252 privatesector P U B L I C P O L I C Y F O R T H E 35101 NUMBER NOTE 2003 Infrastructure Projects JANUARY Clive Harris, John A Review of Canceled Private Projects Hodges, Michael Schur, and Padmesh Shukla In recent years the renegotiation and, even more, the cancellation of private infrastructure projects in developing countries have made the headlines in the world's financial press. For a variety of reasons the renegotiation of projects is not an unusual occurrence. But as this Note explains, only 48 private infrastructure projects in developing NETWORK countries were canceled in 1990­2001, a small fraction of the nearly 2,500 projects that reached financial closure over this period. The 1990s saw a revolution in the provision of Box A definition of cancellation INFRASTRUCTURE infrastructure services as governments world- 1 wide turned to the private sector for financing For the purposes of this Note, a project is considered AND and management expertise. In developing to be canceled if one or more of the following events countries in 1990­2001, nearly 2,500 infrastruc- occurred before the end of the project's expected life (as determined in a contract or license): SECTOR ture projects involved private participation, The private company sold or transferred its economic attracting investment commitments of US$750 interest in the project to the public sector. billion. But the investment flows to such proj- The private company physically abandoned the proj- PRIVATE ects have fallen fairly steadily since 1997, drop- ect (such as by withdrawing all staff from the project). ping to less than half their peak by 2001. At the The private company ceased to provide services to same time the cancellation and renegotiation of all customers or halted construction of the project for some private projects have grabbed the head- GROUP around 20 percent or more of the project's expected lines in the world's financial press. life following the revocation of a license or repudiation Renegotiations of private infrastructure proj- by the relevant contracting or licensing authorities. BANK ects have occurred in many countries, and par- ticularly in some sectors and regions. One study estimates that as many as 74 percent of transport government. But some long-term contracts are concessions and 55 percent of water conces- renegotiated in response to unforeseen or WORLD sions in Latin America were renegotiated in the changing circumstances. Cancellation, where 1990s.1 Some renegotiations are due to oppor- the private sector exits the project before the THE tunistic behavior by the private investor or the end of the contract or license term, has made I N F R A S T R U C T U R E P R O J E C T S A R E V I E W O F C A N C E L E D P R I V A T E P R O J E C T S Canceled infrastructure projects with private participation in developing countries, Table 1990­2001 1 Committed investment Year of (2001 US$ financial Year of Project Sector Country millions)a closure cancellation Electricité et Eaux des Comores Electricity distribution Comoros 12 1997 1998 Cesco, Orissa Electricity distribution India 31 1999 2001 2 Dabhol Power Plant I Electricity generation India 1,050 1996 2001 Dabhol Power Plant II Electricity generation India 1,988 1999 2001 Dieng Geothermal Power Plant Electricity generation Indonesia 508 1996 1998 Karah Bodas Power Plant Electricity generation Indonesia 454 1994 1998 Patuha Power Ltd. Electricity generation Indonesia 717 1997 1998 Almaty Power Consolidated Electricity distribution Kazakhstan 347 1996 2000 Kazakhstan Natural Gas Natural gas distribution Kazakhstan 662 1997 2000 Transmission System Senelec Electricity distribution Senegal 69 1999 2000 Millicom de Costa Rica S.A. Telecommunications Costa Rica 4 1989 1993 Evergrowth Telecom Ltd. Telecommunications India 201 1997 1999 Koshika Telecom Ltd. Telecommunications India 443 1996 1999 Telecom Services Kiribati Ltd. Telecommunications Kiribati 1 1990 2001 France Telecom Mobile Liban Telecommunications Lebanon 183 1994 2001 Liban Cell Telecommunications Lebanon 263 1994 2001 Mobile Telecom Services Limited (MTS) Telecommunications Nigeria 1 1992 1995 Utel Telecommunications Ukraine 285 1996 2001 Rosario Port Ports Argentina 185 1998 2000 Sizarail Rail Congo, 0 1995 1997 Dem. Rep. of Ferihegy Airport Airports Hungary 132 1997 2001 M1/M15 Toll Road Toll roads Hungary 453 1993 1999 Jakarta Outer Ring Road Toll roads Indonesia 260 1995 1998 (E2, E3, and N sections) Jakarta Outer Ring Road Toll roads Indonesia 368 1993 1998 (S and E1 sections) Mombasa Container Terminal Ports Kenya 0 1996 1997 Acapulco-Tierra Colorada-Cuernavaca Toll roads Mexico 2,612 1989 1997 Aquascalientes-León-Lagos de Moreno Toll roads Mexico 430 1990 1997 Cadereyta-Reinosa Toll roads Mexico 439 1990 1997 Campeche-Champoton Toll roads Mexico 117 1991 1997 Chamaca-Lechería Toll roads Mexico 340 1991 1997 Córdoba-Veracruz y La Tinaja- Toll roads Mexico 1,592 1990 1997 Cosoleacaque Culiacán-Mazatlán Toll roads Mexico 739 1990 1997 Guadalajara-Tepic Toll roads Mexico 1,398 1990 1997 Guadalajara-Zapotlanejo Toll roads Mexico 123 1993 1997 Lagos de Moreno-Zapotlanejo Toll roads Mexico 359 1990 1997 Libramiento de Tampico Toll roads Mexico 42 1990 1997 Maravatio-Zapotlanejo Toll roads Mexico 1,460 1992 1997 Monterrey-Nuevo Laredo Toll roads Mexico 293 1990 1997 Puente El Zacatal Toll roads Mexico 57 1994 1997 Saltillo-Torreón Toll roads Mexico 270 1993 1997 Canceled infrastructure projects with private participation in developing countries, Table 1990­2001 (continued) 1 Committed investment Year of (2001 US$ financial Year of Project Sector Country millions)a closure cancellation Bangkok Elevated Road and Toll roads Thailand 857 1990 1997 Train System Tucuman Water and sewerage Argentina 93 1995 1997 3 BA Provincial Water and Sewerage Water and sewerage Argentina 1,009 1999 2001 Cochabamba Water and Sewer System Water and sewerage Bolivia 340 1999 2000 Jiangsu Province Water Supply Potable water China 183 1996 1999 SOGEA Potable water Gambia 0 1993 1995 Indah Wastewater Urban Sewerage Sewerage Malaysia 2,858 1993 2000 Rehabilitation Kelantan Water Supply Potable water Malaysia 12 1995 1999 Total or average 24,237 1994 1998 a. Investment commitments may not be required for some management contracts (such as for Sizarail, SOGEA, and the Mombasa Container Terminal). Source: World Bank, PPI Project Database. headlines, but there has been little attempt to How many private infrastructure projects see how common it is. Do the highly publicized have been canceled? cancellations seen thus far represent wide- Based on the definition of cancellation, 48 proj- spread difficulties in sustaining private infra- ects were judged to have been canceled in structure? Do they herald a return to public 1990­2001. These projects represented only 1.9 provision? percent of the nearly 2,500 infrastructure proj- To assess how widespread the problem has ects with private participation that reached been, a working definition of cancellation was financial closure in developing countries during developed (box 1). This definition uses the cri- that period (see table 1 for a list of the canceled terion of whether the private sector continued projects). The canceled projects had attracted to be active in a project or not, rather than a investment commitments of US$24.2 billion, 3.2 strict legal definition of exit from a project. Thus percent of the total investment (US$754 billion) a project abandoned by a private company but in private infrastructure projects in developing subsequently revived by another would be con- countries in 1990­2001.3 On average, projects sidered canceled. But a project in which a pri- were canceled four and a half years after finan- vate company sold its interest to another without cial closure, relatively early in their life. a cessation in services or abandonment of the More than a third of the canceled projects project would not be. were from the Mexican toll road program. The definition was applied to the projects in Without this program, projects canceled in the World Bank's Private Participation in 1990­2001 would represent only 1.0 percent of Infrastructure (PPI) Project Database to see how the projects that reached financial closure over many were canceled during the period January the same period and 1.9 percent of the invest- 1990­December 2001. The database covers proj- ment commitments for those projects. The 19 ects that have reached financial closure and that canceled toll road projects accounted for about directly or indirectly serve the public in the trans- 5.8 percent of all projects in that sector (table 2). port, electricity, telecommunications, natural The water sector had the second highest rate gas (transmission and distribution), and water of cancellation, with 3.5 percent of projects can- and sewerage sectors in low- and middle-income celed, followed by electricity projects involving countries.2 distribution and sale to final consumers. Other I N F R A S T R U C T U R E P R O J E C T S A R E V I E W O F C A N C E L E D P R I V A T E P R O J E C T S Canceled infrastructure projects with private participation in developing countries, Table by sector, 1990­2001 2 Canceled projects Projects reaching Projects canceled as a share of total financial closure in 1990­2001 in 1990­2001 (percent) Committed Committed By investment investment By investment Sector Number (2001 US$ billions) Number (2001 US$ billions) number value 4 Energya 978 247.7 10 5.8 1.0 2.4 Natural gas 146 34.5 1 0.7 0.7 1.9 Electricity generation 600 150.3 5 4.7 0.8 3.1 Electricity distribution or integrated utilities 220 63.0 4 0.5 1.8 0.7 Telecommunications 650 331.4 8 1.4 1.2 0.4 Transport 662 135.3 23 12.5 3.5 9.3 Airports 82 12.5 1 0.1 1.2 1.1 Ports 177 18.0 2 0.2 1.1 1.0 Rail 76 28.8 1 --b 1.3 0.0 Toll roadsc 327 76.0 19 12.2 5.8 16.1 Water and seweraged 202 39.7 7 4.5 3.5 11.3 Total 2,492 754.1 48 24.2 1.9 3.2 a. Including 12 electricity and water projects. b. No investment commitments were made. c. Including the Bangkok Elevated Road and Train System. d. Including the SOGEA lease, which covered electricity and water. Source: World Bank, PPI Project Database. sectors, including electricity generation, had than 60 percent of projected traffic flows in its lower rates of cancellation. The eight canceled initial years of operation. projects in telecommunications represented In some cases a government's willingness to only a very small share of projects and invest- assume traffic risks may have led to less investor ment commitments in that sector. scrutiny of demand forecasts and so perhaps increased the probability that the project Why are projects canceled? would fall short of projections. The Mexican Many factors that led to the cancellation of a government offered indirect guarantees to the project were specific to the investors, govern- investors and lenders funding the private toll ments, and regulators involved in that project. road program, which may have led private But in each sector some factors appeared in operators and banks to undertake some proj- many of the cancellations. ects that they might otherwise have turned In transport most of the canceled toll road down. projects saw the exit of the private sector because Most water and sewerage projects that were the roads could not attract enough users to meet canceled confronted controversies over price the optimistic traffic forecasts. Consumers were increases and difficulties in collecting from con- often less willing than had been expected to pay sumers. Public water utilities had generally kept for the right to use the toll roads, sometimes prices below costs and had low collection rates. because the effect of alternative toll-free routes Attempts to raise prices or increase collections had been underestimated. For example, more with the shift to private participation led to oppo- than half the Mexican toll roads reached less sition from some consumers and politicians. than 50 percent of the forecasted volumes, and Sometimes project design worsened the situ- the M1/M15 toll road in Hungary achieved less ation by magnifying the adjustments required. In the Cochabamba water concession in Bolivia, Implications for private provision of for example, the local government opted for a infrastructure high-cost bulk water source that exacerbated We should expect to see some cancellations of the need to raise previously subsidized tariffs. private infrastructure projects. Much of the When tariffs were increased by about 35 percent rationale for moving away from public provision almost immediately after privatization, the proj- was based on the assessment that public enter- ect ran into widespread opposition. Moreover, prises did not face real commercial disciplines. all this took place against the backdrop of The "freedom to fail" provides incentives for the broader political opposition to irrigation private sector to be efficient. 5 reforms and the government's coca eradication The projects canceled thus far represent only policy, both of which contributed to the civil dis- a small share of the projects that have encoun- turbances that preceded the cancellation of the tered problems. Most problems are solved by project. adjusting key terms, by renegotiating contracts, Projects canceled in the electricity sector, like or through other means short of cancellation. those in the water sector, had difficulties in Even where substantial macroeconomic shocks enforcing and maintaining cost recovering pric- occurred, most private infrastructure projects ing policies and problems in collecting pay- successfully withstood the impacts. ments owed by consumers or government The small number of cancellations reflects the off-takers. These problems afflicted both distri- incentives for both the government and the pri- bution and generation projects. vate sector to remain in projects rather than walk In some cases, such as with independent away. The private sector may have invested in power producers in Indonesia, macroeconomic sunk assets and thus may be willing to accept shocks may have increased the difficulty of some changes in contractual terms. Govern- implementing reforms and accelerated the fail- ments may want to avoid cancellation because of ure of projects. These shocks led to a contrac- the substantial payments often required in com- tion in real incomes, reducing demand for pensation for breach of contract (including can- infrastructure services. And because most proj- cellation) and because of concerns about service ects were funded by foreign currency loans, the continuity following the exit of the private sector. shocks led to higher debt service costs in local The small number of canceled projects, the currency terms. attempts by governments to reprivatize some of In contrast, the projects canceled in telecom- them, and new private projects in countries that munications saw relatively few disputes over pric- have seen cancellations all suggest that many ing or collection. Instead, most of the projects, governments still view the public sector as less involving cellular services in markets with alter- effective than the private sector in providing native suppliers, were canceled because they infrastructure services. But we may see more failed to attract sufficient customers or because canceled projects in the near future. Many of the government decided to change the market the most active infrastructure operators in structure. developing countries face financial pressures, Some infrastructure projects had major calling into question their ability to continue to problems in the bidding phase. At least four support temporarily unprofitable ventures. And projects were won through high bids for con- the macroeconomic crisis in Argentina, a pio- cession or license fees--but once in operation neer in private participation in infrastructure, failed to provide revenues sufficient to both pay has put many of the private infrastructure canon fees and meet investment obligations. arrangements in that country under great stress. Where projects were not competitively ten- The factors that lead to disputes over proj- dered, the case for more than half the 48 can- ects, and in some cases to cancellation, illustrate celed projects, political and social opposition the complexities and challenges of placing seems to have focused on the lack of trans- infrastructure provision on a commercial foot- parency in the award, with frequent allegations ing. The politics of this reform have proved dif- of corruption and impropriety. ficult in many countries, leading to some of the I N F R A S T R U C T U R E P R O J E C T S A R E V I E W O F C A N C E L E D P R I V A T E P R O J E C T S reversals that we have seen. Ensuring trans- parency in the award and regulation of projects and building a consensus for reform will help reduce opposition to private provision of ser- vices. But difficult adjustments may be unavoid- viewpoint able in some sectors. In the energy and water sectors consumers have historically paid much less for services than it costs to provide them. is an open forum to One way or another someone--whether con- encourage dissemination of sumers or taxpayers--has to pay for these ser- public policy innovations for vices. Governments that do not have strong private sector­led and fiscal positions will be unable to subsidize these market-based solutions for services for most consumers, although they development. The views might be able to do so for particular groups, published are those of the such as the poor. But several steps can be taken authors and should not be to avoid tariff shocks--phasing in price attributed to the World increases, making judicious use of transitional Bank or any other affiliated subsidies, and ensuring that investment obliga- organizations. Nor do any of tions that must be financed by user fees require the conclusions represent only realistic increases in prices. official policy of the World Bank or of its Executive Directors or the countries they represent. Notes 1. Luis J. Guasch, Jean Jacques Laffont, and Stephane To order additional copies Straub, "Renegotiation of Concession Contracts in Latin contact Suzanne Smith, America" (World Bank, Washington, D.C., 2002). managing editor, Room I9-009, 2. For more information on the database, see http:// The World Bank, www.worldbank.org/privatesector/ppi/ppi_database.htm. 1818 H Street, NW, 3. Investment data are from the World Bank's PPI Washington, DC 20433. Project Database. The database records total investment (privatization revenues and license or canon fees), not pri- Telephone: vate investment alone, on a commitment basis in the year 001 202 458 7281 of a project's financial closure. Actual disbursements are Fax: not tracked. All investment figures cited here are in 2001 001 202 522 3480 U.S. dollars. Email: ssmith7@worldbank.org Copyedited and produced by Communications Development Inc. Printed on recycled paper T h i s N o t e i s a v a i l a b l e o n l i n e : h t t p : / / r r u . w o r l d b a n k . o r g / V i e w p o i n t / i n d e x . a s p