85053 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements ©2014 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org Acknowledgements This report was led by Mukami Kariuki, Sector Leader, the World Bank. It was compiled by Guillaume Patricot, WSP consultant, with input from Rosemary Rop, Task Team Leader within the Water and Sanitation Program, Sam Mutono who managed the data collection, field work and liaison with the client, and Midori Makino who was the initial Task Manager within the World Bank. Within the National Water and Sewerage Corporation Eng. Alex Gisagara, the then acting Managing Director, Mr. J. B. Otema, Urban Pro Poor Manager, Eng. Paddy Twesigye Project Manager for Planning and Capital Development, Eng. Jude Mwoga, Senior Manager of Design, Monitoring and Evaluation provided input, data and quality assurance of the output and messages. The authors gratefully acknowledge the role of Joseph Gadek, Xavier Chauvot de Beauchene and Michael Webster for their useful guidance and insight as peer reviewers. Anjali SenGupta, editor, Catherine Njeri designer and Toni Sittoni, regional communications specialist in WSP are gratefully acknowledged for their support to the final production process. Disclaimer This work is a product of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. Foreword Since the late 1990s, Uganda has been undertaking an ambitious program to reduce national poverty, emphasizing pro-poor service delivery and promoting equitable economic growth. The urban water sector, specifically, has been the focus of wide-ranging reforms targeted at reaching the poor. At the institutional level, the national public water authority responsible for service delivery in Uganda’s towns and cities - National Water and Sewerage Corporation (NWSC) - has been implementing a raft of policies explicitly focused on improving delivery of services to the urban poor. This study examines the extent to which these reforms have increased water coverage within Kampala City, specifically in informal settlements where the majority of poor households reside. This collaborative effort by the World Bank and NWSC seeks to identify and document what works and what does not in reform measures aimed at reaching the poor, and highlights the gains and challenges of making service delivery pro-poor and managing service delivery reform to improve effectiveness and accessibility. We hope that this knowledge will contribute more insights to guide reforms aimed at improving access to basic services by the poor, and enhance the effectiveness of pro-poor mechanisms employed in the urban water sector. Jamal Saghir Director Sustainable Development Department iii Executive Summary Background Description of Pro-poor Measures The Government of Uganda (GoU) has i. An Affordable Connections Policy, to undertaken a wide range of reforms in the urban increase coverage in all NWSC areas water sector since the inception of the Poverty (in particular, the poor settlements) by Eradication Action Plan (PEAP) in 1997 (updated lowering the cost of the connection fee in 2000 and 2004). for any customer (domestic and non- domestic) living within 50 meters of the The PEAP provided a framework for development water mains. This policy was to be funded of multisector targets and investment through an 11 percent surcharge in the programs at national, district, and local levels. average water supply tariff, to be collected Implementation of the PEAP included the and ring-fenced into a New Connections development of detailed sector strategies and Fund. programs which were supported by the GoU’s development partners, including the World Bank, ii. A Pro-Poor Tariff Policy, which shifted the via Poverty Reduction Strategy Credits. uniform tariff structure for all consumers to a differentiated structure for (i) domestic In 2004, the Ministry of Water and Environment (ii) non domestic customers, and (iii) (MoWE) unveiled a policy to expand water supply public water points (PWPs). This means services to the poor in urban as well as rural that there is a specific tariff for each of areas. This was updated in 2006. The policy set the NWSC’s customer categories across the target of 100 percent coverage for water all water supply areas: industrial and supply and sanitation (WSS) services in urban commercial, government and institutional, areas by 2015. and domestic. An additional tariff was developed specifically for PWPs. The In response to the policy, in 2004 the National MoWE considers that the differentiated Water and Sewerage Corporation, (NWSC) which tariff structure per customer category is responsible for service delivery in Kampala, ensures equity in the supply and pricing.1 undertook a series of measures to implement the pro-poor policy in urban areas. iii. A Pro-Poor Targeting Project, which the NWSC is implementing since 2008 with This case study assesses the impact of pro-poor support from the Global Partnership for measures implemented by NWSC in Kampala Output-Based Aid (GPOBA) to subsidize since 2004, identifies the key factors that affected water supply connections (including the outcomes of these policies and strategies, shared yard taps and prepaid PWPs/ proposes areas for improvement, and identifies kiosks) in specific poor settlements of lessons that can be learned and shared from Kampala. the experience in Uganda. The primary purpose of the case study is to expand the scope and To deliver these policies, the NWSC adopted two increase the effectiveness of NWSC’s pro-poor main delivery mechanisms: first, it established policy which is described below. a pro-poor branch and, second, it provided 1 MoWE. 2011. Uganda Water and Environment Sector Performance Report 2011. iv Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements multiple levels of service. An Urban Pro-Poor unit (average consumption is 20 liters per was set up and staffed in 2007 to promote, plan capita per day). and support water supply and sanitation service expansion in poor unserved settlements of iii. Nonpayment of water bills by households Kampala. Multiple levels of service were provided and PWP/kiosk operators is causing a to new domestic customers, including a choice significant number of disconnections: 21 of individual house connections, PWPs/kiosks, percent of yard taps and 53 percent of and shared yard taps. In 2008 the NWSC started PWPs. Although it is being addressed by pilot-testing prepaid PWPs/kiosks. Community the NWSC’s Urban Pro-Poor Branch, this participation and demand creation for shared significantly undermines any effort by the yard taps and prepaid PWPs/kiosk pilot-tested NWSC and its partners to expand services in the GPOBA project have now become the into poor settlements at scale using yard NWSC’s approach to expanding services in all taps and PWPs. unserved areas. iv. The NWSC did not set up a dedicated and On the whole, the pro-poor policy and program adequately ring-fenced New Connection have led to a rapid expansion of services to poor Fund in spite of the fact that this was a households. Specifically, 2,405 new yard taps key element of its Affordable Connections and 663 PWPs serving an estimated 81,000 Policy. As a result, the funds collected people were realized. A number of difficulties for the Affordable Connection Policy and challenges, nevertheless, emerged during through an 11 percent tariff surcharge implementation. The World Bank, the Water and were diverted to the funding of regular Sanitation Program and the NWSC collaborated operational activities. in preparing this case to draw lessons from this experience, and inform replication of good v. Water mains extensions in poor approaches at scale. settlements were, and are still, carried out on a piecemeal, project by project Summary of Key Challenges Identified basis because Kampala Water has no clearly defined strategic business plan i. The discounted cost of household that identifies total investment needs and connections at UGX59,000 or $35 in revenue/funding requirements over the 2004 is still considered a significant long term. barrier to access for poor households, as it represents 74 percent of the average Summary of Key Findings monthly income of poor households and is therefore beyond the reach of some Overall, in spite of the these challenges, the case households. study finds that the pro-poor policy has largely been successful due to the following factors: ii. The water tariff is affordable for poor households, but the level of subsidy i. Improved financial sustainability: The reaching the poorest is very low. It is NWSC’s broad-based policy has allowed estimated at UGX2,600 ($1.3) per capita the utility to significantly improve access to per annum for users of PWPs, compared services, while also improving its financial to UGX10,300 ($5.1) per capita for house viability through the more than doubling connections. This is mainly due to the fact of connections and tripling of revenue that poor households share connections between 2004 and 2010. Achieving and use relatively low volumes of water financial sustainability is essential v for utilities to realistically serve poor iv. Increased access to PWPs: The poorest households at scale. However, utilities also consumers in Kampala are currently need to undertake targeted programs to served by PWPs/kiosks. The National reach the poor. Service Delivery Survey (NSDS, 2008) showed that 26 percent of the population ii. Improved operational efficiency: Increased was served by a public tap (PWP/kiosk, efficiency of operations has enabled Presidential taps) in 2008. The three-fold the NWSC to plan for, procure, and increase in the number of public taps, as manage services for a growing number well as expansion of shared connections of customers. This entailed employing (predominantly yard taps) in poor areas, more and better trained staff; setting key therefore, helped increase access by the performance contracts with staff and poor. This also led to a reduction in the with internal independently managed number of privately managed water points, water supply operations; applying for selling water at a significantly higher price. and obtaining an effective tariff structure; installing a new customer billing and v. Creation of a pro-poor branch: The revenue management system; improving creation of an organizational structure community liaison and customer care; and institutional mechanism to initiate, and improving procurement and works facilitate, continuously monitor and management systems. promote pro-poor initiatives, update MIS, and coordinate the activities of different iii. More affordable connections: The interest groups (including the NWSC’s own Affordable Connection Policy increased the operational branches) and organizations number of connections in Kampala across was a very positive step. This allowed the board. Although it was designed dedicated personnel of the NWSC to to increase coverage in poor, unserved concentrate on pro-poor activities, a focus areas, the NWSC broadened coverage that the organization lacked earlier. The of the Affordable Connections Policy to pro-poor branch was directly responsible non-poor domestic, government and for improving the effectiveness and industrial customers as well. As a result, efficiency of the NWSC’s programs in poor NWSC’s rapid increase in connections settlements by reducing the proportion from 59,000 in 2004 to 146,000 in 2009, of inactive PWPs and yard taps from (on average 14,700 new connections per 40 percent in 2007 (when the pro-poor year), led to an increase from 64 percent branch was created) to less than 10 to 72 percent increase in water supply percent in 2009. coverage. 2,500 yard taps and 660 PWPs were installed in poor settlements due to vi. An affordable tariff: The tariff was made the policy.2 This represents approximately more affordable to poor households by 4.3 percent of the total new connections introducing three different measures: First, made between 2004 and 2010.3 These different rates were set for all customer new, predominantly shared connections groups; a specific rate for standposts have benefited 21 percent of the 466,000 which was also applied to yard taps, if additional people served through the these are shown to serve more than three policy. households. Second, a cross-subsidy 2 In total 6,000 yard taps and 1,000 public standposts were constructed in poor settlements from 2004 to 2010. However, approximately 700 yard taps and 60 PWPs were already being constructed every year before the policy. These baseline figures have not been included in the estimate of the additional connections that can be attributed to the Affordable Connections Policy only. 3 If the baseline number of new connections per year had been included the total number of connections made in poor settlements would represent 9.5 percent of the total number of connections made since the Affordable Connections Policy was implemented (that is, over the period 2004–09). vi Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements between different customer categories I. Recommendations that the NWSC can (for example domestic, government/ implement on its own: institutional and commercial/industrial. The third was the 11 percent surcharge in the i. The NWSC and its partners should tariff to fund new connections. develop a long-term strategic business plan: Investments in network expansion vii. Geographical targeting: With support have, to date, been piecemeal and on a from GPOBA, output-based subsidies project-by-project approach due to the are being provided against delivery of lack of a long-term strategic business prepaid PWPs and yard taps. Although plan for Kampala. A long-term business progress has been slow, the approach plan which coherently identifies needs, shows that geographical targeting of specific schemes, and funding sources poor households and performance-based for expansion and sustaining of services subsidies offer significant opportunities to to the whole of Kampala, in particular the increase coverage (including volumes of poor and underserved settlements, is water sold and thus financial sustainability) required. This would catalyze the NWSC, in poor, unserved, areas. The GPOBA GoU, customers and development project is expected to connect 400,000 of partners and would ensure that pro-poor Kampala’s poor. initiatives are brought to the forefront at the corporate level and are large scale. viii. Prepaid water metering: A careful and well-thought through introduction of ii. Pro-poor targets should be set at prepaid water meters at PWPs has led to company, branch, and individual staff its widespread acceptance by consumers, contract level: Although the NWSC’s politicians, nongovernmental organizations objectives include expanding services (NGOs), and even the press. This has in unserved areas, these are not clear largely eliminated the middlemen in water targets, nor are these mentioned in the supplies and made water affordable at the Internally Delegated Area Management official NWSC price. Currently only 4,500 Contracts (which the head office has people have access to prepaid meters set up with each of the towns under the but the NWSC’s intention is replicate their NWSC’s jurisdiction) and staff contracts. It adoption at scale, eventually reaching 75 is considered that specifying clear targets percent of the poor. for expansion of services in unserved areas would incentivize NWSC, Kampala Recommendations on measures to address City Council (KCC), and the Department challenges and build on the above findings are of Water Development to prepare detailed presented below: investment plans and strategies, and thereafter deliver these. These detailed Recommendations targets would need to be included at Business Plan level (see point above). Recommendations are divided into the following categories: measures which NWSC iii. A New Connection Fund: Revenues raised can implement on their own; those they can through the 11 percent surcharge in the implement with the support and/or approval water tariff should be sequestered into of the government; and those that are beyond a properly ring-fenced New Connection NWSC’s direct control or influence. Fund, to improve the effectiveness and vii efficiency of these operations, and give iii. Effectiveness (and benefit) of the cross- the NWSC’s management a clear view subsidy should be improved: Poor on what rate of new connections can households are not purchasing enough realistically be achieved. This should be water from the NWSC and are therefore supported by a clear view of the water not benefiting as much as they should be supply investment needs—to be included from the cross-subsidy. In fact, average in the Business Plan above. Use of the consumption has reduced from 1.3 Fund could be audited by the GoU and m3/day to 0.6 m3/day for all customer development partners. categories, due to a combination of water supply capacity restrictions (although iv. Improve marketing and availability of continuity of service is officially reported prepaid tokens (and customer training to be at 24x7) and the use of alternative on how to use them). Alternatively, new sources of water. It is noted that water payment arrangements (for example, supply tariffs are more affordable to poor use of cell phones) can gradually be households when water resale does not introduced. take place. II. Recommendations that the NWSC can iv. Tariff structure to be improved: Whilst implement with the support and/or approval of the benefits of the differential flat tariff the GoU: structure are recognized, the flat tariff for domestic customers is not i. Legal constraints faced in getting new conducive to water saving. Therefore, connections should be eased: For a it is recommended that an increasing new connection the NWSC insists on block tariff be set for (nonpoor) domestic production of a land title document, a customers. In addition, it is considered lease document or permission of the that the automatic tariff increase should be landowner or lease holder. In the present waived for public standposts. Any revenue land tenure system many people don’t shortfall could thus be funded via the have such land documents and getting a increasing block tariff. water connection is thus not possible for them. Documents such as voters lists, v. Water tariffs in slums (at PWPs) to be municipal identification, community group equalized (normalized): The differentiated guarantees, ration card, and so on, have tariff rates for water for PWPs/kiosks been tried in other countries. Such an (UGX867/m³) and Presidential Pledge approach would potentially lead to more taps (UGX784/m³) should eventually be rapid expansion into unserved areas. standardized. ii. Size (and scope) of the cross-subsidy III. Recommendations that are outside the between different categories of customers NWSC’s control or influence: should be rationalized to maximize total revenues: Poor households are indeed i. Barriers to access caused by the existing benefiting from a differentiated flat tariff land tenure system should be addressed structure. However, the level of subsidy by the KCC: The Mailo/private land tenure appears to be relatively low given that system remains a barrier of access to domestic, industrial, and institutional these services in poor settlements. As consumers are only charged twice the rate owners cannot be forced to provide basic charged at PWPs. urban infrastructure services (including water, sanitation, roads, and so on) this viii Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements limits the utilities’ ability to expand the demand in the poor settlements) limits the network into unserved areas. While special NWSC’s ability to expand consumption. rights can be obtained from relevant Average individual consumption in public authorities, including the KCC and these settlements is less than 20 liters the Ministry of Lands, this is not a simple per capita per day. In addition, many of process. these alternative sources of water are unprotected and therefore unsafe. Thus, ii. Options for managing unsafe, alternative it is important that the GoU and KCC sources of water in Kampala’s poor adopt a clear policy for closing and/ settlements should be decided. The or rehabilitating all of these sources. availability of alternative, unsafe sources This needs to be complemented with a of water in poor settlements (which detailed analysis of remediation works and provide up to 50 percent of all water management options. © WSP ix Acronyms and Abbreviations AFD Agence Française de Développement, a French development agency AFTU1 Urban Development and Services Practice 1 APC Area performance contracts CIDI Community Development Initiatives (an NGO) CPI Consumer Price Index DWD Directorate of Water Development DWRM Directorate of Water Resources Management GPOBA Global Partnership for Output-Based Aid GIS Geographic information system IEC Information education and communication IDAMC Internally delegated area management contracts KCC Kampala City Council KIEMP Kampala Integrated Environment Planning and Management Project KfW KfW Entwicklungsbank, a German development bank KUSP Kampala Urban Sanitation Program lpcd Liters per capita per day MDG Millennium Development Goal MoLHU Ministry of Land, Housing, and Urban areas MTEF Medium Term Expenditure Framework MoWE Ministry of Water and Environment MoWLE Ministry of Water, Lands, and Environment (now MoWE) NGO Nongovernmental organization NSDS National Service Delivery Survey NWSC National Water and Sewerage Corporation OBA Output-based aid O&M Operation and maintenance PEAP Poverty Eradication Action Plan PRSC Poverty Reduction Support Credit PRSP Poverty Reduction Strategy Paper PWP Public water point SWAp Sector-Wide Approach (to Planning) UBOS Uganda Bureau of Statistics UGX Uganda Shillings UNDP United Nations Development Programme UNHS Uganda National Household Survey W&S Water and Sanitation WHO World Health Organization WSS Water Supply and Sanitation WUA Water User Association x Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Contents Foreword iii Executive Summary iv Acronyms and Abbreviations x 1. Introduction 1 1.1 Background 2 1.2 Study Objectives 4 1.3 Methodology Used and Data Collected 4 1.4 Data Collected 6 1.5 Structure of the Report 6 2. Overview of the Water Supply Context in Kampala 7 2.1 Legal, Institutional, and Regulatory Framework 8 2.2 NWSC Internal Reforms 9 2.3 Definition and Incidence of Poverty in Kampala 11 2.4 Land Tenure and Its Impact on Services to the Poor 13 2.5 2002 Baseline of Urban Poor Coverage Areas in Kampala 13 3. Urban Water Sector Policy and Strategy for Improving Services to the Poor 19 3.1 GoU’s Poverty Eradication Action Plans 20 3.2 The Government’s Pro-Poor Water Policies and Strategies 20 3.3 NWSC’s Pro-Poor Policy and Instruments 22 3.4 Principal Components of the Pro-Poor Policy 23 3.5 NWSC’s Pro-Poor Strategy Delivery Mechanisms 25 4. Analysis of the Impact of the Pro-Poor Policies on the Poor in Kampala 29 4.1 Assessment of Coverage Increase 30 4.2 Impact on Financial Viability 42 4.3 Individual Consumption 43 4.4 Distance and Time Spent 50 4.5 Affordability 51 4.6 Social Impact Analysis 55 4.7 Summary of Achievement of the GoU’s Objectives 60 4.8 Review of Results and Lessons Learned 62 5. Conclusions 66 6. Recommendations for Further Study 69 Appendices 71   xi Boxes Box 1: Assessing efficiency, effectiveness, and equity of service provision to the poor 5 Box 2: Typology of water supply connections in Kampala 14 Box 3: Geographic targeting of poor households 25 Box 4: Assessment of the quantity of subsidy that targets poor households 54 Tables Table 1: Development indicators for Uganda (2009 and 2010) 2 Table 2: Population growth in Kampala (1980 to 2010) 12 Table 3: Sources of water in urban areas (National data, NSDS, 2004) 15 Table 4: Distance to nearest water source (Kampala slums, 2002 Census) 16 Table 5: Kampala water prices in 2008–09 25 Table 6: Yearly connection targets of the Affordable Connections Policy (2004) 30 Table 7: Growth in NWSC connections in Kampala (2002/03 to 2009/10) 31 Table 8: Sources of drinking water in urban areas for dry and wet seasons (NSDS, 2008) 33 Table 9: Total number of new connections (per annum and cumulative) from 1998 to 2011 in Kampala 34 Table 10: Actual number of new connections per annum (2004 to 2010) for NWSC 35 Table 11: Estimate of the total number of poor beneficiaries of the Affordable Connections Policy(2004–10) in Kampala only 37 Table 12: Schedule of water sector investment projects in Kampala (2002–10) 41 Table 13: Average unit consumption per customer category (NWSC, 2010) 45 Table 14: Distribution of households by distance to water sources during the dry season (urban areas) 51 Table 15: Average collection time for drinking water in urban areas 51 Table 16: Water market segmentation and individual consumption (2010) 53 Table 17: Summary of FGD on impacts of NWSC pro-poor policy 57 Table 18: Achieving the components of GoU and NWSC policy 60 xii Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Figures Figure 1: Institutional and regulatory framework for urban water supply services in Uganda 10 Figure 2: Total water supply connections in Kampala (2002 to 2010) 32 Figure 3: Total number of new connections (per annum and cumulative) from 1998 to 2011 in Kampala 35 Figure 4: Number and proportion of inactive connections in Kampala (1998 to 2011) 36 Figure 5: Water sector investments in Kampala (2001 to 2008) 40 Figure 6: NWSC financial data (2002–10) 43 Figure 7: Volumes of water produced and sold vs. total number of connections in Kampala (2002 to 2010) 44 Figure 8: Average monthly consumption of poor households using yard tap connections 45 Figure 9: Unit consumption at public water points in Kampala (2009) 46 Figure 10: Unit consumption at prepaid public water points/kiosks 47 Figure 11: Nonrevenue expressed in % of water produced and in m3/km/day (2004–10) 59 Pictures Prepaid public water point/kiosk in use in Kisenyi, Kampala (2011) Picture 1: 48 Prepaid private yard tap in use in Kisenyi, Kampala (2011) Picture 2: 48 Unprotected shallow well in Kisenyi (Kampala, 2011) Picture 3: 48 ‘Protected’ but unsafe spring in Kisenyi (Kampala, 2011) Picture 4: 49 Public and private Sulabh toilets in Kisenyi (Kampala, 2011) Picture 5: 49 xiii xiv Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements 1 Introduction 1 1.1 Background Despite being one of the poorest countries in the world Uganda has, since the late 1990s, 1.1.1 Country and Economic Data developed a coherent poverty reduction Uganda is one of the poorest countries in Africa with strategy which focuses on strong governance, a gross national income (GNI) per capita in 2010 of economic growth, human development, US$500 compared with the sub-Saharan regional and participation of all individuals in the average of $1,170. Uganda’s population growth of over development process. Only 12 percent of its 3 percent per year, one of the highest in the world, puts population lives in urban areas. A selection of a considerable strain on public sector service delivery, development indicators for Uganda (2010) are not just for water and sanitation but also in other areas summarized in Table 1. such as health, education, and transport. Table 1: Development Indicators for Uganda (2009 and 2010) Country overview Population, total (millions) 33.42 Population growth (annual %) 3.2 Surface area (sq. km) (thousands) 241.0 Poverty headcount ratio at national poverty line (% of population) 24.5 GNI, Atlas method (current US$) (billions) 16.55 GNI per capita, Atlas method (current US$) 500 GNI, PPP (current international $) (billions) 41.49 People Income share held by lowest 20% (2009) 5.8 Life expectancy at birth, total (years) in 2009 53 Fertility rate, total (births per woman) in 2009 6.2 Adolescent fertility rate (births per 1,000 women ages 15-19) in 2009 141 Mortality rate, under-5 (per 1,000) 99 Ratio of girls to boys in primary and secondary education (%) in 2009 99 Prevalence of HIV, total (% of population ages 15-49) in 2009 6.5 Economy GDP (current US$) (billions) 17.01 GDP growth (annual %) 5.2 Inflation, GDP deflator (annual %) 9.1 Global links Total debt service (% of exports of goods, services and income) in 2009 2.0 Workers' remittances and compensation of employees, received (current US$) (millions) 778 Foreign direct investment, net inflows (BoP, current US$) (millions) 789 Net official development assistance and official aid received (current US$) (millions) 1,786 Source: www.worldbank.org/data/wdi 2 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements 1.1.2 Water Resources substantial increases in coverage as well as in operational and commercial performance.4 As a whole, Uganda has more than enough Sector reforms in the period 1998–2003 included freshwater. Estimates indicate 66 km³ of the commercialization and modernization of renewable water resources per year, which the NWSC operating in cities and larger towns, correspond to approximately 1,974 m³ per as well as decentralization and private sector person per year. However, the distribution of the participation in small towns.5 These reforms have resource is uneven both in spatial and temporal significantly improved the performance of the terms. Furthermore, freshwater is increasingly urban water sector, and thus attracted significant exploited through population growth, international attention. urbanization, agriculture, and industrialization. However, 40 percent of the population still had The rivers, lakes, and wetlands cover about no access to an improved water source and 57 18 percent of Uganda’s total surface, including percent had no improved sanitation in 2004.6 Lake Victoria, Africa’s largest freshwater lake Low access to urban sanitation and wastewater and one of the major sources of the Nile River, treatment, compared to the progress achieved the longest river of the world. Almost the on urban water supply, remains a key area of entire country lies within the Nile basin. Rainfall concern for the sector. contributes most to the country’s surface and groundwater. The average annual rainfall The water and sanitation sector has been ranges from 900 mm in the semi-arid areas recognized as a key area under the 2004 Poverty of Kotido to 2,000 mm on the Sese Islands Eradication Action Plan (PEAP), by the then in Lake Victoria. Although there is a lack of Uganda’s main strategy paper to fight poverty.7 information on the sustainability of groundwater A comprehensive expenditure framework was recharge, it is considered that groundwater introduced to coordinate financial support by recharge in Ugandan towns meets the current external donors, the national government, abstraction volumes. In addition, although the and NGOs.8 The PEAP estimated that from water level in Lake Victoria has receded by 2001 to 2015, about US$1.4 billion, or US$92 3 percent of its volume between 2003 and million per year, are needed to increase water 2006 (due to a combination of a decline in supply coverage up to 95 percent.9 Since 2010 rainfall and overexploitation), it is considered to the PEAP has been replaced by the National provide enough raw water, despite additional Development Plan (NDP). investments required to extend raw water intakes for Kampala, Entebbe, and Jinja (and Whilst much has been achieved, significant other large cities located around the lake, for investments are required to maintain and replace example, Kisumu in Kenya, and so on). existing infrastructure, in particular in small-town systems and rural areas. Anticipated public funds 1.1.3 Water Supply and Sanitation are inadequate to deal with these dual demands of expansion and replacement. The considerable The Ugandan water supply and sanitation finance gap that exists for water and sanitation (WSS) sector has made spectacular progress means that the positive gains of the past may in urban areas since the mid-1990s, with plateau.10 4 Mugisha, Silver, and Sanford V. Berg. November 2006. Struggling State-Owned Enterprises: NWSC’s Turnaround in Uganda. 5 UN-Water, World Water Assessment Programme. 2006. National Water Development Report: Uganda. Prepared for 2nd UN World Water Development Report, “Water, a shared responsibility”. 6 MDG target for urban water supply coverage in Uganda: 100 percent by 2015. 7 Republic of Uganda, Ministry of Finance, Planning and Economic Development. Poverty Eradication Action Plan (2004/5–2007/8). 8 Ministry of Water and Environment (Uganda). September 2006. Water and Sanitation Sector Performance Report 2006. 9 Rural areas: US$956 million; urban areas: large towns (US$281 million) and small towns (US$136 million). Republic of Uganda, Ministry of Finance, Planning and Economic Development. Poverty Eradication Action Plan (2004/5–2007/8). 10 WSP et al. 2010. Water Supply and Sanitation in Uganda: Turning Finance into Services for 2015 and Beyond. An AMCOW Country Status Overview. 3 1.1.4 Water Supply Services to supply services in poor settlements. The study Poor Households focuses exclusively on the area of Kampala city, the capital of Uganda, as this constitutes In this climate of urban water supply sector the largest service area of the NWSC—with performance improvement, serving the poor approximately 60 percent of its connections has received increased attention not only at the and 64 percent of its revenues. The objective political level but also at the utility level. This has is to undertake a detailed assessment of the meant that Uganda’s urban water sector has workings and implementation of the policy and been able to attract increased financial support strategy, to identify key lessons that can be from development partners. learned from these approaches, and to provide concrete recommendations on how these might The NWSC proposed and implemented a be improved. strategy in 2004 for increasing coverage in unserved areas (including poor and nonpoor In addition, the study seeks to address the urban areas in the country), called the Affordable following: Connections Policy. The policy stipulates that the NWSC will connect all customers (including • Analyze the institutional arrangement domestic and nondomestic) for free if these are for delivery of water to the poor areas located within 50 meters of the NWSC’s water and assess its impact on different mains at the time of application, and will share stakeholders. In particular, identify the the cost of connection above 50 meters with potential barriers to access that prevent these customers. However, domestic customers poor households from obtaining safe and are expected to pay 50,000 as a connection fee cost-effective water supply services. (about $20 in 2004). In addition to this strategy, the NWSC has implemented a number of options • Put the reform challenges in perspective for serving poor households that could not be by reviewing lessons learned. connected to the water mains. These include a range of community kiosks and water points. • Analyze the effect of the program on the whole population and particularly the It is widely considered that this strategy has poor, and determine whether the policy had a significant impact on urban water supply has produced the intended benefits and if coverage and financial sustainability of the resources have been spent effectively and NWSC, and that the lessons that can be learned efficiently. from the progress of the urban water sector and the improvement in services to all customers, • Make recommendations regarding including poor households, in Uganda, can policies, strategies, and design in the be of real value to sector practitioners and distribution and management of water development partners worldwide. supply in informal settlements and low income areas. 1.2 Study Objectives 1.3 Methodology Used and The aim of this study is to assess the impact of Data Collected the pro-poor policies introduced by the NWSC between 2004 and 2010, by analyzing the The assessment of the impact of the pro-poor effectiveness, efficiency, and equity of these policies on the poor in Kampala is based on policies and strategies implemented by the GoU an analysis of the effectiveness, the efficiency, and NWSC to increase coverage of urban water and the equity of these policies since 2004. The 4 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Box 1: Assessing efficiency, effectiveness, and equity of service provision to the poor Impact assessment criterion Description Effectiveness • Primarily related to assessing the number of new connections made in poor settlements versus the intended number of connections. • Indicators: increased coverage. Efficiency • Primarily related to assessing whether the approach and the service delivery mechanisms chosen to increase coverage were the right ones. The approach taken is to determine whether the blanket approach to targeting the poor achieved the intended increase in connections from poor households. • Indicators: increased coverage, NRW, time saved. Equity • Primarily related to assessing whether the poor are getting a similar quality of services than the nonpoor. • Indicators: increased quantity of water available to poor households, affordability. degree to which each of the indicators above supply and collection time. Estimations of contributes to assessing efficiency, effectiveness, this were secured from survey data from the and equity is summarized in Box 1. Uganda Bureau of Statistics (UBOS). Specifically, the study assessed the following: iv. Whether the cost of connection and the water tariff constituted barriers to access i. Whether access to water supply services for poor households. This was done by had increased in poor settlements. This measuring the cost of connections and was calculated by determining the total the tariff as a percentage of the average number of new connections that were household income in poor areas, also using made in poor settlements, compared data from UBOS. with the total number of new connections made throughout Kampala. v. Whether the level of nonrevenue water had reduced following the increased number ii. Whether the NWSC had increased the of customers connected to the network quantity of water sold to poor households. in Kampala. The NWSC’s adoption of This is expressed as average unit customer service lines was also examined consumption at yard taps, PWPs, and as a means of measuring efficiency. Focus prepaid PWPs. group discussions (FGDs) were undertaken with customers and local stakeholders to iii. If increasing coverage had led to a assess customer satisfaction both with reduction in the waiting time that poor the levels of service provided and with households spend in obtaining/fetching the processes involved in providing these water. This is expressed as distance to services. 5 1.4 Data Collected • Chapter 1 (this chapter) is the introduction of the report and presents the objectives and The following data is collected and analyzed: methods used. • Primary data from NWSC on 17,000 • Chapter 2 presents an overview of the domestic customers in areas of Kampala water sector in Kampala, including the legal, that are considered to be poor (principally institutional, and regulatory framework, a consumption and billing data). This was definition of poverty in Kampala and what obtained from the Billing and Commercial this means for poor households. Chapter division of NWSC with support from the 2 also provides the baseline situation (as at Urban Pro-Poor Branch. 2003/04) against which the assessment of the impact of the Affordable Connections • Secondary data, that is reports from the Policy is made. UBOS, Department of Water Development (DWD), KCC, NWSC and the recent • Chapter 3 describes the urban water sector GPOBA funded project in Kampala. policy and the strategy developed by the This included poverty data as well as NWSC to expand services and improve financial and economic cost and sector financial performance. This Chapter also performance data from NWSC Annual presents the policy components and Reports. delivery mechanisms implemented to serve domestic households in Kampala, including • Data obtained from the NWSC’s GIS the creation of a dedicated pro-poor unit, department to identify the location of the the pilot testing of geographical targeting water mains and low income settlements. of poor households via an output-based subsidy from GPOBA, and prepaid PWPs. • Qualitative data obtained from focused interviews with customer representatives, • Chapter 4 analyzes in detail the impact NWSC senior management, DWD staff of NWSC policies and water supply and local NGOs (including Community delivery mechanisms on services to poor Development Initiatives or CIDI, and households in Kampala. The assessment Concern Worldwide). A total of 18 FGDs considers coverage, consumption, distance were undertaken in 17 parishes and 76 to and waiting time at water sources, key informants were interviewed. A variety affordability of new water services, and of stakeholders and consumers were impact on NRW. Chapter 4 includes interviewed, including domestic and PWP specific recommendations for improving the users, male and female users, protected effectiveness, efficiency and equity of the and unprotected spring users, as well as social connection policy. These can be used water resellers and vendors, community by other utilities in Africa who are planning water organizers, community leaders, to implement a social connection policy. local water business entrepreneurs, and politicians. • Chapter 5 contains the conclusions of this study. 1.5 Structure of the Report • Chapter 6 contains recommendations for further analysis which could be undertaken The report is divided into the following six by the NWSC, GoU, and/or the World Bank. Chapters: 6 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Overview of the 2 Water Supply Context in Kampala 7 2.1 Legal, Institutional, and (i) The National Water Policy (1999): Regulatory Framework The objectives of the National Water Policy of The urban water supply sector in Uganda Uganda are: is dominated by the Ministry of Water and Environment (MoWE), which is responsible for • To manage and develop the water setting policy and overall sector direction, and the resources of Uganda in an integrated and NWSC, which is a government-owned parastatal sustainable manner, so as to secure and responsible for service provision in all urban areas provide water of adequate quantity and of Uganda. quality for all social and economic needs of the present and future generations with The reforms in the water and sanitation subsector the full participation of all stakeholders. were carried out between 1998 and 2005 with the objective of ensuring that services are • To provide safe water and hygienic provided and managed with increased attention sanitation facilities within easy reach to performance and cost effectiveness, to based on management responsibility and decrease the government burden (due to low ownership by the users to 75 percent cost recovery) while maintaining the government’s of the population in rural areas and 100 commitment to, and investments in, equitable percent of the urban population by 2015, and sustainable provision of services in the and to ensure that 80 percent–90 percent country. of these facilities are effectively used and remain functional. The underlying policy reforms framework (ii) The National Water and Sewerage Corporation for the WSS sector includes a shift to Act (2000), which sets out the roles and community demand for services; decentralized responsibilities of the NWSC and describes implementation and management of services; the institutional framework within which it government facilitation; regulation and quality operates, including the substantial autonomy assurance; and delivery of goods and services by which the utility benefits from. the private sector. 2.1.2 Institutional Framework For the urban subsector, the reforms emphasized the need for separation of asset ownership The MoWE has three directorates: and operations; commercialization of service delivery; establishment of an effective regulatory • The Directorate of Water Development framework; and engagement of the most (DWD); appropriate form of private sector participation. • The Directorate of Water Resources Management (DWRM); and This Section provides an overview of the legal, • The Directorate of Environment Affairs institutional, and regulatory frameworks which (DEA). led to and underpinned the water sector reform process in Uganda. The DWD is responsible for providing 2.1.1 Legal Framework overall technical oversight for the planning, implementation, and supervision of the delivery The legal documents most relevant to the urban of urban and rural water and sanitation services water sector include: across the country. However, the NWSC is responsible for water supply service provision in 8 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements the 23 large urban centers across the country, • Keeping stakeholders informed about including Kampala, Entebbe, and Jinja. service performance and activities of the service providers through information The DWD has also established 95 performance- dissemination. based contracts with Water Authorities in small towns that are outside the NWSC’s control. Other relevant regulatory functions such as As such, the DWD’s structure includes a abstraction licensing, water quality surveillance, Performance Contract Review Committee. and permits for wastewater discharge are held by other departments within the DWD. The 2.1.3 Regulatory Framework independence of the regulation unit can be questioned; the sector is, however, planning for There is no independent regulator for the urban gradually establishing an independent regulator. water supply sector in Uganda. Instead, the DWD monitors the quality of service provided The structure of the institutional and regulatory by the NWSC through specific targets and key framework is illustrated in Figure 1. performance indicators (KPIs) such as coverage, financial and operating efficiency, unaccounted for water, and customer service indicators. 2.2 NWSC Internal Reforms Since 2009, regulation of water services is carried The GoU initiated the reorganization of NWSC in by out the Regulatory Unit of Urban Water Supply 2000 by amending the NWSC Statute and Act. Department within the DWD. This department This provided substantial operational autonomy was set up after extensive consultations to the NWSC to operate services on a financially among key stakeholders. The previous mode and commercially viable basis. The Corporation’s of regulation was by contracts (as there was no debt service obligations ($100 million prior to regulator). The functions of the Regulation Unit 2000) were temporarily suspended by the GoU include the following: on the condition that the NWSC makes significant financial and operational improvements. The • Setting standards for services provided to government also gave the NWSC the political consumers, including targets in relation to mandate to implement pro-poor reforms. asset management, technical inputs and operational performance, and promotion of pro-poor service delivery, 2.2.1 Key Elements of the Internal Reforms • Monitoring operational performance of The key elements of the turn-around of the service providers (including NWSC) against NWSC, designed to create an efficient and set targets, and ensuring compliance with financially sound organization, were: contractual obligations among the parties to the performance and management i. Clear goals set by the government, contracts, imposing penalties or remedies including more autonomy (less for substandard performance, and interference) and support provided to creation of comparative competition in the the NWSC including additional CAPEX, market, budget, tariffs, internal policies, and strategies. • Reviewing requests and proposals for new tariffs, adjustment of existing tariffs for ii. Strong leadership by NWSC management approval, and ensuring that only approved with the long term vision of being one of tariffs are applied, the leading water utilities in the world. 9 iii. Focusing on strengthening the financial iv. Creation of a professional incentive-based credentials of the Corporation, including structure (at organization, business, cutting costs, reducing operational and staff levels) and associated internal inefficiencies, improving billing and monitoring and evaluation systems. revenue collection through improvement of customer metering and reduction of illegal v. Benchmarking performance with connections, and increasing the number of international best practice. registered connections/customers. vi. Creation of a customer-oriented culture, Figure 1: Institutional and regulatory framework for urban water supply with the NWSC’s current motto: “The services in Uganda customer is the reason we exist”. vii. Improved donor coordination through sectorwide approaches to planning, Ministry of Water and Environment (MoWE) and use of donors for expanding the knowledge base of the corporation. Directorate of Water It is clear from the above that in 2000 there was Development (DWD) no deliberate focus on expanding services to Directorate of Directorate of the poor. Instead, the focus was on improving Water Resources Environment Affairs commercial performance and financial autonomy, Management (DWRM) (DEA) and on reducing operational inefficiencies. Urban Water Supply Department (UWSD) 2.2.2 Focus on Performance and Performance Benchmarking Overall performance based Overall performance based The sector reform process included significant contract supervision contract supervision changes within the NWSC. To increase accountability and performance, the Corporation created and set up Internally Delegated Area National Water and Sewerage Management Contracts (IDAMC) between Corporation (NWSC) Other Water Authorities the head office and each of the urban areas (managed by Town and Municipal Councils as Water in its area of service. These internal branches Head office (based in Kampala) of the NWSC are all financially independent, Authorities performance contract with MoWE) with managers appointed through individual Performance contracts for staff performance contracts, which mimic the business units typical performance clauses (including bonuses and penalties) of private sector participation Performance contracts with contracts. The head office is, thus, overseeing Internal Delegated Area local private operators the performance of 23 IDAMCs, including that of Management Contracts (IDAMCs) (e.g., Kampala Water, Kampala which is the largest. Jinja Water, Entebe Water) The Kampala IDAMC is itself subdivided into Performance contracts for managers 74 Town Councils and 9 Town nine water supply and sewerage branches. The Boards have management managers of each of those branches are also contracts with local private contracted through individual staff performance Business Unit Managers water operators contracts. Typical targets (at branch and at business level) include: 10 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements • Increase sales—including billing and To this date, the NWSC is one of the only utilities revenue collection. in the African region that has adopted automatic (annual) tariff adjustments. However, tariffs do • Reduce unaccounted for water. not yet allow the NWSC to service old debts or • Increase the number of customers—and cover the costs of commercial borrowing. To therefore sales. enable loan servicing the tariff would have to be increased by about 60 percent (according to • Increase productivity—and cut costs. PEAP, 2004). • Increase and ensure customer satisfaction. 2.3 Definition and Incidence of Poverty Expansion of services to poor areas is not in Kampala considered as a contractual KPI—either at business level or at branch level, although the 2.3.1 Definition of Poverty construction of water points is one of the many activities undertaken by each of the branch Poverty in Uganda is measured in absolute terms engineers/operators. by measuring the level of income against the expenditure needed to secure basic food and 2.2.3 Autonomy for Setting Tariffs nonfood items. This is determined through a national household survey. The trends in poverty The NWSC’s autonomy to set its own tariffs (to show a decline in poverty at the national level be reviewed and approved by the MoWE) came from 55.7 percent in 1992 to 31.1 percent in in 2004, when the Corporation was able to 2005–06 (Uganda National Household Survey, implement a differentiated and flat tariff structure. UNHS, 2005). In urban areas, the level of poverty This means that there is a specific flat rate tariff came down to 13.7 percent in 2005–06. In the for each of the NWSC’s customer categories Central Urban region in which Kampala city falls, across all water supply areas: industrial and the number of poor households is estimated to commercial, government and institutional, be 127,755. This number is growing. and domestic. An additional flat rate tariff was developed for PWPs. The MoWE considers that To appreciate the extent of poverty in its supply the flat tariff structure (per customer category) areas (in particular Kampala), and to develop ensures equity in supply and pricing.11 concrete plans to provide these households with sustainable services, the NWSC has developed Although the tariff structures are reviewed and its own definition of poverty. This is based on four approved by the DWD and MoWE, the tariff parameters: levels have, since 2004, been subjected to automatic annual indexation against the domestic • Poor households have monthly incomes Consumer Price Index (inflation), exchange rate, of less than UGX80,000 ($48), which is foreign price index, and electricity tariffs. This about one-third of the individual house was designed to protect the company’s financial connection fee (at $125) and 14 percent equilibrium and to maintain real value for the tariff. of the average cost of connection ($350). These automatic adjustments help ensure that the NWSC is able to recover all its operating and • Poor households live in clustered maintenance costs and to make provision for settlements with a high crowding index of depreciation.12 up to 14 persons per household (although 11 MoWE. 2011. Uganda Water and Environment Sector Performance Report 2011. 12 World Bank. 2010. Uganda Poverty Reduction Support Credits (PRSCs 5, 6, and 7): Implementation Completion and Results Report. 11 the average household size is reported by The main reasons for this high rate of population Uganda Bureau of Statistics as six). growth are increased demand for employment, land for housing, as well as social services • Poor households have very low water and infrastructure that have stimulated spatial consumption of 0–20 liters per capita per urban development and industrialization and, day (lcpd) which equates to 3.6 m3/month as a consequence, a high rate of rural to urban for a household of six), as they use this migration. Additional demographic factors water mainly for drinking and cooking. include high fertility, the decline in mortality, internal migration, and immigration. New informal • Poor households do not have their own settlements have emerged while others are house connection and mainly use PWPs expanding, thereby increasing both the number (including kiosks). and the size of unserved (informal) areas. 2.3.2 Incidence of Poverty in Kampala Civil infrastructure has not kept pace with this explosion of population. This means that The incidence of poverty in Uganda in general, and the living environment of the urban poor can Kampala in particular, is difficult to assess because be characterized as follows: poor sanitation, the last population Census was undertaken in inadequate housing and urban infrastructure, 2002 (a new population Census will be undertaken poorly managed solid and human waste, and in 2012). However, a number of studies have increased environmental pollution. been carried out since the last one, which provide some guidance on the relative percentage of poor The challenge with identifying the number and households in urban areas and in Kampala. size of Kampala’s informal settlements is their fast and ever-changing nature. In 2002, the Kampala has grown very rapidly in the last two United Nations Human Settlements Programme decades. It has experienced a rapid population (UN-Habitat) estimated that 44 percent of the growth of nearly 4 percent per annum from population of Kampala lived in slums.13 This 774,241 in 1991 to 1,189,142 in 2002 (last corresponds well with the 2002 Census which Census). This is summarized in Table 2. The stated that approximately 120,000 households in population forecast for 2010 uses a growth rate of Kampala were poor. These informal settlements about 4 percent per annum. cover up to 25 percent of the total area of the city. Sixty-five percent of the city’s population has Table 2: Population growth in Kampala (1980 to 2010) 1980 1991 2002 2010 (projection) Kampala population 458,503 774,241 1,189,142 1,597,800 National urban population 938,503 1,889,622 2,921,981 3,992,180 National urban growth rate p.a. 4.88% 3.98% 3.98% Kampala as % of national urban Population 48.85% 40.97% 40.69% 40.02% Source: UBOS Census 2002 data (selected). 13 UN-Habitat. 2002. Addressing the urban water challenges in Kampala. 12 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements access to clean and safe piped water, the rest significant challenge for the orderly development gets water from unhygienic and highly polluted/ not only of slums but of Kampala City in contaminated sources. general.15 Several sources converge on the estimation that The main issue with the land ownership system more than half the population of Kampala lives is that approximately 50 percent of Kampala16 is in slums. A study undertaken by the Ministry of held under a freehold (or similar, Mailo) system Lands, Housing and Urban areas (MoLHU) in in which KCC and the central government are 2008, for example, estimated that 64 percent unable to plan, control, and enforce orderly of the population of Kampala lived in slums (an urban development. Freehold (and Mailo) increase of 20 percent over six years), adding landowners (including absentee landlords and to approximately 170,000 households. The secondary, illegal, owners) cannot be forced to MoLHU’s findings are echoed by the World Health provide infrastructure services to the informal Organization (WHO) (2010) which states that 60 residents that have settled on this land. As such, percent of the population of Kampala lived in development in these areas is haphazard and slums and lacked access to utilities and amenities. continues to be informal. Likewise, UN-Habitat (2007) states that over 60 percent of the population of Kampala lives in This has huge impact on the NWSC’s ability slums.14 and incentive to provide sustainable water supply and sanitation services in Kampala, as The NWSC has developed a map of the city the Corporation expects to see documentary of Kampala divided into 500m by 500m block proof of land ownership, or a sale agreement. maps—117 of these block maps have been Thus, squatters are not recognized as potential identified (and are considered) as being poor. The customers. The only way forward for the utility map showing the extent of the distribution network is to provide PWPs on private land (where the and the location of known poor settlements in landlord often becomes the PWP/kiosk operator) Kampala is included in the Appendixes. or to provide services on land which is not under Mailo/freehold ownership, such as that which is likely to be on the outskirts of the poor 2.4 Land Tenure and Its Impact on settlements. Rigid water supply connection Services to the Poor requirements based on formal land tenure are preventing poor households from accessing The region in which Kampala is located has a improved levels of service. complex land tenure system. One system is ‘mailo.’ The basic unit of the mailo system is a square mile, hence the derivation of mailo, which 2.5 2002 Baseline of Urban Poor is also equivalent to 640 acres. The term is used Coverage Areas in Kampala in Uganda to describe a land tenure system that came into effect when the Kingdom of Buganda 2.5.1 Background of Water Supply Services signed an agreement with the British-administered in Kampala’s Poor Areas Uganda Protectorate there in 1900. Other systems are freehold, leasehold and customary tenure. The terrain of Kampala is hilly and has many The Kingdom of Buganda owns 52 percent of the natural springs. The poor have settled in the land in Kampala. The system is described as a low lying swampy areas, which were less 14 See: “Situation Analysis of Informal Settlements in Kampala - Cities without Slums Sub-regional Programme for Eastern and Southern Africa” (Series title) (this is a report prepared in partnership with Kampala City Council). The report focuses in particular on two parishes: Kivulu (Kagugube) and Kinawataka (Mbuyu 1). 15 UN-Habitat. 2007. See Footnote 14. 16 Kampala Urban Sanitation Project (2006). 13 attractive and therefore cheaper than the areas Field visits also found that many of the sources located higher up the hills. A study of 160 had not been labeled as unfit for human springs in Kampala by Chemiphar (2006) in 35 consumption, as was the accepted practice by urban poor parishes indicated that 80 percent KCC staff. In addition, many of the 307 PWPs of the inhabitants in these urban poor parishes and 59 protected springs built by the KCC with obtained domestic water from springs (protected support from the French Development Agency or unprotected); that 79 (49 percent) of these (AFD) between 2002 and 2006 (following cholera sources were totally unfit for drinking, another 79 outbreaks in the late 1990s) had become (49 percent) could be used for drinking once the dysfunctional due to poor operation and water was boiled, and that only two sources (1 maintenance (O&M). The main reason was found percent) were actually safe to drink (but were not to be the backlog of huge arrears accumulating chlorinated). from nonpayment of PWP/kiosk operators to the NWSC. The study also confirmed that other sources of supply were available (not counted in the This meant that the NWSC had to disconnect Chemiphar study), including hand dug wells those nonpaying PWPs and kiosks (in total and protected springs made by charitable between 5 percent and 10 percent of these individuals and/or organizations, and a limited were disconnected). In addition, field surveys number of PWPs. As a consequence, and further undertaken in 2011 showed that the so-called exacerbated by low competition (and perhaps protected springs had been constructed upon— collusion), the cost of piped water to poor sometimes with—latrines, rendering these so- households was expensive due to significant called protected springs unsafe. reselling (and little real competition). The survey also showed limited access to water points, 2.5.2 Water Supply Coverage unaffordable prices charged at the existing water points, and the perception from poor households The typology of water supply services provided by that the alternative sources could provide drinking the NWSC to domestic customers is described in water. Box 2. Box 2: Typology of water supply connections in Kampala • House connection: Individual metered house connection to the water supply network where the point of supply is inside the house, and therefore associated with internal plumbing, showers, taps, sinks, storage tanks, and so on. A house connection is always associated with some kind of drainage system (network, pit, and so on). • Yard tap: Similar to the individual house connection except that the point of supply is in the yard and is usually shared between a small number of households (usually two to three). A yard tap is not usually associated with internal plumbing or a drainage system. • Public water point: Metered connection which is typically a battery of taps fed by one connection and recognized as such by the NWSC. Public water points are usually managed directly by the NWSC or someone appointed by it (for example, a community group, an individual, and so on). • Kiosk: Typically, a battery of taps developed by a private individual (operator or community-based organization) to sell water to the public. Kiosk operators usually charge a mark up on the normal tariff. A kiosk implies a formal structure and attracts the preferential public water point/kiosk tariff. 14 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements According to the 2002 Census, 98 percent of households in Kampala had house of households in Kampala had access to connections in 2002 (this includes a house safe drinking water sources (tap/piped water, connection and/or a yard tap).17 borehole, spring water, gravity flow, and rain water) while 2 percent used water from unsafe • About 35 percent of households, sources. Additional data from the National presumably the poorest, obtain water from Service Delivery Survey (2004) suggest that boreholes, protected springs, and gravity access to improved water sources was also high schemes (40 percent in the wet season). It (but not as high), but that there was seasonal is considered that as much as 50 percent variation and significant use of alternative sources of households used these alternative of water. This data are summarized in Table 3. sources. Key observations are that: • As many households obtain water for • Only a limited number of households in drinking from safe sources as they do urban areas (NWSC-wide) had individual water for other uses than drinking from house and/or yard connections (22 unsafe (unimproved) sources. This could percent). This is consistent with NWSC mean that demand for NWSC water data that show that only 20 percent is potentially only half of total average Table 3: Sources of water in urban areas (National data, NSDS, 2004) Dry season Wet season Water source (urban only) Water for drinking Other uses Water for drinking Other uses Piped water in dwelling 9% 9% 9% 9% Piped water in compound 13% 13% 12% 13% Piped water outside compound 27% 27% 24% 23% Borehole, protected spring, gravity flow scheme 39% 33% 33% 26% Rain water 0% 0% 13% 16% Unprotected source 8% 11% 6% 8% Lake, river, stream, pond, dam 3% 6% 2% 5% Other 1% 1% 1% 1% Total 100% 100% 100% 100% Breakdown (improved /unimproved) Water source (urban only) Water for drinking Other uses Water for drinking Other uses Improved 88% 82% 91% 86% Unimproved 12% 18% 9% 14% Total 100% 100% 100% 100% 17 The Poverty Eradication Action Plan (2004) mentions urban water supply coverage of only 18 percent with house connections and/or yard taps. 15 Table 4: Distance to nearest water source (Kampala slums, 2002 Census) Name of parish/division On % on Up to % up 1 to % 1 5 km % 5 km premises premises 1 km to 1 km 5 km to 5 km and over and over Total Naguru II/Nakawa 2,342 37% 3,906 62% 24 0% - 0% 6,272 Banda/Nakawa 220 7% 3,084 93% 7 0% 1 0% 3,312 Kisenyi II/Central 854 31% 1,943 69% 3 0% - 0% 2,800 Kamwokya II/Central 1,165 26% 3,320 74% 12 0% 9 0% 4,506 Katwe I/Makindye 390 15% 2,150 83% 35 1% - 0% 2,575 Wabigalo/Makindye 730 15% 4,252 85% 20 0% - 0% 5,002 Bwaise III /Kawampe 430 14% 2,544 85% 16 1% 1 0% 2,991 Mulago II/Kawampe 699 17% 3,404 83% 3 0% 1 0% 4,107 Nakulabye II/Rubaga 1,015 17% 4,666 79% 197 3% 8 0% 5,886 Ndeeba/Rubaga 706 14% 4,245 85% 57 1% 1 0% 5,009 Total 8,551 20% 33,514 79% 374 1% 21 0% 42,460 Note: There is no breakdown of distance for less than 1 km demand—particularly in unserved/poor is thus little information available on the level of settlements. suppressed/unmet demand in poor settlements. • A significant number of households use rainwater for drinking (13 percent) and for There are some difficulties with providing water other uses (16 percent) during the wet supply services to poor households in Kampala. season. These are described in more detail below. Table 4 provides some additional insight as it 2.5.3 Income and Expenditure Review shows the distance to the nearest water point for (Affordability and Willingness to Pay) households in the same parishes as above (also based on the 2002 Census). These are shown in The 2002 Census, the 2004 NSDS survey Figure 3. and household surveys as well as studies by a number of consultants have indicated that the The Census 2002 data (see above) show that willingness to pay for improved water services 79 percent of the population in the poor areas was high among poor households. Generally, of Kampala obtained water (drinking water) households were ready to pay the official NWSC from water points located up to 1 km from their rates for water, but there remained a significant homes, and implies that households that do not proportion of households who cited the cost of have access to water on their premises are able water (or perhaps the cost of access the water, to access this water from another piped source that is, the cost of connection) to be high. These such as a PWP/kiosk. households may also have been commenting on the high cost of water obtained from water The 2002 Census also does not comment on vendors (or owners of PWPs) who have applied whether the number of PWPs/kiosks provided significant mark ups. in these poor settlements was sufficient. There 16 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements The weekly expenditures in 2004, for water in The cost of connection in Kampala prior to three large poor parishes Kisenyi I, Kisenyi II, and 2004 included a connection fee of UGX125,000 Ndeeba with large informal settlements, were in the ($75). In addition to that, customers were range of UGX1,000 to 5,000, or 4,000 to 20,000 expected to pay for the full cost of connection per month (Beller Consult, 2006), which represents (service line) from the NWSC’s water mains to 5 percent to 25 percent of average monthly income their customer meter. This represented a total for poor households. This is confirmed by PEAP of $275 to $375 which was not affordable to (2004) which mentions that households in the poor households and meant that these very bottom 20 percent income quintile (who chose to households were incentivized to overlook the use and pay for NWSC water) devoted as much as quality of materials and workmanship used to 22 percent of their income to purchase water from construct their service lines. In turn, this caused PWPs/kiosks. significant losses to the NWSC from these poorly constructed service lines. In addition, 69 percent to 76 percent of the households surveyed were ready to pay for improved water services. Since this price level was The NWSC management, therefore, discussed already higher than that of NWSC standpost rates and implemented a number of strategies to of supply, it can be assumed that the households address the low coverage and high levels of would welcome domestic connections for clean losses. These strategies are discussed in more water supply (Sanitation Strategy and Master Plan detail in the following Section. for Kampala, 2004). © WSP 17 18 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements © Samson Gitonga Urban Water Sector Policy and Strategy 3 for Improving Services to the Poor 19 This Chapter presents the different elements of the The GoU also pointed out that the financing of pro-poor policy of the GoU, and describes the pro- urban water was a public responsibility and that poor strategy that was implemented in 2004 by the the investment plan would be difficult to finance NWSC. within the GoU’s existing available budget/ envelope. The PEAP was, therefore, used as The GoU’s first focus on the poor came with the a key input into the World Bank’s PRSP and Poverty Eradication Action Plan (PEAP), which lead associated PRSC. The strategy for achieving to developing and informing the Poverty Reduction PRSC targets for the urban water sector was Strategy Papers (PRSP) and Poverty Reduction based on the following components: Strategy Credits (PRSC) and specific amendments to policies and government Acts. Thus, the PEAP • Sector-Wide Approach (SWAp) to is the cornerstone of all pro-poor interventions in planning. the country. • Autonomous management of service provision by professionals. 3.1 GoU’s Poverty Eradication Action Plans • Emphasis on commercialization, financial viability, cost recovery, and operational In 1997 Uganda was the first country to prepare efficiency. a comprehensive and participatory national • Use of the private sector in construction development strategy, the PEAP. One of the main and management of water supply objectives of the GoU, as declared in its successive schemes. PEAPs, is to improve the equitable distribution of water resources in the urban areas, and thus • Ownership of assets vested in the public to improve coverage and the volume of water sector. available to households, in the spirit of “less for all • Independent regulation. rather than all for some”. This Section outlines the objectives and pillars of the PEAPs as well as the • Sustainability and replicability of national planning framework that they supported. approaches. • Pro-poor approaches. The first PEAP outlined a basic change in the government’s strategy to deliver basic services The overall national framework supported by to the poor by recognizing the multidimensional the first PEAP (1997) lead the GoU to develop a nature of poverty and adopting a multisectoral number of sector specific policies and strategies, approach for its eradication. Under this plan, including the National Water Policy (1999) and Uganda is being transformed into a modern the Pro-Poor Strategy of the Ministry of Water economy in which people in all sectors can and Environment (2004). These are described in participate in economic growth. more detail here. Priority actions for the water sector are identified in PEAP (2004) as follows: 3.2 The Government’s Pro-Poor Water Policies and Strategies • Review the water sector targets in light of the Long Term Expenditure Framework Following the preparation and review of PEAPs, (LTEF) projections. sector stakeholders managed to find little evidence of water supply and sanitation service • Prioritize provision of water and sanitation delivery to poor households in urban areas. services to the rural population and the With support from budget support grants and urban poor. 20 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements credits from the corresponding PRSCs, the GoU carried out an assessment of past policies and developed the National Water Policy (1999) and strategies in 2004, and published an updated later the Pro-Poor Strategy of the MoWE in 2004. strategy in 2006. This is described in more detail below. 3.2.1 National Water Policy Targets and MDG Goals 3.2.2 GoU’s Pro-Poor Policy The National Water Policy of 1999 was A review of the pro-poor performance of current developed to provide further guidance on the policies, strategies, and practice by the DWD objectives, targets, and processes envisioned in (undertaken in 2004) revealed that, while water the first PEAP. The National Water Policy also lays sector policies were generally committed towards down the timeline for the water and sanitation improving the social and economic conditions of sector as follows: all Ugandans, the policies had a weak orientation towards improving actual services to the poor “Sustainable provision of safe water within easy and were generally unclear about the role of reach and hygienic sanitation facilities, based on the GoU and implications of these previous management responsibility and ownership by policies on the poor.18 The DWD’s assessment the users to 77 percent of the population in rural found that flat tariffs charged at PWPs were not areas and 100 percent of the urban population benefiting the poor because of significant mark- by the year 2015, with an 80 percent–90 percent ups charged by middlemen (for example, kiosk effective use and functionality of facilities” operators) and that the intended consumption (MoWE, 2004). subsidies embedded in the tariff were actually benefiting customers who were already It is noted here that the GoU’s water supply connected to the network, and not the poor.19 coverage target of 100 percent in urban areas by 2015 is more ambitious than the Millennium The government’s policy for serving the poor Development Goal (MDG) target of halving the (2004) therefore emphasized the following points: proportion of people without sustainable access to safe drinking water and improved sanitation by i. To ensure equitable allocation of subsector 2015. budget. As per the 2008 National Service Delivery Survey, ii. To improve overall sector performance. rural water supply coverage was about 67 percent, and the urban water coverage was 94 iii. To enhance access by increasing the percent, making it a combined national coverage density of network/points and expanding of 71.6 percent, thus meeting the United Nations to unserved areas. Development Programme (UNDP)-assessed iv. To target vulnerable groups such as MDG target of 62 percent, but still falling short of women, people with disabilities and HIV/ the final PEAP/ministry targets. AIDS. Although the national target was set at 100 v. To improve sanitation and hygiene percent there was no specific policy and strategy practice. on how to expand services to unserved (poor) settlements, nor were any investment programs vi. To monitor the impact of water supply and identified to deliver such services. The MoWE sanitation services on the poor: This led to 18 DWD/MoWE. 2004. Pro-poor strategy of the water and sanitation sector. The assessment was undertaken in 2004 by the DWD (Finance Thematic Subsector Working Group) but the actual policy was written in 2006. The full report is included in the Appendixes. 19 These were common weaknesses of consumption subsidies in sub-Saharan Africa. 21 the monitoring and evaluation process of activities that were considered to have significant the MoWE’s Sector Performance Report. positive impacts on the quality and sustainability of services to poor households. The main vii. To encourage participation of the poor and components of the NWSC’s pro-poor strategy giving them a voice, through mechanisms were: such as Water User Associations (WUA). • Component 1: An Affordable Connections viii. To establish more PWPs in poor Policy20 which helps all customers, settlements. including the poor, to connect to the ix. To create a pro-poor tariff structure: Set network by subsidizing the cost of the up cross-subsidies between different connection. The policy is supported by a customer categories. New Connection Fund which is funded through an 11 percent surcharge in the x. To subsidize yard connections serving water tariff. as authorized PWPs: An arrangement where the water authority would work with • Component 2: A differentiated flat tariff existing yard tap owners to use the yard structure which allows for consumption taps as designated PWPs, managed by subsidies to be targeted to poor the respective owners on behalf of the households, and a certain level of cross authority. subsidy between industrial and commercial customers on the one side, and domestic xi. To recognize that the poor only benefit customers on the other. from the consumption subsidies if they are connected to (or have access to) • Component 3: A Pro-Poor Targeting the network. This policy is supported by Project which the NWSC has been specific activities designed to increase implementing since 2008 with GPOBA water supply coverage support. This project is providing yard taps, PWPs and prepaid PWPs in targeted poor xii. To continuously monitor water quality. settlements of Kampala, and is therefore a shift from the Affordable Connections Following this new policy, the NWSC developed Policy which is not specifically targeting the its own pro-poor strategy and pro-poor service poor. delivery mechanisms. The NWSC used different delivery mechanisms to 3.3 NWSC’s Pro-Poor Policy implement its pro-poor strategy. These include: and Instruments • Mechanism 1: Setting up a dedicated The NWSC’s specific approach for providing pro-poor unit (called the Urban Pro-Poor services to unserved (poor) settlements was Branch) in Kampala. developed in 2004 and officially published in • Mechanism 2: Giving customers a choice 2006 to follow the MoWE’s Pro-Poor Water and levels of service: Sanitation Sector Policy. PWP/kiosks; Presidential Pledge taps; However, the NWSC’s approach for serving Yard taps; and poor households is not enshrined in a specific Prepaid connections—only pro-poor strategy. Rather, NWSC management implemented in 2008. developed and implemented a number of 20 The policy is actually called “The New Policy of Providing and Maintaining Service Lines to Customers”, but for ease of reference shall be referred to as Affordable Connections Policy throughout this report. 22 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements These components and mechanisms for represented an annual investment of delivering pro-poor services are described in UGX4 billion21 ($2 million). more detail below. • The costs of constructing connection lengths greater than 50 meters were 3.4 Principal Components of the shared between the customer and Pro-Poor Policy the NWSC. In addition, the cost of reconnection (for example, following 3.4.1 Component 1: The Affordable disconnection or illegal connections) Connections Policy was reduced to UGX75,000 or $45 (with the same policy based on length of The NWSC’s Affordable Connections Policy was connection). adopted on July 1, 2004. The principal objectives of the policy were to: The NWSC recognized that, to be financially sustainable, the connections policy needed to be • Increase the number of connections for internally funded. Therefore: all customer categories (domestic and nondomestic), thereby making water • The Affordable Connections Policy was accessible to the population, especially the funded via a 10.7 percent tariff surcharge urban poor. applied to domestic, commercial/ industrial, and institutional/government • Ensure standardization in the quality of customers,22 with most of the subsidy materials used for connection, and thus coming from commercial and industrial contribute to reducing technical losses due customers and government institutions. to leaks and bursts within the service lines. Previously consumers were responsible for • Revenue from the tariff surcharge was paying for their own connection to NWSC to be ring-fenced in a dedicated New mains, and therefore had an incentive to Connections Fund and used exclusively use cheaper/nonstandard materials which for new connections in line with the policy. led to leakages and loss of water. The Affordable Connections Policy is derived from Highlights of the policy include: the following policy documents (NWSC Board Paper 571 of 2004): • The connection fee for a standard half- inch connection was reduced from • The NWSC Corporate Plan: 2003–2006 UGX125,000 ($75) to 59,000 ($35). Strategic Goals, in which it is stated Rates were similarly cut across the that “the NWSC is to create funds for board for other sizes of connections (and maintenance, new connections, and nondomestic customers). network extensions as part of its strategy to ensure that investments and new • The NWSC should meet all the costs of connections are implemented on a timely connections within a radius of 50 meters basis”. from the NWSC supply point. Thus the NWSC mobilized and provided all • The Performance Contract with the GoU: materials (pipes and fittings) and labor 2003–2006, which states that “The NWSC (trenching and pipe laying) for construction will undertake the introduction of a “New of the service line up to the customers’ Connection Fund” mechanism in the tariff meter. The NWSC estimated that this policy that will help accelerate provision 21 The breakdown as estimated by the NWSC is shown in the Appendixes. 22 However, large industrial consumers using more than 1,500 m3 per month were not included in the 10.7 percent tariff surcharge (and therefore were not included in the cross-subsidy pool). 23 of services to the poor and also curb the went a step further and also implemented a incidence of unaccounted-for water”. water tariff specific to public water points—as a means to subsidize water consumption to poor • The Poverty Reduction Support Credit households. (PRSC IV) targets which provide that within the period April 2004 to March 2005, the Thus PWPs/kiosks were given the lowest NWSC will develop a water tariff structure water rates, followed by domestic connections aimed at encouraging new connections and commercial tariffs (which were the most through the establishment of a Connection expensive). The tariff was designed such that Fund. This would also address the access large consumers, including domestic customers, of safe water to the poor. government/institutions, and commercial/ industrial customers using up to 500 m3/month The Affordable Connections Policy developed cross subsidize the water supply tariff at PWPs. and implemented by the NWSC since 2004 was It was considered that industrial consumers not based on a detailed assessment of needs, using more than 1,500 m3/month should not be capital investments, and funding sources by included in the tariff surcharge—as they might water supply zone or area in Kampala per se, decide to develop their own sources of water and but rather on the Corporation’s estimate of the thus no longer use the NWSC’s. number of new connections it considered it could deliver annually and on its ability to raise enough The water tariff has been indexed on the funding for new connections through a tariff Consumer Price Index and on fuel and energy surcharge. prices also since 2004, so as to protect the NWSC’s ability to recover its O&M costs from In addition, the Affordable Connections Policy collected revenue.23 The water tariff for 2008–09 enabled the NWSC to plug a major loophole in is shown in Table 5. the institutional and operational framework as it agreed to adopt, and therefore operate and With these automatic increases, the NWSC is maintain, all customer service lines between the one of the very few African utilities that is able NWSC mains and the customer meter. Previously to protect its revenue generation from a large (that is, before the change in ownership and increase in operational costs, for example, due to management responsibility) the NWSC was not rising electricity prices. responsible for these lines—and neither were the customers. That meant that losses from bursts Yard taps are considered as domestic and leaks on these lines remained relatively connections and therefore charged the domestic uncontrolled and explains the NWSC’s high level tariff. However, the NWSC analyzes the average of unaccounted for water in 2004 (at more than monthly consumption at the yard taps to 40 percent of water produced). determine whether these are shared between many families and thus used as PWPs. In that 3.4.2 Component 2: Differentiated and case, the NWSC automatically informs the Pro-Poor Tariff household under whose name the yard tap is A key element of the NWSC’s pro-poor policy registered that they will be charged the PWP was a differentiated tariff structure that allowed tariff. Therefore, yard taps that are used by two or for different flat rates for different types of three households are charged the domestic tariff. customers (essentially domestic, institutional/ government, and commercial/industrial). This The effectiveness of the tariff structure on was implemented in 2004. However, the NWSC services the poor is discussed in Chapter 4. 23 However, NWSC data on operational costs and collected revenues showed that the Operating Cost Coverage Ratios (OCCRs) for 2004, 2005, and 2006 were 77 percent, 76 percent, and 83 percent, respectively, suggesting that the NWSC required significant external subsidies to cover basic operation and maintenance (O&M) costs. However, these increased to 134 percent, 124 percent, and 134 percent in 2007, 2008, and 2009, respectively. 24 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Table 5: Kampala water prices in 2008–09 Type of connection Water tariff/m³ Water tariff/m3 (UGX) (US$) Presidential Pledge (no service charge) 784 0.31 Public water points (PWP) 867 0.34 Domestic (individual household) connection 1,341 0.53 Institutional/government 1,651 0.65 Commercial less than or equal to 1500 m³/month 2,085 0.82 Commercial greater than 1500 m³ /month 1,697 0.67 Average weighted water tariff 1,561 0.61 3.4.3 Component 3: Pro-Poor Targeting Box 3: Geographic targeting of poor households Given the need to continue to improve service provision (and coverage expansion to meet the The NWSC, with support from the GPOBA, is currently pilot testing a geographically targeted program in which subsidies are provided MDGs) to poor settlements in Kampala, and to the Corporation on the basis of the provision of agreed number of given that the Affordable Connections Policy outputs, including public and domestic yard taps in the areas where implemented by the NWSC since 2004 was not the poor live, with the added condition that they should be working deliberately focused on poor households, the at the end of 12 months of service. The project was initiated in 2008 NWSC has developed a new way of targeting to demonstrate that performance-based subsidies are an efficient poor households with support from the GPOBA, demand-based financing mechanism to extend services to the poor in designated (geographically targeted) areas. a World Bank administered trust fund (see Box 3). This project was initiated in 2008. A description The target is to provide 19,070 yard taps, 205 conventional PWPs, of the approach is included to illustrate the and 615 PWPs with prepaid meters in different slum areas. Increasing continued improvements made to the NWSC the access of piped water to the poor will lead not only to economic pro-poor strategy. benefits by reducing the opportunity cost of water but also result in better health and hygiene for the poor. 3.5 NWSC’s Pro-Poor Strategy The project uses a grant of $2.85 million and will reimburse 60 percent of the amount spent on the project in the form of OBA subsidies (upon Delivery Mechanisms successful delivery of outputs), while 10 percent is paid for by the beneficiaries as user contribution, and the remaining 30 percent will 3.5.1 Mechanism 1: Pro-Poor Unit be financed by the NWSC. However, the NWSC has to prefinance all investment costs (that is, 90 percent of the total cost), consistent with Following the implementation of the Affordable OBA principles. The Corporation is also responsible for funding all new Connections Policy, there was limited network reinforcements or secondary network extensions. coordination of the NWSC’s activities in the unplanned and unserved settlements. In fact, a To date 6,000 of the intended yard tap connections have been made situational analysis report undertaken as part of as the poorest households (targeted by the subsidy mechanism the German Development Agency (KfW) funded available to the NWSC through the GPOBA) are living amongst nonpoor or less poor households (for which no subsidy is available). project in Kisenyi-Ndeeba by Beller Consult Given the difficulty with conventional PWPs, the NWSC is going to (2006) states that: continue implementing yard taps and prepaid PWPs only. The likely benefit to poor households is expected to be high as the project is • The NWSC was focused then only on specifically targeting poor households in designated settlements of commercial efficiency and profit. Kampala. 25 • There was no specific pro-poor policy. 3.5.2 Mechanism 2A: Public Water Points • The poor are perceived as risky customers and Kiosks for a number of reasons including PWPs and water kiosks were the initial response vandalism, low consumption, need for by the NWSC and the KCC to the challenge watermains extensions, and difficulties of delivering water services to slums. They with land tenure certificates. constituted the first refuge of the slum population when they were afflicted with diseases such as With support from donors and development cholera and diarrhea. The establishment of PWPs partners, the NWSC established an Urban Pro- and kiosks helps in expanding access of water Poor Branch in 2006 to address the following services to the poor as they generally cannot objectives: pay for connections due to high costs or legal • Develop and implement tangible pro- difficulties associated with land tenure (discussed poor policies, procedures, indicators, and earlier in the report). Initially, a PWP is supposed targets. to cater for 25 households, serving between • Facilitate provision of adequate 150 and 200 people. The NSDS survey of 2008 infrastructure in informal settlements. shows that 26 percent of the people in urban areas obtain water from public taps, which • Set up strategic alliance with stakeholders includes PWPs, kiosks, Presidential taps and and increase networking with development prepaid taps. partners. • Implement a pro-poor approach in NWSC The initial trend of construction of kiosks was structures and provide an internal formal started by the KCC in the 1990s with the help feedback mechanism. of donor agencies and the connections were provided by the NWSC but, as the trend caught The NWSC’s Urban Pro-Poor Branch was set on, the number of kiosks increased fast. Other up to promote, plan, coordinate, and manage agencies and NGOs also started to make their service provision to unserved (poor) settlements own kiosks. All these are connected by the throughout the NWSC’s 23 service areas. The NWSC and provided with a water meter. unit/branch is set up as an individual internal advisory business and therefore has internal Often kiosks are managed by a private individual targets, both at business and staff levels. (or kiosk operator), who is generally the person who has donated the land, or his/her agent. The The Urban Pro-Poor Branch has a limited operator generally sells water at the market price number of seven staff, including O&M engineers, (not the NWSC price) but this is not regulated, billing and revenue collection specialists, and the usual unit of water is a 20-liter jerrican. socioeconomists, and customer care staff. The team is led by a socioeconomist. Field visits revealed that the price charged per jerrican of water ranged from UGX50 to 30024 Since its inception, the Branch has become ($0.02 to 0.12 per jerrican, or between $1.0 and a coordination point for the organizations $6.0 per m3)—depending on availability. The price (including donors, local authorities, and NGOs) is highly influenced by continuity of water (that is, working for the poor and takes up the issues interruptions), seasonal variations (and presence of the poor with the commercial branches. The of alternative sources), and competition from branch coordinates and oversees the pro- other water sellers). The average price charged poor mechanisms deployed by NWSC (see per jerrican was between UGX50 to 100 ($1 to paragraphs 3.5.2-3.5.4) 24 This equates to $0.02 to $0.12 per 20-liter jerrican, or $1.0 and $6.0 per m3. 26 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements $2 per m3), compared with the NWSC’s tariff at than on the street or on public land like PWPs. PWPs/kiosks of UGX867 per cubic meter ($0.35) The NWSC considers that the ratio of shared to or UGX17 ($0.007) per jerrican. Thus the poor individual yard taps is approximately 50:50. were charged a disproportionate amount for purchasing water at privately managed kiosks, A domestic connection is supposed to cater and were not benefiting from the intended for a household of an average of six persons. consumption subsidy available at PWPs/kiosks However, a yard tap is usually used by up to 15 through the differentiated tariff structure. people. Most of the connections are located outside the house and water from these is sold to 3.5.3 Mechanism 2B: Presidential Pledge the citizens at the prevailing market rate. (Yellow Water) Taps The NWSC is able to monitor the consumption Another delivery mechanism was launched during of yard taps and can decide, based on this the presidential election campaign in Uganda level of consumption, whether the yard tap is (2006), called the Presidential Pledge taps (yellow actually used by a large number of people or not. in color). This assessment is followed by a site visit by a representative of the Urban Pro-Poor Branch. If Presidential Pledge taps are similar to PWPs/ it is deemed that the yard tap is, in fact, being kiosks in all respects except that they are exempt used like a PWPs (that is, it serves more than from service (meter rental) charge (but not just two or three households), then the NWSC basic water charge). Thus, one cubic meter of automatically applies the PWP tariff so that the water costs UGX784 ($0.31) as compared with intended consumption subsidy reaches the poor. UGX867 ($0.34) at a conventional PWP managed on behalf of the NWSC. However, it is not clear 3.5.5 Mechanism 2D: Public Water Points why the meter rental charge was dropped only with Prepaid Water Supply Meters for Presidential Pledge taps and not across the board in poor settlements. Given the difficulty in collecting revenue from some of the PWPs/kiosks (whether these are According to the NWSC, the Presidential community taps or yard taps), the NWSC started Pledge (yellow) taps were obtained following pilot-testing the use of approximately 300 prepaid recommendations by political (elected and not water supply PWPs/kiosks in July 2008. elected) leaders in the communities. The “PP” denomination is only visible on the NWSC’s This was carried out with funding and technical customer database according to the date at assistance from KfW under the Kampala Water which the PP tap was applied for and made. and Sanitation Program phase 1. The premise is that customers have to obtain electronic tokens 3.5.4 Mechanism 2C: Yard Taps and ensure that these are credited, in exchange of which they can redeem this credit for drinking There are two types of yard taps: domestic yard water at designated prepaid PWPs/kiosks. taps are those used only by a household or a few families; shared yard taps are similar to PWPs Early results show that the prepaid water and thus shared by many more households. supply system, whilst more expensive than a Effectively shared yard taps are the same as conventional PWP, offers significant benefits PWPs in that they are used by many households, both to NWSC and poor customers. An but they are still located in someone’s yard rather early assessment of these prepaid supplies 27 indicates that the supply is available 24x7 to the prepaid supplies are installed. The exact poor households and that they are an effective arrangements are, however, still being tested. means of transferring the consumption subsidy to poor households, by removing the middle More commentary on the perceived strengths man effect. Security and cleanliness of the apron and weakness of the prepaid metering system is is ensured by NWSC staff in conjunction with included in Chapter 4. representatives of the communities within which © Samson Gitonga 28 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Analysis of the Impact of the Pro-Poor 4 Policies on the Poor in Kampala 29 This Section presents the findings of the The methodology used and data collected for assessment of the impact of the pro-poor analysis are presented in Section 1.3 of this policies on the poor in Kampala, including an report. assessment of the following: coverage increase, unit consumption, distance to supply and waiting time, affordability, NRW, and socioeconomic 4.1 Assessment of Coverage Increase impact. The NWSC’s Affordable Connections Policy Whilst the NWSC was able to exceed the targets provides yearly targets for increasing the number it set out in the Affordable Connections Policy of connections, by customer category. This is Document (see below), it has been unwilling or shown in Table 6, for the whole of the NWSC unable to set up the dedicated and ring-fenced service area.25 New Connections Fund, probably because it was not receiving sufficient revenue from customers 4.1.1 Overall Growth in Total Connections to cover all O&M costs, but also because it was and Population Served not forced to do so by the MoWE. In addition, whilst the number of new PWPs and yard taps Overall there has been a significant increase has exceeded their initial targets, the NWSC is in the total number of connections (including still facing a huge issue with nonpayment in poor individual house connections, PWPs, and yard settlements, so much so that the Corporation is taps) throughout Kampala: from 59,000 in now planning to retrofit all PWPs to prepaid water 2004 to 146,000 in 2010. This is a significant supply PWPs. This is discussed in more detailed achievement and is a testimony to the NWSC’s later in this Section. ability to adapt the size and management of its operations to increasing numbers of customers, Where appropriate, recommendations are made improve billing and revenue collection, and at the end of the Section on how the NWSC and manage larger operations (a larger network GoU could improve the effectiveness, efficiency, with more bursts, greater losses, and more and equity of the various elements of the pro- maintenance work). The Corporation increased poor strategy such that more benefit accrues to the size and the efficiency of its operations in poor households. parallel with increasing the number of customers. Table 6: Yearly connection targets of the Affordable Connections Policy (2004) Customer category Number of new New connections connections as % total Public water points 280 2% Domestic 11,900 85% Government/institutions 420 3% Commercial/industrial 1,400 10% Total 14,000 100% Note: No specific targets are included for yard taps as these are considered to be similar to domestic connections. In practice, therefore, the number of yard taps will be included in the total number of new domestic connections. 25 Note that no specific data is available for Kampala only. 30 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements It is considered that the organizational structure increase in connections from 2004 onwards is of the NWSC (with IDAMCs and individual staff double that of the pre-2004 period (14,500 and performance contracts) has contributed to these 7,000 new connections per annum, respectively). improvements. However, the rate of increase is dependent on The growth of NWSC connections from 2002/03 demand and the NWSC’s ability to implement to 2009/10 in Kampala area is shown in Table these new connections, both from the technical 7, in parallel with a population growth of about 4 (procurement, construction, availability of percent (except between 2007 and 2008 where sufficient water supply in the area) and financial it exceeded 10 percent). Coverage of piped (mostly the availability of funds in the new water supply has increased from 62 percent to connection fund) aspects. NWSC management 74 percent during this period, and has remained set (and therefore limited) yearly targets of stagnant at 74 percent since 2008, despite the approximately 10 to 15,000 new connections significant increase in connections. This may be per year (post-2004). This is considered prudent due to high population growth or to the fact that given the limitations on the water network and fewer households are sharing connections. water production capacity, as well as the NWSC’s own ability to carry out these connections. In However, in net terms this means that the NWSC addition, it is considered that much higher rates has managed to increase coverage at a rate of 8 of connections would have been achievable if percent points greater than population growth. there had been large scale investment programs for mains rehabilitation, mains extensions, and This data is presented in Figure 2. The graph water supply capacity increases—in parallel with clearly shows the increase in total water supply the new connections. This is discussed in Section connections from 2004/05, which is directly due 4.2. to the Affordable Connections Policy. The rate of Table 7: Growth in NWSC connections in Kampala (2002/03 to 2009/10) Year Total no. of Total % growth in Population % growth in Water supply connections population population served population served coverage 2002/03 (*) 50,302 1,208,544 749,297 62% 2003/04 59,022 1,255,341 17% 810,865 4% 65% 2004/05 72,231 1,302,138 22% 872,433 4% 67% 2005/06 93,305 1,351,619 29% 946,133 4% 70% 2006/07 105,531 1,402,981 13% 999,892 4% 71% 2007/08 119,897 1,554,818 14% 1,150,565 11% 74% 2008/09 132,882 1,635,744 (*) 11% 1,177,735 (*) 5% 72% 2009/10 146,109 1,716,669 10% 1,277,407 5% 74% Increase in population served (2004-2010) = 466,542 Note: * Data obtained from NWSC’s Annual Report 2009/10, calculated average between two successive periods. 31 Figure 2: Total water supply connections in Kampala (2002 to 2010) section of those consumers who could not pay one time (upfront) connection charges, could now afford safe on-premise piped water. This led to a doubling of the number of connections over the 160,000 period 2004 to 2009. 140,000 Table 8 and Figure 3 summarize the results of 120,000 the National Service Delivery Survey which was undertaken in urban areas26 of Uganda by the 100,000 UBOS in 2008. Overall, compared to the same Z survey that was prepared in 2004 prior to the 80,000 start of the pro-poor policy, it seems that: Y 60,000 • The proportion of households using X unimproved sources of water in urban 40,000 areas has reduced—both during the dry and wet season. 20,000 • The proportion of households using - 2001 /02 2002 /03 2003 /04 2004 /05 2005 /06 2006 /07 2007 /08 2008 /09 2009/10 piped water in their dwellings (that is, house connections and/or yard taps) has reduced—from 9 percent in 2004 to 5 X Rate of new connections (pre 2004): 7,000 per year percent in the dry and wet seasons of Y Implementation of the Affordable Connections Policy (2004) 2008, respectively. Z Rate of new connections (post 2004): 14,500 per year • The proportion of households that have access to water in their compounds has increased from 13 percent to 15 percent in Before 2004, the NWSC used to charge 125,000 2008. Similarly there has been a significant Shillings (US$75 at the then exchange rates) as increase (+16 percent) in the proportion of the basic connection charge and request users households with access to water “outside to pay for and undertake their own length of their compound”: this is interpreted as connection. This would cost another $200–300, PWPs/kiosks. thus amounting to a total cost of $275 to $375. It is considered that the NWSC’s customer base • The proportion of households using grew only slowly prior to 2004 because of these alternative sources of water (unimproved) large costs. has been reduced by half, from 12 percent in 2004 to 6 percent in 2008, with less As the charges for the basic half-inch (standard marked difference between the dry and domestic) connections were reduced from wet season. UGX125,000 to 59,000 (basic charge of UGX50,000 + 18 percent VAT, representing The following Sections present a more detailed US$35), the most important result of this assessment of coverage increase from yard taps, Affordable Connections Policy was that a large PWPs, and prepaid water supply PWPs. 26 Neither the NSDS of 2004 nor that of 2008 presents results for Kampala only. 32 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Table 8: Sources of drinking water in urban areas for dry and wet seasons (NSDS, 2008) National Service Delivery Survey Dry season Wet season Water for drinking Difference Water for drinking Difference Water source (urban only) since 2004 since 2004 Piped water in dwelling 5% -4% 5% -4% Piped water in compound 14% 2% 13% 2% Piped water outside compound 16% 16% 15% 15% Public tap 26% 24% Borehole, protected spring, gravity flow scheme 31% -8% 24% -9% Rain water 0% 0% 12% -1% Unprotected source 4% -4% 3% -3% Lake, river, stream, pond, dam 1% -2% 1% -2% Other 1% 1% 1% 1% Total 100% 100% Breakdown (improved/unimproved) Water for drinking Difference Water for drinking Difference Water source (urban only) since 2004 since 2004 Improved 94% 5% 95% 4% Unimproved 6% -5% 5% -4% Total 100% 100% 4.1.2 Growth in Domestic Yard Taps however, 6,793 yard taps and 1,032 PWPs and PWPs were constructed; see Table 8 and Figure 3— both show that the rate of new connections Data were obtained from the NWSC’s Billing and almost doubled in 2005–06 when the Affordable Commercial Department on the number and type Connections Policy was rolled out. However, from of new connections in each of the 117 known 2009 the rate of new connections fell back to its poor areas of Kampala from 1998 to 2011—as pre-2004 level although the total number of new a means to assess the average rate of new connections in Kampala (domestic, industrial, connections in poor settlements before and after and institutional) increased rapidly year on year. the pro-poor policies. The data included the date In fact, the annual target of 280 new PWPs in of installation of each connection (installation poor settlements projected by the Affordable dates older than 1998 are shown as having been Connections Policy has been met only over the made in 1998). period 2005–09 and has since then fallen back to pre-2004 levels, to about 70 PWPs par year. As shown in Table 8, over the period 1998–2011, Table 8 and Figure 3 also show that the NWSC at the NWSC carried out 16,198 connections in the time was already connecting about 700 yard the poor settlements of Kampala: 14,668 yard taps and 60 PWPs per annum. taps and 1,530 PWPs. Between 2004 and 2010, 33 Table 9: Total number of new connections (per annum and cumulative) from 1998 to 2011 in Kampala Year Total number of Total number of yard taps per annum PWPs per annum 1998 3,271 103 1999 499 27 2000 581 69 2001 700 69 2002 585 56 2003 758 65 2004 728 57 2005 1,269 129 2006 1,538 430 2007 1,303 240 2008 1,063 91 2009 851 66 2010 769 76 2011 753 52 Grand total 14,668 1,530 From the ACP 2,405 663 Thus, the Affordable Connections Policy has connections in the poor settlements were inactive: directly led to the construction of approximately 21 percent of yard taps and 53 percent of PWPs. 2,400 yard taps and 660 PWPs between 2004 Although the percent of inactive connections is and 2010. This means that approximately 4 reducing from one year to the next due to the fact percent of total new connections over the period that apparent debts are written off by the NWSC, directly benefited poor households—or potentially the percent of inactive PWPs has frequently been approximately 144,000 persons. greater than 60–70 percent in 2005–06 (and up Table 10 shows the average number of new to 80 percent in earlier years, for example, 1999 connections per annum from 2004 to 2010. and 2000). This highlights the difficulties faced The Table also shows that the NWSC’s targets by the NWSC in the management of PWPs and, (shown in Table 6) have consistently been met therefore, extending water supply services to low and/or exceeded, for all customer categories. income settlements via PWPs in Kampala. However, the figures shown in Table 10, in The main challenges faced by the NWSC when particular for PWPs, hide the difficulty which the providing PWPs is the nonpayment of the monthly NWSC faces with providing working solutions water bill, which leads to immediate disconnection in poor settlements in Kampala. Figure 4, which while the accounts stay on the customer is based on NWSC billing data, shows that as database until the debt is deemed irrecoverable and written off. of end of 2010 a noticeable proportion of new 34 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Figure 3: Total number of new connections (per annum and cumulative) from 1998 to 2011 in Kampala 1800 1600 New connections per annum 1400 1200 Yard taps 1000 800 Public standpipes 600 400 200 0 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 16000 14000 Cumulative number of connections 12000 10000 Yard taps 8000 Public standpipes 6000 4000 2000 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Table 10: Actual number of new connections per annum (2004 to 2010) for NWSC Target/year Actual 2004–10 Connection Number of new New Number of Number of Rate per New category connections connections connections, connections, year connections as % total 2004 2010 as % total PWPs 280 2% 3,114 7,748 772 3% Domestic 11,900 85% 85,717 194,848 18,189 77% Government/institutions 420 3% 4,261 6,686 404 2% Commwercial/institutions 1,400 10% 11,850 36,977 4,188 18% Total 14,000 100% 104,941 246,259 23,553 100% Note: Based on NWSC Annual Reports (although data for 2004 were extrapolated). The total number of connections relates to the whole of NWSC’s service area. 35 Figure 4: Number of inactive connections in Kampala (1998 to 2011) 350 300 250 Number of inactive connections 200 Yard taps 150 100 Public standpipes 50 0 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 100% 90% Proportion of inactive connections 80% 70% 60% % inactive yard taps 50% % inactive PSPs 40% 30% 20% 10% 0% 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 3000 2500 Cumulative of inactive connections 2000 Yard taps 1500 Public standpipes 1000 500 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 45% 40% Cumulative of inactive % 35% 30% % inactive yard taps 25% 20% % inactive PSPs 15% 10% 5% 0% 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 36 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Table 11: Estimate of the total number of poor beneficiaries of the Affordable Connections Policy (2004–10) in Kampala only Level of service Increase from 2004 to 2010 % inactive People served Beneficiaries Yard taps 2,405 18% 18 35,498 PWPs 663 36% 150 63,648 Sub-total= 99,146 Total increase in population served (see Table 7) 466,542 Total % poor beneficiaries over total beneficiaries 21% Table 11 shows an estimate of the total number Branch, the Corporation is spending more of beneficiaries of new PWPs and yard taps time planning for the provision of PWPs. in poor settlements since 2004. The estimate This has been followed by a significant takes into account the fact that the NWSC was reduction in the number and proportion of already providing approximately 700 yard taps inactive accounts (see Figure 4). and 60 PWPs per annum before the Affordable Connections Policy was implemented. • Inadequate network coverage in the poor settlements, which means that the number Thus Table 10 before this shows that, whilst only of yard taps and PWPs is also low in these 4 percent of the new connections provided in areas. The new connections need to be Kampala via the Affordable Connections Policy supported by corresponding water mains were made in poor settlements, 21 percent of extensions and/or water supply capacity the beneficiary population was in fact in poor increases that are not materializing as settlements. quickly as anticipated/needed. It is considered that had so many PWPs and • Difficulties with providing poor households yard taps not been disconnected/were inactive in informal settlements with connections— due to nonpayment, the impact of the Affordable yard and/or PWPs, because of the Connections Policy in poor settlements may have unavailability of suitable land. been much bigger (at about 31 percent of the total beneficiaries). In view of these challenges, the NWSC is currently piloting prepaid water supply systems The assessment concluded that the marked with a view to replicating these at scale—and slow-down in the rate of new connections in poor gradually replacing all yard taps and PWPs with settlements after 2009 is due to a combination of prepaid water supply PWPs. the following: 4.1.3 Prepaid Water Supply PWPs • Reduced willingness from the NWSC to provide PWPs without more involvement The NWSC is currently pilot-testing the from the community (for example, installation and use of, as well as customer socioeconomic surveys, ATP and WTP satisfaction with, prepaid water supplies, studies, customer education, and mainly in Kisenyi I and II, and Ndeeba. This new community participation). However, since approach started in 2008. To date, 300 prepaid 2007 and the setup of the Urban Pro-Poor water supply PWPs have been installed. This is 37 very slow progress and is attributed (in part) to However, the study consultations on the prepaid the long lead line in procuring (importing) prepaid water supply system highlighted a number of water supply meters. Prepaid water supply PWPs challenges, including: have principally been installed in areas where there was limited number of yard taps and a high • The major drawback is the very high number of inactive conventional PWPs. price of each of the prepaid water supply systems (about $1,350 compared to $380 Although it is too early to analyze the impact of for a PWP and $170 for a yard tap) and these prepaid water supply PWPs, some lessons the need to fund them through grants (that are appearing. The advantages of the prepaid is, the cost of the prepaid meters cannot PWPs include: yet be recovered through the pro-poor tariff). In the long run a suitable funding • The system is convenient and more mechanism is required if the prepaid water customer-friendly as it eliminates the need supply PWPs are to be used at scale.27 for an operator who is likely to apply a mark-up on the NWSC tariff, and who • Difficulty to locate the prepaid PWPs on is unlikely to work 24x7. Therefore, the suitable land. Due to the private nature prepaid PWPs ensure that poor households of land in most slums, it was not always really benefit from the consumption subsidy possible to find a willing landlord who embedded in the tariff. would offer free land for establishing a prepaid PWP. This may have been due • It virtually eliminates disconnection of to the fact that the initial prepaid PWPs the community due to nonpayment of a were not seen as usable given that the single individual (for example, the kiosk number of tokens was limited. In addition, operator), thus making it a favored option they may have been considered as lost for the NWSC. From the point of view income—since they are not managed by of the consumer it eliminates the fear of individuals. Later as the number of tokens inflated bills, or paying bills in arrears. In increased the reluctance of landowners to the daily cash economy of the poorest of part with land reduced. the poor, prepaid water supply systems are considered as a relief by consumers as • Difficulty with setting up the supply chain captured from the FGDs. of tokens such that these are sold en masse at a large number of locations/retail • For the NWSC it represents an opportunity outlets. to achieve 100 percent collection efficiency which is a significant improvement on the The initial success of prepaid PWPs has been poor performance of conventional PWPs. such that the contractual outputs of the GPOBA- The NWSC is considering using the prepaid funded project has been recast to support approach with large government and more prepaid PWPs. The project is now going institutional customers—to improve cash to establish 2,300 PWPs with prepaid meters, flow from these customers. estimated to serve in total 350,000 poor people in Kampala. As a comparison, the project is also • Finally, it is considered that the prepaid going to provide 6,000 new yard tap connections supply systems encourages water demand (serving 90,000 people) and 200 new PWPs management on a daily basis in the (serving 30,000 people). This is summarized household and discourages wasteful use of below. water. 27 However, the GPOBA-funded project is showing that prepaid PWPs are more cost effective than yard tap connections: average of $5.5 per capita for prepaid PWPs against $7 for shared yard taps. 38 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Outputs Initial OBA Progress made in 2010 in Revised OBA outputs the target areas (2010) 28 outputs since 2010 Yard taps 19,000 119 6,000 Public water points 400 7 200 Prepaid public water points 400 3 2,300 4.1.4 Effectiveness of the New connections may have been significantly larger, Connections Fund particularly in poor settlements. However, it is difficult to quantify how much more could have NWSC senior management indicated that the been achieved given that the number of new New Connections Fund had not been ring- connections (of all types), and the ability of the fenced, and that revenues collected through NWSC to deliver these, are linked to water the surcharge were not necessarily utilized for network characteristics (coverage, condition, expansion of connections. The Corporation and performance—affecting flow) and treatment reports that the funds were not earmarked for capacity (affecting volume). new connections because the approach met with competing demands of a commercial nature, The effectiveness of the Affordable Connections including assets renewal (buying pumps), rising Policy may also have been improved had cost of chemicals, electricity, and inflation. adequate financial support been earmarked sufficiently early on in the process to expand the It is precisely because utilities always have water distribution network to poor (unserved) competing needs for available resources settlements, although increasing the number of (especially when the utility is not recovering all its connections would have undoubtedly contributed operating costs from customer’s bills—as was to increasing revenue and stabilizing financial the case with the NWSC from 2004 to 2006) that performance. the fund needed to be ring-fenced. The NWSC is expected to have collected approximately $12 Initial lessons learned from this process are: million from 2004 to 2010. • To ensure that new connection funds The fact that the NWSC had identified an (once agreed by government) are set up opportunity of collecting an additional UGX4 and properly ring-fenced, and that clear billion (about $2 million) per annum from tariffs procedures (including third party counter alone meant that the size of the fund (in real approval for and monitoring and reporting terms) reduced from one year to the next due of their use) are agreed, in particular to the rising cost of materials. Thus the amount when the utility has competing needs for of funds available to pay for connections may available resources. have reduced from one year to the year precisely due to inflation—whereas the discounted cost • Once adequately set up and ring-fenced, of connection was kept at $35 (for domestic ensure that a mechanism exists for customers). protecting the value of the fund against inflationary increases in the actual cost Had an effective working New Connection of connection, as this will lead to the Fund been established, the number of new net reduction of the number of new 28 Data on progress are obtained from the Mid-Term Report on the Impact and Management of Piped Water Services in Uganda (GPOBA, 2010). 39 connections that can be made per annum. prevent the managers of the fund from Two cases appear: having to subsidize basic utility operations instead of planned expansion. Protecting the value of the surcharge collected via the tariff—by indexing the 4.1.5 Capital Investments in the Water Sector tariff on inflation. This the NWSC had of Kampala implemented as a means to protect its ability to recover its operating costs To be effective, any new connection fund needs from customers’ bills to be complemented by a capital investment Protecting the value of the funds program to construct new infrastructure/water against increases in the cost of mains. connection—which the NWSC had not implemented. This Section therefore looks at the different capital investment projects that were implemented • The NWSC could also have stimulated in the water sector in Kampala since 2004 to customer demand, in particular in poor provide some indication on whether these were settlements, by offering a discounted cost sufficiently mobilized (in focus and in scope) to of connection for a time limited period. support the Affordable Connections Policy and services to the poor in particular. • Ensure that the utility is already in a financially stable position before any new In urban areas the strategy of the government connection fund is operationalized—to is to increase the coverage of population with improved water sources and provide money for investment in increasing capacity and distribution Figure 5: Water sector investments in Kampala (2001 to 2008) while stressing the commercial viability of the utility services. However, in large towns, the government found that there was extra capacity 40 and the priority was to invest in distribution networks rather than extra capacity. As the 35 situation in the urban water sector has improved over the years, the priority of the government has 30 shifted to the rural areas and the annual funding in the urban water supply has started to decline for Total investments ($ m, 2009) 25 the past two years. The annual financial outlay of the government in the water and sanitation sector 20 and in grants to the NWSC since 2001–08 are shown in Figure 5. 15 The NWSC estimated (also in Board Paper 571 of 2004) the potential costs of connecting all 10 customers (domestic and nondomestic) within a 50-meter radius on either side of distribution 5 mains. However, these costs estimates were not matched with a detailed assessment of the level - of investments required to extend the network to 2001 2002 2003 2004 2005 2006 2007 2008 unserved areas. This assessment came later and was mostly undertaken through donor assisted Donor funded investments ($ m) GoU investments ($ m) projects. These are listed in Table 12. Source: NWSC 40 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Table 12: Schedule of water sector investment projects in Kampala (2002–10) Project name Period Funding source Funding amount (US$ million) Kampala Urban Pro-Poor Project 2002–2007 KFW and GoU 3.3 Kampala Peri-Urban project 2002–on-going NWSC 1.5 Kampala Network Rehabilitation Project 2003–on-going KFW 4.4 Mukono Water Supply Project 2003–2008 NWSC 11.1 Kampala Transmission Mains 2006–2009 KFW, NWSC and GoU 11.6 Mukono Water Supply Project (phase 2) 2008–2009 NWSC 7 Urban Pro-Poor Project (Kisenyi, Ndeeba) 2008 KFW and NWSC 2.7 Urban Pro-Poor Project (Kagugube) 2009 AfDB 1.2 Kampala Network Restructuring Project 2009–2010 NWSC and KFW N/A Gaba Offshore Intake Pipeline Project 2009–2010 AFD and NWSC 7 OBA Water Supply to the Urban Poor 2008–on-going GPOBA, NWSC 4 Kampala Water (Lake Victoria) Water and 2011–2016 NWSC, EIB, AFD, KFW, EoU 250 (212m euros) Sanitation Project and EU African Water Facility Source: Data from NWSC Corporate Planning division. The specific pro-poor investments above were i. The fact that poor households cannot geared at increasing coverage in the poor afford individual connections and thus settlements using PWPs/kiosks as principal share access to yard taps and/or PWPs. delivery mechanisms. According to Table 12, the investments in water mains expansion (to unserved ii. Difficulties with the high level on settlements) were mobilized in 2007—three years nonpayment in poor settlements, which after the start of the Affordable Connections Policy. resulted in 18 percent of yard taps and 36 percent of PWPs having to be 4.1.6 Summary of Findings on Assessing disconnected. This is considered to Increased Access significantly affect any effort by the NWSC to increase coverage in these settlements. The NWSC delivered 16,198 new connections in poor settlements of Kampala between 1998 and iii. The NWSC’s incentive to connect as 2011. However, the real impact from the Affordable many high consuming and able paying Connections Policy was 2,500 new yard taps customers, rather than the poor first. The and 660 new PWPs. Of the 466,542 who gained total number of connections in 2004 was access between 2004 and 2010, however, it is low, with many government, industrial, and considered that 21 percent, that is, 99,146 of domestic customers not yet connected. them, were in poor settlements. iv. The lack of adequate water supply The relatively low number of additional connections networks supplying low income in poor settlements is based on a number of settlements. This is coupled with the factors including: fact that actual investments in new 41 watermains, and funding to deliver the Policy, 40 percent in 2007 and 2008, additional yard taps and/or kiosks/PWPs 30 percent in 2010, and 10 percent in came only in 2007. The total number of 2011). NWSC’s current strategy is now to pro-poor connections may also have a replace all the PWPs with prepaid water reached a plateau given that additional supply PWPs. funding for network expansion stopped in 2009 and that additional water supply viii. Difficulties with connecting poor capacity was required (this capacity issue, households due to the complicated land however, is being addressed). tenure system. However, in order to be more effective, the NWSC and GoU v. The focus on expanding services to poor should have tackled these obstacles much (unserved) settlements is still considered earlier as these are common issues linked only on a project-by-project basis instead to the extension of services in informal of at a corporate level. A long-term settlements, where most people live. strategic business plan is good practice, and would enable the NWSC to plan for the whole of Kampala for both water 4.2 Impact on Financial Viability supply and sanitation needs. Such a plan would also enable coordinated and Over the period 2004 to 2010, the NWSC sectorwide development support to the managed to significantly increase the total NWSC, in particular for poor settlements. number of customers which, in turn, positively impacted revenue generation. For example, from vi. The NWSC’s decision to not set up 2004 to 2010, revenues rose from UGX54 to and use a properly ring-fenced New 113 billion (equivalent to US$27 million to $56 Connections Fund principally because million in 2010 exchange rates. This resulted in a the additional revenue collected through corresponding doubling of operational costs from the tariff surcharge was being used to UGX41 to 86 billion, or $20 million to $43 million subsidize basic operations (in particular in the same period. between 2004 and 2006 when the NWSC was recovering only 80 percent However, given that the NWSC did not set up a of its operational costs). In addition, the ring-fenced New Connection Fund it is difficult size of the fund seems to have excluded to comment on how much of the additional 11 any provision for inflation, meaning that percent tariff surcharge was (a) kept as revenue; the actual number of new connections (b) used to fund new connections; or (c) reused in was being eroded by increasing cost of operations. materials, and so on. This significant revenue increase is due to the vii. Difficulties with using PWPs as a increase in new customers, but also in large sustainable delivery mechanism for water part due to the fact that the water supply tariff supply services in poor settlements, was set to ensure an appropriate level of cost principally due to the fact that these recovery, and that it was kept in line with sector PWPs are privately registered and that price indices and inflation through yearly and NWSC experienced a significant level of automatic price increases. This has also meant nonpayment of bills. As a consequence that the NWSC’s operating cost coverage ratio a large proportion of PWPs were has remained more or less constant at around disconnected (80 percent in 2000 before 130 percent.29 This is shown in Figure 6. 29 This is excellent performance as the internationally accepted benchmark for OCCR is between 130 percent and 160 percent, meaning that the utility’s revenue covers between 130 percent and 160 percent of its operating cost, thus allowing the utility to adequately maintain and rehabilitate the infrastructure. 42 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Figure 6: NWSC financial data (2002–10) 160,000 120 140,000 100 Volume of water produced Total number of connections (i) Increased connections and revenue 120,000 and sold (Mm3) 80 100,000 Total water connections 80,000 60 Operating revenue (UGX billion) 60,000 40 Operating costs (UGX billion) 40,000 20 20,000 - 0 2001 /02 2002 /03 2003 /04 2004 /05 2005 /06 2006 /07 2007 /08 2008 /09 2009/10 160,000 120 140,000 100 (ii) Increased revenue and OCCR Volume of water produced Total number of connections 120,000 and sold (Mm3) 80 100,000 Total water connections 80,000 60 Operating revenue (UGX billion) 60,000 Operating costs (UGX billion) 40 40,000 20 20,000 - 0 2001 /02 2002 /03 2003 /04 2004 /05 2005 /06 2006 /07 2007 /08 2008 /09 2009/10 This means that the NWSC is more and more important to assess the range of individual able to focus on expanding and providing domestic consumptions and to identify what sustainable services to the poor. factors appear to be limiting demand. 4.3.1 Overall Reduction in the Unit 4.3 Individual Consumption Consumption of Water One of the tenets of the government’s policy is Figure 7 shows the total volume of water to equitably provide services to all households produced and sold, and the total number of (some for all rather than all for some). It is thus connections since 2002. 43 Figure 7: Volumes of water produced and sold vs. total number of connections in Kampala (2002 to 2010) 160,000 60 140,000 Volume of water produced and sold (Mm3) 50 120,000 Total number of connections 40 Total water connections 100,000 Water production (Mm3) 80,000 30 Water sales (Mm3) 60,000 20 40,000 10 20,000 - 0 2001 /02 2002 /03 2003 /04 2004 /05 2005 /06 2006 /07 2007 /08 2008 /09 2009/10 It appears from the above that water sales have The rate of network expansion and new increased at a much slower pace than new connections is constrained by the capacity/ connections between 2003/04 and 2009/10, availability of water. It seems that this situation will suggesting that whilst some additional water start improving post-2012 with the improvements was provided, the bulk of the water has been to the Gaba I and II water treatment capacity, as distributed amongst a greater number of well as the extension of the Gaba offshore intake customers. However, it is difficult to comment on pipeline. whether this has reduced illegal connections or commercial losses. Therefore, this means that 4.3.2 Unit Consumption by the average unit consumption per connection Customer Category has reduced since 2004. In fact, the average unit consumption (all connections included) has This Section considers unit consumption at reduced by half from 36.4 m3/month to 17.3 m3/ individual house connections (poor and nonpoor), month between 2002 and 2010. PWPs/kiosks and yard taps. This is illustrated (for the whole of the NWSC’s operations) in Table 13. The graph also shows that whilst there has been a significant increase in connections, which have The Sections below present and discuss continued to be provided at a discounted price NWSC data on unit consumption at yard taps by the NWSC (within 50 meters of the mains) and PWPs. It is considered that none of the and constructed by NWSC staff, NRW is still an households in the poor settlements of Kampala issue in Kampala, despite the two recent network has an individual house connection. rehabilitation projects in Kampala (in 1994–98 with $18.5 million funding from Austria, and 2003–ongoing with $4.4 million from KfW). NRW is discussed in more detail in Section 4.5. 44 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Table 13: Average unit consumption per customer category (NWSC, 2010) Customer category Number of % total Volume of % total Average unit connections connections water billed (m ) 3 billed consumption (litres/day) Public water points 7,748 3.1% 2,210,307 4.7% 781 Domestic (incl. yard taps) 207,350 83.7% 22,966,603 48.9% 303 Institutional/government 5,664 2.3% 9,781,786 20.8% 4,731 Industrial/commercial 27,088 10.9% 11,992,093 25.5% 1,213 Total 247,850 100.0% 46,950,789 100.0% 519 Note: This data is for NWSC as a whole. 4.3.2.1 Unit Consumption at Yard households using yard taps in Kampala is shown Tap Connections in Figure 8. The graph also shows the cumulative frequency distribution of monthly consumption for The average monthly water consumption of poor pro-poor household connections. Figure 8: Average monthly consumption of poor households using yard tap connections 100% 90% 60% 80% 50% 70% 40% 60% 50% 30% 40% 20% 30% 20% 10% 10% 0% 0% Less than 3.6 m 3 From 3.6 to 5 m 3 From 5 to 12 m 3 Greater than 12 m 3 0 10 20 30 40 Monthly water consumption for poor household using yard Monthly consumption (m3) taps (NWSC, 2009) Source: NWSC customer database (2009). 45 From the above, it appears that 57 percent limits water used for flushing and bathing. The of poor households that have a yard tap NWSC reports adequate network pressures connection use less than or equal to 3.6 m3 and 24x7 supply, which suggests that customer per month—this equates to 20 liters per capita demand is fully met. Poor households are per day (lcpd) for a family of six.30 If 20 lcpd therefore using the amount of water that they is the minimum recommended daily water need and can afford. This corresponds to consumption then this data show that poor approximately 120 liters per day for a household households are able to access the minimum of six. volume of water recommended by the WHO. However, this also shows that poor households 4.3.2.2 Unit Consumption at Public using yard taps are restricting their water Water Points demand to 20 lcpd. Approximately 20 percent of households use more than 10 m3 per month, Only 18 percent of the PWPs show suggesting that many of these households are consumptions of less than 3.6 m3 per month also reselling water. (20 lcpd for a family of six), because they are shared between a large number of households. Low consumption can be attributed to a number This explains why 60 percent of PWPs report of reasons, including the cost of water, network consumptions greater than 10 m3 per month pressure, interruptions, and the availability of (55 lcpd). Unit consumption is summarized in alternative, less expensive sources of water, as Figure 9. These figures show that whilst individual well as the lack of internal household plumbing household consumption might not be high, the for customers using yard taps which in turn low consumption at PWPs is due to the fact that these are shared. 4.3.2.3 Unit Consumption at Prepaid Public Figure 9: Unit consumption at public water points in Kampala (2009) Water Points/Kiosks The NWSC is gradually replacing all PWPs/ 70% kiosks operated by private individuals with 60% prepaid (therefore unmanned) kiosks as a means to ensure that consumption subsidies 50% that are embedded in the tariff structure actually 40% benefit the poor. Prepaid PWPs also improve 30% the NWSC’s bottom line by achieving virtual 100 20% percent revenue collection efficiency. Previously, as described in earlier Sections of the report, 10% private PWP/kiosk operators used to charge 0% significant mark-ups on the PWP/kiosk tariff. Less than 3.6m3 From 3.6 to 5m3 From 5 to 10m3 Greater than 10m3 The unit consumption at PWPs/kiosks is shown in Figure 10. Monthly consumption for poor households using public water points (NWSC, 2009) 30 In addition, 80 percent of households use less than 10 m3, and 90 percent less than 20 m3. 46 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Figure 10: Unit consumption at prepaid public water points/kiosks 60% 50% Monthly consumption for poor households using 40% prepaid PSP/kiosks (NWSC, 2009) 30% 20% 10% 0% Less than 3.6 m3 From 3.6 to 5 m3 From 5 to 10 m3 Greater than 10 m3 The profile of unit consumption at prepaid PWPs/ In addition, it is considered that low unit kiosks and yard taps is markedly different. consumption at PWPs is due to the presence of a large number of alternative sources of water Similar to yard tap connections, 57 percent of the which are free or cheaper than the NWSC’s piped prepaid PWPs report consumptions of less than drinking water. Thus, if individual household 3.6 m3 per month—which equates to 20 lcpd for consumption at PWPs is low—it is only reflective a family of six (120 liters per day per household). of the volume of water which households This is low given that the prepaid PWPs are purchase for drinking and cooking. Other uses, shared between many households, but may be including washing and ablution, are met with due to the fact that the prepaid water supply other sources of water, including but not limited PWPs are still at the pilot scale and thus not yet to wells, boreholes, springs, and public ablution fully integrated within the communities they serve. blocks (for example, public and private Sulabh toilets/showers). However, close to 30 percent of prepaid PWPs report consumptions greater than 10 m3 per Pictures 1 and 2 illustrate prepaid public water month (55 lcpd for a family of six), similar to points/kiosks and yard taps in Kisenyi, a low normal PWPs. income but formal settlement in Kampala. The technology used for yard taps and prepaid PWPs One factor that may impact on individual can be the same: the only difference is in the consumption at prepaid PWPs is the availability location of the yard tap/PWP (public thoroughfare of tokens to purchase water from these PWPs, or gated private land), or the number of people and the ease with which users are able to keep using it, and under what name the yard tap/PWP the tokens topped up with credit. Individual is registered with NWSC. consumption at prepaid PWPs may actually be higher once more such PWPs and more tokens are provided. 47 1 2 1 Picture 1: Prepaid public water point/ kiosk in use in Kisenyi, Kampala (2011) 2 Picture 2: Prepaid private yard tap in use in Kisenyi, © WSP Kampala (2011) 3 3 Picture 3: Unprotected shallow well in Kisenyi (Kampala, 2011) Note: In this case the unprotected shallow well was 4–5 meters in diameter and 3 meters deep, acting like a below- ground storage reservoir. The water level was © WSP 2.5 meters deep. 48 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements 4 Picture 4: 4 ‘Protected’ but unsafe spring in Kisenyi (Kampala, 2011) Note: In this case the spring is deemed unsafe due to the presence of houses and latrines directly on top of the spring catchment. © WSP 5 © WSP © WSP 5 Picture 5: Public and private Sulabh toilets in Kisenyi (Kampala, 2011) Note: Above left is a public Sulabh toilet, which comprises separate toilets and showers for men and women. Above right is a private Sulabh toilet which is shared by 14 families. The building comprises a toilet and a shower, shared between men and women. 49 Pictures 3, 4, and 5 illustrate some of the It is considered that alternative water supply alternative sources of water in Kisenyi (Kampala). sources in Kampala (most of which are not safe, see Section 2.5.1) are providing a significant The NSDS survey of 2008 states that the average disincentive for households to purchase treated water availability in urban areas is 70 liters per water from the NWSC (treated water that is used day for a household; this corresponds to 2.1 m3 for uses other than drinking, that is, washing per household per month, or less than 20 lcpd clothes, and so on), and therefore an equally for a family of six. important disincentive for the NWSC to provide anything other than PWPs to the poorest Some parishes or subcounties have become households. better connected to improved water sources, while others lag behind. Branchwise analysis of It is recommended, therefore, that clear policy variation within Kampala of estimated domestic direction be given regarding the future of all consumption by the poor of improved water available nonpiped drinking water sources in sources in informal settlements is shown below. Kampala. This needs to take into account the negative health impact of poor households The average monthly domestic consumption for using these sources, as well as a realistic the poor differed between branches and ranged estimate of the cost of either physically closing from 4.98 to 19.40 m3 per household. Higher unit or rehabilitating (including protecting the source consumption confirms the extent of water resale, catchment) all these sources. in particular in the City Centre, Kansanga and Nakulabye areas. The lowest unit consumption was encountered in the Nansana and Mukono 4.4 Distance and Time Spent areas located on the outskirts of the city, in areas where watermains coverage is the lowest. Distance to supply and collection time are two important socioeconomic factors in water supply 4.3.3 Summary of Findings on Assessing services as they generally affect the livelihood of Unit Consumption only one type of customers: women and girls. The NSDS 2008 confirms that 65 percent of The NWSC’s significant expansion program was individuals who carry water from the PWPs to not supported by an equally significant capital the house are women. Distance to supply and investment program to increase water supply the time spent waiting in the queue at the PWP capacity, despite some significant investments therefore prevent individuals, and women in being committed in 2011. This means that particular, from entering into an economic activity. average unit consumption has reduced from 36 Reducing the time taking to fetch water is, m3/month to 17 m3/month (including all customer therefore, intimately linked with reducing poverty. groups). 4.4.1 Distance to Water Sources Unit consumption at yard taps seems to be the same as that at PWPs. This would suggest that The NSDS 2008 survey provides useful, although the only difference between yard taps and PWPs limited, information on the average time and is proximity to dwelling (and thus convenience). distance to supply in urban areas in Uganda In both cases the majority of households (about (not Kampala specifically). The report states 60 percent) seemed to be using 20 lcpd, which is that the average distance to a water sources the minimum amount of water per capita per day was 0.9 km and 0.6 km during the dry and wet proposed by the WHO. seasons, respectively. This is assumed to exclude 50 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements all households with individual connections and/ Table 14: Distribution of households by distance to water or yard taps, that is, the poorest households. sources during the dry season (urban areas) In 2004 a similar assessment showed that the average distance was 1.1 km and 0.9 km, respectively. The slight increase during the dry Distance in km 2004 2008 season confirms that households are able to Less than 0.5 km 75.1 86.5 use alternative and seasonal sources at different 0.5 to 1km 14.0 8.3 times of the year, and that consumption at the NWSC’s points of supply (in particular at PWPs 1 km to 1.5 km 1.7 0.3 which individuals have to walk up to) is seasonal. 1.5 km to 3 km 7 4.2 The distribution of households by distance to Greater than 3 km 2.6 0.8 water source is shown in Table 14. Total 100% 100% The fact that 7 and 4.2 percent of individuals in urban areas walked up to 3 km to a water Table 15: Average collection time for drinking water in urban areas source in 2004 and 2008, respectively, confirms that there are some pockets with only limited, or nonexistent, water supply. Individuals in these 2004 2008 areas can only walk up to 3 km to obtain their Waiting time (in minutes) Dry Wet Dry Wet season season season season water. The situation has somewhat improved though but would illustrate that there is a need to At water source 38 25 22 10 increase the number of water mains extensions in urban areas in general. Time taken to and from water source 22 17 13 10 4.4.2 Collection Time for Water Total water collection time 60 42 35 20 The NSDS 2008 survey provides useful information on average collection times for water the pressure (and therefore the convenience) of in urban and rural areas. However, no data is water available at these water points may also available for Kampala only. Water collection time have increased. Whilst this is not an outcome is proportional to distance and to the number of of any of any of the pro-poor approaches individuals that are sharing a particular access implemented by the NWSC, it may be the result point (for example, a water point). Table 15 of improved operational efficiency in the system. estimates the total water collection time based on (a) waiting time at source; and (b) time taken to and from the water source. 4.5 Affordability Table 15 shows that there has been a significant This Section considers affordability of the reduction in the total water collection time discounted connection fee and the water tariff, between 2004 and 2008, in both dry and wet and discusses the impact of the consumption seasons: this has been halved. However, time subsidy on the poor. The World Bank reference spent at the water sources has reduced the document for assessing the targeting efficiency most: 60 percent in the wet season and 42 of connection and consumption subsidies was percent in the dry season. This suggests that published in 2005 (Komives et al.).31 It considers whilst the number of water points has increased, the following: 31 Komives, K., V. Foster, J. Halpern, and Q. Wodon. 2005. “Water, Electricity and the Poor: Who Benefits from Utility Subsidies?” Published in Directions in Development, by the World Bank. 51 i. Benefit incidence, which measures how reduced since 2008, as the actual unit cost of well the subsidy instrument benefits the connection (that is, the additional length of service poor versus nonpoor households. line) increased from year to year with inflation. ii. Beneficiary incidence, which assesses 4.5.2 Affordability of the Water Tariff whether poor households actually receive the subsidy, and the degree to which According to the NSDS (2008) survey, the average poor households (that could benefit) are monthly household expenditure on water in excluded. Kampala is UGX10,960. However, the report also states that the people were willing to pay only iii. Materiality, which measures the value of UGX5,610 per month. The NSDS 2008 did not the subsidy received by the poor as a present any findings relating to average monthly percentage of household income. income and/or expenditures, so it is not possible However, assessing these indicators requires to comment on affordability from the NSDS 2008 more information than is available in this case report. study, including the number of households However, according to the UNHS (2005–06) (poor and nonpoor) that live within 50 meters report, the average monthly household of the water mains and thus have access to expenditure in Kampala in 2005–06 was the subsidy, and the number of households UGX333,704 which, using a CPI inflation rate (poor and nonpoor) that, once they had access, index of 128 percent (from 2005 to 2009), actually connected to the water mains. corresponds to an average monthly income/ This Section assesses the affordability of the expenditure of UGX425,803 per household. This connection fee, expressed as a percent of in turn corresponds to an average household monthly income, and of the tariff, expressing the expenditure on water of 2.6 percent of monthly monthly expenditure on water as a percent of income/expenditure. This is considered to be an monthly income. acceptable level of expenditure. 4.5.1 Affordability of the Connection Fee The analysis of the prepaid data shows that at the lowest official rates of water that is, UGX867 The average monthly income of poor households per month, for achieving this minimum average of in Kampala is $48 (or UGX80,000), according to 20 liters/day a household has to pay UGX3,121 the NWSC and GPOBA.32 The NWSC charges ($1.56) per month. For an average poor family domestic customers UGX59,00033 which is 74 living below the minimum poverty level of $1.25/ percent of the average monthly income of poor day, this constitutes 3.9 percent of the monthly households. This is a significant proportion of expenditure. This is also considered to be an monthly household income and a significant acceptable level of expenditure, although the level barrier to access to water supply services—in of consumption is low (20 liters per capita per particular individual connections and yard taps. day). The NWSC, therefore, has to pay all the However, there are only a limited number of additional cost increases for pipes and materials prepaid meters. This means that poor households due to inflation. Whilst the NWSC has protected are continuing to purchase water from PWPs that itself against increases in operational costs, are managed by private individuals who charge the fixed discounted connection fee is proving significant mark-ups, particularly when water is costly to administer and may explain that the scarce. A survey of the slum areas of Kisenyi I and rate of new connections in poor settlements has II and Ndeeba, undertaken by Beller Consult in 32 This equates to daily income of $0.26 per capita per day assuming a household size of six. 33 This is equal to UGX50,000 plus 18 percent VAT. 52 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements 2006, revealed that the poor spend UGX4,000– However, the cross-subsidy in the case of the 20,000 per month on water. Thus those below NWSC is not significant as large commercial/ the poverty line spend up to 5–25 percent of their industrial and government/institutional consumers monthly incomes on water. This is confirmed by are only charged twice the PWP tariff.34 This the NSDS (2008) survey results: 41.3 percent of is further compounded by the fact that poor the urban households in Kampala reported high households use a minimum amount of water. costs to be a major constraint limiting access to safe water, probably because they are being NSDS (2008) shows that 26 percent of the urban charged mark-ups by middlemen. population uses PWPs or kiosks. These PWPs constitute 4.7 percent of the total volume of water This is acting as a further incentive for poor consumed and account for 2.2 percent of the total households to continue to use alternative sources revenue of the NWSC (Annual Report 2009/10). of water that are not safe. This is summarized in Table 16. The NWSC is planning to replace all privately Given the data below, it is considered that poor managed PWPs by prepaid meters—to ensure households are unlikely to make the most of the that poor households can benefit from the pro- intended public health benefits of the consumption poor consumption subsidy embedded in the tariff subsidy as they are only consuming a minimum structure. amount of water. This is explained in more detail in Box 4. 4.5.3 Impact of the Consumption Subsidy on Poor Households Box 4 shows that the poor, who are purchasing water from PWPs, capture a significantly smaller The differential tariff has had a positive result in subsidy than households that are connected two respects. The first is that the intended tariff individually, including normal house connections subsidy is passed on to the poor consumers, and yard taps. Furthermore, the subsidy that and the second is that additional revenue was accrues to poor households is limited by the small collected and reinvested into a self-funded new volume of water that they use (on a per capita connections program. basis). Table 16: Water market segmentation and individual consumption (2010) Customer category Number of % total Volume of % total Revenue % total connections connections water billed billed collected revenue (m3) (UGX ’000) Public standposts 7,748 3.1% 2,210,307 4.7% 2,682,881 2.2% Domestic 207,350 83.7% 22,966,603 48.9% 45,487,035 37.3% Institutional/government 5,664 2.3% 9,781,786 20.8% 31,706,781 26.0% Industrial/commercial 27,088 10.9% 11,992,093 25.5% 42,072,459 34.5% Total 247,850 100.0% 46,950,789 100.0% 121,949,156 100.0% 34 For comparison: the ONEA in Burkina Faso has a subsidy factor of five between industrial customers and public standpipes. But the ONEA is reported to have exclusive right to the use, and management, of water resources in a dry environment, whereas NWSC customers can obtain water from alternative sources, if this becomes more economical. 53 Box 4: Assessment of the quantity of subsidy that targets poor households Using NWSC data presented in Table 16, it is possible to estimate the total quantity (in UGX) which goes to poor households, making an assumption that the proportion of yard taps over domestic customers is similar in the NWSC’s service area as it is in Kampala. By calculating the average tariff per customer category (using total revenue/customer category divided by volume billed) it is possible to estimate the flow of funds that customer categories receive or contribute. Thus it is possible to estimate the level of subsidy that poor households receive—per annum. This is summarized below. Customer category Average tariff including Flow of funds Subsidy per service charge (from volume sold & connection average tariff) per annum Public water points 1,214 3,058,131,641 394,699 Domestic 1,981 14,166,016,880 68,319 Institutional/government 3,241 (6,299,741,581) (1,112,243) Industrial/commercial 3,508 (10,924,406,940) (403,293) Total 2,597 - - The table suggests that PWPs receive on average UGX395,000 per annum subsidy, or UGX2,631 per person per annum (assuming one PWP serves 150 people). Whilst it is not possible to estimate the average subsidy received for yard taps compared to nonpoor connections, the average subsidy for connections is UGX68,319 per annum or UGX10,295 per capita (assuming a weighted average of 6.6 people per connection), which is nearly four times as much. The consumption subsidy thus unequally favors nonpoor households instead of the poor (PWPs and yard taps). This is due to the fact that they only use a small amount of water and that they are not connected. Furthermore, it is considered that a flat rate of drainage and sanitation in poor settlements— tariff structure does not provide any signals to an important aspect which the GoU and NWSC consumers (all categories) to conserve (and not have completely sidelined. waste) water.   To ensure that more of the consumption subsidy 4.5.4 Summary of Findings on Affordability goes to poor households, the NWSC would need to ensure either that (a) the poor are able The simple assessment of affordability has to purchase a greater volume of water; or (b) the showed the following results: differential tariff for high consumers is increased, or both. Increasing the tariff charged to industrial i. Despite the significant discount offered and commercial customers may lead to their by the NWSC on its connection fee, the developing their water supply systems, and discounted connection fee (UGX59,000 thus no longer contributing to the total subsidy or $35) is still a major hurdle for the amount. However, government/institutional poorest households to access water customers may be forced to contribute more to supply services within their premises, as it the total amount of subsidy. represents 74 percent of average monthly incomes of poor households. Options The most likely option for helping the poor use for improving this situation may include more water may be to simply improve the quality (a) making the fee payable in monthly 54 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements installments; and (b) giving customers households to use more water—and a time-limited window of opportunity limiting the availability of alternative sources to connect at a cheaper cost for the of water, or a combination of the two. poorest. This might incentivize increased connections—provided households have access to a water network. 4.6 Social Impact Analysis ii. The differential and flat rate water tariff, This Chapter presents the results of FGDs that in particular the specific PWP tariff, is were undertaken with stakeholders and key affordable for poor households as it informants, including consumers, water vendors represents about 3–4 percent of their and NGOs. Interviewees were asked to provide monthly income/expenditure. However, some feedback on the positive and negative this consumption subsidy is only impacts of the NWSC’s pro-poor policies and affordable for poor households because strategies. they use, on average, 20 liters per day per capita. The group representing domestic customers acknowledged the increase in water availability in iii. The targeting of the consumption subsidy their settlements as well as the overall reduction on poor households, despite the limits on of the price of water charged at PWPs. Although available data, has been found to be very the increase in the number of connections in poor—precisely because poor households the poor settlements directly attributable to the use only 20 liters per capita per day. As pro-poor policies and strategies implemented such the greatest beneficiaries of the by the NWSC is only 4 percent of the total consumption subsidy are the nonpoor number of new connections (as demonstrated domestic customers that can also use a earlier), increasing the number of PWPs has flat rate tariff structure which encourages had a significant impact on poor households’ liberal usage of water. Poor households expenditure on water, and can be considered that use PWPs benefit from a subsidy of to have directly impacted the poor. In addition, about UGX2,600 per capita per annum the analysis of unit consumption at PWPs and (only $1.3), whereas domestic customers yard taps suggests that poor households use, on on average benefit from about UGX10,300 average, 20 liters per capita per day—whether per capita per annum ($5.15). However, they purchase water from PWPs or yard taps. given that the total number of yard taps This is a significant finding from this study and is (10,993 as per NWSC data) represents confirmed through the social impact analysis and only 5 percent of the total number of group discussions. Discussions also showed high connections, the difference is likely to be customer satisfaction with the prepaid PWPs, much greater. despite the initial difficulties with launching the pilots. iv. Options available to the NWSC to improve subsidy targeting include: The assessment also confirms a number of • Increasing the degree of cross subsidy already documented concepts that are still to poor households, either from nonpoor very relevant to urban pro-poor water sector domestic, industrial and governmental or a practitioners, including: combination of the above; or • Ensuring that poor households use • The social connection program generated more water. This may be achieved by a lot of demand from poor households, in improving the sanitary conditions in poor particular, households that had an illegal settlements—and therefore permitting connection. 55 • Customers fetching water, most of whom driven out of business. Further, all the pipes were women and young children, felt became the property of the NWSC and the use reassured that many of the PWPs and of standardized materials for connection reduced kiosks were managed by women. NRW from 37.6 percent (Kampala 44.7 percent) in 2003–04 to 33.5 percent (Kampala 40 percent) • One of the principal drawbacks of PWPs in 2008. NRW is shown below in Figure 11, is that they are usually not available after expressed in percent of water (produced) lost working hours. This confirms one of the and in m3 per km per day. key strengths of prepaid PWPs (which are unmanned and thus available 24hx7). Whilst the Affordable Connections Policy has provided a large number of well built • Water vendors and existing kiosk (and therefore leak-free) connections, water operators are likely to see their water losses are potentially not coming from leaking revenues reduce if water supply coverage connection pipes but from distribution is increased in the poor settlements and transmission mains. To this effect the within which they work. To prevent them Corporation is implementing water network from potentially vandalizing the new rehabilitation programs. Commercial losses are infrastructure, it is important that they also being incurred through PWPs that have (as well as other more formal community been disconnected due to nonpayment, and structures) be involved early on in the thus becoming inactive. However, overall, this process is considered to have a minimal effect on the These are presented by customer category/group NWSC’s operations as less than 5 percent of the in Table 17. volume of water sold in Kampala is supplied to the poor settlements. 4.6.1 Impact of Pro-poor policy on other utility Priorities Commercial losses have been reduced significantly through the installation of a modern One of the stated objectives of the Affordable customer billing and revenue collection system. Connections Policy was to reduce NRW, This is further likely to be improved if the NWSC including technical (leaks) and commercial is able to replicate prepaid PWPs in poor losses (metering, customer billing, and revenue settlements. In addition, it is understood that collection). This Section considers the reduction the GoU wants all institutional and government of technical losses only. customers to receive water on a prepaid basis. This is likely to be implemented in 2012. The Affordable Connections Policy has had an undeniably positive effect on the NWSC’s Improvements in NRW for Kampala (in particular, operations and on customers, but NRW is still the level of technical losses) are expected to a key issue for the Corporation, in particular in improve when the old network, especially in Kampala. the Kampala city centre, is refurbished. This is expected to happen when the project under As the number of connections and the volume way supported by the European Investment of water supplied increased and water became Bank, KfW, and the European Union (EU) is more freely available, the prices of resellers implemented from 2012 to 2016. came down until a large number of them were 56 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Table 17: Summary of FGD on impacts of NWSC pro-poor policy Views and opinions Positive Impacts Negative Impacts of stakeholders Domestic • New connections (domestic) have led to an increase • No assessment of sanitation conditions in newly connected consumers in water availability and a reduction in monthly households which are able to use and purchase a lot more expenditure. water (and hence generate proportionately volumes of • When additional kiosks were provided this, first, wastewater). reduced the unit price of water (supply/demand) and • Consumers living on the outskirts of the city, where there are also limited the historical seasonal increase in tariff to fewer watermains as the network had not yet been expanded UGX50–100 instead of UGX100–300 in more distant to reach these areas, where network pressure is low (for (and less well served) areas. example, hilly areas) or service is discontinuous, may not have • Significant improvements in customer satisfaction benefited from the pro-poor policy. from all newly connected poor households; some • In all the above, the unit tariff for water (for example, per 20-liter mentioned that the new water supply connections jerrican) is likely to have increased. provided some legitimacy to the household. • Some of the connected customers (who did not benefit from • The social connection policy generated a lot of the free connection policy) were critical of the tariff increase, demand from households that were either poor although general dissatisfaction was low as the monthly water and/or had an illegal connection to the mains. bills were still very affordable. • Increase in number of connections helped in saving • Some households which consumed a lot of water (and/or were women’s and girls’ walking, queuing, and waiting reselling water) complained about being charged a substantially time to collect water. higher rate by the NWSC (Domestic rate: UGX1,341) as • Customers were very satisfied with prepaid water compared with the PWP rate of UGX867, although they were kiosks/PWPs. also providing service. • The presence of a majority of female kiosk • Increasing the number of PWPs/kiosks and Presidential operators (43 percent of kiosk operators are Pledge taps, with two different tariff structures, leads to a lot of women) increases the acceptability of collection confusion. points to the majority of water carriers who are still • Some of the new PWPs/kiosks were not available outside of women and children. normal working hours (and this is a strong reason for promoting prepaid, unmanned, PWPs/kiosks). • Major issue with prepaid water kiosks/PWPs was the poor availability of tokens—both in total number and in selling/ purchasing points: some customers reported that the unavailability of tokens led to their being sold at higher prices (completely distorting the intention and impact of the new system). • Social connection policy works only where there are existing watermains or where extensions can be built. It is not a solution for illegal settlements. • Social connection policy (free connections, and so on) needs to be accompanied by a strong campaign to prevent households from using unsafe (unprotected) sources, in particular when 57 these sources are available (as is the case in Kampala). 58 Other consumers • No positive impacts mentioned, although approximately 10,000 • The level of cross-subsidy has increased (from industrial to (for example, new industrial connections were made from 2004 to 2009. domestic) and some industrial customers are considering industrial) developing their own water resources (and thus stop being NWSC customers). • This could significantly affect the sustainability of the whole urban water sector, in particular in areas where there is a significant proportion of such high-paying customers (Kampala, Entebbe. and Jinja), and is a major risk in the development and setting up social connection policies where new connections are funded directly from an increase in the water tariff (cross- subsidies across customer categories). Vendors and kiosk • The role of the operators was very crucial to the success of • Vendors and kiosk operators will be losing revenue, in particular in areas where customers are able to obtain free (discounted) of service delivery in Kampala’s informal settlements operators the PWP/kiosk scheme. These operators were generally poor people and were often appointed as operators as they had new domestic connections. Do pro-poor policies increase water coverage? An analysis donated the land on which the PWP was located. They would • This may lead to, or increase, vandalism. work for more than 10 to 12 hours per day or, if their houses were located next to the kiosk, almost round the clock. In most areas they were given stiff competition by the domestic yard tap owners in slum areas. Most of the domestic yard tap owners (more than the 50 percent assumed by the NWSC) were found to be selling water in the informal settlements. • Water resellers constituted an important part of the distribution system and increasing the availability, albeit at a price. In certain areas the vendors used to carry water from distances as far as 1 km and then sold it at UGX300 per jerry can. Community- • The NWSC had used community organizers for promoting based and prepaid systems, who were highly effective in locating the areas nongovernmental with water deficits and convincing communities in drafting the organizations applications for a PWP. • Most of the NGOs were found to be small in capacity and could take only small operations in the area. However, generally, they were appreciative of the change in the water situation, and of prepaid PWPs in particular. Vendors of prepaid • Prepaid token vendors are a key link in the system as they tokens recharge the prepaid tokens of the consumers by remaining mobile in the area. However, as there are only eight vendors and their place of business not fixed, the accessibility of these vendors was a key issue as many customers could not meet them in times of need. Figure 11: Nonrevenue water expressed in % of water produced and in m3/km/day (2004-2010) 50% 60 45% Volume of water produced and 50 Water losses (NRW) in % 40% 35% (i) NRW as % of water produced 40 sold (Mm3) 30% NRW (%) 25% 30 20% Water production (Mm3) 20 15% Water sales (Mm3) 10% 10 5% 0% 0 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 4,000 23 Water losses (NRW) in m3/km/day 3,500 20 3,000 18 (ii) NRW as m3/km/day Mains length (km) 15 2,500 13 Length of watermains (km) 2,000 10 NRW (m3/km/day) 1,500 8 1,000 5 500 3 0 0 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 59 4.7 Summary of Achievement of the provided on the achievement of each of the GoU’s Objectives components of the GoU’s and NWSC’s policy. Where relevant, recommendations have also been This Section makes an assessment of the degree made on how each of the policies and strategies to which the GoU sector policy has been achieved. could be met and/or improved. Each of the components of the GoU’s and NWSC’s pro-poor policies and strategies are commented The findings are presented in Table 18 in the same on, in the order in which they have been discussed order as they were presented in Section 3.2.2. (in Section 3.2.2 of the report). Comments are Table 18: Achieving the components of GoU and NWSC policy GoU policy NWSC policy measure Comments 1. Equitable allocation Affordable Connections Policy. Allocation of sector budget to poor households has improved of subsector budget. but is not considered equitable yet as the NWSC still has much lower per capita budget for poor households than nonpoor (and nondomestic). In addition, the amount of subsidy captured by the poor is much lower than normal domestic connections (including yard taps). This could be corrected if the NWSC developed a detailed business plan that covers all areas of Kampala, especially the poor settlements, and if a strategy is developed to help poor households get more water. 2. Improve overall Cut costs, improve revenues. Overall sector performance has been improved: the NWSC has sector performance. significantly increased its customer base, has an appropriate (indexed) tariff structure, and is collecting much more revenue. 3. Enhance access by Not part of the NWSC’s policy but Access in poor settlements has improved but significant mains increasing the density linked to the Affordable Connections extensions are still required, in particular in the poor areas that of network/points Policy. are on the outskirts of the city (where the poor also live). This and expanding to could be corrected if the NWSC developed a long term business unserved areas. plan for improving water supply (and sanitation) services in the whole of Kampala. This business plan would identify specific needs, schemes, and secure funding sources. 4. Target vulnerable No special policy of preference to See the point above. The pro-poor policy was initially targeted groups such as these groups. to serve the poor. Whilst it did benefit the poor, the majority women, and people of beneficiaries of the Affordable Connections Policy were, in with disabilities and fact, nonpoor domestic and other nondomestic categories (as HIV/AIDS. evidenced by the fact that 4 percent of the new connections are directly attributable to the connection policy). This was recognized by the NWSC and its partners and led to the implementation of the Po-Poor Targeting Project with support from the GPOBA. 5. Monitoring impact of Pro-poor service delivery The Urban Pro-Poor Branch was set up in 2007 to promote, W&S services on the mechanism: Setting up the Urban plan, and improve service provision to the poor settlements of poor. Pro-Poor Branch. Kampala. This report is, however, the first assessment of the impact of water supply policies and strategies on services to the poor. The Ministry of Water and Environment has instituted large scale and broad-based monitoring and evaluation processes called the GoU policy MoWE’s Sector Performance Reports. 60 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements GoU policy NWSC policy measure Comments 6. Encourage Pro-poor service delivery The Urban Pro-Poor Unit was set up in 2007 and has been participation of the mechanism: Setting up the Urban encouraging participation of poor men and women since then. poor and give them a Pro-Poor Branch. However, to increase the impact of the Affordable Connections voice. Policy on the poor, additional investments are needed both in water supply capacity and water network expansion. 7. Establishing public Affordable Connections Policy. Although a large number of PWPs were constructed. it is water points. considered that as much as 50 percent of them are inactive. These are disconnected because private operators do not pay the monthly bill to the NWSC. This is having a negative impact on poor households (who continue to use expensive PWPs or alternative sources of water) despite the NWSC’s recent testing of prepaid PWPs. The Corporation is gradually going retrofit all PWPs to prepaid as a way to prevent disconnection of a community due to nonpayment of an individual. 8. Pro-poor tariff Pro-Poor Tariff. The differentiated flat rate tariff is allowing poor households structure. to access water from PWPs and yard taps. The NWSC has set up a system that automatically considers yard taps with high consumption as PWPs. However, domestic customers are capturing a large amount of the subsidy that was in fact designed to help the poor, because of the limited volume of water sold to poor households (total of less than 5 percent per annum for about 50 percent of the population). 9. Subsidizing yard Pro-Poor Tariff. There has been a significant increase in the number of yard taps. connections serving However, this has been limited in time as the number of new as authorized public yard taps has reduced post-2007 (to pre-2004 levels). This is water points. considered to be due to the lack of water network coverage and water supply capacity. The unmet additional demand from yard taps is considered to be significant and can only be met with additional and equally significant capital investment programs. 10. Subsidy policy. Pro-Poor Tariff. The cross-subsidy could be improved, as it is currently low— with government, institutions, and commercial customers paying twice the unit rate charged at water points/kiosks. In addition, the poor are not benefiting as much as they could from the subsidized consumption subsidy because they are not using much water. The GoU and NWSC to address alternative sources. 11. Continuous Regular monitoring of water quality. Water quality at the NWSC’s taps is considered to be adequate. monitoring of water However, poor households are not able or willing to use more quality. than the minimum amount of 20 lcpd and thus use alternative sources that are unsafe. This is considered to be a significant public health risk as well as a disincentive for the NWSC to serve them until clear policy guidance is given on the future of all alternative sources. 61 4.8 Review of Results and Component 1: Affordable Connections Policy Lessons Learned Increasing the number of connections across This Section is divided into two subsections: a the board: The Affordable Connections Policy description of the elements of the policy which was an intelligent move to increase the domestic have been successful, and elements which could consumer connections base of the NWSC—it be improved. led to economies of scale and increased revenue and infrastructure investments in the long term. Each subsection is subdivided into the following The policy was implemented through a reduction Sections: general comments on overall utility in connection fees by more than 50 percent while performance; comments on the Affordable reducing reconnection fees by 75 percent. These Connections Policy; comments on the new connections were funded by a 10 percent differentiated and pro-poor tariff; comments on surcharge on the water tariff. It is considered pro-poor targeting; and finally, comments on the that 4 percent of the new connections that four delivery mechanisms (pro-poor unit, PWPs/ were made through the social connection kiosks, Presidential Pledge taps/PWPs, yard taps policy between 2004 and 2009 actually went and prepaid PWPs). to poor households, but that 21 percent of the beneficiaries were the poor. 4. 9.1 Elements of the Pro-Poor Policy Which Have Been Successful Component 2: Differentiated and pro-poor tariff/cross-subsidy in tariffs: Three types of The pro-poor policy was successful due to the cross-subsidies were utilized: differential tariff following factors: rates with lower rates for the poor; cross-subsidy across different areas; and subsidizing of new General: connections with the 10 percent surcharge in i. Financial sustainability as a key tariff. requirement for serving poor households: It is considered that the NWSC’s broad- Component 3: Pro-poor targeting based policy has allowed the utility to significantly improve its financial Geographical targeting: The recent project sustainability (tripling revenue between implemented by the NWSC, with support 2004 and 2010), and that this is a key from the GPOBA, is proving that geographical requirement for utilities before they can targeting of poor households offers significant realistically serve poor households. opportunities to increase coverage (volumes of water sold and thus financial sustainability) ii. NWSC’s improved operations: The in poor, unserved areas, and demonstrates NWSC has had to significantly increase that performance-based subsidies are an the efficiency of its operations (both efficient demand-based financing mechanism technical and financial) to plan for, to extending services to poor households in procure, and manage the increase in designated areas. customers. This has required employing more and better trained staff, setting up Delivery Mechanisms key performance contracts with staff and with internal independently managed i. Creation of a pro-poor branch in the water supply operations, applying for and NWSC: Creation of an organizational obtaining an effective tariff structure, as structure and institutional mechanism well as installing a new customer billing to initiate, facilitate, and continuously and revenue management system. 62 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements monitor pro-poor initiatives, update at business plan level (see point above). MIS and coordinate the activities of Such pro-poor targets could then be different interest groups (including the monitored and reported on at Sector NWSC’s own operational branches) and Performance Report level by the MoWE organizations was a very positive step. and (eventually) the new independent This allowed NWSC dedicated personnel water sector regulator (to be created). to concentrate on pro-poor activities, a focus that the organization lacked earlier. iii. Remove barrier to access caused by the existing land tenure system: The ii. Prepaid water system: A careful and well- Mailo/private land tenure system in which thought out introduction of prepaid PWPs owners cannot be forced to provide basic has led to its widespread acceptance by urban infrastructure services (including consumers, politicians, NGOs, and even water, sanitation, roads, and so on) is the press. This has largely eliminated the acting as a major barrier of access in middlemen in water supplies and made poor settlements. It is considered that water affordable at the official price of the a special law should be obtained from NWSC. However, these still need to be relevant public authorities, including the replicated at scale. KCC and the Ministry of Lands. Some countries, for example, Casablanca in 4.9.2 Recommendations for Improving the Morocco, have created a specific legal Pro-Poor Policy status for the “households with no status”, whereby a household without property General: title demanding service and committing to paying the monthly bill (and some of i. Development of a long-term business the connection costs) can be contracted plan: The NWSC and its partners serve by a service provider. The Municipality in the poor on a piecemeal and project- this case enters into a contract with the by-project basis. Instead, a long term service provider and the household and business plan is required which coherently commits to pay the service provider in identifies needs, specific schemes, and case the household defaults. In practice, funding sources for expansion of services after five years of implementation, the to the whole of Kampala. This would collection rates are above the service catalyze the NWSC, GoU, customers and provider’s average and there has hardly development partners. been any default. Such an approach could have made the poor greater beneficiaries ii. Setting pro-poor targets for of pro-poor policies. decentralized units and individual staff level: Although the NWSC has objectives iv. Determine the future of alternative to expand services in unserved areas, sources of water in poor settlements: the these are not clear targets, nor are availability of alternative sources of water these mentioned in the IDAMC and staff in poor settlements (up to 50 percent of contracts. It is considered that specifying drinking water demand) is considered to clear targets for expansion of services be a significant disincentive for NWSC in unserved areas would incentivize the to provide new house connections. In NWSC, KCC and the DWD to prepare addition these alternative sources of water detailed investment plans and strategies, are often unprotected and unsafe. Thus it and thereafter deliver these. These is considered that GoU and Kampala City detailed targets would need to be included Council adopt a clear policy for closing 63 and/or rehabilitating all these sources. vii. Cost of connection fees for the poor: A This needs to be complemented with a number of options can be implemented detailed analysis of remediation works and to better help poor households obtain management options. water connections. For example, payment of the connection fee can be made in Component 1: Affordable Connections Policy installments (instead of paid upfront) which are reimbursed via a separate v. Proper ring-fencing of the New system through the water bill. This may be Connections Fund: Revenues raised applicable to individual connections and/ through the 10 percent surcharge in the or yard taps. Alternatively, the NWSC may water tariff should be sequestered into a also propose a connection subsidy that properly ring-fenced New Connections is time-limited, to incentivize connections. Fund, as this would greatly improve the Even if the total number of applications effectiveness and efficiency of these became too high for the NWSC to operations, and would give NWSC manage, the subsidy could be made management a clear view on what rate available until the Corporation is able to of new connections can realistically be connect households that have already achieved. This is to be supported by expressed their interest. a detailed understanding of the water supply investment needs—to be included Component 2: Differentiated and pro-poor tariff in the business plan above. The New Connections Fund can only realistically viii. Improvement of the size of the cross- be set up once the utility is already subsidy: Poor households are indeed financially stable (to prevent funds from benefiting from a differential and flat tariff being used for basic operations) and structure. However, the level of subsidy needs to be protected against inflationary appears to be relatively low given that price increases as this will lead to the domestic, industrial and institutional net reduction of the number of new consumers are only charged twice the connections that can be made per annum. PWP rate. In addition, it appears that large industrial customers who use more than vi. Legal constraints faced in getting new 1,500 m3/month are not included in the connections: For a new connection cross-subsidy. the NWSC insists on production of a land title document, a lease document ix. Improvement of the effectiveness of or permission of the landowner or the cross-subsidy: Poor households are lease holder. In the present land tenure not purchasing enough water from the system many people don’t have such NWSC and are therefore not benefiting as land documents and getting a water much as they should be from the cross- connection is not possible for them. The subsidy. This is due to a combination of liberalized NWSC connection policy has water supply capacity restrictions and the started to take advantage of kibanjas (land availability of alternative sources of water. without titles) for connection, yet it does It is considered that water supply tariffs not meet the requirements of the poor. are affordable to poor households except Documents such as voters lists, municipal when water reselling takes place. identification, ration cards, and so on, have been tried in other countries. Such an approach would lead to faster results. 64 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements x. Improvements to the tariff structure: Delivery Mechanisms Whilst the benefits of the differential flat tariff structure are recognized, it is considered i. Improve marketing of, and customer that the flat tariff for domestic customers is training on how to use, prepaid tokens: not conducive to water saving. Therefore, Prepaid meter are proving to be very it is recommended that an increasing successful in Kampala (both with the block tariff be set up at least for domestic utility and its customers). A mixture of customers. In addition, it is considered conventional and prepaid connections that the automatic tariff increase should be seems to be the way out for many slum waived for public standposts. Any revenue areas of Uganda. shortfall could thus be funded via the increasing block tariff. ii. Increase the market for retail of credits for prepaid meter tokens: In parallel with xi. Unequal water tariff in slums: The improving marketing of, and training on differential tariff rates for water for PWPs/ how to use, prepaid tokens, the market for kiosks (867/m³) and Presidential Pledge retail of credits for prepaid meter tokens taps (784/m³) are not equitable for poor needs to be increased. households. It is considered that the current PWP/Presidential Pledge tariffs be normalized. © UN Photo/B Wolff 65 5 Conclusions 66 Since 2004, the NWSC has implemented a commensurate increase in water supply capacity large-scale new connections program targeting and availability (mains). all customer categories in Kampala, including the poor. To do that the NWSC has also had to Unit consumption over the period has reduced improve the efficiency of its operations (technical which suggests that, while there was some and financial), invest in network rehabilitation increase in water production and water sold, and expansion projects, create a bespoke Urban the bulk of the water was distributed among Pro-Poor unit, and set up performance-based more customers. The assessment found that contracts with the government and with its own approximately 60 percent of the poor are using staff. However, both the NWSC and GoU have less than 20 lcpd, and that they are purchasing not ensured that a ring-fenced New Connection this water from PWPs. The NWSC is currently Fund was set up. It is considered that the pro- developing additional water supply, transmission, poor policy may have had a much bigger impact and distribution capacity. on the poor had the Fund been set up and used. Alternative sources of water are widely available Overall, the Affordable Connections Policy and in Kampala ad are used to supplement NWSC the differentiated and pro-poor tariff structure supply. Coupled with a complicated system have had a positive impact on poor households of land tenure, this acts as a disincentive for in Kampala. While 21 percent of the beneficiaries the NWSC to provide anything other than lived in poor settlements, this policy was shared connections in poor settlements as eventually used to broaden coverage across unit consumption is low. There is a need for the board. There are opportunities to continue clear policy, customer education campaigns, to learn from the lessons in implementing the and concrete actions to remediate all springs, Affordable Connection Policy, and to improve boreholes, and wells in poor settlements, and for the planning and implementation of service realistic management arrangements. expansion in unserved areas. Water supply coverage has increased significantly over the period, from 62 percent to 74 percent, despite a population growth rate of 10 percent. The total number of connections in Kampala has increased from 59,000 in 2004 to 146,000 in 2010. Whilst most of the beneficiaries have been the nonpoor (and industries), this increase has had positive impacts on the poor by leading to a reduction in the price of resold water. However, the rate of new connections in poor settlements has sharply reduced from 2009 onwards. This is considered to be due to the fact that 18 percent of yard taps and 36 percent of PWPs are inactive. This has prompted the Urban Pro-Poor Branch, with support from development partners (principally the KfW and GPOBA) to look for innovative service delivery mechanisms, including prepaid PWPs. This new approach is expected to generate significant benefits to unserved areas © WSP and the NWSC, providing that it is coupled with a 67 The differentiated tariff structure (including the stagnant 40 percent of water produced lost, from PWP tariff) is considered to be affordable even 2004 to 200. This is in spite of NWSC taking to the poorest households. However, poor ownership of service lines and paying for and households are using low volumes of water for undertaking all connection works. It is considered drinking and are therefore not benefiting from the that major network rehabilitation works, coupled cross-subsidy as much as they could. with investments to increase total supply capacity, are required. The level of the discounted cost of connection is considered to be unaffordable for the poorest The report demonstrates that large scale households as it represents up to 75 percent connection programs in poor settlements can of the average monthly income. Options to be funded directly through a tariff surcharge improve the level of uptake of connection, once once a sufficient number of large consumers adequate supply capacity is provided to unserved are able to contribute to the cross-subsidy settlements, include paying for the connection fee mechanism. However, utilities need to be able (of $20 for the poor) in installments. to invest in additional water supply capacity and distribution networks, and also need to develop Finally, this assessment confirms that utilities specific programs that specifically target poor which want to expand water supply services to households. Ultimately, these strategies and unserved areas first need to focus on improving programs need to be detailed in a strategic the efficiency of their operations. NRW continues business plan which financiers, public and to be a key issue for he urban water sector private, can help implement. in Uganda, and Kampala in particular, with a 68 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Recommendations 6 for Further Study 69 This Section of the report presents options iii. Models for token distribution and for undertaking more detailed research on circulation for automated prepaid PWPs; specific aspects of the planning, funding, and as described, one of the challenges implementation of the pro-poor policy: observed with the automated prepaid systems were adequate distribution and i. Options to overcome land barriers to circulation of tokens. It will be important to water connections; the complicated land develop a model to understand how these tenure system locks out investment for challenges can be overcome to inform poor urban households. Policy options on scale up of prepaid PWPs. how these land barriers can be overcome need to be further explored. iv. Detailed review of new connection applications versus new connections ii. More detailed analysis of actual water undertaken since Affordable Connections use by source and consumption patterns Policy was implemented—to be able to to inform targeting. It is important to determine whether specific areas of the understand the impact of the alternative network/customers were not connected, sources of water on customer demand or were deliberately connected at a and on the NWSC’s incentive to serve later stage (for example, once the major poor areas. More detailed assessment of customers were connected). customer demand in poor settlements, v. The analysis of benefit incidence, with guidance on how unit water beneficiary incidence, and materiality of consumption in poor settlements may be consumption subsidies. improved to ensure the expected health benefits. 70 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Appendices 1. Pro-poor Services Areas in Kampala 2. NWSC Pro-Poor Strategy (2004) 71 Appendix 1: Pro-poor Services Areas in Kampala Kawempe Nakawa Central Rubaga Makindye N 2 0 2 Km KCCA slums GIS unit, 2010 72 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Appendix 2: NWSC Pro-Poor Strategy (2004) National Water and Sewerage Corporation Information paper dated 19th July 2004 The new policy of providing and maintaining service lines to customers 1.0 Introduction • Reduce the level of unaccounted for water by reducing the incidence of leaks and This paper provides clarification on the new bursts within the service lines. policy of providing and maintaining service lines to all NWSC customers. The policy was • Ensure standardization in the quality of approved by the Hon Minister of Water, Lands materials. and Environment in a statutory instrument to be published shortly and takes effect from 1st July This policy is in line with the Millennium 2004. Development Goal (MDG) of halving the number of persons without access to clean water by the As you are aware, until now, the policy of year 2015; and improving efficiency in service connecting a new customer is that the customer delivery. The proposal is also in line with the pays a connection fee and is also responsible for principle of “some for all and not all for some”. supplying all materials such as pipes and fittings necessary for connection between the supply line and the meter. 2.0 The basis for the new policy Under the new policy, the customer shall only The policy of providing and maintaining the be required to pay the normal connection fee, service lines to customers was derived from the while the NWSC shall ensure that the customer following policy documents: is connected by purchasing all the pipes, fittings and carrying out trenching and pipe laying up to • The NWSC Corporate Plan: 2003–2006 the customer’s meter. Strategic Goals: in which it is stated that “the NWSC is to create funds for The principal objectives of the new policy are to: maintenance, new connections and network extensions as part of its strategy • Increase the number of customer to ensure that investments and new connections thereby making water connections are implemented on a timely accessible to the population especially the basis” urban poor. 73 • The Performance Contract with the GoU: connections by which the poor will get 2003–2006: which states that “The NWSC water at the true NWSC rates. The policy will undertake the introduction of a “new therefore aims at increasing the number of connection fund” mechanisms in the tariff yard taps and domestic connections. policy that will help accelerate provision of services to the poor and also curb the • The industrial sector which is one of incidence of unaccounted for water” the key drivers in ensuring continued industrialization, urbanization, and • The Poverty Reduction Support Credit growth of the country’s economy. The (PRSC IV) targets which provides that implementation of the policy is expected within the period April 2004 to March to facilitate access of water services to the 2005, the NWSC will develop a water industrial sector. tariff structure aimed at encouraging new connections through the establishment ii) Need to curb the high levels of of a connection fund. This would also unaccounted for water (UfW) of which a address the access of safe water to the significant proportion has been attributed poor. to poor materials and workmanship by customers when carrying out new connections. 3.0 Rationale for providing and maintenance of service lines to iii) Need to standardize materials for customers connection. Water is both an economic and social good. iv) Need to adequately maintain service Hence a balance needs to be established pipes up to the meter and thus continually between access and costs of service delivery. improve the system network. The following are the main reasons behind the policy for providing and maintaining customer v) The policy for providing and maintaining service lines. service lines to customers is widely accepted as best practice in service i) The need to increase access of services delivery and has been adopted by a to all consumer categories including the number of countries including South following: Africa, Ivory Coast, Senegal and Tanzania. • The Urban poor. A study carried out by Maxwell and Stamp in 200235 established 4.0 The cost of providing and that the poor did not actually benefit maintenance of service lines to from the subsidized rate at the stand customers posts due to the middle man effect. As a consequence it was recommended that a The NWSC currently makes an average of about strategy be developed to reduce reliance 14,000 new water connections per annum on water points and encourage domestic according to the following breakdown: 35 Utility Reform Unit Ministry of Finance Planning and Economic Development. 2002. Poverty Impact Assessment of Private Sector Participation in the Water Sector, Kampala, Uganda. 74 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Table showing estimated number of new connections per annum Connection category Number of new New connections connections as % total Public water points 280 2% Domestic 11,900 85% Government/institutions 420 3% Commercial/industrial 1400 10% Total 14,000 100% 4.1 Cost implications: Based on cost scenarios generated and presented for different lengths of service pipes, the Board accepted the average distance of 50 meters from NWSC’s service point as ideal in terms of cost implications. Cost per connection: The cost break down for a typical domestic water connection based on the distance of 50 meters is given in the table below: Estimated cost per connection (typical domestic ½” connection) Item Measurement Unit cost (shs) Total cost (shs) Pipe size 15 mm and less (DN 15,PN 10) 50 meters 600 shs/ meter 30,000 Fittings, tees, bends etc 30,000 30,000 Saddle 6 inch 150,000 150,000 Labour 50 meter 660 shs/meter 33,000 Total 243,000 The average cost per connection of a ½ inch pipe supply is therefore estimated at Shs 243,000. This cost does not include the cost of a meter which has always been an NWSC cost. Based on the model for the domestic cost estimate, the cost for the other customer categories were estimated as follows: Water points/standpipes: Shs 400,000 Government Institutions: Shs 350,000 Commercial Industrial: Shs 450,000 The following table indicates the estimated annual cost of providing service lines for 14,000 new connections based on the mean distance of 50 meters from the NWSC service point. 75 Table Showing Annual Cost of Service Lines Within A Distance Of 50 Meters From NWSC Service Point. Connection category Number new Average cost of Total cost connection new connection Shs (000) Shs (000) PWP 280 400 112,000 Domestic 11,900 243 2,891,700 Government Institutions 420 350 147,000 Commercial Industrial 1400 450 630,000 Total 14,000 3,780,700 From the table above, the total costs to cover the provision and maintenance of service lines to customers within a distance of 50 m would be about Shs. 3.78 billion. 5.0 Implications on the tariff In order to raise the Shs. 3.78 billion for the provision and maintenance of service lines, the tariff has been increased by an average of about 10% as shown in the table below. It should be noted that in line with the policy of re-balancing of the tariff, large industrial customers (consuming over 500 cu.m of water per month) have been excluded from the tariff increase. Table Showing Implications on Tariff Customer category Water tariff Water tariff Water tariff % Increase 2003/04 2004/05 after 2004/05 5% indexation after adjustment Standpipes 449 471 521 10.7% Domestic 693 728 806 10.7% Government Institutions 854 897 993 10.7% Commercial/ Industrial Up to 500 m3/month 1187 1,246 1,379 10.7% 500 – 1,500 m3/month 1,421 1,421 1,421 0% Above 1,500 cu.m/month 1,324 1,324 1,324 0% Average tariff 836 873 957 10% It should be noted that the change in tariff is solely due to the transfer of responsibility of providing and servicing consumer lines from the user to the Corporation. For purposes of clarification, the table below shows the impact of the change on the price of a 20 liter jerrican of water. 76 Do pro-poor policies increase water coverage? An analysis of service delivery in Kampala’s informal settlements Change in Price per Jerrican Change in Water bills/Jerrican/Household: Coverage 50 meters Band Current cost Cost/jerrican Cost/jerrican Typical bill 30 Typical bill 30 Typical bill 30 per jerrican Cu.m with after fund (D)/200 (C) cu.m (D)/200 (C) cu.m (D)/200 (C) cu.m 2003/04 tariff indexation Shs Shs per month per month after per month after Current tariff Shs indexation Shs adjustment Stand pipe 8.98 9.42 10.42 13,470 14,130 15,630 Domestic 13.86 14.6 16.12 20,790 21,840 24,180 Comm< 500m3 23.74 25 27.6 237,400 249,200 275,800 6.0 Implementation The Board in its two sittings on the 18th and 24th June 2004 approved the implementation of the new policy based on the following tenets: • The policy covers only water connections. • The policy will cover all customers within a distance of 50 meters from the NWSC service point. All materials and other costs including road reinstatement to be met by the Corporation. • The cost in excess of 50 meters to be met by applicant. • Customers will continue to pay the connection fee (e.g. shs 50,000 for domestic customers) as a commitment and contribution to the service. • The tariff is to be adjusted by an average of 10% to defray the additional costs of providing and maintenance of service lines up to the customers meter. • As in the case of the electricity industry, all service lines will henceforth belong to NWSC. This will allow more consumers to be connected to the service lines. Following the Board approval, the Hon Minister of Water, Lands and Environment assented to the policy through a statutory instrument (copy attached) which shall shortly be published in the Uganda Gazette. As you will note the effective date of implementation is 1st July 2004. 77 Cover photo © WSP