mw. \21476 [O wnpoint Private Infrastructure Q, Melissa Houskamp and Are the Trends in Low-income Countries Different? Nicola 7ynan The PPI Project Database This Note, based on the World Bank's Private Participation in tracks inorastnctureprnoects Infirastructure (PPl) Project Database, reviews trends in infrastruc- newvlv owned or managed by ,,,uc pnivate companies that ture projects with private participation in low-income countries. Four reached financial closure in reacd inanca rn main conc]usions arise. Surprisingly, the proportion off countries with *: 1990-99 in energy (electnic- ity and naturalgas trans- at least one project-81 percent-is higher among low-income than misseon andmdistribution) middle-income countries. As in middle-income countries, most invest- telecommunications, tratis- port, and water. Seepage2 ment has been in telecommunications or energy projects. However, in for more information on the low-income countries, well over half the projects are greenfield. And database. the scale off private participation in low-income countries lags far behind that in middle-income countries (figure I). Since 1990, a growing number of low-income Cumulative investment in infrastructure with LW developing countries have encouraged private private participation in developing countries, = operators in infrastructure.' Between 1990 and [ 1990-99 1999, the proportion of low-income countries with * Energy OTelecom jTransport jWater at least one private infrastructure project grew lU5$ billions O from nearly 20 percent to more than 80 percent- Income level 0 50 100 150 200 50 countries (table 1), exceeding the percentage of lower-middle-income countries (77 percent). : Low Z Investment in projects with private participation I in low-income countries rose almost every year _ during the 1990s and peaked in 1997 at US$33.6 Lower-middle E- I -' billion, almost as much as the US$35.1 billion O invested in lower-middle income country projects Upper-middle thatyear (figure 2). m WU After 1997, investments fell, mainly because of the financial crisis in East Asia. Although Sure: PPI Project atabase. PRIVATE INFRASTRUCTURE: ARE LOW-INCOME COUNTRIES DIFFERENT? Database coverage: * Divestiture. A private consortium buys an equiq stake in a . Projects that have reached financial closure and directly or indirectly state-owned enterprise. The private stake may or may not imply serve the public. private management of the company. . Projects in electricity, natural gas (transmission and distribu- tion), telecommunications, transport, and water sectors, but not Definition of financial closure. For greenfield projects, movable assets, incinerators, stand-alone solid waste projects, and and for operations and management contracts with major capital small projects such as windmills. expenditure, financial closure is defined as existence of a legally 2 . Low- and middle-income developing countries in 1999, as binding commitment of equity holders or debt financiers to pro- defined and classified by the World Bank. vide or mobilize funding for the project. The funding must account for a significant part of the project cost, securing the Definition of private participation. The private com- construction of the facility. For operations and management con- pany must assume operating risk during the operating period or tracts, a lease agreement or a contract authorizing the com- assume development and operating risk during the contract mencement of management or lease service must exist. For period. A foreign state-owned company is considered a private divestitures, the equity holders must have a legally binding com- entiq. mitment to acquire the assets of the facility. Definition of a project unit. A corporate entity created Recording investments. Investments, privatization rev- to operate infrastructure facilities is considered a project. When enues, license fees, and canon commitments generally have been two or more physical facilities are operated by the corporate recorded on a commitment basis in the year of financial closure entity, all are considered as one project. (for which data are typically readily available). Actual disburse- ments have not been tracked. Where privatizations and new Project types investments are phased and data were available at financial clo- a Operations and management contract. A private entity takes sure, they are recorded in phases. When license fees and canon over the management of a state-owned enterprise for a given commitments were due over the concession period, their net pres- period. This category includes management contracts and leases. ent values were recorded in the year of financial closure. * Operations and management contract with major capital expenditure. A private entity takes over the management of a Sources: World Wide Web, commercial databases, specialized pub- state-owned enterprise for a given period during which it also lications, developers, sponsors, and regulatory agencies. assumes significant investment risk. This category includes concession-type contracts such as build-transfer-operate, build- Web site: http://www.worldbank.org/html/fpd/privatesector/ lease-operate, and build-rehabilitate-operate-transfer contracts as PPIDBweb/lntro.htm applied to existing facilities. a Greenfield project. A private entiq or a public-private joint Contact: The database is maintained by the Private Provision of venture builds and operates a new facility. This category includes Public Services Group of the World Bank. For more information, build-own-transfer and build-own-operate contracts as well as contact Shokraneh Minovi at 202 473 0012 or sminovi@ merchant power plants. worldbank.org. Latin Middle East Asia Europe and America East and Income Sub-Saharan and Central and the North South All group Africa Pacific Asia Caribbean Africa Asia regions Low 76 Io0 83 100 100 67 81 Lower-middle 67 70 86 81 67 100 77 Upper-middle 80 50 88 73 80 N/A 76 All income groups 76 77 86 79 73 75 78 N/A Not applicable. Source: PPI Project Database. Total investment in infrastructure projects with private participation in developing countries, Figure by income group, 1990-99 2 1999 U5$ billions - Upper-middle income 280 - 8 - Lower-middle income 70 6 All low-income 60 countries 50 - Low-income 40 countries excluding China and India 3 30 - - - Low-income countries -/ , \: excuding China, India, 20 - Indonesia, and Pakistan 10 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Source: PPI Project Daiabase. investment in projects with private participation in nications and energy. If all developing countries upper-middle income countries continued to rise with at least one project are taken as a group, in 1998, partly lifted by the major telecommunica- telecommunications receives a higher percentage tions privatization in Brazil, by 1999 investment of total investment than energy. In low-income was below its peak for all income groups. countries, however, the percentage invested in energy is higher (46 percent) than telecommuni- Greenfield projects predominate cations, and in lowest-income countries it is higher Greenfield projects predominate in developing again at 50 percent (see figure 1). countries, particularly in lowest-income countries. Investment in the energy sector in low-income They account for 65 percent of projects in low- countries increased rapidly during the 1990s, income countries compared with 37 percent for exceeding that in lower-middle-income countries developing countries as a whole, mainly because but not in upper-middle-income countries. Only low-income countries have so little infrastructure nine lowest-income countries have energy projects in place. with private participation, even though energy Greenfield projects are common in the receives 50 percent of investment in this income telecommunications sector, especially for new group. The number of low-income countries imple- wireless technologies, and the energy sector, menting private telecommunications projects rose where non-sovereign guarantees encourage pri- from 3 in 1990 to 40 (65 percent of this group) in vate investment in new infrastructure. In low- 1999. Of the 29 lowest-income countries, 19 had pri- income countries, only 21 percent of projects vate telecommunications projects by 1999. Most of involve operations and management contracts these projects use low-orbit satellites and wireless with major capital expenditure-a small propor- technology. tion compared to upper-middle-income Private water and sanitation projects have been countries-and more than half such projects are rare in low-income countries: 30 such projects were in China. There are also fewer divestitures and implemented in six countries (four in Africa, two operations and management projects in low- of which are in the lowest-income group) in income countries, though twice as many low- 1990-99 with a total investment of US$1.8 billion income as upper-middle-income countries have at (only 2 percent of projects with private participa- least one operations and management project. tion in low-income countries). Most of the earliest private water projects were operations and man- Fastest growing sectors agement contracts involving little or no investment In low-income and middle-income countries, pri- risk on the part of the private operator. (Private vate participation is concentrated in telecommu- operators play a significant role in small-scale water PRIVATE INFRASTRUCTURE: ARE LOW-INCOME COUNTRIES DIFFERENT? and sanitation projects in a number of developing gate. China, India, Indonesia, and Pakistan all countries, but because they do not operate under have projects with private participation in at least formal government contracts and they are small three infrastructure sectors (India and Pakistan do they are not included in the PPI Project Database.) not have projects in water and sanitation) and In the transport sector, 20 low-income coun- together account for 91 percent of investment in viewpoint tries (32 percent) implemented over 190 projects low-income countries. Countries that have experi- with total investment of US$23 billion between mented with private participation in one infra- 1990 and 1999. Eighteen were in lowest-income structure sector now need to allow greater private is an open forum to countries with a total investment of US$594 mil- entry to reap the efficiency and service rewards encourage dissemination of lion. China, with 116 projects, accounts for 75 per- being reported for projects in many developing public policy innovations for cent of private transport investment in low-income countries. private-sector led and countries (US$17.2 billion). market-based solutions for development. The views Four countries dominate published are those of the Countries in Latin American and the Caribbean Notes authors and should not be invested nearly U$300 billion in projects with pri- 1. Income groups are defined in World Bank 2000 in attributed to the World vate participation from 1990 to 1999. Only three of terms of 1998 GNP per capita: low-income, US$760 or Bank or any other affiliated the countries involved were low-income and none less; lower-middle income, from US$761 to US$3,030; organizations. Nor do any of were lowest-income. By contrast, Sub-Saharan upper-middle income, from US$3,031 to US$9,360; and the conclusions represent Africa has the largest number of low-income high income, US$9,361 or more. official policy of the World countries (78 percent are low-income and 43 per- This Note defines another group-lowest-income Bank or of its Executive cent are lowest-income) and received only 2 countries-as those whose 1998 per-capita GNP was less Directors or the countries percent of investment. Nevertheless, three-quarters than US$365. They are: Angola, Bangladesh, Burkina Faso, they represent. of low-income countries in Sub-Saharan Africa have Burundi, Cambodia, Central African Republic, Chad, implemented at least one project since 1990 (see Democratic Republic of the Congo, Eritrea, Ethiopia, To order additional copies table 1). Kenya, Kyrgyz Republic, Lao People's Democratic contact Suzanne Smith, About two-thirds of investment in low-income Republic, Madagascar, Malawi, Mali, Mozambique, Nepal, managing editor, countries went to the two largest countries in the Niger, Nigeria, Rwanda, Sierra Leone, Tajikistan, Room 19-216, group, China and India. Figure 2 compares the Tanzania, Togo, Uganda, Vietnam, Yemen Republic, and The World Bank, trends in investment in 1990-99 for the whole Zambia. For a full breakdown of all countries by income 1818HStreet,NW, group and for the group excluding China and group classification, see World Bank 2000, pages 290-91. Washington, D.C. 20433. India. The difference is particularly striking in 1997, when investment continued in China but fell References eo 1 202 458 7281 in other East Asian countries because of the finan- WAlorld Bank. 2000. Entering the 21st Century: World Fax: cial crisis. Investment in China and India fell Development Report 1999-2000. New York: Oxford 001 202 522 3181 sharply during 1998, but started to rise again in University Press. Email: 1999. Two other low-income countries, Indonesia ssmith7@worldbank.org and Pakistan, also exhibit levels and types of invest- Melissa Houskamp (houskamp_melissa@jpmorgan.com), JR ment similar to middle-income countries. These Morgan, New York, and Nicola Tynan (ntynan@gmu.edu), four countries are among the ten developing George Mason University, Virginia. Printed on recycled paper countries, of all income groups, with the highest investmenit in projects with private participation in 1990-99. U Conclusion Most low-income countries have some form of pri- vate participation in at least one sector, and the proportion is even higher among lowest-income countries. Four low-income countries stand out as being most like developing countries in the aggre- This Note is available online: www.worldbank.org/html/fpd/notes/