69032 Rev Water Papers No 69032 | April 2012 A CASE STUDY of PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM Researched and written by Thelma Triche Managed and supervised by Mukami Kariuki and Midori Makino AFTUW Africa Region The World Bank Water Papers are published by the Water Unit, Transport, Water and ICT Department, Sustainable Development Vice Presidency. Water Papers are available on-line at www.worldbank.org/water. Comments should be e-mailed to the authors. Approving Manager Alexander Bakalian Contact Information The paper is available online at http:www.worldbank.org/water. The author may also be contacted through the Water Help Desk at whelpdesk@worldbank.org. Disclaimer – World Bank © 2012 International Bank for Reconstruction and Development/International Development Association or The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff 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 boundar- ies, 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 endorse- ment or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemina- tion of its knowledge, this work may be reproduced, in whole or in part, for noncommercial pur- poses 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. About the Authors This case study was sponsored by the Water Partnership Program (WPP) and the World Bank Africa Region. Mukami Kariuki, Water Supply and Sanitation Sector Leader, and Midori Makino, Senior Financial Analyst, supervised, managed and otherwise contributed to the preparation of the case study which was researched and written by Thelma Triche, Consultant and Author, who can be reached at tatriche@gmail.com. Acknowledgements The study would not have been possible with- Their sincerity and openness was greatly ap- out the participation and support of Mr. Ar- preciated. chard Mutalemwa of DAWASA, Mr. Alex Kaaya of DAWASCO, and their staffs who assisted The author also benefited from the inputs the author and responded to numerous re- of many others who are too numerous to list. quests for detailed information. Mr. Haruna Peer reviewers were Vickram Cuttaree, Se- Masebu, Director General, and Mr. Felix M. nior Infrastructure Economist, the World Ngamlagosi, Director of Regulatory Economics Bank; Jan Janssens, consultant; and Philippe at EWURA clarified the regulatory framework Marin, Senior Water Supply and Sanitation and decisions. Specialist, The World Bank. The author sin- cerely hopes to have presented an accurate Mr. Ato Brown, Mr. Alain Locussol, Mr. Ash- and balanced picture and remains respon- ley Roe, Mr. John Sitton and Mr. Cliff Stone sible for the content and any errors contained provided historical insights and perspectives. in this report. Editorial Note When discussing legal and other official docu- example, lease contract is capitalized when it ments and terms contained in them, capitaliza- refers to either the Lease Contract with CWS or tion poses a challenge and might appear to the the Lease Contract with DAWASCO. Likewise, reader to be arbitrary or inconsistent. The au- operator tariff is capitalized when it refers to thor has adopted the following approach: The the Operator Tariff specified in the legal docu- titles of specific legal or official documents, and ments. When the terms are used in a generic terms that are capitalized in legal or official sense they are not capitalized. documents, are capitalized in this report. For List of Acronyms and Abbreviations AFD French Development Agency AfDB African Development Bank BGT Biwater Gauff Tanzania CEO Chief Executive Officer CWS City Water Services DAWASA Dar es Salaam Water and Sewerage Authority DAWASCO Dar es Salaam Water and Sewerage Corporation DCW Delegated capital works DfID U.K. Department for International Development DWSSP Dar es Salaam Water Supply and Sanitation Project EDAMS Electronic Data and Asset Management System EIB The European Investment Bank EWURA The Energy and Water Utility Regulatory Authority of Tanzania FTNDWSC First Time New Domestic Water Supply Connection GOT Government of the United Republic of Tanzania ICSID International Center for the Settlement of Investment Disputes IDA International Development Association of the World Bank IMF International Monetary Fund MOF Ministry of Finance MOW Ministry responsible for water, currently the Ministry of Water and Irrigation NUWA National Urban Water Authority NRW Non-revenue water NWP National Water Policy O&M Operation and maintenance POG Procurement of Goods Contract PPP Public-Private Partnership PSRC Parastatal Sector Reform Commission PWP Priority Works Program SIPE Supply and Installation of Plant and Equipment Contract STM Super Doll Trailer Manufacture Company Limited TANESCO Tanzania Electricity Supply Company Ltd. TICTS Tanzania International Container Terminal Services viii PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM Tshs Tanzanian shillings TTCL Tanzania Telecommunications Company Ltd. UFW Unaccounted-for water UNCITRAL United Nations Commission on International Trade Law WHO World Health Organization Table of Contents Executive Summary.................................................................................................................xiii 1  Introduction............................................................................................................................... 1 1.1  Geography and Brief History................................................................................................. 1 1.2  The Government of Tanzania’s Privatization Program......................................................... 1 1.3  Methodology..........................................................................................................................2 Policy and Institutional Reform of the Urban Water Supply and 2  Sewerage Sector........................................................................................................................5 2.1  The Urban Water Supply and Sewerage Sector....................................................................5 2.2  Institutional and Regulatory Reforms..................................................................................6 2.3  Support of the Reform by International Development Agencies.........................................8 3  The Pursuit of a Private Operator for Dar es Salaam................................................... 11 3.1  First Bidding Process, 1997–1998........................................................................................11 3.2  The Adoption of the Lease Contract Model.........................................................................11 3.3  Second Bidding Process, 1998–2000................................................................................. 12 3.4  Third Bidding Process and Selection of the Operator, 2001–2003................................... 13 4  The Lease Contract Framework.........................................................................................15 4.1  Key Terms of the Lease Contract......................................................................................... 15 4.2  The Restructuring of DAWASA and its Development Contract......................................... 19 4.3  Financial Projections........................................................................................................... 19 4.4  Arrangements for Serving the Poor................................................................................... 20 x PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM 5  CWS’s Performance.............................................................................................................. 23 5.1  CWS’ Shareholders...............................................................................................................23 5.2  CWS’ Management..............................................................................................................24 5.3  Commercial Operations.......................................................................................................24 5.4  Compliance with the Terms of the Lease Contract.............................................................25 5.5  Implementation of the Priority Works Program and Delegated Capital Works................26 5.6  Impacts on Staff and Productivity.......................................................................................27 5.7  Termination of the Lease Contract......................................................................................27 5.8  The Role of the World Bank and International Partners During the Crisis.......................29 5.9  Results of International Arbitration...................................................................................29 United Nations Commission on International Trade Law.................................................29 International Center for the Settlement of Investment Disputes...................................... 30 6  The DAWASA-DAWASCO Arrangement..........................................................................31 6.1  DAWASCO........................................................................................................................... 31 Governance.......................................................................................................................... 31 Terms of DAWASCO’s Lease Contract with DAWASA....................................................... 31 Management and Staff.........................................................................................................32 Operational and Commercial Performance........................................................................33 Comparison of the Performance of CWS and DAWASCO..................................................35 6.2  DAWASA..............................................................................................................................35 DAWASA’s Development Contract and Memorandum of Understanding........................35 Oversight of DAWASCO......................................................................................................35 Implementation of Capital Investments funded under DWSSP.......................................36 6.3  Financial Performance and Viability of the Arrangement.................................................37 7  Regulation............................................................................................................................... 39 7.1 The Interim Regulator......................................................................................................39 7.2 EWURA’s Regulatory Authority.......................................................................................39 7.3 EWURA’s Tariff Decisions................................................................................................39 Appendix A – Detailed Tables on Tariffs, Access to Service and Affordability 61 xi 8  Coverage and Access of the Poor to Service....................................................................41 8.1 Results of the NBS Impact Surveys: Coverage and Availability....................................... 41 8.2 Kiosks................................................................................................................................42 8.3 First Time Connection Fund.............................................................................................43 8.4 Community Water Supply and Sanitation Program........................................................44 8.5 Affordability of Service.....................................................................................................44 9  Conclusions and Lessons.................................................................................................... 47 9.1 The Policy Environment...................................................................................................47 9.2 Consultation with Customers...........................................................................................48 9.3 Measures for Serving the Poor..........................................................................................48 9.4 Strategies for Dealing with Vested Interests....................................................................49 9.5 Selection Process and Consultation with Bidders............................................................50 9.6 Appropriateness of the Lease Contract Model................................................................. 51 9.7 Financial Viability and Allocation of Risks.......................................................................52 9.8 Tariff Setting Framework..................................................................................................53 9.9 System for Monitoring Operator’s Performance..............................................................54 9.10 Conflict Resolution Mechanisms......................................................................................55 9.11 The Role of the World Bank..............................................................................................56 Appendix A – Detailed Tables on Tariffs, Access to Service and Affordability................................. 61 Appendix B – Summary of the Key Provision of the Lease Contract with City Water Services.......67 Executive Summary Overview to evaluate how customers, especially the poor, were affected. In 2003, following a long and costly selection process, City Water Services (CWS), a private operator was engaged under a lease contract ES.1 Background with the Dar es Salaam Water and Sewerage Authority (DAWASA) to provide water sup- Located on the eastern coast of Africa, the Unit- ply and sewerage services in Tanzania’s larg- ed Republic of Tanzania, including the former est city. CWS’ performance was disappointing Tanganyika on the mainland, and the semi-au- and it encountered serious financial difficulties tonomous islands of Pemba and Zanzibar, has a early. Within two years, the contract collapsed population of 40 million. It is one of the world’s with the dramatic expulsion of its expatriate poorest countries. During the 1990’s the gov- managers from the country. Two international ernment embarked on a liberalization program arbitration tribunals ensued. that included the restructuring of state-owned enterprises, private participation in infrastruc- Following the departure of the private opera- ture services and the divestiture of about 350 tor in June 2005, a public corporation, the Dar parastatal entities. es Salaam Water and Sewerage Company (DA- WASCO) took over operation of the services In 1997, the Dar es Salaam Water and Sew- and encountered many of the same constraints erage Authority (DAWASA) was created to de- that CWS had. Despite new financial injec- velop and operate the water supply and sewer- tions and an increase in the Operator’s Tariff in age services in the metropolitan area of Dar es 2006/07, its financial performance was unsat- Salaam, Kibaha and Bagamoyo. Funding for isfactory and, during the first two years, opera- maintenance and repairs had been inadequate tional results were poor. After five years, oper- for some years and the infrastructure that ational performance had improved but further DAWASA took over was in need of rehabilita- progress was still needed. tion. In 2002, because of leaks, non-metered connections and illegal usage, most of the The case raises compelling questions about water produced was lost. Service was avail- the preparation of the public-private part- able for only a few hours and a few days per nership (PPP), the selection process, the al- week in most areas. The water that did reach location of risks in the contract, expectations consumers often was not safe for drinking. A regarding financial viability and service im- large percentage of households bought water provements, the effectiveness of the public- from neighbors, tankers or other vendors, for public partnership that has existed since the which they paid rates that were considerably private operator departed, and how to struc- higher than the DAWASA tariff. Low salaries ture institutional relationships to ensure ac- made it difficult for DAWASA to hire and re- countability. It also provides an opportunity tain qualified staff. Less than 10 percent of the urban population was connected to a sewerage xiv PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM system. Another 20 percent, mostly in the The AfDB provided US$ 48 million; EIB pro- middle and upper-income brackets, used sep- vided US$ 34 million and the IDA provided tic tanks and the remainder depended on pit US$ 61.5 million. latrines. Most wastewater treatment facilities were not functioning and pit emptying servic- es were inadequate. The city suffered periodic ES.3 The Pursuit of a Private Operator outbreaks of cholera and other water-borne diseases. While embarking on a program of reforms in the policy and institutional environment for By the late 1990’s, driven by chronic water water supply and sanitation, the Government shortages in Dar es Salaam and other parts launched a parallel effort to engage a private of the country, the poor and declining quality operator for water and sewerage services in of services, and the mounting backlog of un- Dar es Salaam. In 1997 the British company Bi- served households, pressure to improve water water International Ltd. approached GOT with supply and sanitation services caused the Gov- a proposal for a joint venture with DAWASA in ernment to mount a significant effort to reform which GOT would have assumed most of the the policy environment and create an institu- risk. Hoping to attract more private sector in- tional framework that would promote greater vestment through a concession, GOT opted to accountability. invite Biwater and three other firms, Saur In- ternationale, Vivendi Environnement (which became Veolia Environnement in 2003), and ES.2 Policy and Institutional Reforms Northumbrian Water Group PLC, to submit bids. However, the four bidders each proposed The National Water Policy (NWP) adopted in a different approach and the bids varied so 2002 called for the development of an enabling much, that the Parastatal Sector Reform Com- environment and incentives for the delivery of mission (PSRC) decided it could not evaluate reliable, sustainable and affordable services, them in a fair manner and did not award a con- and the introduction of full cost recovery and tract. financial autonomy. It laid the foundation for reform and public-private partnerships (PPP) PSRC subsequently decided to request bids in water and sewerage services, and as a result, for a lease contract in which Biwater, Saur and DAWASA was authorized to appoint an opera- Vivendi expressed interest. However the pre- tor under a concession or lease contract. The bid process was marked by controversy and policy also endorsed independent, transparent concerns over the riskiness of the undertaking. and fair regulation of services and emphasized Ultimately, only Saur and Vivendi submitted the need to improve services for low-income bids, and their bids contained qualifications households. The Energy and Water Utility aimed at reducing the operator’s risks. In May Regulation Authority (EWURA) was created 2000, the PRSC decided the bids deviated from (on paper) with the mandate to regulate service the requirements set out in the request for bids quality and tariffs. and declared the bids non-compliant. A third attempt to secure bids for DAWASA led to a The African Development Bank (AfDB), formal prequalification process organized by the European Investment Bank (EIB) and the PRSC. Although, Saur and Vivendi fully met World Bank (IDA) supported GOT’s plan to re- the prequalification criteria, Biwater Gauff structure DAWASA and financed investments Tanzania Limited (BGT)—a joint venture of in infrastructure under the Dar es Salaam Wa- Biwater International and the German engi- ter Supply and Sanitation Project (DWSSP). neering firm, HP Gauff Ingenieure GmbH—did Executive Summary xv not fully meet the criteria but was nevertheless million to be on-lent by DAWASA to the op- pre-qualified. erator—and repaid by the end of the ten-year contract period. After reviewing the revised bidding docu- ments the pre-qualified bidders, particularly The Lease Contract was signed in February Saur and Vivendi, still had concerns and re- 2003. A transition period ensued and CWS quested several amendments to deal with began operations in August, 2003. Under risks. These requests were not accommodated the terms of the contract, DAWASA, as Les- and, ultimately, only BGT submitted a propos- sor, remained responsible for financing capi- al—with a bid equal to the minimum allowed tal investments and executing most of them. Operator’s Tariff. The World Bank project CWS assumed responsibility for operations team expressed concern about the sound- and maintenance and the PWP and DCW. ness of BGT’s proposal, but following a review CWS would collect tariff revenues, retain its meeting in which representatives of BGT pro- share (the Operator Tariff), remit to DAWASA vided clarifications, the World Bank issued a its share (the Lessor Tariff), and deposit the no-objection to the award of the contract. BGT First Time Connection Tariff into a fund that and a Tanzanian investor, Super Doll Trailer would be used to subsidize connections for Manufacture Company Limited (STM), sub- low-income households. In addition CWS was sequently created the operating company, City required to pay a monthly Rental Fee to DA- Water Services Limited (CWS) in which Biwa- WASA. The financial projections BGT submit- ter, the lead professional partner, held only a ted with its bid showed that CWS would lose 41 percent interest, Gauff held 10 percent, and Tshs 7.9 billion (about US$ 7.9 million at the STM, with no experience in water supply ser- time) over the first two years of operation, and vices, owned 49 percent. would not reach a break-even point until year seven. The operator would require US$8.5 mil- lion in equity from CWS’s shareholders and the ES.4 The Lease Contract Framework US$ 5.5 million line of credit from DAWASA to stay afloat in the early years. Targets for some The drafting of the lease contract incorpo- performance indicators were specified in the rated lessons of experience from Western Af- contract but many (for example, those for water rica. For example, the operator would be re- losses and continuity of supply) would be speci- sponsible for implementing a Priority Works fied only after baseline data were established in Program (PWP) and Delegated Capital Works the course of the contract. (DCW) which together included rehabilita- tion of the production works and the network, The lease contract framework included a De- procurement of bulk and customer meters, velopment Contract between DAWASA and the construction of water kiosks in low-income Ministry of Finance which (among other provi- neighborhoods, rehabilitation of thousands of sions) defined the capital works program, pro- spaghetti connections and the installation of vided for its financing and committed GOT to new connections. As is often the case in such ensuring that DAWASA was adequately funded undertakings, financial projections showed at all times. When the Lease Contract became that the operator would incur losses during effective, most of DAWASA’s staff was trans- the early years of the contract. Its sharehold- ferred to CWS. The CEO of DAWASA (who ers would therefore be required to contribute had also been CEO of DAWASA’s predecessor, sufficient equity to keep the operator solvent. NUWA) and all senior managers remained To help bridge the early financial gap, IDA in place. A staff of about 65 was retained to would finance a line of credit of up to US$5.5 manage the planning and implementation of xvi PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM investment projects and to oversee the Lease than DAWASA’s had been in 2002/03. By May Contract with CWS. 2005 Government arrears for water and sewer- age services amounted to US $1.5 million, but Three mechanisms were designed to improve CWS could not take advantage of the Govern- services for the poor. CWS was expected to con- ment’s guarantee to pay, because many bills struct 250 new water kiosks and use the First were disputed and could not be verified. Fur- Time Connection Fund to connect low-income thermore, CWS did not pay the Rental Fee to households to the network. In addition, DA- DAWASA regularly, periodically withheld Les- WASA would implement the Community Wa- sor Tariff collections to pay its own operating ter Supply and Sanitation Program (CWSSP) costs and failed to deposit First Time Connec- which would support the construction of up to tion Tariff into the account for that program. 50 small water supply and sanitation schemes The number of employees remained excessive in mostly low-income communities not served because BGT had declined the government’s by the networks. offer that would have enabled the operator to limit staff to the number required. Productiv- ity was further compromised bcause CWS em- ES.5 CWS’ Performance and Termination of ployees devised few incentives to change cul- the Lease Contract ture or improve performance. Within months of its start, CWS was already facing serious Management, Operational and Commercial cash-flow problems and, by March 2005, its Performance accumulated losses amounting to about US$ 12.3 million. Initiation of the DCW was delayed When CWS assumed operations in August and the PWP, which should have been executed 2003, it faced a challenging agenda. To meet its within 18 to 24 months, was incomplete when targets, it would have to identify and regularize CWS departed. thousands of unregistered connections, intro- duce a new billing system, reduce water losses Termination of the Contract from an estimated 70 percent to 44 percent in the first three years, and more than double During June–August 2004, the Minister of monthly collections within 12 months. Achiev- Water, DAWASA and CWS discussed these is- ing these targets would require strong lead- sues but failed to agree on how to move for- ership and a change in culture on the part of ward. In September 2004, CWS requested an operating staff and many of the customers. Un- Interim Review of the Operator’s Tariff, but fortunately, CWS was unable to meet many of its an Independent Assessor determined that targets and obligations almost from the start. grounds for a revision of the Operator’s Tar- Its shareholders failed to inject their agreed iff could not be established. In January 2005, equity contributions; the new billing software CWS formally requested renegotiation of its system was only completed in March 2005 and contract. In February 2005, DAWASA decided turned out to be woefully inadequate; limited to call the performance bond for the amount of progress was made in cleaning up the customer unpaid penalties, Lessor Tariff revenues and database; values for baseline data were not pro- Rental Fees, but was surprised to learn that the posed; and the number of connections rehabili- performance bond could not be called for less tated with meters installed was less than one than its full value. By March 2005, a mediator quarter of the targets for the first two years of had been engaged to assist the parties to revise the contract. In addition, CWS’ average month- the contract. While GOT and DAWASA were ly collections in 2004/05 reached only 52 per- willing to accommodate most of CWS’s propos- cent of projections and were 21 percent lower als, they were not willing to extend the Lease Executive Summary xvii Contract for an additional five years unless Contract and ordered CWS to pay DAWASA CWS’s collection performance improved. CWS damages of Tshs 6,990 million (about US$ 5.6 insisted on the extension but refused to com- million) and one half the cost of the arbitration mit to higher collection targets, and a compro- proceedings. But as of August 2010 DAWASA mise could not be reached. The Expert Panel had been unable to collect from the bankrupt process which was foreseen in the Lease Con- CWS. tract and which would have bound both parties to its decision was never invoked. In August 2005, BGT initiated a case against the United Republic of Tanzania at the Interna- On 12 May 2005, DAWASA called the perfor- tional Center for the Settlement of Investment mance bond in its entirety. After CWS failed Disputes (ICSID). BGT claimed that the termi- to reinstate the performance guarantee as re- nation of the contract, deportation of CWS’s quired under the Lease Contract, DAWASA senior managers and seizure of the company’s served a notice of termination of the contract assets constituted a breach of a trade treaty be- on May 25, 2005. Fearing that CWS was finan- tween the U.K. and the Republic of Tanzania cially unable to continue to operate and that in- and argued that it should receive compensation terruption of services was imminent, DAWASA of about US$ 20 million. In July 2008, the Ar- asked CWS to cooperate with a speedy transi- bitration Tribunal by a vote of 2 to 1 found that, tion. However, CWS would not agree to the ter- although GOT had violated the treaty, these vi- mination of the contract. The Minister of Water, olations did not cause the losses and damages in the face of declining public support for pri- for which BGT claimed compensation and no vate participation and impending elections in damages were awarded. November 2005, decided to end the stalemate. On 1 June 2005, CWS’ three British managers were deported. A few days earlier GOT had cre- ES.6 The DAWASA-DAWASCO Arrangement ated a new public operating company, the Dar es Salaam Water and Sewerage Corporation DAWASCO’s Performance (DAWASCO), which assumed management of the services that same day. The new public operator, DAWASCO took over operations on 1 June 2005 and the GOT con- During the crisis, The World Bank, AfDB and tributed Tshs 2 billion to its equity. A ten-year EIB played a neutral but supportive role. The lease contract with DAWASA was signed on 2 World Bank made DWSSP funds available to September 2005 and revised on 12 December pay for the Independent Assessor and for the 2005. The Lease Contract was almost identical mediator. After the contract was terminated, to the contract with CWS. However, although it the World Bank decided that, in the interest provided for a performance bond, none was re- of ensuring the continuation of services, it quired. Likewise, although DAWASA’s author- would not suspend disbursements and would ity to impose financial penalties for non-com- continue to support the project within the new pliance with performance targets was asserted DAWASA-DAWASCO contractual framework. in the Lease Contract, in practice DAWASA never imposed them. The lack of reliable base Two International Arbitration Cases data continued to make it impossible to specify all targets but those that were specified were In 2007, an Arbitration Tribunal convened by considered to be realistic. the United Nations Commission on Interna- tional Trade Law (UNCITRAL) ruled that DA- Like CWS before it DAWASCO failed to meet WASA was justified in terminating the Lease its targets in the early years but it took some xviii PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM promising steps. New salary scales based on were state enterprises with their CEOs and job level and performance rather than seniority Boards appointed by the Minister of Water were established in 2007. The number of staff undermined the authority of DAWASA (the was reduced from 1320 in 2005 to 885 in 2009. principal) over DAWASCO (the agent). In the By July 2009 all area managers had signed in- absence of a performance bond or other finan- ternal delegated management contracts under cial incentives, there was no mechanism to which they could earn performance bonuses. enforce performance targets. In addition, the In 2008/09 DAWASCO’s performance with re- collection of performance data for this case gard to some key performance indicators, such study revealed that there was a lack of central as water quality, was very good. Performance direction and coordination of data reporting on other indicators lagged but was improving. which contributed to DAWASA’s inability to ef- The installation of new connections picked up fectively monitor performance and enforce the significantly in 2009/2010 and the total num- Lease Contract. In an effort to correct this, the ber of active connections that were being billed Lease Contract was amended in late 2010, but reached 100,000 in June 2010—an important the DAWASA Act and its Regulations also re- benchmark though still well below the project quired revisions and, as of January 2011, they target of 170,000. Because of long delays in had not been amended. the installation of bulk and district meters, the most serious problem remained non-revenue Implementation of Capital Investments water which, at 53 percent in May 2010, was still unacceptably high. The decentralization of area Delays in the completion of capital works and management in 2007 and 2008 had brought new connections slowed down the pace of op- significant improvements in performance. In erational improvements. The closing date of 2009 the introduction of text message remind- DWSSP was extended three times to allow for ers, cell phone bill payments and other inno- completion of investments, including some vations by area managers resulted in a notice- that were added in the course of the project. able improvement in collection efficiency but, The final closing date was November 2010. The because of the high percentage of non-revenue PWP was completed in March 2007. DCW for water, total collections fell far short of projec- which CWS was responsible were taken over by tions. DAWASCO hoped that the completion of DAWASA when CWS left and then delegated bulk metering in late 2009, a district meter- again to DAWASCO in 2008. While it was orig- ing and water loss reduction program, and the inally conceived as a five-year program to be planned completion of the metering and reha- completed by June 2008, the DCW was com- bilitation of connections aimed would reduce pleted in June 2010. Similarly, the non-delegat- losses to about 40 percent by late 2010. ed works for which DAWASA was responsible experienced delays. Critical activities, such as The DAWASA-DAWASCO Relationship the installation of zonal bulk meters were not completed until late 2010. In a lease contract arrangement it should be quite clear that the lessor is the principal and Financial Viability of DAWASA and DAWASCO the lessee is the agent that is accountable to the principal. Unfortunately, while the Lease DAWASA’s and DAWASCO’s financial per- Contract designed for CWS was transferred to formance was undermined by several fac- DAWASCO with few changes, the governance tors, most prominently DAWASCO’s failure arrangements for DAWASA and DAWASCO to achieve billing and collection targets. In resulted in ambiguity regarding the principal- every year since DAWASCO began operating, agent relationship. The fact that both entities collected Operator Tariff revenues have been Executive Summary xix inadequate to cover its core recurrent expens- After EWURA began to function, DAWASA es. To cover the gap between expenses and in- submitted a tariff proposal that included a real come, DAWASCO received subsidies of Tshs 7.3 increase in the tariff (as foreseen in the Lease billion (about US$ 6 million) over 2006/2007– Contract) and the indexation of the tariff for 2008/2009 and borrowed from DAWASA. 2006/2007. This was approved by EWURA. However, EWURA subsequently declined to In 2008/2009, although DAWASA was able approve the annual indexation of tariffs pro- to cover its core recurrent expenses with its posed by DAWASA for both 2007/2008 and share of revenues from water and sewerage 2008/2009, citing DAWASCO’s poor perfor- services, it was unable to fund other obliga- mance, DAWASA’s failure to enforce perfor- tions, such as counterpart funding, and relied mance targets, and their failure to submit an on cash transfers from GOT. DAWASA’s finan- updated basis for calculating flat tariffs for un- cial weakness was cause for concern because metered customers. In fact, DAWASA and DA- in July 2008 it was expected to begin servicing WASCO were unable to produce a reliable ba- its debt to GOT. DAWASA would also have to sis for estimating consumption by unmetered repay GOT for the amounts on-lent to the two customers because of the lack of sufficient bulk operators. GOT hoped to recoup at least part of and zone metering. the sub-loan to CWS through bankruptcy pro- ceedings but, given DAWASCO’s poor financial DAWASA appealed its case in Tanzania’s Ad- performance, it would not be able to repay its ministrative Court, but the latter ruled in favor sub-loan for some time. In light of this, the of EWURA. However, DAWASA’s subsequent Ministry of Finance rescheduled DAWASA’s request for indexation in 2009 was approved initial debt repayments for two years until July and resulted in a 30 percent increase in the 2010 but in June 2010, when it was clear that customer tariff for 2009/2010. In its decision, DAWASA would be unable to initiate payments, EWURA set several standards for improve- it requested that the interest be reduced to five ments in performance prior to the next tariff percent; CWS debt be cancelled; and the ini- review. EWURA’s decisions were controversial tial debt repayment be rescheduled again un- because it did not recognize the concept of rou- til July 2013. These results were disappointing tine annual indexation and subjected a request since the achievement of financial viability was for indexation to the same rules and procedures one of the major objectives of the restructuring as a full tariff review. The fact that tariffs were of the services and of the DWSSP. not indexed as projected contributed to DAWA- SA and DAWASCO’s failure to meet financial targets but, on the more positive side, the regu- ES.7 Regulation lator showed itself to be committed to holding them accountable for their performance and The Minister of Water served as Interim Regu- had an positive impact in this regard. lator during the entire period of the Lease Con- tract with CWS and the first year of DAWAS- CO’s Lease Contract. While CWS was operator, ES.8 Coverage and Access of the Poor to the Minister approved the tariff indexation for Service 2004/2005. After DAWASCO became opera- tor in June 2005, the Minister deferred the ap- Coverage proval of the tariff indexation for 2005/2006 to EWURA. As a result, there was no indexation A 2009 survey showed that 54 percent of house- for the year 2005/06 because EWURA did not holds in the DAWASA service area had access begin to function until a year later, in mid-2006. to piped water supply from DAWASCO as their xx PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM main source of drinking water—a percentage Users Associations (WUAs) and contribute 5 that was well below the DWSSP end-of-project percent of the cost of the schemes. By late 2010, target of 80 percent. With regard to the loca- 50 small water supply schemes and ten sanita- tion where piped water was accessed, about 17 tion facilities were completed. DAWASA esti- percent of households had a house connection mated that 275,000 people were benefitting. In or yard tap; about 26 percent of households re- addition to the DWSSP-funded schemes, DA- ported they bought piped water from neighbors; WASA rehabilitated 35 existing small systems eight percent bought piped water from vendors and about 237,000 people benefitted from and less than two percent of households bought these. piped water from a DAWASCO kiosk. Once the systems were complete, WUAs op- Services for the Poor erated and maintained them and charged us- ers for their use. In the large majority of cases, Three mechanisms were designed to improve CWSSP schemes functioned well. However, services for the poor: (i) the First Time Con- on-going training and support was needed be- nection Fund; (ii) the construction of DAWASCO cause of frequent turnover in management and water kiosks; and (iii) the DAWASA Community staff of WUAs and occasional technical issues. Water Supply and Sanitation Program (CWSSP). DAWASA had created a Community Liaison Unit to manage the organization of WUAs and Although the First Time Connection Fund construction of CWSSP schemes and it con- was established at the start of the DWSSP, giv- tinued to function after the project. In 2010, en the nature of settlement in Dar es Salaam, in recognition of the need to step up efforts to the criteria for its use made it difficult to apply serve the poor, DAWASCO established a Pro- and it was therefore not used during the period Poor Unit that was expected to assume respon- under review. Throughout the project period, sibility to supervise kiosk services, promote the Fund remained in a separate account that connections using the First Time Connection could be used to make connections for low-in- Fund, and provide technical support to off- come households in the future. network schemes such as those created by the CWSSP. In June 2010 DAWASA reported that, of 294 new kiosks built with DWSSP funds under the Affordability DCW, 249 were receiving water and function- ing. Assuming an average of 500 people was In 2009, a basic supply of water from a DA- served by each kiosk, about 124,500 additional WASCO connection, a DAWASCO kiosk or a people were being served by DAWASCO kiosks WUA-managed scheme was quite affordable by the end of the project. (These numbers ex- for a low-income household with two mini- clude those served by kiosks built under the mum incomes. Unfortunately, the large major- CWSSP, which is considered separately.) DA- ity of low-income households without access WASA also reported that although a number of to adequate supply from one of those sources constraints and delays were encountered dur- paid considerably more than the DAWASCO or ing the implementation of CWSSP (which cre- WUA tariffs. Moreover, while the cost of water ated borehole-based schemes with kiosks and from a direct connection was affordable for a connections), it eventually achieved its targets very poor household with an income of about and was considered one of the more successful three quarters of one minimum wage, the cost components of the DWSSP. As CWSSP systems of a basic supply of water from a DAWASCO ki- were independent from the network, the tar- osk was more than twice the generally accepted get communities were required to form Water benchmark of four percent of income, with the Executive Summary xxi result that such households could not consume DAWASA revised the criteria for the a desirable volume of water from safe sources. FTC program and DAWASCO was able to It was clear that a strategy to improve the af- start using the fund. fordability of water would have to focus on in- zz When introducing community-managed creasing the number of connections and con- schemes, planners should anticipate the veniently located kiosks with bulk water from complexities of working with communi- DAWASCO. ties. Strategies need to be developed to avoid or dispel the disruptive behavior of political factions, deal with high turn- ES.9 Lessons over of WUAs and manage community disputes. The history of the sector reforms and lease zz On-going training and support for WUAs contracts for water supply and sewerage in Dar and kiosk managers after completion of es Salaam provides a number of lessons that the schemes is required to resolve tech- could be useful for policy makers and profes- nical problems, strengthen skills and sionals who are planning such interventions. make up for frequent turnover in the ex- ecutives and staff. Policy Environment Vested Interests zz Preparation of the policy environment for sector reform and PPP should include zz Proven experience in introducing cul- a strategy for dealing with the failure of ture change and using performance in- the PPP and plans for quickly putting in centives and other controls to eliminate place a new, if only temporary, arrange- illicit behaviors and improve efficiency ment and for developing a longer-term should be a pre-requisite for managers solution. (whether public or private) who are hired to turn around a non-performing water Consultation with Potential Customers and supply and sewerage utility. Services for the Poor zz The completion of a comprehensive cus- tomer survey and introduction of billing zz The effective management of kiosks and and collection systems that enable man- the connection of poor households re- agement to control rent-seeking behav- quires special attention during the turn- iors are high-priority elements of a strat- around of a water utility when the focus egy to improve the performance of water of management is on system wide reha- utilities and cannot be neglected. bilitation, reduction of water losses and overall commercial performance. Creat- Bidding and Selection Process ing a dedicated unit/team to manage the CWSSP worked well in DAWASA and this zz Transaction preparation teams should led to the establishment of a unit in DA- avoid letting a potential bidder take WASCO toward the end of the DWSSP. control of the process and meticulously zz Better consultation with low-income avoid even the appearance of preferential households, evaluation of housing condi- treatment of any of the bidders. tions and an assessment of willingness zz Prequalification criteria should be ap- to pay might have resulted in the design propriate to the challenge of the under- of a more practical and effective First taking and, once adopted, should be en- Time Connection program. Belatedly, forced. xxii PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM zz Transaction preparation teams should the arrangement in two phases with a less consult meaningfully with all qualified risky management contract for the first 18 bidders and take their concerns seriously. to 36 months, during which time the base data are established, and a transition to Selection of the Lease Contract Model the riskier lease contract thereafter. zz To protect Lessor Tariff revenues from zz Preparation of a lease contract transac- misappropriation, all revenues should be tion should be guided by careful consid- deposited into a joint lessor/operator ac- eration of the conditions that have been count and allocated by a mechanism that associated with success elsewhere and protects each party’s rights. adequate attention to sustaining those zz A detail such as the inability to call a conditions throughout the contract. performance bond in part can result in zz Lack of adequate prior experience of delays and the worsening of a dispute to the lead professional partner and /or its the point that it is irreconcilable and the failure to take a predominant share in cost of failure extremely high. the operating company may result in a zz The feasibility of obtaining parent com- failure to mobilize the managerial and pany guarantees for the performance of financial resources necessary to ensure an operator needs to be examined. the success of a lease contract. zz A lease contract between two public en- Tariff Setting Framework tities is not likely to be successful un- less there is an effective principal-agent zz The regulatory framework and the cho- framework. Examples from other coun- sen management model should be mutu- tries seem to indicate that public-public ally consistent. Independent regulation arrangements have been successful when can be consistent with the lease contract the entities involved were owned by dif- model if the regulator recognizes the ferent levels of government or where one lease contract as an essential element of was a subsidiary of the other. the regulatory framework and endorses the tariff adjustment and other regula- Financial Viability and Allocation of Risks tory provisions of the contract. Including a pre-specified methodology for revising zz When preparing a PPP on the basis of tariffs that the Regulator endorses might unreliable existing data, it is important also reduce the regulatory risks. not only to recognize the need for veri- zz Many regulatory and PPP experts believe fication and resetting of performance that in a lease contract framework the targets, but also to provide protection regulator should set or approve the over- against the financial consequences of all Customer Tariff but not the Operator the operator’s not being able to achieve Tariff which should be set by agreement operational and commercial targets that between the asset holding company (les- were based on faulty base data. sor) and the operator (lessee). This is con- zz In situations where losses will be in- sistent with the principle of the operator’s curred in the early years, the commit- being subsidiary to the asset holder. ment of the lead professional partner to the long-term objectives needs to be con- Monitoring the Operator’s Performance firmed. An alternative approach that is favored by international private operators zz The performance of water and sewer- but not yet widely tested is to structure age services is not likely to improve in Executive Summary xxiii the absence of rigorous data reporting and require parties to use the pre-speci- and effective information management fied dispute-resolution mechanisms. systems that are used to enforce per- zz All parties have an interest in acting formance standards and guide manage- sooner rather than later to enforce ac- ment decisions. countability, invoke dispute resolution zz Audits and impact surveys should be de- mechanisms and resolve disputes before signed to produce meaningful data and they become irreconcilable and extreme- analysis that can be used by managers ly costly. and planners. When a national statis- tics service conducts impact surveys, the The Role of the World Bank participation of an experienced water supply professional would enhance the zz The current more balanced approach to quality of the results and help to build lo- promoting the efficiency of both private cal capacity. and public operators needs to be further refined by a better understanding of the Mechanisms for Resolving Conflicts conditions that favor positive outcomes for different types of institutional ar- zz To the extent practical, steps necessary rangements and more rigorous efforts to to initiate a dispute-resolution mecha- ensure that these conditions are created. nism (for example, the selection of an zz Care needs to be taken by both interna- Expert Panel) should be completed prior tional lenders and implementing agencies to contract effectiveness so as to enable to ensure that monitoring and verifying speedy invocation of the mechanism meaningful indicators of coverage, afford- when needed. ability and service quality are sustained zz Political and regulatory authorities and not overshadowed by the pressing should avoid getting involved in disputes business of project implementation. 1 Introduction 1.1 Geography and Brief History that included the restructuring of state-owned enterprises, private participation in infrastruc- Located on the eastern coast of Africa, the ture services and the divestiture of about 350 United Republic of Tanzania, including the parastatal entities. former Tanganyika and the semi-autonomous islands of Pemba and Zanzibar, has an area of 1.2 The Government of Tanzania’s about 947,000 square kilometers and a popu- Privatization Program lation of 40 million. With a per capita Gross National Income of US$ 410 on the mainland, At the end of the 1980’s, many of the 425 state- Tanzania is one of the world’s poorest coun- owned enterprises in Tanzania were losing tries. Total population grew at 2.7 percent over money and required subsidies through direct the period 1990–2006 but, due to rural-urban budgetary transfers or soft credit. In response migration, urban population grew at 4.4 per- to this unsustainable situation, the Govern- cent. Twenty-five percent of the population ment of Tanzania (GOT) decided to restruc- now lives in urban areas.1 Dar es Salaam, the ture and/or privatize virtually all state-owned largest city and the commercial center of the enterprises. The Presidential Parastatal Sec- country, and one of the major ports in East Af- tor Reform Commission (PSRC) was cre- rica, had an estimated population of 3.4 mil- ated in 1993 to manage this effort. This was lion in 2009.2 Dodoma, a centrally located city, a remarkable decision for a country that until was chosen as the new capital in 1973 and the then had followed a socialist model of govern- National Assembly moved there in 1996, but ment. A high degree of consensus about the many government offices remain in the com- need for reform was driven by dissatisfaction mercial capital, Dar es Salaam. Survey data with the scarcity of goods and inadequacy of show that about 16 percent of the population services throughout the economy. A study of of Dar es Salaam lives below the poverty line, the impact of the infrastructure privatization and it is estimated that more than one half of program that was carried out in 2005 noted the urban population lives in unplanned or that “By the early 2000’s, the privatization squatter areas.3 From independence in 1961 until the mid 1 The World Bank, World Development Indicators, 2008; 1980’s, Tanzania was a one-party socialist and United Republic of Tanzania, National Bureau of Sta- state. Political and economic reforms were tistics, Tanzania in Figures 2008. introduced after 1985 and the first national 2 The population of Dar es Salaam was estimated (interpo- lated) on the basis of the medium population projections multi-party elections were held in 1994. Agri- presented in DAWASA, Water Supply Improvement Plan, culture dominates the economy and exports in- Final Report, by Dr. Ahmed Abdel Warith Consulting En- clude coffee, tea, cotton, cashews, sisal, cloves gineers, June 2008. 3 United Republic of Tanzania, Household Budget Survey and pyrethrum. During the 1990’s the govern- 2007, Ministry of Finance and Economic Affairs and Na- ment embarked on a liberalization program tional Bureau of Statistics Tanzania, 2009. 2 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM of manufacturing and commercial parastatal 1.3 Methodology enterprises was virtually complete. The pro- gram has been a solid success. It has made a This case study was prepared during June significant contribution to Tanzania’s strong 2009–May 2010. It is based on extensive inter- macroeconomic performance in the last de- views with the key actors as well as a review of cade… However, infrastructure enterprises original documents, consultants’ reports, ben- continued to perform poorly… and little head- eficiary surveys, and reports and performance way was made in their reform. Thus, in the data provided by DAWASA and DAWASCO. early 2000s the Government, through PSRC, Sections 2 through 8 focus on the history of has focused its attention on the privatization the sector reform; the selection of the private of infrastructure.�4 operator, City Water Services (CWS); the ter- mination of the contract; the subsequent cre- Given the success of the privatization of man- ation of the public operator, DAWASCO; the ufacturing and commercial sectors, it was ex- performance of the key actors; and impacts pected that public private partnerships (PPP) on the population, especially the poor, from in infrastructure would lead to significant im- 2003–2010. provements in services. By 2003, some form of PPP had been introduced in five infrastructure In carrying out the case study, the author at- enterprises: DAWASA (Dar es Salaam Water tempted to answer the questions listed below Supply and Sewerage Authority), TTCL (tele- which were chosen for their relevance to the communications), TANESCO (power), TICTS case. The findings and lessons associated with (container terminal) and Air Tanzania. In 2005, each question are summarized in section 9. the above-mentioned study concluded that while two of the transactions (a management i. Was the policy environment appropriate contract for TANESCO and the concession of for successful private participation and for TICTS), could be rated as clear successes, two the subsequent public-public partnership? (the partial divestures of TTCL and of Air Tan- ii. Was the design of the reform and the lease zania) had produced mixed results, and the contract guided by adequate consultation lease contract for water supply and sewerage with existing and potential customers and services was a failure. In fact, it collapsed after an assessment of willingness to pay? less than two years and GOT subsequently cre- iii. Were appropriate measures for serving ated a public operating company, DAWASCO, the poor incorporated? which assumed responsibility for operations iv. Were vested interests identified and were under a similar lease contract with DAWASA. strategies developed to bring them on- board? The objective of the present study is to review v. Was the selection process well-managed, the evolution of the institutional framework for fair and transparent? Were potential bid- water supply and sewerage services in Dar es ders for the contract with a private opera- Salaam from the late 1990s to 2010, assess the tor consulted and their inputs taken into performance of the key actors, and extract les- account in designing the transaction? sons that might be useful to GOT, DAWASA, the World Bank, other international lenders and sector practitioners. Frequent reference is made to the Dar es Salaam Water Supply and 4 United Republic of Tanzania, Privatization Impact As- Sanitation Project (DWSSP) which supported sessment, Infrastructure, July 21, 2005, p. 6. In this report, the term privatization referred to any form of public-pri- the institutional reforms and investments in vate partnership from a management contract to complete the services from 2003–2010. divestiture. Introduction 3 vi. Was the lease contract model (first with Did it provide adequate predictability, the private operator and subsequently transparency and credibility? with the public operator) appropriate to ix. Was a rigorous system for monitoring and the context? auditing operational, commercial and fi- vii. Were the financial and operational projec- nancial performance of the operator es- tions and underlying assumptions realis- tablished and implemented? tic and reliable? Were the tariffs, sharing x. Were the legal contracts well-drafted and of revenues, financial terms and allocation did they include procedures and condi- of risks appropriate and did they allow for tions necessary to ensure smooth relation- both partners to achieve financial equilib- ships and resolve disputes? rium within a reasonable time? xi. Could or should the World Bank and other viii. Was the tariff-setting framework appro- international lenders have managed their priate for the lease contract framework? support for the transaction differently? 2 Policy and Institutional Reform of the Urban Water Supply and Sewerage Sector 2.1 The Urban Water Supply and Sewerage and treatment plants on the Upper and Lower Sector Ruvu River and a small surface water scheme in Mtoni Area within Temeke Municipality, for In 1981 the National Urban Water Authority a total installed production capacity of 273,000 (NUWA) was established under the Ministry of m3/day. The production plants were in need of Water and Irrigation. It became operational in repair; treatment filters had not worked in 20 1984 and, though it had a mandate to develop years; pumps and transformers broke down reg- and manage urban water supply on the main- ularly and there were no bulk meters. Two long, land, it focused its efforts primarily in Dar es and vulnerable, transmission lines of 50 km and Salaam.5 Until 1997, sewerage and sanitation 55 km connected the Upper and Lower Ruvu services were provided by the Sewerage and treatment plants to the distribution network but Sanitation Department of the Dar es Salaam City only a fraction of the total production reached Commission. Throughout the 1980s and 1990s, the city because of (often illegal) offtakes, in- as a result of low tariffs, inefficient operations, cluding for agricultural use. The distribution lack of investment and poor management of bill- network was thought to include about 100,000 ing and collections, NUWA was unable to satisfy connections but there was no map of the sys- demand and the infrastructure deteriorated. tem and the physical address of many connec- tions was unknown since bills went to post office The Waterworks Act, as most recently boxes. Many of the accounts in the registry were amended in 1997, ushered in partial decentral- duplicates—set up by staff to permit customers ization of responsibility for urban water supply to avoid paying arrears—and only about 15,000 and combined the responsibility for water sup- customers paid bills regularly. Only 1000 con- ply and sewerage services through the creation nections were metered and most billing was on of Water and Sewerage Authorities. Its Regu- the basis of estimated consumption.7 The sewer- lations endorsed the principles of autonomy age systems consisted of about 140 km of sewers, and the recovery of O&M costs. Until the En- several stabilization ponds that received septage ergy and Water Utilities Regulatory Authority as well as sewerage and a marine outfall, all at (EWURA) was created, the Boards were re- various stages of disrepair.8 sponsible for setting consumer tariffs.6 5 DAWASA, Quarterly Progress Report No. 24 (A), May 2009, p.1. As part of this reform, the water supply func- 6 United Republic of Tanzania, Chapter 272, The Water- tions of NUWA and the sewerage functions of the works Act of 1949, as amended most recently by Act. No. City Commission were transferred to the Dar es 8 of 1997; and The Waterworks Regulations, G.N. No. 371 of 1997. Salaam Water and Sewerage Authority (DAWA- 7 Dar es Salaam Water Supply and Sewerage Project SA) which became responsible for piped water (DWSSP), Project Appraisal Document, The World Bank, supply and sewerage services in the metropoli- April 2003, pp. 3–5. 8 John Sitton, Emerging Markets Group, Ltd., What Went tan area of Dar es Salaam, Bagamoyo and Kiba- Wrong with the DAWASA Lease Contract? PowerPoint ha. DAWASA took over two water abstraction Presentation, 2006. 6 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM Given the dilapidated condition of the as- 2.2 Institutional and Regulatory Reforms sets and the lack of adequate resources for im- proving operations and management, it is not The Dar es Salaam Water Supply and surprising that performance did not improve Sewerage Authority Act of 2001 after DAWASA was established. Although in 2002 about 73 percent of the households in The DAWASA Act reconfirmed DAWASA’s legal Dar es Salaam was said to have access to piped status as an autonomous corporate body, speci- water from house connections, yard taps, fied its corporate governance structure and up- standpipes and neighbor’s taps, service was dated DAWASA’s functions, powers and privi- available for only a few hours and a few days leges to make them consistent with the new per week in most areas. The water that did regulatory framework (see below). Of particu- reach consumers often was not safe for drink- lar interest is the provision giving DAWASA the ing. About 16 percent of households got water power to appoint an operator under a conces- from boreholes, protected wells and springs sion or lease contract, effectively making DA- and 8 percent from unprotected sources. It WASA an asset holding company. It is notewor- was estimated that more than 45 percent of thy that the Act does not require DAWASA to households bought water from neighbors, appoint an operator and allows it to continue tankers or other vendors, for which they paid to operate services.12 DAWASA is not a decen- rates that were considerably higher than the tralized entity though there is local representa- DAWASA tariff.9 The water sold by vendors tion on its Board. The Chairman of its Board is came from the both the piped system and pri- appointed by the President of the Republic and vate boreholes. the Chief Executive Officer (CEO) is appointed by the Minister of Water, upon the recommen- Low salaries made it difficult for DAWASA dation of the Board. Other Board members in- to hire and retain qualified staff. Customer clude representatives of the Ministry of Water, service employees engaged in illicit activities Ministry of Finance, Dar es Salaam City Coun- such as accepting bribes for illegal connec- cil, two large water consumers, the Community tions, opening duplicate accounts to allow Water User Organization, the Coast Region ad- customers to avoid payment of arrears, and ministration and the private sector. retaining collected revenues.10 They and other vested interests could be expected to resist Both DAWASA and any operator appointed by efforts to control unaccounted-for water, in- DAWASA are subject to regulation by EWURA troduce management control systems and ex- and each must obtain a license from EWURA. pand legitimate access to low-cost piped water The DAWASA Act affirms that the regulator services. has authority to approve tariffs, monitor ser- vice quality, performance standards and envi- As is often the case, sanitation services were ronmental impacts, and lay down or approve less developed than water supply. Less than 10 standards and codes of conduct governing the percent of the urban population was connect- ed to a sewerage system. Another 20 percent, mostly in the middle and upper-income ranks, 9 United Republic of Tanzania. 2002 Population and used septic tanks and the remainder depended Housing Census, as reported in Baseline Survey on Im- pact Assessment of the Dar es Salaam Water Supply and on pit latrines. Most wastewater treatment fa- Sanitation Project, Report. National Bureau of Statistics, cilities were not functioning and pit emptying December 2008. services were inadequate. The city suffered 10 John Sitton, op. cit. 11 DWSSP Project Appraisal Document, op. cit., p. 4. periodic outbreaks of cholera and other water- 12 United Republic of Tanzania, The Dar es Salaam Water borne diseases.11 Supply and Sewerage Authority Act, No. 20 of 2001. Policy and Institutional Reform of the Urban Water Supply and Sewerage Sector 7 relationship between the operator and custom- appointed by the Minister of Water and Irriga- ers. Under the DAWASA Act, disputes of a reg- tion. Its Director-General/CEO is appointed ulatory nature arising between DAWASA and by the Board upon nomination by a committee an operator (e.g., regarding tariffs) are to be that includes the Permanent Secretary of the submitted to EWURA. Ministry and two representatives of the private sector and consumers. EWURA’s Consumer The Energy and Water Utilities Regulatory Consultative Council is appointed by the Min- Authority ister to represent the interests of consumers. Under its 2001 enabling legislation, EWURA is Though its enabling legislation was approved a corporate body responsible for regulation of in 2001, EWURA was not actually established electricity, petroleum and natural gas pipeline until 2005 and did not begin to exercise its transmission, natural gas distribution, and wa- functions until 2006. During the interim, ter and sewerage services. It is independent in which included the period during which the that it is funded by license fees and regulation lease contract with the private operator for levies of no more than one percent of the gross water supply and sewerage services in Dar es operating revenues of regulated suppliers. Its Salaam was in effect, the Minister served as In- mandate is to promote competition and efficien- terim Regulator. cy; protect the interests of consumers; promote the availability of services to all including low- EWURA’s legislation and its own rules and income, rural and disadvantaged populations; guidelines allow service providers some flex- protect the financial viability of efficient suppli- ibility in calculating proposed tariffs and give ers; enhance public knowledge and awareness EWURA a great deal of discretionary authority about regulation and customers’ responsibili- in reviewing and approving tariffs. EWURA’s ties and rights; and take into account the need Tariff Application Guidelines 2009 include de- to protect and preserve the environment.13 tailed procedural rules for tariff reviews and a formula for calculating revenue requirements. Subject to sector legislation, EWURA is- The Guidelines do not mention indexation to sues licenses, establishes standards for goods reflect inflation and foreign exchange move- and services, regulates rates and charges and ments as a distinct category of tariff adjust- monitors the performance of service provid- ment. EWURA’s discretionary authority and ers. It also facilitates resolution of complaints approach to tariff reviews is somewhat at odds and disputes, disseminates information and with the lease contract model of private par- consults with customers. Water sector legisla- ticipation in water services that was chosen for tion (the DAWASA Act of 2001 and the Water Dar es Salaam, an issue that will be discussed Supply and Sanitation Act of 2009) specify that later. EWURA regulates Water Authorities and op- erators engaged by the Authorities (whether The National Water Policy public or private), but not community-owned water supply organizations. The DAWASA Act The National Water Policy (NWP) adopted in (Article 26) gives EWURA regulatory author- 2002 laid the foundation for reform and pub- ity to examine and approve tariffs chargeable lic-private partnerships (PPP) in water supply for the provision of water supply and sewerage and sewerage services. Acknowledging the services as submitted by DAWASA. deteriorated condition of assets and the poor The Chairman of EWURA’s Board is appoint- 13 United Republic of Tanzania, The Energy and Water Reg- ed by the President. Other Board members are ulatory Authority Act, No. 11 of 2001. 8 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM performance of the urban water supply and lease contract. In addition, as this case study sewerage services, the NWP specified that will demonstrate, it would take time for the new GOT’s policy objectives for the urban water institutions to start up and begin functioning, supply and sewerage sector were to: guide the for cultural and historical attitudes and habits development and management of efficient, ef- to change, and for reliable data to be generated fective and sustainable urban systems; create —all of which are important for delivering ef- an enabling environment and incentives for fective services and creating accountability. the delivery of reliable, sustainable and afford- able services; develop an effective institutional framework; promote full cost recovery; ensure 2.3 Support of the Reform by International financial autonomy; and enhance water de- Development Agencies mand management and wastewater disposal. The policy also endorsed independent, trans- In 1999, the African Development Bank parent and fair regulation of services, empha- (AfDB), the European Investment Bank (EIB), sized the need to improve services for low-in- the French Development Agency (AFD) and the come households, and encouraged private sector World Bank had expressed an interest in fund- participation and the recognition of small-scale ing investments in Dar es Salaam. Ultimately, service providers. when the Dar es Salaam Water Supply and Sanitation Project (DWSSP) became effective The Water Supply and Sanitation Act, 2009 in 2003, AfDB provided US$ 48 million; EIB provided US$ 34 million and the World Bank This Act, which updates sector legislation to be (IDA) provided US$ 61.5 million. AFD with- consistent with the NWP, governs the creation drew from the project in September 2002. and functions of Water Authorities (other than DAWASA, which is covered by the DAWASA AfDB had initiated project preparation in Act of 2001) and community-owned water sup- 1997 under the assumption that DAWASA ply organizations, and decentralizes their gov- would continue to operate the services but put ernance. It makes Water Authorities subject to it on hold when GOT began to search for a pri- regulation by EWURA and allows them to ap- vate operator. About that time the World Bank point service providers. also began discussions with GOT and AfDB about a joint investment program. In May Conclusion 1998, the World Bank advised DAWASA and the Minister of Water that it would appraise At the time of the award of the lease contract to the project only after certain conditions were CWS, the legal and regulatory framework was met, including the selection of a private opera- generally supportive of public-private partner- tor under a lease contract and a commitment ships (PPP). It provided a basic foundation to by the GOT to set tariffs at levels that would support efficiency, financial viability, avail- fully cover O&M costs and contribute to capital ability of services and accountability. How- costs. Private participation in DAWASA was a ever, the role of the independent regulator in condition of both the World Bank Adjustment the context of the lease contract had not been Program and the International Monetary Fund clearly worked out. Because EWURA had not (IMF). The GOT and DAWASA were already yet been established when the private operator committed to private participation, although a was engaged, there was no regulatory history consensus on the legal form of the arrangement or precedent to provide predictability. In fact, was not reached until late 1998 when the lease the regulator’s discretionary authority created contract model was adopted. At about the same discomfort among some of the bidders for the time, the U.K. Department for International Policy and Institutional Reform of the Urban Water Supply and Sewerage Sector 9 Development (DfID) funded a utility reform secondary distribution pipes and the adop- advisor, a regulation advisor for PSRC and a tion of commercial policies that would favor public awareness campaign to prepare the pub- household connections; (c) enhancing finan- lic for the changes to come. cial viability by raising tariffs to levels that would cover operational and maintenance DWSSP was conceived as the first phase of a costs, service long-term debt and contribute program of investments to improve water sup- the capital expenditure program, and eventu- ply, sewerage and sanitation services in Dar ally approximate the long-run marginal cost es Salaam. The rationale behind the project of the services. It was anticipated that the pri- design was that before creating more produc- vate operator would upgrade commercial op- tion capacity, it was essential to improve the erations to industry standards, reduce physi- efficiency of operations and the effectiveness cal and commercial unaccounted for water of the existing infrastructure and put in place and increase the collection of bills. The project mechanisms to ensure that investments led to also included support to construct community improvements in services for the poor. Thus, water supply systems and on-site sanitation DWSSP was designed to address prolonged facilities for about 50 low-income communi- under-investment and inadequate mainte- ties that would not be served by the DAWASA nance by (a) rehabilitating DAWASA’s existing system in the near future. It was expected water production, transmission, storage and that subsequent projects would expand water distribution facilities and wastewater collec- storage, production, transmission and main tion and treatment facilities; (b) extending distribution to meet growing demand and piped water supply service to un-served ar- DWSSP provided funding to begin planning eas through the construction of primary and for these future investments. 3 The Pursuit of a Private Operator for Dar es Salaam 3.1 First Bidding Process, 1997–1998 hired to present the options at the workshop. As mentioned above GOT had hoped to secure sig- The effort to engage a private operator for wa- nificant private investment through a conces- ter and sewerage services in Dar es Salaam fol- sion. However, citing the lack of baseline data lowed a long and costly path that dragged out and recent experiences with PPP, the consul- for six years before a contract was awarded. As tants recommended a management contract. the following account reveals, the expectations They argued that this would allow DAWASA and objectives of the various interested parties to inventory the system and clean up the cus- often clashed. The effort began in 1997 when tomer base and prepare the way for a second the British company Biwater International phase arrangement in which the private sector Limited approached GOT with a proposal for would assume greater risks. World Bank staff a joint venture with DAWASA (in which GOT agreed with the consultants that a long-term would have assumed most of the risk) to devel- concession involving significant private invest- op the system. Hoping to attract more private ment was unlikely to attract sufficient interest sector investment through a concession, GOT among bidders but, referencing the success- invited Biwater and three other firms, Saur In- ful lease contracts in West Africa, the World ternationale, Vivendi Environnement (which Bank team favored a lease contract, in which became Veolia Environnement in 2003), and operational responsibility and commercial Northumbrian Water Group PLC, to submit risk would be transferred to the private opera- bids. However, the four bidders each proposed a tor. Under a lease contract, the operator would different approach, among them (respectively) provide equity for working capital while most a joint venture, a concession, a lease contract, of the investment capital would be provided by and a management contract. At this point, re- donors. The DAWASA Divestiture Team chose sponsibility for the transaction was transferred the lease contract option. to PSRC. Since the four bids varied significant- ly, PSRC decided it could not evaluate them in a At this point it is worth considering the fol- fair manner and did not award a contract. lowing conditions that favor a successful lease contract and whether they existed in Dar es Salaam or could be incorporated into the lease 3.2 The Adoption of the Lease Contract contract.14 Model zz Government commitment and a policy In mid-1998, with the assistance of the World environment that are supportive of PPP, Bank, the DAWASA Divestiture Team (with zz The support of key stakeholders, includ- representatives of PSRC, DAWASA, the Minis- ing existing and potential customers, try of Finance and the Ministry of Water) or- ganized a workshop to examine the options for 14 The conditions listed here are those that the author found private sector participation. Consultants were to be most relevant to this study. The list is not exhaustive. 12 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM existing alternative service providers, promoting desirable changes in behav- staff of the water utility, ior, assessing customer needs and keep- zz Tariffs and demand levels that are likely ing key stakeholders informed. to result in adequate revenues to cover O&M costs and depreciation in the near Many of these conditions already existed or term, and gradually contribute to capital could have been incorporated into the lease con- costs, tract framework. Most of them were explicitly zz The availability of capital finance from addressed during the planning process. How- government and international lenders to ever, subsequent events showed that not all the rehabilitate and expand infrastructure, conditions were given adequate attention dur- zz A competent asset holding company with ing preparation or sustained throughout the life sufficient authority, funding and staff of the two lease contracts that ensued. to manage major capital investments, supervise the operator and enforce the lease contract, 3.3 Second Bidding Process, 1998–2000 zz Periodic (annual or more frequent if warranted) adjustments of the operator’s Following a competitive tendering process, tariff to reflect inflation and changes in PSRC hired a transaction advisor (under an ex- exchange rates, isting World Bank supported Privatization Proj- zz A transparent, predictable and fair regu- ect) to assist with preparation of bidding docu- latory framework for major tariff revi- ments, including a draft lease contract, and the sions and rebalancing the sharing of tar- selection of an operator. It was decided that the iff revenues between the asset holding four bidders that had participated in the earli- company and the operator, er process would be invited to bid on the basis zz A well-drafted contract that provides ad- of the share of the customer tariff the opera- equate protection for both parties (the tor would retain to cover its costs (the operator asset holder and the operator) against tariff). Northumbrian withdrew and the other risks they cannot control and clearly three firms attended two pre-bid meetings. specifies each party’s obligations, zz Mechanisms for resolving conflicts that The three interested bidders supported the are used when needed and respected by selection of the lease contract model. However a all parties, number of problems developed. Biwater, a Brit- zz A requirement that the lead professional ish company, alleged that the structuring of the partner of each bidder have relevant op- lease contract and the bidding process favored erational experience in similar environ- the French firms, Saur and Vivendi. During the ments, that the lead professional partner preparation phase, the bidders attended two will own a predominant share (70 to 90 pre-bid meetings where all, but particularly percent) of the operating company and Saur and Vivendi, conveyed concerns about the has the capacity to provide significant operator’s risks and asked for revisions to the working capital, especially in the early bidding documents to alleviate these, but PSRC years when tariff revenues are not likely turned down their requests. In the end, Biwater to be adequate to cover all costs, did not submit a bid. In early 2000, Saur and zz Strategies for serving neighborhoods Vivendi submitted bids, but both contained where piped services are not likely to be qualifications that were aimed at reducing the available in the near term, operator’s risks and covering situations such as zz Effective and sustained consultation non-payment of bills by government agencies. and public relations programs aimed at In the interest of avoiding further delays in the The Pursuit of a Private Operator for Dar es Salaam 13 award of the contract, the World Bank advised and commercial strategy.15 Accordingly, it was GOT to evaluate the qualifications and make a agreed that during the initial five years of the selection. Biwater objected and requested a re- lease contract, the operator would design, pro- opening of the bidding process. In May 2000, cure and supervise Delegated Capital Works PRSC ruled it would be impossible to conduct (DCW) aimed at network improvements. The a transparent evaluation of the two bids and operator would also be awarded two other declared them non-compliant. It decided that contracts with DAWASA aimed at the reha- the lease should be rebid to ensure fairness and bilitation of the two main production units and transparency and increase competition. transmission lines within the first two years of the lease contract, and the supply of about 170,000 customer meters. These two contracts 3.4 Third Bidding Process and Selection of comprised the Priority Works Program (PWP). the Operator, 2001–2003 Altogether, the management of DCW and the two PWP contracts would be worth about US$ PSRC organized a formal prequalification pro- 33 million to the operator. DAWASA would im- cess in hopes of attracting additional bidders. plement Non-delegated Works aimed at major Only Saur and Vivendi met the prequalification rehabilitation and new investments. AfDB, EIB criteria. Biwater International and the German and IDA would finance the DCW, PWP, Non- engineering firm, HP Gauff Ingenieure GmbH delegated Works, and technical assistance and formed a joint venture, Biwater Gauff Tanza- training for DAWASA through the DWSSP. Fi- nia Limited (BGT), in which Biwater held an nancial projections carried out as a basis for 80 percent interest and Gauff held 20 percent. the lease contract predicted that the operator Though both companies had extensive inter- would incur losses during the early years of the national experience and had executed water contract. It was therefore agreed that in order supply and sanitation projects and smaller op- to bridge the period of losses, shareholders erational contracts satisfactorily, BGT did not would be required to provide sufficient equity. fully meet the criteria. It was nevertheless al- In addition, a line of credit of up to US$ 5.5 mil- lowed to participate, presumably in the interest lion would be on-lent by DAWASA to the opera- of enhancing competition. tor on concessional terms.16 There was a delay of several months while Bidding was again based on the operator PSRC renegotiated the international transac- tariff. In light of experience in other countries tion advisors’ contract and the advisors’ team where winning bidders had submitted unreal- remobilized. Preparation of a third set of bid- istically low bids only to call for a renegotiation ding documents was initiated in January 2001. of the operator’s fee a year or two into the con- Experience with lease contracts in other coun- tract, the minimum bid was set at Tshs 322 per tries showed that delays in completing key re- cubic meter in 2002 prices, an amount judged habilitation and network expansion works by the asset holding company could prevent the 15 Matar Fall, Philippe Marin, Alain Locussol and Richard operator from achieving its income targets. Verspyck, Reforming Urban Water Utilities in Western This had happened in Senegal and, as a result, and Central Africa: Experiences with Public-Private Part- the asset holding company had to compensate nerships, The World Bank, Water Sector Board Discussion Paper No. 13, June 2009, pp. 44–45. the operator for loss of revenues. In Niger, to 16 Funds for the line of credit were included in IDA’s funding mitigate this risk, some of the rehabilitation for DWSSP. They were included in that part of the DWSSP works were delegated to the operator, allow- financing that was structured as a loan to DAWASA from Ministry of Finance and were on-lent by DAWASA to the ing it to design and execute a rehabilitation Operator at an interest rate of 11.5 percent, over a ten-year plan that supported its operational priorities term with a five-year grace period. 14 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM to be adequate to cover the operator’s costs While these requests appeared to be reason- within two or three years. The bidding docu- able, many were not incorporated into the final ments also specified the rental fee, amounts to bid documents. Ultimately, World Bank project be paid to the operator for the PWP and DCW team members were disappointed with the bid and the maximum amount of DAWASA’s line of documents and felt that the PSRC and its trans- credit to the operator. To improve their chanc- action advisor had not dealt with the bidders’ es, the bidders could propose an increase in the comments in a satisfactory manner. The trans- rental fee to be paid to DAWASA starting in action advisor, for its part, continued to prefer year six of the contract. a management contract and told World Bank staff and PSRC that it would be difficult for DA- The international operator chosen as the pre- WASA to provide effective oversight of a lease ferred bidder would be required to form a local contract that it considered to be very complex. company in which the preferred bidder would In addition, the transaction advisor could not own at least 51 percent of shares and Tanzanian convince PSRC to agree to the protections that shareholders would own at least 20 percent. The the bidders had requested. In the end, it was minimum subscribed share capital of US$ 2.5 PSRC that made the final decisions about the million would be paid up as of the effective date documentation. of the contract, and shareholders were required to contribute additional equity as necessary to In July 2002, against the advice of the Bank, keep the operator solvent during the initial years. PSRC issued a request for bids. Only Biwater Gauff Tanzania Limited (BGT) submitted a The redrafted bid documents were given to proposal—with a bid equal to the minimum the pre-qualified bidders for comment and, allowed operator’s tariff and a commitment following the first pre-bid meeting in March that shareholders would contribute US$ 8.5 2002, the bidders submitted 250 comments million in equity. BGT’s bid also offered a sub- and requests for clarification. At a second pre- stantially increase of the rental fee in years bid meeting in June 2002, Saur and Vivendi six through ten. The World Bank project team still had concerns. They requested several expressed concern about the financial sound- amendments, such as: ness of BGT’s proposal and its ability to meet performance expectations. The transaction zz Protection of the lease contract from in- advisor also had doubts. Nevertheless, fol- appropriate discretionary decisions of lowing a review meeting in which representa- the Regulator, tives of BGT provided clarifications, the World zz A partial risk guarantee against non- Bank issued a no-objection to the award of the payment of water and sewerage bills by contract. BGT and a Tanzanian investor, Su- government agencies (a protection that per Doll Trailer Manufacture Company Limit- had been proposed by the World Bank ed (STM), subsequently created the operating staff but rejected by GOT), company, City Water Services Limited (CWS) zz Clarification of the respective roles of in which BGT owned 51 percent (the mini- DAWASA and the operator with regard mum required of the bidder) and Super Doll, to the financing of repairs (often a point 49 percent. As a partner in BGT, Biwater, the of uncertainty and conflict in lease con- lead professional partner held only a 41 per- tracts, but one that can be addressed ad- cent interest in CWS. The local investor with equately through detailed specification), no experience in water supply services owned zz Revision of the valuation of the PWP to a greater share than Biwater. be executed by the operator which the French firms considered too low. 4 The Lease Contract Framework When the lease contract became effective, DA- zz The Lessor leased the existing and new WASA became the asset holder and Lessor and assets to the Operator for use and main- CWS became the Operator. Figure 1 shows the tenance during the term of the contract, relationship of the parties in the lease contract while ownership of all assets including framework. small equipment (i.e., pumps, meters, etc.) remained vested in the Lessor. zz The Lessor’s employees, excluding cer- 4.1 Key Terms of the Lease Contract tain recently appointed directors, were available for employment by the Opera- The ten-year lease contract between DAWA- tor. The Lessor remained responsible SA (the Lessor) and CWS (the Operator) was for their pension programs and bore signed in February 2003. A transition period the cost of retaining or retrenching any ensued and CWS began operations in August, employees that were not hired by the 2003. The key terms of the contract included: Operator. zz The Operator’s Managing Director would Organizational Matters be provided from the staff or officers of the majority shareholder. zz The total subscribed share capital was set at US$ 8.5 million. Relationship of the Parties in the Lease Contract Framework Figure 1  Min of Finance Development Contract DAWASA License Delegated Capital Regulator Lease Contract Works & Priority Works Program CWS Ltd License 51% 49% BGT Ltd STM 80% 20% Biwater Gauff Adapted from a Diagram by TRC Global Management Solutions, Ltd., 2005 16 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM Obligations of the Operator and the Lessor consulting engineer and was respon- sible for procurement in accord with zz The Operator assumed the exclusive lender-approved procurement plans, right and obligation to provide water site supervision and commissioning supply and sewerage services and main- of works. Neither the Operator nor tain assets within the specified Opera- any associated entity of the Opera- tor’s Area in conformity with the stan- tor was allowed to bid for the works. dards and conditions of the contract and DCW were to be implemented during relevant legislation.17 DAWASA remained the first five years of the contract. responsible for funding and implement- || Non-Delegated Works included ma- ing capital investments. jor rehabilitation of the water treat- zz At its expense, the Operator was respon- ment units, reservoirs and mains sible for maintenance and repair of as- and new main pipes. It would be sets, up to specified limits (e.g., water procured separately by the Lessor. pipes and mains up to 300 mm in diam- (Firms affiliated with the Opera- eter and six meters in length), replace- tor would be allowed to compete for ment of small equipment, and purchase these contracts.) and replacement of operating equipment (vehicles, tools, office equipment, etc.). Performance Monitoring and Enforcement zz The Operator was required to prepare and maintain an asset register and an zz Key Performance Targets and Other Per- asset management plan. formance Targets were specified and fi- zz The Lessor was responsible for financing nancial penalties were set for failure of of the capital investment programs (with the Operator to comply with Key Perfor- the assistance of international and bilat- mance Targets. eral funding institutions) and arranging zz An Enhanced Monitoring Period of up for implementation. The capital invest- to 18 months was allowed for ascertain- ment program for the first five years had ing the base value of and performance three components: targets for parameters for which reliable base values were not available. Perfor- || The Priority Works Program (PWP) mance Targets that were dependent on included works to improve the reli- the implementation of the capital invest- ability of production at the major ment program would also be agreed dur- source works, the supply of bulk ing the Enhanced Monitoring Period. meters, and the supply and instal- zz The Operator was required to create and lation of customer meters. (Bulk maintain records on its performance, meters would be installed under the create and maintain an asset manage- DCW.) The PWP would be executed ment plan and records, and submit spec- by the Operator under two separate ified annual and quarterly reports to the contracts during the first 18 to 24 Lessor. months of the lease contract period. || Delegated Capital Works (DCW) pro- gram included rehabilitation of the 17 The Operator’s Area included the Water Supply Area distribution networks, rehabilita- and the Sewerage Area, the latter being smaller than the tion of connections, construction of former since the sewerage system did not cover the entire area. There were some minor exceptions to the Operator’s new connections and installation of exclusivity including existing private and community wells bulk meters. The Operator served as and boreholes. The Lease Contract Framework 17 zz As a condition precedent for commence- were applied to 80 percent of the volume ment, the Operator was required to ex- of water consumed. The Customer Tariff ecute a performance guarantee in the for sewerage included: amounts of US$ 3 million and Tshs 2.147 || The Lessor Tariff (sewerage) per cu- billion (equal to US$ 2.15 million) which bic meter – Tshs 30 (US$ 0.03) per it was required to maintain throughout cubic meter and the contract period. || The Operator Tariff (sewerage) per cubic meter – Tshs 106 (US$ 0.11) per Financial Provisions cubic meter. zz The Customer Tariff for water supply, set Figure 2 shows how the Customer Tar- at Tshs 451 (US$ 0.45 ) per cubic meter18 iffs for water and sewerage were shared in year one of the contract, included: among the various components during || The Lessor Tariff (water supply) per the first year of the contract. cubic meter – Tshs 89 (US$ 0.09) zz The Operator’s revenues included: per cubic meter, which was intended || The Operator Tariff for water supply to cover DAWASA’s administrative, and the Operator Tariff for sewerage planning and oversight costs,19 collected from customers, || The Operator Tariff (water supply) || Charges for the provision of ancillary per cubic meter – Tshs 337 (equal and incidental services (tanker water, to about US$ 0.34 ) per cubic meter, which was intended to cover CWS’s operating costs, and || The First Time New Domestic Wa- Figure 2  Components of the Customer ter Supply Connection Tariff (First Tariffs, 2003/04 Time Connection Tariff) – Tshs 25 (US$ 0.02) per cubic meter, which 500 was paid into a fund to cover the cost 400 of new domestic connections with up 300 Tshs/m3 to three water points. 200 zz Domestic customers paid only the Oper- 100 ator’s Tariff for the first five cubic meters 0 consumed in each month. Water Supply Sewerage When the contract was awarded, non- domestic customers paid a higher tariff FTNDWSC Fund Lessor Tariff of Tshs 725 (US$ 0.73) per cubic meter. Operator Tariff Once the Customer Tariff reached or ex- ceeded that level as a result of indexation and increases, non-domestic customers would pay the same Customer Tariff as domestic customers. 18 The Customer Tariff values shown here are the indexed values that were took effect in 2003/04, and are about 4.4 percent higher than the 2002-price-based tariffs shown in zz Sewerage charges, initially set at Tshs the Lease Contract. 136 (US$ 0.14) per cubic meter for sew- 19 The Operator also paid a Rental Fee out of its own rev- enues which, in theory, was intended to fund DAWASA’s erage, were charged only to those cus- contribution to capital costs and debt service on capital in- tomers with a sewerage connection and vestments. 18 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM disposal of liquid or solid waste from Regulatory Levy, which would be set by sewage tankers, etc.), EWURA. || A fee for installing First Time New zz The Operator Tariffs for water supply and Domestic Water Supply Connections sewerage, the fixed Rental Fee and the set at Tshs 145,000 (US$ 145) in Operator Fee for First Time Connections 2002 prices, to be paid to the Opera- were to be adjusted annually to reflect tor out of the fund for this purpose, inflation in accordance with adjustment || Charges paid by customers for the formulas specified in the lease contract. cost of installing new non-domestic The Lessor Tariff and First Time Con- water supply connections, including nection Tariff were to be adjusted annu- the cost of the meter, and new sewer- ally to reflect the index of retail prices. age connections, zz During year five of the contract, the Op- || Payments by customers for private erator Tariff, the indexation formula and extensions. performance targets would be reviewed. || Ten percent of any debt owed to the Revisions would be proposed for con- Lessor that was collected by the Op- tract years six through ten and submit- erator; ted to the Regulator for approval. In ad- || Ten percent of the value of the DCW, dition, conditions that could result in an for design and supervision; Interim Review prior to year five were specified. zz The Operator was required to pay to the Lessor on a monthly basis: Resolution of Disputes || The Lessor Tariff for water supply and The Lessor Tariff for sewerage zz An Independent Assessor would be ap- collected from customers, pointed in the event of a disagreement || The amount by which the water sup- over the value of certain data that could ply tariff for non-domestic custom- demonstrate a material change in cir- ers temporarily exceeded the Cus- cumstances that would warrant an In- tomer Tariff, terim Review of Tariffs. || The (fixed) Rental Fee to be paid zz Disputes for which an amicable solution monthly in advance out of the Op- could not be found would be referred to erator’s own resources (see Table 1), an Expert Panel of three members for a || Ninety percent of the value of cus- determination. tomer debt owed to the Lessor and zz If a dispute relating to the Tariffs were collected by the Operator. not resolved amicably or through the Expert Panel mechanism, it would be re- zz The Operator was also required to col- ferred to the Regulator for resolution. lect and remit to the Regulator the zz If a dispute relating to a matter other than the Tariffs were not resolved amica- bly or through the Expert Panel mecha- Table 1 – Monthly Rental Fee nism, it would be settled by Arbitration in accordance with the UNCITRAL Arbi- Contract Year Rental Fee in Tshs millions tration Rules and in accord with English 1 50 Law. 2 75 3–5 100 A more comprehensive summary of the con- 6–10 435 tract terms is provided in Appendix B. The Lease Contract Framework 19 4.2 The Restructuring of DAWASA and its 4.3 Financial Projections Development Contract Financial modeling carried out in prepara- When the private Operator was engaged, DA- tion for the transaction identified conditions WASA signed a Development Contract with the necessary to ensure the financial viability of Ministry of Finance which (inter alia): DAWASA and the Operator and the two enti- ties’ financing requirements. Tariffs had been zz Specified the capital works program and increased substantially to the equivalent of provided for its financing; US$ 0.43 per cubic meter in 2002. A further zz Specified DAWASA’s responsibility for real increase of six percent was planned for planning, financing and implementing year two of the contract. This would allow for investments, and monitoring and report- a substantial increase in the Lessor Tariff but ing on its own performance and that of no increase in the Operator Tariff. No further the Operator; real tariff increases were foreseen until after zz Provided procedures for the retrench- the Five-Year Review. However, it was assumed ment of staff not hired by the Operator; that annual adjustments to reflect inflation and zz Committed GOT to ensuring that DA- exchange rate movements would maintain the WASA was adequately funded at all real value of tariffs over time. times and specified that the Lessor Tariff should allow DAWASA to recover its op- The model assumed that water losses would erating costs, debt service and its coun- be reduced from an estimated 70 percent to 33 terpart contributions to investment proj- percent within five years, and that, during the ects; same period, the customer base would be ex- zz Specified that the Customer Tariff would panded rapidly through the rehabilitation and be adjusted in accordance with the lease regularization of 132,500 existing connections contract, that adjustments to tariffs (legal and illegal) and the addition of 35,000 would be subject to the approval of the new connections. Uncertainty about the ac- Regulator, and that a major review of the curacy of some of DAWASA’s data, including tariffs, indexation formula and perfor- inconsistencies between available technical mance targets would be carried out in and commercial data, was considered a risk. year five of the Lease Contract; Because of the weakness of the existing data zz Committed GOT to budgeting for and on operational and commercial performance, paying its water and sewerage bills.20 many performance targets were not specified in advance and some that were specified, such In August 2003, when the Lease Contract as the targets for water losses, were subject to became effective, most of DAWASA’s staff was revision. These targets were to be established transferred to CWS. The current CEO of DA- on the basis of data to be collected by the Op- WASA (who had also been CEO of DAWASA’s erator during the first 12 to 18 months of the predecessor, NUWA) and all senior managers contract. remained in place. A staff of about 65 was re- tained to manage the planning and implemen- The reduction of transmission and distribu- tation of investment projects and to oversee the tion losses and expansion of the customer base Lease Contract with CWS. Since CWS hired all the staff not required by DAWASA, the latter did not have to implement the retrenchment provi- 20 United Republic of Tanzania, Development Contract Be- tween the Government of the United Republic of Tanzania, sions. DAWASA staff benefitted from a gener- Represented by the Ministry of Finance, and DAWASA, ous training budget funded by the DWSSP. 2003. 20 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM Figure 3  Flow of Funds within the Lease Contract Framework Loan Repayments WB, AfDB MOF & EIB Loan Proceeds Loan Loan Proceeds Repayments DAWASA Capital investments Repayments Subsidiary – Rental fee Loan Loan – Lessor tariff FTC Fee First Time Loan CWS BGT Equity FTC Tariff Connection Fund (Working capital) Customer Tariffs STM Customers Adapted from a diagram by TRC Global Management Solutions, Ltd., 2005 were considered essential for the operator’s investments.21 Both DAWASA and the Opera- success. Both depended on the timely comple- tor were expected to be able to begin servicing tion of rehabilitation works. Another critical their loans in year six. Figure 3 shows the flow factor was a significant improvement in billing of funds within the lease contract framework. and collection efficiency. This would require that the Operator rapidly develop an effective commercial system. Even assuming these im- 4.4 Arrangements for Serving the Poor provements progressed as planned, cash flow from operations was expected to be negative Three mechanisms were designed to improve in the first two years and very weak in year service for the poor: (i) CWS would construct three. Contributions of US$ 8.5 million to eq- 250 new water kiosks under the DCW; (ii) First uity by the operator’s shareholders and the line of credit from DAWASA would be required to keep the Operator solvent. As for DAWASA, its 21 Proceeds of the IDA, AfDB and EIB credits that were dis- viability depended on timely remittance of the bursed by the Ministry of Finance to DAWASA for DWSSP collected Lessor Tariff revenues and the Rent- were structured partly as grants and partly as loans to DA- WASA at an interest rate of 11.5 percent to be repaid over al Fee by the Operator, and disbursements of 15 years (IDA and AfDB) or 25 years (EIB) with five years loans and grants from GOT to finance DWSSP grace. The Lease Contract Framework 21 Time Connection Tariff revenues would be standpipe and kiosk operators would cover used to create a Fund to subsidize new connec- their costs, including the manager’s salary, by tions for domestic customers with up to three selling water to end-users for about Tshs 15 per water points on the property; and (iii) DAWA- 20-liter can in 2003/04, about 2.25 times the SA would implement a Community Water Sup- lifeline rate. ply and Sanitation Program (CWSSP) which would support the construction of up to 50 In addition DAWASA also financed the con- small water supply and sanitation schemes in struction of 50 small CWSSP schemes in com- mostly low-income communities not served by munities not served by the networks. These the networks. schemes were designed in partnership with beneficiary communities which were required The lease contract specified that, to the ex- to contribute five percent of the cost and form tent possible, standpipes and water kiosks water user associations (WUAs) to manage were to be located no more than 200 meters the schemes. The CWSSP, which was funded from the people to be served and that the Op- by the DWSSP, was to be implemented by DA- erator would be responsible for the manage- WASA with the involvement of specialized in- ment, maintenance and repair. Whenever pos- ternational non-governmental organizations sible, the Operator was expected to subcontract (NGOs) and their local partners. The NGOs standpipe and kiosk management to local man- helped each WUA to establish tariffs to be paid agers and communities. The lease contract also by customers of CWSSP schemes at levels ad- specified that standpipe and kiosk customers equate to cover operation, maintenance and would pay the domestic lifeline tariff, the same repairs. tariff domestic customers paid for the first five cubic meters.22 The term customers as used in 22 United Republic of Tanzania, Water and Sewerage Lease the lease contract was intended to mean the Contract between DAWASA and CWS Ltd., 19 February local kiosk operators. It was expected that the 2003, Articles 1 and 24 and Appendix K. 5 CWS’s Performance When CWS assumed operations in August subcontracted entirely to affiliates of Biwater 2003, it faced a very challenging situation. To and Gauff, leaving it with few of the short-term meet its targets, it would have to identify and benefits and little say in day-to-day manage- regularize thousands of unregistered connec- ment. STM had difficulty raising its equity tions, introduce a new billing system, reduce contribution and, even more troubling to the water losses from an estimated 70 percent to BGT transition team leader, STM’s expecta- 44 percent in the first three years, and increase tion of short-term gains was inconsistent with monthly collections from Tshs 720 million (the the financial profile of the undertaking. De- amount collected in August 2003) to Tshs 1.8 spite these mismatched expectations, the BGT billion within 12 months. Achieving these tar- transition team leader attempted to establish a gets would require a strong management team rapport with STM’s owners. However follow- and a change in culture on the part of the op- ing the end of the transition period, and the ap- erating staff and many of the customers. Such pointment of the new CEO, relations soured.23 changes in payment culture and collections STM later complained that the CWS manage- have been brought about successfully in other ment team was incompetent and refused to pay similar environments. They are more likely to the full amount of its equity share.24 occur when service is improving and the num- ber of connections is increasing. Unfortunate- The financial projections BGT submitted ly, CWS was unable to create these conditions. with its bid showed that CWS would lose Tshs In fact, the Operator failed to meet many of its 7.9 billion (about US$ 7.9 million) over the first obligations almost from the start and, by De- two years of operation, and would not reach a cember 2003, was already facing serious cash- break-even point until year seven. CWS would flow problems. require a significant injection of capital to en- able it to survive the first two years of oper- ating losses and the four more years it would 5.1 CWS’ Shareholders take to recoup those losses. That capital was to come from two sources: US$8.5 million in While neither the bid documents nor the equity from CWS’s shareholders, and the line MOU signed after the selection of BGT as pre- of credit from DAWASA which was conditioned ferred bidder specified Tanzanian sharehold- on the equity contributions. By the end of year ers, during the contract negotiations BGT was one, CWS’s shareholders should have contrib- encouraged to associate Super Doll Trailer uted US$ 5.5 million in equity and should have Manufacture (STM) as its local shareholder. The relationship between BGT and STM was 23 Interview with Cliff Stone, 22 November 2009. somewhat awkward from the beginning. STM 24 International Center for the Settlement of Investment had no experience in the water supply sector Disputes (ICSID), Case No. ARB/05/22, Biwater Gauff but nevertheless expected to have a share of (Tanzania) Ltd., Claimant v. United Republic of Tanza- nia, Respondent, Award, dispatched to the parties July 24, the contracts for the PWP and was surprised 2008, p. 45; and Larry Magor Testimony, Hearing Tran- to learn that the contracts had already been script, Day 1, pp. 149–150 24 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM borrowed US$ 4 million from DAWASA, giv- and expansion works managed by CWS (i.e., ing the company a total of US$ 9.5 million of the PWP and DCW programs) did not appear to debt and equity capital, but shareholders had be driven by operational objectives, reinforcing contributed only about US$ 3.1 million and the impression that Biwater and Gauff might had borrowed only US$ 2.2 million for a total have been more interested in the construction of US$ 5.3 million.25 As a result, and because contracts than in operations.26 One year into of its failure to achieve collection targets, the the contract, the government was clearly dis- company suffered a severe cash shortfall. illusioned and the Minister of Water publicly criticized the CEO on television. The leader of While the BGT transition team leader seemed BGT’s original transition team, an experienced to have a realistic appreciation of the long- manager, was brought back in as CEO in Sep- term nature of the project, it is not clear that tember 2004. Biwater’s marketing team had approached the project with the same commitment. It has been speculated that Biwater and Gauff’s marketing 5.3 Commercial Operations teams were more attracted by the PWP con- tracts and the possibility of winning the large The data base of customers that CWS inher- contract for the rehabilitation of the treatment ited included 115,000 registered accounts, works (which in fact they did not win). That but many were moribund or duplicates that is, they, like STM, were more interested in the had been set up to avoid payment of arrears. construction contracts than in the operational It was estimated that only 22,000 to 25,000 aspects of the lease contract and its long-term were active and potentially billable, and it was financial commitment. well-known that there were thousands of ille- gal connections that were not in the data base. Installation of a new billing software system 5.2 CWS’ Management that was completed in March 2005 turned out to be woefully inadequate, and subsequently The initial CWS management teams were not had to be replaced. CWS made limited prog- up to the challenge and there were three turn- ress cleaning up the customer database. It fo- overs in leadership within the first 13 months. cused efforts on regularizing and expanding In August 2003, a retired Biwater engineer was customers in an area that was hydraulically appointed to serve as interim CEO. The perma- separate, and therefore relatively controllable, nent CEO who arrived three months later had but the lease contract was terminated before worked for Biwater under contract but had nev- CWS was able to realize the benefits of these er served as a CEO. An independent observer efforts. In fact, CWS’ average monthly collec- pointed out that CWS was incredibly slow in tions in 2004/05 reached only 52 percent of getting started and management did not seem projections and were 21 percent lower than to approach the situation as a turn around proj- DAWASA’s had been in 2002/03. (See Table 2) ect. During the first year, the financial director Government arrears for water and sewerage failed to focus on the urgent need to increase services were a problem and by May 2005 cash flow and the commercial director did amounted to US$ 1.5 million, the equivalent not seem to have a strategy for dramatically improving billing and collection quickly. Rev- 25 ICSID, op. cit. pp. 37-38 and 44-45; and Minutes of the enues fell in the early months because there meeting of 20 August 2004 between the Interim Regulator were no performance incentives for commer- and CWS. (Amounts cited in Tanzanian shillings in the IC- SID Award have been converted to US dollars at the 2004 cial staff to step up their performance—or re- average exchange rate.) mit all the funds they collected. Rehabilitation 26 Interview with John Sitton, 11 September 2009. CWS’s Performance 25 Average Monthly Amounts Billed and Collected Table 2 –  (in millions of Tanzanian Shillings) DAWASA CWS 2002/03a 2003/04a 2004/05b Average amount billed 1,537 1,949 1,533 Average amount collected 1,193 1,075 939 Projected average amount to be collected 1,512 1,536 1,817 Ratio: amount collected/projected 0.79 0.70 0.52 a August–July b August–May Source: DAWASA of about 40 percent of CWS’ total billings.27 of the Rental Fee to DAWASA were irregular While CWS had an agreement with Govern- and, after June 2004, payments of the Rental ment that ensured payment, there were many Fee ceased altogether. In addition, beginning disputed bills (claims of no water, inaccurate in December 2003, CWS periodically withheld volume, etc.) and, because of meter malfunc- Lessor Tariff revenues to pay its own operating tion due to irregular supply and the preva- costs. This created a serious financial crisis for lence of meter vandalism, verification was DAWASA. In April 2004, DAWASA imposed problematic. a penalty for non-remittance of the Rental Fee—which CWS failed to pay. By June 2005, Given the historically poor coverage and CWS owed DAWASA Tshs 3.2 billion in Rental quality of piped water services, a thriving tank- Fees and Lessor Tariff revenues (about US$ 2.9 er industry existed. CWS considered the tanker million). Moreover, CWS made only one depos- industry a potential partner and took steps to it to the First Time New Domestic Water Sup- develop a constructive relationship by promot- ply Connection Fund—in October 2003—and ing self-regulation through membership in made no further deposits over the remaining a water tanker operators’ association and en- period of the contract.28 Instead it used First couraging tankers to move their operations to Time Connection receipts amounting to Tshs neighborhoods not slated for immediate piped 820 million (about US$ 726,000) to cover its water supply improvements. To promote the operational expenses. latter, CWS proposed to establish convenient water filling points for registered tankers. The Lack of data and reliability issues made it more reputable tanker operators were recep- difficult to assess CWS’ performance on key tive to this initiative but it had not yet devel- indicator targets set in the lease contract. oped to the implementation stage before the CWS submitted an Annual Report at the end lease contract was terminated. of its first year of operations and several re- ports in year two. GOT claimed that CWS failed to submit all required reports but it was 5.4 Compliance with the Terms of the Lease not possible to verify the facts on this point. Contract CWS failed to fulfill several of its contractual 27 ICSID, op.cit., p. 164. obligations. During the first year, payments 28 Ibid., p. 54. 26 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM Table 3 – CWS Performance on Key Indicators 2003/04 2005/06 Target Actual Target Actual Collection efficiency (collected as percent of billed) 60 55 70 61 Number of connections rehabilitated and metered 16,500 5,237 25,000 4,268 Source: DAWASA None of the reports could be obtained by the expanding connections. The plan was for it author of this study. Besides collection effi- to be completed in the first 18 to 24 months ciency, the only other indicator for which data of the contract but it ultimately took much were available was for the number of connec- longer. Start-up was delayed for six months tions rehabilitated and metered. As Table 3 until January 2004 because CWS wanted to shows, CWS’ performance was well below the change the advance payment bond language target values. Moreover, none of the new con- that had been specified in the bid documents nections qualified as First Time Connections and because CWS’ bank was unable to receive (for low-income households) because the limi- the contract-specified currencies. Eventually, tation on the use of the First Time Connection a total of 104,000 revenue meters were pur- Fund for connections with no more than three chased out of an original target of 173,000. It taps within the premises made it difficult to was subsequently determined that the num- apply.29 ber procured was sufficient for the foresee- able future, given the delay in making new Baseline data for several other indicators connections. While a World Bank supervision should have been established no later than mission reported that satisfactory progress 18 months after Commencement, but by May was made on the rehabilitation of production 2005, CWS had not proposed values for any of works and transmission lines,32 another ob- the missing data. As a result, targets could not server pointed out that the failure to install be set nor could performance be monitored for bulk meters early in the contract period was important indicators such as water losses and an inexplicable strategic mistake that made it the average number of hours customers re- impossible to reduce water losses and increase ceived water supply services.30 collections.33 CWS began to develop an asset recording One of the objectives of the DCW program system and a detailed hydraulic model of the was to lay new mains, eliminate spaghetti distribution network, but the asset register was connections and install new connections not complete and the model had not been vali- so as to reduce water losses and improve dated prior to termination of the contract.31 It was subsequently completed by DAWASA. 29 DAWASA, Dar es Salaam Water Supply and Sanitation Project, Mid-Term Review Report, 15 December 2006, pp. 5.5 Implementation of the Priority Works 15, 20–21, 48–61. 30 Ibid., p. 15. Program and Delegated Capital Works 31 Ibid., p. 61. 32 DWSSP Joint Supervision Mission, Nov. 7–9, 2005, Aide Memoire, p. 3; and DWSSP Aide Memoire and Mid-Term The PWP was designed to improve technical Review Report, Nov. 2006, p. 4. operating conditions and create the basis for 33 Interview with John Sitton, 11 September 2009. CWS’s Performance 27 collections.34 Another objective was to extend experience in managing changes in the be- the distribution system into unserved areas, havior and attitudes of staff. When they tried including some middle- and high-income to introduce a new culture of responsibility, areas that would provide reliable revenues. independence and punctuality, some employ- Gauff Ingenieure was responsible for the ees adapted, but most did not. Language was a design, procurement and supervision of the problem and the Tanzanian human resources program. Disagreements between Gauff and director often had to serve as translator. Prob- DAWASA over technical standards resulted lems were compounded by the fact that BGT in delays: Gauff favored steel pipes while DA- had passed up a critical opportunity to improve WASA favored PVC pipes produced in Tanza- productivity when, in an effort to strength- nia. There were allegations of vested interests en its bid, it committed to employ and retain on both sides. throughout the contract almost all existing operational staff—even though the bid docu- Consultants hired by DAWASA found that, ments and draft contract would have allowed although the volume and quality of works com- for significant layoffs. BGT planned to use the pleted was consistent with expectations, CWS excess staff for the DCW but, in fact, the excess did not appear to have a strategic approach to employees were never transferred to the DCW the DCW program and the works that were sub-contractors. Consequently, at the end of completed were not accompanied by comple- May 2005, there were 1320 employees, result- mentary actions that were necessary to bring ing in a very high ratio of between 20 to 50 staff about the expected outcomes.34 It was also ob- per 1000 connections.37 Saddled with the costs served that the construction contracts were of excessive staff that did not have the required managed separately from operations by affili- skills, CWS lacked the flexibility to hire more ates of Biwater and Gauff. As a result construc- qualified staff. Perhaps the biggest failure was tion decisions seemed to favor the components that CWS gave staff few incentives to improve that were most attractive to the contractors, their performance and identify with CWS. The without regard to the impact of the works on former employees of DAWASA did not in fact commercial performance.35 For example, pipes ever transfer their allegiance from DAWASA to were extended to districts where there was lit- CWS. tle or no water supply and, though new mains were laid, little was done to remove and replace old spaghetti connections. However, the CEO 5.7 Termination of the Lease Contract of CWS at the time of termination pointed out that this was due to delays in commissioning Early in the contract CWS was in serious trou- (final inspection and approval by the supervis- ble. Monthly collections were much lower than ing engineer) of the pipes, rather than lack of a projected and operational costs were much strategy. When CWS’ contract terminated, none higher. By March 2005, accumulated losses of the DCW pipes had been commissioned, and illegal customers could not be regularized until 34 Spaghetti connections are those with connection pipes the pipes were commissioned.36 longer than 30 meters, sometimes directly into transmis- sion mains. Efficient operation requires shorter connec- tions into secondary or tertiary mains. 5.6 Impacts on Staff and Productivity DAWASA, Mid-Term Review, op. cit., p. 26. 35 36 Interview with John Sitton, 11September 2009. 37 Interview with Cliff Stone, 22 November 2009. The staff of DAWASA had little preparation 38 Estimates of the number of connections varied, making it impossible to specify the ratio precisely. CWS had about for the transition to the private sector and the 60,000 accounts in its data base in 2005, but only about early CWS management teams seemed to lack 25,000 were being billed. 28 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM amounted to about US$ 12.3 million. During be called for less than its full value. In March the pre-bid process, Biwater had asserted that 2005, a mediator was engaged to assist the the data presented in the draft bidding docu- parties to work out a compromise. While GOT ments was acceptable, but, in meetings with and DAWASA were willing to accommodate MOW during June–August 2004, CWS ar- most of CWS’s proposals for revision of the gued that the collection projections had been contract, they were not willing to extend the based on inaccurate information, collecting lease contract for an additional five years un- from government agencies was more problem- less CWS’s collection performance improved. atic than expected, and a drought had affected CWS insisted on the extension but refused to water production. It requested an adjustment commit to higher collection targets, and a com- to the economic terms of the lease contract. In promise could not be reached. At that point, or response, GOT promised to make government even before, the Expert Panel foreseen in the agencies pay their bills but, otherwise, DAWA- Lease Contract should have been engaged but it SA argued that CWS shareholders should con- was not—in part because CWS thought it could tribute equity to cover its cash requirements. rely on the Minister to resolve the matter. After CWS requested an Interim Review of the On 12 May 2005, DAWASA called the perfor- Operator’s Tariff in September 2004, an Inde- mance bond in its entirety and was paid Tshs pendent Assessor appointed by the Minister 5.5 billion (about US$ 5 million) by the guar- (in his role as Interim Regulator) determined anteeing bank. After CWS failed to reinstate whether a material change in circumstances the performance guarantee as required under that would provide grounds for a revision of the Lease Contract, DAWASA served a notice the Operator’s Tariff in accordance with the of termination of the contract on May 25, 2005. lease contract had occurred. The auditor found The Lease Contract provided for a 30-day no- there were inadequate data to demonstrate a tice of termination during which period the material change in circumstances as defined in Operator was required to cooperate with DA- the lease contract. WASA to ensure a smooth transition and con- tinued operation of the services. The transition In late 2004, the Chairman of Biwater lobbied period would have expired on 24 June. DAWA- MOW to renegotiate the contract. While DA- SA, fearing that CWS was financially unable WASA argued that only tariff disputes should to continue to operate and that interruption of be brought to the Minister in his capacity as In- services was imminent, requested an early ter- terim Regulator, the Minister saw the problems mination and asked CWS to cooperate with a with the Lease Contract as a political problem speedy transition. CWS claimed that the notice for him and did not separate his political and of termination was invalid and would not agree regulatory roles as he should have. In January to an early termination.39 Though the CEO of 2005, CWS formally requested renegotiation. DAWASA was in favor of continuing to negoti- Its requests included: an increase in the Op- ate the termination date with CWS, the Minister erator’s Tariff, reduction in the equity require- of Water, in the face of declining public support ments, reduction in the Rental Fee, revision of for PPP and impending elections in November Performance Targets, and the freedom to re- 2005, ended the negotiations. On 1 June, three duce staff. In addition it later requested an ex- tension of the term of the Lease Contract for an additional five years. In February 2005, when 39 The sequence of events presented here is as reported in DAWASA tried to call the performance bond the report of UNCITRAL Arbitration Case No. UN6761, for the amount of unpaid penalties, Lessor Tar- City Water Services Ltd., Claimant/Counterrespondent v. Dar es Salaam Water and Sewerage Authority, Respon- iff revenues and Rental Fees, it was surprised dent/Counterclaimant, Award, 31 December 2007, pp. to learn that the performance bond could not 3–8; and in John Sitton (2006), op. cit.. CWS’s Performance 29 British managers were deported—an act which CWS shareholders were contentious and the Biwater labeled an “unlawful… repudiatory shareholders had not even met since early breach of contract.� 40 A few days earlier GOT in the company’s existence. With the encour- had quickly created a new public operating com- agement of the World Bank team leader, CWS pany, the Dar es Salaam Water and Sewerage shareholders agreed to meet and space for the Corporation (DAWASCO), which assumed man- meeting was provided at the World Bank’s Lon- agement of the services that same day. don office. Since the DWSSP was conditioned on the engagement of a private operator, the Bank DAWASA’s relations with CWS were strained had a strong interest in sustaining the contract by the operator’s financial problems and its fail- but, after the contract was terminated, the World ure to pay the Rental Fee and Lessor Tariff rev- Bank decided that, in the interest of ensuring the enues and comply with other contractual com- continuation of services, it would not suspend mitments. This prevented the development of disbursements and would work with the new the cooperative partnership that was essential DAWASA-DAWASCO arrangement. for success. Some observers thought DAWASA should have responded to CWS’s failures earlier in the contract and, in particular, required the 5.9 Results of International Arbitration shareholders to make their equity contributions and appoint a more effective management team. Two international arbitration cases ensued, It is not clear how DAWASA could have done so one under the aegis of the UN Commission on other than by calling the performance bond, International Trade Law and the other in the which it attempted to do in February 2004, only International Center for the Settlement of In- six months into the contract. An advisor to DA- vestment Disputes. WASA thought that, had the parties been able to come to an agreement, CWS’ performance, United Nations Commission on International which had begun to improve, would have con- Trade Law tinued to improve and the contract could have been completed successfully, but that political After negotiations over the revision of the lease pressures prevented an agreement. contract broke down in early May 2005, CWS filed a Notice of Arbitration under the rules of United Nations Commission on International 5.8 The Role of the World Bank and Trade Law (UNCITRAL), as provided for in International Partners During the Crisis the contract. Following the subsequent early termination of the contract and deportation of During the crisis, CWS shareholders asked the CWS’s managers, CWS failed to meet the dead- World Bank to intervene on their behalf with line for filing a statement of its claim and filed GOT. However, the World Bank, AfDB and for liquidation. In the meantime, DAWASA EIB avoided taking sides and played a neutral filed a counterclaim against CWS for damag- supportive role. The World Bank team leader es, including: overdue Lessor Tariff revenues, based in Dar es Salaam worked to promote Rental Fees, unpaid First Time Connection communication among all the parties. DWSSP Tariff funds, principal and interest for the line funds were made available to pay for the In- of credit sub-loan, penalties for non-compli- dependent Assessor and for the mediator. The ance and stocks and amounts due for operating Minister of Water visited World Bank head- quarters in Washington to discuss the issues and potential solutions. A contributing factor 40 Financial Times, 29 June, 2005, as cited in Privatization to the crisis was the fact that relations among Impact Assessment, op.cit., p. 32. 30 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM equipment DAWASA had sold to CWS at con- of CWS’s senior managers and seizure of the tract commencement. These were partially off- company’s assets constituted a breach of the set by the value of the Performance Guarantee Agreement between the U.K. and the Republic which DAWASA had called and amounts owed of Tanzania for the Promotion and Protection to CWS by DAWASA. At the end of an arbitra- of Investments of 7 January 1994. BGT argued tion process that lasted two years, the Arbitra- that it should receive compensation of about tion Tribunal ruled that DAWASA was justified US$ 20 million. In July 2008, the Arbitra- in terminating the lease contract and ordered tion Tribunal found that, although GOT had CWS to pay DAWASA damages of Tshs 6,990 committed an expropriation, its violations of million (about US$ 5.6 million) and one half the treaty did not cause the losses and dam- the cost of the arbitration proceedings. 41 But as ages for which BGT claimed compensation, of June 2011 the ruling in its favor had been a and no damages were awarded. Each party pyrrhic victory; DAWASA had been unable to was required to pay its share of the cost of the collect a single shilling from the bankrupt op- case. One member of the Tribunal filed a Con- erator. CWS is still in liquidation and the court- curring and Dissenting Opinion arguing that appointed liquidator has been notified of DA- even though the violations of the treaty per se WASA’s award. did not cause BGT’s losses, BGT should have received some compensation on account of the International Center for the Settlement treaty violations.42 of Investment Disputes In August 2005, BGT initiated a case against the United Republic of Tanzania at the Inter- 41 UNCITRAL Arbitration Case No. UN6761, City Water Services Ltd., Claimant/Counterrespondent v. Dar es Sa- national Center for the Settlement of Invest- laam Water and Sewerage Authority, Respondent/Coun- ment Disputes (ICSID). BGT claimed that terclaimant, Award, 31 December 2007. the termination of the contract, deportation 42 ICSID, op. cit. 6 The DAWASA-DAWASCO Arrangement 6.1 DAWASCO on 12 December 2005. The terms of DAWAS- CO’s contract were almost identical to those of Governance the contract with CWS. Although the contract provided for a performance bond, none was re- DAWASCO was created by an Order signed by quired. Likewise, although DAWASA’s author- the President of the Republic on 17 May 2005, ity to impose financial penalties for non-com- gazetted on 20 May 2005, under the Public pliance with performance targets was asserted, Corporations Act of 1992. Under the terms of in practice DAWASA never imposed them. The the Order, DAWASCO’s Chairman, other Board Lease Contract gave DAWASA the authority members, and the CEO are appointed by the to terminate the Lease Contract under certain Minister responsible for Water. Its share capi- conditions, including non-compliance, though tal was set at Tshs 2 billion (about US$ 1.8 mil- it was unlikely DAWASA would exercise this lion) to be held entirely by the Tanzanian Trea- power except with the approval of the Minister. sury. The Order states that DAWASCO shall have exclusive right…to provide public water The lack of reliable base data continued to supply and sewerage service in the Designated make it impossible to specify all targets. How- Area. The Order makes no mention of a lease ever, those that were specified were considered contract to be celebrated with DAWASA. to be realistic. DAWASCO was expected to in- crease average monthly collections from Tshs On 9 November 2005, in his capacity as In- 946 million to Tshs 1.7 billion in year one and terim Regulator, the Minister of Water and to Tshs 2.3 billion in year two.43 These targets Livestock Development granted DAWASCO a assumed gradual improvements in reducing license for a period of ten years, which states, inter alia, that the licensee is required to com- ply with the terms of the lease contract with Table 4 – DAWASCO’s Monthly Rental Fee DAWASA and that the license may be revoked Contract Year Rental Fee in Tshs millions if the licensee is in material breach of the lease 1 50 contract, or following termination of the lease contract. The license also reiterated the author- 2 75 ity of EWURA to assume regulatory authority 3–5 100 once EWURA became operational. 6–10 500 Terms of DAWASCO’s Lease Contract with DAWASA 43 It is significant that, instead of the collection ratio (i.e., DAWASCO took over operations on 1 June the ratio of collection to billing values) that was specified as the measure of CWS’ collection efficiency, the total value 2005. A ten-year lease contract with DAWASA of collected revenue was specified as a performance target was signed on 2 September 2005 and revised for DAWASCO. 32 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM water losses, annual tariff indexations and a was an insufficient number of qualified middle real increase in tariffs in 2005. The schedule of level managers and supervisors. By 2009, DA- Rental Fees is shown in Table 4. WASCO had taken a number of steps to im- prove the morale and performance of staff. An GOT contributed Tshs 2 billion to equity in organizational restructuring resulted in the re- 2006/2007. To assist with the cash shortfalls trenchment of 432 employees in January 2007, expected in the early years of the contract, DA- another minor retrenchment later in the year, WASCO had access to a line of credit of up to and the recruitment of qualified middle level US$ 6 million from DAWASA using funds pro- managers. In June 2009, the total number of vided under DWSSP. employees was 885; the staffing ratio was 6 per 1000 registered accounts, or 12 per 1000 active A key difference between DAWASCO’s con- connections, and the level of competence had tract and that of CWS was that initially DA- improved. 44 Management expected the ratio WASA assumed responsibility for the DCW would continue to improve as more connec- program. However in early 2008 it was de- tions were added or regularized, and as staff cided that coordination could best be ensured competencies and skills were strengthened. by transferring responsibility for planning and execution of DCW to DAWASCO, though DA- New salary scales based on job level and WASA remained responsible for procurement. performance rather than seniority were es- tablished in 2007 and individual performance Management and Staff contracts were put in place for commercial staff. Such changes met with some resistance While the reform of public utilities is usually but were eventually accepted. The Director of accompanied by a change in top management, Human Resources felt that having a Tanzanian this was not the case in Dar es Salaam. As was CEO who spoke Swahili helped to promote bet- the case in DAWASA, most of the senior man- ter communications than had been possible in agers appointed to DAWASCO in 2005 were all CWS. In July 2008, four area managers agreed veterans of DAWASA (prior to the lease con- to internal delegated management contracts tract with CWS) and most of them remained in under which they could earn performance bo- their positions at the end of 2009. The CEO had nuses that could be shared with staff. By July focused attention on three high priority perfor- 2009 all 12 area managers had such contracts. mance areas: production, reduction of water losses and customer relations. Team work was Unfortunately, turnover among the most emphasized and performance incentives were qualified staff was frequent as salaries were put in place for managers as well as staff. Staff not competitive with the private sector. In morale appeared to have improved relative to addition, DAWASCO’s resources for training the CWS era and a sense of cohesion was taking and technical assistance were very limited. hold. Although management’s expressed objec- The DWSSP did not originally include funds tives were sound, management control and ex- for training operational staff and managers, ecution did not always match expectations. as that would have been the responsibility of the private operator, but DAWASCO’s tight In 2005, DAWASCO took over a staff of 1320 – many of whom lacked appropriate skills and qualifications and/or had not yet adopted 44 This ratio was obtained using DAWASCO staff only; the ratios are about 6.5/1000 and 13/1000 respectively when a new work culture. In fact, little had changed 65 DAWASA employees are included. Efficient water utili- while the private operator was in place. While ties in Africa typically employ about 6 persons per 1000 the overall number of staff was too high, there connections. The DAWASA-DAWASCO Arrangement 33 cash-flow situation did not allow adequate The Operational Rescue Plan was intended funds for human resources development. It to enable DAWASCO to get off to a strong start seemed unlikely that much needed improve- under very unfavorable conditions. Its focus ments in efficiency and customer relations was justifiably on actions that could be taken would be realized without an intensive training within a short period and on creating a new program. In early 2009, funds from DWSSP corporate culture. Meeting the longer-term were reallocated to fund DAWASCO’s training performance targets specified in the Lease program. By the end of 2009, more than half Contract would prove to be more elusive, in the staff had received instruction in a variety part because DAWASCO was unable to sustain of topics, including preventive maintenance, the high level of effort over the longer term, but laying sewers, audit techniques, customer care, also because some of the targets could not be supervision skills and technical operations achieved until the fundamental weaknesses of skills In addition, the Millennium Challenge the system were addressed—for example, the Corporation (MCC) Program (funding by the lack of an accurate customer registry and a USA) also began to provide funding for capac- map of the assets. Removing those constraints ity building in late 2009. took longer than hoped. In 2007, DAWASCO requested that several targets be revised to Operational and Commercial Performance  reflect reality. Since the targets had been set as DWSSP objectives as well as in the Lease With the assistance of NWSC External Servic- Contract, changing them would have required es-Uganda (the consulting affiliate of Uganda’s a revision of both, but no action was taken by publicly owned National Water and Sewer- either DAWASA or the DWSSP financing agen- age Corporation), DAWASCO prepared and cies. As a result, the targets became somewhat executed an Operational Rescue Plan during irrelevant. In addition, the collection of reliable July through September 2005. The program data for some indicators was a challenge. Data established short-term quantitative targets for provided by DAWASA and DAWASCO did not revenue collection, leakage control, customer always agree and it was impossible to deter- care, water production and response to sewer- mine which, if either, was reliable. age blockages. There were also qualitative ob- jectives for improving staff commitment and DAWASCO’s operational performance with managerial autonomy, measuring and using respect to several indicators in 2009/10 is performance data, and improving the cleanli- shown in Table 5. Performance on potable wa- ness of the premises. To create incentives for ter quality standards was very good. The qual- improving performance, results were moni- ity of wastewater effluent remained well below tored and tabulated for each of DAWASCO’s ten standards until 2010 when rehabilitation of the service areas and staff of the areas with best ponds was completed. The reliability of the po- overall performance received cash rewards at table water treatment works gradually improved the end of the program. The Final Performance due to the completion of rehabilitation works, Evaluation Report of the Plan claimed that it and in 2009/10 capacity utilization of the two had been remarkably successful with regard to large treatment works averaged 99 percent. The the achievement of most, though not all targets. most serious problem remained water losses. In addition, a number of important constraints The target for water losses by the end of year were identified: an unreliable billing system, five (June 2010) was 33 percent. On the basis lack of customer location maps, inadequate network rationalization, poorly performing wa- ter treatment units, lack of skilled staff in some DAWASCO, Operational Rescue Plan, Final Performance 45 areas, delays in procurement of inputs, etc.45 Evaluation Report (July–September 2005). 34 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM Table 5 – DAWASCO’s Performance in 2009/10 Indicator Targeta Actual Percentage of water samples meeting standards 100 97 Percentage of effluent samples meeting standards 95 83 Treatment plant output as percent of capacity 95 99 Non-revenue water (not billed as % of production) b 33 53 Collection Efficiency (collected as % of billed) n.s. 79 Cumulative number of new and rehabilitated connections 166,475 112,329 Number of registered water accountsc 124,600 161,902 Number of active water accounts (bills sent) n.s. 97,290 Number of registered sewerage accounts d 14,000 20,000 Number of active sewerage accounts (bills sent) n.s 17,590 Percentage of metered customers billed on meter readings 95 79 Percentage of customers receiving 24-hour water supply 70 30 a Targets are those in the Lease Contract and, in some cases, vary from those for the DWSSP. b The target reflects separate targets set for transmission and distribution. c The number of registered water accounts includes duplicate accounts, large numbers of suppressed accounts and cus- tomers in dry zones. The 2010 target in Appendix N of the Lease Contract (used here) is 124,600 but, in the Technical and Financial Assumptions table of the Lease Contract, the value assumed for June, 2010 is 139,500. d The number of registered sewerage accounts includes duplicate accounts. There are no inactive sewerage accounts. Source: DAWASA of production and billing data (the latter based duplicates and many were moribund or in dry in part on estimated consumption because of zones. In fact, 97,290 water customers were the absence of meters) total non-revenue water being billed and counted as active. By Decem- amounted to 53 percent in 2009/10. Long delays ber 2010, that number had risen to 110,000. in the installation of bulk and district meters Although its collection ratio was 79 percent in compromised DAWASCO’s ability to measure 2009/10, DAWASCO billed customers for only water losses and reduce non-revenue water. The 47 percent of the water it produced. In 2009/10 completion of bulk metering, a district metering collected revenues were 58 percent of the tar- and water loss reduction program, replacement get. Only 30 percent of customers received wa- of key pumps, and the completion of the meter- ter 24 hours per day, compared to a target of 70 ing and rehabilitation of connections were ex- percent set for the DWSSP project. pected to improve this situation. Decentralization of area management in In the first four years of the contract DAWAS - 2007 and 2008 had brought significant im- CO fell far short of targets for new and reha- provements in performance. Training funds bilitated connections. In fact, the targets which were made available to DAWASCO from the were cumulative had become irrelevant and DWSSP in 2009, and 60 staff members of the should have been revised. It began to catch up Commercial Department received training in 2010. In May 2010, the commercial depart- in customer care. Energetic and creative area ment was still trying to clean up the customer managers promoted team spirit among their data base: though there were 161,902 regis- staff, improved response time to customer com- tered connections in the data base, some were plaints and inquiries. It was reported that theft The DAWASA-DAWASCO Arrangement 35 of collected funds had been virtually eliminat- the new Lease Contract with DAWASCO. In ed through better oversight and the introduc- addition, it signed a Memorandum of Under- tion of a new electronic billing and collection standing (MOU) with the Ministry of Water system. In 2009 the Commercial Department and Irrigation that, among other provisions, and area managers introduced text message established specific performance standards reminders and cell phone bill payments. These and required DAWASA to prepare an annual and other innovations by area managers result- report including data on its performance. ed in a noticeable improvement in collections. Oversight of DAWASCO Comparison of the Performance of CWS and DAWASCO The collection of performance data for this case study revealed that central direction and coor- There are only a few performance indicators for dination of data reporting was lacking. Manag- which CWS’ performance and that of DAWAS- ers often were unable to retrieve internal and CO can be compared and the comparison must published reports and there did not appear be limited to the first two years of the contracts to be a strategic or integrated approach to its since CWS operated for only two years. Both verification and analysis for the purpose of im- operators failed to reach the targets set in the proving overall performance. Data differed not DWSSP and their lease contracts. The number only from one report to another, but sometimes of connections that were rehabilitated by the two there were inconsistencies even within the was very similar. For both operators, the delay same document. There appeared to be no sin- in installing bulk, zone and customer metering gle audited set of operational and commercial made it impossible to measure and reduce water data that was shared by DAWASA, DAWASCO losses. In both cases, collections fell far short of and EWURA. The dismal performance of CWS projections though DAWASCO’s performance with regard to data collection and verification exceeded that of CWS and improved steadily explains the poor quality of information in the while CWS’s actually declined in year two. In early years after DAWASCO took over (2005– fact, DAWASCO encountered many of the same 2006), but four years later there was still a lack challenges CWS had faced in its first two years of adequate management control over key per- but, while CWS’s shareholders were unwilling to formance data. To some extent this could be sustain further losses, DAWASCO, as a state en- attributed to the inauguration of new informa- terprise, had a more indulgent shareholder that tion systems and the lag in the installation and was willing to inject subsidies. In addition, there debugging of data management systems in all was no performance bond that DAWASA could three entities, but it was clear that more atten- call and no financial penalties. Its performance tion needed to be focused on coordination and gradually began to improve in response to pres- verification, and the key actors needed to avoid sure from EWURA, the Ministry and the inter- issuing reports with inconsistent data. national lending agencies. The lack of a functioning information man- agement system contributed to DAWASA’s in- 6.2 DAWASA ability to monitor the Operator and enforce the Lease Contract. In a lease contract arrange- DAWASA’s Development Contract and ment is it should be quite clear that the lessor Memorandum of Understanding is the principal and the lessee is the agent, ac- countable to the principal. Unfortunately, their In 2005, DAWASA’s Development Contract governance arrangements resulted in ambigu- with Ministry of Finance was updated to reflect ity regarding the principal-agent relationship. 36 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM Average Monthly Amounts Billed and Collected by DAWASCO Table 6 –  (in millions of Tanzanian Shillings) 2005/06a 2006/07a 2007/08a 2008/09a 2009/10a Average amount billed 2,007 2,322 2,099 2,051 2,763 Average amount collected 1,217 1,525 1,675 1,745 2,180 Projected avg. amount to be collected 1,708 2,299 3,180 3,709 3,750 Ratio: amount collected/projected 0.71 0.66 0.53 0.47 0.58 a July–June Source: DAWASA The fact that both entities were state enter- of these attributable to CWS and others to DA- prises; both CEOs were appointed by the Min- WASA itself. ister of Water; and each had access to political channels undermined the authority of DAWA- The PWP, which included urgent rehabilita- SA over DAWASCO. In the absence of a perfor- tion of production works and the procurement mance bond or other financial incentives, there of bulk meters was originally supposed to be were no mechanisms to enforce performance implemented by CWS but, as mentioned earlier, targets. The development of an effective system startup was delayed several months. After the for monitoring the performance targets of the departure of CWS (which caused further de- Lease Contract lagged. DAWASCO was required lays) DAWASA took over the subcontracts for to report performance data to DAWASA which the incomplete components. PWP was declared in turn forwarded it to EWURA, but there ap- complete in March 2007. DCW, which included peared to be little follow-up except during the the rehabilitation and extension of the second- tariff review process. Dynamic collaboration ary and tertiary network, installation of bulk between DAWASA and DAWASCO to develop meters, replacement of old connections and in- effective plans for improving operational per- stallation of new connections - originally con- formance was not evident. As we shall see, the ceived as a five-year program to be completed only entity that was able to take a strong stance by June 2008—was completed in June 2010. with regard to the performance of both DAWA- SA and DAWASCO was the regulator, EWURA. The non-delegated works, for which DAWASA In an effort to correct this, the Lease Contract had implementation responsibility, included the was amended in late 2010 but the DAWASA Act rehabilitation of the water production/treatment and its Regulations also required revisions and, units, the transmission lines, reservoirs and the as of January 2011, they had not been amended. primary distribution network; construction of a new reservoir; and rehabilitation, replace- Implementation of Capital Investments ment and expansion of sewerage and wastewa- funded under DWSSP ter components. Delays of up to two years were experienced by several components but all were Overall, DAWASA performed well in imple- successfully completed in 2009. The replace- menting the capital works funded by DWSSP ment of raw and treated water pumps and ra- for which it was responsible over the period tionalization of high voltage power supply at the 2003–2010 but there were several delays in treatment plants, both of which were added to completion which had negative impacts on op- DWSSP during the course of implementation, erational and commercial performance—some were completed in 2010. The CWSSP was also The DAWASA-DAWASCO Arrangement 37  AWASCO’s Core Recurrent Revenues and Expenses Table 7 – D (in Tanzanian Shillings)   FY06/07 FY07/08 FY08/09 Core Recurrent Revenue (Tshs) a 17,440,731,140 17,011,716,288 18,498,894,577 Revenue Collected (Tshs) b 12,397,978,730 15,003,487,780 15,403,646,629 Core Recurrent Expenses (Tshs) c 20,503,364,432 18,448,714,647 22,210,081,095 Core Recurrent Expenses as % of Revenue collected 165% 123% 144% a Revenue from Operator’s Tariff component only, billed prior to provisioning for bad debt b Operator’s Tariff component only c Excluding non-cash expenses such as bad debt and depreciation Source: DAWASCO’s Audited Financial Statements delayed and was completed in late 2010. (Imple- continued to improve in 2009/10 but were still mentation of CWSSP is discussed in section 8.3.) far short of what was needed to cover costs. To allow for completion of the project, the clos- ing date for DWSSP was extended three times, As Table 7 shows, in every year since DAWASCO first until December 2009, then until June 2010 began operating, collected Operator Tariff reve- and finally until November 2010. nues were inadequate to cover its core recurrent expenses. This was despite a temporary decrease in expenses that resulted from staff reductions 6.3 Financial Performance and Viability of in 2007/2008. To cover the gap between expens- the Arrangement es and income, GOT, which had already contrib- uted Tshs 2 billion in equity, provided subsidies DAWASA’s and DAWASCO’s financial perfor- of Tshs 7.3 billion over 2006/2007–2008/2009. mance was undermined by several factors, In addition, DAWASCO borrowed funds from most prominently the latter’s failure to achieve DAWASA over and above the amounts allowed billing and collection targets. Table 6 shows for in its line of credit. that while DAWASCO’s collection ratio im- proved steadily, the amount billed remained Although mechanisms were put in place to stagnant, making it impossible to achieve tar- ensure that DAWASCO paid the Lessor Tar- gets. A three-year delay in the approval of the iff revenues and the Rental Fee to DAWASA annual tariff indexation by EWURA was also on time, the inadequate growth in collections a significant factor in the revenue shortfall. In had a significant negative effect on DAWASA. 2009 DAWASA and DAWASCO undertook a Table 8 shows that, although DAWASA was Joint Financial Recovery Plan but the results able to cover its core recurrent expenses with were disappointing—billed amounts actually its share of revenues from water and sewerage fell slightly (due to the cleaning up of the cus- services, it was unable to fully fund other obli- tomer data base and elimination of duplicate gations. Under the terms of DWSSP, DAWASA accounts) and the increase in amount collect- was required to contribute a minimum of ten ed was insignificant. In 2008/09, while DA- percent of the cost of capital investments. Be- WASCO collected 85 percent more than CWS cause of the shortfall in its tariff revenues, it had collected in 2004/05, it achieved only 47 relied on cash transfers from GOT and these percent of the collection target. Collections were not always paid in a timely manner. 38 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM  AWASA’s Core Recurrent Revenues and Expenses Table 8 – D (in Tanzanian Shillings)   FY06/07 FY07/08 FY08/09 Core Recurrent Revenue (Tshs)a, b 8,247,000,000 7,203,000,000 6,722,548,200 Revenue Collected (Tshs) a 4,327,057,012 4,556,998,831 4,633,920,927 Core Recurrent Expenses (Tshs) c 3,825,000,000 3,976,000,000 4,555,130,091 Core Recurrent Expenses as % of Revenue collected 88% 87% 98% a Lessor Tariff Component plus Rental Fees b Revenue billed prior to provisioning for bad debt plus Rental Fees c Excluding rescheduled interest payments due to GOT and contribution to CAPEX Source: DAWASA’s audited financial statements DAWASA’s core recurrent revenue actu- amounts on-lent to the two operators. There ally decreased over the three-year period seemed to be little chance of fully recouping the 2006/2007 to 2008/2009. The main reason sub-loan to CWS and, given DAWASCO’s poor for this was that in 2006/2007, the tariff struc- financial performance, it would not be able ture was revised and the percentage of the to repay its sub-loan for some time. To allevi- Customer Tariff allocated to the Lessor Tariff ate the burden on DAWASA, the international decreased from 24 percent to 21 percent. The lenders had suggested that the sub-loans be decision to restructure the tariff was based on converted to equity and debt be rescheduled. a projected significant increase in revenues, GOT preferred not to take any action with re- which—had it materialized—would have pro- gard to CWS’ sub-loan, as it still hoped to col- vided adequate revenue for both DAWASA and lect the damages awarded by UNCITRAL. The DAWASCO. Another factor was that, prior to Ministry of Finance agreed to postpone initial 2007/2008, DAWASA had received the excess repayments for two years until July 2010. In collected revenues that resulted from the dif- June 2010, DAWASA again requested that the ference between the historically higher Non- CWS sub-loan be converted to equity. Still un- Domestic Tariff and the Customer Tariff. In able to service its debt, it also requested that 2003 the difference had been significant but the interest be reduced to five percent and that over time the difference diminished as the debt repayments be rescheduled again until Customer Tariff increased. Although the Rent- July 2013. These results were disappointing al Fee was rising during this period, it was not since the achievement of financial viability was sufficient to offset the decrease in DAWASA’s one of the major objectives of the restructuring tariff revenues. of the services and of the DWSSP. While revenues were declining, DAWASA’s A tariff study that was initiated in 2009 ex- core recurrent expenses increased. In addition, amined the need and justification for tariff in July 2008 it was expected to begin servic- increases and for restructuring the sharing of ing its debt for that part of capital investment tariff revenues, but it was clear that tariff in- funding that had been structured as a loan creases alone would not be sufficient or justifi- from GOT. The outstanding sub-loans to CWS able. Cost control and improvements in opera- and to DAWASCO created a looming burden as tional efficiency would have to be part of the DAWASA would also have to repay GOT for the strategy to improve financial performance. 7 Regulation 7.1 The Interim Regulator regional water and sewerage authorities, 101 dis- trict and small town water and sewerage author- The Minister of Water served as Interim Reg- ities and several bulk water suppliers. In 2010, it ulator during the entire period of the Lease began to take an active role in the regulation of Contract with CWS and the first year of DA- the retail prices charged by kiosk operators and WASCO’s Lease Contract. During that period in the registration of private water tankers. its key regulatory actions included the ap- proval of the tariff indexation for 2004/2005 EWURA Guidelines issued in June 2009 and the decision to defer the approval of the specify rules and procedures for tariff appli- tariff indexation for 2005/2006 until EWURA cations and a formula for calculating revenue began to function. Though the Lease Contract requirements. There is no provision for routine gave the Regulator a role in resolving disputes annual indexation of tariffs, the implication over tariffs, the Minister’s involvement in the being that indexation is treated like any other renegotiation and termination of the contract tariff revision and is subject to the same appli- with CWS went beyond its regulatory mandate. cation process. The absence of detailed rules Had DAWASA and CWS called upon the Expert for calculating tariffs allows regulated entities Panel as foreseen in the contract, they might some latitude in proposing tariffs, and EWU- have been able to reach a more satisfactory out- RA enjoys substantial discretionary power in come. Instead, rather than invoking the Expert determining whether to approve the proposals. Panel, CWS appealed to the Minister, appar- ently hoping for a political settlement. 7.3 EWURA’s Tariff Decisions 7.2 EWURA’s Regulatory Authority While the annual indexation of the tariff which was foreseen under the terms of the Lease Con- EWURA began to regulate energy and water tract was implemented as expected in 2004/05 utilities in mid-2006. Its main role with re- when the Minister was Interim Regulator, it gard to water supply and sewerage services is was not routinely permitted after EWURA be- to approve—or disallow—changes in tariffs and gan to function. EWURA’s approach to annual approve codes governing utility-customer re- indexation of tariffs was controversial because lations. It also educates and consults with cus- it was considered by many to be at odds with tomers and resolves customer complaints not the terms of DAWASA’s Development Contract resolved by the utilities through its Consumer and DAWASCO’s Lease Contract. The latter Consultative Council. EWURA considers ser- clearly distinguishes between annual adjust- vice quality and performance in making tariff ments to reflect changes in price and foreign decisions and may require utilities to improve exchange indices, and periodic revisions of tar- quality and performance as a condition of tariff iffs to reflect changes in economic or technical revisions. By 2009, EWURA was regulating 20 operating conditions. Appendix K specifies the indexation formulas for the three components 40 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM of the tariff and Article 35.2 specifies that the favor of EWURA. Since EWURA did not recog- indexation formula shall be applied to the Op- nize the concept of routine annual indexation erator Tariff on the commencement date and of tariffs, each annual request for indexation annually thereafter. Article 40 outlines the took on the characteristics of a full tariff re- conditions under which a review of the tar- view in which the operator’s performance be- iff may be undertaken to take into account came a determining factor. In preparing its ap- changes in economic and technical conditions plication for indexation for 2009/10, DAWASA and verified base data. It also specifies that any was diligent in following EWURA’s Guidelines review of the Operator Tariff is without preju- and in addressing the inadequacies in its own dice to the Operator’s right to require the ap- and DAWASCO’s performance. EWURA ap- plication of the indexation formula adjustment. proved the indexation in July 2009. This re- This seems to imply regulatory discretion in sulted in a 30 percent overall increase in the approving real changes in the tariff but very customer tariff that took into account the price little, if any, discretion in approving the annual indices for previous three years. In its decision, indexation—other than to verify that the for- EWURA required DAWASA and DAWASCO to mulas have been applied correctly. increase the number of metered connections to 104,000, distribute Customer Codes of Prac- In June 2005, DAWASA submitted a request tice, develop a plan for reducing unaccounted- to the Interim Regulator for the annual tariff for water and initiate a pro-active program of indexation for the year 2005/06 and, in Janu- customer outreach. As the discussion of DA- ary 2006, applied for a real increase in the tariff WASCO’s performance in Section 6 showed, that was foreseen under the terms of DAWAS- these requirements represented significant but CO’s Lease Contract. The Minister, anticipat- achievable targets. It was agreed that a more ing that the newly established EWURA would complete tariff study would be undertaken as a commence functioning earlier than it actually basis for the 2010/11 tariff proposal. 46 did, deferred a decision on both. A new appli- cation for both the indexation and the tariff in- The advantage of annual indexation is that crease was submitted to EWURA in May 2006 it allows tariffs to be maintained at their real and was approved in July for 2006/07. For the value without sudden large increases that can years 2007/08 and 2008/09, EWURA refused be politically difficult. Because tariffs were not DAWASA’s request for tariff indexation, citing indexed during 2007–2010 their real value ac- the poor performance of the services. In ad- tually decreased but, while the 30 percent in- dition, in its 2008/09 application, DAWASA crease only restored customer tariffs to their had failed to submit an updated basis for cal- real 2006 level, it was likely to be perceived culating flat tariffs for unmetered customers. by customers as a very large increase. Such a In fact, DAWASA and DAWASCO were unable jump in the tariffs could have had a negative to produce a reliable basis for estimating con- effect on demand and collections but that did sumption by unmetered customers because of not in fact appear to happen—perhaps because the lack of sufficient bulk and zone metering. demand still outstripped DAWASCO’s ability to deliver water to customers and tariffs were After the 2008/09 decision, DAWASA and still far lower than those of alternative sup- its international lenders argued that an index- pliers. Appendix A, Table A1 shows the evolu- ation was not, strictly speaking, an increase tion of nominal and real tariffs over the period in the tariff and was intended to be automatic 2002/03–2009/10. under the terms of the Lease Contract. On this basis DAWASA appealed its case in Tanzania’s Administrative Court, but the latter ruled in 46 EWURA Order No. 09-008, 10 July 2009. 8 Coverage and Access of the Poor to Service 8.1 Results of the NBS Impact Surveys: Figure 4  H  ousehold Access to Water from Coverage and Availability Different Sources The National Bureau of Statistics carried out 3% a baseline survey of the DAWASA service area 5% in February 2006 and a follow-up survey in November 2009. Data were also available from an NBS Household survey in 2002. The stated objectives of the 2006 and 2009 surveys were to determine the sources of water supply used by the population, the locations where they ob- 38% tained the water, the reliability of supply, quan- 54% tities of water used, prices paid, existence of sewerage and willingness and ability to pay. 47 (The surveys distinguished between sourc- es, e.g., piped, borehole, well, spring, surface source, etc., and locations where water was obtained, e.g., own house or yard, neighbor’s house or yard, kiosk, institution, or vendor). Piped (DAWASCO) Despite the stated objectives, the survey report Private or Community Borehole did not present information on quantities of Protected Well or Spring water used by any of the households, or prices Other paid for water from kiosks or other alternative Source: Tanzania, National Bureau of Statistics, Follow- suppliers. Up Survey on Impact Assessment of the DWSSP, 2009. With regard to source, the results showed that, in 2009, 54 percent of households in the service area had access to piped water sup- not within reach of the network. Borehole- ply from DAWASCO as their main source of based community schemes were initiated in drinking water. This percentage represented a the late 1990s and were scaled up through the significant decrease from 73 percent in 2002, CWSSP. and 60 percent in 2006 and was far below the DWSSP end-of-project target of 80 percent. 47 National Bureau of Statistics, Baseline Survey on Impact The decrease was due to the combined effect of Assessment of the Dar es Salaam Water Supply and Sani- population growth, inadequate growth in the tation Project, Report, December, 2008; and Follow-Up number of connections and a deliberate policy Survey on Impact Assessment of the Dar es Salaam Water Supply and Sanitation Project, 2009, Final Report, May on the part of DAWASA to introduce borehole- 2010. (Some of the figures presented here have been de- based water supply for communities that were rived on the basis of data presented in the survey reports.) 42 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM 2006 and in 2009 are presented in Appendix Figure 5   Household Access to Water by Source A, Tables A2(a) and A2(b). and Location, 2009 1% With regard to the location where piped wa- 1% 1% ter was accessed, in 2009 about 17 percent of households had a house connection or yard tap 8% 5% (up from 11 percent in 2006). About 26 per- 8% 9% cent of households reported they bought piped water from neighbors (down from 35 percent 5% in 2006). Another eight percent bought piped water from vendors (down slightly from 10 percent in 2006). Less than two percent of households got DAWASCO piped water from a 26% kiosk (double the 2006 share of less than one 23% percent) and about one percent got DAWASCO piped water from other locations. Regarding the reliability and continuity of 8% supply, in 2009, about 67 percent of house- 2% 2% holds that had in-house connections reported 1% that water flowed daily. On days when water DAWASCO – house connection was available, 80 percent of households with DAWASCO – yard connection in-house connections reported that it flowed DAWASCO – neighbor's connection for six hours or more. This represented a con- DAWASCO – kiosk siderable improvement since 2006. However, DAWASCO – vendor frequency and duration of supply for yard taps DAWASCO – other Borehole – house or yard was considerably lower than for house connec- Borehole – neighbor's tions. Only 37 percent of households with yard Borehole – kiosk taps reported daily availability and, when wa- Borehole – other ter was available, only 59 percent of households Protected well with yard taps got water for six hours or more. Unprotected well, spring, etc. (See Appendix A, Table A3) Bottled Unknown Source: Tanzania, National Bureau of Statistics, Follow- 8.2 Kiosks Up Survey on Impact Assessment of the DWSSP, 2009. In November 2010 DAWASA reported that 294 Most of the remaining 46 percent of house- new kiosks had been built with DWSSP funds holds that did not use piped water as their (other than those built under the CWSSP, which primary source of drinking water got water is discussed below) and all were receiving wa- from boreholes or protected wells. Access to ter—though some depended on tanker supply. improved water as defined by the World Bank Assuming an average of 500 people was served Water and Sanitation Sector Board is relatively high at about 83 percent.48 48 Improved water includes water from piped service, bore- holes and protected wells, which is available in the house Detailed data on the sources and locations or yard, or from a kiosk, but not from mobile vendors. It of water used by the surveyed households in excludes bottled water. Coverage and Access of the Poor to Service 43 by each kiosk, about 124,500 people could have difference. At 2.4 times the lifeline rate, this been served by these kiosks, but it was impos- was in line with the original expectations of sible to confirm this number or to assess the project designers. However, when water was quality of service at the kiosks because no ben- not available from DAWASCO, some kiosk op- eficiary survey was conducted in 2010. In No- erators resorted to buying tanker water which vember 2009 DAWASA had reported that 169 cost considerably more and had to be resold at kiosks were functioning (potentially serving a considerably higher rate than water bought about 84,500 people) but the NBS survey con- from DAWASCO. In mid-2010, EWURA is- ducted that month found that only 1.6 percent sued an order limiting the price of water sold of households (the equivalent of about 54,000 at DAWASCO kiosks to Tshs 20 per 20 liters persons) reported using DAWASCO-supplied (1.6 times the lifeline rate). Of this, DAWASCO kiosks as their primary source of drinking wa- could retain Tshs 12.74 (i.e., the lifeline rate) ter.49 Another 4.5 percent of households (about and the remaining Tshs 7.26 would be used 152,000 people) used kiosks that sold water to construct and maintain kiosks. It appeared from boreholes which could have been either that DAWASCO would be required to pay kiosk privately-owned or part of CWSSP schemes. For operators separately from its other revenues. a quarter of of kiosk customers, water was avail- able for up to 9 hours per day. For 60 percent While the number of functioning kiosks had it was available for 10 to 14 hours per day. The increased and the availability of water was ex- NBS survey found that 40 percent of households pected to continue to improve, it was recog- using kiosks reported that there was no con- nized that increasing the number of well-man- gestion or that it seldom occurred; 49 percent aged kiosks and sustaining good service would reported that congestion occurred sometimes require stepped up efforts. In 2010 DAWASCO and 11 percent reported that it was common. established a Pro-Poor Unit that, among its The survey report did not distinguish between other duties, would be responsible for the over- DAWASCO-supplied kiosks and other kiosks as sight of kiosk services. regards number of hours water was available.50 A WaterAid-funded study carried out in 2009 8.3 First Time Connection Fund suggested that finding appropriate sites for ki- osks was difficult and, as a result, a large per- Like CWS, DAWASCO found that the crite- centage of kiosks were located on private prop- ria for using the First Time Connection Fund erty. In such cases, the owner of the property made it difficult to apply. The criteria were was usually appointed as manager—a situation which the study suggested might make it dif- ficult to replace a poorly performing manager. 49 A WaterAid survey conducted in March–July 2008 found that, of 184 kiosks that had been built to date, only 79 were The study also reported that long lines tended receiving water from DAWASCO, and of those only 26 were to form at the kiosks that had the most reliable operating. John Mfungo and Ben Taylor, Mapping Public supply.51 Water Kiosks in Dar es Salaam, WaterAid Tanzania, Sep- tember 2008. 50 National Bureau of Statistics, 2010, op. cit. Tables 4.2 Until 2010, the price of water sold at DA- and 4.4.5. WASCO kiosks was not regulated. DAWASCO 51 Paula Tibandebage and Festo Maro, Strengthening In- clusion in Investments in Urban Water and Sanitation sold water to the kiosk operators at the lifeline Services: A Case Study of the DWSSP, February 2010, pp. rate (the equivalent of Tshs 12.74 per cubic 23–27. (unpublished background study for the WaterAid- meter at the 2009/10 tariff). Kiosk operators funded study, Strengthening pro-poor targeting of invest- ments by African utilities in urban water and sanitation typically sold it for about Tshs 30 per 20 liters, – the role of the International Development Association of covering their own management costs with the the World Bank, July 2010.) 44 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM relaxed in 2007 to allow all connections in schemes were based on boreholes, seven de- specified (primarily low-income) wards in Dar pended on bulk water supply from DAWASCO es Salaam to qualify. However, for all practical and in a couple of cases start-up was delayed purposes, the fund was irrelevant during the because DAWASCO had not completed the bulk period covered by this study, as all new connec- water supply connections. Needless to say, such tions made under the DCW were free but not a situation disappointed the affected commu- targeted specifically at the poor. At the end of nities. In the large majority of cases, once the 2009, the Fund remained in a separate account schemes were completed, they functioned well. and in mid 2010, DAWASCO began to use it to However, the implementing NGOs reported connect eligible households in a bid to improve that on-going training and support would be commercial performance. needed for WUAs because of frequent turnover in the executives and staff. To meet this need, DAWASCO’s Pro-Poor Unit, once it became ful- 8.4 Community Water Supply and ly established, was expected to provide techni- Sanitation Program cal support to off-network community-based schemes, such as those constructed by CWSSP. The Community Water Supply and Sanitation Program (CWSSP) funded small water supply By December 2010, 50 small water supply and on-site sanitation facilities for communi- schemes were completed, and DAWASA re- ties within the DAWASA service area that were ported that all were operational. Ten sanita- not likely to be reached by DAWASA systems in tion facilities were constructed by WUAs with the near future. It was managed by DAWASA’s DWSSP funding, including public toilets for lo- Community Liaison Unit. CWSSP was the most cal markets and promotional EcoSan latrines. successful pro-poor component of the DWSSP, Once the schemes were completed, WUA’s op- though it too encountered a number of chal- erated and maintained them and charged us- lenges. Three specialized non-governmental ers for their use. CWSSP exceeded the origi- organizations (NGOs), Plan Tanzania, Water nal DWSSP estimate of 170,000 beneficiaries Aid and CARE International, were engaged by —by December 2010, DAWASA reported that DAWASA to mobilize low-income communities 275,000 people were served by CWSSP wa- and assist them to prepare grant requests and ter schemes. In addition to the new schemes develop the capacity to manage the schemes funded by DWSSP, 35 existing schemes were following construction. The targeted commu- rehabilitated by DAWASA and an additional nities (with populations ranging from about 237,000 residents were said to have benefited 2,000 to 20,000) were required to form Water from them. Users Associations (WUAs) and contribute 5 percent of the cost of the schemes. 8.5 Affordability of Service A number of challenges were encountered by the program. As is often the case, mobili- Unfortunately the NBS survey did not pro- zation took longer than planned. The NGOs vide data on the volume of water used and the reported that, particularly during election amount households paid for water from sourc- years, representatives of the opposition tried es other than the DAWASCO piped system. to block the program, making mobilization of Information on the prices of water from alter- communities more difficult. As a result of this native suppliers was reported by WaterAid.52 and delays caused by other factors, implemen- tation took five years instead of the projected three years. Though most of the water supply 52 Ibid. p. 30. Coverage and Access of the Poor to Service 45 Appendix A, Table A4 compares the monthly would have been able to stay within the afford- cost and affordability of a basic supply of 30 able limit of four percent of income if it con- liters per capita per day for a household of five sumed slightly less than half the basic mini- persons from various sources. For a household mum volume. The cost of water obtained from with two minimum wages, a basic supply of other suppliers was well above the four percent water from a DAWASCO connection was quite limit and well out of reach of the very poor. It is affordable and, even for a very poor household clear that a strategy to improve the affordabil- with an income of about three quarters of one ity of water should focus on increasing connec- minimum wage, cost less than four percent of tions and the number of conveniently located income. Unfortunately, as mentioned above, kiosks with bulk water from DAWASCO, and only about 17 percent of households had con- improving the availability of water at kiosks. nections and the 83 percent that did not have A significant increase in the number of con- connections paid considerably more than the nections and well-managed kiosks would not DAWASCO tariff. For those that bought from only give more households access to affordable DAWASCO kiosks, the basic volume would water supply, but would increase competition have been quite affordable for a household with with households that sell water. two minimum incomes. A very poor household 9 Conclusions and Lessons The failure of the Lease Contract with CWS per se was found to be justified, a more grace- and the delays in achieving the objectives of the ful departure of the private management would sector reform and the DWSSP were costly to all have been in everyone’s interest and in keeping concerned, most especially the government with Tanzania’s international trade treaty with and people of Tanzania, so it is appropriate to the U.K. As a result of this highly publicized return to the questions posed at the outset of and dramatic event and the failure of PPPs in this case study and draw some lessons from other sectors, Tanzania’s image as a country the events and outcomes that have been re- that supported private investment was at least viewed. The most relevant findings associated temporarily tarnished and reinforced the ero- with each question and the lessons that can be sion of public support for PPP. drawn are discussed below. In contrast to the deliberate preparation of the PPP, the events that followed the collapse 9.1 The Policy Environment of the PPP were not foreseen and suffered from inadequate policy preparation. The most nota- Was the policy environment appropriate ble problem was that following the replacement for successful private participation and the of the private operator with the public opera- subsequent public-public partnership? tor, DAWASCO, the principal-agent relation- ship between DAWASA and DAWASCO was The Tanzanian policy environment embodied undermined by the fact that both had the same in official policy statements, laws and political owner (GOT), a situation that interfered with actions provided a reasonably good foundation the channels of accountability that are normal- for the sector reform and for the restructuring ly created by a contract between two entities. of institutional arrangements. GOT generally While this was primarily an institutional issue, demonstrated a commitment to strengthening its resolution required policy action. the autonomy and financial viability of service providers, although delays in GOT’s payment Lessons: of subsidies and contributions to equity cre- ated problems for DAWASA. Policy makers zz Preparation of the policy environment did not interfere in the decisions of EWURA for sector reform and PPP should include despite pressure from international lenders. a strategy for dealing with the failure of (The separate question of whether an indepen- the PPP and plans for quickly putting in dent regulator is appropriate in the context of a place a new, if only temporary, arrange- lease contract is discussed below.) ment and for developing a longer-term solution. A few political actions complicated the situ- zz Sector reform efforts can be derailed or ation with CWS, the most notable being the undermined by a weak accountability expulsion of the British managers of CWS in framework—particularly when a public- 2003. While the termination of the contract public contractual relationship exists. 48 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM 9.2 Consultation with Customers sell, consumers are likely to devise cop- ing mechanisms (e.g. illegal connections) Was the design of the reform and the lease which undermine the utility. contract guided by adequate consultation with existing and potential customers and an assessment of willingness to pay? 9.3 Measures for Serving the Poor Surveys carried out prior to the transaction Were appropriate measures for serving the showed that, because of DAWASA’s dismal per- poor incorporated? formance, there was broad support for PPP, but no willingness-to-pay study was undertaken. Three mechanisms were designed to improve A 2002 survey of households reported that 73 services for low-income households and com- percent of the city’s population was using piped munities: the First Time Connection Fund, the water but DAWASA was not being paid for most construction of kiosks, and CWSSP. Of these of it and it was not clear whether or how much the most successful was the CWSSP, which households were paying. It was assumed that supported the construction of 50 small water willingness-to-pay and demand for lower-cost supply schemes (most based on boreholes) and piped water would be high among residents 10 sanitation facilities in communities that who were paying high rates for water from were unlikely to be reached by DAWASA sys- neighbors and vendors. For them the key re- tems in the near future. NGOs invested consid- quirement was that a connection be affordable erable time in mobilizing communities, orga- and that reliable services be available. How- nizing WUAs, creating management capacity ever, households and other vested interests and providing feedback to DAWASA. In fact, it that were stealing water could be expected to was found that more time was needed for these resist a more disciplined commercial approach. activities than had originally been allowed. DAWASCO has recently made a big push to in- Political factions that opposed the incumbent crease connections and improve reliability of government sometimes disrupted progress. supply but it remains to be seen whether it will It was also found that on-going support was pay off. Surveys carried out by the National Bu- needed after the water and sanitation schemes reau of Statistics in 2006 and 2009 to assess began to operate because of frequent turnover the impact of the DWSSP provided informa- in the executives and staff of the WUAs. Over- tion on access to service, but they did not throw all CWSSP was relatively successful. It pro- much light on attitudes and willingness-to-pay. vided useful lessons and DAWASA hoped to The fact that all new connections made under expand the program. the DCW were free also makes it difficult to as- sess effective demand. While 294 kiosks were constructed, few if any had water 24 hours per day and some depend- Lessons: ed on tanker-supplied water. Unfortunately no survey was conducted after the completion of zz A more strategic approach that linked the kiosks and the improvements in the avail- expansion and connection programs to ability of water that occurred in mid-2010 so it household preferences and willingness- is not possible to evaluate the the quality and to-pay might have paid off. availability of service. EWURA’s recent deci- zz Consultations should be conducted when sion regarding tariffs to be charged kiosk users it is certain that services can be pro- was an indicator of an increasing awareness of vided. If kiosks are established in an the need to expand affordable services to the environment where there is no water to poor. However, the significant percentage of Conclusions and Lessons Conclusions and Lessons 49 households (26 percent in 2009) that bought time for the mobilization of communities water from neighbors would not benefit from should be allowed and strategies to avoid this decision. In addition, this action left DA- or dispel the disruptive behavior of po- WASCO responsible for the cost of managing litical factions are needed. kiosks, and it remained to be seen how man- zz Completion of community-managed agement would be funded and whether it would schemes should be coordinated with be satisfactory. The first priority seemed to be complementary components, such as to ensure reliable water supply. electricity service or the supply of bulk water to those that depend on it. Alterna- The First Time Connection Fund proved dif- tively, communities should be informed ficult to use and was somewhat irrelevant ini- in advance that there may be a delay in tially because all new connections made under delivery of essential components. the DCW were free, though not necessarily tar- zz On-going training and support for geted at the poor. The collected funds remained WUAs after completion of the schemes available and could be used by DAWASCO’s is required to resolve technical prob- Pro-Poor Unit once it becomes fully functional. lems, strengthen skills and make up for frequent turnover in the executives Going forward, improvements in service for and staff of Water User Associations. the poor will depend on DAWASCO’s improv- In urban communities with high rates ing the availability of water generally and on of turnover and weak community cohe- the Pro-Poor Unit’s ability to mobilize resourc- sion other management models may be es and promote the effective management of needed. kiosks. Lessons: 9.4 Strategies for Dealing with Vested Interests zz Adequate attention is unlikely to be giv- en to establishing effectively managed Were vested interests identified and were kiosks and connecting poor households strategies developed to bring them onboard? during the turn-around of a water util- ity when management must focus on During preparation and implementation, more rehabilitation, reducing water losses attention could have been paid to motivating and overall commercial performance. changes in the behavior of vested interests and/ Creating a dedicated unit to manage the or bringing them onboard. Eventually the illicit CWSSP worked well in DAWASA and behavior of staff, the unregulated water tanker might improve DAWASCO’s attention to industry and large-scale theft of water had to be the poor. addressed. In lease contract arrangements, pri- zz Better consultation with low-income vate operators typically attach a high priority to households, evaluation of housing condi- culture change among staff. They usually create tions and an assessment of willingness to controls and performance incentives soon after pay might have resulted in the design of taking over operations, but CWS did not take a more practical and effective First Time effective action early in its contract. Once DA- Connection program. WASCO assumed operations, it reduced the ex- zz When introducing community-managed cess staff, introduced performance incentives schemes, NGOs can play an important and eventually installed commercial manage- role in ensuring adequate consultations ment systems that practically eliminated illicit and appropriate designs. Sufficient lead activities among staff. Other vested interests 50 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM received less attention until much later. In the develop light-handed or self-regulation last year of its tenure, CWS had begun to de- and constructive relationships is eventu- velop a constructive relationship with tank- ally necessary for both political and op- ers, but this initiative was not continued by erational reasons. DAWASCO. In 2010, in response to high level political pressure, EWURA, DAWASA and DA- WASCO were looking into ways of regulating 9.5 Selection Process and Consultation tankers and promoting a collaborative relation- with Bidders ship with them. The unexplained delay in the completion of bulk metering and initiation of Was the selection process well-managed, fair an effective water loss reduction program, par- and transparent? Were potential bidders ticularly along the transmission lines, allowed for the contract with a private operator illegal use of large volumes of water to thrive for consulted and their inputs taken into account many years. In 2009 DAWASCO began to make in designing the transaction? some progress in identifying and closing down large-scale offtakes. During preparation, potential bidders were consulted several times but the outcome was Lessons: not particularly beneficial. Perceptions of un- fair treatment were raised by all the bidders. zz Vested interests need to be assessed early Saur and Vivendi had the impression that in the reform. Explicit strategies should PSRC and GOT were responsive to one bidder be developed and supportive actions to while ignoring the concerns of others, while Bi- be taken by each of the key players (poli- water claimed that the World Bank favored the cy makers, asset holder, operator, regula- French. The latter allegation was damaging and tor, police, etc.) should be identified. contributed to the Divestiture Committee’s re- zz Proven experience in introducing cul- sistance to visit successful lease contract cases ture change and using performance in- in francophone Western Africa. Whether valid centives and other controls to eliminate or not, these allegations and the behaviors that illicit behaviors and improve efficiency prompted them undermined the quality, trans- should be a pre-requisite for managers parency and competitiveness of the process and (whether public or private) who are hired significantly raised the cost of the transaction. to turn around a non-performing water In hindsight, the contentious selection process supply and sewerage utility. seemed to presage the difficulties that followed. zz The completion of a comprehensive cus- PSRC’s failure to take seriously the requests for tomer survey and introduction of billing risk mitigation by two highly experienced po- and collection systems that enable man- tential bidders resulted in their not submitting agement to control rent-seeking behav- bids. As a result, in the final round only one bid iors are high-priority elements of a strat- was received – from the least experienced bid- egy to improve the performance of water der, BGT. Subsequent events showed that the utilities and cannot be neglected. other bidders’ concerns were valid. zz Identifying and eliminating large-scale illegal water use are also essential ele- BGT made several mistakes in its bid. Its bid ments of a turn-around strategy and re- was overly optimistic. It did not take advantage quire the cooperation of political author- of GOT’s offer to fund retrenchment of staff. Fi- ities, regulators, police and the utility. nally, the assumptions underlying its bid were zz Working with tankers and other private worse than those used by DAWASA, making its alternative water service providers to overly optimistic projections less credible and Conclusions and Lessons 51 also making it difficult for CWS to subsequent- bidders complained that the draft contract did ly justify a material change in circumstances. not protect the operator from risks over which it had little or no control. Another condition Lessons: that was neglected was the requirement that a lead professional partner own a predominant zz Transaction preparation teams should share (70 to 80 percent) of the operating com- avoid letting a potential bidder take pany and be able and willing to contribute the control of the process and meticulously necessary equity/working capital. The Lease avoid even the appearance of preferential Contract with CWS failed in part because the treatment of any of the bidders. lead professional partner did not seem to be zz Prequalification criteria should be ap- fully committed to the operational aspects of propriate to the challenge of the under- the contract and failed to appoint a manage- taking and, once adopted, should be en- ment team that was capable of dealing with the forced. operational challenges. A key mechanism for zz Transaction preparation teams should enforcing the contract and controlling DAWA- consult meaningfully with all qualified SA’s risks, the performance bond, proved inad- bidders and take their concerns into ac- equate because it could not be called in part. A count. key mechanism for resolving disputes, the Ex- zz The link between a bidder’s assumptions pert Panel, was not invoked because CWS and and a Material Change in Circumstances the Interim Regulator circumvented them. should be clearly stated. In the current public-public context, the Lease Contract is dysfunctional because the 9.6 Appropriateness of the Lease Contract asset holder is not able to hold the Operator ac- Model countable. DAWASCO was created as an emer- gency response to the collapse of the contract Was a lease contract, first with the private with CWS. In that context, there was little time operator and subsequently with the public to consider whether the DAWASA-DAWASCO operator, appropriate to the context? arrangement would be optimal. In fact it was considered to be a temporary arrangement by Many of the conditions that favor a successful many of the stakeholders. The lack of a prin- lease contract already existed in Dar es Salaam cipal-agent framework undermines the emer- in the late 1990s and others could have been in- gence of effective strategic corporate direction, corporated into the lease contract with the pri- management incentives and accountabil- vate operator. While some observers have ar- ity. This brings the current separation of asset gued that the lease contract should have been management from operations into question. preceded by contracts to clean up the customer At the very least, repair of the principal-agent data base and create an asset registry, others framework is needed. cite the successful implementation of lease contracts in similar situations where base data Lessons: was equally unreliable. One thing is clear: not all the pre-conditions listed in Section 3.2 of zz Preparation of a lease contract transac- this report were addressed adequately or sus- tion should be guided by careful consid- tained. The failure of the Lease Contract with eration of the conditions that have been CWS might have been avoided had the drafting associated with success elsewhere and of the lease contract and the selection process adequate attention to sustaining those been managed better. In particular, some of the conditions throughout the contract. 52 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM zz The lead professional partner’s lack of Operator’s shareholders would have to con- adequate prior experience and/or its tribute equity sufficient to keep the company failure to take a predominant share in solvent. In hindsight, it appeared that Biwa- the operating company may result in a ter and Gauff’s marketing personnel were not failure to mobilize the managerial and committed to the long-term operational objec- financial resources necessary to ensure tives of the contract but were more interested the success of a lease contract. in the short-term income from the PWP and zz A lease contract between two public en- the possibility of winning the large contract for tities is not likely to be successful un- rehabilitation of treatment works. less there is an effective principal-agent framework. Examples from other coun- Though the conditions that could justify a tries seem to indicate that public-public tariff review were specified, the methodol- arrangements have been successful ogy to be applied to determine the change in when the entities involved were owned tariffs was not specified. However, it was as- by different levels of government or sumed that annual indexations would be made where one was a subsidiary of the other. to reflect inflation and exchange rates and This subsidiary relationship should be the annual indexation and the formulas to be respected by political and regulatory au- used were also specified in the Lease Contract. thorities. Given the acknowledged weakness of the base- line data and the lack of a specified method for resetting tariffs to protect the financial 9.7 Financial Viability and Allocation of equilibrium, it was a fairly risky undertaking. Risks This was confirmed by the fact that two of the pre-qualified bidders found it too risky and did Were the financial and operational not bid. In addition, the financial soundness of projections and underlying assumptions BGT’s bid was questioned by the World Bank realistic and reliable? Were the tariff level, team. In hindsight, it is not possible to deter- sharing of revenues, financial terms and mine to what extent the failure of the Lease allocation of risks appropriate and did they Contract with CWS was due to misjudgments allow for both partners to achieve financial in the preparation process and to what extent equilibrium within a reasonable time? it was due to CWS’ management errors. Both undoubtedly played a role. The financial modeling carried out in prepara- tion for the PPP was based on best estimates Risks to the Lessor were also not adequately and existing data of questionable reliability. addressed. When CWS encountered financial In fact, because of the weakness of the exist- difficulties, it failed to remit Lessor Tariff reve- ing data on operational and commercial per- nues and to deposit First Time Connection Tar- formance, many performance targets were iffs into the fund for that purpose. A major flaw not specified in the Lease Contract and some in the arrangement was that the performance that were specified, such as the targets for guarantee could not be called in part. As a re- water losses, were subject to revision after re- sult, DAWASA had to wait until CWS owed it liable base data had been collected. Even as- the full amount of the guarantee. This delay in suming that significant improvements in op- enforcing accountability contributed to the col- erational and commercial performance would lapse of the contract. In addition, the fact that be achieved during the first five years of the BGT was a joint venture of Biwater and Gauff contract, cash flow from operations was ex- shielded the two parent companies from risk. pected to be negative in the early years and the Although a World Bank review of the draft Conclusions and Lessons 53 lease contract had recommended making the or political action that undermines the bidders’ parent companies responsible for the sustainability of a contract may could performance guarantee, that advice was not also be used to balance risks. heeded. Lessons: 9.8 Tariff Setting Framework zz When preparing a PPP on the basis of Was the tariff setting framework appropriate unreliable existing data, it is important for the lease contract framework? Did it not only to recognize the need for veri- provide predictability, transparency and fication and resetting of performance credibility? targets, but also to provide protection against the financial consequences of Private operators who seek predictability and the Operator’s not being able to achieve protection from political interference and reg- operational and commercial targets that ulatory failure are likely to be wary of regula- were based on faulty base data. tors with a high degree of discretion. Two of zz In situations where losses will be incurred the potential bidders for the lease contract in in the early years, the commitment of the Dar es Salaam were concerned that the new lead professional partner to the long-term regulator, EWURA, would have the authority objectives needs to be confirmed. to make tariff decisions that were inconsistent zz An alternative approach that is favored with the terms of the Lease Contract, thus ren- by the international private operators dering tariff-related clauses of the Lease Con- but not yet widely tested is to structure tract meaningless. Because of these concerns PPP in two phases with a less risky man- and others, those firms did not submit bids. In agement contract for the first 18 to 36 fact, EWURA, did not begin to function during months, during which time the base data the two-year period when CWS was the Opera- are established, and a transition to the tor. The Minister of Water, acting as Interim riskier lease contract thereafter. Regulator, honored the terms of CWS’s Lease zz To protect Lessor Tariff revenues from Contract. However, after DAWASCO became misappropriation, all revenues should be the Operator and EWURA assumed its func- deposited into a joint lessor/operator ac- tions, it did indeed make tariff decisions that count and allocated by a mechanism that seemed to be at odds with the intent of the protects each party’s rights. Lease Contract. EWURA argued, exactly as zz Special purpose revenues (such as the the two non-bidders had feared, that the Lease First Time Connection Tariff) should be Contract was part of a larger legal context in placed in a trust with clear disbursement which EWURA had the authority to disallow rules. tariff adjustments, even the annual indexation. zz A detail such as the inability to call a performance bond in part can result in EWURA claimed that its decisions motivat- delays and the worsening of a dispute to ed DAWASA and DAWASCO to improve their the point that it is irreconcilable and the performance, and there was some evidence cost of failure extremely high. that is true. It is a positive development when zz The feasibility of obtaining parent com- a national institution takes the lead in forcing pany guarantees for the performance of service providers to be accountable—despite an operator needs to be examined. complaints from influential institutions such zz Likewise, sovereign guarantees to pro- as the World Bank. This demonstrates that the tect the operator from regulatory failure institutional framework is maturing. On the 54 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM other hand, EWURA’s decisions revealed the operator’s being subsidiary to the asset inconsistency between the lease contract mod- holder. el and the regulatory model.53 This inconsisten- zz Other mechanisms to alleviate regula- cy is certain to be a barrier to attracting private tory risk can also be used, for example, operators for a future lease contract unless it a guarantee mechanism through which could be alleviated in some way. In fact, the ex- the operator would be compensated for istence of a regulatory body does not preclude any loss of revenue due to a regulatory a lease contract. There are independent regu- decision that is inconsistent with the lators in developing countries that function in terms of its contract. a manner that is consistent with regulation by contract. The Water Regulatory Council in Mo- zambique is an example. 9.9 System for Monitoring Operator’s Performance Since EWURA has broader authority to pro- mote competition and efficiency and promote Was a rigorous system for monitoring the the interests of consumers, it could get more performance of the operator established? involved in addressing vested interests that impede the effectiveness of the utilities. Evi- While CWS was required to submit a number of dence from elsewhere (e.g. tanker associations reports to DAWASA and the Regulator at speci- in Ghana) shows that certain vendors can be fied intervals, the reporting regime reportedly recognized as a legitimate and necessary com- broke down early in the contract. One of the ponent of service delivery and brought into the most serious lapses was CWS’ failure to deter- formal sector through contracts with the utility mine values for the baseline data. In addition, or light-handed regulation. DAWASA did not appear to have established a rigorous system for recording and verifying Lessons: performance data during the period of CWS’ contract and could not provide complete data zz The regulatory framework and the cho- for the purpose of this case study. In fact, the sen management model should be mutu- weakness of the reporting and data manage- ally consistent. Independent regulation ment systems persisted throughout the period can be consistent with the lease contract under review. Four years after the departure of model if the regulator recognizes the CWS, management control over the reporting lease contract as an essential element of and verification of key performance data was the regulatory framework and endorses lacking. This contributed to DAWASA’s inability the tariff adjustment and other regula- to hold DAWASCO accountable and to the over- tory provisions of the contract. Including all weakness of management and strategic plan- a pre-specified methodology for revising ning. However there was hope for improvement. tariffs that the regulator endorses might A shared system for reporting and monitoring also reduce the regulatory risks. data among EWURA, DAWASA and DAWASCO zz Many regulatory and PPP experts believe was supposed to become functional in 2010. that in a lease contract framework the regulator should set or approve the over- all Customer Tariff but not the Operator 53 It is noteworthy that those who designed the lease con- Tariff which should be set by agreement tract model seemed to think that the role of the regulator between the asset holding company would be to review and advise on the adjustment and re- visions of the tariffs. See World Bank, Proposed DWSSP (lessor) and the operator (lessee). This Preparation Mission, September–October, 1998, Aide is consistent with the principle of the Memoire, pp. 4, 8. Conclusions and Lessons 55 Improvements were needed in other con- of Water was acting as Interim Regulator with tributions to the database as well. The impact regard to Tariffs seemed to encourage this ap- surveys carried out by NBS on behalf of DA- proach and the Minister did not limit his in- WASA were not well-designed and the data tervention to tariff issues. This undermined collected were not presented and analyzed in the use of the conflict resolution mechanisms a meaningful way. Independent auditors fre- provided the Lease Contract. In fact, the Ex- quently reported that the operational data re- pert Panel should have been used for all dis- ported by the Operator were unreliable, but putes short of Arbitration (including disputes failed to provide more reliable estimates. In over the Operator Tariff). Selection of the panel short, the audits and surveys seemed to have prior to the effectiveness of the contract might been carried out in a perfunctory manner to have expedited the process. comply with laws or with international lenders’ requirements rather than to produce useful in- Unfortunately, the parties were not able to re- formation. solve their differences in a timely manner and the Operator’s performance worsened to the Lessons: point that, ultimately, political authorities be- came impatient and intervened, citing the need zz The performance of water and sewer- to ensure the provision of essential services. age services is not likely to improve in The disputes then were taken to two interna- the absence of rigorous data reporting tional arbitration bodies, UNCIT and ICSID. and management systems that are used Though these proceedings brought a certain to enforce performance standards and degree of closure, and minor moral victories guide management decisions. for both parties, they were extremely costly to zz Audits and impact surveys should be de- both sides and neither collected any monetary signed to produce meaningful data and damages. CWS was not awarded any and GOT analysis that can be used by managers has been unable to collect the amounts owed by and planners. While there are good argu- the bankrupt CWS. ments for allowing the national statistics entity to carry out impact surveys, a wa- Lessons: ter sector professional should be engaged to participate in the design of surveys zz To the extent practical, steps necessary and analysis of the data. to initiate a dispute-resolution mecha- nism, such as the selection of an Expert Panel, should be completed in advance of 9.10 Conflict Resolution Mechanisms contract effectiveness. zz Political and regulatory authorities Did the legal contracts include procedures should avoid getting involved in disputes and conditions necessary to ensure smooth and require parties to use the pre-speci- relationships and resolve disputes? fied dispute-resolution mechanisms. zz All parties have an interest in acting The Lease Contract with CWS contained sever- sooner rather than later to enforce ac- al mechanisms for resolving disputes, review- countability, invoke dispute resolution ing tariffs and performance targets and for the mechanisms and resolve disputes before termination of the contract. However, CWS they become irreconcilable and extreme- tended to ignore the mechanisms and, instead, ly costly. appealed to the Minister of Water in hopes of a political solution. The fact that the Minister 56 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM 9.11 The Role of the World Bank the DAWASA-DAWASCO institutional frame- work and the completion of DWSSP before the Could or should the World Bank have closing date. While significant attention was managed its support for the transaction given to tracking the financial performance of differently? DAWASA and DAWASCO and compliance with World Bank-mandated environmental and so- The role of the World Bank is to promote the cial safeguards, this case study showed that the success of the investment operations and as- measurement and verification of operational sociated transactions that it supports. It does performance indicators, and service quality this by (for example) working with implement- was somewhat neglected until close to the end ing agencies to design and execute cost-effec- of the DWSSP. Not only is this type of infor- tive and sustainable operations that promote mation essential for strategic management and economic development and benefit the poor, sustainability but, in addition, accurate data on ensuring that principles of fair competition access to services is necessary for tracking the and transparency are honored, and bringing achievement of the Millennium Development to bear its extensive experience to reduce the Goals. In recognition of this, the World Bank risks of failure. At the time the Dar es Salaam Water and Sanitation Sector Board recently transaction was being prepared, the policies focused more attention on establishing consis- of GOT’s international development partners tent definitions of indicators and on the collec- led by the World Bank strongly favored PPPs. tion and analysis of data on access and service Several successful cases, including those in quality. Western Africa, had created confidence that PPP could lead to significant improvements in Lessons the financial viability and quality of services everywhere. The Bank’s preparation team tried zz The World Bank should consider aban- to bring the experience of the successful cases doning an operation to support PPP to bear on the transaction and advised PSRC when the preparation process is marred to address certain operator risks but, in the by allegations of favoritism and contro- end, the team still had some misgivings about versial decisions such as the pre-quali- the viability of the Lease Contract with CWS. fication of bidders who do not meet the Despite these issues, World Bank no objections specified pre-qualification criteria. allowed the transaction to move forward. Since zz The current more balanced approach to that time, the World Bank’s enthusiasm for PPP promoting the efficiency of both private has evolved in light of experience. The failure of and public operators needs to be further several PPPs, examples of highly effective pub- refined by a better understanding of the lic operators and a growing appreciation for the conditions that favor positive outcomes role of small local service providers have led to for different types of institutional ar- the adoption of a more balanced policy of pro- rangements and more rigorous efforts to moting the efficiency and financial viability of ensure that these conditions are created. public operators while promoting a wide vari- zz Both international lenders and imple- ety of forms of PPP where feasible. menting agencies need to devote more sustained attention to monitoring and Because of the difficulties that arose, the re- verifying meaningful indicators of cov- sources of World Bank teams were absorbed by erage, affordability and service quality. the more highly visible matters such as the me- This includes monitoring the design and diation of the dispute between DAWASA and validity of beneficiary surveys and using CWS and, after that failed, the sustainability of the results as a guide in planning. List of Sources Government of the United Republic of DAWASA. Draft Institutional and Organisa- Tanzania Reports tional Arrangements Review, 2006. United Republic of Tanzania. Bank of Tanza- DAWASA. Draft Institutional Review of the nia. Economic and Operations Report for the Water Supply Sector in Dar es Salaam, Novem- Year Ended 30 June 2003. ber 2006. United Republic of Tanzania. Ministry of Wa- DAWASA. Dar es Salaam Water Supply and ter. National Water Sector Development Strat- Sanitation Project, Interim Technical Review egy, 2005–2015. Report, 31 May 2004. United Republic of Tanzania. Ministry of Wa- DAWASA. Dar es Salaam Water Supply and ter and Livestock Development. National Water Sanitation Project, Mid-Term Review Report, Policy - July 2002. 15 December 2006. United Republic of Tanzania. National Bureau DAWASA. Water Supply Improvement Plan. Dr. of Statistics. Baseline Survey on Impact As- Ahmet Abdel Warish Consulting Engineers. Fi- sessment of the Dar es Salaam Water Supply nal Report, June 2008. and Sanitation Project [2006]. Report dated December, 2008. DAWASCO, Annual Technical Reports for 2005/06, 2007/08, and 2008/09. United Republic of Tanzania. National Bureau of Statistics. Follow-up Survey on Impact As- DAWASCO Operational Rescue Plan and Final sessment of the Dar es Salaam Water Supply Performance Evaluation Report (July–Sept., and Sanitation Project, 2009. Draft Report 2005), October 2005. (Included in World Bank dated January, 2010. Aide Memoire of Nov. 2005) United Republic of Tanzania. National Bureau of Statistics. Tanzania in Figures, 2004, 2005, EWURA Documents 2006, 2007 and 2008. Annual Report for the year ended 30 June United Republic of Tanzania. Privatization Im- 2008 pact Assessment, Infrastructure, July 21, 2005. EWURA Guidelines for the Evaluation of Tariff Applications, 18 Oct. 2006. DAWASA and DAWASCO Reports EWURA Guidelines for Tariff Applications, DAWASA. Dar es Salaam Water Supply and June 2009. Sanitation Project Quarterly Progress Reports. 58 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM EWURA Order No. 08-001, 8 April 2008. Water and Sewerage Lease Contract between Dar es Salaam Water and Sewerage Authority EWURA Order No. 09-007, 6 April 2009. (The Lessor) and City Water Services (The Op- erator), 19 Feb. 2003. EWURA Order No. 09-008, 10 July 2009 Water and Sewerage Lease Contract between Dar es Salaam Water and Sewerage Author- Legal Documents, Contracts and Rulings ity (The Lessor) and Dar es Salaam Water and Sewerage Corporation (The Operator), 2 Sep- Dar es Salaam Water Supply and Sewerage Au- tember 2005. thority Act, No. 20 of 2001. The Energy and Water Utilities Regulatory Au- World Bank Project Documents thority Act, No. 11 of 2001. Dar es Salaam Water Supply and Sanitation International Center for Settlement of Invest- Project, Identification Mission, May 1998. ment Disputes, Case No ARB/05/22, Biwater Gauff (Tanzania) Ltd. v. United Republic of Dar es Salaam Water Supply and Sanitation Tanzania, Award, dispatched 24 July 2008. Project, Preparation Mission, Oct. 1998. International Center for Settlement of Invest- Dar es Salaam Water Supply and Sanitation ment Disputes, Case No ARB/05/22, Biwater Project, Project Appraisal Document; Report Gauff (Tanzania) Ltd. v. United Republic of No. 25249-TA, April 10, 2003. Tanzania, Concurring and Dissenting Opinion, Gary Born, 18 July 2008. Dar es Salaam Water Supply and Sanitation Project, Supervision Mission Aide Memoires, License granted under the Dar es Salaam Wa- June 2005, Nov. 2005, March 2007 (Mid-Term ter and Sewerage Authority Act, 2001, by the Review), Dec. 2007, May, 2008, Dec. 2008, Minister for Water and Livestock Development June 2009, Dec. 2009. to the Dar es Salaam Water and Sewerage Au- thority, July 2003. Locussol, Alain. Note to Quality Assurance Group (QAG) members, 2003. License granted under the Dar es Salaam Wa- ter and Sewerage Authority Act, 2001, by the World Development Indicators, 2008. Minister of Water and Livestock Development, Dar es Salaam Water and Sewerage Corpora- tion (DAWASCO), Nov. 2005. Other Reports Memorandum of Understanding [2008–2011] Approaches to Private Participation in Water between Ministry of Water and Irrigation and Services, A Toolkit. The World Bank, 2006. DAWASA, Jan. 2009. Brocklehurst, Clarissa, and Jan G. Janssens. UNCITRAL Arbitration Case No. 6761, City Wa- Innovative Contracts, Sound Relationships: ter Services Ltd. Claimant/Correspondent v. Dar Uran Water Sector Reform in Senegal. The es Salaam Water and Sewerage Authority Re- World Bank, Water Sector Board Discussion spondent/Counterclaimant, Award, 31 Dec. 2001. Paper No. 1, January 2004. List of Sources 59 de Waal, Dominic, and Brian Cooksey. Why pro-poor targeting of investments by African did City Water fail? The rise and fall of private utilities in urban water and sanitation – the sector participation in Dar es Salaam’s water role of the International Development Asso- supply. WaterAid, May 2008. ciation of the World Bank, Case studies from Burkina Faso, Ghana and Tanzania, by Peter Fall, Matar, Philippe Marin, Alain Locussol, Newborne and Josephone Tucker of the Water and Richard Verspyck. Reforming Urban Wa- Policy Programme at the Oversees Develop- ter Utilities in Western and Central Africa: ment Institute and Kate Bayless of the School Experiences with Public-Private Partner- of Oriental and African Studies of the Univer- ships. Vol. 1: Impact and Lessons Learned. The sity of London, July 2010.) World Bank, Water Sector Board Discussion Paper No. 13, June 2009. Triche, Thelma, Sixto Requena, and Mukami Kariuki. Engaging Local Private Operators in Halpern, Jonathan, Charles Kenny, Eric Dick- Waer Supply and Sanitation Services, Initial son, David Ehrhardt, and Chloe Oliver. Deter- Lessons from Emerging Experience in Cam- ring Corruption and Improving Governance in bodia, Colombia, Paraguay, The Philippines, the Urban Water Supply & Sanitation Sector, A and Uganda. Vol. 1: Overview of Experience. Sourcebook. Water Sector Board Water Working The World Bank, Water Sector Board Working Note No. 18, The World Bank, December 2008. Note No. 12, December 2006. Marin, Philippe. Public-Private Partnerships for Urban Water Utilities, A Review of Expe- Websites riences in Developing Countries. The World Bank and PPIAF, Trends and Policy Options EWURA: www.ewura.go.tz Series No. 8, 2009. DAWASA: www.dawasa.co.tz Marin, Philippe, Jean-Pierre Mas and Ian Palmer. Using a Private Operator to Establish DAWASCO: www.dawasco.com a Corporatized Public Water Utility, The Man- agement Contract for Johannesburg Water. United Nations Economic Commission for Af- The World Bank, Water Sector Board Water rica: www.uneca.org Working Note No. 20, June 2009. United Republic of Tanzania, National Bureau Tibandebage, Paula, and Festo Maro. Strength- of Statistics: www.nbs.go.tz ening Inclusion in Investments in Urban Wa- ter and Sanitation Services: A Case Study of United Republic of Tanzania, Ministry of Water the DWSSP. [unpublished], WaterAid Tanza- and Irrigation: www.maji.go.tz nia, February 2010. (Background study for the WaterAid-funded study, Strengthening The World Bank: www.worldbank.org Appendix A Detailed Tables on Tariffs, Access to Service and Affordability Table A1 - Water Supply and Sewerage Tariffs in Dar es Salaam, 2002/03 to 2009/10 Water Supply Tariff for Metered Connections (Tshs per cubic meter) 1, 2 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 Operator Tariff 322 337 359 359 488 488 488 637 Lessor Tariff 86 89 121 121 137 137 137 176 FTDWSCF Tariff 24 25 26 26 29 29 29 37 DOMESTIC TARIFF 322 337 359 359 488 488 488 637 (up to 5m3 and at kiosks) TOTAL DOMESTIC WATER TARIFF 432 451 506 506 654 654 654 850 (over 5m3) NON-DOMESTIC WATER TARIFF 725 725 725 725 725 725 725 850 Sewerage Tariff (Tshs per cubic meter)3 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 Operator Tariff 100 106 111 111 133 133 133 174 Lessor Tariff 29 30 36 36 41 41 41 53 TOTAL SEWERAGE TARIFF 129 136 147 147 174 174 174 227 Price Index and Exchange Rate with US Dollar 2003 2004 2005 2006 2007 2008 2009 2002–2009 4 CPI Inflation Rate 3.5 4.7 5.0 7.3 7.0 10.3 12.1 63.1 Exchange rate ofTshs/US$1.00 999 1089 1029 1251 1255 1178 1293 1 Tariff rates were obtained from DAWASA and from EVI/URA Order No. 09-008. 2 For non-metered connections the tariffs are applied to an assessed water consumption that varies from 20 m3 to 45 m3 per month depending on the service zone. 3 The sewerage tariff is applied to 80 percent of the water supply consumption. 4 United Republic of Tanzania, National Bureau of Statistics, Tanzania in Figures 2005 and Tanzania in Figures 2008; and Bank of Tanzania Monthly Economic Report  January 2010. Appendix A 63  ercentage of Population Using Water from Selected Sources and at Selected Locations in Table A2 (a) – P Dar es Salaam, 2006 Location of Access to Water In In own own Neighbor’s Mobile To1al using each Type of Water Source house yard house/yard Kiosk lnstitution vendor Other source Piped (DAWASCO 5.4 5.3 34.8 0.7 3.0 9.9 0.3 59.6 or Community) Borehole 0.6 1.8 14.2 5.1 1.7 2.7 0.1 26.2 Protected Well 0.1 1.5 3.4 1.0 0.5 0.6 0.0 7.1 Unprotected Well 0.0 0.3 1.2 0.0 0.0 0.0 0.1 1.6 Protected Spring 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 Unprotected 0.0 0.1 0.2 0.0 0.0 0.0 0.8 1.0 Spring Surface Source 0.0 0.0 0.0 0.0 0.0 0.0 0.3 0.3 Covered Rainwater 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.1 Tank Bottled Water 0.0 0.0 0.0 0.0 0.0 0.6 0.1 0.7 Unknown 0.1 0.0 0.0 0.0 0.0 3.4 0.0 3.4 Total with access 6.2 9.0 53.9 6.9 5.1 17.2 1.7 100.0 at each location Source: United Republic of Tanzania, National Bureau of Statistics, Baseline Survey on Impact Assessment of the DWSSP (2006), Report, December 2008. 64 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM Percentage of Population Using Water from Selected Sources and at Selected Locations in Table A2 (b) –  Dar es Salaam, 2009 Location of Access to Water Type of Water In own In own Neighbor’s Mobile Total using Source house yard house/yard Kiosk Institution vendor Other each source Piped (DA- 8.1 9.3 25.6 1.6 0.7 8.4 0.1 53.8 WASCO or Community) Borehole 0.7 1.5 23.0 4.5 3.4 3.0 1.3 37.3 Protected Well 0.0 0.4 2.0 0.6 1.4 0.4 0.3 5.1 Unprotected 0.0 0.0 0.5 0.0 0.1 0.2 0.3 1.1 Well Protected 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Spring Unprotected 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 Spring Surface Source 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 Covered Rain- 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 water Tank Uncovered 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Rainwater Tank Bottled Water 0.0 0.0 0.0 0.0 0.0 0.4 0.9 1.3 Unknown 0.0 0.0 0.0 0.3 0.0 0.9 0.1 1.2 Tolal with 8.8 11.3 51.1 6.9 5.7 13.2 3.2 100.0 access at each location Source: United Republic ofTanzania, National Bureau of Statistics, Survey on Impact Assessment of the DWSSP, 2009, Drafl: Report, January 2010. Appendix A 65 Table A3 - Reliability and Duration of Piped Water Supply Service in Dar es Salaam, 2006 and 2009 Percentage of Households With Each Type of Connection In House Connection Yard Tap Frequency of Water Availability 2006 2009 2006 2009 Daily 39.5 66.9 44.9 37.3 Every 2–3 days 32.6 9.3 24.4 21.3 More than once a week 1.2 10.2 10.3 14.7 Once a week 3.5 0.8 9.0 4.7 Atleastonce or twice a 1.2 0.0 2.6 5.3 month Unpredictable 20.9 11.9 9.0 16.7 Never 1.2 0.0 0.0 0.0 Total 100.0 100.0 100.0 100.0 In House Connection Yard Tap Duration of Water Flow 2006 2009 2006 2009 When Available 24 hours n.a. n.a. 32.5 32.0 6 or more hours 60.5 79.7 24.7 26.7 4–5 hours 11.6 5.1 10.4 14.0 Less than 4 hours 12.8 10.1 26.0 17.3 Unpredictable 15.1 5.1 6.5 10.0 Total 100.0 100.0 100.0 100.0 Source: NBS, DWSSP Impact Assessment Surveys, 2006 and 2009 66 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM  ypothetical Monthly Cost and Affordability of Basic Water Supply From Various Sources in Table A4 – H 2009/2010 (Assuring 30 Liters per Capita per Day by Household of Five Persons) DAWASCO DAWASCO Tanker-supplied Tanker Connection Kiosk Kiosk Neighbor Vendor (bulk) Monthly 2,914 6,862 22,875 11,438–30,875 57,187 16,013 Cost(Tshs) Monthly Cost as Percent of Two Minimum Wages 1.4 3.3 10.9 5.4–14.7 27.2 7.6 Monthly Cost as 3.9 9.2 30.5 15.3–41.2 76.2 21.4 Percent of Very Poor Household Income Percentage of 17.4 1.6 n.a. 51.1 13.2 n.a. Households that get water primarily from this location Notes and Sources: Prices for water bought from vendors, neighbors and tankers are indicative, as no comprehensive survey has been carried out Domestic water supply tariff of Tshs 637/m3 for the first 5 cubic meters. DAWASCO kiosk price of Tshs 30 per 20 liters, or Tshs 1,500/m3 (DAWASCO) Tanker-supplied kiosk price of Tshs 100 per 20 liters, or Tshs 5, 000/m3. (WaterAid, 2010) Water bought from neighbors price of Tshs 50 to 130 per 20 liters, or Tshs 2, 500 to 6, 500/m3. (Author’s estimate) Vendor water price of Tshs 250 per 20 liters, or Tshs 12,500/m3. (WaterAid, 2010) Tanker water price of Tshs 22,000 to 45,000/10m3. (WaterAid, 2010). An average of Tshs 35,000 was used. At 30 liters per day per capita, a household of 5 persons would consume 4.575 cubic meters of water per month. The minimum wage in 2008 was Tshs 100,000/month (NBS, Tanzania in Figures, 2008). Assumed minimum wage of 105,000 in 2009 The average household income in poor districts is assumed to be Tshs 75,000/month, based on the Dar es Salaam Com- munity. Upgrading Infrastructure Program reports and others. The average month has 30.5 days (365 + 12). Average household size is 4.5 persons (NBS, DWSSP Impact Assessment Survey , 2006) Source of data on percentage of households that get water from each source is NBS DWSSP Impact Assessment Survey, 2009 n.a.: not available Appendix B Summary of the Key Provisions of the Lease Contract with City Water Services 68 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM The Lease Contract for Water Supply and Sewerage in Dar es Salaam Summary of Key Provisions 1 Topic Contract Terms Reference2 Duration/ Duration of the contract is ten years from the Commencement Date and it Art. 3 Termination is renewable for ten additional years. At termination, Operator shall coop- erate with the Lessor to ensure smooth continuation of services, provide specified information, and return assets as required under the contract. Licenses As a Condition Precedent for Commencement, the Lessor and the Opera- Art. 1.7(m), tor are each required to obtain a license from the Regulator. The Lessor 5.3(b), (j) and the Operator must each also obtain other consents and permits as required by law. Performance As a Condition Precedent for Commencement, the Operator is required to Art. 1.7(e); Bond execute the Performance Guarantee providing for payment of up to US$ 3 47; Sch. 4 million in respect of the non-performance of the Operator or failure of the Operator’s majority shareholder to hold at least 51% of the shares of the Operator. Service Area The Operator’s Area including the Water Supply Area and the Sewerage Art. 7.1 Area is specified and distinguished from the DAWAWA Designated Area Appendix A Exclusivity and Operator has the exclusive right (subject to exceptions such as existing Art. 1.6(b); Obligations private and community wells and boreholes) and the obligation to provide 6.1; 7.2 piped potable water and sewerage services within the specified Operator’s Art. 15, 16, Area in conformity with the terms of the contract. The Operator’s specific 17 obligations with regard to the operation and management of services, maintenance and repairs, connections and revenue meters are listed. Obligation to Subject to the terms of the contract and provided it is reasonably possible Art. 12.3; Supply Service and in accord with good practice, the Operator is required to supply water Art. 6.3, 6.4, and sewerage services to any customer in the Operator’s Area who applies 6.5 for such service within the times specified, and subject to specified condi- tions. The Operator is not required to accept the disposal of any waste conveyed by tankers or other vehicles at any location except as designated by the Operator and agreed by the Lessor and on terms agreed by the Operator and the person wishing to discharge waste. Ownership and Lessor leases the existing Assets and new Assets after they have been Art. 8; 9 transfer of Assets created and makes them available to the Operator for use and maintenance. and Access to Ownership of the Assets, including Small Equipment (such as small pumps, valves, Land standpipes), remains vested in the Lessor and shall be returned to the Lessor in conditions consistent with good industry practice at the termination of the contract. The Lessor ensures the Operator will have access to all necessary land, property and water sources within the Operator’s Area. Physical Asset Within 12 months of Commencement the Parties shall establish the Art. 8.2 (b), Register and Agreed Physical Asset Register and the Operator shall maintain an up (c); 8.3 Asset to date comprehensive version of the Asset Register. The Operator is Art. 15.2(f) Management required to survey and quantify Assets in order to prepare and implement Plan a maintenance and replacement schedule (Asset Management Plan) to be updated annually. (continued on next page) Appendix B 69 The Lease Contract for Water Supply and Sewerage in Dar es Salaam Summary of Key Provisions 1 (continued) Topic Contract Terms Reference2 Transferred Personnel Lessor’s employees (not including certain recently appointed di- Art. 1.7, rectors) will be available for transfer to the Operator [for employ- 14.1; App. ment by the Operator]. Lessor will continue to provide for statu- D; DC 3 tory or voluntary pension programs of all transferred employees. Employment of Transferred Employees is subject to the DAWASA Act. Human Resources The Operator is required to train persons employed by it or its Art. 14.3(a), Development agents or contractors for the purposes of providing services within (b) the [Operator’s] Area and ensure that only qualified and certified personnel are employed for tasks of potential risk to the health or safety of the person. Management As a Condition Precedent, the Operator must complete the Mobi- Art. 1.7(a); Personnel and lization which includes the appointment of specified management App E; Technical Assistance personnel. The Operator’s Managing Director must be provided Art. 4.1(d), from the staff or officers of the majority shareholder. The Op- 4.5(b) erator may enter into technical assistance or sub contracts as required. Customer Contracts Within 12 months of Commencement, as agreed with the Lessor, Art. 12.1 following consultation with the Regulator and Customers and with the consent of the Regulator, the Operator shall publish a standard Customer Contract (model form in Appendix G) and issue Customer Contracts to all Customers. The deadline may be extended if required to obtain the consent of the Regulator. First Time New First Time New Domestic Water Supply Connections on proper- Art. 17.3 Domestic Water ties with up to three water points and no farther than twenty Supply Connections meters from the water main will be funded by the Fund for that purpose and the Operator will be paid the amount specified in Ap- pendix K (Tshs 145,000 at a July 2002 price base). Codes of Practice Within 12 months of Commencement (unless otherwise agreed Art. 13 and authorized) the Operator is required to publish Codes of Prac- tice for Customers, Customer Service Standards, Leakage Control, Disconnections, Customer Complaints Procedures, Entry into Land and Pollution Prevention, with certain contents as specified. Maintenance and At its own expense, the Operator is responsible for maintenance Art. 4.2(a), Repairs and repair of Assets, up to specified limits (see below) as re- (b); 15.1; quired, in accordance with Approved Technical Standards, Good 16.3; 16.4; Industry Practice and Good Industry Practice, to ensure their 16/5 reliability and performance. The responsibility of each Customer for maintenance and repair of water service pipes and sewerage drains within his property is specified. The Lessor is responsible for major repairs and replacements (e.g., water mains over 300 mm in diameter and over 6 meters in length, sewers over 450 mm in diameter and 4 meters in length and structural failures) and, in case the cost of any repair or replacement exceeds Tsh 10 million, the Lessor is responsible for the amount over Tsh 10 million. The Operator is responsible for the maintenance and repair of the Ocean Outfall after it has been rehabilitated as part of the Non- delegated Works. (continued on next page) 70 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM The Lease Contract for Water Supply and Sewerage in Dar es Salaam Summary of Key Provisions 1 (continued) Topic Contract Terms Reference2 Investment Planning The Immediate Capital Investment Program and a summary of App. J the proposed Medium Term Capital Investment Program are Art. 18.2; 20.1 shown in Appendix J. The Operator is responsible for preparing Art. 61.1, 61.2 the Medium Term Capital Investment Program, including pro- grams for rehabilitation and renewal of Assets, strategic business plans, network modeling, and annual investment programs, for agreement with the Lessor. The Lessor is responsible for under- taking surveys to determine demand and willingness to pay for the purposes of investment planning, demand forecasts, studies of the expansion of the water supply system, environmental assess- ments, and a strategic sanitation plan. Reporting: The Operator shall prepare an Annual Business Plan including an annual capital investment plan and the Immediate Capital Investment Program [including the Priority Works Program, Delegated Capital Works and Non-Delegated Works to be carried out within the first five years of the contract] detailing assets to be replaced or rehabili- tated and areas for expansion of service. Responsibility for The Lessor is responsible for all works concerning the extension and Art. 5.2; Financing, rehabilitation of the system, except as otherwise specified. (Art. 5.2) 8.2; 18.1; Procurement and The Operator shall review and provide written comments on the 18.3; 20 Implementation of preliminary or detailed designs for new assets to be created under Capital Investments the Capital Works Program. (Art. 8.2) The Lessor is financially responsible for the procurement of works for the Immediate Capital Investment Program included in Ap- pendix J, including Non-Delegated Works (financed and imple- mented by the Lessor), Delegated Capital Works (financed by the Lessor and implemented by the Operator), and the Priority Works Program (financed by the Lessor and constructed by the Operator). (Art. 18.1, 18.4) However, the Operator is responsible for financ- ing operating equipment; leak detection equipment; and cleansing equipment and safety devices for the sewerage network. (18.3) The Lessor shall endeavor to secure external finance to fund the Medium Term Capital Investment Program. Non-Delegated Works The NDW program includes major rehabilitation and new works Art. 18.4, (NDW) (excepting those in the PWP – see below) that are financed and 18.6 procured by the Lessor in accord with the procurement rules of the respective international or bilateral funding institutions. The Operator is not eligible to bid but affiliated companies or share- holders are eligible to bid (subject to conditions). The Lessor and the Operator consult and cooperate regarding design, implemen- tation and commissioning. Delegated Capital The DCW program includes components of the distribution net- Art. 18.4, Works (DCW) works and connections and is financed by the Lessor. The Operator 18.5; 20.2 serves as consulting engineer and is responsible for procurement in accord with lender-approved procurement plans (including prepa- ration of preliminary designs, detailed design and preparation of bid documents, bid evaluation and signing of contracts), site super- vision and commissioning of works. Neither the Operator nor any associated entity of the Operator is permitted to bid for the works. (continued on next page) Appendix B 71 The Lease Contract for Water Supply and Sewerage in Dar es Salaam Summary of Key Provisions 1 (continued) Topic Contract Terms Reference2 The Lessor has the right to review and approve or reject bidding documents prepared by the Operator and the Operator’s recom- mendations for the award of contracts under the DCW. The Lessor arranges for payment of contractors. Priority Works PWP includes high priority rehabilitation works and equipment, Art. 1.7, Program (PWP) rehabilitation of connections, and supply and installation of bulk 18.4, 18.7; and customer meters. It is implemented by the Operator under 20.3; the Supply and Installation of Plant and Equipment Contract (SIPE) and Procurement of Goods Contract (POG) contracts en- tered into by DAWASA and the Operator as a condition precedent. Replacement of Small Operator is responsible for replacement of Small Equipment (e.g., Art. 4.2(b) Equipment small pumps, valves, bulk and zonal meters, standpipes). Installation or Lessor funds and Operator procures and installs revenue me- Art. 17; Replacement of ters under the POG contract. Operator funds all other required Appendix C Connections and materials and the installation or replacement of meters on exist- Revenue Meters ing, registered, sound connections. Lessor funds and Operator procures/provides pipes, accessories and installation required for rationalizing existing legal and illegal connections under the SIPE contract or the DCW program. For new domestic connections, all materials (excluding meters) and the cost of installation are funded by the First Time New Domestic Water Supply Connection Fund. All materials for and installation of new non- domestic con- nections are funded entirely by the customer. Operating Equipment Operator is responsible for provision and replacement of Operat- Art. 4.2(b) ing Equipment (e.g., vehicles, tools, office equipment, and infor- mation technology hardware and software), and for the provision, maintenance and routine upgrading of any proprietary software or systems. Extensions to the Extension funded by Customers shall be implemented by the Art. 19 Network Operator. Developers may build extensions at their own ex- pense, subject to the prior approval of the Operator. The Lessor is responsible for funding extensions and reinforcements to meet urban demand or population growth. Standpipes The Operator is responsible for the management, maintenance and repair of standpipes and kiosks and whenever possible, shall subcontract management to standpipe agents and communities. There is no obligation to provide water free of charge, except for fire fighting. Service and Quality The Operator’s obligations are comprehensive, clearly defined Art. 25 – 33; Obligations and and specified in detail. Water and sewage quality standards and App H; App Performance Targets monitoring requirements are specified in Appendix H. Perfor- N; mance targets during the first five years of the Contract for 11 key performance parameters (such as drinking water quality, efflu- ent quality, number of meters installed, new connections, water losses, repair times, data collection and pressure) and for 16 other parameters (such as volume of water produced, continuity of service, number of active accounts, response times, billing based on meter reading), and penalties for non-compliance with the Key Performance Targets are specified in Appendix N. (continued on next page) 72 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM The Lease Contract for Water Supply and Sewerage in Dar es Salaam Summary of Key Provisions 1 (continued) Topic Contract Terms Reference2 In cases where the base value for a parameter is not reliable, transition values are specified. The base values will be ascertained by the Operator and agreed with the Lessor during an Enhanced Monitoring Period and target values will be agreed between the two Parties. Targets that are dependent on the implementation of the Immediate Capital Investment Program will also be agreed during the Enhanced Monitoring Period. Customer Tariff The Customer Tariff (Water Supply) includes the Operator Tariff Art. 34, 35; (Water Supply), the First Time New Domestic Water Supply App. K Connection Tariff (to be paid into a fund for that purpose) and the Lessor Tariff (Water Supply). The Customer Tariff (Sewer- age) includes the Operator Tariff (Sewerage) and the Lessor Tariff (Sewerage). The (average) Customer Tariffs per cubic meter for Water Supply and Sewerage for the first five contract years are specified. These tariffs are set at July 2002 prices. The three components will be adjusted annually: the Operator Tariff by the specified Indexation Adjustment Formula and the Lessor Tariff and the First Time Connection Tariff by a retail price index. The tariff structure (differentiated tariffs paid by different catego- ries of customers) is established by the Lessor in accord with the Development Contract between GOT and the Lessor. All custom- ers pay the full Operator Tariff on all volumes consumed. Domes- tic customers with household connections who consume more than 5 cubic meters per month pay the Operator Tariff, Lessor Tariff and First Time Connection Tariff on consumption above 5 cubic meters. Non-domestic customers pay all three components on all volumes consumed. Regulatory Levy The Operator levies and collects the Regulatory Levy from Art. 39 customers and pays all collected amounts to the Regulator. [not enforced until EWURA was established] Operator’s Revenues3 The Operator’s revenues include: Art. 36; 39; – The Operator Tariff for water supply and sewerage collected 43; 45; App. from customers (from which the fixed monthly Rental Fee K; App. P; must be paid to Lessor); – Charges for the provision of ancillary and incidental services (tanker water, disposal of liquid or solid waste from sewage tankers, etc.); – Operator Fee for First Time New Domestic Water Supply Connections as specified in Appendix K and indexed annu- ally; (This fee is withdrawn by the Operator from a Fund into which all FTNDWSC Tariff revenues are deposited by the Operator.) – 10% of any debt owed to the Lessor and collected by the Operator; – For design and supervision of the DCW program, a total of 10 % of the value of the DCW; – Payments by Customers for private extensions under agree- ment with the Operator. (continued on next page) Appendix B 73 The Lease Contract for Water Supply and Sewerage in Dar es Salaam Summary of Key Provisions 1 (continued) Topic Contract Terms Reference2 Payments to the Lessor The Operator pays the Lessor: Art. 39; 43; – The Lessor Tariff for water supply and sewerage collected 48; 49 from customers; App. K; – The amount by which any water supply tariff (for non-domes- App. N tic customers) temporarily exceeds the specified Customer Tariff; – The (fixed) Rental Fee as specified, and annually indexed, to be paid monthly in advance. – Financial penalties (if applicable), subject to specified limits; – Customer debt owed to the Lessor and collected by the Op- erator, minus 10%; – Any Customer security deposits held by the Operator at the termination of the contract; Review of the Operator The Lessor shall review the Operator Tariff and the Indexation Art. 41 Tariff Formula to take account of changes in relevant conditions and recommend any changes to the Regulator for approval. – The Major Review will be undertaken to set the Operator Tar- iff and Performance Targets for Contract Years six to ten. – An Interim Review may be undertaken at the request of either the Lessor or the Operator, in the event of a Material Change of Circumstances. – An Annual Review may be requested if either Party believes the percentage weightings of the Indexation Formula need revision, as a result of a change in Base Values and Perfor- mance Targets during the Enhanced Monitoring Period. Billing and collection Customer relations, Customers must pay a deposit which will be held by the Opera- App. G customer contracts and tor and returned to the Customer at termination of the Customer customer rights contract after any outstanding charges have been deducted. Reporting, record The Operator is required to: create records of information as Art. 58; 59; keeping and audits required for facilitating efficient management and supervision of 60; services and for informing the Lessor and Consumers; maintain financial, technical and commercial information in formats that permit regular audits; create and develop an Asset Management Plan and a Schedule [record] of Assets; submit reports as required by the Lessor; and have its records audited [in addition to audits that may be undertaken by the Lessor]. (continued on next page) 74 PUBLIC-PRIVATE AND PUBLIC-PUBLIC PARTNERSHIPS IN WATER SUPPLY AND SEWERAGE SERVICES IN DAR ES SALAAM The Lease Contract for Water Supply and Sewerage in Dar es Salaam Summary of Key Provisions 1 (continued) Topic Contract Terms Reference2 Regulator The Regulator: (Until EWURA was established, the Minister of DC: 11; Art. Water served as Interim Regulator.) 12, 13, 29, – Approves Customer Contracts and codes of practice relevant 30, 34.4, to customers; 41.1(b), – Approves any adjustments to the Customer Tariffs and the 41.3(b), Operator’s charges for other services; 41.4(b), – Approves any changes in the Operator Tariff that result from 66.3; the Major Review, an Interim Review or an Annual Review; – Is involved in the Major Review and may direct the Parties to undertake an Interim Review; – Receives monthly raw, potable and effluent water quality monitoring reports and reports of any major failure in quality standards from the Operator; – Resolves disputes between the Lessor and the Operator relating to the Tariff, if the Parties cannot resolve it through amicable means or through submission to the Expert Panel. Resolution of disputes The Parties will endeavor to settle disputes or differences ami- Art. 66 cably. Failing that, a dispute will be referred to an Expert Panel. If the Parties fail to agree on the recommendation of the Expert Panel, a dispute which relates to a matter other than the Tariff may be referred to arbitration. An unresolved dispute that relates to the Tariff will be referred to the Regulator. 1 This summary is not exhaustive, nor does it include all exceptions and conditions that may apply to the contractual pro- visions that are covered. In some instances, the contractual language has been shortened and summarized and American English spelling has been applied. 2 References refer to articles (Art.) of the Lease Contract and its Schedules, Appendices (App.) and Annexes. References to articles of Annex 1, the De velopment Contract, are indicated by the abbreviation DC. 3 The Operator’s Revenues are those specified in the Lease Contract. Revenues for implementation of the POG and the SIPE are specified in those contracts.