments. 3 Reduction in non-revenue water may require upfront investments and hence may not be the first efficiency improvement to focus on. 5 B.6 Promote sustainability in tariff setting while safeguarding the affordability of access The Government has set up different cost recovery policies for the different water services. Urban water authorities are required to achieve full cost recovery. In the case of rural water supply, communities are required to pay the full operation and maintenance costs (and costs for any service levels higher than the standard), while managing their own schemes. o There is some space for tariff increases, especially in urban areas, where richer consumers disproportionally use higher levels of piped water services. All house and yard connections are used by the 40 percent of the country‘s richest households, showing that individual piped connections constitute a service used by richer consumers. o Nevertheless, affordability remains a critical issue when assessing changes in tariff levels and structures. It is possible to include more discriminatory tariff policies whereby users of lower levels of service pay less than those who use better quality services. Evidence from a recent study on water and electricity subsidies4 found that subsidies on the basis of self-selection of service levels tend to be more pro-poor than the more common consumption- or connection- based subsidies. o However, before making any changes to the tariffs, utilities should first address their low billing and collection efficiencies. Fixing gaps in the utilities‘ billing and collection systems will generate additional revenue, and reduce the pressure to increase tariffs to ensure the sustainability of the water supply service. o There are limits to the level of cross-subsidization in the tariff systems approved in December 2009. The tariff structure that utilities use is heavily skewed towards non-residential users resulting in large cross-subsidies in the tariff structure. Although cross-subsidies have their advantages, the problem with very extensive cross-subsidies is that they provide non- residential users with an incentive to find alternative sources. This is already happening in the largest five utilities where the number of non-residential connections declined between 2002 and 2007 despite rapid economic growth in the country during that period. C. Systemic issues The Government has to address a series of systemic issues, most of which are outside the control of the water sector, in order to increase the release of funds allocated to the water sector. They are mainly related to: (i) improving budget procedures; (ii) decentralizing public works allocations; (iii) the need to harmonize procurement, disbursement, and monitoring procedures; (iv) the need to find a more sustainable solution for counterpart and VAT funding; and (v) improving the predictability and reliability of donor funding to the sector. 4 Komives,K, et al.,2005. Water, Electricity and the Poor: who benefits from utility subsidies? World Bank, Washington D.C. 6 o Reflect the priority status of the sector by increasing domestic and decentralized funding to the water sector. The priority status of the water sector should be reflected in its domestic funding allocation. The high dependence on donor funding leaves the sector vulnerable because a significant part of the recurrent funding, especially in DNA, is externally funded. Under the existing system, allocations for district level spending cannot be transferred through sectoral budgets. They can be transferred to provinces, as these are deconcentrated rather than decentralized branches of government, but onward transfer from the provinces is essentially not possible. o Improve the use of the Medium-Term Fiscal Framework as a reference point for longer-term planning and budgeting. Thus far, the annual deviations from the MTFF show that the multi- year approach to planning and budgeting needs improvement. To make this tool more effective at the sectoral level, the sector will need to improve overall coordination of planning and budgeting, so that it can submit a detailed plan and budget to be included in the MTFF. o Government should address the remaining deficiencies in the legal and regulatory framework for procurement. Despite significant improvements in the legal and regulatory framework for public procurement, including a new legal framework introduced in 2005, several weaknesses remain, which are documented in a recent Country Procurement Assessment Review 5. These include onerous requirements for foreign bidders and a lack of clarity in some areas of the bidding process, as well as a lack of transparency, in particular through the absence of an independent forum through which to appeal the decision of the procurement entity. o Government and donors should also agree on a harmonization of procurement, disbursement, and monitoring procedures to reduce the number of parallel systems in place that constrain already scarce capacity. As is demonstrated by the regular assessments of the donor group, progress in harmonization is still very limited. Harmonization of procurement, disbursement and monitoring requirements could also help to ensure economies of scale. The proposed Rural Water Supply and Sanitation Common Fund may be a place to start these donor harmonization efforts. o Government and donors should come to an agreement on how to deal with the counterpart and VAT issue, as the current method of partial solutions is not sustainable. Better planning at the sector agency, the Budget Directorate, and by donor agencies is required to ensure that the predictability of funding improves significantly. It will also require a realistic assessment of absorptive capacity in the sector. If the large levels of donor funding cannot be adequately matched by domestic resources to fund VAT and counterpart funding, an assessment should be made whether to relax the need for domestic resources or reduce the levels of donor funding that correspond with the availability of domestic resources. Donor agency assumption of the project‘s full tax burden can go a long way to redress further accumulation of arrears in new infrastructure projects. Such a measure would accelerate implementation, avoid special rules and their potential for abuse, and also reduce the cost of water sector infrastructure. 5 Republic of Mozambique and The World Bank Joint Report. Update of the Country Procurement Assessment Review (CPAR) Draft Report. June 2008. 7 o Government and donors should agree on a code of conduct6 with regard to the fragmentation of donor funds, requiring that: (i) all donor funding to the water sector be communicated to the Government (Ministry of Finance) to ensure a better understanding of what funding is actually available in the sector; (ii) Government set a minimum threshold of at least US$5 million for donor funding commitments to reduce the transaction costs associated with management and operation of a large set of very small projects (in May 2009, 29 of the 50 projects identified in the sector had a value of less than $5 million); and (iii) any donor funding that does not meet the minimum threshold will not be accepted as an independent project, but pooled with other funds in, for instance, the rural Sector-Wide Approach (SWAp). o Donors should be improving the predictability and reliability of their funding to the sector. Although part of the unpredictability of donor funding is linked to procedures and capacity bottlenecks, another part of the unpredictability is linked to a set of issues that donors can improve upon. The use of parallel systems is one of the issues already mentioned. A second relates to administrative bottlenecks in donor countries. A survey of aid donors mentioned that 29 percent of delayed or lost disbursements were due to administrative problems in donor countries. A third way in which donors can improve the predictability and reliability of funds is to register all their funding on-budget rather than putting some on- and some off-budget. These inefficiencies are further exacerbated by the fragmentation of donor and project funding, which results in high transaction costs for donors and government alike. 6 This code of conduct is different from the one that the Government of Mozambique has agreed with a number of donors regarding the Sector Wide Approach for the Rural Water and Sanitation Sector. 8 EXECUTIVE SUMMARY Reform is definitely under way in the water sector. Mozambique‘s goals and macro strategies are laid out in the Poverty Reduction Action Plan (PARPA) which is the main reference for water supply and sanitation policies. A water policy was established in 2007 which sets national targets for water supply and sanitation coverage, specifically, to achieve the Millennium Development Goals. The sector also has a National Strategy for the Management of Water Resources. In addition, the different water subsectors have generated a large set of additional plans and strategies. The concurrence of these planning documents (including Visao 2025, Plano Quinquenal do Governo, the Economic and Social Plan (Plano Economico e Social, PES), the Medium-Term Fiscal Framework (MTFF) with specific proposals for each institution/agency in the sector, and the Economic and Social Plan for Rural Agriculture (Programa de Agua Rural, PESA-rural). The concurrence of these planning documents at the provincial and local levels requires strong coordination to ensure that the sector uses an integrated approach that balances the interests of the different sub-sectors with that of the overall sector. Yet, currently the sector still lacks a strong coordination mechanism. As a result of these policy changes, the Government‘s actual expenditure for the sector increased from $27 million in 2002 to $95 million in 2008; this is unlikely to be the total of funding available in the sector, but what could be substantiated from the government accounts. Donors supply most of funding in the form of investment funding. This Public Expenditure Review focuses on the quantity and quality of Mozambique‘s public funding process to assess how well budget allocations for the water sector actually translate into better water and sanitation service delivery. 1. Overview of the Water Sector While Mozambique is not considered to be water scarce, it is highly water vulnerable. Water resources management is constrained by increasing uncertainty in the national water resources base. In the area of water supply and sanitation, policy reforms anchored in the National Water Policy and National Water Development Program of 1995 have helped progress in the sector but gaps remain. Key water-related challenges for Mozambique can be summarized as follows:  High dependence on shared water resources, and vulnerability to hydrological variability and water shocks. Over half of Mozambique‘s water resources originate outside the country. This high dependence on shared water resources is an important factor affecting national water vulnerability. Development and use of the country‘s water resources is also affected by the highly variable climate with large variations in annual and inter-annual rainfall, and frequent flood and drought events. Floods and droughts have a significant impact on the country‘s economic performance. The poor, who depend largely on subsistence agriculture, are often the most vulnerable to these shocks. The uneven geographic distribution of water resources across the country and growing competing future demands by water–dependent economic sectors in many river basins represent serious challenges to long-term growth. As an interesting comparison, an analysis undertaken over the period 1981-2004 suggests that growth in Gross Domestic Product (GDP) is cut by 5.6 percent on average when a major water shock occurs. In this analysis, the total costs of water shocks over the period under review were estimated at about US$1.75 billion. 9  Gaps in provision of water and sanitation services remain an issue. Access to improved water and sanitation services is still low, while the reliability and quality7 of water services still need improvement. Despite the recent reforms which have helped to strengthen sector institutions and the regulatory framework, improvements in the use of water supply services is showing only modest improvements, and gaps remain in the quality of rural and urban water services. In the area of sanitation, progress has been made but much work still remains to be done to ensure that the adverse impacts from the lack of improved sanitation can be avoided. 2. Review of Water Sector Spending Water spending has increased rapidly since 2002. Between 2002 and 2008, actual expenditure in the water sector (excluding the own-generated funding) grew from MZN 636 million in 2002 to MZN 2.3 billion – an increase of 255 percent. In real terms, the growth rate was less spectacular, but the expenditure on water still increased by a still impressive 140 percent. Current levels of spending are increasing as a percentage of GDP, but remain low in absolute terms. The Africa Infrastructure Country Diagnostic (AICD) found that central government expenditures for water supply and sanitation ranged from 0.29 to 0.83 percent in Eastern African countries. Compared to these numbers, the total government expenditures (including the allocations to provincial directorates and districts through the ‗ambito central‘) added up to 1 percent in 2008 – slightly above the numbers found by the AICD study. This is unlikely to cover all the funding in the sector, as ―off-budget‖ funding is likely to be substantial as anecdotal evidence shows, while a new generation of donors is becoming active in the sector (China, India, South Korea, and South Africa among them). This is well below the 5 percent of GDP needed to meet the MDGs according to the AICD base scenario. In addition, funding tends to be very volatile, with large annual changes in flows between years. Improvements in access to improved water sources have been slow. Although some progress has been made since 2002, progress has been slow. Rapid urbanization rates have resulted in an absolute increase in the urban population served, but the share of urban dwellers using improved water sources has increasing very slowly since 2002. In rural areas, where population growth has been slow, the use of improved sources has also shown slow growth rates. In addition, the quality of service (as measured by the level of service provided) is declining. In urban areas, piped water provided through house and yard connections or standpipes decreased from 51 to 47 percent between 2000 and 2008. Use of protected groundwater declined from 15 to 9 percent, while the use of non- protected groundwater increased from 16 percent in 2000 to 27 percent in 2008. Dependence on other sources (often including water from neighbors – which can and often does include piped water) increased from 17 to 20 percent. In rural areas, a similar trend is noticeable – with an increasing share of households depending on groundwater sources. The use of unprotected groundwater sources is especially prevalent among the poorest households; hence the burden of a deficient water service is mainly affecting the poor. 7 World Bank, 2007. Project Appraisal Document on a Proposed Credit to the Republic of Mozambique for a Water Services and Institutional Support Project. Washington D.C.: World Bank. 10 Use of Improved Water Sources in Mozambique, various years Source: Household Surveys and Census Data, various years 3. Assessment of Funding Flows The water sector budget is almost entirely comprised of development expenditures. In 2008, development expenditures constituted about 86 percent of water sector total expenditures. Recurrent expenditure does not yet make up 14 percent of total expenditures. Most of this development expenditure is provided through donors, which in 2008 funded almost 88 percent of total development expenditures. The vast majority of donor funds are directly linked to projects. Total domestic resources dedicated to the sector (recurrent plus domestic investment) decreased in 2008, reversing a positive trend between 2004 and 2007; this trend is incompatible, however, with the status of the water sector as priority sector. The reduction in domestic resources has especially affected the recurrent budget which in 2008 was lower than in 2006 despite the larger than ever investment volume going into the sector. This lack of local funding for recurrent expenditure has serious implications. The first is an extreme dependence of water sector agencies on external funding and the subsequent vulnerability of its core functions and the delivery of water and sanitation services to potential cuts in external aid. A second implication is that the funding for operations and maintenance of water sector agencies becomes a major issue once donor funding is fully allocated, and the sustainability of services can be compromised. 11 Composition of Nominal Public Expenditure in the Water and Sanitation Sector by Economic Category (000s MZN) Source: data from water sector institutions and authors’ calculations 4. Efficiency of Funding 4.1 Allocative Efficiency Allocative efficiency measures whether sector expenditures are going toward realizing the national objectives and MDGs as set out in the national objectives. The indicators used to determine allocative efficiency are consistent with the basic objectives set out in the sector strategy. Progress in achieving these objectives has been slow. The National Water Policy of 1995, the subsequent Water Policy of 2007 in the context of the MDGs, and the experiences with sector reform include the following objectives: (i) achievement of the MDGs in the rural and urban water and sanitation sector; (ii) consolidation of the process of sectoral decentralization; (iii) improve the sustainability of urban water systems in the long run (through the use of a minimum level of service in water and sanitation access and a focus on cost recovery in tariff setting, among other measures); and (iv) adoption of adequate hygiene practices in households, communities, and schools. Achieving the MDGs is an ambitious goal that is unlikely to be met under current scenarios. While access to improved water and sanitation is increasing as was shown in Chapter 4, there is a gap between the current trend and what will need to achieved by 2015, according to the Joint Monitoring Program. While sector expenditure, though volatile, has seen an upward trend in the last five years, it is unlikely that the current levels of investment in the sector will be sufficient to meet the sector objectives of achieving the MDGs. 12 4.2 Budget Efficiency Despite increases in the past years in the total water sector budget, actual expenditures fall below the budgeted estimates. Although actual sectoral expenditures have increased rapidly, they have generally lagged behind budget allocations and have had a rather volatile performance in their execution. In the past four years, the execution rate of sector expenditures that are registered on- budget has vacillated significantly between and among the different sector agencies. Delays in budget implementation are mainly concentrated in executing the development budget. Development expenditures tend to be used with more discretion and are the first to be affected when budget cuts need to be made. In addition, because of the high donor dependency in the development budget, budget implementation is largely determined by how efficiently external funds are being spent. Budget execution rates are lowest and most volatile for donor-funded development expenditures. During the last three years, on average only 40 percent of the external development funding budgeted was disbursed (according to the Overseas Development Assistance Mozambique Database, ODAMOZ), whereas this disbursement ratio fluctuated widely over that period. The high unpredictability of recurrent and development funds can provide perverse incentives to the sector. The succession of higher- and lower-than expected allocations renders strategic and operational planning ineffective. In addition, because the predictability of funds affects all types of expenditures, and there is no guarantee that recurrent costs will be covered and agencies in the sector have an incentive to divert funds from their development budgets to supplement their recurrent budgets. Under this dynamic, the first and foremost preoccupation of implementing agencies might not be service delivery per se but securing the financial viability of the agency in the short-run. The gaps between budget allocations and actual expenditure are caused by a combination of factors. As will be discussed in the remainder of this chapter, the divergence between budgeted and actual expenditure is due to a combination of issues in the budgeting, planning, procurement, and disbursement processes. The above-mentioned inefficiencies are exacerbated by the high levels of donor fragmentation. Between 2003 and 2008, the sector institutions benefited from at least 78 projects. There were 24 different donors active in the sector (excluding Non-Governmental Organizations (NGOs)). DNA had to deal with the largest number of donors – 19 different donors in 2008. Yet, with the decentralization process starting, the number of sector agencies with funding has increased. Almost all of these donors use project funding as their main vehicle. Another indicator of donor fragmentation is the small size of projects. The ODAMOZ database shows that the average project size was slightly over US$4 million; hence, the sector deals with a large number of very small projects which adds to the transaction costs. This large fragmentation also places a very heavy administrative burden on a sector with serious capacity constraints, thus shifting the focus of the sector‘s central institution (DNA) away from sector-wide functions such as coordinating, planning, and monitoring and enforcement (M&E), into project management. 13 4.3 Efficiency of Investments We will measure the efficiency of investments using two indicators: (i) the relationship between the volume of expenditures and the improvements in access; and (ii) the subsidies channeled through the sector. These indicators are only indicative as the data is incomplete. The link between the expenditures and improvements in use of improved sources has been weak. There are many reasons for this performance. An increase in funding usually does not translate into a rapid increase in use of water supply services given the long preparation and implementation periods of many projects. In the case of Mozambique, matters are further complicated by the fact that much of the investment is focused on rehabilitation of existing infrastructure (which essentially crowds out expansion of access, focusing instead on keeping existing infrastructure in working condition) whereas high levels of funds are allocated for capacity building and only affect expansion of infrastructure indirectly and in the long term. Efficiency in rural water supply investments is low. One explanatory factor for the less than optimal efficiency is the high breakdown rates of rural water points. In view of a lifetime of assets in the rural water sector (i.e., 7 years for an individual water point and 25 years for a piped water system), breakdowns of between 4 percent and 14 percent should be considered a benchmark for acceptable performance depending on the mix of individual water points and piped water systems used in rural areas. Obviously, this figure is likely to be significantly below the current level of non- functionality of at least 33 percent. Secondly, the efficiency in rural water supply investments is adversely affected by the high cost of investment. Inflated contract prices result from insufficient contractor competition, a multiplicity of donor-specific tendering and procurement systems, accumulation of tax arrears, slow counterpart funding, and a general lack of accountability of expenditure. The investment costs are further inflated by the market structure (with only a limited number of drilling companies active) and at times challenging hydrogeological conditions In the urban areas, efficiency is improving, but more needs to be done. A first indicator of possible inefficiencies of urban water systems is the size of the subsidies provided through the urban water sector. The good news is that these operating subsidies have been declining between 2000 and 2008, as more urban water utilities started to cover basic operation and maintenance costs because of tariff increases and modest improvements in reducing inefficiencies in service delivery. Apart from the direct subsidies provided to the urban water sector, in the form of operating and capital subsidies, there are also implicit subsidies or hidden costs. These implicit subsidies are caused by the mispricing of water services, collection inefficiencies, overstaffing and non-revenue water losses. All these forms of inefficiencies in water supply delivery result in distortions. Tariff regimes that do not allow for full cost recovery and collection inefficiencies provide implicit subsidies to existing utility consumers. Non-revenue water losses are also implicit subsidies but do not necessarily benefit existing consumers only. All of these types of inefficiencies result in lack of maintenance, under-investment, and deteriorating service levels. 5. Financing Future Water Investments The Africa Infrastructure Country Diagnostic investment requirements far exceed actual funding for the sector. Though budget allocations have increased, actual expenditures have never surpassed one percent of GDP. Though household fees can cover the cost of operations and maintenance in urban 14 areas, the overall sector is faced with a significant investment gap. This is further aggravated by the high dependency on donor funding and the common practice of many donors to disburse funds with large delays. Given the high unit cost for connecting and servicing households, especially in rural areas, the level of resources and timing needed to meet the MDGs are highly sensitive to efficiency improvements. Efficiency can be realized in several different ways:  Sector-specific issues play a major role in explaining the performance of the sector in translating funds into actual outcomes. The sector currently requires significant subsidies for rehabilitation, crowding out the space for expansion investments and as such delaying improvements in access to water services. Improving the efficiency with which resources are used, would go a long way to improve the absorptive capacity of the sector, increase the efficiency of public expenditure, and ensure that access to and use of services is improved. To that effect, the following measures are suggested: (i) establish a strong sector coordination mechanism; (ii) improve sector investment planning; (iii) improve the procurement and disbursement capacity in the sector; (iv) put a sharper focus on including incentives in the allocation of funding; (v) promote efficiency gains that lower operating cost and/or capital costs; and (vi) promote sustainability in tariff setting while safeguarding the affordability of access.  The Government has to also address a series of systemic issues in order to increase the release of funds allocated to the water sector. Systemic issues are largely outside the control of the water sector. They are mainly related to: (i) improving budget procedures; (ii) decentralizing public works allocations; (iii) the need to harmonize procurement, disbursement, and monitoring procedures; (iv) the need to find a more sustainable solution for counterpart and VAT funding; and (v) improving the predictability and reliability of donor funding to the sector. 15 1. INTRODUCTION Improving access to and quality of water supply and sanitation (WSS) services is a key objective in poverty alleviation and a target of the Millennium Development Goals (MDGs). The achievement of the MDGs will, in many countries, including Mozambique, require a large investment program that will help increase access to safe and sustainable water and sanitation services. Yet, though additional resources may be urgently needed, research in other social sectors (health and education) has also shown that higher public expenditures do not necessarily result in better social outcomes.8 Gaps in achieving outcomes can be due to:  Sub-optimal spending, due to inefficient allocation of resources (delays and lack of predictability), discretionary reallocation of resources (lack of transparency in allocation of resources), inappropriate policies and institutional incentives, or poor targeting of resources;  Low quality of service delivery due to inefficiencies;  Lack of demand from certain segments of the population. A lot of effort has been dedicated to increasing resources to achieve the MDGs, and Mozambique has seen its resource allocation for the water sector increase significantly in the past few years. However, the size of the required investments can be reduced if the efficacy, efficiency, and quality of public expenditures in the water and sanitation sector can be increased. Looking into the efficiency of public expenditure programs in the water sector is complicated. More so than in other social sectors (health and education), the water sector is a subset of sectors that are often managed by different agencies within the country. In general, the water sector includes the provision of water supply, sanitation and irrigation services, but also water resource management and at times hydropower. In this analysis, we only have included the subsectors managed by the Ministry of Public Works and Housing (MoPH) that include water supply, some sanitation, and water resource management. All other subsectors have been excluded from the review. Due to differences in the availability of data, the focus of this report is tilted towards the water supply sub- sector, though information on sanitation and water resources is included wherever possible. Objectives of the Public Expenditure Review. The objective of this Public Expenditure Review (PER) is to assess the quantity and quality (as measured by allocative and technical efficiency9) of the transfers of public funds to the water sector. The study aims to gain insight into how budgeted allocations for the water sector translate into actual water service delivery, and to understand what affects the links between budget allocations and actual water service delivery. 8 See for instance: Lant Pritchett and Deon Filmer, The Impact of Public Spending on Health: Does Money Matter?". Social Science and Medicine 49(10): 1309-23, 1999. 9 Allocative efficiency is here defined as whether the sector expenditures are going toward realizing the national objectives and MDGs, as set out in its policy objectives. Technical efficiency measures whether the funding is spent efficiently given the allocative decisions. 16 2. METHODOLOGY The methodology aims to be comprehensive in the sense that it tries to cover relevant budgetary and non-budgetary areas of water sector spending. There are five sources of funding for on-budget expenditure in the water sector. The first are Treasury Funds, which include donor aid channeled through the Treasury (the so-called ASAS10 funds are provided to DNA by the Government of the Netherlands, the only donor to provide such funds to the water sector), and general tax revenue, used for both recurrent and investment expenditures. A second source of funding comes from donor-funded projects which appear on-budget, with quarterly execution rates reported to the sector institutions. External investment is almost entirely financed through these funds. In addition, there are smaller, donor-funded projects managed by the sector institutions (on-budget, off-treasury). A fourth source of funding is provincial, municipal, and district allocations that go directly to these respective levels of government and are used for many purposes including water expenditure. These allocations have been excluded from this analysis, as there is insufficient available information to determine what portion of the allocation is actually spent on water. A final source of on-budget funding is own-revenue generated through the collection of water-specific taxes (ARA-Sul, ARA-Centro and CRA) or concessions to private operators (FIPAG), which has been captured where data was available. Off-budget funding has also been considered in the analysis wherever it was possible to gain insight in these flows. Off-budget funding can take the form of capital investment subsidies and operation and maintenance subsidies to water utilities from sector entities, as well as funding from donors that is not registered in the government‘s budget. Sometimes this ―off-budget‖ funding is registered and its size is known, as it is registered in other information sources. Often, the actual size of off-budget funding is unknown because it bypasses sector agencies completely, as is the case for services funded by the many NGOs active in the water sector. Yet, these data are not consistently available to determine their actual impact on the sector, and hence the actual spending in the water sector is likely to be underestimated. The data collection process raised a number of methodological issues that were dealt with as carefully as possibly. First, data availability and quality dictate to a large extent what type of analysis of budget allocation and expenditure can be conducted. Second, special care was taken to ensure that expenses were analyzed and classified according to their economic use, either as capital or current expenditure (Box 1). Third, it was important to avoid double-counting of transfers from the central government to parastatals and sub-nationals by careful matching-up of the accounts. Finally, for the analysis of water service outcomes, the use of the terms ―access‖ and ―use‖ of water services was carefully reviewed. DNA has been collecting data (especially for rural water supply coverage) to measure ―access‖ based on engineering definitions of access.11 In the framework of the MDG monitoring, use of services has become an increasingly important concept. ―Use of services‖ refers to the number of people using certain water supply and sanitation services. Unlike engineering 10 ASAS funds are funds that are provided by donors to the sector agencies, in this case DNA, and are registered on- budget, and on the Single Treasury Account 11 These data are collected on the basis of outputs delivered in the sector, such as the number of water points built or rehabilitated or number of connections. These outputs are then multiplied by the number of people that are considered to be accessing these outputs, resulting in the number of beneficiaries having access to these services. 17 estimates, these data are collected through household surveys and census. The ―use‖ data tend to be generally lower than ―access‖ data as it captures non-functioning water points. For the sake of clarity, this report presents the survey numbers which are here also designated as ―use‖. Notwithstanding these efforts, it is important to be aware of the data limitations. These limitations should be borne in mind when interpreting the results of the analysis. The creation of FIPAG, CRA, and now AIAS have contributed significantly to the availability of urban water performance data. Similarly, the SISTAFE law12 has increased the availability of expenditure data. However, some gaps remain. First, data on donor contributions to the sector is not always consistent and there is little overlap between different data sources, for example between the OECD database of donor contributions and the ODAMOZ database in Mozambique. This is likely due, in part, to off- budget donor funding. In addition, some on-budget donor funding is not channeled via the Treasury single account (―off-CUT‖). This has implications for analyzing public expenditures in that the only source of data available at the time of the analysis for the years 2007 and 2008 were the Budget Execution Reports (BER). The BERs typically underestimate the total budget execution in the sector due to delays in incorporating the on-budget, off-CUT donor funding into official reports. Box 1: Evidence of misclassification of expenditures across budget types Mozambique uses a dual budget system aiming at separating capital expenses –in principle, recorded in the development budget— and current expenses recorded in the recurrent budget. The data collection process took care to examine whether individual budget lines were correctly classified according to their economic nature into capital versus current spending; regardless of whether the budget line belonged originally to either budget. Few of the sector institutions routinely report their expenditures following the economic classifier. Most of them limit their financial accounting practices to keeping some records of project disbursements in Excel worksheets as reported by the project implementers. In doing so, they do not always comply with Government regulations establishing the use of 13 the official chart of public accounts and the economic and organic classifiers as mandatory. Misclassification of expenses is not limited to external projects. The only institution that makes a comprehensive use of the economic classifier across all its expenses is ARA-Sul. The most obvious misclassification is that of recurrent and development expenditures. Development expenditures seem to include a significant portion of recurrent expenditures. This misclassification is a result of the sector responding to incentives created by less flexible criteria for allocating discretionary shares of the recurrent budget than for shares of the development budget. Because of data issues, there are differences in the depth of analysis between the different water sub-sectors, with limited information on water resources. Basic sanitation is included where possible, but is quite likely seriously underestimated as this type of expenditure is spread over many different entities – including the Ministry of Education through its school sanitation programs. 12 SISTAFE is a financial administrative system that was introduced by the Government to better manage the government’s expenditures. The electronic version of SISTAFE is called ESISTAFE. 13 Diploma Ministerial 23/2004 establishing a new chart of accounts, classification of revenues by their source and a more detailed organic classification including at the district level. 18 3. SECTOR BACKGROUND Mozambique has seen substantial economic growth over the past decade following the devastating civil war that ended in 1992. Economic expansion at an average annual rate of 8 percent between 1996 and 2008 was made possible by overall macroeconomic stability, sound policy reforms, growth in agriculture, post-war reconstruction, mega-projects, and strong support from development partners.14 However, despite this strong economic growth, Mozambique remains one of the poorest countries in the world, ranking 172 out of 177 countries in the Human Development Index (an index of income, education and life expectancy) for 2007/2008.15 The World Bank Country Partnership Strategy of 2007 articulates that water resources are a major natural asset for Mozambique and are instrumental in all economic sectors. Future and sustained growth is expected to rely on the natural resource base, including the country‘s water resources. Similarly, provision of reliable water supply and sanitation services is a critical element of Mozambique‘s long-term growth and poverty reduction plans, with effects on the overall well-being of people through impacts on health, education, gender equality, and productivity. Water-related issues have been identified as a priority area in the Government’s National Poverty Reduction Action Plan (Plano de Acção para a Redução da Pobreza Absoluta, PARPA I and II). Similarly, the World Bank Country Partnership Strategy of 2007 includes among its development outcomes for Mozambique increased access to potable water (Outcome 8) and increased sustainable management of water resources (Outcome 17).16 While Mozambique is not considered as water scarce, it is highly water vulnerable. Water resources management is constrained by increasing uncertainty in the national water resources base. In the area of water supply and sanitation, policy reforms anchored in the National Water Policy and National Water Development Program of 1995 have helped progress in the sector but gaps remain. Key water-related challenges for Mozambique can be summarized as follows:  High dependence on shared water resources, and vulnerability to hydrological variability and water shocks. Over half of Mozambique‘s water resources originate outside the country. This high dependence on shared water resources is an important factor affecting national water vulnerability. Development and use of the country‘s water resources is also affected by the highly variable climate with large variations in annual and inter-annual rainfall, and frequent flood and drought events. Floods and droughts have a significant impact on the country‘s economic performance. The poor, who depend largely on subsistence agriculture, are often the most vulnerable to these shocks. The uneven geographic distribution of water resources across the country and growing competing future demands by water–dependent economic sectors in many river basins represent serious challenges to long-term growth. In an interesting comparison, an analysis undertaken over the period 1981-2004 suggests that GDP growth is cut 14 World Bank, 2007. Country Partnership Strategy for the Republic of Mozambique. Washington, D.C.: World Bank. 15 Watkins, K., 2007. Human Development Report 2007/2008: Fighting climate change: Human solidarity in a divided world. New York: UNDP. 16 World Bank, 2007. Country Partnership Strategy for The Republic of Mozambique. Washington, D.C.: World Bank. 19 by 5.6 percent on average when a major water shock occurs. In this analysis, the total costs of water shocks over the period under review were estimated at about US$1.75 billion. Figure 3.1. Mozambique: Real GDP Growth Rates and Rainfall (1988-2002) Source: Mozambique Country Water Resources Assistance Strategy: Making Water Work for Sustainable Growth and Poverty Reduction. World Bank (2007)  Gaps in provision of water and sanitation services remain an issue. Access to improved water and sanitation services is still low, while the reliability and quality17 of water services still needs improvements. Despite the recent reforms which have helped to strengthen sector institutions and the regulatory framework, improvements in the use of water supply services is showing only modest improvements, while the gaps in the quality of rural and urban water services remain. In the area of sanitation, progress has been made but much work still remains to be done to ensure that the adverse impacts of lack of improved sanitation can be avoided. 17 World Bank, 2007. Project Appraisal Document on a Proposed Credit to the Republic of Mozambique for a Water Services and Institutional Support Project. Washington D.C.: World Bank. 20 4. INSTITUTIONAL CONTEXT AND LEGAL FRAMEWORK 4.1 Institutional Context and Legal Framework There are a number of strategies and policies guiding Mozambique’s water sector. Mozambique‘s goals and macro strategies are laid out in the Poverty Reduction Action Plan (PARPA) which is a main reference for water supply and sanitation policies. Yet, the PARPA does not include national targets for service coverage (although it includes a set of output indicators for the period 2006 to 2009), funding mechanisms or other sectoral policies in water and sanitation. A water policy was established in 2007 which sets national targets for water supply and sanitation coverage, specifically, to achieve the Millennium Development Goals. In addition, the sector has a Rural Water and Sanitation Strategic Plan, and a National Strategy for Management of Water Resources, whereas an Urban Water and Sanitation Strategic Plan is currently under preparation. The provision of water supply and sanitation in Mozambique is carried out through various institutions. Water supply in rural areas is provided by the National Directorate of Water (DNA) located in the Ministry of Housing and Public Works (MoPH), through its ten Provincial Directorates of Water (DAS). There are approximately 300 small rural water supply systems managed by the District Administrations. The DAS, which answer technically to DNA but administratively to the provincial government, are responsible for the organization and implementation of rural water supply and sanitation projects including the construction and rehabilitation of wells and boreholes. The DAS also coordinate and monitor activities of contractors and the consultants supervising the works. Currently, decisions on water supply investments in rural areas are taken by DNA and implemented by DAS, which results in the provinces having limited decision- making powers over investments. Yet, it should be noted that districts have autonomy and do often assign funds to water supply investments.18 As such, the current model is a centralized model of allocation and provision, with DAS having little involvement in the actual decision making-process. As the decentralization processes evolves, it foreseen that the DAS will become more dependent on the province and less so on DNA. This change will constitute a large change in the planning and implementation of rural water supply investments. Other rural water supply systems are managed by administrative posts and there are boreholes and wells equipped with handpumps which are managed by communities. In addition, many NGOs are active in the country, especially in the rural areas, and a significant number of them are involved in the provision of water supply. 18 These district investments are not always very efficient as districts usually cannot offer drilling contracts of an economically viable size for contractors, and they often lack the capacity for procurement and technical supervision. 21 Figure 4.1 Flow of Funds in Water Supply and Sanitation Sub-Sector Other Ministries MoPH MoF Donors Donor Projects (including health) DNA FIPAG CRA Regional DAR DAU Government (rural) (urban) DAS Local Rural systems managed by DAS Districts Government Rural systems managed by Urban systems not Urban operator communities and (delegated management) under FIPAG administrative posts Users Over the last ten years, the provision of water supply in urban areas has become dominated by the institutions set up to support the Delegated Management Framework (Quadro de Gestão Delegada do Abastecimento de Água, DMF) which was created by Decree no 72/98. It initially supported private water supply operations in the five cities of Maputo, Beira, Quelimane, Nampula and Pemba. The principle behind the framework is the separation of water operations (which are delegated to autonomous private sector entities) from asset ownership (which remains in public hands). There are three main actors in the Delegated Management Framework: the Water Assets and Investment Fund (Fundo de Investimentos e Patrimônio do Abastecimento de Água, FIPAG), the Council for the Regulation of Water Supply Services (Conselho de Regulação do Abastecimento de Água, CRA), and the independent (private or public) operators managing the utilities. FIPAG was created by Decree no 73/98 (and subsequently enlarged under Decree no 19/09) to perform the role of custodian of the water assets; it enters into contracts with private operators in the designated cities. Currently, FIPAG manages 14 water systems in urban areas, including the system of Maputo city and all other major urban centers. CRA was created by Decree 74/98. Its main role is to protect the interests of consumers, while at the same time ensuring that a balance is maintained between the interests of consumers, the Government of Mozambique, and the operators. CRA regulates lease and management contracts signed between FIPAG and private operators. The third group of agents is the private operators – responsible for operating and maintaining the water systems and whose responsibilities depend on the type of contract. In Maputo, there is a lease contract under which most management and operations activities are delegated to a private operator. In the other cities, FIPAG is the operator for a transitional period, but is working to put the systems under private 22 management. Figure 4.2 depicts the relationship between FIPAG, the CRA, and private operators under the Delegated Management Framework. Figure 4.2 Delegated Management Framework GOVERNMENT OF MOZAMBIQUE INTERNATIONAL FUNDING AGENCIES MoF MoPH On-lending Performance Stakeholders Forum Agreement Contract FIPAG FIPAG-Vitens CRA Lease Contract Technical and Management Assistance Contract Contract OPERATORS Subscription Contract CONSUMERS Source: FIPAG The five Regional Water Basin Agencies are responsible for water resources development and management, including collecting hydrological information, controlling irrigation systems and collecting water fees. The Regional Water Basin Agencies (Administraçoes Regionais de Aguas, ARAs) were established through Decree no 16/91 and formally established through the Ministerial Diploma no 134/93 on decentralized and integrated management of water resources at river basin level. The ARAs have administrative, organizational and financial autonomy, but report to the MoPH. ARA-Sul (Southern Region) and ARA-Centro (Central Region) are currently the only two basin agencies that are fully operational though the others have all been established and are receiving various levels of support. Within the ARAs, Basin Management Units carry out the management functions of the ARAs and Basin Committees are formed at the local level to enable consultation with stakeholders. In areas not yet covered by an (operational) ARA, DNA and the Provincial Directorates of Public Works and Housing (DPoPH) are gradually handing over responsibility for water resources management to the ARAs as their capacity grows. 23 Figure 4.3 Flow of Funds in WRM Sub-Sector MoPH MoF Donors DNA ARA-Zambezi, ARA-Sul & Regional Centro & ARA-Centro Government Norte Basin Management DPoPH Unit Local Basin Committee Government Users The ongoing decentralization process in the country may, over time, significantly impact the provision of water services. Decentralization so far has been mainly an exercise of de- concentration of expenditures, whereby the provincial directorates and the districts absorb a small portion of the flow of sector funds through the MoPH. Thus, in terms of resource allocation, the centralized model of rural water supply service provision still remains, with DNA planning investments and allocating resources to their provincial DAS which are implementing and supervising investments in the provinces. Similarly, FIPAG‘s expenditure is still fully centralized. None of their funding goes directly through provincial directorates or through municipalities. Although FIPAG is not the owner of district water assets, it has taken over some of the municipal water assets. Where FIPAG has taken over municipal assets, responsibility has been effectively removed from the Municipal Council. Decentralization has so far mainly impacted rural water supply provision with some responsibilities transferred to districts and provinces. Districts will become responsible for the maintenance of their water sources, while provinces will be made responsible for the construction and rehabilitation of these sources. The existing legal framework on municipalities‘ legal authority19 states that municipalities are responsible for water supply, while FIPAG is responsible for the water supply infrastructure. These two provisions obviously overlap and lead in practice to the lack of clarity. The Government has established the AIAS - Water and Sanitation Infrastructure Management Agency. The AIAS would have a structure similar to FIPAG, but operate those water supply and sanitation systems in urban areas outside the scope of FIPAG through provincial water and sanitation boards to be set up by the individual provincial governments. The role of municipal 19 Municipal Development in Mozambique: Lessons from the First Decade, page 6. 24 governments and district administrations with respect to AIAS and the provincial bodies is not yet well defined. Table 4.1 Organizations and their Roles and Responsibilities in the Water Sector Organization Responsibilities Ministry of Public Works Lead agency for water and sanitation and Housing (MoPH) Ministry of Education The Ministry of Education plays a role in the provision of rural water and sanitation through construction of primary schools in rural areas fitted with water supply and school sanitation infrastructure. Ministry of Planning and Planning, budgeting and funds allocation. Responsible for on-lending agreements under DMF Development Ministry of Finance Treasury disbursements, audits and fiduciary control under DMF; Channels funding from donors to the water sector. Ministry of Health The Ministry of Health has a sanitation section which leads some sanitation initiatives and provides a low level of promotional effort and sets parameters for water quality in the country. The Ministry is not involved in the financing of new infrastructure. Water Regulatory The CRA is the regulatory body in the urban water and sanitation subsector, and reports directly Council (CRA) to the Cabinet. National Directorate for DNA (located in MoPH) is the water sector focal point. Its roles include policy-setting, planning, Water Affairs (DNA) monitoring, and reporting for rural water supply and sanitation, urban water outside FIPAG (for now), and some urban sanitation. The DNA, along with FIPAG, puts forward funding proposals and new sector initiatives. In addition to its role in WSS, DNA has responsibility for international rivers, hydraulic works, and water resources management. Water Supply FIPAG is an investment fund with legal responsibility for provision of urban water supply services Investment Fund (FIPAG) in 19 municipalities and 3 towns. In these cities and towns, FIPAG owns the water utility assets. It is a semi-autonomous agency under the MoPH and acts in concert with DNA to put forward funding proposals and new sector initiatives. Regional Water ARAs are the regional water administration entities responsible for water resources management. Resources Management There are five ARAs which have administrative and financial autonomy and report to the MoPH. Agencies (ARAs) The ARAs are responsible for collecting hydrological information, providing water to irrigation systems, and collecting water fees. Water and Sanitation With a structure similar to FIPAG, the newly created AIAS would operate water supply and Infrastructure sanitation systems in urban areas outside the scope of FIPAG through provincial water and Management Agency sanitation boards to be set up by the individual provincial governments. (AIAS) Provincial Directorates This is the provincial government branch with the water sector within its purview. In areas not yet of Public Works and covered by an ARA, the DPOPH is the authority responsible for water resources management in Housing (DPOPH) the province. Provincial Directorates The DAS are provincial water and sanitation departments under the DPOPH. Provincial of Water (DAS) governments are responsible for province-wide planning, the application of national policies and regulations, sector monitoring, some small town system construction and oversight, and some rural water point construction and rehabilitation District Governments District governments own all public water supplies within their jurisdictions and are responsible for needs identification, annual planning and promoting additional access. Most districts also manage their capital’s water system. DNA’s current policy is that the district ensures the maintenance of existing infrastructure while provincial governments assume the responsibility for access expansion, though this policy is not strictly enforced. Municipal Governments The municipalities have legal responsibility to assure water supply within their area of jurisdiction. 25 Organization Responsibilities In the larger ones, FIPAG has been assigned ownership of and responsibility for ensuring management of the water supply systems, creating an overlapping of authority, resulting in some loss of clarity regarding the division of responsibilities. In the case of sanitation, authority for provision and O&M of infrastructure associated with sewerage systems is with the municipality. User groups Rural water points are often managed by community groups with little or no government intervention or regulation Utilities and Large cities and towns are served by public and private utilities. But in peri-urban areas of Maputo independent providers and Matola there are more than 400 independent providers serving 300,000 clients (approximately 10% of the population) CFPAS Water and Sanitation training center 4.2 Budget Cycle The budgeting process in the water sector draws on a number of planning tools and involves a great number of different actors. The budget calendar and key activities that take place in the elaboration of the Mozambique budget are outlined in Table 4.2. Table 4.2 Budget Cycle February - April Update of the MTFFfor the following year to determine the overall resource envelope and define initial sector ceilings. Drafting of municipality, district and provincial PES (Socio-Economic Plan) and Budget Proposals begins. May MTFF envelope and ceilings approved. Initial ceilings sent to sectors and other budget units with the methodology for formulating proposals for the annual state budget (OE) and Economic and Social Plan (PES). May – July 31 Based on budget ceilings, sectors, and other budget units, prepare budget proposals and submit them to the DNO (National Department for the Budget within the Ministry of Finance) and the Ministry of Planning and Development (MPD). August Final update of resource envelope and allocations proposed in MTFF and state budget including donor commitments. DNO/MPD hold budget discussions to compile and consolidate proposals for OE and PES. 15 September Draft OE and PES are sent to the Economic Council and then to the Council of Ministers for approval. 30 September Draft OE and PES are submitted to Parliament. 15 December Deadline for parliamentary approval of OE and PES. Source: McCoy, Simon. Provincial Budget Allocations in the Health, Education and Water sectors of Mozambique: An analysis 2003-2006. January 2008. The state budget uses the classification for recurrent and investment expenditure. In practice, there is some misclassification of expenditure whereby a portion of the investment budget ends up financing operating costs that are not covered under the recurrent expenditure.20 20 Source: McCoy, Simon. Provincial Budget Allocations in the Health, Education and Water sectors of Mozambique: An analysis 2003-2006. January 2008. 26 Recurrent expenditure in the state budget is set by extrapolation of the budgetary execution of the previous year. While sectors and provincial authorities are free to recommend deviations from initial budgetary ceilings, changes to total flows by province are difficult to achieve. The result of this is that past patterns are frequently maintained and perpetuated. Provincial and sector budget limits are communicated to the Provincial Directorates of Planning and Finance (DPPF) and to the central ministries. Following this, there is an opportunity for those in receipt of funds to recommend deviations from the budget ceilings in budget negotiation meetings with the DNO following the writing of the State Budget (Plano Económico Social, PES). In practice, changes, if any, are minimal. This makes the budget allocations very stable, but also provides little incentives for cost reductions, innovations, and budget compliance, while it can also result in new priorities in the sector not being matched with financial resources. In the case of investment expenditure, the state budget is divided into two inter-related parts: the external component and the internal component. The external component is essentially a list of investment projects financed directly by donors. Within the external component, there is a further distinction between projects that are in accordance with the Governments‘ plans for the water sector (―on-plan‖), those that are written into the state budget (―on-budget‖), those whose funds are actually channeled via the Treasury single account (―on-CUT‖), and funds which are allocated at the discretion of donors (―donor-discretional‖). While the Government may have some success capturing donor projects on the budget, persuading them to channel funds via the Treasury has remained a challenge. Such ―on-budget‖, ―off-CUT‖ projects have some transparency in terms of projects and allocations, however, actual execution is difficult to track as funds bypass the Ministry of Finance and move directly to the beneficiary institutions. The bypassing of the Ministry of Finance has reduced over time, but is still a major issue and requires setting some guidelines for donors to ensure that the Government has insight into what is happening in the sector and can ensure that this information results in improved decision making as to where scarce resources will be used. The internal component of investment expenditure is funded largely by non-earmarked General Budget Support. 27 5. WATER SUPPLY AND SANITATION SECTOR PERFORMANCE In this chapter, the performance of Mozambique‘s water supply (Section 5.1) and sanitation (Section 5.2) sectors is presented focusing on the use of services as measured by census and household surveys. The overall management of water resources has not been assessed here, as there is insufficient information to track progress over time. 5.1 Access to Improved Water Sources Use21 of improved water sources has been stagnant over the past two decades. Household survey data22 show that use of improved water sources increased marginally between 1997 and 2008. Of the 43 percent of the population in Mozambique using improved sources in 2008, 17 percent used piped water, 6 percent water provided by neighbors and another 20 percent other protected sources (mostly groundwater). Figure 5.1 Use of Improved Water Sources in Mozambique, various years Source: Household Surveys and Census Data, various years While there has been an absolute increase in the urban population served, the share of urban dwellers using improved water sources has been decreasing. Growth patterns are distinct 21 The data presented here follows the Government’s nomenclature for survey data, and refers to “use”, rather than access, as measured by the number of water points and the number of people for which such water points are constructed. 22 The categories used were slightly different between the three surveys: use of a neighbor’s piped water supply was , not considered separately before 2007 . 28 between rural and urban areas. Use of improved water sources by populations in urban areas declined between 1997 and 2008, from 83 percent to 70 percent, respectively. In 2008, only 60 percent of the urban population used piped water (including water provided to neighbors) compared to 71 percent in 1997. In the capital city, the share of the population using improved sources has been steadily declining since 1996/97– with usage decreasing from close to 100 percent in 1996/97 to 92 percent in 2008 due to rapid urban population growth. However, even though use of improved sources declined in terms of the percentage of population served, 1.7 million more people living in urban areas gained access during this period. Of these 1.7 million additional people using improved sources, more than 1.4 million of them used piped water between 1997 and 2008. Use of improved water sources in rural areas is increasing, albeit slowly. In rural areas, use of improved water sources increased to almost 30 percent in 2008 from 23 percent in 1997. Most of that increase was brought about by increasing dependence on other improved water sources, especially groundwater. In contrast, the importance of piped water in rural areas, especially in the form of standpipes, has declined rapidly. As population growth in rural areas is low, increasing the number of people using improved water supply sources tends to be easier to achieve than in urban areas. Between 1997 and 2008, about 1.3 million people in rural areas gained use of improved water sources. Spatial use of improved water sources shows wide variations. The household survey data show significant variations across regions and across time. In 2000, the difference in use between the province with the lowest access and the province with the highest access rates was more than 80 percent – with the lowest rate of access to improved water supply sources measured in Zambezi and the highest in Maputo City. In 2008, the gap had decreased, but was still more than 60 percent – with the same provinces on the top and bottom of the list. Interestingly, use of improved water supply sources tends to vary by geographical location, with use of improved water sources increasing from the north to south – with the exception of Inhambane and Zambézia provinces. 29 Figure 5.2 Use of Improved Water Sources shows large geographical variation Source: INE, Household Surveys, 2000/01 and Multiple Indicator Cluster Survey (MICS) 2008 There are striking differences in the use of improved water sources across time. In some provinces, use of improved water sources has increased significantly over the years: in Nampula, it almost doubled between 2000 and 2008, and in Maputo Province, Niassa, and Zambezi, significant increases were registered. Yet, at the same time, Manica, Cabo Delgado, Maputo City, Gaza, and Tete show decreases in the use of improved water services. Figure 5.3 Change in Use of Improved Water Sources between 2000 and 2008 by Province Source: Household Surveys, 2000/01 and Multiple Indicator Cluster Survey (MICS) 2008 30 The quality of service (as measured by the level of service provided) is declining. In urban areas, piped water provided through house and yard connections or standpipes has decreased between 2000 and 2008 from 51 to 47 percent. Use of protected groundwater declined from 15 to 9 percent. At the same time, use of non-protected groundwater increased from 16 percent in 2000 to 27 percent in 2008, whereas dependence on other sources (often including water from neighbors – which can and often does include piped water) increased from 17 to 20 percent. Hence, the level of service has been declining. In rural areas, a similar trend is noticeable, with an increasing share of households depending on groundwater sources. Use of piped water among rural households decreased as a percent of rural use of total improved sources from 52 percent in 2000 to 15 percent in 2008. As the utilities cannot keep up with rapid urban growth, urban households look for alternative forms of water supply. In 2008, households in urban areas did not depend as much on piped water as they did before (47 percent of population), but they were increasingly using non-protected sources (27 percent) and water provided through neighbors (18 percent). The use of neighbor water is especially high in the wealthier income quintiles suggesting that those households are the most capable of finding alternative sources when utilities cannot keep up in providing services. At the same time, use of unprotected groundwater sources is especially prevalent among the poorest households; hence the burden of a deficient water service is mainly affecting the poor. 31 The poor have less access to improved water supply services than the non-poor. Use of improved water sources is highly skewed towards the richest households, as can be seen in Figure 5.4. In the wealthiest quintile (Quintile 5), 15 percent of the population does not use improved water sources compared to 87 percent in the poorest quintile (Quintile 1). Hence, use of improved water sources is highly unequal, especially compared to other countries in the region. Not only do richer households have much higher access to improved sources, they also tend to have much higher levels of service. Access to house and yard connections is almost exclusively the domain of the wealthiest households in the country. However, this wealth is relative, and reflects in large part the higher wealth of urban residents, as compared with predominantly very poor rural dwellers. Figure 5.4 Use of Improved Water Supply Services by Consumption, Quintiles in 2008 Source: Multiple Indicator Cluster Survey (MICS) 2008 32 The service gap in access to improved water sources between poor and non-poor has been increasing over the past years. The population in the lowest income quintiles saw little progress in achieving access to improved water sources. Although the differences in data definitions between the household surveys of 2002/03 and 2008 are stark, it appears that almost all the gains in access to improved water sources have been made in the three highest income quintiles at the expense of those in the lowest income quintiles. Table 5.1 Use of improved water source by income quintile Income quintiles 2002/03 2008 Access to Piped Access to all Access to Piped Access to all Water Improved Water Improved (excl. neighbors) Sources (excl. neighbors) Sources Q1-poorest 6.3% 28.5% 0.0% 12.6% Q2 9.3% 26.3% 0.5% 22.8% Q3 10.3% 30.3% 9.4% 44.4% Q4 15.9% 37.6% 20.3% 49.9% Q5-wealthiest 31.3% 50.3% 54.1% 85.1% Source: Household Budget Surveys, 2002/03 and Multiple Indicator Cluster Survey (MICS) 2008 Note: the quintile calculation as used in the Household Budget Survey and the MICS is not similar and hence not fully comparable. Note: The “other” category in access to water sources is quite distinct between earlier and later years and may skew the results, as it is possible that neighbor’s water is re-classified between surveys from unimproved to improved services. Service levels differ widely between poor and non-poor. The poor almost exclusively depend on non-piped water sources, and access to piped water increases with expenditure level. Nevertheless, wealthier households are depending less on piped water than they did in the past. Another interesting feature is the increased dependence on other improved water sources, mainly protected wells and boreholes in rural areas and neighbor water in urban areas. Table 5.1 shows that in 2002/03, piped water made up about 22 percent of the total use of improved water sources for the poorest quintile. By 2008, the poorest households only used point sources. For the richest quintile, the trend is the opposite –use of groundwater declined from 19 percent in 2000/01 to 7.5 percent in 2008, whereas dependence on other sources (including water provided through neighbors) increased. Less than half of the households surveyed live within 30 minutes of the nearest drinking water source. Overall, 48 percent of the households in Mozambique had access to a drinking water source within 30 minutes in 2008 – with the average time to go and return from the water source being 49 minutes. Yet, almost 83 percent of the households in the wealthiest quintile are located within 30 minutes of the nearest drinking source, compared to less than 35 percent of the households in the poorest quintile. Households in the richest quintile spend about half the time that poorer households spend to go and return from their water source. The long distance to a drinking water source is especially prevalent for poorer rural households. There are also wide variations by province. In Maputo City, only 8 percent of households spend more than 30 minutes to go and return from their 33 water source in contrast to the province of Gaza where 72 percent of the households spend more than 30 minutes to haul water. Table 5.2 Distribution of time to the nearest drinking water source by income quintile, 2008 Income quintiles Less than 30 Between 30 - 60 More than 1 hour minutes minutes Q1-poorest 34.7% 30.4% 33.9% Q2 37.4% 31.5% 29.9% Q3 42.9% 28.2% 28.3% Q4 46.7% 23.2% 29.2% Q5-wealthiest 82.6% 9.3% 7.6% Source: Household Surveys, 2008 5.2 Performance of Water Supply Sector 5.2.1 Rural water supply sector performance Less than one third of Mozambique’s rural population uses improved water sources. As is shown in Section 5.1, most households in rural areas use non-piped water, which mainly refers to wells with or without hand pumps, boreholes, and springs. The performance of the rural water sector is lagging seriously behind that of the rest of Sub-Saharan Africa and its growth is similar to that of other low-income countries, but lower than that of other countries in Sub-Saharan Africa. Table 5.3 Use of Improved Water Sources in Rural Areas between 1990-2006 Growth Rate of Population with Access 1990- 1990 2006 2006 Mozambique 24% 26% 2.0% Sub-Saharan Africa 35% 46% 3.6% Low-Income Countries 50% 60% 1.9% Source: UNICEF-WHO Joint Monitoring Program, 2008 34 Functionality of water access points is an issue. Although no specific studies are available about the functionality of rural water points in Mozambique, data from National Water Directorate (DNA) seems to point to low functionality. In general, more than 40 percent of the total water points added by DNA each year are, in reality, rehabilitation of existing points, as can be seen in Table 5.4. Table 5.4 Annual Production of Water Points between 2003 and 2008 Number of water points 2003 2004 2005 2006 2007 2008 Total number of new water points 696 523 615 734 1,529 1,368 Rehabilitated water points 508 355 574 714 913 1,204 Total water points23 9,200 9,896 10,419 11,034 11,768 13,297 Rehabilitated water points (% of total) 42.2% 40.4% 48.3% 49.3% 37.4% 46.8% Breakdown rate of water points 30% 31% 33% 34% 34% 33% Source: DNA/MPOH Annual Reports and Annual Strategic Plans, 2004 till 2009 Although no longer time series are available, the high level of rehabilitation in the total annual production suggests a relatively high breakdown rate, and a focus on rehabilitation investments. Combining data on the annual production of water points with other data from DNA on new and rehabilitated water points suggests a breakdown rate of about 33 percent in 2007. If one assumes the average lifetime of a well with a pump at about 7-10 years, a breakdown rate that is much higher than 10 to 14 percent assumes that the efficiency of investments in rural water supply are low. The low functionality rate can be explained by many different factors – some which are beyond the control of the sector. These factors include:  Hydrological conditions play an important role. In districts with consolidated sedimentary formations, boreholes exhibit high depths (>60 meters) and consequently lower functionality rates of hand pumps. The ratio of rehabilitation and construction is highest in the provinces of Gaza, Inhambane and Sofala. These provinces also tend to have the highest drilling costs, suggesting that hydrological conditions are less than optimal for hand pumps installation, operation, and maintenance.  A lack of an adequate market for spare parts and lack of effective arrangements for ensuring regular maintenance by trained mechanics may adversely affect functionality. The current policy of leaving the former to the private sector and the latter to communities to organize has not been very effective and new approaches are needed.  The capacity in the districts to operate and maintain water supply systems may differ significantly.  The geographic location of water points may also affect their sustainability. Figure 5.5 shows that water point location might be largely informed by proximity to a road, and less by 23 Based on calculating the total number of new and rehabilitated water points and the total water points in 2003. 35 population density. Thus, household survey data suggest that the assumption of 500 people per water point may not necessarily be very accurate. Given data on protected water points by DNA, the household survey data on rural population, and depending on protected groundwater sources, the actual number of people per protected water point is about 290. Figure 5.5 Distribution of water points in a province in Mozambique, 2004 36 Water point data from the National Institute of Statistics, (Instituto Nacional de Estaticas, INE) collected in 2003 in several provinces shows, when combined with population density and other infrastructure data, some interesting trends. First, it is not clear whether the Government is meeting the rule of thumb of providing an improved water point per 500 people.. Using the example of the Province of Zambézia, INE reports that the number of boreholes is are far less than what one would expect based on the Household Survey access data. This could imply that either survey data on water points is deficient or the number of people per water point exceeds the Government‘s target. Yet, the data in the particular districts show that water points are at times built in clusters, which is quite visible in the district of Namacurra (Figure 5.5). This clustering results in some villages (depicted in grey dots) having several water points (depicted in red dots) while others have no access whatsoever resulting in high inequalities in access as mentioned earlier in this Chapter. Secondly, as can be seen in the example of the District of Morrumbala (Figure 5.5), the location of water points is correlated with the presence of roads. A recent assessment of the national drilling sector cites that ―transport is most problematic during the rainy season. Therefore drillers prefer small works along major roads during the months of January to March. 24‖ The logistical challenges to drill boreholes in rural areas with no proximity to a road further add to the inequalities in access in rural areas. 5.2.2 Urban water and sanitation performance Most households in urban areas have access to some form of piped water. We will use two indicators to measure the performance of Mozambique‘s urban water sector. The first indicator measures access to improved water sources in urban areas of Mozambique compared to that of other countries, while the second indicator measures the sustainability of water utilities‘ piped water services. 5.2.2.1 Performance of the Urban Water Sector using International Benchmarks Mozambique’s urban water sector reflects similar trends seen in Sub-Saharan Africa as a whole. While Mozambique‘s urban population has increased rapidly since 1990, and more than 1.7 million urban users have been served with improved water supply, access rates have not kept up with the pace of urbanization. As a result, the percent of urban population using improved water supply actually declined during the period 1990-2006. Table 5.5 Use of Improved Water Sources in Urban Areas between 1990-2006 1990 2006 Growth Rate of Population with Access between 1990-2006 Mozambique 83% 71% -5.1% Sub-Saharan Africa 82% 81% -4.2% Low-Income Countries 87% 84% -1.0% Source: UNICEF-WHO Joint Monitoring Program, 2008 24 WE Consult. Assessment of the National Drilling Sector Capacity for Rural Water Supply in Mozambique. WSP, DNA: June 2006. 37 5.2.2.2 Sustainability of Urban Water Utilities Sustainability of urban water utilities can be measured in many different ways. In this section, we will look at the financial sustainability of FIPAG, the asset holder of Mozambique‘s water utilities, and the financial and operational performance of individual water utilities in the country. The existing government policy in the urban sector is to have centralized or regionalized utilities which can be financially sustainable without fully decentralizing. Under such a policy, it is assumed that the cities within the region are cross-subsidizing one another. An analysis of the Operating Cost Coverage of FIPAG as a whole shows that, on average, the utilities under the delegated management framework are able to sustainably serve their current customer base in the short run, as the operating cost coverage ratio has been increasing over time, notwithstanding the larger number of utilities managed under the delegated management framework. Table 5.6 Operating Cost Coverage Ratios for FIPAG, 2000-2008 2000 2001 2002 2003 2004 2005 2006 2007 2008 Operating Cost Coverage Ratio 2.5 1.3 0.9 0.8 1.2 1.2 1.0 1.0 1.1 Operating Cost Coverage Ratio less operating subsidies 1.0 0.4 0.4 0.4 0.7 0.8 0.9 0.9 1.0 Source: FIPAG Box 2 A simple of rule of thumb on financial sustainability  Short term sustainability assumes an Operating Cost Coverage Ratio (OCCR) (operating revenues divided by operating expenditures) larger than 1, but smaller than 1.45. In this case, the utility is capable of ensuring that the current customer base can be served in the short- run;  Medium-term or current system sustainability assumes that the utility not only covers O&M but also depreciation (similar to an operating cost coverage ratio of more than 1.45). In this case the utility will be able to replace worn out assets and can therefore serve the current customer base over time;  Long-term sustainability assumes an OCCR of more than 2 with all costs covered (including debt service) and which leaves some revenue for expansion of the network, and which, in the context of Mozambique’s rapidly growing urban population, with the funds to expand its customer base. 38 However, on a utility-by-utility basis, over half of these utilities cannot cover their operation and maintenance costs through their revenues. This compromises the utilities‘ capacity to serve their current customer base if no subsidies are provided. As can be seen in Figure 5.6, five of the utilities have an operating cost coverage ratio greater than one. None of the utilities is yet able to generate additional cash flow to ensure that they will be able to replace worn out assets and can therefore serve their current customer base in the long run. The possibilities for cross-subsidization between utilities are not very high. Even though many utilities are able to cover their operating and maintenance costs through their revenues, the additional funds they generate are small and not enough to ensure their medium-and long-term sustainability. And hence, any cross-subsidization in the short run will have to be matched by government funding to pay for the investments needed for most, if not all replacement and expansion of the network. Figure 5.6 Operating Cost Coverage Ratio for 13 utilities in 2008 (with exception of Maputo data referring to 2007) Source: FIPAG (Maputo data were not available for 2008) Inefficiencies in the way utilities operate further compromise financial viability. Large inefficiencies still do exist in the utilities managed under the delegated framework. An example is the high level of non-revenue water that averages about 56 percent, whereas the best performing utility has a non-revenue water of about 40 percent in 2007. The high level of non-revenue water is likely to be mostly of an administrative nature. It seems to be especially concentrated in the utility of Maputo which heavily influences any average because of its sheer size compared to the other 39 utilities. A large part of that non-revenue water is likely linked to the use of water by neighbors (including vendors), and the inability to control illegal connections. Preliminary data from the census for Maputo City and Maputo Province suggest that more people use piped water supply (house or yard connection or standpipes) than the data from the utility indicates. The gap may be explained by the presence of independent small-scale providers25 serving urban dwellers (which may count as piped water in the surveys) and the high number of illegal connections. Figure 5.7 Non-Revenue Water (%) in the utilities under FIPAG in 2009 Source: FIPAG 25 The rapid increase in tariffs may have resulted in households using more alternative sources that they can more readily control. 40 Despite recent improvements, over-employment remains another source of inefficiency. The best performing of the utilities in Mozambique has over 6 employees per thousand connections. This is twice as much as the average African utility (Africa Infrastructure Country Diagnostic (AICD) 2008) and much higher than the benchmark of two employees per thousand connections frequently used as the international benchmark for developing countries. In the other utilities the number of employees per 1,000 connections can run up to 20. The smallest utilities as measured by number of customers served suggest that part of the inefficiency is linked to lack of economies of scale26. Figure 5.8 Staff Ratio per 1000 connections in 13 utilities in Mozambique in 2004 and 2008 Source: FIPAG 26 It is also possible, as the AICD 2008 study noted, that utilities are used as a vehicle for employment creation that signifies a very inefficient way to transfer funds to the population. 41 Water rates are currently insufficient to expand services beyond the current customer base, but the rates are not low compared to many other countries in the region. As can be seen in Figure 5.9, the average revenue per cubic meter water sold (a proxy for the average tariff) has increased rapidly over time in US dollar terms. The tariff structure that utilities use is heavily skewed towards non-residential users, resulting in large cross-subsidies in the tariff structure. Although cross-subsidies have their advantages, the problem with such extensive cross-subsidies is that they provide non-residential users with an incentive to find alternative sources. This is already happening in the five utilities where the number of non-residential connections declined between 2002 and 2007 despite rapid economic growth in the country during that period. In Maputo and Beira, this decrease is especially linked to a decline in institutional connections; in the other three cities the decrease is a reflection of commercial and industrial connections. Figure 5.9 Tariff Levels in 13 utilities in Mozambique in 2000 and 2008 as measured by average revenues, per m3 water sold (US$) Source: FIPAG Note: In the above figure, “Original Utilities” are Maputo, Beira, Quelimane, Nampula, Pemba; “Second Generation” are Xai Xai, Chokwe, Inhambane, Maxixe; “Third Generation” are Chimoio, Mancica, Gondola, and Tete. 42 5.3 Sanitation Performance Access to improved sanitation is low. Only 19 percent of the population has access to improved sanitation according to the latest Multiple Indicator Cluster Survey (MICS) survey. Although that is still very low, it is a significant improvement compared to 2000/01, when less than 10 percent of households reported having access to improved sanitation. More impressive is the rate with which households without any access to a toilet have decreased: from 57 percent in 2000/01 to 42 percent in 2008. Open defecation has decreased, and more households have moved up the sanitation ladder. Figure 5.10 Access to Improved Sanitation 2000/01 and 2008 Source: INE, Household Surveys, 2000/01 and Multiple Indicator Cluster Survey 2008 Disparities in access to basic sanitation between rural and urban areas are significant. Some 55 percent of the population in rural areas report not having any toilet in 2008, as compared to less than 15 percent in urban areas. Toilets connected to a sewerage network or a septic tank (including ―com autoclismo‖ water born toilets and ―sem autoclismo‖ on-site solutions) are virtually nonexistent in rural areas (less than 1 percent), compared to 14 percent of the urban population using such toilets. Progress in reducing open defecation has succeeded – reducing the percentage of households that have no toilet from 71 percent in 2000/01 to 55 percent in 2008. Yet, a subsequent increase in access to improved sanitation has not yet taken place, suggesting that the affordability of sanitation services may be an issue. 43 Figure 5.11 Sanitation facilities in urban and rural areas Source: INE, Household Surveys, 2000/01 and Multiple Indicator Cluster Survey 2008 There are also significant differences in access to basic sanitation across time and space. There is a lot of heterogeneity in the use of sanitation facilities across regions. In 2008, the percentage of the population without any toilet varies from 0 percent in Maputo City to almost 72 percent in Zambézia Province. Figure 5.12 Households with access to basic sanitation by type of service by region in 2008 Source: INE, Multiple Indicator Cluster Survey , 2008 44 Income matters in gaining access to basic sanitation. As can be seen in Figure 5.13, the more well-off households are, the more likely they are to have access to improved sanitation. In 2008, only households in the three upper quintiles had access to some form of improved sanitation. Yet, even in these quintiles, a number of households do not have access to any toilet and have to depend on open defecation. Figure 5.13 Access to Sanitation by Income Quintile 2008 Source: INE, Multiple Indicator Cluster Survey 2008 Almost all increases in access to improved sanitation took place in the upper income quintiles. Figure 5.14 shows that the percentage of households with no access to sanitation services decreased rapidly for the upper quintiles – the higher the income quintile, the better the performance with regard to reducing dependence on open defecation. Yet, for households in the poorest quintile, access to some level of service declined – households in the poorest quintile increasingly have no access to any sanitation service. Figure 5.14 Percentage of households with no access to sanitation service between 2002/03 and 2008 Source: INE, Household Survey, 2002/03 and Multiple Indicator Cluster Survey 2008 Note: the quintile calculation as used in the Household Budget Survey and the MICS is not similar and hence not fully comparable. 45 6. WHAT IS BEING SPENT ON WATER SUPPLY AND SANITATION? 6.1 Introduction There are five main sources of funding for on-budget expenditure in the water sector. The first are Treasury Funds, which include donor aid channeled through the Treasury (the so-called ASAS funds are provided to DNA by the Government of the Netherlands, who is the only donor to provide such funds to the water sector), and general tax revenue, used for both recurrent and investment expenditure. A second source of funding comes from donor-funded projects which appear on-budget, with quarterly execution rates reported to the sector institutions. External investment is almost entirely financed through these funds. A third source includes smaller, donor- funded projects managed by the sector institutions (on-budget, off Treasury) used for a variety of purposes including operating costs, technical assistance, and small investment projects. A fourth source of funding is provincial, municipal, and district allocations that go directly to these respective levels of government and are used for many purposes including water expenditure. These allocations have been excluded from this analysis, as there is insufficient available information to determine what portion of the allocation is actually spent on water. A final source of on-budget funding is own-revenue generated through the collection of water-specific taxes (ARA-Sul, ARA- Centro) or fees from private operators (FIPAG and CRA). These funds are commonly used to support operations and running costs of the levying institution, and will be discussed separately in Section 6.5. Water sector spending is higher than the government budget assumes. This Public Expenditure Review (PER) is only a partial analysis due to a number of data issues. Although we were able to determine some of the sources of funding, others were not captured because either no data are available, or the data are not available at a sufficient level of disaggregation to ensure that the water components can be captured. The current analysis includes data on water spending from the Ministry of Public Works and Housing and other water sector agencies (FIPAG, CRA, ARA-Centro and ARA-Sul). Yet, there are many other funds outside these agencies that also affect water sector spending. Some funding is allocated through other Ministries, but not included in this analysis because of lack of disaggregated data. Other Ministries involved in the sector include the Ministry of Education and the Ministry of Health, which are responsible for a share of total funding going to rural water and sanitation. Another non-sectoral entity spending in water and sanitation is the National Institute of Disaster Management (INGC) which provides safe water for internally displaced populations as a result of recurrent floods in the Zambezi, Pungue, Sabe, and Incomati basins.27 The Provincial Directorate of Tourism of Cabo Delgado is scaling up its water investments to promote tourism in the province. In addition, the Millennium Challenge Account (MCA) is a new important source of funding for the water sector, but is not linked to central government water 27 The INGC has recently embarked on a program to extend its services beyond the transit camps for internally displaced population by constructing permanent settlements for people vulnerable to floods. 46 agencies. In 2007, 46 percent of total donor commitments to the water sector, as reported by ODAmoz, were from MCA (Figure 6.1).28 Although hardly any of this funding had been spent at the time this report was being written (less than 1 percent), its sheer size will require that this funding is monitored and included in any water expenditure in future years. Figure 6.1 Millennium Challenge Account & Other Donor Commitments in ODAmoz Database, 2007 (US$millions) Source: ODAMOZ database Off-budget funding is a significant source of funding in the sector but difficult to track. There are several forms of off-budget funding: (i) expenditures by water utilities; (ii) expenditures made in the sector by donors that are not registered in the Government‘s budget; and (iii) expenditures in the sector made by NGOs (whose budgets at times tend to be funded by donors).  Water Utilities. Funding from FIPAG and a small part of DNA funding is channeled to water utilities in urban areas which they provide with capital investment subsidies and operation and maintenance subsidies. We have presented data on the financial performance of water utilities in Chapter 5 but have not included data in the government budget data to avoid double counting, as data cannot always be disaggregated in sufficient detail to allocate costs to the individual public utilities. Current expenditure and revenue levels (2008) for 13 utilities (excluding Maputo) show that these utilities on average are able to cover their operation and maintenance costs through their revenues (which include some operating grants). 28 Despite the ODAMoz classification, total MCA funds not only go to the water and sanitation sectors but also to other areas such as roads infrastructure and the Nacala Dam. 47  Off-budget funding by donors. The ODAmoz database shows that many bilateral donors and also some multilateral organizations (such as UNICEF) are registered in the ODAmoz database as ―off-budget‖. Very often bilateral donors are registered for some projects on-budget and for some other projects off-budget. Of the total donor funding registered in ODAmoz, 24 percent of donor commitments in the period 2006-2008 were registered as ―off-budget‖. On top of this ―off-budget‖ funding, is the funding provided through projects that are not registered in the ODAmoz database, such as funding approved by the new generation of donors like India.  Funding through Non-Governmental Organizations (NGOs). Another important source of funding is financing provided by NGOs and civil society organizations. The Provincial Directorates of Public Works and Housing feed these NGO-led developments into their periodic monitoring and evaluation documents (the Balanço do PES, coordinated by the DNA). While these NGO projects tend to be small, they have an important impact in this largely under-funded sub-sector. As an illustration, DNA does not report the number of boreholes rehabilitated and created by DNA funds. Instead, it reports the total number of boreholes rehabilitated and created by both DNA and other (including NGO) funding, thus rendering it difficult to analyze the actual flow of funds in the sector, its outputs and effectiveness. In the water sector, any public expenditure analysis is impeded by the fact that the sector’s main agency is not an independent budgetary unit. DNA sits within the Ministry of Public Works and Housing (MoPH) but is not itself an independent budgetary unit (in contrast to FIPAG and CRA), which means that it does not receive an explicit allocation. Funds are allocated from the state budget to MoPH which, mainly based on implementation rates from the previous year and other variables (such as staffing patterns in the case of recurrent expenditure), allocates resources to DNA. DNA then allocates funds internally to its departments including its Rural Water Department (DAR), Urban Water Department (DAU) and Sanitation Department (DES). 48 6.2 Total Flow of Funds in the Sector Water spending has increased rapidly since 2002. Between 2002 and 2008, actual expenditure in the water sector (excluding the own-generated funding) grew from MZN 636 million in 2002 to MZN 2.3 billion – an increase of 255 percent. In real terms, the growth rate was less spectacular, but the expenditure on water still increased by a still impressive 140 percent. Figure 6.2 Increase in Actual Nominal Water Expenditure between 2002 and 2008 (MZN millions) Source: Ministry of Finance, Expenditure in PARPA Priority Sectors Note: While the source for 2002-2006 data is the State Final Accounts (which incorporate the execution of all funds, internal and external registered on the State Budget), the only available source for 2007 and 2008 is the Budget Execution Report (BER). This has implications since typically BER figures for sectors with large external funding not channeled through the State’s Treasury or directly managed by the sector (such as health or indeed water and sanitation) might underestimate total budget execution in the sector due to delays incurred in incorporating external execution data into official reports. Current levels of spending are increasing as a percentage of GDP, but remain low in absolute terms. The Africa Infrastructure Country Diagnostic (AICD) found that water expenditures from central government for water supply and sanitation ranged from 0.29 to 0.83 percent in Eastern African countries. Compared to these numbers, the total central government (including the allocations to provincial directorates and districts through the ‗ambito central‘) added up to 1 percent in 2008. However, this is well below the Maputo Pledge by African Ministers to increase public expenditures in water to 10 percent of GDP. In contrast, the AICD study estimated the annual investment needs (for both water supply and sanitation) under different scenarios to be from 5 to 8 percent of GDP, as can be seen in Table 6.1 49 Table 6.1 Annual Required Investment Needs to Meet MDGs in Water Supply and Sanitation As percentage of GDP Scenario Water Supply Sanitation Total WSS Expansion Rehabilitation O&M Expansion Rehabilitation O&M Expansion& rehabilitation without with O&M O&M Pragmatic 0.84% 0.81% 0.62% 1.22% 1.00% 0.25% 3.87% 4.74% Base 1.63% 0.81% 0.83% 1.12% 1.00% 0.23% 4.56% 5.62% High-end 2.23% 0.81% 1.00% 2.23% 0.81% 1.00% 6.08% 8.08% Source: Africa Infrastructure Country Diagnostics Water Supply and Sanitation Spending Model (2008) Note: In summary, the pragmatic scenario assumes that all new customers will be connected using the lowest level of service available, the base case scenario assumes that new customers’ service level will follow the current distributions of service levels, while the high-end scenario assumes that an increase in service levels will be available for both existing and new customers. Figure 6.3 Evolution of Expenditure as % of GDP for Selected Priority Sectors Source: Ministry of Finance The allocations to the water sector do not yet fully reflect the priority status of the sector. The Government has classified water as one of the priority sectors (which also include health, education, agriculture, infrastructure, governance, and social protection). Yet, of all priority sectors, water and sanitation continues to be the second least-funded (after social protection and employment). Water and sanitation‘s share in the total priority sector budget has only marginally increased from 4.9 percent in 2002 to 5.6 percent in 2008. It is likely that the current funding will not suffice to achieve the MDGs. 50 Apart from the lower levels of funding compared to other priority sectors, the funding for water and sanitation tends to be very volatile – with large annual changes. When compared against some of the other priority sectors, annual real growth in water and sanitation has been considerably more volatile. Some years have even seen reductions in the sector‘s allocation (2003 and 2007) with no apparent reason to explain these falls in available funding. Moreover, shrinking allocations have not been uncommon in the sector‘s central institution (DNA: 2004 and 2006). Figure 6.4 Annual Real Change in Expenditure for Selected Priority Sectors Source: Ministry of Finance Note: Annual growth rates in reported actual expenditure adjusted by annual inflation rates (as measured by the CPI for Maputo City used by the Government for macroeconomic programming and by the IMF Missions). The larger gap between BER figures and sector figures for 2007 and 2008, as noted before, results from the fact that the sector benefits from on-budget external funds not channeled through the Treasury whose execution is not incorporated into the Budget Execution Reports. The existence of these extra-Treasury resources may be playing against the sector when it comes to the annual allocation of the state budget. 51 6.3 Flows by Institution There are five major institutions in the water and sanitation sector with specific budget lines: the National Directorate of Water (DNA) through the MoPH, the Water Assets and Investment Fund (FIPAG), the Administration for Water Resource Management of the Southern Region (ARA-Sul), the Administration for Water Resource Management of the Central Region (ARA-Centro) and the Water Regulatory Council (CRA). For some of these institutions there are no available records of public expenditure during the initial years of analysis. However there is enough data to track FIPAG from 2000, DNA from 2003 and all five institutions from 2005 onwards. The expenditure patterns are markedly different for each of the institutions in the sector. Since at least 2003, FIPAG has consistently been the recipient of the largest share of budgetary funds, followed by DNA. Together, flows to FIPAG and DNA constitute 90 percent of central government flows into the sector. The flow of funds for FIPAG has been increasing since 2000 and it only experienced a decline in 2006. In contrast, the flow of funds going to DNA shows much more volatility, although since 2006 its expenditures have increased rapidly. Figure 6.5 Total Nominal Expenditure in the Sector by Institution (‘000 MZN) Source: As reported by Water Sector Institutions 6.4 Expenditure by Sub-sector Expenditure by institution does not equate to expenditure by subsector. While FIPAG is solely devoted to one sub-sector (urban water), DNA is responsible for expenditure in the following sub- sectors: urban water, urban sanitation, rural water, and rural sanitation. DNA is also responsible for water resource management of the Northern Region and the Zambezi basin until these administrations are operational to manage the water resources in these areas and are given specific lines in the state budget. The smaller institutions, ARA-Sul and ARA-Centro, are only responsible for water resources management in the Southern and Central regions, respectively (and fall outside 52 the MoPH budget), and CRA is the agency responsible for regulation and consumer protection. Moreover, expenditures that other administrative levels (districts, municipalities, and provinces) might incur in water and sanitation are not captured by sector reports, but are included in general funds that do not provide details on how these funds were allocated and spent. The reliability of the sub-sectoral data is far from optimal because of deficiencies in the availability and quality of data. Because sector expenditure is dominated by donor-funded projects, which often cut across various sub-sectors (for example by mixing water supply and sanitation), it is cumbersome to exactly determine the volumes going to each of these sub-sectors. For example, to calculate the total amount devoted to urban water, one must include FIPAG‘s budget, DNA‘s central funds going to urban water projects, DNA‘s decentralized funds transferred to the provinces going to urban water, DNA‘s donor-funded urban water projects, government participation in donor-funded urban water projects, ASAS funds allocated to urban water projects, fractions of cross-cutting projects going to urban water, as well as unreported expenditures such as those incurred by municipalities and districts and urban water off-budgets. As the data tends to be incomplete, the data presented here by sub-sector should only be seen as indicative. While the quality of data has improved in recent years, improvements are still needed to facilitate strategic decision making in the sector. Because it is difficult to track, for example, how much funding actually goes to rural water, it is also difficult to estimate how much more should be allocated to rural water if the sector is to achieve a specific coverage target as set out in the strategic plan for rural water. This, in turn, raises questions about the financial reliability of urban and rural water and sanitation strategic plans. Figure 6.6 Total Nominal Expenditure by Sub-sector (‘000 MZN) Source: Data from water sector institutions and authors’ calculations 53 Nevertheless, a clear pattern of expenditure by sub-sector emerges with the majority of funds being allocated to urban water supply. In 2008, urban water received 49.2 percent of total reported funding going to the sector, down from 70.8 percent in 2006. Although its relative weight in total expenditure has drastically fallen since 2003, in absolute terms expenditure in the sub-sector has doubled from 2003 reaching MZN 1.4 billion. Effectively, in 2008, DNA virtually ceased to be an implementing actor in urban water, but its role is expected to widen starting in 2009 with the Millennium Challenge Corporation (MCC) project setting up two project implementation units in DNA.29. Rural water and sanitation expenditures have increased rapidly in the last two years (albeit starting from a much lower base). DNA remains the only sector agency involved in rural water (managed by DNA‘s department of rural water) and urban and rural sanitation (both managed by DNA‘s department of sanitation). Expenditure on rural water peaked in 2008, receiving 12 percent of total available funding to the sector, despite more than two thirds of Mozambican population leaving in rural areas. Sanitation has increased rapidly, representing almost 15 percent of the sector budget. This relatively high level of spending for sanitation is mainly driven by a European Development Funds urban sewerage project and a UNICEF project to extend sanitation. 6.5 Allocation across Expense Categories The water sector budget is almost entirely made up of development expenditures. In 2008 development expenditures constituted about 86 percent of the total expenditures. Recurrent expenditure does not yet make up 14 percent of total expenditures. Most of this development expenditure is provided through donors, which in 2008 funded almost 88 percent of total development expenditures. The vast majority of donor funds are directly linked to projects. Domestic investment largely represents counterpart funding and contribution to pay projects‘ taxes. Since 2007, domestic investment volumes increased as a result of stronger efforts to alleviate the sector‘s debt burden of Value Added Tax (VAT) arrears and counterpart contributions to external projects. However, the VAT arrears remain a major problem and will be discussed in the following section. Total domestic resources to the sector (recurrent plus domestic investment) decreased in 2008, thus reversing a positive trend between 2004 and 2007. This trend is incompatible with the status of the water sector as priority sector. The reduction in domestic resources has especially affected the recurrent budget which, in 2008, was lower than in 2006, despite the larger than ever investment volume going into the sector. Recurrent expenditure has continuously declined since 2003. The share of expenses categorized as recurrent has fallen from 33.7 percent in 2003 to 13.7 percent in 2008. This decline in recurrent expenditure is partially due to inadequate monitoring of expenditures. Investment data include both 29 At least two more PIUs will be established for other components of the MCC compact: in the national roads authority and in the Ministry of Planning and Development. 54 recurrent and development expenditures. Because external projects (often not controlled by any government institution) tend not to adopt the government‘s accounting procedures and classifiers, which require a breakdown of expenditure by economic category, the actual recurrent data is likely to underestimate the magnitude of actual recurrent expenditures. In DNA, for instance, sector budget support is used to pay part of DNA staff salaries and benefits. In 2008, only 9 percent of DNA‘s total recurrent expenditures were classified as recurrent expenditure; the rest is included in the recurrent portion of investments. This shows an extreme dependence of DNA on external funding. DNA is not the only organization that has much higher recurrent costs than the official recurrent expenditure suggests – it is a pattern that shows up in all water agencies. This lack of local funding for recurrent expenditure has serious implications. The first is an extreme dependence of water sector agencies on external funding and the subsequent vulnerability of its core functions and the delivery of water and sanitation services to potential cuts in external aid. A second implication is that the funding for operation and maintenance of water sector agencies becomes a major issue once donor funding is completed, and the sustainability of services can be compromised. Figure 6.7 Composition of Nominal Public Expenditure in the Water and Sanitation Sector by Economic Category (MZN 000s) Source: data from water sector institutions and authors’ calculations The gap in reporting between the Ministry of Finance and sector agencies is narrowing. The larger gap between BER figures and sector figures for 2007 and 2008, as noted before, results from the fact that the sector benefits from on-budget external funds not channeled through the Treasury whose execution tends not to be incorporated into the Budget Execution Reports. But the amounts reported by the sector and the Ministry of Finance for the last year when State Final Accounts are available (2006) are very similar; a result that if confirmed by the State Final Accounts for 2007 and 2008 (not released yet) will speak of the notable improvements in financial reporting brought about by the reform of the country‘s financial systems (including the adoption of the SISTAFE system). 55 6.6 Spatial Allocation A very large part of the development budget was allocated to urban areas. Before 2006 almost all development expenditure was geared to provide services in urban areas. However, much of the large increase in commitments in 2007 is from a large commitment from the Millennium Challenge Account (MCA) which is an outlier in the basic donor funding trends. Further, the major sanitation element is mostly for urban drainage. Figure 6.8 Donor Commitments to Water Subsectors, 2005-2007 (US$) Source: ODAMOZ database and author’s calculations 56 The per capita rural water allocation is much smaller than the per capita urban water allocation. In 2007, the per capita urban water expenditure exceeded that of the rural water sector by almost twelve times – with about MZN 177 (equivalent to about US$7) being spent on water per urban dweller compared to only about MZN 15 (equivalent to about $0.60) per rural inhabitant, as shown in Figure 6.9. Part of that difference is related to differences in unit costs. In general, it costs less to supply water services in rural areas than in urban areas, as shown by the 2008 African Infrastructure Country Diagnostic. But differences in unit costs (and corresponding service levels) cannot alone explain this large difference. Figure 6.9 Urban and Rural per Capita Water Expenditure, 2007 (in MZN) Source: Water Sector Institutions and authors’ calculations In view of the Government’s interest in achieving the MDGs, it is likely that a larger part of the development budget should be allocated to rural areas. It should be noted, however, that even if funding were equitably distributed between rural and urban areas, the average cost of supply to an urban resident with improved water, especially piped water, tends to be higher than the cost to supply a rural resident. Moreover, in most urban centers the water sources close to the urban centers are already tapped and the next generation of urban water projects will transport water from very distant sources, increasing substantially the marginal cost of each additional cubic meter to be distributed in urban centers. The higher per capita cost of water supply provision in urban areas is directly linked to the choice of technology for service delivery. Piped water systems tend to be associated with higher costs than lower-cost technologies such as the use of wells or boreholes equipped with hand pumps, which are mostly used in rural areas. 57 6.7 Water Agencies’ Own Revenues Most water agencies generate some own revenues. Utilities generate revenues through water charges. The operating revenues generated by utilities are used to pay for the operation and maintenance costs of providing services to utility customers, while in several utilities these revenues also cover part of the depreciation. The water asset holder (FIPAG), the regulator (CRA) and the water basin agencies (ARA-Sul and ARA-Centro) also raise revenues through the collection of water-related taxes. FIPAG and CRA raise their revenues by charging utilities. Yet, for each of these three water sector agencies, resources still make up a very small part of the total revenues of these water institutions. These self-generated revenues have been hard to track. Even when agencies like CRA and FIPAG did report these self-generated revenues to the Ministry of Finance, they are not automatically reported in the official budget execution reports. For the water sector, the Inspection Services of the Ministry of Finance recommended that ―a comprehensive compilation of own revenues be collected, and that these revenues should be channeled to the delegations of the Ministry of Finance.30 The Ministry of Finance, however, still does not track these funds. The total own-generated revenues from the different water agencies add up to about MZN 350 million in 2008. Yet, this data is incomplete, as it does not include all own-generated revenues from Maputo utility, but only the income generated by the utility net of operators‘ income; nor does it include revenues generated by ARA-Sul for which data on water taxes was not available. Table 6.2 Own Revenue at Select Public Water Entities, 2000-2008 (MZN millions) 2000 2001 2002 2003 2004 2005 2006 2007 2008* FIPAG 47.4 45.4 61.7 109.8 165.4 198.5 315.1 351.1 325.8 Revenue from Maputo 39.8 34.0 34.6 29.5 73.1 69.1 98.5 88.5 81.4 Revenue Billed from 16 Towns 7.6 11.4 27.1 80.3 92.3 129.4 216.6 262.5 244.4 ARA-Centro - - - 0.5 2.0 3.0 2.4 3.2 2.4 Water Tax - - - 0.5 2.0 3.0 2.4 3.2 2.4 CRA 2.7 5.8 8.0 8.4 11.7 12.8 10.7 16.2 21.4 Regulatory Taxes 2.7 5.3 7.6 8.1 11.7 10.9 10.7 15.9 21.4 Other - 0.5 0.4 0.3 - 1.9 0.1 0.3 - Totals 50.1 51.3 69.7 118.7 179.0 214.4 328.2 370.4 349.6 *2008 numbers were not final at the time of the analysis. Source: Sector Institutions and author’s calculations Note: Revenue from Maputo is net of Maputo’s operator’s income 30 Auditoria de Desempenho do Sector de Aguas, IGF, Ministerio das Financas, 2006; presentation in March 2008. 58 6.8 Donor Funding in the Sector Donor funding has increased over the past few years and is concentrated in two sector agencies. Around 90 percent of the external funding goes to FIPAG and DNA. In 2008, DNA surpassed FIPAG as the largest recipient of external funds. However, in reality the bulk of these funds were projects, which are often not directly managed by DNA. Yet, the number of other agencies that get funding has been increasing in the past few years, mostly because of the decentralization of responsibilities, and the effects of the delegated management framework (where funding is directly channeled to operators). The Provincial Directorates, the Municipal Councils of Maputo and Nampula, and the Ministry of Planning are also recipients of water sector donor funding. Figure 6.10 Nominal External Funding by Recipient Institutions (‘000 MZN) 2.500.000 2.000.000 CRA 1.500.000 ARACENTRO ARASUL 1.000.000 FIPAG DNA 500.000 0 2005 2006 2007 2008 Data includes external projects and non-project funding (ASAS) Source: water and sanitation institutions and authors’ calculations. Almost all investment expenditure in the sector is funded by donors. The water sector‘s development budget is highly dependent on donor funding; it is estimated that about 88 percent of the total capital funding is foreign funded. This high percentage of donor funding in a sector that is classified as a priority sector is an issue – especially as the internal component of capital funding tends to be under-spent. A significant proportion of development aid is not on-budget. Data sources are very incomplete, but some data still show that a large part of the funding is not on-budget. Data from the Organization for Economic Co-operation and Development (OECD) database and ODAmoz show little overlap. Different donors in the OECD database are not showing up in the ODAmoz database. 59 This could be a sign of inefficient accounting, but it is likely that some of the funding is off-budget and is actually channeled through NGOs as many of these donors are showing up in the OECD database with very small commitments. As can be seen in Table 6.3, records of bilateral aid in the OECD and ODAmoz database vary significantly. Table 6.3 Bilateral Aid data from various sources (US$ million) Bilateral Aid 2005 2006 2007 Bilateral aid from OECD donors – as recorded by ODAmoz 10 72.6 189.9 Bilateral aid from OECD donors – as recorded by OECD 68.8 81.2 33.2 ODAmoz as a % of OECD records 15% 89% 571% Note: In 2007, about $88 million from Italy (off budget), $27 million from the Netherlands (on-budget) and $39 million from EC (on-budget) were not picked up by the OECD database. The number of non-traditional donors is increasing. The number of donors from countries that do not belong to the OECD is increasing rapidly. These donors include China, India, and the Southern African Development Community (SADC). These donors tend to provide funds as they are registered in the DNA reports but, apart from India, no details on the actual funds that have been disbursed are available. Table 6.4 Execution of projects from non-OECD donor countries (‘000 MZN) 2005 2006 2007 Registered Reported Registered Reported Registered Reported on Budget Execution on Budget Execution on Budget Execution India 0 0 22,090 Unavailable 117,829 Unavailable China 5,260 0 0 0 0 0 South Africa 0 0 3,545 Unavailable 0 0 SADC 0 0 52,635 Unavailable 0 0 Source: DNA 60 7. HOW EFFICIENT IS SPEND 7.1 Allocative efficiency Allocative efficiency measures whether sector expenditures go toward realizing established national objectives and MDGs. The indicators used to determine allocative efficiency are consistent with the basic objectives set out in the sector strategy. Progress in achieving these objectives has been slow. The National Water Policy of 1995 and the subsequent Water Policy of 2007 in the context of the MDGs and the experiences with sector reform include the following objectives: (i) achievement of the MDGs in the rural and urban water and sanitation sector; (ii) consolidation of the decentralization process in the sector; (iii) improve the sustainability of urban water systems in the long run (through, among others approaches, the use of a minimum level of service in water and sanitation access and a focus on cost recovery in tariff setting); and (iv) adoption of adequate hygiene practices in households, communities, and schools. Achieving the MDGs is an ambitious goal that is unlikely to be met under current scenarios. While access to improved water and sanitation is increasing, as was shown in Chapter 4, there is a gap between the current trend and what will need to achieved by 2015, according to the Joint Monitoring Program. While sector expenditure, though volatile, has seen an upward trend in the last five years, it is unlikely that the current levels of investment in the sector will be sufficient to meet the sector objectives of achieving the MDGs. Figure 7.1 MDG Targets for Water and Sanitation and Sector Expenditure in Various Years Source: Expenditure in PARPA Priority Sectors (MoF) and UNICEF-WHO Joint Monitoring Program The total water budget is below the threshold guideline of 5 percent of total budget that was set by the Government. Even though funding has increased over time, the sector has been less able 61 than other priority sectors to increase its budget correspondingly. Expenditure has never reached the 5 percent government guideline for the sector. According to Poverty Reduction Action Plan (PARPA) guidelines, expenditure in the water sector should be reaching between 5.3 and 5.9 percent of total budget expenditures during the period between 2006 and 2010. This is well below countries in the region such as Tanzania, which devoted 6.5 percent of its government budget in the 2007/08 fiscal year, whereas in 2008 only 3.5 percent of the Mozambique Government‘s budget was spent on water. The process of decentralization is still in its infancy in the sector. Up to now, most of the decentralization in the sector has been effectively a de-concentration of expenditures whereby the Provincial Directorates and districts absorb small amounts of funds that previously were spent by central agencies. So far, only DNA has channeled funds through Provincial Directorates. With the competence of water and sanitation service provision accruing to the Provincial Directorate of Public Works and Housing, the degree of control of DNA on provincial expenditure is diminishing. In the case of urban water supply, decentralization has not altered the delegated management framework and FIPAG has thus far been unaffected. Yet, decentralization makes it more important to have an overall sector mechanism in place that can help to make the trade-offs between subsectors and agencies. Progress has been made in achieving the goal of sustainable urban water supply, but a long road still lies ahead. In the utilities that are operating under the delegated management framework, the capacity to cover basic operation and maintenance costs through revenues has increased over time. On average, the 13 utilities have seen progress. Unfortunately, this data does not include performance data on how the country‘s largest water utility is doing. The current levels of cost recovery have implications for the effectiveness of the Government in expanding access to water supply and sanitation services, as too much funding is needed to ensure basic access to water services for existing users. By just being able to pay for basic operation and maintenance costs, the sector does not generate enough funds to pay for basic rehabilitation, let alone expansion of services. Although this is rather common phenomenon in countries building up their urban water sector, it may raise equity concerns with large sums of funds going to the urban water sector. Improvements in adoption of adequate hygiene practices are difficult to measure, but using proxies for this progress has been made. There are a number of different agencies with fragmented responsibility for hygiene promotion including the Sanitation Department of DNA, the Ministry of Health and the Ministry of Education for school sanitation. Budgets for hygiene promotion are difficult to separate from sanitation budgets. Hygiene practices are not monitored systematically and hence it is difficult to measure progress. However, open defecation as a practice has been declining, suggesting that progress has been made in adopting more appropriate hygiene practices. Between 2000 and 2008, reliance on open defecation as a means of sanitation decreased from 57 percent to 42 percent. In summary, significant progress has been made in achieving the Government’s basic objectives. As progress has been made, it is also easy to see that there is still a long road ahead in achieving these goals. It is another reminder that sector reform takes time and is a process that requires many years to complete. 62 7.2 Technical Efficiency Technical efficiency measures whether funding is spent efficiently given the allocative decisions. To that effect, we will use two indicators to measure technical efficiency: (i) budget efficiency through budget execution rates; and (ii) the efficiency of investments. 7.2.1 Budget Efficiency Despite increases in the past years in the total water sector budget, actual expenditures fell below the budgeted estimates. Although actual expenditure has increased rapidly, it has generally lagged behind budget allocations. The sector shows a rather volatile performance in the execution of expenditure. In the past four years, the budget execution rate of the sector that is registered on- budget has vacillated significantly – between and among the different sector agencies. Large differences in the ability to spend recurrent expenditure exist between the different water agencies. On average, recurrent expenditure is slightly below the allocated budget, with big differences in budget execution patterns between the key water agencies. FIPAG has consistently received and executed much more funding from the Treasury than it budgets for – either because of under-budgeting its recurrent funds, or because of an over-allocation of government funds throughout the year, or a combination of both. In 2008, FIPAG spent more than twice the recurrent budget allocated in the State Budget. In the case of DNA, the opposite is happening. Execution rates of recurrent expenditure in DNA have constantly been only partially used with a mere 70 percent in 2008 being spent on recurrent expenditure. This would seem to imply that more than one third of the recurrent needs of DNA (salaries, etc) are not realized. The reasons for this low recurrent budget efficiency are linked to the lack of financial autonomy of DNA. The National Water Directorate has to bid for a recurrent budget with other agencies within MoPH, which decides the amounts that will be allocated to the different MoPH agencies. DNA shows low execution rates of its recurrent expenses when compared against the budgeted amounts requested to MoPH, but when compared against the amounts actually disbursed by MoPH, DNA execution rates are much higher. Because the MoPH allocation does not cover the full extent of DNAs recurrent costs, DNA makes use of external project and non-project funding to fund part of its recurrent costs. Yet, the fungibility of DNA external funding may be creating an incentive and the fiscal space for MoPH to allocate more resources to other directorates under its jurisdiction. 63 Table 7.1 Performance Budget Execution Ratios show wide variations on an annual basis 2005 2006 2007 2008 Recurrent Expenditure DNA 93% 80% 432% 65% FIPAG 215% 173% 163% 229% ARA-Sul 93% 97% 89% 90% Recurrent Expenditure 203% 167% 165% 178% Development Expenditure – Domestic Resources DNA 53% 79% 65% 71% FIPAG 100% 100% 100% 100% ARA-Sul 65% 97% 76% 90% Development – Domestic 60% 83% 75% 81% Development Expenditure – External Resources DNA 65% 33% 53% 48% FIPAG 89% 66% 63% 99% ARA-Sul 97% 99% 94% 98% Development – External 76% 51% 58% 65% Source: DNA Delays in budget implementation are mainly concentrated in executing the development budget. Development expenditures tend to be used with more discretion and are the first to be affected when budget cuts need to be made. In addition, because of the high donor dependency in the development budget, budget implementation is largely determined by how efficiently external funds are being spent. As can be seen in Table 7.1, budget execution rates are lowest and most volatile for donor-funded development expenditure. Donor funding, the major source for funding the development budget, shows that actual aid disbursements lag significantly behind aid commitments. Although donor commitments have increased over the past years, actual disbursements have lagged behind. During the last three years, on average only 40 percent (according to ODAMOZ) of the external development funding budgeted was disbursed, whereas this disbursement ratio fluctuated widely over that period 31. In addition 31 Actual donor funding from OECD reporting is, in general, underestimated, as the donor commitments and disbursements of multilateral agencies tend to be significantly underreported, while funding through NGOs is only registered when NGOs are funded by official donors. 64 there were wide variations between water sector agencies, which suggest that some are better in disbursing funds than others. Figure 7.2 Donor funding – cumulative actual commitments and cumulative actual disbursements (In MlnUS$) Source: ODAmoz Database of Donor Funding Unpredictability of donor funding is a major issue and an important reason for the delays in implementation. It is reflected in the quarterly aid disbursements. Quarterly disbursement of donor funding fluctuates over time, with no discernable pattern as to when funding will be available. If aid were distributed more efficiently, the pattern of quarterly disbursements would show a much smoother pattern, as reflected in the moving average in Figure 7.3. This lack of predictability is also shown in a recent World Bank study that discussed how the unpredictability of aid, as measured by the difference between aid commitments and disbursements, can be harmful, as every aid dollar withheld reduces government investment by 12 percent, while every dollar unexpectedly given reduces government consumption by 64 percent.32 Many different factors underlie this erratic pattern of disbursements, but it is clear that when implementation is smooth, efficiency of government spending in a sector where donor dependency is high could increase significantly. 32 Celasun, O, and J. Walliser, "Predictability of Aid: Do Fickle Donors Undermine Aid Effectiveness?" Economic Policy, 23, 545-594 65 Figure 7.3 Actual Disbursements on a Quarterly Basis in US$ million Source: ODAmoz data and authors’ calculations The reasons for unpredictability of donor funding are related to inefficiencies in budget administration and processes of the Government of Mozambique and donor countries. The Performance Assessment Framework of the G19 donors shows that predictability is low, as measured by a set of indicators.33 Unpredictability is linked to: (i) the planning and implementation of donor funding which is not aligned with the Government‘s budget calendar; (ii) the use of parallel systems that make it hard to obtain full information on the expected assistance flows; and (iii) the seasonality in project implementation. In addition, a survey of aid donors34 mentioned that 29 percent of delayed or lost disbursements were due to administrative problems in donor countries, because of re-allocations of donor funding and administrative delays. The high unpredictability of recurrent and development funds can provide perverse incentives to the sector. A much higher actual allocation than the initial one (as FIPAG benefited from in 2007, for instance) can constitute an important source of inefficiency, as the spending institution will not necessarily be able to ensure that the additional funds can be translated without much notice into additional outputs and outcomes. On the other hand, a much lower allocation is likely to have an adverse impact on service delivery and will reduce the future capacity of the sector to execute funding. Such is the case of DNA, where funding in future years is determined by previous capacity to spend funds. The succession of higher and lower than expected allocations renders strategic and 33 These indicators include (i) % PAPs with multi-year agreements of not less than 3 years, (ii) Commitments of General Budget Support (GBS) for year n+1 made within 4 weeks of the Joint Review Process in year “n” ; (iii) Disbursement of confirmed GBS commitment in the fiscal year for which it was scheduled, according to quarterly disbursement schedule as agreed with GoM ; (iv) % of PAPs Overseas Development Assistance (ODA) that is recorded in the government budget; and (v) PAPs ODA disbursed as percentage of aid recorded in government budget. 34 Celasun, O, and J. Walliser, "Predictability of Aid: Do Fickle Donors Undermine Aid Effectiveness?" Economic Policy, 23, 545-594 66 operational planning ineffective. In addition, because the predictability of funds hits all types of expenditures, and as such, there is no guarantee that the recurrent costs will be covered, the agencies in the sector have an incentive to divert funds from their development budgets to supplement their recurrent budgets. Under this dynamic, the first and foremost preoccupation of implementing agencies might not be service delivery per se, but securing the financial viability of the agency in the short-run. The gaps between budget allocations and actual expenditure are caused by a combination of factors. As will be discussed in the remainder of this chapter, the divergence between budgeted and actual expenditure is due to a combination of issues in the budgeting, planning, procurement, and disbursement processes. 7.2.1.1 Budget Procedures Concerns surrounding the comprehensiveness and credibility of the budget center on the fact that budget allocations can be altered upwards and downwards at any point during the fiscal year. According to assessments conducted in 2004 and 2006, budget credibility is the only dimension of Mozambique‘s Public Expenditure and Financial Accountability that worsened between 2004 and 2006.35 The release of domestic sourced sector funds can be unpredictable due to these budget reallocations. When funding is unpredictable there is a gap between strategic planning and finance, because planning is based on certain assumptions about the timing and availability of finance. When the assumptions are not realized, it is easy to divert from strategic goals and targets. Part of the unpredictability is due to the discretion the Ministry of Finance has to alter the Budget Law approved by the Parliament in December each year. This commonly implies that the initial allocation can be altered throughout the fiscal year. The amended allocation can be above or below the initial one and the sector is not notified in advance of the deviation, as is illustrated with data from ARA- Sul whose domestic funding allocation varies significantly over the years. This high flexibility is obviously detrimental for implementation and is a direct consequence of the absence of planning to ensure that resources are allocated in a prioritized manner, and the need to adjust due to delays in the release of external funds. 35 Mozambique PEFA Reports 2004 and 2006. 67 Figure 7.4 ARA-Sul’s Allocations of Domestic Funding (‘000 MZN) Source: CGE 2000-2004 (Ministry of Finance) The registration procedures of externally-funded projects on-budget are partly responsible for low budget execution rates. The Budget Law authorizes the Budget Directorate to register new external projects on the budget anytime throughout the fiscal year. The Ministry of Finance has requested that donors work with the beneficiary sectors so that projects can be registered in May of each year, at the start of the budgeting process for the subsequent year. Donors should also forecast the time when project execution will start, as well as the amount of resources they will disburse to minimize distortions in the sectors‘ and Treasury‘s financial programming. Donors tend to register projects on-budget well before they start implementation (sometimes the gap is up to 9 months) which brings sector‘s execution rates down. The reason for this early registration is linked to tax exemptions. Once projects are registered on-budget, quarterly execution reports need to be submitted before the government‘s quarterly execution reports are published (i.e., 45 days after the end of the quarter). Yet, this data often is delayed, resulting in information not being included in the official quarterly execution reports – and possibly underestimating the actual flows in the sector by quarter. 7.2.1.2 Investment Planning The link between investment planning and budget is weak. There is little harmonization between long-term strategies, medium-term policies, and annual budgets. This problem is reinforced by an overall lack of investment plans to prioritize interventions according to resource availability and recurrent cost implications at the subsector and sector level. Each subsector agency needs to plan its investments to ensure that the investment plans are linked to (sub) sector strategies. Yet, at the same time, the sector leader will need to compile the different (sub) sector strategies and coordinate planning and budgeting among the different sub-sectoral agencies while making explicit trade-offs between the sub-sectors. Even though DNA is the sector leader, its planning department is not 68 making many of the trade-offs (such as the balance between urban and rural water supply, or between water resources and water supply). The fragmentation of donor funding further exacerbates the weak links between planning and budgeting. To align projects with the government planning and budgeting cycle, by March of each year, donors should report three-year forecasts of disbursements that can be integrated in the Medium-Term Fiscal Framework proposal submitted by the Ministry of Planning and Development to the Council of Ministers. This is especially important for multi-year projects to ensure proper planning. However, much work needs to be done to get this information included in the budgets. Although progress has been made as more sector strategies have been developed, the fundamental weaknesses of sector plans have not been addressed. These include the lack of reliable budget projections for future years, infrequent use of economic and financial tools in strategic planning (currently based on proposed activities not linked to budget commitments) to determine the costs and benefits of different investment projects, and the lack of consideration of the recurrent cost implications of investment projects. A recent assessment of fiscal transparency practices in Mozambique36 found that estimates of new revenue and expenditure initiatives and the recurrent costs of government policies are not clearly identified in the budget documentation. In addition, the high number of plans and strategies does not help to clarify sector planning. The concurrence of many central planning documents for water supply and sanitation37 (Visao 2025, Plano Quinquenal do Governo, PARPA, PES, MTEF with specific proposals for each institution/agency in the sector, Politica de Aguas, PESA-rural, Programa de Agua Rural) with additional planning documents at the provincial and local levels, does not facilitate sector investments, as all these planning documents differ in their objectives, assumptions, and targets. Hence, coordination is clearly needed to ensure that the sector has an integrated approach that balances the interests of the different subsectors with that of the total sector. 36 International Monetary Fund. 2008. “Republic of Mozambique: Report on Observance of Standards and Codes – Fiscal Transparency Module.” IMF Publication Services (Washington, D.C.). 37 Despite the large number of planning documents, the sector still lacks a formalized urban water sector strategy and the largest share of sector expenditure is allocated to the urban water sector. 69 Figure 7.5 Planning and Control Instruments in Mozambique’s Water Sector National Level Water Sector Visão 2025 MDGs (long-term) PQG (5 years) PARPA (4 years) Rural Strategy Urban Strategy WRM Strategy CFMP (under (3 years) preparation) PES & Budget (OE) FIPAG Inputs (annual) DNA Source: Based on IMF, “Republic of Mozambique: Report on Observance of Standards and Codes” As a result of this fragmentation in planning, the sector does not have an integrated long-term plan. The sector has not yet presented an integrated, sector-wide, MTEF proposal to the Ministry of Planning and Development. The two main sector agencies have each presented their own proposals. In addition, DNA has developed a rural water and sanitation program whereas an urban program is under development. Given the institutional set up of the two key sector institutions (the combination of the DNA, ministerial directorate, and FIPAG, a fully autonomous asset holder) the formal status, attribution of specific responsibilities and enforceability of the strategic plans, PARPA and how they are interlinked, remains largely unclear. 7.2.1.3 Procurement The sector deals with severe procurement bottlenecks. There are no analyses reporting compliance rates of Mozambican institutions with procurement regulations. Yet, a recent Inspectorate General for Finance (IGF) audit concluded that sector institutions were failing to comply with the country‘s procurement regulation (Decree 42/89 of 28th of December 1989). The 2008 Country Procurement Assessment Report mentions four different elements that explain the procurement bottlenecks.38 The first element is the deficiencies in the legal and regulatory framework. Despite significant improvements in the legal and regulatory framework for public procurement, including a new legal 38 Republic of Mozambique and The World Bank Joint Report. Update of the Country Procurement Assessment Review (CPAR) Draft Report. June 2008. 70 framework introduced in 2005 (Decree 54 of December 2005: Country Procurement Regulations) as part of the SISTAFE law, several weaknesses remain. These include onerous requirements for foreign bidders and a lack of clarity in some areas of the bidding process, a lack of transparency, in particular through the absence of an independent forum through which to appeal the decision of the procurement entity. The second element is the lack of a link between procurement planning and budget preparation. This may be a contributing factor for the high variances between budget allocation and actual expenditure and the poor quality of future projections. A third reason for bottlenecks in procurement is the lack of capacity to conduct procurement. The capacity to conduct procurement remains weak, especially at the decentralized level. Given the weak capacity, the procurement regulations are not consistently applied and the quality of records is not up to standards. The lead time for procurement is quite long and one of the bottlenecks often quoted is the length of approvals that need to be obtained from the Supreme Audit Institution (Tribunal Administrativo, TA) – according to the law, the TA approves all government-financed contracts irrespective of size. Finally, information on procurement is scarce. A final issue is the transparency of public procurement. Public procurement is frequently cited as an area most susceptible to corruption. While Mozambique has an Anticorruption Strategy, and there are provisions in the 2005 procurement regulations, there is not much evidence of prosecution of corrupt practices. The procurement delays result in high costs. Slow procurement has large economic impacts on the budgets and on performance targets of the sector agencies. These costs are explicit in the form of increased penalties and arrears, because of the modification and re-negotiation of contracts. In addition, procurement delays result in inefficiencies, such as lack of transparency and accountability, and poor procurement management, that together increase the cost of procurement and, in turn, of the services delivered. The opportunity costs (forgone returns to investment) of delayed completion of projects can also be substantial. 7.2.1.4 Audit Processes The use of multiple financial systems in the sector results in an inordinate administrative burden for the existing sector agencies. Because of the heavy dependence on donors, the sector uses three different financial systems: (i) SISTAFE; (ii) projects implemented by sector agencies outside SISTAFE; and (iii) donor project implementation units (PIUs) using donor systems. Each of these three systems are governed by different planning, control, procurement, implementation and audit rules that constitute a heavy administrative burden on the sector‘s already limited human resource base. Yet, the importance of SISTAFE is increasing. SISTAFE has been established in FIPAG, ARA-Sul and, after substantial negotiation by the joint donor-sector water and sanitation working group, a unit for the direct execution of funds (with some limited functionalities) has also been created in the DNA, which previously had to share its access to SISTAFE with other departments in the MoPH. The constraints in human resource capacity in sector institutions are further complicated by the different reporting requirements for the various agencies. Monitoring is complicated by the 71 fact that the water sector is funded through many different donors which very often depend on their own reporting systems. In a country with scarcities of skilled staff, the multitude of donors and variations in their procurement, disbursement, audit and monitoring requirements puts a heavy administrative burden on the few staff available. DNA, for instance, had 24 different donors to deal with from 2003 to 2008. The first performance audit concluded that it was impossible to follow the audit trail.39 The link between physical outputs, contracts, and financial accounts is non-existent, which makes it impossible to determine the efficiency of investments. 7.2.1.5 The Value Added Tax Tax arrears and government counterpart funding (“co-participation”) has been another bottleneck. Within the sector, the problem of Value Added Tax (VAT) arrears affects DNA, FIPAG and ARA-Sul. While on-budget projects are not subject to import duties (which may explain the phenomenon of registering projects on-budget before it is sure that they will materialize), they are subject to VAT and government counterpart funding.40 An inventory of the country‘s accumulated arrears in VAT and counterpart funds41 amounted to MZN 457 million (equivalent to 26 percent of its 2006 expenditures). Table 7.2 shows, that the institutions bearing the brunt of VAT arrears and counterpart payments are those benefiting from large funding. Table 7.2 Accumulated Arrears in VAT and Counterpart Government Contributions up to end-2006 VAT Arrears Counterpart Taxa Liberatoria Other Total Funds (Consultancies) DNA 16 75 - - 91 FIPAG 20 153 7 - 180 ARA-SUL 172 14 - 0 186 Total MZN (millions) 208 242 7 0 457 Total US$ (millions) 8.3 9.7 0.3 0 18.3 Source: DNA Note: data in millions of Meticais except last line, in US$ Millions. The reasons for the VAT and counterpart arrears are mainly the result of inadequate planning.42 Many donors that provide funds for infrastructure projects do not accept the payment of fiscal charges which recipient governments would impose under normal circumstances. Since recipient governments are trying to limit fiscal exemptions, they pay these taxes out of general government funds. In an environment where resources are scarce and where there are many donors, 39 Auditoria de Desempenho do Sector de Aguas (2006), IGF, Ministério das Finanças. 40 The Government counterpart funding varies by donor. 41 Payment of taxes (or accumulation of arrears) is difficult to obtain as it is not published in the official budget execution reports. 42 This section draws on Dieter Orlowski, “Donors’ principles and resulting stress: Indirect taxes and donor-funded public works”. Oxford Policy Management, 2007 72 financial planning requires timely information. VAT and counterpart funding is funded through the domestic investment component. When funding falls short, recurrent budget expenditures are usually prioritized with too-little funding available for VAT and counterpart funding. Even if sufficient funding is available at the beginning of the fiscal year, reallocations during the budget year may result in reductions in the funds available for VAT and counterpart funding. If the sector agencies are unable to secure sufficient revenues, projects will incur delays, as donors keep on disbursing funds despite insufficient counterpart and VAT funding being available. These delays result in increased payback installments, longer amortization periods, and forgone income for the country and the funding agency. Tax arrears have had pervasive effects on the sector. Sector institutions have historically incurred large arrears of VAT and investment co-participation which has translated into the contractors not fulfilling their tax obligations with the State. This has had pervasive effects: (i) project implementation has been halted, with the delays typically having an explicit cost-augmenting effect (penalties, interests, storage, cancelation of contracts, etc); and (ii) contractors have come to factor potential delays into their cost calculation, which can have an inflationary impact on the costs of infrastructure projects. Some partial measures have been implemented to mitigate the accumulation of VAT and counterpart funding arrears. One of these measures includes the partial redemption of the sector‘s VAT and co-participation arrears by the Treasury. This has mitigated the debt burden for water institutions but has not yet addressed the structural causes of arrears accumulation. A more direct measure to deal with the accumulation of arrears is the partial exemption of VAT on inputs of large water projects. This measure benefits large projects that tend to be disproportionally managed by foreign contractors and suppliers. Yet, water and sanitation projects funded with domestic resources (which could be implemented by Mozambican companies) would not qualify for such exemptions, which results in a serious bias against local firms. A third measure is the full assumption of the project‘s tax burden by the donor agency, which has gone a long way in redressing further accumulation of arrears in new infrastructure projects. To deal with the VAT and counterpart arrears will require better planning. Obviously, better planning at the level of the sector agency, the Budget Directorate, and donor agencies is required to ensure that the predictability of funding improves significantly. It will also require a realistic assessment of absorption capacity in the sector. If the large levels of donor funding cannot be adequately matched by domestic resources to fund VAT and counterpart funding, an assessment should be made whether to relax the need for domestic resources or reduce the levels of donor funding that correspond with the availability of domestic resources. Flexibility is needed on the part of the donors to adjust their rules that were implemented in an environment where balance-of-payment gaps and not budgetary constraints were the main focus of donors. Full assumption of the project‘s tax burden by the donor agency can go a long way in redressing further accumulation of arrears in new infrastructure projects. The latter measure would accelerate implementation, avoid special rules and their potential for abuse, and also reduce the cost of water sector infrastructure. Another option is that in case donors are unable or unwilling to make these changes, to ensure that the disbursements of donor funds are linked to the disbursements of counterpart funding and VAT. 73 7.2.1.6 Donor Fragmentation The above-mentioned inefficiencies are exacerbated by the high levels of donor fragmentation. According to the ODAmoz, between 2003 and 2008 the sector institutions benefited from at least 78 projects. There were 24 different donors active in the sector (excluding NGOs). DNA had to deal with the largest number of donors – 19 different donors in 2008. Yet, with the decentralization process starting, the number of sector agencies with funding has increased. Almost all of these donors use project funding as their main vehicle.43 Another indicator of donor fragmentation is the small size of projects. The ODAmoz database shows that although the average project size was slightly over US$4 million, the median project size is about US$ 3.5 million; hence, the sector deals with a large number of very small projects which adds to the transaction costs. This large fragmentation also places a very heavy administrative burden on a sector with serious capacity constraints, thus shifting the focus of the sector‘s central institution (DNA) away from sector-wide functions such as coordinating, planning, and M&E, into project management. Figure 7.6 Number of water sector projects by financial size as measured in US$ (as of May 2009) Source: ODAmoz 43 The water sector has the smallest share of sector budget support of all priority sectors. Currently, there are multi- donor “common funds” in the following sectors: agriculture, health, education, tax authority, Administrative Tribunal, Technical Unit for the Reform of the Financial System, and Technical Unit for the Reform of the Public Sector, yet none exists in the water sector. 74 Table 7.3 Number of Donors and Institutions benefiting from donor funding to the water and sanitation sector (2003-2008) Total Bilateral Multilateral Sub-sectors number of donors donors donors DNA 24 14 10 All FIPAG 7 3 4 Urban Water ARA-Sul 6 3 3 Water Resource Management ARA-Centro 5 3 2 Water Resource Management CRA 5 2 3 Regulation Operators* 2 2 - Urban Water DPOPHs* 5 5 - Rural Water and Sanitation Municipal Councils* 2 2 - Urban Water and Sanitation Other* 1 1 - Urban and Rural Water and Sanitation Source: DNA and ODAmoz Note: Data with a star (*) is for the years 2006-2009. Other includes other institutions benefiting from funding to the sector including the Maputo operator (Aguas de Moçambique), the provider of technical assistance (Vitens) on other towns under the DMF, provincial directorates, the Municipal Councils of Maputo and Nampula and the Ministry of Planning. Box 3 The Effects of Fragmentation of Donor Funding in the Water Sector In contrast to the aid profile in most sectors in Mozambique, where donors are increasingly channeling their aid through the single Treasury account, most financial assistance to the water and sanitation sector is either managed by donor agencies or transferred from donor accounts to sector accounts in commercial banks (charging administration fees and commissions). These accounts are controlled by donors or by staff in beneficiary institutions with little oversight by the Ministry of Finance. When funds are not channeled through the single Treasury account, the Government does not know its cash position and hence cannot transfer funds in a timely manner to deliver public services. This lack of information is exacerbated when the projects are not registered on the State Budget. This practice results in liquidity shortages during which state institutions are unable to purchase goods and services (such as materials for infrastructure maintenance) resulting in slow execution of Government programs and donor projects. On the other hand, it can result in liquidity overflows, where cash sits in bank accounts for months, while simultaneously there are shortages of cash to pay for services financed through different accounts. These liquidity overflows may result in inflationary effects in the implementation of infrastructure projects. Effectively, the fragmentation of donor funds in different bank accounts compromises the value of public spending and accountability. Source: S.R. Goicoechea (2008) “Making Aid More Predictable and Transparent Through the Use of National Treasury Systems: The Case of Mozambique’s Single Treasury Account. A Case Study for DFID Mozambique” 75 Part of the inefficiencies is directly linked to the poor performance of donors in the administration of their funds. Almost all donors in the sector are members of the Group of 19 donors (except for the US-funded Millennium Challenge Corporation44) attempting to improve donor harmonization in the country as signatories of the Paris Declaration on Aid Effectiveness. These G19 donors have signed a Memorandum of Understanding with the Government whereby the Government of Mozambique and the G19 will jointly assess their respective performance each year. The overall improvement in harmonization has proved relatively elusive so far, however, and should be urgently addressed. 7.2.1.7 Project Fragmentation The sector is also characterized by project fragmentation. As a result of the myriad project contributions, the water sector is the most project-based priority sector. It is also the one with the smallest share of sector budget support of all priority sectors. While ASAS funding is not ear- marked and DNA has discretion to decide on its allocation across DNA‘s portfolio of responsibilities, the only donor contributing budget support funds is the Government of the Netherlands. A major element of the poor performance is due to inadequate planning which hinders implementation and severely affects accountability. The practice of registering projects on- budget with little planning of project funding availability is a major problem. There is evidence of projects taking more than 15 Quarters (i.e., close to 4 years) to kick- start execution of implementation. The unpredictability of aid and high level of fragmentation contribute to the procurement bottlenecks that the water sector is facing. Procurement bottlenecks are severe, as discussed earlier in this Chapter. Because of the unpredictability of funds, water sector agencies tend to procure on a more ad-hoc basis, which may result in smaller-sized procurement contracts, a lack of possibilities to bundle smaller procurements and realize economies of scale, and increased procurement transaction costs. In addition, the high number of donors and projects which the sector has to deal with and the subsequent large number of different procurement, audit, and other reporting requirements, puts a high burden on the human resource capacity of the sector. Pooling resources may reduce the transaction costs of donor funding. Given the sector‘s high dependence on external funds, a solid public expenditure management system requires that donors improve the predictability of their support and make progress on streamlining and harmonizing administrative procedures. Registration from donor funding should be improved upon, both at the Ministry of Finance and at donor level. Increasing the size of donor funding will help to reduce the transaction costs of managing donor funds. The current focus on multi-donor initiatives that pool funds to provide general support for a sector-wide program of interventions may be a good alternative to deal with the costs of aid fragmentation. 44 MCC funds are directed to the Ministry of Planning and Development and registered in an ad-hoc budget line. As a result, the percentage of the funds going to the sector that will not be managed by water sector institutions will increase dramatically. 76 7.2.2 Efficiency of Investments It is difficult to measure the efficiency of investments in the sector for a set of reasons. The funding data, as mentioned before, are incomplete. Moreover, as has been discussed previously, a significant part of the funding is off-budget which means that any analysis of expenditures tends to be incomplete. Off-budget funding results in “miracle premiums” making it very difficult to determine the actual links between expenditures and outputs and outcomes. When on-budget investment funding is not readily available, the large levels of ―off-budget‖ funding make it difficult to assess the link between expenditures and outputs and outcomes. The off-budget funding results in a so- called ―miracle premium‖ for the sector. When on-budget funding increases, it is not clear whether the total level of funding increases. If the total funding (whether it is registered as on-budget or whether it is off-budget) remains the same, the impact of access is by definition limited. As such, this analysis tends to be indicative only and will focus on the water supply and sanitation sector due to difficulties in tracking water resources management (WRM) outcomes over time. We will measure the efficiency of investments using two indicators: (i) the relationship between the volume of expenditures and the improvements in access; and (ii) the subsidies channeled through the sector. These indicators are indicative only due to the issues with the data availability as mentioned above. Figure 7.7 Water Spending in Real MZN million Source: Expenditure in PARPA Priority Sectors – Water and Sanitation (MoF) & Household Surveys (INE) 77 The link between the expenditures and improvements in use of improved sources has so far been weak. There are many reasons for this performance. An increase in funding usually does not translate into a rapid increase in use of water supply services due to the long preparation and implementation periods of many projects. In the case of Mozambique, matters are further complicated by the fact that much of the investment is focused on rehabilitation of existing infrastructure (which essentially crowds out expansion of access as it focuses on keeping existing infrastructure in working condition) whereas high levels of funds are allocated for capacity building and only affect expansion of infrastructure indirectly and in the longer run. 7.2.2.1 Efficiency in Rural Water Supply and Sanitation High breakdown rates of rural water points shows that the efficiency of rural water supply investments can be improved. In view of the expected lifetime of assets in the rural water sector (i.e., 7 years for an individual water point and 25 years for a piped water system), between 4 and 14 percent should be considered a benchmark for more acceptable levels of performance depending on the mix of individual water points and piped water systems used in rural areas. Obviously, this figure is likely to be significantly below the current level of non-functionality of at least 33 percent. This lack of sustainability in water supply infrastructure is also reflected in the large gap in access to improved water sources in rural areas. Data from DNA estimated that 48.5 percent of the rural population in 2007 (estimated by the total number of water points each of which is to serve 500 people) had access to improved water sources. Yet, the most recent household survey which estimates the use of improved water points (boreholes and protected) wells at only 25 percent in 2007. The large difference is mainly explained by two factors; (i) the inclusion of non—operational water points that are assumed to deliver some water but to a smaller number of people than the norm of 500 people per water point; and (ii) the norms being used to determine access to services are likely overestimating the number of people per water point, as discussed in Chapter 5. 78 The efficiency in rural water supply investments is also adversely affected by the high cost of investment. Inflated price of contracts are a result of insufficient contractor competition, multiplicity of donor-specific tendering and procurement systems, accumulation of tax arrears, slow counterpart funding, and general lack of accountability of expenditure. Although little information is available, it is likely that prices in Mozambique are higher than in other parts of the region. A simple unit cost comparison of a borehole in South Africa and Mozambique seems to imply this is due to market structure (the number of drilling companies is limited in Mozambique) 45 and hydrogeological conditions. This study also mentioned the higher prices for rural water supply equipment in Mozambique, as compared to neighboring countries. Table 7.4 Unit Costs Comparison of a 43m borehole in Mozambique and South Africa Country Cost in US$ South Africa 5,774 Mozambique 7,691 Difference 33% Source: WSP and information from South Africa drilling firms Note: Generator costs are not included, but both countries face serious energy bottlenecks. 7.2.2.2 Efficiency of Investments in Urban Areas A first indicator of possible inefficiencies of urban water systems is the size of the subsidies provided through the urban water sector. The good news is that these operating subsidies declined between 2000 and 2008, as more urban water utilities started to cover basic operation and maintenance costs, because of tariff increases and modest improvements in reducing service delivery inefficiencies. Subsidies come in different forms, such as operating subsidies, maintenance and rehabilitation investments, as well as the investment subsidies that are standard in the sector. Operating subsidies in utilities under the DMF have been declining between 2002 and 2008. This decline is mainly brought forward by the improvements in utilities‘ capacity to cover an increasing part of their operating and maintenance costs through their revenues. Apart from the direct subsidies provided to the urban water sector in the form of operating and capital subsidies, there are also implicit subsidies or hidden costs. These implicit subsidies are caused by the mispricing of water services, collection inefficiencies, overstaffing, and non- revenue water losses. All these forms of inefficiencies in water supply delivery result in distortions. Tariff regimes that do not allow for full cost recovery and collection inefficiencies provide implicit subsidies to existing utility consumers. Non-revenue water losses are also implicit subsidies but do 45 WSP and DNA, Assessment of the National Drilling Sector Capacity for Rural Water Supply in Mozambique. June 2006 by WE Consult, consultants for Water, Environment and Geoservices. 79 not necessarily benefit existing consumers only. All of these types of inefficiencies result in lack of maintenance, under-investment, and deteriorating service levels. Box 4 Hidden cost or Implicit Subsidies Hidden cost (or implicit subsidy) estimates provide a measure of the costs of a policy action (or inaction) that could, but do not necessarily, have a direct fiscal cost. These hidden costs put a value on the failure either to implement public policy or to provide an offsetting payment. When looking at the question of tariff enforcement, a government could fail to enforce tariffs but instead provide the utility with an operating subsidy to offset its failure to collect tariffs. It could also not provide the operating subsidy, resulting in lack of maintenance and eventually the early demise of infrastructure. The intention in developing this model has been to provide a single measure of the combined impact of implicit subsidies to contribute to a better understanding of the impact of reform in the water sector. The secondary objective of this model is to ensure that this measure can be easily calculated, tracked and reported, and furthermore that trends can be monitored and benchmarked across sectors and countries without the need for intensive data collection efforts. The model for estimating these implicit subsidies, the Hidden Costs Calculator, has specifically been formulated to provide an insight into different components that are considered key to the infrastructure sectors: (i) poor rates of bill collection; (ii) excessive losses in water systems (as measured by non-revenue water) that result from operational or administrative inefficiencies from the network; (iii) tariffs that are set below the cost recovery rate (i.e. the amount needed for long-run operations and maintenance costs, plus allowance for investment while assuming a certain level of efficiency by using normative (instead of actual) losses); and (iv) over-employment in utilities. This model looks at the difference between actual revenues and revenues that could be anticipated in a system that is operating with tariffs that cover costs, where bills are paid, where losses are within normal levels expected for networks of that age and design, and with reasonable levels of staff productivity. The size of hidden costs46 in Mozambique’s urban water utilities is significant. The hidden cost analysis is a static calculation of what happens when efficiency improves and prices are at least covering operation and maintenance. In a dynamic context, therefore, changes in efficiency will for instance the need for tariff increases. Yet, the hidden cost analysis helps to understand where significant improvements can be made. An estimate of implicit subsidies was calculated for the 12 utilities under the FIPAG framework for which data were available for 2008 – this excludes the largest utility of Maputo for which data are only available up to 2006, and for which a separate calculation is made. The hidden cost analysis is based on the assumption that each utility would at least cover its operating and maintenance costs If a utility is able to cover its operation and maintenance costs from its operating revenues, it is at least able to guarantee that the utility can supply its current customer base in the short-run. This is a minimum standard. Interestingly, most of these 12 utilities are able to cover their basic operation and maintenance costs. Yet, although covering basic operation and maintenance costs means that the utilities can keep on covering their current client base in the short run, the utilities will not be 46 The hidden cost analysis is a static calculation of what happens when efficiency improves and prices are at least covering operation and maintenance. Yet, improvements in efficiency obviously affect the need for price increases, for instance, so in a dynamic context the impact on total revenues and costs may be different. Yet, the hidden cost analysis helps to understand where most of the improvements can be made. 80 able to cover the cost of replacement let alone increase access to services over time. In addition, it was assumed that each utility would perform in line with the best performing Mozambique utility in terms of employment (the minimum staff ratio in the 12 utilities is 12 employees per 1,000 connections, compared to 1-2 as best practice, and about 6 in Maputo; as best practice we have used the 6 employees per 1,000 connections) and non-revenue water (the minimum non-revenue water (NRW) in the 12 utilities is 25 percent of water production, with global best practice between 15-25 percent). It is further assumed that all utilities are able to collect all their billed revenues. The total hidden cost in the 12 utilities is equivalent to MZN 151 million (about US$6 million). Yet, there is wide variation in the size of these hidden subsidies. In the first generation utilities under the FIPAG framework, the overall size of hidden subsidies is significant but not excessive. Yet, in the second generation utilities under the FIPAG framework, these subsidies are much larger, and the smaller the utility is (in terms of number of connections), the larger the hidden subsidies, especially those linked with underpricing. The lack of economies of scale in the smallest utilities obviously makes it difficult to achieve financial sustainability when the utility is currently having only a very small number of connections. Starting up new utilities in smaller towns will require large subsidies in the short-to-medium-term depending on how demand is evolving in these small towns. In recent years, the increase in towns being managed under the FIPAG framework has increased rapidly. Yet, in view of the large subsidies needed to support utilities in small towns, the need for expansion should be balanced against concerns of financial viability. Data for Aguas de Mozambique (Maputo) shows that the pattern of hidden costs in Maputo in 2006/07 is similar to that of the 12 utilities with regard to NRW, which is very high in Maputo at 57 percent. Labor productivity shows a more positive trend, as Maputo, through its sheer size, can capitalize its economies of scale much better than many of the other utilities. Labor productivity is 6 staff per 1,000 connections (which is still high compared to best practice but about half of what is observed in the 12 utilities). Collection efficiency is lower than in the 12 utilities at about 85 percent, whereas the low operating cost coverage ratio shows that underpricing is still an issue in Maputo. As a result, the hidden costs in Maputo are high and amount to more than MZN 560 million (about US$21.7million). What is clear is that these hidden costs are sizeable with regard to the actual revenues generated by the water utilities. The biggest gains could be made if the utility providing water and sewerage services in Maputo and its surrounds, would be able to improve its performance. 81 Figure 7.8 Hidden costs in Urban Water Utilities (excluding Maputo) against Best Practice in Mozambique Source: FIPAG, 2008 data for 12 utilities Figure 7.9 Hidden costs in Maputo against Best Practice in Mozambique (2006) Underpricing of piped water services is a source of implicit subsidies when urban utilities are benchmarked against the best performing utility. Tariffs do not always cover basic operation and maintenance (O&M), although the trend toward O&M cost recovery shows a positive trend. In 82 February 2010, nominal tariffs were increased by 20 percent, although in real prices, tariffs are declining over time. Underpricing is not the most important issue in the 12 smaller utilities (although the individual utilities show large differences in the level of underpricing – with some larger utilities able to cover more than the operation and maintenance costs and others much less), Yet, the current water tariffs are relatively high; in Maputo and the first generation utilities put under the FIPAG framework, median revenues per cubic meter of water sold (as a proxy for tariffs) of about $0.70 (in 2008, the average exchange rate stood at MZM 24.3 per US dollar), which is a price level relatively commonly seen in many African utilities. Most of the hidden costs are linked to the high levels of non-revenue water. It is likely that a significant amount of non-revenue water is linked to inadequate billing systems. The data from the 2007 Census for Maputo City and Maputo Province (Matola) suggest that large gaps exist between the billing data that the utility provides and that arising from the household surveys and the census. According to the Multiple Indicator Cluster Survey (managed by UNICEF),47 more than 1 million people of the total population of 2.2 million people in Maputo City and Maputo Province,48 had access to piped water, compared to only 0.64 million people with access to piped water as registered by the utility. Part of this difference may be made up by urban populations with access to piped water outside of Matola City, but it is likely that this urban population is rather small. Another part of the difference may be made up by vendors and small-scale private operators that use piped water as their source, while illegal connections may also explain part of the difference. Obviously more detailed data will be needed from the Census that can help explain this large difference. Yet, these data seem to suggest that the billing system of Aguas de Mocambique may require updating – a conclusion that is also supported by the high levels of non-revenue water that characterizes the utility. 47 The Census data are not yet sufficiently disaggregated with regard to piped water supply service provision, but the numbers of the census are at first sight higher than that of the latest MICS household survey 48 The urban population in Maputo Province is 828,678 people according to the 2007 Census. Of these, the majority (675,422 people) live in Matola City adjacent to Maputo City, and part of the service area of Aguas de Mocambique. The remaining part of the urban population lives in district towns. 83 REFERENCES Africa Infrastructure Country Diagnostics, 2008. Water Supply and Spending Needs Model. Washington D.C. (mimeo). Banerjee. S. et al., 2008. , Ebbing Water, Surging Deficits: Urban Water Supply in Sub-Saharan Africa. AICD, Background Paper 12 (Phase 1), Washington D.C. Briceño-Garmendia, Cecilia, Karlis Smits, and Vivien Foster, 2008. Financing Public Infrastructure in Sub-Saharan Africa: Patterns and Emerging Issues. AICD, Background Paper 15 (Phase 1), Washington D.C. C.N. Castel-Branco. 2008. ―Mozambique Programme Aid Partners Performance Reviews 2006 and 2007.” IESE, Maputo. Celasun, O, and J. Walliser, "Predictability of Aid: Do Fickle Donors Undermine Aid Effectiveness?" Economic Policy, 23, 545-594 DNA & WSP. 2006. ―Assessment of the National Drilling Sector Capacity for Rural Water Supply in Mozambique‖. DNA. 2009. Cenário Fiscal de Médio Prazo 2008-2011 and 2009-2012, Propostas, Maputo. DNA and Water and Sanitation Working Group (Joint Note 27/03/2007). DNA, Análise do Relatório de Execução do Orçamento de Estado de 2008, Sector de Aguas FIPAG. 2008. Brochura Quadro de Gestao Delegada, Maputo. FIPAG. 2009. Cenário Fiscal de Médio Prazo 2008-2011 and 2009-2012, Propostas, Maputo. FIPAG Consolidated Income Statement up to 2007 (―Profit and Loss Statement‖), received March 2009. Foster, Vivien and Cecilia Briceño-Garmendia (editors), 2008. Africa Infrastructure: a time for transformation. AICD, Washington D.C. Goicoechea, S.R. 2008. ―Making Aid More Predictable and Transparent Through the Use of National Treasury Systems: The Case of Mozambique‘s Single Treasury Account.‖ DFID, Mozambique. Governo de Moçambique, Conselho de Ministros, Cenários Fiscais de Médio Prazo, 2008-2011 and 2009-2012. Governo de Moçambique, Contas Gerais do Estado (2001-2005) (extracts), Ministerio das Finanças. Governo de Moçambique, Diploma Ministerial xx/2001 and 23/2004. Governo de Moçambique, Ministerio das Finanças, ATM (2008), Apresentaçao sobre o Acordo Fiscal Relativo ao Programa Millennium Challenge Account – Moçambique, Maputo. Governo de Moçambique, Ministerio das Finanças, IGF. 2006. Audiotria de Desempenho do Sector de Aguas, Maputo. Governo de Moçambique, Orçamentos do Estado 2007 and 2008, Ministério das Finanças. Governo de Moçambique, Relatórios de Execuçao Orçamental (2000-2008), Ministerio das Finanças. Governo de Moçambique, Resoluçao do Conselho de Ministros Nº 7/95, August 8th 1995. International Monetary Fund. 2008. ―Republic of Mozambique: Report on Observance of Standards and Codes – Fiscal Transparency Module.‖ IMF Publication Services (Washington, D.C.). Instituto Nacional de Estatística. 2005. ―Atlas Sócio-Demográfico da Província da Zambézia.‖ Maputo. Komives,K, et al.,2005. Water, Electricity and the Poor: Who benefits from utility subsidies? World Bank, Washington D.C. 84 Lawson, A. et al. 2006. ―Assessment of Public Finance Management in Mozambique 2004/05 based on the PEFA Methodology.‖ Oveseas Development Institute (London) and SAL (Maputo). Lawson, A. et al. 2007. ―Assessment of Public Finance Management in Mozambique, 2006 based on the PEFA Methodology.‖ Fiscus Limited (Oxford) and SAL (Maputo) . McCoy, Simon. 2008. ―Provincial Budget Allocations in the Health, Education and Water Sectors of Mozambique: An analysis 2003-06 - Final Draft: 7th January 2008.‖ Morella, E., et al, 2008. Climbing the Ladder: The State of Sanitation in Sub-Saharan Africa. AICD, Background Paper 13(Phase 1), Washington D.C. Orlowski, Dieter. 2007. ―Donors‘ principles and resulting stress: Indirect taxes and donor-funded public works.‖ Oxford Policy Management. undefined. "Estatísticas de Moçambique." 3º Recenseamento Geral da População e Habitação . Instituto Nacional de Estatistica. 5 October 2009. . _____. ―Official Development Assistance to Mozambique.‖ ODAmoz. 5 October 2009. . _____. ―Harmonizing Efforts for Reducing Poverty in Mozambique.‖ Programme Aid Partnership Mozambique. 5 October 2009. . _____. ―The International Benchmarking Network for Water and Sanitation Utilities.‖ IBNET. 5 October 2009. . Watkins, K. 2007. Human Development Report 2007/2008: Fighting climate change: Human solidarity in a divided world. New York: UNDP. World Bank. 2007. "Country Partnership Strategy for the Republic of Mozambique." Washington, D.C.: World Bank. ________. 2007. ―Mozambique Country Water Resources Assistance Strategy: Making Water Work for Sustainable Growth and Poverty Reduction.‖ Washington, D.C.: World Bank. ________. 2009. ―Mozambique - Municipal Development in Mozambique : lessons from the first decade (Vol. 2 of 2) : Full report.‖ Washington, D.C.: World Bank. World Water Assessment Programme. 2009. The United Nations World Water Development Report 3: Water in a Changing World. Paris: UNESCO, and London: Earthscan. 85 BASIC DATA 2002 2003 2004 2005 2006 2007 2008 Official exchange rate (LCU per US$, period average) 23.68 23.78 22.58 23.06 25.40 25.84 24.30 GDP deflator (base year varies by country) 95.04 100.00 107.47 116.91 127.78 137.38 147.98 Population, total 19,259,465 19,783,627 20,310,610 20,834,379 21,353,466 21,869,362 22,382,533 Rural population 13,054,065 13,258,987 13,457,810 13,646,518 13,819,963 13,983,270 14,136,808 Urban population 6,205,400 6,524,640 6,852,800 7,187,861 7,533,503 7,886,092 8,245,725 Population density (people per sq. km) 24.49 25.16 25.83 26.49 27.15 27.81 28.46 Source: World Bank, Data Development Platform 86 DATA ANNEX TABLE 1 - SECTOR EXPENDITURES IN TOTAL MOF BUDGET Budgetary Expenditure on Selected Priority Sectors - Nominal (bln MZN) Sectors 2002 2003 2004 2005 2006 2007 2008 Water and Sanitation 636 410 618 1,320 1,694 1,366 2,261 Roads 2,859 2,619 2,911 4,409 5,073 5,046 5,192 Health* 2,630 2,520 2,978 4,329 5,560 6,833 6,304 Education 3,919 4,804 5,909 7,267 8,797 11,920 14,859 Total Budget Expenditure 24,871 25,044 29,503 36,912 48,569 58,256 64,067 * Excluding HIV/AIDS Source: Expenditure in PARPA Priority Sectors (MoF) Share in Total Budget Expenditure of Selected Priority Sectors Sectors 2002 2003 2004 2005 2006 2007 2008 Water and Sanitation 2.6% 1.6% 2.1% 3.6% 3.5% 2.3% 3.5% Roads 11.5% 10.5% 9.9% 11.9% 10.4% 8.7% 8.1% Health* 10.6% 10.1% 10.1% 11.7% 11.4% 11.7% 9.8% Education 15.8% 19.2% 20.0% 19.7% 18.1% 20.5% 23.2% * Excluding HIV/AIDS Source: Expenditure in PARPA Priority Sectors (MoF) 87 Expenditure on Selected Priority Sectors as Percentage of GDP Sectors 2002 2003 2004 2005 2006 2007 2008 Water and Sanitation 0.64% 0.37% 0.48% 0.87% 0.96% 0.68% 1.00% Roads 2.87% 2.36% 2.26% 2.91% 2.87% 2.51% 2.29% Health* 2.64% 2.27% 2.31% 2.85% 3.14% 3.39% 2.78% Education 3.94% 4.33% 4.59% 4.79% 4.98% 5.92% 6.55% * Excluding HIV/AIDS Source: Expenditure in PARPA Priority Sectors (MoF) Budgetary Expenditure on Selected Priority Sectors - nominal growth rates Sectors 2002 2003 2004 2005 2006 2007 2008 Water and Sanitation -35.59% 50.72% 113.75% 28.31% -19.39% 65.55% Roads -8.40% 11.15% 51.44% 15.07% -0.52% 2.88% Health* -4.18% 18.17% 45.39% 28.43% 22.88% -7.74% Education 22.60% 22.99% 22.99% 21.04% 35.51% 24.65% Total State Budget 0.70% 17.81% 25.11% 31.58% 19.95% 9.97% * Excluding HIV/AIDS Source: Expenditure in PARPA Priority Sectors (MoF) Budgetary Expenditure on Selected Priority Sectors - real growth rates Sectors 2002 2003 2004 2005 2006 2007 2008 Water and Sanitation -43.23% 33.82% 100.83% 13.31% -25.47% 51.33% Roads -19.26% -1.32% 42.28% 1.62% -8.03% -5.96% Health* -15.55% 4.92% 36.60% 13.41% 13.61% -15.67% Education 8.06% 9.20% 15.55% 6.89% 25.28% 13.94% * Excluding HIV/AIDS Source: Expenditure in PARPA Priority Sectors (MoF) 88 Expenditure in Selected Priority Sectors vs Total State Budget (bln MZN) Sectors 2002 2003 2004 2005 2006 2007 2008 Water and Sanitation 636 410 618 1,320 1,694 1,366 2,261 Roads 2,859 2,619 2,911 4,409 5,073 5,046 5,192 Health* 2,630 2,520 2,978 4,329 5,560 6,833 6,304 Education 3,919 4,804 5,909 7,267 8,797 11,920 14,859 All Priority Sectors 10,044 10,353 12,416 17,325 21,124 25,165 28,615 All Other Sectors 14,827 14,691 17,088 19,587 27,445 33,092 35,452 Total Budget Expenditure 24,871 25,044 29,503 36,912 48,569 58,256 64,067 Note Water as % of Priority Sectors 6.3% 4.0% 5.0% 7.6% 8.0% 5.4% 7.9% Water as % of All Sectors 2.6% 1.6% 2.1% 3.6% 3.5% 2.3% 3.5% * Excluding HIV/AIDS Source: Expenditure in PARPA Priority Sectors (MoF) 89 TABLE 2 - TOTAL SECTOR EXPENDITURES Expenditure in the Sector ('000 MZN) 2000 2001 2002 2003 2004 2005 2006 2007 2008 Recurrent 49,011 140,116 278,300 336,119 225,169 350,332 424,642 486,474 395,896 DNA 8,108 9,526 9,926 9,428 13,021 37,020 FIPAG 45,866 132,988 269,464 318,122 201,466 322,370 398,590 447,508 326,568 ARASUL 4,630 3,250 3,232 7,596 ARACENTRO 941 2,430 1,506 1,888 2,852 2,563 2,811 7,553 9,108 CRA 2,204 4,698 7,329 8,001 11,325 10,842 10,563 15,160 15,605 Domestic Investment 9,365 15,416 18,463 41,679 97,392 104,967 129,594 263,550 281,159 DNA 30,829 76,323 71,707 94,995 139,372 148,525 FIPAG 8,870 14,782 18,144 10,718 20,729 19,000 23,000 87,972 85,656 ARASUL 14,153 11,461 35,476 41,144 ARACENTRO - - - - - - - - - CRA 495 634 319 131 339 107 138 729 5,834 External Projects 14,541 73,519 200,593 570,409 556,227 1,364,616 911,252 1,614,954 2,008,769 DNA 375,049 222,039 535,150 192,078 748,529 847,237 FIPAG 14,541 73,519 200,593 168,682 310,186 778,697 640,168 752,886 1,009,159 ARASUL 28,686 43,372 32,229 91,822 ARACENTRO - - - 26,678 19,001 14,874 13,096 17,618 922 90 CRA - - - - 5,001 7,208 22,538 57,862 51,754 Non-Project External Funding - - - 48,357 18,985 132,786 120,975 140,840 200,122 DNA 48,357 18,985 132,786 120,975 140,840 200,122 FIPAG - - - - - - - - - ARASUL - - - - - - - - - ARACENTRO - - - - - - - - - CRA - - - - - - - - - Taxes (ARA Sul) - 8,082 5,830 7,875 Total Reported by Sector 72,917 229,052 497,356 996,564 897,772 1,952,700 1,586,463 2,505,817 2,885,946 Total Reported by MoF 636,192 409,779 617,633 1,320,210 1,694,006 1,365,518 2,260,653 Type of funding 2000 2001 2002 2003 2004 2005 2006 2007 2008 External Funding 19.9% 32.1% 40.3% 62.1% 64.1% 76.7% 65.1% 70.1% 76.5% Internal Funding 80.1% 67.9% 59.7% 37.9% 35.9% 23.3% 34.9% 29.9% 23.5% Type of funding 2000 2001 2002 2003 2004 2005 2006 2007 2008 External Projects 19.9% 32.1% 40.3% 57.2% 62.0% 69.9% 57.4% 64.4% 69.6% Non-Project External Funding 0.0% 0.0% 0.0% 4.9% 2.1% 6.8% 7.6% 5.6% 6.9% Domestic Investment 12.8% 6.7% 3.7% 4.2% 10.8% 5.4% 8.2% 10.5% 9.7% Domestic Recurrent 67.2% 61.2% 56.0% 33.7% 25.1% 17.9% 26.8% 19.4% 13.7% 91 Total Expenditure by Agency ('000MZN) Sector Agency 2000 2001 2002 2003 2004 2005 2006 2007 2008 DNA - - - 462,343 326,873 749,569 417,476 1,041,762 1,232,904 FIPAG 69,277 221,289 488,202 497,523 532,381 1,120,068 1,061,758 1,288,367 1,421,383 ARASUL - - - - - 47,468 66,165 76,767 148,436 ARACENTRO 941 2,430 1,506 28,566 21,853 17,437 15,907 25,171 10,030 CRA 2,699 5,332 7,648 8,132 16,665 18,158 33,239 73,751 73,193 Total 72,917 229,052 497,356 996,564 897,772 1,952,700 1,594,544 2,505,817 2,885,946 92 TABLE 3 - EXPENDITURES BY SUBSECTORS Expenditure by Sub-sector ('000 MZN) 2003 2004 2005 2006 2007 2008 Rural Water 23,092 12,094 233,070 97,474 206,938 352,763 DNA - Agua Rural 23,092 12,094 233,070 97,474 206,938 352,763 Urban Water 773,552 784,353 1,611,898 1,088,312 1,365,892 1,449,632 DNA - Agua Urbana 276,029 251,972 491,830 26,554 77,525 28,249 FIPAG 497,523 532,381 1,120,068 1,061,758 1,288,367 1,421,383 Water Resource Management 57,610 51,646 105,218 134,059 183,588 244,440 DNA - GRH 29,044 29,793 40,312 51,988 81,650 78,099 ARA Sul 0 0 47,468 66,165 76,767 148,436 ARA Centro 28,566 21,853 17,437 15,907 25,171 17,905 Sanitation 4,561 7,243 22,346 17,661 478,973 431,052 DNA 4,561 7,243 22,346 17,661 478,973 431,052 Institutional Management 23,713 51,004 337,942 147,575 183,657 294,870 DNA - Gestao, Admin, etc 23,713 51,004 337,942 147,575 183,657 294,870 Regulation 8,132 16,665 18,158 33,239 73,751 73,193 - Unallocated 105,904 25,233 -375,931 76,225 13,019 39,996 - DNA 105,904 25,233 -375,931 76,225 13,019 47,871 ARA Centro - - - - - -7,875 TOTAL 996,564 897,772 1,952,700 1,594,544 2,505,817 2,885,946 * note ARASul only since 2005 93 Expenditure by Sub-sector (%) 2003 2004 2005 2006 2007 2008 Rural Water 2.3% 1.3% 11.9% 6.1% 8.3% 12.2% Urban Water 77.6% 87.4% 82.5% 68.3% 54.5% 50.2% Water Resource Management 5.8% 5.8% 5.4% 8.4% 7.3% 8.5% Sanitation 0.5% 0.8% 1.1% 1.1% 19.1% 14.9% Institutional Management 2.4% 5.7% 17.3% 9.3% 7.3% 10.2% Regulation 0.8% 1.9% 0.9% 2.1% 2.9% 2.5% Unallocated 10.6% -2.8% -19.3% 4.8% 0.5% 1.4% 94 TABLE 4 - EXPENDITURES OF DNA DNA - DAR - Expenditure in Rural Water by Province ('000 MZN) Province 2003 2004 2005 2006 2007 2008 Maputo Province 1,250 4,102 6,096 5,616 Gaza 949 3,519 4,900 4,710 Inhambane 394 0 0 0 Manica 2,093 10,202 13,934 12,625 Sofala 518 9,303 11,019 362 Tete 0 2,471 5,143 942 Nampula 961 8,459 23,471 5,213 Zambézia 961 10,930 28,614 6,155 Cabo Delgado 417 1,003 0 17,434 Niassa 1,162 2,365 0 0 Total 8,705 52,353 93,178 53,056 DNA: Expenditure ('000 MZN) 2003 2004 2005 2006 2007 2008 Recurrent & Operational Costs* 8,108 9,526 9,926 9,428 117,480 150,849 Domestic Investment 30,829 76,323 71,707 94,995 100,348 130,442 External Investment (Projects) 423,406 241,024 667,936 313,053 823,934 951,613 Total Expenditure 462,343 326,873 749,569 417,476 1,041,762 1,232,904 * Note that this category includes re-classified recurrent & operational costs that were misclassified as investment expenditure 95 DNA Expenditure by Economic Category (%) 2003 2004 2005 2006 2007 2008 Recurrent & Operational Costs 1.8% 2.9% 1.3% 2.3% 11.3% 12.2% Domestic Investment 6.7% 23.3% 9.6% 22.8% 9.6% 10.6% External Investment (Projects) 91.6% 73.7% 89.1% 75.0% 79.1% 77.2% Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% DNA: Execution Rates ('000 MZN) 2003 2004 2005 2006 2007 2008 Recurrent & Operational Costs 91.2% 89.6% 93.4% 80.0% 431.6% 65.0% Domestic Investment 41.6% 69.7% 11.5% 11.1% 8.8% 7.7% External Investment (Projects) 54.1% 25.1% 87.6% 87.9% 89.6% 82.7% Total 228.0% 159.5% 63.6% 38.4% 59.8% 51.5% 96 TABLE 5 - EXPENDITURE BY FIPAG FIPAG: Expenditure ('000 MZN) 2000 2001 2002 2003 2004 2005 2006 2007 2008 Recurrent & Operational Costs 45,866 132,988 269,464 318,122 201,466 322,370 398,590 447,508 326,568 Domestic Investment 8,870 14,782 18,144 10,718 20,729 19,000 23,000 87,972 85,656 External Investment (Projects) 14,541 73,519 200,593 168,682 310,186 778,697 640,168 752,886 1,009,159 Total Expenditure 69,277 221,289 488,202 497,523 532,381 1,120,068 1,061,758 1,288,367 1,421,383 Distribution of Expenditures (%) 2000 2001 2002 2003 2004 2005 2006 2007 2008 Recurrent & Operational Costs 66% 60% 55% 64% 38% 29% 38% 35% 23% Domestic Investment 13% 7% 4% 2% 4% 2% 2% 7% 6% External Investment (Projects) 21% 33% 41% 34% 58% 70% 60% 58% 71% Total 100% 100% 100% 100% 100% 100% 100% 100% 100% FIPAG: Execution Rates (%) 2000 2001 2002 2003 2004 2005 2006 2007 2008 Recurrent & Operational Costs 386% 358% 309% 375% 243% 215% 173% 163% 229% Investments 97% 102% 96% 99% 93% 89% 67% 65% 99% Domestic Investment 100% 100% 100% 100% 100% 100% 100% 100% 100% External Investment (Projects) 96% 103% 95% 99% 93% 89% 66% 63% 99% Total 288% 256% 214% 275% 150% 126% 107% 100% 129% 97 TABLE 6- EXPENDITURE BY ARA-SUL ARA-Sul: Expenditure ('000 MZN) 2005 2006 2007 2008 Recurrent & Operational Costs* 26,216 20,763 44,401 50,691 Domestic Investment 1,504 4,348 8,214 11,529 External Investment (Projects) 19,748 24,891 18,322 78,342 Total Expenditure 47,468 50,001 70,937 140,561 * Note that this category includes re-classified recurrent & operational costs that were misclassified as investment expenditure ARA-Sul: Distribution of Expenditures 2005 2006 2007 2008 Recurrent & Operational Costs 55% 42% 63% 36% Domestic Investment 3% 9% 12% 8% External Investment (Projects) 42% 50% 26% 56% Total 100% 100% 100% 100% ARA-Sul: Execution Rates (%) 2005 2006 2007 2008 Recurrent & Operational Costs 93% 97% 89% 90% Domestic Investment 65% 97% 76% 90% External Investment (Projects) 97% 99% 94% 99% Total 93% 98% 88% 95% 98 TABLE 7 - EXPENDITURE BY ARA-CENTRO ARA-Centro: Expenditure ('000 MZN) 2000 2001 2002 2003 2004 2005 2006 2007 2008 Recurrent (Internal) 941 2,430 1,506 1,888 2,852 2,563 2,811 7,553 9,108 Domestic Investment - - - - - - - - - External Investment (Projects) - - - 26,678 19,001 14,874 13,096 17,618 922 TOTAL 941 2,430 1,506 28,566 21,853 17,437 15,907 25,171 10,030 ARA-Centro: Distribution of Expenditures (%) 2000 2001 2002 2003 2004 2005 2006 2007 2008 Recurrent (Internal) 100.0% 100.0% 100.0% 6.6% 13.1% 14.7% 17.7% 30.0% 90.8% Domestic Investment 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% External Investment (Projects) 0.0% 0.0% 0.0% 93.4% 86.9% 85.3% 82.3% 70.0% 9.2% TOTAL 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% ARA-Centro: Execution Rates ('000 MZN) 2000 2001 2002 2003 2004 2005 2006 2007 2008 Water Tax - - - 454 1,962 3,022 2,412 3,209 2,377 Treasury Funds 1,067 2,484 1,423 2,963 5,218 5,218 3,390 8,791 27,799 99 TABLE 8- EXPENDITURE BY CRA CRA: Expenditure ('000 MZN) 2000 2001 2002 2003 2004 2005 2006 2007 2008 Recurrent 2,204 4,698 7,329 8,001 11,325 10,842 10,563 15,160 15,605 Domestic Investment 495 634 319 131 339 107 138 729 5,834 External Projects 0 0 0 0 5,001 7,208 22,538 57,862 51,754 Total 2,699 5,332 7,648 8,132 16,665 18,158 33,239 73,751 73,193 Total without external projects 2,699 5,332 7,648 8,132 11,664 10,949 10,700 15,889 21,439 2000 2001 2002 2003 2004 2005 2006 2007 2008 Regulation Tax Receipts 2,699 5,332 7,648 8,132 11,664 10,949 10,700 15,889 21,439 Other Revenue 0 496 355 251 0 1,896 48 275 0 Treasury Funds 0 0 0 0 0 0 0 4,076 9,522 Total 2,699 5,829 8,003 8,383 11,664 12,846 10,748 20,240 30,961 Distribution of Expenditures (%) 2000 2001 2002 2003 2004 2005 2006 2007 2008 Recurrent & Operational Costs 82% 88% 96% 98% 97% 99% 99% 95% 73% Domestic Investment 18% 12% 4% 2% 3% 1% 1% 5% 27% External Investment (Projects) 0% 0% 0% 0% 43% 66% 211% 364% 241% Total 100% 100% 100% 100% 143% 166% 311% 464% 341% 100 TABLE 9 - USE OF IMPROVED DRINKING WATER SOURCES Principal Sources of Drinking Water (in percent) Improved water Improved Water Sources Unimproved Water Sources sources Total borehole with Bottled water Unprotected In the home River, lakes Outside the hand pump hand pump Standpipe / Neighbors Rainwater house/tap Protected Without a well or Other kiosk well By Selected Characteristics Total 2.1 5.6 9.2 6.2 19.8 0.2 4.5 36.1 0.1 16.0 0.2 100.0 43.1 Location Urbano 6.2 16.7 19.7 17.8 9.3 0.3 6.6 20.2 0.2 2.8 0.1 100.0 70.0 Rural 0.1 0.3 4.1 0.5 25.0 0.2 3.5 43.8 0.0 22.4 0.2 100.0 30.2 Province Niassa 1.1 0.4 5.3 1.8 35.4 0.0 4.7 27.4 0.0 23.7 0.1 100.0 44.0 Cabo Delgado 0.5 1.4 5.1 3.0 19.9 0.1 6.4 47.3 0.0 16.2 0.1 100.0 30.0 Nampula 2.1 3.9 7.6 5.3 24.3 0.0 2.8 41.4 0.0 12.6 0.2 100.0 43.2 Zambezia 0.3 0.9 5.1 2.7 14.7 0.1 2.0 49.5 0.0 24.8 0.0 100.0 23.8 Tete 0.0 0.5 8.6 1.6 23.4 0.1 2.2 35.7 0.0 27.8 0.2 100.0 34.2 Manica 0.8 1.4 5.4 1.8 22.7 0.0 5.0 36.8 0.0 25.4 0.7 100.0 32.1 Sofala 2.7 7.3 14.0 8.1 15.9 0.0 1.5 38.5 0.0 12.0 0.0 100.0 48.0 Inhambane 0.6 3.4 6.0 3.9 20.9 2.4 8.9 49.2 0.0 4.5 0.0 100.0 37.2 Gaza 1.3 6.4 16.2 5.1 31.7 0.3 8.0 17.5 0.0 11.8 1.7 100.0 61.0 Maputo Provincia 4.1 21.9 12.0 19.8 9.7 0.0 12.3 10.3 0.2 9.5 0.2 100.0 67.5 Maputo Cidade 14.5 29.1 24.4 24.2 1.0 0.0 3.9 1.7 1.2 0.0 0.0 100.0 93.2 Income Quintiles Poorest 0.0 0.0 0.0 0.0 12.6 0.0 0.0 58.8 0.0 28.6 0.0 100.0 12.6 Second 0.0 0.0 0.5 0.0 22.3 0.0 1.4 51.5 0.0 24.3 0.1 100.0 22.8 Third 0.0 0.0 9.4 1.0 34.0 0.1 6.9 33.3 0.0 15.0 0.2 100.0 44.5 Fourth 0.4 2.4 17.5 6.7 22.8 0.5 8.3 30.2 0.0 10.8 0.4 100.0 50.3 Richest 9.9 25.8 18.4 23.1 7.5 0.5 5.8 6.8 0.4 1.3 0.4 100.0 85.2 Source: UNICEF, Multiple Indicator Cluster Survey 2008 101 TABLE 10- USE OF IMPROVED SANITATION SERVICES improved sanitation Type of Sanitation Used by the Household (in percent) population with Percentage of Improved Sanitation services Services Unimproved Sanitation Services Total Unimproved information Traditionall y Improved septic tank septic tank Toilet with Improved No mato Without without Latrine Beach latrine latrine Toilet Other Selected Characteristics Total 2.4 2.5 6.8 7.6 38.3 1.5 40.3 0.0 0.6 100.0 19.3 Location Urban 7.0 7.3 17.9 14.9 37.9 2.6 11.9 0.0 0.5 100.0 47.1 Rural 0.2 0.2 1.4 4.0 38.5 0.9 54.1 0.0 0.6 100.0 5.8 Province Niassa 0.6 0.1 4.5 10.3 62.6 0.1 21.0 0.0 0.8 100.0 15.4 Cabo Delgado 0.1 0.4 2.6 2.6 64.7 4.3 25.3 0.0 0.1 100.0 5.6 Nampula 2.2 1.1 4.7 7.3 38.3 5.6 40.8 0.0 0.0 100.0 15.2 Zambézia 0.2 0.2 3.0 4.1 18.1 0.1 71.6 0.0 2.6 100.0 7.6 Tete 0.1 0.0 2.3 0.9 38.5 0.0 58.0 0.0 0.1 100.0 3.4 Manica 0.2 1.5 7.3 5.3 33.9 0.0 51.4 0.0 0.4 100.0 14.2 Sofala 3.7 5.4 7.3 5.6 23.6 0.0 54.1 0.0 0.3 100.0 22.0 Inhambane 1.2 0.3 5.3 8.7 55.9 0.2 28.1 0.0 0.2 100.0 15.5 Gaza 1.3 1.9 11.9 8.7 56.2 0.3 19.7 0.0 0.0 100.0 23.8 Maputo Província 5.4 6.9 10.5 24.5 38.1 0.0 14.5 0.1 0.1 100.0 47.3 Maputo Cidade 17.3 17.8 31.4 18.3 15.2 0.0 0.0 0.0 0.0 100.0 84.6 Income Quintile Poorest 0.0 0.0 0.0 0.0 7.2 1.0 91.6 0.0 0.1 100.0 0.0 Second 0.0 0.0 0.0 0.1 38.6 1.4 58.8 0.0 1.1 100.0 0.1 Third 0.0 0.0 0.8 7.0 60.8 0.8 29.8 0.0 0.7 100.0 7.7 Fourth 0.0 0.0 6.1 10.9 59.9 2.2 20.2 0.0 0.7 100.0 17.0 Richest 11.9 12.6 27.1 20.0 25.2 2.0 1.0 0.0 0.2 100.0 71.5 Source: UNICEF, Multiple Indicator Cluster Survey 2008 102