infrastructure: achievements, challenges, and opportunities · infrastructure: achievements,...

24
I Infrastructure: achievements, challenges, and opportunities Infrastructure servicesincluding power, trans- port, telecommunications, provision of water and sanitation, and safe disposal of wastesare central to the activities of households and to economic pro- duction. This reality becomes painfully evident when natural disasters or civil disturbances destroy or disable power stations, roads and bridges, tele- phone lines, canals, and water mains. Major in- frastructure failures quickly and radically reduce communities' quality of life and productivity. Con- versely, improving infrastructure services enhances welfare and fosters economic growth. Providing infrastructure services to meet the demands of businesses, households, and other users is one of the major challenges of economic devel- opment. The availability of infrastructure has in- creased significantly in developing countries over the past several decades. In many cases, however, the full benefits of past investments are not being realized, resulting in a serious waste of resources and lost economic opportunities. This outcome is frequently caused by inadequate incentives embod- ied in the institutional arrangements for providing infrastructure services. While the special technical and economic characteristics of infrastructure give government an essential role in its provision, domi- nant and pervasive intervention by governments has in many cases failed to promote efficient or re- sponsive delivery of services. Recent changes in thinking and technology have revealed increased scope for commercial principles in infrastructure provision. These offer new ways to harness market forces even where typical competition would fail, and they bring the infrastructure user's perspective to the forefront. This Report focuses on economic infrastructure: the long-lived engineered structures, equipment, and facilities, and the services they provide that are used in economic production and by households. This infrastructure includes public utilities (power, piped gas, telecommunications, water supply, sani- tation and sewerage, solid waste collection and dis- posal), public works (major dam and canal works for irrigation, and roads), and other transport sec- tors (railways, urban transport, ports and water- ways, and airports). Social infrastructure, often en- compassing education and health care, represents an equally important although very different set of issues that are not analyzed in this Report (see World Development Report 1993: Investing in Health). As defined here, infrastructure covers a complex of distinct sectors that, by any measure, represent a large share of an economy. Taken togethei the ser- vices associated with the use of infrastructure (mea- sured in terms of value added) account for roughly 7 to 11 percent of GDP (Table 1.1), with transport being the largest sector. Transport alone commonly absorbs 5 to 8 percent of total paid employment. A sample of developing countries shows that infra- Table 1.1 Value added of infrastructure services by country group (percentage of GDP) Transport, Note: At market prices. At factor cost (for which fewer observations are available), the values are slightly higher. Figures in parentheses are number of observations. Data are for 1990 or latest available year Source: World Bank national accounts data. 13 storage, and 5.34 6.78 9.46 communications (9) (26) (3) Gas, electricity, and water 1.29 (22) 2.24 (36) 1.87 (5) Low-income Middle-income High-income Sector countries countries countries

Upload: dangcong

Post on 27-Jul-2018

266 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

IInfrastructure: achievements,challenges, and opportunities

Infrastructure servicesincluding power, trans-port, telecommunications, provision of water andsanitation, and safe disposal of wastesare centralto the activities of households and to economic pro-duction. This reality becomes painfully evidentwhen natural disasters or civil disturbances destroyor disable power stations, roads and bridges, tele-phone lines, canals, and water mains. Major in-frastructure failures quickly and radically reducecommunities' quality of life and productivity. Con-versely, improving infrastructure services enhanceswelfare and fosters economic growth.

Providing infrastructure services to meet thedemands of businesses, households, and other usersis one of the major challenges of economic devel-opment. The availability of infrastructure has in-creased significantly in developing countries overthe past several decades. In many cases, however,the full benefits of past investments are not beingrealized, resulting in a serious waste of resourcesand lost economic opportunities. This outcome isfrequently caused by inadequate incentives embod-ied in the institutional arrangements for providinginfrastructure services. While the special technicaland economic characteristics of infrastructure givegovernment an essential role in its provision, domi-nant and pervasive intervention by governmentshas in many cases failed to promote efficient or re-sponsive delivery of services. Recent changes inthinking and technology have revealed increasedscope for commercial principles in infrastructureprovision. These offer new ways to harness marketforces even where typical competition would fail,and they bring the infrastructure user's perspectiveto the forefront.

This Report focuses on economic infrastructure:the long-lived engineered structures, equipment,

and facilities, and the services they provide that areused in economic production and by households.This infrastructure includes public utilities (power,piped gas, telecommunications, water supply, sani-tation and sewerage, solid waste collection and dis-posal), public works (major dam and canal worksfor irrigation, and roads), and other transport sec-tors (railways, urban transport, ports and water-ways, and airports). Social infrastructure, often en-compassing education and health care, representsan equally important although very different set ofissues that are not analyzed in this Report (see WorldDevelopment Report 1993: Investing in Health).

As defined here, infrastructure covers a complexof distinct sectors that, by any measure, represent alarge share of an economy. Taken togethei the ser-vices associated with the use of infrastructure (mea-sured in terms of value added) account for roughly 7to 11 percent of GDP (Table 1.1), with transportbeing the largest sector. Transport alone commonlyabsorbs 5 to 8 percent of total paid employment. Asample of developing countries shows that infra-

Table 1.1 Value added of infrastructure servicesby country group(percentage of GDP)

Transport,

Note: At market prices. At factor cost (for which fewer observations areavailable), the values are slightly higher. Figures in parentheses arenumber of observations. Data are for 1990 or latest available yearSource: World Bank national accounts data.

13

storage, and 5.34 6.78 9.46communications (9) (26) (3)

Gas, electricity,and water

1.29

(22)

2.24

(36)

1.87

(5)

Low-income Middle-income High-incomeSector countries countries countries

Page 2: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

Figure 1.1 Public infrastructure investmentis a large fraction of both total and publicinvestment in developing countries.

Percentage of investment allocatedto infrastructure60

50

40

30

20

10

Total investment Public investment

Low-income countries

Middle-income countries

Sample: Twelve low-income and eight middle-incomecountries; unweighted averages, 1980-89.Source: Easterly and Rebelo 1993.

structure typically represents about 20 percent oftotal investment and 40 to 60 percent of public in-vestment (Figure 1.1). In round figures, public infra-structure investment ranges from 2 to 8 percent (andaverages 4 percent) of GDP. Even these shares un-derstate the social and economic importance of in-frastructure, which has strong links to growth,poverty reduction, and environmental sustainability.

Infrastructure's impact on development

Links to economic growth

Infrastructure represents, if not the engine, then the"wheels" of economic activity. Input-output tablesshow that in the economies of Japan and the UnitedStates, for example, telecommunications, electricity,and water are used in the production process of

14

nearly every sector, and transport is an input forevery commodity. Users demand infrastructure ser-vices not only for direct consumption but also forraising their productivity by, for instance, reducingthe time and effort needed to secure safe water, tobring crops to market, or to commute to work.

Much research in recent years has been devotedto estimating the productivity of infrastructure in-vestments (Box 1.1). Many studies attempting tolink aggregate infrastructure spending to growth ofGDP show very high returns in a time-series analy-sis. Some cross-national studies of economic growthand infrastructure notably, one using public in-vestments in transport and communications and an-other using capital stocks in roads, railways, andtelephonesalso show that infrastructure variablesare positively and significantly correlated withgrowth in developing countries. In both types ofstudies, however, whether infrastructure invest-ment causes growth or growth causes infrastructureinvestment is not fully established. Moreover, theremay be other factors driving the growth of bothGDP and infrastructure that are not fully accountedfor. Neither the time-series nor the cross-sectionalstudies satisfactorily explain the mechanismsthrough which infrastructure may affect growth.

Sectoral studies focusing on rural infrastructure'seffect on the local economy in certain developingcountries have revealed more about the nature ofthe apparent benefits. Studying data over time fromeighty-five districts in thirteen Indian states, re-searchers found that lower transport costs increasedfarmers' access to markets and led to considerableagricultural expansion and that modern irrigationmethods brought higher yields. At the same time,because improved communications (through roads)lowered banks' costs of doing business, banks ex-panded lending to farmers, and farmers used thefunds to buy fertilizer, further increasing yields. Ac-cording to a household- and village-level surveyconducted in Bangladesh, villages classified as"most developed" in terms of access to transport in-frastructure were significantly better off than the"less developed" villagesin terms of agriculturalproduction, incomes and labor demand, and health.(It is difficult, however, to verify whether theBangladesh study took into account all possible in-tervening factors, such as unobserved differencesamong the communities in natural endowments.)

What is evident is that a strong association existsbetween the availability of certain infrastructure -telecommunications (in particular), power, pavedroads, and access to safe waterand per capitaGDP (Figure 1.2). An analysis of the value of infra-

Page 3: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

Box 1.1 Returns on infrastructure investmenttoo good to be true?

Recent studies in the United States suggest that the im-pact of infrastructure investments on economic growthrepresents startlingly high rates of return (up to 60 per-cent). Too good to be true? Possibly. The results presentedin Box table 1.1 may overestimate the productivity of in-frastructure for two reasons. First, there may be a com-mon factor that causes growth in both output and infra-structure that is not included in the study. Second, it maybe that growth leads to infrastructure investment, andnot that investment produces growth. A number of stud-ies have found that causation runs in both directions. Yetmore sophisticated estimates that address these issues ei-ther have concluded that the positive results were notmuch affected by different econometric methods or havefound no noticeable impact of infrastructure on growth.Neither findingof an extremely high impact or of anegligible impactis entirely credible, and research ef-forts continue in an attempt to refine the methodology.

An alternative approach estimates the impact of in-

Box table 1.1 Results from studies of infrastructure productivity

United StatesUnited States48 states, United States5 metro areas, United StatesRegions, JapanRegions, FranceTaiwan, China

Korea

Israel

Mexico

Multicountry, OECDMulticountry, developingMulticountry, OECD

and developingMulticountry, developing

0.39 600.34 60

0 00.08 -0.20 960.08 12

0.24 77

0.19 51

0.31-0.44 54-70

0.05 5-7

0.070.07

0.01-0.16

frastructure on production costs. Studies (summarizedin Aschauer 1993) found that infrastructure significantlyreduces production costs in manufacturing in Germany,Japan, Mexico, Sweden, the United Kingdom, and theUnited States. One estimate suggests that three-quartersof U.S. federal investment in highways in the 1950s and1960s can be justified on the basis of reductions in truck-ing costs alone.

While there is still no consensus on the magnitude oron the exact nature of the impact of infrastructure ongrowth, many studies on the topic have concluded thatthe role of infrastructure in growth is substantial, signifi-cant, and frequently greater than that of investment inother forms of capital. Although the indications to dateare suggestive, there is still a need to explain why thefindings vary so much from study to study. Until thisproblem is resolved, results are neither specific nor solidenough to serve as the basis for designing policies for in-frastructure investment.

Aschauer 1989Munnell 1990Holtz-Eakin 1992Duffy-Deno and Eberts 1991Mera 1973Prud'homme 1993Uchimura and Gao 1993

Uchimura and Gao 1993

Bregman and Marom 1993

Shah 1988, 1992

19 Canning and Fay 199395 Canning and Fay 1993- Baffes and Shah 1993

0.16 63 Easterly and Rebelo 1993Percentage changes in output with respect to a 1 percent change in the level of infrastructure.Ratio of discounted value of increase in dependent variable to discounted value of investment in infrastructure.

Nonmilitary public capitalNonmilitary public capitalPublic capitalPublic capitalIndustrial infrastructurePublic capitalTransportation, water, and

communicationTransportation, water, and

communicationTransportation, power, water, and

sanitationPower, communication, and

transportationTransportationTransportationInfrastructure capital stocks

Transportation and communication

structure stocks indicates that their compositionchanges significantly as incomes rise. For low-in-come countries, more basic infrastructure is impor-tant such as water, irrigation, and (to a lesser ex-tent) transport. As economies mature into themiddle-income stage, most of the basic consump-tion demands for water are met, the share of agri-culture in the economy shrinks, and more transportinfrastructure is provided. The share of power andtelecommunications in investment and infrastruc-

ture stocks becomes even greater in high-incomecountries. Data for 1990 indicate that, while total in-frastructure stocks increase by 1 percent with each1 percent increment in per capita GDP, householdaccess to safe water increases by 0.3 percent, pavedroads increase by 0.8 percent, power by 1.5 percent,and telecommunications by 1.7 percent.

These relationships suggest that infrastructurehas a high potential payoff in terms of economicgrowth, yet they do not provide a basis for prescrib-

15

Implied rateSample Elasticity' of returnb Author/year Infrastructure measure

Page 4: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

Figure 1.2 Per capita availability of major infrastructure is closely related to income levels.

Telephone main linesper thousand persons500

100

20

5

1

1,000

Jordan.*

Bolivia..Honduras I

Zambia .

Malawi' .- . Rwanda

Chad

CongoS

Indonesia

Bangladesh

:'Gabon

250 400 700 1,000 3,000 5,000 8,000GDP per capita (PPP dollars)

Kilometers of paved roadsper million persons3,000

Zimbabwe.

Zambia

Mali ? Dominican Republic

250Peru Colombia

S Bolivia

Algeria

Morocco.

20250 400 700 1,000 3,000 5,000 8,000

GDP per capita (PPP dollars)

Middle East and North AfricaLatin America and the CaribbeanEast Asia and Pacific

Uruguay

Mexico

Percentage of householdswith electricity100

50 Senegal Costa Rica VenccuelaNigeria. S Peru ,s

. Guatem._%exicoS El Salvador

BoliviaMalawi Honduras Indonesia

50

20

Mozambique.

Burundi

Percentage of populationwith access to safe water100

Congo

250 400 700 1,000 3,000 5,000 8,000GDP per capita (PPP dollars)

5250 400 700 1,000 3,000 5,000 8,000

GDP per capita (PPP dollars)

Sub-Saharan AfricaSouth AsiaEurope and Central Asia

Note: Axes are logarithmic; infrastructure quantities and GDP are for 1990; purchasing power parity (PPP) dollars are valuedin Summers and Heston 1985 international prices.Source: WDI table 32; Summers and Heston 1991.

ing appropriate levels, or sectoral allocations, for in-frastructure investment. Other evidence confirmsthat investment in infrastructure alone does notguarantee growth. Many studies reveal muchsmaller returns for infrastructure than those sug-gested in Box 1 .1closer, in fact, to the return onprivate investments. These disparities may be due to

16

differences in the efficiency of investment acrosscountries and over time. For example, a study of theeconomic returns to individual World Bank projectsshows that, when overall economic policy condi-tions are poor, the returns to infrastructure invest-ment decline. Returns are lower by 50 percent ormore in countries with restrictive trade policies than

Chad S Bangladesh

Burkina Faso

Indonesia

1001 Sudan

10

0.5

Page 5: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

in countries where conditions are more favorable.Infrastructure spending cannot, therefore, overcomea weak climate for economic activity. Nearly twenty-five years ago, the Brookings Transport ResearchProject evaluated the impact of transport projects inseveral developing countries and concluded simi-larly that, although the investments generally hadreasonable rates of return, success depended largelyon economic policy.

Another approach to assessing the economic re-turns from infrastructure investment is to examinethe rates of return in a large sample of completedWorld Bank projects. The average economic returnon infrastructure projects, reestimated after loan dis-bursement (completion of project construction), hasbeen 16 percent over the past decadejust abovethe World Bank project average of 15 percent (Table1.2). Returns have been lowest (and declining) for ir-rigation and drainage, airports (for a very smallsample), railways, power, water supply, and sewer-age. Why should this be so, given the expected ben-efits of such investments in developing countries?

Some of the causes relate to implementationproblems (discussed below under "The record ofperformance") and others to project identificationand design. A common pattern discovered in proj-ect completion reviews of water, railway, and powerprojects is the tendency at the time of appraisal tooverestimate the rate of growth in demand for newproduction capacity and, therefore, of revenues. Forthe power projects in the sample, demand was over-estimated by 20 percent on average over a ten-yearoperating period. In water projects, overestimationof rates of new connections and per capita con-sumption also averaged about 20 percent. In thecase of railways, until recent years projects often as-sumed recovery in demand even where railwayswere continually losing traffic to roads offering bet-ter service. In twenty-nine of thirty-one cases,freight traffic failed to reach its projected level, andin one-third, traffic actually declined.

One important explanation for the misjudgmentsduring appraisal is inadequate procedures for as-sessing demand (including the effects of tariff in-creases). Oversizing and inappropriate design of in-vestments then occur, resulting in financial burdenson the project entities concerned. Although Bankprojects may not be entirely representative, they aresubject to more careful evaluation than many infra-structure investments in developing countries andso may have achieved better performance than av-erage public investments in these sectors.

Infrastructure is a necessary, although not suffi-cient, precondition for growthadequate comple-

Table 1.2 Average economic rates of returnon World Banksupported projects, 1974-92

Not available.a. Rates are financial, not economic, rates of return.Source: World Bank data.

ments of other resources must be present as well.The growth impact of infrastructure investmentsalso depends on the timing and location of addi-tions to capacity, and on the existing imbalance be-tween supply and demand. Because much infra-structure consists of networks, relieving bottlenecksat certain points of the system can produce veryhigh returns. Box 1.2 illustrates the repercussions inChina's economy from critical constraints in thetransport of coal needed for power generation.

Adequate quantity and reliability of infrastruc-ture are key factors in the ability of countries tocompete in international trade, even in traditionalcommodities. In part because of infrastructure prob-lems, shipping costs from Africa to Europe are 30percent higher for plywood (and 70 percent higherfor tuna) than those from Asia to Europe. Thesecosts have to be borne by exporters.

The competition for new export markets is espe-cially dependent on high-quality infrastructure.During the past two decades, increased globaliza-tion of world trade has arisen not only from the lib-eralization of trade policies in many countries butalso from major advances in communications, trans-port, and storage technologies. These advances cen-ter on the management of logistics (the combinationof purchasing, production, and marketing func-tions) to achieve cost savings in inventory andworking capital and to respond more rapidly to cus-tomer demand. About two-thirds of production andsales in the OECD countries are processed directly

17

(percent)

Sector 1974-82 1983-92

Irrigation and drainage 17 13

Telecommunications 20 19

Transport 18 21

Airports 17 13

Highways 20 29

Ports 19 20Railways 16 12

Power 12 11

Urban development 23Water and sanitationa 7 9

Water supply a 8 6Seweragea 12 8

Infrastructure projects 18 16

All Bank operations 17 15

Page 6: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

Box 1.2 The importance of infrastructure to economic development: an examplefrom China

The fact that infrastructure provides critical support tothe growth of an economy can be clearly seen when bot-tlenecks arise. One of the most striking examples is thatof China's intercity transport system, with its links to thesupply of raw materials, coal, and electricity.

The coverage of China's intercity transport networksis one of the thinnest in the world: the total route lengthper capita or per unit of arable landfor highways orrailwaysis similar to, or lower than, that in Brazil,India, and Russia. This has resulted mainly from chronicunderinvestment in China's transport infrastructure.China's transport investments amounted to only 1.3 per-cent of GNP annually during 1981-90, a period of rapidgrowth in transport demand.

Since the onset of China's open door policy in 1979,economic growth averaging 9 percent a year has resultedin an unprecedented expansion in intercity trafficwithgrowth averaging 8 percent a year for freight and 12 per-cent a year for passengers. This traffic growth has im-posed tremendous strains on the transport infrastruc-

ture, as manifested by the growth of bottlenecks in therailway network, the severe rationing of transport capac-ity on railway lines, and the poor quality of service expe-rienced by shippers and passengers.

Transport shortages have adversely affected the sup-ply of coal in particular. Coal is the source of some 73percent of China's commercial energy and representsabout 43 percent of the total tonnage of freight handledby the railways. The shortage of coal has in turn ad-versely affected supplies of electricity, about 76 percentof which is generated by thermal plants. In 1989, Chinawas experiencing a shortfall in available power of about20 percent of industrial electricity requirements. Centraland local authorities established quotas for allocatingelectricity and rationed new connections, but power cutshave nevertheless been frequent.

A conservative estimate is that the annual economiccosts of not having adequate transport infrastructure inChina during the past several years amount to about 1percent of China's GNP.

to order, and "just-in-time" delivery of products hasbecome the norm in many sectors. Because about 60percent of their exports are directed to OECD mar-kets, developing countries must meet these stan-dards. Virtually all the improved practices designedto reduce logistics costs, including those in trans-port, have been based on information technologiesusing telecommunications infrastructure. Cost re-ductions and the increased speed of freight move-ments over the past few decades have also been in-creasingly based on multimodal transport involvingcontainerization, which requires intensive coordina-tion by shippers across rail, port, air, and roadfreight modes.

For developing countries wishing to compete inglobal markets, or to participate in "global sourc-ing" (the linking of businesses in several countriesproducing different components for a final prod-uct), not just any kind of transport and telecommu-nications infrastructure will do. Manufacturing as-sembly operations in Mexico and horticulturalexports from Kenya are examples of the diversifica-tion of trade permitted by appropriate logisticalsupport and multimodal facilities. During the 1980s,the proportion of garments, shoes, and handicraftexports shipped by air from northern India quintu-pled because land and ocean transport systemswere no longer able to meet demanding delivery re-quirements. Because India's ports have been slow to

18

adapt to containerization and are subject to regula-tory delays, freight transport to the United States isone-third more expensive from Indian ports thanfrom Bangkok or Singapore.

The availability of infrastructure services valuedby users is also critical for the modernization anddiversification of production. The growth of elec-tronic data exchange involving telecommunica-tionsinformatics--is central to efficient opera-tions in manufacturing, services, the financial sector,arid government. Availability of power allows sub-stantial improvements in workers' productivity (forexample, in the transition from foot-powered toelectrically powered sewing), while internationaltelecommunications, facsimile services, and rapidtransport of goods permit the artisan to produce toorder for a computerized global market. A higherquality of water and sanitation is required to shiftfrom production of raw agricultural commodities toprocessed foods. Surveys of prospective foreigninvestors over a wide range of countries show thatthe quality of infrastructure is an important factorin ranking potential sites for location of directinvestment.

The nature of an economy's infrastructure is cen-tral to its ability to respond to changes in demandand prices or to take advantage of other resources.The formerly socialist countries (particularly thosein Central and Eastern Europe and the former So-

Page 7: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

viet Union) provide a clear illustration of how thepatterns of supply and demand imposed by centralplanning affect infrastructure development. Thesecountries showed an extremely high transport andenergy intensity (owing to noneconomic decisionson location of production units, underpricing andinefficient use of energy, and an emphasis on heavyindustry and raw materials production). They alsoshowed a greater reliance on rail than on road trans-port than did countries with similar conditions, andon long- over short-haul public transport facilities.With market reforms, the location and compositionof demand will alter, giving a greater role in theseeconomies to light industry, to services such asdomestic distribution, and to the diversification ofexternal trade. Small enterprises and consumerswill become a more important source of demand.These trends require corresponding modificationsin infrastructure, with greater attention to the qual-ity and variety of services.

Public spending on infrastructure constructionand maintenance can be a valuable policy tool toprovide economic stimulus during recessions. Aslong as quality and cost-effectiveness are not com-promised, labor-based approaches to infrastructuredevelopment can also be an important instrumentfor employment-intensive economic growth. In de-ciding on public spending for infrastructure, policy-makers have frequently not looked sufficientlybeyond the near-term impacts, and many govern-ments have been attracted to the political benefits ofthe highly visible structures created. When publicspending on infrastructure is not wisely deployed, itcan crowd out more productive investment in othersectors. At the same time, short-term fiscal con-straints have often led to disproportionate cutbacksin infrastructure, thereby sacrificing an importantimpetus to renewed growth following adjustment(Box 1.3).

Sometimes the least-cost approach to improvingthe supply of infrastructure services would requireinterregional (cross-country) integration of infra-structure networks, for example, power grids. Suchan agreement would call for not only coordinationof investments but, equally important, cooperationto maintain efficient policies governing the trade inservices. Most countries, however, resist dependingon others for a supply of services deemed to be ofstrategic importance; therefore, importing power tomeet the base load demand is less acceptable thanacquiring only peak load from abroad. Internationalagreements have been more common for cross-bor-der transport, which is a particularly importantissue for landlocked countries. Often, the quality of

I

I

Box 1.3 Throwing infrastructureoverboard

When times are hard, capital spending on infra-structure is the first item to go, and operations andmaintenance are often close behind. Despite thelong-term economic costs of slashing infrastructurespending, governments find it less politically costlythan reducing public employment or wages. Stud-ies of fiscal adjustment and expenditure reductionfind that capital expenditures are cut more thancurrent expenditures, with infrastructure capitalspending often taking the biggest reduction. More-over, within current expenditures, nonwage expen-ditures (which include operations and mainte-nance) are cut by more than the wage bill.

The decline in investment, at least in the initialphases, is not altogether undesirable as it often in-duces a rationalization and strengthening of coun-tries' project portfolios. Cutbacks in operations andmaintenance expenditure, however, are worrisome.A World Bank review of countries' adjustment ex-perience found that reductions in nonwage opera-tions and maintenance and a marked deteriorationin infrastructure services were common. For in-stance, in Costa Rica during the 1980s current non-wage expenditures (principally operations andmaintenance) fell from 1.6 percent of GDP to a mere0.3 percent, and the share of the national and can-tonal road network in poor to very poor conditionrose to 70 percent.

transport infrastructure on an international corridoris less of a problem than are institutional con-straints. For example, one-third of the time requiredto ship freight between landlocked Mali and neigh-boring ports in Lomé (Togo) and Abidjan (Côted'Ivoire) is due to delays in customs clearance. Re-moving inefficient regulation of road transport andprivatizing transport operations, and deregulatingpower generation and distribution (as discussed inlater chapters), may facilitate some international ex-change of services in these sectors.

To summarize, infrastructure investment is notsufficient on its own to generate sustained increasesin economic growth. The demand for infrastructureservices is itself sensitive to economic growth,which is notoriously difficult to predict. The eco-nomic impact of infrastructure investment variesnot only by sector but also by its design, location,and timeliness. The effectiveness of infrastructureinvestmentwhether it provides the kind of ser-vices valued by users (responding to "effective de-mand")depends on characteristics such as quality

19

Page 8: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

and reliability, as well as on quantity. Matching sup-ply to what is demanded is essential. Finally, the ef-ficiency with which infrastructure services are pro-vided is also a key to realizing potential returns.

Links to poverty

Infrastructure is important for ensuring thatgrowth is consistent with poverty reduction, a topiccovered extensively in World Development Report1990: Poverty. Access to at least minimal infrastruc-ture services is one of the essential criteria for defin-ing welfare. To a great extent, the poor can be iden-tified as those who are unable to consume a basicquantity of clean water and who are subject to un-sanitary surroundings, with extremely limited mo-bility or communications beyond their immediatesettlement. As a result they have more health prob-lems and fewer employment opportunities. Theburgeoning squatter communities surroundingmost cities in developing countries typically lackformal infrastructure facilities, a condition arisingfrom their nonpermanence of tenure. In India theproportion of the urban population living in slumareas grew during 1981-91, while the share of thepopulation living in poverty (estimated using tradi-tional poverty measures based on income and foodconsumption) declined. The lack of access to infra-structure is a real welfare issue.

Different infrastructure sectors have different ef-fects on improving the quality of life and reducingpoverty. Access to clean water and sanitation hasthe most obvious and direct consumption benefitsin reducing mortality and morbidity. It also in-creases the productive capacity of the poor and canaffect men and women differently. For example, thepoorwomen in particularmust commit largeshares of their income or time to obtaining waterand fuelwood, as well as to carrying crops to mar-ket. This time could otherwise be devoted to high-priority domestic duties, such as childcare, or to in-come-earning activities. Such gender-specific effectsneed to be considered in the evaluation of proposedprojects.

Access to transport and irrigation can contributeto higher and more stable incomes, enabling thepoor to manage risks. Both transport and irrigationinfrastructure have been found to expand the op-portunities for nonfarm employment in rural areas,often in indirect ways (Box 1.4). A seeming develop-ment dilemma is that while rural poverty reductionrequires higher incomes, raising farmgate foodprices could make urban poverty worse. By raisingthe productivity of farms and of rural transport,

20

both an increase in the incomes of rural workers anda reduction in food prices for the urban poor can beachieved. The green revolution (with irrigationplaying a central role) demonstrated that the wagesof, and demand for, low-skilled agricultural laborersrise in step with more intensive cultivation and in-creased yields. Over twenty years, one closely ob-served Indian village saw yields increase almostthreefold and agricultural laborers' wages rise from2.25 to 5 kilograms of wheat a day. Improved ruraltransport can also ease the introduction of improvedfarming practices by lowering the costs of moderninputs such as fertilizer. An adequate transport net-work reduces regional variations in food prices andthe risk of famine by facilitating the movement offood from surplus to deficit areas.

The benefits of transport and communications in-clude the access they provide to other goods andservices, especially in cities. Where the poor are con-centrated on the periphery of urban areas, as inmany developing countries, the costs and availabil-ity of public transport become key factors in theirability to obtain employment. Access to secure andreliable public transport has been identified inhousehold surveys in Ecuador as influential in de-termining the ability of low-income girls andwomen to participate in evening training classes.

The construction and maintenance of some infra-structureespecially roads and waterworkscancontribute to poverty reduction by providing directemployment. Civil works programs (as carried outin Botswana, Cape Verde, and India), which ofteninvolve the provision of infrastructure, have alsobeen important in strengthening famine preventionand providing income.

Links to the environment

Infrastructure provision results from the efforts ofindividuals and communities to modify their physi-cal surroundings or habitat in order to improve theircomfort, productivity, and protection from the ele-ments and to conquer distance. Each sectorwater,power, transport, sanitation, irrigationraises is-sues concerning the interaction between man-madestructures (and the activities they generate) and thenatural environment. Environment-friendly infra-structure services are essential for improving livingstandards and offering public health protection.With sufficient care, providing the infrastructurenecessary for growth and poverty reduction can beconsistent with concern for natural resources andthe global environment (the "green" agenda). At thesame time, well-designed and -managed infrastruc-

Page 9: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

Box 1.4 Infrastructure's direct and indirect effects in rural India

A study of two villages in rural Karnataka state, south-ern India, offers a glimpse of the full impact that infra-structure can have on rural living standards. The re-searcher, who studied the Wangala and Dalena villagesin the 1950s and 1970s, described how the two villageshad been similarly poor and backward until a large-scaleirrigation project brought Wangala into a canal networkwhile Dalena's high elevation left it unirrigated.

Although canal irrigation directly promoted rapid in-tensification of cultivation in Wangala, institutions andthe villagers' way of life were relatively unaffected oth-erwise. In contrast, Dalena did not benefit directly fromthe canal. Its villagers were compelled to adjust theirway of life significantly in order to capture the indirecteconomic benefits from the irrigation project. The vil-lagers purchased land outside Dalena, sought positionsin the Public Works Department and a nearby sugar mill,and became involved in the transport of irrigated vil-

lages' sugarcane to the mill. Dalena quickly establisheditself as a service center in the region, and its residentsintegrated themselves into a much wider economicsphere than did those in Wangala.

The research emphasized how many Dalena villagerstraveled daily between their homes in the village andtheir places of work in nearby towns. This observation,echoed in other studies, suggests that the developmentprocess need not entail migration from rural areas tourban centers. In the Uttar Pradesh village of Palanpur,per capita living standards rose between 1957 and 1993in the face of population growth, in part because of ex-panding nonfarm employment. Residents of Palanpurcommute daily to the towns of Chandausi and Morad-abad, largely by rail. This type of rural commuting morecommonly occurs along roads by foot, bicycle, motorcy-cle, bus, or car.

ture can promote the environmental sustainability ofhuman settlements (the "brown" agenda). World De-velopment Report 1992 focuses on environmentalissues, including those of infrastructure sectors, indetail.

The relationship between each infrastructure sec-tor and the environment is complex. The most posi-tive impacts of infrastructure on the environmentconcern the removal and disposal of liquid andsolid wastes. But much depends on how disposal fa-cilities are planned and executed. Underinvestmentin municipal sewerage relative to water supply indensely populated cities such as Jakarta has beenfound to lead to harmful contamination of water re-serves, to exacerbate flooding, and to reduce thehealth benefits from water investments. Provision ofsewerage without wastewater treatment can lead tosevere downstream pollution and public healthproblems where receiving waters are used fordrinking-water supply or for recreation, irrigation,and fisheriesas illustrated by the cholera out-breaks in Peru and neighboring countries in recentyears. Poor management of solid waste complicatesurban street drainage and has been linked with theproliferation of disease-bearing mosquitos in stand-ing water. The growing problem of hazardous andtoxic wastes as countries industrialize poses partic-ular concerns about safe disposal. For example, un-controlled dumping has led to soil contamination inthe Upper Silesian industrial region of Poland andto subsequent food crop contamination.

Power plant and vehicle emissions are importantcontributors to air pollution, so their air quality im-pacts deserve careful analysis when facilities are ex-panded. In developing countries, almost one-thirdof commercial energy is devoted to electricity gener-ation, which is the fastest-growing component of theenergy sector. By the year 2000 Asia may well sur-pass all of Europe in sulfur dioxide emissions, andby 2005 it may surpass Europe and the United Statescombined in power plant emissions. Vehicles are asignificant source of airborne toxic pollutants, ac-counting for up to 95 percent of lead contamination.In Central and Eastern Europe, road transport is es-timated to account for 30 to 40 percent of total emit-ted nitrogen oxides and hydrocarbons. AlthoughOECD countries account for three-quarters of theworld stock of motor vehicles, a rapid increase in ve-hicle use is expected in parts of Central and EasternEurope, East Asia, and South America. In large andgrowing developing country cities, such as Bangkokand Jakarta, vehicle congestion already gives rise toconsiderable environmental and economic costs, ForBangkok, it is estimated that if reduced traffic con-gestion permitted a 5 percent increase in peak-hourvehicle speeds, the value of travel time saved wouldamount to more than $400 million a year. A 20 per-cent improvement in air quality in Bangkok, as a re-suit of a reduction in pollutants related to vehicle orpower plant emissions, would produce annualhealth benefits valued at between $100 and $400 percapita for Bangkok's 6 million residents.

21

Page 10: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

Expansion of transport infrastructure can reducetotal pollution loads as congestion falls, average ve-hicle speeds rise, and routes are shortened. But roadimprovements can also encourage vehicle use andincrease emissions. Therefore, additions to infra-structure capacity are oniy part of the solution. Im-proved management of traffic and land use and pro-motion of nonmotorized modes, cleaner fuels, andpublic transport are also needed (see Chapter 4). In-tegrated urban planning and transport policy canlead to more efficient use of both land and transportcapacity with favorable environmental results. Inthe city of Curitiba, Brazil, an emphasis on encour-aging enterprises and residential developments tolocate around carefully designed public transportroutes has contributed to low gasoline consump-tion, low transport costs relative to household in-comes, and very low rates of traffic accidents de-spite one of the highest rates of private vehicleownership in the country.

Beyond urban areas, overuse of water for irriga-tion (which accounts for about 90 percent of waterwithdrawals in most low-income countries) dam-ages soils and severely restricts water availabilityfor industry and households, which often have ahigher willingness to pay for the quantities of waterthey use. The inefficient burning of biomass fuel(plant and animal waste) for household energy con-tributes to deforestation and thus to erosion and lossof soil nutrients, as well as to indoor air pollution.Some infrastructure investments, especially roadconstruction, can put unspoiled natural resources atrisk and threaten indigenous communities. Reser-voirs associated with hydroelectric projects, floodcontrol, or irrigation can give rise to environmentalproblems, both upstream (inundation of land) anddownstream (sedimentation).

Origins of the public sector role in infrastructure

Infrastructure's large and varied potential impactson development derive from certain technologicaland economic characteristics that distinguish itfrom most other goods and services. These charac-teristics make infrastructure subject to special pol-icy attention.

Production characteristics

Historically, society's needs for water supply, irriga-tion and flood control, and transport have led to theconstruction of engineered physical worksmanyof them quite large, elaborately designed, and en-during. Today's distinctively modern infrastructure

22

sectors are the result of a technology-driven "infra-structure revolution" that has changed the way inwhich age-old demands for water, lighting, commu-nications, and waste disposal are met.

Not until the invention of cast-iron pipes andsteam-driven pumps did extensive water infrastruc-ture spread, beginning with a piped water networkin London in the 1850s. This lowered costs (espe-cially in urban areas) and dramatically increaseduse. Before the development of gas networks at thestart of the 1800s, infrastructure for lighting wasrare. The invention of alternating-current transmis-sion near the end of the century lowered costs ofelectricity and led to new and expanded uses ofelectric power, especially in urban transport.

The history of other infrastructure sectors is sim-ilar. The public telegraph and telephones replacedhand-carried messages, and piped sewerage re-placed individual disposal of wastes in many com-munities. Irrigation and transport have for centuriesutilized networks of irrigation canals and roads, al-though development of alternative modes of trans-port (including inland canals and railroads) has pro-ceeded since the early 1800s.

The most general economic characteristic ofmodern infrastructure is the supply of servicesthrough a networked delivery system designed toserve a multitude of users, particularly for publicutilities such as piped water, electric power, gas,telecommunications, sewerage, and rail services.The delivery system is in most cases dedicated, thatis, it carries only one good. Investments in the deliv-ery system (such as underground water pipes orelectric wires) are mostly irrecoverable because theycannot be converted to other uses or moved else-whereunlike the investment in a vehicle, for ex-ample. Once paid, these costs are said to be "sunk."Because the delivery system is networked, coordi-nation of service flows (traffic, electricity, communi-cations signals) along the system is critical to its effi-ciency. This interconnectedness also means that thebenefits from investment at one point in the net-work can depend significantly on service flows andcapacities at other points.

The scope for competitive supply of infrastruc-ture varies greatly across sectors, within sectors, andbetween technologies. Where the unit costs of serv-ing an additional user decline over a wide range ofoutput, economies of scale are createdan impor-tant source of "natural monopoly." This is a com-mon term, although one best used cautiously be-cause many infrastructure monopolies are in factunnatural, driven by policy and not technology. Butsectors differ greatly in the range of declining costs.

Page 11: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

For example, the optimal dimensions of a high-volt-age transmission grid may well be national, but thevolume-related unit cost savings for water can berealized at the municipal or submunicipal level.Even within sectors, different production stageshave different characteristics. In power, size savingsfor generation are often exhausted at a capacity thatis small relative to the size of a well-developed mar-ket. Activities also differ in the importance of sunkcosts, another potential source of natural monopoly.In railways and ports, for example, sunk costs areless significant for investments in rolling stock orfreight-handling equipment than for the fixed facili-ties. It is easier for firms to enter and exit activitieswith a relative absence of sunk costs and therebychallenge one another's potential market power.Such activities are said to be "contestable." Techno-logical and economic differences in production cre-ate the possibility of "unbundling" the componentsof a sector that involve natural monopoly fromthose that can be provided more competitively.

Many infrastructure services can be produced byvery different technologies. Sanitation based onimproved latrines or septic tanks provides thesame underlying service as does seweragedis-posal of wastes, but without networked invest-ments. Small-scale irrigationparticularly irriga-tion based on wells or boreholesand small-scalerenewable-energy-based power generation (such asmicro-hydro schemes) also need not involve inter-cormections with large networks but can provideservice highly responsive to users. Telephone ser-vices can be provided over wire-based networks orthrough radio-based systems.

Consumption characteristics

As seen earlier, the demand for infrastructure ser-vices derives from the activities of both industriesand individuals. Ensuring a flow of services of atleast minimum quality and quantity is often consid-ered by governments to be of strategic importance,since any interruption or restriction of supplywould be seen as a threat to society. However, be-cause infrastructure investments are often "lumpy"(new capacity must be created in large increments),it is difficult for planners to match the availability ofsupply with demand at all times. Costly episodes ofover- or undercapacity often result.

Beyond consuming an "essential minimum" ofcertain infrastructure services, users have very di-verse demandsalthough the output of large-scale,monopoly providers is often not sufficiently differ-entiated to meet these demands. For example, a

steel mill and a residential community may both de-rive water from the same supplier, but each usergroup values the quality of the water in quite differ-ent ways. Yet, because many infrastructure facilitiesare locationally fixed and their products are non-tradable, users cannot readily obtain substitute ser-vices that better suit their needs. Moreover, it isoften difficult for users to obtain information aboutservice alternatives or characteristics. They cannot,therefore, "shop around" for the best source of sup-ply and are vulnerable to any abuse of monopolypower. With many infrastructure activities, how-ever, supply can be better tailored to differences indemand once suppliers understand themfor ex-ample, transport can be offered at varying serviceand fare levelsand provided that consumers haveadequate information to declare their choices. Ser-vice markets can also be opened to alternative sup-pliers and technologies in order to provide a differ-entiated product (such as cellular and enhancedservices in telecommunications).

Many infrastructure services are almost (al-though not perfectly) private goods. Private goodscan be defined as those that are both "rival" (con-sumption by one user reduces the supply availableto other users) and "excludable" (a user can be pre-vented from consuming them). In contrast, "publicgoods" are neither rival in consumption nor exclud-able. Markets work best in providing pure privategoods or services. Most of the services that the infra-structure sectors produce are excludable in a specificsensetheir use depends on gaining access to a fa-cility or network, for example by connection to thepiped water, gas, or sewer system, and service usemay be metered and charged for. In the case of rail-ways, ports, and airports, access to the entire infra-structure can be restricted. However, once a user isconnected to the network utility or gains access tothe transport facility, the degree of rivalry with otherusers depends on the costs (including congestion)imposed on existing users or on the service supplierwhen an additional service unit is consumed.

It has been common in many countries not tocharge users for the volume of some utility servicesconsumed because the marginal supply cost wasconsidered negligible, congestion was absent, ortechnological constraints (such as the absence ofwater meters) prevented volume pricing. However,recent developments, such as the increased scarcity(and supply cost) of water, growing congestion asnetwork capacity becomes fully utilized, and techni-cal innovations in metering consumption, havemade it possible and desirable to price these ser-vices like other private goods.

23

Page 12: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

Roads are not private goods, although for rea-Sons that differ with the type of road. Rural roads (atypical public good) and uncongested interurbanroads are not completely rival because an additionaldriver does not reduce the value of anyone else'suse of the road. Access to some interurban roadscan be prevented by making them toll roads (a clas-sic "club" good, i.e., a good that is excludable butnonrival). By contrast, urban roads are congestedduring peak periods, but until recently it has beendifficult to exclude users from urban roads or tocharge users different amounts during peak and off-peak periods. New electronic techniques of moni-toring road use may eventually make it technicallyfeasible to treat many urban roads almost as privategoods.

Water outside of piped networks is ofteninpractice and in principlea "common property"resource. While water consumption is rival betweenusers, monitoring the use of groundwater from un-derground aquifers or from other natural sources isdifficult and costly, and therefore groundwater useis rarely excludable. By the same token, controllingthe consumption of common property resources isalso difficult. How much the extraction of water(from aquifers or natural flows) affects other poten-tial users depends on location-specific hydrologicalfeatures that are important in water policy.

Although most infrastructure goods are private,they produce spillovers or external effectsmanyof which (as shown earlier) affect the environment.Ignoring the important negative externality of emis-sions from fossil fuel power generation could leadto excess power being produced with the wrongmix of fuels. By contrast, some cities have neglectedto develop a well-designed public transport system,even though such a system can have positive envi-ronmental effects and also promote social equity. Toensure that society obtains positive benefits suchas public health benefits from water and sanita-tionthe private goods must also be delivered ef-fectively.

Thus, although infrastructure services differ fromother goods, they also differ among themselves(Figure 1.3). The characteristics of various infra-structure activities have important implications forhow services should be provided. To the extent thatspecific infrastructure activities entail natural mo-nopoly or depend on a network characterized bynatural monopoly, they will not be provided effi-ciently by an unfettered market. The network com-ponent can, however, be separated (unbundled)from the more competitive activities of the sector,with regulation to ensure fair access to the network.

24

Infrastructure activities that create externalities orproduce essential services to captive users may alsowarrant some regulation, but this can be narrowlyfocused on these market imperfections while per-mitting wide scope for competition in other compo-nents of the sector.

Certain characteristics of infrastructure also cre-ate challenges in financing. Where a minimum levelof consumption of a particular service (such aswater, heating, or power) can be identified as a "life-line" for some users, society may judge that theyshould not be excluded if they cannot afford to pay.Financing strategies also have to be designed to takeaccount of the risk that arises because many infra-structure investments are large and long-lived,while the revenue stream is often slow to develop.Such characteristics can justify some public financ-ing of infrastructure from general revenues, but tosupplementnot entirely substitute forthe rev-enues obtained from users and commercial sourcesof finance.

Public sector dominance in infrastructure

Infrastructure clearly represents a strong public in-terest, and so merits the attention of governments.However, the special characteristics of infrastruc-ture do not explain or justify the fact that govern-ments and public sector agencies have dominatedalmost all aspects of this sector in developing coun-tries in recent decades. Private participation wasimportant in the nineteenth century and the firsthalf of the twentieth century in many countries -and some pockets of private provision still re-mainbut the overwhelming trend until the early1980s was government or parastatal provision,largely through vertically integrated, monolithic en-tities. By then, only a small percentage of the powersector was in private hands. Virtually no privatetelecommunications firms existed, and most earlyprivate railways had disappeared with nationaliza-tion. Although toll roads played a part in the earlyhistory of many countries, they also became rare,and road construction (and especially maintenance)was executed largely by government employees, orforce account. Other serviceswater, sewerage,waste disposal aisci tended to be both owned andoperated by governments at either the national orthe local level.

The dominant public sector role in infrastructurehas arisen for a number of reasons: recognition ofinfrastructure's economic and political importance;a belief that problems with the supply technologyrequired a highly activist response by governments;

Page 13: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

Figure 1.3 Infrastructure services differ substantially in their economic characteristicsacross sectors, within sectors, and between technologies.

Excludable Nonexcludable

10

Interurban highways(toll roads)

Club goods

Lower .i4

Urban bus

Rural sanitation(on-site disposal)

Rail, airport, and port services

High-voltage transmission

Rail, port, and airport facilities

Fossil fuel power generation

Piped water supply

Surface water irrigation

Sanitary landfill

Urban sewerage

Commonproperty

GroundwaterUrban roads

Rural roadsStreet sweepingTraffic signaling

Public goods

Note: Excludable means that a user can be prevented from consuming the good or service. Rival means that consumption byone user reduces the supply available to other users.

and a faith that governments could succeed wheremarkets appeared to fail. Many countries made im-pressive strides in infrastructure expansion underthe earlier stages of this public leadership. But morerecent experience has revealed serious and wide-spread misallocation of resources, as well as a fail-ure to respond to demand. Moreover, the blunt in-struments of public ownership, financing, andoperation have not demonstrated any advantage inachieving poverty reduction goals or environmentalsustainability. These deficiencies in performance arenot happenstancethey are embedded in the pre-vailing system of institutional incentives for thesupply of infrastructure.

The record of performance

Achievements

Although the data are spotty, impressive expansionin infrastructure has been achieved in recentdecades, as measured by stocks and production ofservices (Table 1.3). In low-income economies, tele-communications, sanitation, and water supply reg-istered the highest rates of increase in availabilitybetween 1975 and 1990, starting from a very lowbase in each sector. In middle-income economies,growth in this period was concentrated mainly inthe power and telecommunications sectors, where

25

Externalities Higher

Rival

A

Non-rival

Page 14: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

Table 1.3 Expansion of infrastructure coverage in low-, middle-, and high-income economies,recent decades

capacity more than doubled between 1975 and 1990.Even in middle-income economies, however, accessto water and sanitation is still lacking for significantshares of the populationfor water, one-quarter ofthe population in this group remains unserved, andfor sanitation, one-third. The most dramatic expan-sions in paved roads occurred during 1960-75 forboth groups, after which growth slowed.

Infrastructure coverage has increased in bothrural and urban areas. Urban populations are signif-icantly better served than rural populations in ac-cess to drinking water, sanitation, and power. Thegaps in coverage for water and power have beennarrowing (Figure 1.4). Rural and urban areas donot have the same effective demand for infrastruc-ture services and thus may require different rates ofinfrastructure coverage to achieve desired develop-ment benefits. There is an economic case for provid-ing relatively more power and telecommunicationsconnections, and more extensive transport net-works, in locations with a higher density of popula-tion and industry.

Urbanization in itself is an important factor stim-ulating demand for infrastructure. When infrastruc-hire capacity in water supply, sanitation, power,telecommunications, roads, and public transport isinadequate in expanding urban areas, serious con-straints on (environmentally sustainable) economicgrowth and on poverty reduction result. In therapidly growing periurban (and, in many cases,unauthorized) settlements that ring many cities,conventional delivery of formal services is oftenprevented by legal, topographical, or economic con-straints. Projected growth in urbanization in comingdecadesespecially in Africa and South and EastAsiawill inevitably increase pressures for greateraccess to infrastructure. However, some rural-to-

26

urban migration may be forestalled through policiesthat provide appropriate infrastructure in ruralareas and that prevent the degradation of natural re-sources (especially soils, forests, and water sup-plies).

An analysis of how countries measure up on in-frastructure coverage compared with other mea-sures of performance is revealing. Although cover-age tends to be correlated with GDP, efficiency andeffectiveness of infrastructure provision are not.Plots of coverage against performance in water,powel telecommunications, roads, and railwaysshow little relationship across a wide sample of low-and middle-income countries (summarized in Fig-ure 1.5). Moreover, there is no close correlation be-tween a country's efficiency of provision in one sec-tor and its performance in another. These findingsindicate that efficiency and effectiveness of infra-structure provision derive not from general condi-tions of economic growth and development butfrom the institutional environment, which oftenvaries across sectors in individual countries. Thissuggests that changes in the institutional environ-ment can lead to improved performance, even whenincomes are low, because in each sector some low-income countries perform well. As a corollary, a re-cent OECD review of infrastructure noted that evenmany high-income countries encounter the perfor-mance issues described below.

Challenges

To determine future demand for infrastructure, it isnecessary to consider the efficiency with which ex-isting capacity is being used and how well the ser-vices generated are responding to users. Althougheach sector has special problems, there are com-

Sector

Low-income economies Middle-income economies incomeeconomies:coverage,

1990

CoverageAnnual

percentageincrease

CoverageAnnual

percentageincrease1975 1990 1975 1990

Power-generating capacity (thousand kilowatts permillion persons) 41 53 1.6 175 373 4.7 2,100

Telecommunications (main lines per thousand persons) 3 6 3.2 33 81 5.6 442

Sanitation (percentage of population with access) 23 42 3.8 44 68 2.7 95+

Paved roads (kilometers per million persons) 308 396 1.6 1,150 1,335 0.9 10,106

Water (percentage of population with access) 40 62 2.7 54 74 2.0 95+

Note: Percentage increases are compound growth rates.Source: Appendix tables A.1 and A.2.

Page 15: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

Figure 1.4 The rural-urban gap in access topower and water in developing countriesnarrowed over the past decade.

Percentage of population withaccess to infrastructure90

60

30

mon patterns operational inefficiencies, inade-quate maintenance, excessive dependence on fiscalresources, lack of responsiveness to users' needs,limited benefits to the poor, and insufficient envi-ronmental responsibility.

INEFFICIENCY OF OPERATIONS. The broadest indica-

tor of inefficient performance by an infrastructuresystem is the extent of output lost in delivery. Unac-counted-for water (that portion of supply for whichconsumption is not recorded, largely because oftechnical and managerial failures) is typically two tothree times higher in developing country systemsthan in countries that achieve the industry stan-dards. In 1987 one-quarter of the power utilities indeveloping countries had losses of electricity in thetransmission and distribution network that weretwice those in efficiently operated systems. In some

African countries, spending $1 million to reduceline losses could save $12 million in generating ca-pacity. Irrigation efficiency (the proportion of waterdelivered to the field) in developing country proj-ects is typically 25 to 30 percent, compared with 40to 45 percent under best practice.

Inefficient use of labor is especially common andcostly in infrastructure. At various periods, two-thirds of the labor in railways in Tanzania and Zaire,80 percent of port staff in Argentina (before recentprivatizations), and one-quarter of highway depart-ment staff in Brazil have been estimated to be re-dundant. The combination of overstaffing and un-derpricing of railway services produced a wage billalmost as large as (and sometimes larger than) totalrailway revenues in Argentina (before recent re-forms) and in Colombia, Egypt, Nigeria, Turkey,and Uruguay. Overstaffing is also common inwater, power, and telecommunications. At thesame time, in the production of public works andrural infrastructure, developing countries oftenuse equipment-based methods of construction andmaintenance rather than employment-intensive ap-proaches that can produce high-quality results,while being more consistent with relative capitaland labor costs.

INADEQUATE MAINTENANCE. Closely related to op-erating inefficiencies is lack of maintenance: roadsdeteriorate, irrigation canals leak, water pumpsbreak down, sanitation systems overflow, installedphone lines fail, and power generators are not avail-able when needed. Capacity is then lost, output de-clines, and substantial additional investment isneeded simply to sustain existing levels of service.

In the road sector, inadequate maintenance im-poses large recurrent and capital costs. The engi-neering and physical properties of paved roads aresuch that, as a road begins to deteriorate, lack of reg-ular routine maintenance will hasten deterioration.Neglect of (relatively inexpensive) routine mainte-nance can compound problems so much that the en-tire surface of a road has to be replaced. Examina-tion of completed Bank highway projects showsthat, on average, estimated returns on projects in-volving primarily maintenance are almost twice ashigh as those on projects involving mainly new con-struction. Yet, in Sub-Saharan Africa, almost $13 bil-lion worth of roadsone-third of those built in thepast twenty years have eroded because of lack ofmaintenance. In Latin America, for every dollar notspent on maintenance, $3 to $4 are estimated to berequired for premature reconstruction. Maintenanceexpenditures often are not allocated by economic

27

1980 1990 1980 1990

Urban areas Rural areas

Power

Water

Sanitation

Source: Israel 1992; WHO 1980, 1990.

Page 16: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

Figure 1.5 Efficient and effective delivery of infrastructure services does not alwaysaccompany increased availability.

0 50 100 150

Telephone main lines per thousand persons0 20 40 60 80 100

Percentage of households with electricity

28

Faults per 100 main lines Percentage of power delivered0 100- -

Egypt TurkeySaudi Arabia

.Carneroon Zambia.Zlmbabwe.

Trinidad & TobagoMexico

50 Syria . 80Bolivia #t a

S

SS .

S Colombia.,Panama'

o ArgentinaChile 0 .

100Tunisia

0 SJordan Romania 60 Dominican- Republic

150 Chad40

Mauritius

200.

20Zimbabwe

250 0

100

Percentage

Kilometers of paved roadsper million persons

of accounted-for water

Rail passenger-kilometers +ton-kilometers per million US$ of GDP

Cole divoireNicaragua

80 0 TogoIndonesia Gabon Brazil S

Somalia .0 5Jamaica Middle East and North Africa

60 .Ethiopia

.Nepal

SS

Ecuador . Jordan Latin America and the Caribbean

40 Philippines East Asia and PacificSub-Saharan Africa

20 South AsiaEurope and Central Asia

0

Percentage of paved roads Percentage of dieselin good condition locomotive availability100 100Mauritius' S Algeria

80

Mexico Botswana- Côte dIvoire Trinidad & Tobago 80

'EgyptMorocco

MalawiS Pakistan

India SSouth Africa

KoreaOmanPoland S

Mexico60 - :.. Tunisia 60 i(Portugal' $ Argentina

'Zimbabwe

40Colombia Chile

$ .'Argentina Panama. Uruguay

Brazil . Roinania

40 - 'ZambiaColombia

Sudan

20 Peru Costa Rica 20 Nigeria

Sri Lanka Jamaica

0 (1

0 500 1,000 1,500 2,000 2,500 0 200 400 600 800 1,000

0 20 40 60 80 100

Percentage of populationwith access to safe water

Source: WDI table 32.

Page 17: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

priorities. For example, Cameroon, which still has apredominantly rural population, has neglected its30,000-kilometer unpaved road network over thepast ten years in favor of investment in and mainte-nance of 3,700 kilometers of intercity paved roads.The result is that some 80 percent of the unpavednetwork requires either complete reconstruction orheavy reshaping and compaction.

In railways, inadequate maintenance (as well asother operating deficiencies) is evident in the smallshare of locomotives available for service. In 1991only 60 percent of all locomotives were available forservice in Latin America and 70 percent in the Mid-dle East and North Africa region, compared with 90percent in North America. Such deficiencies causesome railways to turn away freight traffic, which inturn compounds the sector's financial difficulties.

In irrigation, too, poor maintenance is costly andresults in distribution channels filling with silt andweeds, canal linings cracking at an increasing rate,and outlets breaking or being bypassed. Drainagealso fails, causing salt buildup in the soil. In Chinaalmost 1 million irrigated hectares have been takenout of production since 1980, and in the former So-viet Union, even with continuing investment in irri-gation, almost 3 million hectares were lost between1971 and 1985one-quarter of the new irrigatedarea. Worldwide, works covering 60 percent of theirrigated area require upgrading to remain in goodworking condition.

In both rural and urban water supply and in thepower sector, inadequate maintenance is a commonproblem. A study of water and sewerage in Bogotafound that the costs of unaccounted-for waterarising in part from poor maintenance of the distrib-ution systemwere 42 percent of the supplier'stotal operating income. Poor maintenance practicesaccount for some of the low availability of power-generating capacity which averages less than 60percent for thermal plants in many developingcountries, compared with more than 80 percent insystems operated at best-practice standards.

Sometimes problems of operation and mainte-nance are rooted in the initial design or constructionof infrastructure. For example, a recent review ofcompleted World Bank irrigation projects foundthat basic design flaws (such as inappropriate trans-fer of desert technologies to tropical monsoon cli-mates) were widespread. Operations and mainte-nance can be made more difficult by inappropriatedesign standards that increase the requirements forskills in short supply or involve heavy dependenceon imported spare parts where foreign exchange isscarce. Poor construction and design of power and

water treatment plants, or inappropriate location,make it difficult to carry out operations and mainte-nance and to meet environmental objectives. Thereare also many examples of investments that wereeconomically nonviable to begin with and thatshould never have been madesuch as over-designed or "gilt-edged" roads and power plants.

Procurement problems are often a factor in weakoperational performance. Systematic delays in pur-chasing by sector entities and inadequate supervi-sion of contracts are estimated to increase costs ofimported materials to some African countries by 20to 30 percent. Contracting and bidding proceduresmay also favor large-scale enterprises, which tendto use more equipment-based methods of construc-tion and maintenance than is appropriate given rel-ative factor costs. The lack of standardization ofequipment, such as water pumps obtained from di-verse foreign donors, creates delays in repair and in-creases the costs of replacement parts. There is needfor donors to standardize their procurement rules toease the administrative burdens on recipient coun-tries. Donor aid that excludes finance for local costscan also bias the choice of technology for publicworks in favor of capital-intensive methods that areunsustainable for the recipient country.

FINANCIAL INEFFICIENCY AND FISCAL DRAIN. Poor

infrastructure policies and inefficient provision ab-sorb scarce fiscal resources and damage macroeco-nomic stability. Because prices are often held wellbelow costs, the subsidies flowing into public infra-structure enterprises and agencies have been enor-mous in many countries. In Bangladesh, India, In-donesia, Pakistan, and the Philippines, irrigationreceipts have been well below the costs of opera-tions and maintenance. During the 1980s power tar-iffs in developing countries were on average aboutone-half the costs of new supply and were muchlower than in OECD countries. (The record on pric-ing is discussed further in Chapter 2.) In recentyears, 60 percent of Ghana Railway revenues con-sisted of government subsidiesa not-uncommonperformance for this sectorand recurrent subsi-dies to railways have amounted to as much as 1 per-cent of GDP in a number of countries. In Zambia thetotal cash shortfall in transport absorbed 12 percentof the government's current revenue in fiscal 1991.Telecommunications tends to be an exception to thegenerally poor cost recovery elsewhere in infra-structure, although its revenues are often siphonedoff by government for other uses, leaving the sectorunderfunded. Inadequate tariffs are often com-pounded by poor financial management. In a sam-

29

Page 18: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

pie of Latin American water utilities, collection ofaccounts receivable took almost four months on av-erage, compared with good-practice standards offour to six weeks. In addition to creating an addedburden on taxpayers, poor financial performance bymany infrastructure providers means a loss ofcreditworthiness for the entity concerned. It alsoresults in a low reliance on internal revenues to fi-nance investment and therefore an inability (andlack of incentive) to expand or improve service.

UNRESPONSIVENESS TO USER DEMAND. The result of

inefficiency and poor maintenance is low-quality,unreliable service, which alienates users. Reliabilityis a critical aspect of user satisfaction that is often ig-nored. Even where users have telephones, high callfailure rates (more than 50 percent in many cases)and high fault rates drastically diminish the value ofthe service. Unreliable quantity or quality of waterleads to enormous investments in alternative sourcesthat are especially costly to those who can least af-

Box 1.5 Households' responses to unreliability of water supply

In 1991, micro-level research on household responses todeficient water supply by public utilities was under-taken in Faisalabad (Pakistan), Istanbul (Turkey), andJamshedpur (India). These surveys revealed that nearlyall households in the three cities are dependent on multi-ple sources of water, including house taps, wells, tube-wells, public taps, rivers, and street vendors. Not all al-ternatives are available to all households. Because accessto a source increases with income, poorer householdsbear a disproportionate share of the burden of deficientinfrastructure. The private expenditures incurred forwater supply indicate consumers' willingness to pay forreliable water.

In Istanbul, the poorest households surveyed spend alarger share of their income (about 5 percent) to supple-ment inadequate water supply than do wealthier ones(which spend about 1 percent). These expenditures oninformal sources of water, including self-provision fromwells or storage facilities, are in addition to the usercharges for publicly supplied water, which amount to 1to 2 percent of annual income.

In Jamshedpur, the connection charges for pipedwater vary between $1.66 and $16.66. The residents ofthe periurban areas, served by the local municipal au-thorities, incur capital costs of $50 to $65 in installingtubewells and $150 to $300 in digging wells to avoid de-pendence on the (unreliable) public water supply. De-spite the existence of a piped water system, at least 17percent of the population meets 90 percent of its waterneeds from wells and handpumps. Over and above themonetary costs that consumers bear, households inJamshedpur spend, on average, two hours a day fetch-ing and storing water. The burden of these activities fallsin nearly all cases on women.

The pattern of private augmentation of the publicwater supply at substantial private costs to consumers isobserved also in Faisalabad, Pakistan. Less than 20 per-cent of the households with piped water use this sourceexclusively; 70 percent have motor pumps and 14 per-cent have handpumps.

Box 1.6 Public failures raise private costs

According to a 1988 study of Nigerian manufacturers, 92percent of the 179 firms surveyed owned electricity gen-erators. In the face of chronically unreliable public ser-vices, many had also acquired radio equipment for com-munications, vehicles to transport personnel and freight,and boreholes to assure their own private water supply.For firms with fifty or more employees that could prac-tice economies of scale, these extra costs amounted tosome 10 percent of the total machinery and equipmentbudget. For small firms, the burden could be as high as25 percent. Yet because Nigerian regulations preventfirms from selling their excess power capacity, busi-nesses both large and small were operating private gen-erators and water systems on average at no more than 25percent of capacity.

Of 306 Indonesian manufacturers recently polled, 64percent had generators and 59 percent (compared withNigeria's 44 percent) had boreholes for their own water

supply. Indonesia's largest companies invested as muchas 18 percent of their capital in private infrastructurealmost twice Nigerian manufacturers' level of 10 per-centyet their generators, too, were underused and op-erating at about 50 percent of capacity.

Today in Indonesia, as in Nigeria, firms too small toafford private power or water are at the mercy of unreli-able public utilities and subject to chronic and costly in-terruptions in service. Yet while the largest Indonesianfirms pay $0.07 per kilowatt-hour to produce electricity(not far above international norms), self-provided elec-tricity costs the smallest firms $1.68 per kilowatt-hourtwenty-four times as much.

Thailandwhere public electric utilities are effi-ciently runhas been able to break this pattern. Of the300 manufacturers polled, only 6 percent had privategenerators and 24 percent had private water supplies.

30

Page 19: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

ford it (Box 1.5). In Indonesia and Nigeria, privatebusinesses incur heavy costs in order to guaranteepower supply: 92 percent of firms sampled in Nige-ria and 64 percent in Indonesia had installed privategenerating capacity (Box 1.6); in Thailand, only 6percent of companies needed generators. These largedifferences in self-provision reflect the performanceof the formal suppliers. In Nigeria, only 43 percent ofinstalled capacity was in service by 1990 (despitemassive overinvestment in public power-generatingcapacity throughout the 1980s); in Thailand, thepower utility is efficiently run.

In telecommunications, unmet effective demandcan be roughly measured, because in many coun-tries users must apply for connection, which ofteninvolves paying a heavy initial fee. Of ninety-fivedeveloping countries, more than one-third werefound to have a waiting period of six or more yearsfor a connection, compared with less than a monthin most industrial countries (Figure 1.6). Countriesthat can deliver service in less than a year includesome with little current pressure on available capac-ity (such as Bolivia) as well as others in which in-vestment is proceeding rapidly (Malaysia). In addi-tion to the shortage of basic connections, in manycountries providers fail to offer differentiated ser-vices matching types of use. For example, busi-nesses increasingly require telecommunications fa-cilities that accommodate high-speed datatransmission as well as voice signals. A muchhigher priority could be given in many developingcountries to the provision of pay phones to extendbasic access to improved communications to alarger share of the population.

Excess demand for infrastructure, coupled withvery low rates of compensation to infrastructurestaff, breeds corruption in both service and invest-ment decisions. Where connections are rare andservice is poor, employees often demand side pay-ments from users to install or repair connectionsespecially in telecommunications, irrigation, andwater supply.

NEGLECT OF THE POOR. The poor typically usefewer infrastructure services than the nonpoor, butnot only because of low incomesthey also havevery low access. In Peru, for example, only 31 per-cent of the poorest fifth of households are connectedto a public water network and 12 percent to a publicsewer compared with 82 percent of the top fifthfor water and 70 percent for sewerage. The poorgenerally have less access than the rich in urbanareas as well (Table 1.4).

Many countries have introduced subsidiesthrough low tariffs with the aim of improving the

Figure 1.6 There is very high unmetdemand for telephone connection.

Number of countries25

20

15

10

0

Less 1.0- 3.0-than 1 2.9 5.9

Waiting time for telephone main lines, 1992(years)

Sample: 95 developing countries.Source: ITU 1994.

6.0-- 9.0- 12.0 or8.9 11.9 more

poor's access to infrastructure services, but most ofthese subsidies have been captured by middle- andhigh-income households (as documented in Chap-ter 4). In addition, the providers often are not ade-quately compensated for the subsidies, so that over-all expansion of service is constrained, The structureof tariffs can be an additional barrier. In Brazil, localtelephone call rates are low, but connection chargesare high. This prevents lower-income users fromgetting service. Flat-rate electricity charges in ruralIndia have benefited mainly richer households, be-cause the poor lack the income to purchase thepumps and consumer appliances that account formost electricity use.

While failure to reach the poor has often been as-sociated with flawed infrastructure pricing policies,too little emphasis has been placed on providing thepoor with suitable options for the kinds of servicesof most value to them (and for which they are will-ing to pay). For example, municipal sanitation agen-cies often promote technical designs for conven-tional sewerage that are unaffordable and evenenvironmentally unsuitable in some low-incomesettlements. In large cities such as New Delhi, thereliance of the poor on foot travel is a serious con-

31

Page 20: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

straint to their mobility (Figure 1.7) A study oftransport options in Latin American cities foundthat in São Paulo, Brazil, personal travel by the poorhad declined more sharply than for any other in-come group over a decade-partly because publictransport services were ill designed for low-incomeusers. The poorest residents on the periphery of Riode Janeiro spent more of their income than the richfor transport, with longer waits, less frequent ser-vice, and more time spent in crowded vehicles.

Appropriate services for the poor are often lack-ing when decisions on investment and service aredriven by assumptions about a "needs gap" ratherthan by an assessment of effective demand. In theMakete District in Tanzania, a survey of householdsundertaken to determine their transport needs inpreparation for a proposed investment project re-vealed that improvement of the road network alonewould benefit only a few residents and that com-plementary measures were needed-includingsupport to transport services (the introduction ofnonmotorized means of transport to replace head-loading), simple improvements to paths and tracks,and rehabilitation of grinding mills. A retrospectiveevaluation carried out after completion of the proj-ect found that these low-cost improvements werehighly successful-and would likely have been leftout of the project if no inquiry into the actual de-mand of the communities had been undertaken.

NEGLECT OF THE ENVIRONMENT. The impact of in-frastructure on the environment has often been verynegative (Box 1.7 recounts one of many examples,and one where regional cooperation is needed to de-velop a solution). The highly visible effects of cer-tain large-scale facilities-such as dams and roadsin sensitive ecological areas or where resettlementoptions are unsatisfactory to populations -have at-

32

Not available.Source: Glewwe 1987a, b; Glewwe and Twum-Baah 1991; World Bank 1993e.

tracted understandable public attention. Yet equallyserious, and more pervasive, is the damage or lossof potential benefits to the environment because offailure to control unnecessary emissions and waste-ful consumption of water. This is due in particularto the underpricing of power, vehicle fuels, andwater for irrigation and municipal uses and to theneglect of maintenance. Inadequate maintenancepractices leading to inefficient thermal power gen-eration account for a large share of energy-relatedpollution. Neglect of sound environmental manage-ment practices in transport-including safe han-dling of hazardous cargos and appropriate disposalof waste from ships, port dredging, and vehiclemaintenance-is a common failing. Unregulated,badly designed, or poorly managed municipalwater and sanitation infrastructure has often beenone of the biggest sources of urban environmentalpollution. The focus of public spending on urbansolid waste management often stops at collection-few developing country cities meet environmentalstandards for sanitary landfills.

Many of the problems in infrastructure perfor-mance are mutually reinforcing, creating seriouseconomic and financial costs that make it more diffi-cult for countries to achieve greater coverage andmore modern services to better meet social and en-vironmental goals. Systemic problems point to sys-temic causes-and solutions.

Diagnosis and directions for change

The conditions for improved performance: causesand cures

Where infrastructure is operated inefficiently anddelivers poor service, the solution cannot be simplyto tell suppliers to do more maintenance and to

Table 1.4 Percentage of the poorest and richest population quintiles with access to infrastructure,various countries

Country/area

Access to public water supply Access to sewers Access to electricityPoorestquintile

Richestquintile

Poorestquintile

Richestquintile

Poorestquintile

Richestquintile

National areasCôte d'Ivoire (1985) 2.4 62.1 3.4 57.0 13.2 74.8Ghana (1987-88) 10.5 30.6 0.5 14.6 5.6 46.0Guatemala (1989) 46.9 86.8 16.1 86.1Mexico (1989) 50.2 95.0 14.2 83.2 66.2 99.0Peru (1985-86) 31.0 82.0 12.3 70.0 22.8 82.5

Urban areasBolivia (1989) 84.8 89.9 52.6 87.4Paraguay (1990) 53.7 88.8 10.4 62.2 94.5 99.2

Page 21: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

Figure 1.7 Walking is a transport mode used frequently by the poor.

80

60

40

20

Delhi, India

Percentage a

100

Squatters Low-income Middle-income

LII Share who walk

Low-income

LI Share who take bus or paratransit

a. Percentage of urban residents in each group using each mode.Source: Survey data from Jaiswal 1992, as cited in Serageldin 1993.

Kingston, Jamaica

Lower-middle-income

Upper-middle-income

consult users. The weaknesses in infrastructure pro-vision are inherent in the incentives built into cur-rent institutional and organizational arrangements,in which outputs and inputs are not closely mea-sured, monitored, or managed, and suppliers donot depend on user satisfaction for reward. Aproper set of incentives would make managers ac-countable to users and to others who own and fi-nance infrastructure facilities. It would also givemanagers autonomy in making decisionsand re-sponsibility for success or failure. This Report's re-view of experience with infrastructure, in both thepublic and the private sectors, suggests that threeelements are essential in creating the right incen-tives for efficient and responsive delivery of ser-vices. These are management based on commercialprinciples, competition, and involvement of usersand other stakeholders.

COMMERCIAL PRINCIPLES. Infrastructure must be

conceived of as a "service industry," providinggoods that meet customers' demands. Such a com-mercial orientation contrasts sharply with the situa-tion in most government departments and state-owned public utilities, which suffer from multipleand conflicting objectives and inadequate account-

ing for costs or financial risk, and which put littleemphasis on revenues collected and the quality ofservice delivered. Managers have little motivationin such circumstances to satisfy customers or toachieve a reasonable return on assets through effi-cient operation and adequate maintenance. Typicalproviders of infrastructure are subject to pervasiveinterference by political authorities, which ad-versely affects operational decisions on investment,pricing, labor, and technological choices. It is com-mon to view certain infrastructure services (such aspower, water, ports, railways, airports, and telecom-munications) as potentially "commercial" becausethese are the services for which it is easiest to re-cover the costs of provision through user charges ortariffs. In fact, almost all infrastructure (even roadsand sanitation) can be operated with a business ori-entation. The basic conditions for this are limitedand well-focused performance objectives, financialand managerial autonomy (with a hard budget con-straint), and clear accountability both to customersand to providers of capital.

COMPETITION. Competition promotes efficiencyand provides users with options that, in turn, makeinfrastructure providers more accountable. Govern-

33

Page 22: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

Box 1.7 Infrastructure activities threaten the Black Sea environment

The Black Sea is fed by a basin of more than 2 millionsquare kilometers, covering parts of seventeen countriesin Central and Eastern Europe, the former Soviet Union,and Turkey. It receives the inflows of several majorrivers, including the Danube, Don, Dnieper, andDniester. As an almost fully enclosed water body, theBlack Sea is especially vulnerable to changes in the quan-tity and quality of inflows from these rivers. The Donand Dnieper, in particular, have been highly developedfor irrigation and other purposes through a chain ofreservoirs.

Increasing pollutant loads from these riversespe-cially the nutrients nitrogen and phosphorushave ledto algal blooms and the destruction of important nurseryareas for fish. In addition, damming of the major riversfor navigation, flood control, water supply, and, aboveall, for irrigation, has considerably altered the seasonalflow patterns of these rivers. The damming has also de-creased the total inflow to the Black Sea, resulting in anincrease in salinity in critical coastal and estuarine areas,especially in the Sea of Azov, which creates further prob-lems for fish breeding. The overall result is a 90 percent

decline in the once-productive Black Sea fishery over thelast thirty years.

With assistance from the Global Environment Facil-ity, the six Black Sea countries (Bulgaria, Georgia, Roma-nia, Russia, Turkey, and Ukraine) have begun a regionalprogram to analyze the causes of observed environmen-tal degradation and to propose solutions. Actions in thebasin to regulate fertilizer use and to control pointsources of pollution are expected to result in reductionsof nutrient inflows. Pilot projects are proposed to restorefish production under the new salinity conditions.Under the Bank-supported Environmental ManagementProject for Russia, a study of the Lower Don Basin willinvestigate ways to alter the operating rules for themajor reservoirs to promote greater fish regenerationdownstream.

Given the size of the problem and the importance ofthese reservoirs in the agricultural economies of Ukraineand Russia, it would be unrealistic to expect dramaticchanges. Nevertheless, recognition of the problem andthe development of mechanisms for regional coopera-tion now make progress much more likely.

ments in most countries have not taken advantageof the potential for competition, even in activitieswhere a natural monopoly does not exist, such asroad freight transport or solid waste collection.Today competition can be used directly in more in-frastructure activities because of technologicalchanges. In telecommunications, satellite, mi-

crowave, and cellular radio transmission of tele-phone signals is revolutionizing the industry, mak-ing the economies of scale with cable-basedtransmission less important. In power generation,combined-cycle gas turbines operate efficiently atlower output levels than other generation technolo-gies. While open competition for users in the marketis still not feasible in many infrastructure areas,there are other ways of obtaining the benefits ofcompetition. For activities with high sunk costs,competing for the right to operate a monopoly cancapture many of these benefits. Even where thenumber of operators is necessarily limited, regula-tion can compel them to compete against perfor-mance benchmarks ("yardstick" competition).

INVOLVEMENT OF USERS AND OTHER STAKEHOLDERS.

In many infrastructure activities, market signalscannot be relied on to provide information aboutdemand or to gauge performance. Where users arelocked into a delivery network, they cannot expresstheir preferences or dissatisfaction through choice.

34

In such circumstances, other means of making sup-pliers accountable to users are needed. Throughvarious mechanisms designed to broaden participa-tion in decisionmaking and to provide wide accessto information on infrastructure provision, usersand other key stakeholders can be represented in(and sometimes take responsibility for) the plan-ning, financing, and delivery of services.

Opportunity knocks

Many of the above notions are not new, and somehave been accepted in principle by policymakers ifnot yet put successfully into practice. Three fac-torstechnological change, more pragmatic atti-tudes, and a greater sensitivity to infrastructure'simplications for poverty and environmental sus-tainabilityhave created a new climate for re-form. Innovative techniques for drawing on pri-vate financing for investment create a furtherchallenge to traditional ways of providing infra-structure. Many countries are now taking advan-tage of all these opportunities to test new ideasand approaches, discussed in later chapters of thisReport.

TECHNOLOGY. Technological changes are creatinga variety of new opportunities for changing the wayinfrastructure is provided in almost every sector-

Page 23: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

in particular, by making the unbundling of diverseactivities more feasible. Microelectronic monitoringdevices and nondestructive testing techniques canfacilitate the assessment of infrastructure facilities(at reduced cost), often permitting testing by anagent other than the operatorsuch as the owneror regulator. Remotely controlled devices for in-specting pipe networks and the shift from analog todigital telephone switching have greatly simplifiedand reduced maintenance costs. Electronic informa-tion systems, including geographic mapping, im-prove the planning and design of investments andthe coordination of network operations. Technolo-gies that are clearly more efficient, robust, and flexi-ble than earlier methods enable developing coun-tries to "leapfrog" sectoral transitions experiencedearlier by high-income countries. For example,Brazil based its telecommunications expansion inthe 1970s on emerging digital equipment andthereby facilitated the development of information-based industries. Policy-induced inefficienciesslowed the modernization of the sector in the 1980s,however.

NEW PRAGMATISM. A new attitude, stemmingfrom an enhanced understanding of the relativestrengths and weaknesses of governments and mar-kets, is also creating opportunities for reform of in-frastructure provision. In the 1980s, the efforts ofmany countries to reduce the size of their over-extended public sectors led to a better realization ofwhat governments and markets can and cannot do.Worldwide liberalization of markets and experi-ments with different forms of private sector partici-pation in many sectors have provided a new body ofexperience to reinforce this pragmatic attitude. Theo-retical and institutional advances have also revealedwhen regulation is necessary and how to refine itsapplication. All this leads to two main conclusions.First, there are fewer infrastructure activities requir-ing government intervention than once believed.Second, when required, government interventioncan be exerted through less distorting instruments ofpublic policy than those traditionally used.

RENEWED COMMITMENT TO SOCIAL AND ENVIRON-

MENTAL CONCERNS. Political developmentsinclud-ing the trend in many countries toward democra-tization, pluralism, and decentralizationhavefueled a concern with finding more affordable andenvironmentally friendly solutions in infrastructure.This commitment has led to greater appreciation ofthe need to consult local communities, the poor, andgroups affected by environmental factors. At the

same time, increased efforts are being made to de-volve responsibility for infrastructure provision tolocal governments, to increase participation, and tofoster self-help.

Awareness that the poor (and future generations)are constituencies that must be answered to hasstimulated a search for alternative ways of provid-ing services or managing demands so as to broadenaccess while avoiding environmental problems. Rel-atively simple changes in design parameters forsewerage and improved design of latrines havemade sanitation affordable to low-income commu-nities while permitting private initiatives in financ-ing, maintenance, and manufacture of parts. Anincreasing range of technical, economic, and institu-tional alternatives to conventional wastewater treat-ment can reduce the need for costly filtration plants.Countries are adopting alternatives to large-surfaceschemes in irrigation such as drip, bubble, andsprinkler systems and low-level canals with low-liftpumpsthat are highly responsive to farmers'needs for water and are also environmentally sus-tainable. There is renewed interest in nonmotorizedmeans of transport, including bicycles and handcarts, and simple road improvements that enhancemobility in both rural and urban areas. Recognitionof the need to conserve scarce resources has led toefforts to avoid unnecessary infrastructure invest-mentsfor example, by promoting recycling andrecovery of solid waste materials; reducing wasteand effluents at the source; and managing demandfor water, power, and transport (Chapter 4). Indus-trial and developing countries are learning fromeach other in these areas.

The way ahead: a road map of reform

Awareness of past mistakes, together with new op-portunities, demands that a fresh look be taken atthe roles that governments or other public agenciesand the private sector should play in providing amore efficient and more responsive infrastructure.The challenge is to determine those areas in whichcompetitive market conditions can work and thosethat require public action. Within these broad pa-rameters, there is a menu of institutional optionsthat allow governments, public sector agencies, andprivate groups (both for-profit and nonprofit) to as-sume responsibility for different aspects of serviceprovision. The choices among the options will varyamong countries, on the basis of their economic, in-stitutional, and social characteristics. The spectrumof options is broad, but four main approaches can beidentified:

35

Page 24: Infrastructure: achievements, challenges, and opportunities · Infrastructure: achievements, challenges, and opportunities ... sanitation, and safe disposal of wastesare central

Option A: Public ownership and operation,through a public enterprise or government depart-ment

Option B: Public ownership but with privateresponsibility for all operation (and for financialrisk)

Option C: Private ownership and operationOption D: Community and user provision.

The remainder of this Report discusses howmore efficient and responsive provision of infra-structure can be achieved by improving incen-tivesthrough stronger mechanisms of account-ability and autonomy. Chapter 2 discusses ways tocreate accountability in a public agency or gov-ernment department (Option A) by establishingcommercial principles and through organizationalrestructuring (corporatization). It also reviews con-tracting instruments to permit better monitoringand performance of operations, and appropriatemechanisms for achieving financial autonomy.

Commercial principles are often very difficult toinstill permanently in the absence of effective com-petition. Chapter 3 discusses the scope and tech-niques for marshaling market forces to createaccountability through competition andwherecompetition alone is insufficientregulation. Chap-ter 3 also examines experiences with public owner-ship and private operation (Option B), in whichcompetition for the market is used, as well as pri-vate ownership and operation (Option C). Both of

36

these arrangements require appropriate sectoral re-structuring to maximize the opportunity for compe-tition and to ease the regulatory burden.

Chapter 4 examines issues that neither commer-cialization nor competition alone can addressproblems of externalities (particularly environ-mental), distributional equity, and the need forcoordination of investments. It discusses ap-

proaches for assessing and creating accountabilityto social and environmental concerns, through de-centralization of governmental responsibilities, par-ticipation by users and stakeholders (includingthrough "self-help" schemes, Option D), and plan-ning. Chapter 5 reviews how mechanisms of financ-ing infrastructure can create incentives for efficiencyby providing the disciplinary pressure of private fi-nancial markets. Because different aspects of infra-structure provision involve different kinds of risks,the chapter considers how a suitable packaging offinance using alternative sources and instruments(private and public) can lead to better risk manage-mentin addition to mobilizing increased fundsfor infrastructure investment. Chapter 6 returns tothe menu of options and shows how these can beapplied in different infrastructure sectors and coun-tries. The conditions for successful implementationof these options are also outlined. The chapter closeswith a broad assessment of the economic and finan-cial benefits that countries can gain by following thereform agenda presented in this Report.