a4 t 1018 M LU z 0 LU 0 w](pwpoit PRural Water Servce 7'hiscasestiid is pail of N a Private HationaD Operator a ViabUe Business MoJdeO? a series Yeviewziig business models for priv(ate In C6te d'lvoire and Senegal water service is provided by privately comnpanies providing operated national water utilities operating under enhanced lease infrastructwle services to r,urali customers in contracts. While the national operators have performed relatively developing coutriei 7 well within their service areas, their ability to expand service to case studies aiieijfactors dlrivingthepeeiforinaince rural areas has been limited. This Note focuses on the key factors ofprivate companies in a shaping rural service expansion: reliance on cross-subsidies, the variety of rural contexts andsectorsandunder limited transfer of commercial risk to the private sector, and the diveise legal and lack of competition for serving new population centers beyond the c: egulatory rlzrzangemientis utilities' existing areas of exclusivity. It compares the national model 7'lie objective I to identify lessoni thait can inforin with those of more decentralized service provision for rural areas. the design of private rural iafiastiiicture scheinei Soci6t6 de distributLion d'eau de la C6te the renegotLiation of its contract in 1987 tlhani it infiastpiuctitre schlenes oke 71priiaiyJundingfor d'lvoire (SODECI), a subsidiary of the French would under a traditional affermage contract. tihis series camefrom the water company SAUR, has been operating in SODECI is responsible for drafting a five-year | Publc-Private C6te d'Ivoire since the country's independ- investment plan, managing a special fund for Infiastructlure Advisory ence in 1960 (box 1). Senegalaise des eaux investments in service expansion (finaniced by a Facility, a intltidonor (SDE), also a subsidiary of SAUR, has been dedicated surcharge on water tariffs), ancl carry- techiiical assistanceJund present in Senegal only since 1996, following a img out most investments (although any invest- h7he oil(ld Bank task comprehensive sector reform. Both companies ment above a certaim threslhold Is supposed to be teatii for the case stutdies are national operators Nvith exclusive service tendered) In practice, however-, the companiy wzas Alan 7 ownsend, areas and serve a large number- of secondary estimates that the Water Ministry, whlich is utltl- Clive Harris, Lorenzo towns, thanks to a long-established system of mately responsible for strategic investnent deci- Bertoliiii, AlVichiael Sihur, cross-subsidization.1 But outside the service sions in the sector, approves only about 50 Juan Navas-Saobater, and areas managed by these private operators, the percent of its proposed Investments =1 John iVeistead rollout of rural water service has lagged. In recent years, amid social tensions stirred by 0 Both companies operate under comparable a cotip in 1999, the system's perfor-mance has affermage contracts, with limited responsibility started to dechlne fi-om previously enviable levels. for investment In Cote d'Ivoire SODECI has in Losses now stand at an average of 23 percent of theory had more control over investnment since water production (compared with a 15 percent RURAL WATER SERVICE IS A PRIVATE NATIONAL OPERATOR A VIABLE BUSINESS MODEL? Cote d'lvoire nationale d'exploitation des eaux du Senigal (SONEES). Sector Societe de distribution d'eau de la C6te d'lvoire (SODECI) reform became necessary as SONEES faced financial difficulties began its operations in Cote d'lvoire's capital, Abidjan, and a and heavy investments were required to address Dakar's long- few other major towns. Later its responsibilities were extended term supply problems. to the entire national territory, including maintenance services In 1996 private sector participation was introduced through in rural areas. a 10-year affermage contract between the government and In 1987 SODECI renegotiated its contractual arrangements Sinegalaise des eaux (SDE). An asset holding company, Sociiti 2 with the government and signed a 20-year concession contract. nationale des eaux du Senegal (SONES), was set up to own the Its responsibilities in rural areas were terminated and trans- assets and carry out investments in the urban water sector ferred to rural communities, while the government kept invest- under a concession contract with the government. The ment responsibilities in those areas. Senegalese contract is an "enhanced" affermage contract, giving But SODE(I's service area is continuously expanding. It now SDE responsibility for maintenance and some well-specified includes 583 centers, reflecting growth of more than 50 per- investments in rehabilitation (17 kilometers of network, 6,000 cent since contract renegotiation in 1987. The size of these connections, and 14,000 water meters every year). centers ranges from a few thousand inhabitants to 3 million in At the time the contract was signed the SONES-SDE service Abidjan. (The population of Cite d'lvoire is around 14.8 mil- area included 56 urban centers and 100 villages along the lion, and just over 50 percent is urban.) main pipeline routes, and it has not expanded since. These centers are usually larger than those in Cite d'Ivoire, reflecting Senegal the higher threshold for adding new population centers to the By contrast, in Senegal the water utility's geographic coverage network. In Senegal the threshold is close to 10,000 inhabi- grew during times of public ownership. Although a private tants, while that in Cite d'lvoire is 3,000 (although there are operator was in place at independence in 1960, water services exceptions in both countries, with incorporation often depend- were nationalized in 1971 and transferred to a national com- ing on administrative status or political factors). In Senegal just pany responsible for serving the main urban centers, Soctiit over 50 percent of the population of 9.8 million is rural. contr-actuial target), and a thi-d of the production within the existing service area than on expand- centers, many of them in the inter-ior-, no longer ing senvices beyond it. meet thc World Health Oirganiizatioin's water This approach has yielcied improved opera- quality standar-ds. The government has not tions. Leakage was cut fi-om 31 percent to 22 per- applied sanctionis to SODECI for failing to meet cent (althougih the 15 percent contractual target its (rather loosely defined) contractual targets, seems difficult to reach), and bill collection acknowledging that the failure to invest is as improved dramatically, to 97 percent. A big part mLich its responsibility as that of the private oper- of this was the success in persuading governmient ator. Total sales reventue and profitability have institutions to pay their bills. By contrast, in C6te also recently stagnated and declined. d'lvoire lowv payment rates by government insti- In Senegal funds for investimienit are chan- tutions remilain one of the main obstacles to mov- neled through Socicte nationiale des eaux dii ing to a full concession. In both coutntries, Senegal (SONES), the asset holding company, however, there is a high rate of disconnection whichi has also received abtindanit fusnding from for nonpayment (about 10 percent of connec- international donos-s in recent years. SONES tions in Senegal and 20 percent in parts of C6te also moniitors SDE's contract, and althougih d'lvoire). Although the performance measures SDE's contractual responsibilities for invest- in Senegal are based on averages across the ser- ments have led to some misunderstandings, they vice area, performance appears comparable in are mnuch better specified than in Cote d'lvoire. all centers regardless of size and location. These responsibilities-added to a set of coII- tractual performance inclicator-s wNith clearly What drives rural performance? defined targets and corresponiding finanacial The factors most affecting the expansioni of incenitives and sanctions-have focused SDE's service into rural areas are the design of cross- incen-tives more on improving performance subsidized tariffs, the degree of risk transfer to the private sector, and the extent of competi- porateci into the service area for fear that they tiOln for- serving new populationi centers, will pay more than thyc neecl to. Cross-subsidization Transfer of commercial risk In both coulitlies regionial cross-subsidization Both countries tiansfer limiited coiiirn eicial risk was the maini financial instrumlaent for support- to the private sector. Under both contracts the ing rtiral expansion. In the Francophone legal private operator receives a fixed payment foir tradition futindamental principles of public ser- each cubic meter ofwater solcl, regar-dless of the vice call for treating tusers equally across the tariff at which it was sold. At the margin, sucih nationial territory, including chaiginig a single arrangements have strong pro-poor properties tariff. This principle was applied in both coun- because the private operator receives the same tries so that rural constimer-s pay the same price r emuner-ation for water sold at the social tariff for water seivice as urban consumilers, regardless as for water sold at the much higher- industr-ial of the costs they impose on the syslem. In addi- tariff even thotigih the costs of the service may tion,l both couLitries iave adopted a social tariff differ. These properties benefit custoiners in (the lowest block of an increasing-block tariff peri-urban areas and secondary centers, which structire), and seconciary centers appear to accotlnt for the largest shar-es of social tariffcLIs- ccouLIt for a larger share of'the consumers pay- tomers. In the long tem, howe0Ver, sucIh pro- ing the social tariff thani do the capital cities. 2 poor properties are tunclerminecl by the fact that In C6te d'Ivoire none of the centers served the fixed remulIer-ation per cubic meter is cal- outside the capital city wotild be profitable on culatecl on the basis of an average tariff If thie its ovn, and all these centers are financed share of social tariff customers increases, the thiough surpluses generated in Abidjan. In average tariff will iall andl so will the private 2001 Abidjan accounted for 48 percent of con- operator's remuner-ation. nectiois, 65 percent of sales revenue, and only Such contr-actual arrangements also runi the 52 percent of productioni costs. Water produc- risk of utndermininiig the long-ter-m sustainabil- tion costs are lower in Abidjan largely because it ity of the sector. Under- these arr-anigemilenits the sits on a large aquifer and abstracted groulid- diflfe-erice between sales reveniule an d the total water- reqtmires less treatmeit, while most of the r emunuer-ationi of the privacte operator- goes back secondclay centers use plentiful (though not to the asset holder for futtL-e investments. If very clean) suLface waters, vhichi are more sales reveniue cdeclines because of a risinig pro- expensive to treat. portion of social tariff ctmstomels, fewer- fuLicis In Senegal, determining w'ho wfins and who will be available for investment. loses fromi cross-sLibsidization is much more dif- ficult because of the lack of an analytical cost Competition for serving new centers accounltilng system. Although Dakar and its Differences betwveen the two countr-ies in the regioni accounted for 57 percent of connections potential for cross-subsidizatior, in contr ol over and 72 percenit of sales r evenue in 2001, water investment policies, and in incentive strtctur-es costs are likely to be substanitially highier- in this have led to differenit results from government region than elsewhere because its water source efforts to expand service in rural areas. Both is 150 kilometers away. have experimented with village water comnuit- No new munlicipal center-s have been acdded tees, whicih are supposecd to manage and oper- to the service area in Senegal since privatization. ate the local water infirastructure. And both are A qualitative review suggests that some second- turning to the national utility to provide service ary towns and rur-al areas might be cross- in villages that lhave outgrowvn1 community miianl- stibsidizing urban centers and would be better agement. But in C6te d'Ivoire community water off with a local, atutonomous stipply system than service has not proved very sticcessftil. Abotut 20 with service by the national operator. And percent of village wells are out of service, and 40 uniless new center-s know how much subsidy they percent of village vater committees arc not func- vill receive, they mighit be reluctanit to be incor- tioniig. Villagers often have great cliflitculties in RURAL WATER SERVICE IS A PRIVATE NATIONAL OPERATOR A VIABLE BUSINESS MODEL? finanicillg their required contribution to invest- tages because of their technical expertise and ment costs, and installationis are poorly maini- economiiies of scale, but competition for subsi- tained. Since any village with more than 3,000 dies would force them to demonstrate those inhabitants is eligible to be included in advantages more actively. SODECI's service area and since there is a track viewpoint record of new centers being regularly incorpo- Conclusion rated, many villages prefer to wait for incorpora- Results in C6te d'Ivoire and Senegal suggest that tion rather than invest in their own facilities. That a private national utility can provide water ser- Abidjan's customers largely subsidize SODECI's vice to small population centers through cross- encourage dissemination of service contributes to that preference. subsidization. But the present arrangements in public policy innovations for By contrast, in Senegal community water ser- these countries have shown limited potential for private sector-led and vice has been more dynamic, with some villages improving service provision or ensuring its sus- market-based solutions for financing all the initial investments (mostly tainability in rural areas. Transferring greater development. The views through migrant associations). This could be risk to the private sector would sharpen private published are those of the because in Senegal the water utility has less of a companies' incentives to expand access and authors and should not be track record of expanding its service area and would also improve sustainability by relieving attributed to the World because scarcer water sources mean greater the pressure on stretched government funds. Bank or any other affiliated incentives for commtnities to organize supply. A Moreover, with support by a strong, committed organizations. Nor do any of pilot project run by the French bilateral aid government, introducing competition from the conclusions represent agency, Agence francaise de developpement small local operators with targeted subsidies official policy of the World (AFD), is tuying to lift some of the biggest barri- could help promote transparency and foster the Bank or of its Executive ers to greater development of these local solu- expansion of service in rural areas. Directors or the countries tions. It is grantng an explicit legal status to they represent. village water committees (allowing them to open a bank account and get credit) and organizing a To order additional copies clear system for service delegation, which could Note contact Suzanne Smith, lead to the emergence of local small-scale private 1. As the term is usecd here, a nIational utilhty is national managing editor, providers. These solutions have been met with in the sense that it operates all over the country but with- Room 19-009, great enthusiasm by small towns that may be eli- out a uniiversal access obligation Within its service area The World Bank, gible for incorporation into the SONES-SDE the utility has exclusivity, and as its serice area expands, so 1818 H Street, NW, sernice area but would prefer controlling their docs its exclusivity. Beyond the national utility's service Washington, DC 20433 water service through a local solution. The area, competition is permitted. Telephone knowledge that local solutions may be less expen- 2. In C6te d'lvoire the social tariff covers consumption 00 202 458 7281 sive may also contribute to this attitude. Even so, upto 18cubicmeters,atUS$0.22perctibicmeter. In 2001, Fax. these towns simply prefer a local solution, which 21 percent of consumilption was billed at this tariff in 001 2025223181 they see as a way to bLtild local capacity and com- Abidjan and 35 percent in the regions. In Senegal the Email munity control. Indeed, some would even prefer social tariff covers consumption up to 20 cubic meters, at ssmith7@worldbank.org to pay higher tariffs if the benefits could be redis- US$0.23 per cubic meter Both counitr-ies subsidize the cost tributed to other community activities. of connection for consumers eligible for the social tariff Copyedited and produced by As water service is increasingly decentralized, (90 percent in C6te d'lvoire and 100 percent in Senegal). Communications competition for new markets and access to sub- Development Inc sidy fuinds could be taken further-to limit the So)phie 7remolet is a consultant at Environmental Resources de facto monopoly granted to national utilities Management in London (sof)hie tremolet@er7m.com). Printed on recycled paper over the management of funds from cross- PUBLIC-PRIVATE subsidization. For example, small-scale pro- INFRASTRUCTURE viders could compete for contracts with the main operator and receive subsidies from a water sector development fund, perhaps on the basis of output-based performance. National tltilities would retain many competitive advan- This Note is available online: www.worldbank.org/viewpoint/