37548 noTE no. 12 ­ SEpT. 2006GRIDLINES Sharing knowledge, experiences, and innovations in public-private partnerships in infrastructure Using management and lease- affermage contracts for water supply How effective are they in improving service delivery? Klas Ringskog, Mary Ellen Hammond, and Alain Locussol A s governments in developing countries Of these, 52 percent were in Eastern Europe seek to expand access to water supply and Central Asia, 17 percent in Africa, and 13 and improve the quality of service, percent in Latin America and the Caribbean. they are considering a range of options for public-private partnerships. Yet proposals to Medium-term management and lease-affermage involve the private sector have often met with contracts best match the current risk-averse concerns--about tariff hikes, staff cutbacks, stance of private operators, no longer willing to and ability to reduce inefficiency and expand assume the high financial and foreign exchange access to service among the poor. And experi- risks of long-term concessions. Under manage- ment contracts the financial risks to private ence with the more than 300 contracts bringing operators are rather low because their fees are private operators into water supply in 1990­ usually financed by government budgets or loans 2004 has been mixed. How have water utilities from international financial institutions. Under performed under management and lease-affer- lease and affermage contracts the private opera- mage contracts? Eight case studies across tors bear a substantial commercial risk because developing regions show that performance they have to generate an operating surplus suffi- has generally improved--across a range of cient to cover their remuneration. (Lease and measures. affermage contracts differ mostly in the way the Widespread problems in the urban water sectors commercial risk is shared between the operator of developing countries impose big burdens and the owner of the contract.) on households and businesses. Poorly served customers are driven to invest in costly alterna- How well have management and lease-affermage tive sources of supply. Poor households are often contracts performed? A recent study addressed excluded from public service and have to resort that question using empirical data. The study to unsafe and ultimately much costlier alterna- collected information on 11 contracts selected tives--if they can afford them at all. on the basis of geographic coverage, a popula- tion coverage of more than half a million, and In the early 1990s many governments launched a duration of at least four years. The data were reforms of their urban water sector aimed at gathered through desk research of secondary improving service. Publicly owned and managed sources, supplemented by limited primary data utilities undertook a range of institutional, collected from owners and operators through a operational, and governance reforms. But questionnaire. improvements were limited and often unsustain- able. Some governments therefore turned to the The analysis centers on the eight cases for which private sector. comprehensive data were available--two each from Africa, Eastern Europe, Latin America, and In 1990­2004, about 340 contracts involving public-private partnerships in water supply were PUBLIC-PRIVATE INFRASTRUCTURE ADVISORY FACILITY Klas Ringskog is a consultant, Mary Ellen Hammond a closed in developing countries.1 About a fifth junior professional associate, and Alain Locussol a lead were management and lease-affermage contracts. water supply specialist, all at the World Bank. Helping to eliminate poverty and achieve sustainable development through public-private partnerships in infrastructure PUBLIC-PRIVATE INFRASTRUCTURE ADVISORY FACILITY and Zambia (Table 2). The growth in access to TAblE 1 water supply was substantial in Barranquilla, Contracts and span of data assessed Cartagena, and Senegal, where it was supported Type of contract and Period by strong incentives for private operators and location studieda adequate investments funded by public sector utility owners. Management Amman (Jordan) 2000­05 Gaza (West Bank and Gaza) 1995­2005 Access to piped water also grew among the poor. Zambia mining townships 2000­04 In Senegal the connection rate for poor house- Affermage holds rose from 22 percent in 1996 to 60 percent Antalya (Turkey) 1996­2001 in 2005. Driving this big increase were a social Gdansk (Poland) 1992­2005 connection fund offering direct connections to Senegal (55 cities including Dakar) 1996­2005 low-income customers at a highly subsidized price lease-affermage (funded in part through a surcharge on customer Barranquilla (Colombia) 1990­2005 tariffs) and regular extensions of the distribution Cartagena (Colombia) 1995­2005 network. In Cartagena, where water customers a. The first year in the range is the year before the contract took are classified by the value of their property (land effect ("before"); the second, the last year for which data were and house), the connection rate in the poorest assessed ("after"). areas is estimated to have risen from 35 percent in 1995 to 60 percent in 2005. the Middle East (Table 1). Most contracts were Even where poor households are not explicitly signed with private operators in the late 1990s. classified, increasing the overall connection Five contracts are ongoing, and three have been rate disproportionately favors these households or soon will be concluded. because they usually are the last to be connected to piped water. The study computed indicators of performance in five areas for the year before the private opera- More water--and better quality tor started operations and for either the last year Water supply also improved. Initially all cases of the contract (for contracts that have lapsed) except Gdansk had less than 24 hours of water or the most recent year available (for ongoing Private supply a day. During the contract period supply contracts): increased by about 5­7 hours a day, and in Anta- operators · Service coverage (including for poor households lya, Barranquilla, Cartagena, and Senegal it expanded where this information was available). increased to nearly 24 hours a day. access to · Quality of service (hours of water supply a day, Compliance with water quality standards, already good in all the cases analyzed, improved further, water-- share of drinking water samples testing negative for pathogens). though by marginal amounts. But the data gener- including ally refer to the quality of water at production sites, and pollution occurs mostly in distribution among the · Operating efficiency (share of nonrevenue water, share of customers metered). networks when empty pipes create a vacuum. poor So it is likely that the increase in hours of water · Employee productivity (employees per 1,000 supply reduced water contamination in the distri- water connections). bution networks. · Sustainability of service (working ratio). Gains in efficiency The review finds that while performance varies, Under private management and operation, water the eight case studies show positive trends in the utilities achieved greater commercial and opera- extent and quality of water service and the effi- tional efficiency. Nonrevenue water declined in all ciency and sustainability of operations. eight cases as a result of a combination of measures adopted by the private operators--including bulk More people with access to water and individual metering, leak detection, and repair Access to piped water increased under the and maintenance. Gains ranged from a mere 3 contracts. The share of households with water percentage points for Gdansk--reflecting the connections rose in five cases, fell slightly in increasing difficulty of further reducing already Gaza, and remained at 100 percent in Gdansk low levels of nonrevenue water--to 31 percentage points for Cartagena (Figure 1). Using management and lease-affermage contracts for water supply TAblE 2 Access to and hours of water supply before and after the contract Households with piped water connections (percent) Hours of supply per day Case before After before After Amman 90 100 4 9 Antalya 93 95 19 23 Barranquilla 60 89 19 23 Nonrevenue Cartagena 74 95 17 24 water Gaza 58 56 -- 8 Gdansk 100 100 24 24 declined in all Senegal 59 73 16 22 Zambia 100 100 13 18 cases--but expectations -- Not available. Source: Survey data from service providers. of rapid reduction Nonrevenue water dropped more sharply where able variations. The average tariff increased (in must be levels were very high before the private opera- constant 2000 prices) in four cases--in Antalya by tor took over (Barranquilla, Cartagena, Zambia) 129 percent, Barranquilla by 43 percent, Gdansk carefully and where the private operator faced particularly by 150 percent, and Senegal by 7 percent. But it managed strong incentives to reduce it (Barranquilla, Cart- actually decreased in three others--in Amman by agena). Nonrevenue water can also be expected 7 percent, Cartagena by 46 percent, and Zambia to drop substantially if the management or lease- by 53 percent (figure 3). affermage contract is implemented in parallel with programs to rehabilitate distribution On the cost side, employee productivity networks or to replace individual connections, (measured by employees per 1,000 connections) where typically 70 percent of physical leaks can improved in six of the eight cases. Except for be found. Barranquilla, where the operator is jointly owned by the municipality and a private investor-oper- Expectations of rapid reduction in nonrevenue ator, there was no retrenchment of staff after water must be carefully managed, however. In the signing of the contract. Instead, most of the almost all the cases the pace of reduction was productivity gains came through a combination slower than targeted in the contract. The failure to of staff reduction through normal attrition and realize initial expectations of a sharper decrease in an increase in connections. nonrevenue water may have been one reason that three of the contracts were not renewed. Two cases, Amman and Cartagena, illustrate the efficiency gains achieved (Figure 4). Both private Growing sustainability of service operators were able to reduce their working ratio Financial sustainability improved under the public- private partnerships, through gains in operating efficiency and cost recovery. Utilities typically fiGurE 1 reduce their working ratio (the ratio of cash operat- Private operators cut commerical losses ing costs to cash collections) through higher tariffs, Nonrevenue water as a percentage of the total better collections, and better control of operating 80 costs. In six of the eight cases the working ratio 70 improved by 55 percentage points on average 60 (Figure 2). This allowed operating surpluses to be 50 applied to investments, reducing the pressure on 40 30 the public budget. (No data were available for the 20 other two cases.) 10 0 Several factors influencing the working ratio are Amman Antalya Barranquilla Cartagena Gaza Gdansk Senegal Zambia under the private operator's control. But the tariff Before After level is controlled by the public utility owner or Source: Survey data from service providers. the government, and it can account for consider- fiGurE 2 fiGurE 3 Private operators improved financial Tariffs rose in phased moves toward sustainability cost-recovering levels Working ratio Average tariff (US$ per cubic meter) Note: No data available for Gaza and Gdansk. Note: Tariffs are in constant 2000 prices. No data available Source: Survey data from service providers. for Gaza. Source: Survey data from service providers. fiGurE 4 big gains in productivity in Amman and Cartagena Amman Cartagena Source: Survey data from service providers. through measures to reduce costs through higher though this did not systematically translate into productivity--even as the tariff was reduced in increases. The average tariff rose (in constant constant terms. 2000 prices) in four cases--in Antalya by 129 percent, Barranquilla by 43 percent, Gdansk GRIDLINES Several factors influencing the working by 150 percent, and Senegal by 7 percent. But ratio are under the private opera- thanks to efficiency gains, it actually decreased tor's control. But the tariff level in three others--in Amman by 7 percent, Carta- Gridlines share emerging knowledge on PPP is controlled by the public utility gena by 46 percent, and Zambia by 53 percent and give an overview of a wide selection of owner or the government, and projects from various regions of the world. Past (Figure 3). it can account for consider- notes can be found at www.ppiaf.org/gridlines. Notes Gridlines are a publication of PPIAF (Public-Private able variations. Tariffs moved PUBLIC-PRIVATE INFRASTRUCTURE ADVISORY FACILITY Infrastructure Advisory Facility), a multidonor toward full cost recovery, 1. Data are from the World Bank and PPIAF Private Participation in Infrastructure (PPI) Project Database (ppi.worldbank.org). technical assistance facility. Through technical assistance and knowledge dissemination PPIAF supports the efforts of policymakers, nongovernmental organizations, research institutions, and others in designing and implementing strategies to tap the full c/o The World Bank, 1818 H St., N.W., Washington, DC 20433, USA potential of private involvement in infrastructure. The PHoNE (+1) 202 458 5588 fAX (+1) 202 522 7466 views are those of the author(s) and do not necessarily GENErAl EMAil ppiaf@ppiaf.org wEb www.ppiaf.org reflect the views or the policy of PPIAF, PUBLIC-PRIVATE INFRASTRUCTURE ADVISORY FACILITY the World Bank, or any other affiliated organization.