88928 JUNE 2014 • Number 149 Sharing Prosperity by Closing South Asia’s Infrastructure Gap Dan Biller, Luis Andres, Matias Herrera Dappe, and Ashma Basnyat While South Asia’s average gross domestic product (GDP) growth in the past decade was only slightly lower than in East Asia and the Pacific, the region significantly lags behind in terms of infrastructure access. In some cases, South Asia’s access levels are more similar to sub-Saharan Africa, although considerable heterogeneity exists across the South Asia region (SAR)—some provinces, infrastructure types, and countries have much higher access levels than others. If SAR can’t close its US$1.7 trillion to US$2.5 trillion infrastructure gap, its growth performance is likely to stall. This note examines the type and magnitude of the infrastructure gap, and also looks at inequality of access to infrastructure across physical, poverty, and income spaces. Keeping in mind five key principles, there are several options for policy makers to consider for closing the infrastructure gap while also improving the equality of access. Despite recent rapid growth and poverty reduction, many between infrastructure access for all, regardless of where indi- people living in SAR remain unconnected to a reliable electri- vidual households are located, and quality access to where cal grid, a safe water supply, a sanitary sewerage disposal sys- most households are located, is a real policy challenge in its tem, and reliable roads and transportation networks. Further- extreme case. Yet, are there countries in South Asia that do a more, the region requires significant infrastructure better job of providing infrastructure access to their poor? Are investment (roads, rails, power, water supply, sanitation, and there infrastructure sectors that tend to be more regressive telecommunications) not only to ensure basic service delivery than others? What is happening with access to infrastructure and enhance the quality of life of its growing population, but services at the household and individual levels? also to avoid a possible binding constraint on economic In an effort to answer these questions, this note, based on growth. substantial research, critically reviews the status of infrastruc- The infrastructure gap is compounded by the fact that ture in SAR compared to other regions, as well as among and the SAR’s population suffers from unequal access to infra- within SAR countries. Several analyses then explore inequali- structure. It is commonly asserted that the poor have less ac- ty of access to infrastructure services across South Asia’s space cess to infrastructure than the rich, similar to the case of pri- (namely physical space, poverty space, and income space) and vate assets. In effect, a nonregressive access to infrastructure across time (how access of the young will influence future op- services would mean no correlation between actual access and portunities). These analyses help provide an estimate of the different poverty-related measures (such as households below total cost of regional infrastructure needs, along with the in- poverty lines and certain income and consumption levels). frastructure investment trends in SAR countries. Finally, this Whereas this may be desirable theoretically—especially for in- note examines ways to better use existing resources by re- frastructures with high public good characteristics—the choice thinking infrastructure service provision—including the role 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Table 1. SAR Lags Behind All But SSA in Access to Infrastructure Services Access to Access to Average GDP Telecommunication Electricity access improved sanita- improved water growth Urbanization rate access (per 100 (% of population; tion (% of (% of population; Region (2000–2012, %)a (2012) people; 2011)b 2010)c population; 2011)d 2011)e EAP 8.9 50 98 92 67 91 ECA 4.4 60 157 100 94 95 LAC 3.1 79 125 94 81 94 MNA 4.2 60 105 94 89 89 SAR 6.7 31 72 71 39 90 SSA 4.7 37 54 35 30 63 World 2.5 53 103 78 64 89 Source: World Development Indicators, unless otherwise noted. a. The average GDP growth for MNA is for the period 2000–2009. b. Telecommunications access is defined as the number of fixed and mobile lines. c. World Energy Outlook 2010, International Energy Association. d. Improved sanitation is defined as connection to a public sewer, a septic system, pour-flush latrine, simple pit latrine, or ventilated improved pit latrine. e. Improved water is defined as household connection, public standpipe, borehole, protected dug well, protected spring, or rainwater collection. of the private sector—as well as policy options to help the physical (“spatial”), poverty, and income spaces—an assess- poorest gain better access to infrastructure. ment that has never been conducted for this region.1 Mapping access to infrastructure with characteristics of households How South Asia Compares to provides a better understanding of issues that limit access for Other Regions some population groups, and allows for a better design and tar- Structural change in South Asian countries has been relative- geting of policies to expand access to infrastructure services. ly slow compared to that of East Asian countries. In 1990, Inequality across physical space they had similar urbanization rates (SAR, 25 percent; East One way of measuring inequality is to analyze whether Asia and the Pacific [EAP], 28 percent) and were close in some countries in South Asia are better at making access to terms of infrastructure service provision. But while they both infrastructure more spatially equal in-country than others. enjoyed high growth rates over the next two decades, EAP has This can be measured by zooming in to a lower administra- seen rapid urbanization (50 percent in 2012), while SAR has tive level (such as a district or province) and measuring in- remained the least urbanized region in the world (31 percent, equality with Gini coefficients (a Gini coefficient of zero compared to the world urbanization rate of 53 percent). represents perfect equality, while a coefficient of one repre- At this point, SAR’s access to infrastructure services sents maximal inequality). The goal is to come up with a closely resembles sub-Saharan Africa (SSA), even though its country-level measure of spatial inequality of infrastructure economic growth is second only to EAP (table 1). access adjusted by household spatial distribution, which is presented in table 2. How Access to Infrastructure Varies within The results show quite a varied picture, with Maldives South Asia having the lowest—and Afghanistan the highest—inequality So who has access to each type of infrastructure? The analysis of access to infrastructure services in the region. These results starts with a look at inequality of access across SAR, weighing are not surprising, especially in the case of Afghanistan, given Table 2. Gini Coefficients of Access to Infrastructure in South Asia Highlight Tremendous Inequality (Gini coefficients for access, adjusted by household distribution) Service Afghanistan Bangladesh Nepal Sri Lanka India Bhutan Pakistan Maldives Improved water 0.12 0.01 0.04 0.01 0.06 0.00 0.02 -0.10 Improved — -0.01 0.06 0.01 0.29 0.18 0.02 0.01 sanitation Electricity 0.49 0.11 0.04 0.04 0.15 0.10 — 0.00 Cooking gas 0.50 0.49 0.24 0.33 0.35 0.22 0.00 0.01 Phone 0.28 0.15 0.05 0.03 0.20 0.19 0.03 0.00 Source: Authors’ calculations based on surveys presented in Andrés, Biller, and Herrera Dappe (2013). Note: The Gini coefficients are estimated over a sample of administrative subdivisions selected on each country. 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise its level of development and years of conflict. What is surpris- Sri Lanka—is: do poorer regions within a country have less ing, however, is Pakistan’s relatively low levels of spatial in- access to key public services as a whole? This can be an- equality. One possible explanation is the country’s higher ur- swered by constructing two infrastructure indexes that en- banization rate, with access to infrastructure services more compass only the basic infrastructure services that have the skewed to its cities relative to other countries in the region. highest impact on welfare (such as water, sanitation, and Are some infrastructure services better than others in electricity) and then using these to generate maps identify- how they are distributed within countries to households? ing where poverty and the location of infrastructure ser- Again, the results show big variations. The most unequally vices intersect. distributed service throughout SAR is cooking gas (LPG). Results show that leading regions (those with a lower Heavy use of biomass for cooking rather than the cleaner LPG poverty level) generally have better access, but lagging regions implies an intergenerational and intragender trade-off be- (those with a higher poverty level) do not necessarily have less cause the significant pollution risks of indoor air contamina- access. As expected, in India, the lagging states have lower ac- tion affect mostly children and women. The most equally cess to basic infrastructure in contrast with the leading states. distributed service throughout SAR is improved water, which This is intuitively expected. The curious exception is the is of course important in terms of welfare impacts such as northeast area bordering Bangladesh, Bhutan, China, and health and hygiene. The other sectors—improved sanitation, Myanmar. The exception found in India is more prevalent in electricity, and phones—fall somewhere in between the two Sri Lanka, where basic infrastructure seems to be more inclu- extremes. sive. Access is widely spread, and the quality of these services Inequality across poverty space in the country is generally good. It is clear that the leading re- Another way of judging inequality is to analyze how pockets gion—the Western Province—enjoys better access and a lower of poverty fit into the picture, in effect, introducing a socio- poverty rate. Yet, for the lagging provinces, the story is more economic variable. The analysis assumes that a country mixed, except for those areas more affected by Sri Lanka’s 30- with a higher poverty rate will have worse access to infra- year conflict.2 structure services than a country with a lower poverty rate. Inequality across income space But how strong is that link? To ascertain whether this hy- How equitably is access distributed across different income pothesis holds, the district poverty rate (percentage of peo- levels—that is, for those with access, is it the richer individu- ple in each district that live under the poverty line) is corre- als who have the bulk of the access, or do richer and poorer lated with the district rate of access to infrastructure individuals tend to have more similar access? This matters (percentage of households that have infrastructure in each because it enables policy makers to better target policies to district) for India, Afghanistan, and Sri Lanka. In this analy- expand access. The answer can be found by comparing in- sis, high regressivity means that poor districts typically have come quintiles and access rates for Afghanistan and Sri less access compared to richer districts, whereas low regres- Lanka. sivity means access is more equitably distributed among Results show that in Sri Lanka and Afghanistan, the rich poor and rich districts alike. enjoy better access than the poor, but the countries differ Results show that, overall, the link is regressive, but the greatly in how equal that access is across incomes. In Sri Lan- strength of that link varies greatly among the countries and ka, the difference in access across quintiles is small—all quin- among sectors. tiles are close to the mean—meaning that there is an almost • India shows strong regressivity of infrastructure service equal share of access to infrastructure regardless of income access except in water. The water exception applies in the quintile. The opposite story is true in Afghanistan. However, other countries as well. some services (like water) are more equitably distributed than • Sri Lanka shows relatively weak regressivity for infra- others (like cooking gas) among those with access. In fact, in structure service access. The exceptions to this trend are Afghanistan, the equality of access across income quintiles for cooking gas and phones, which show a much stronger improved water is particularly striking. link. How Access to Infrastructure Affects • Afghanistan shows relatively strong regressivity, except in Opportunities for Youth water and in Kabul (its capital), an area with one of the country’s lowest poverty rates. Cooking gas seems less re- So, how unequal is access to infrastructure among children? gressive in Afghanistan than in India or Sri Lanka. After all, infrastructure investment choices made today to fill Inequality across physical and poverty space the gap affect current and future generations. Moreover, not Yet another question to be explored—this time by bringing addressing the infrastructure gap threatens both welfare and together poverty data and physical access data for India and economic growth in the medium and long term. 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise The Human Opportunity Index: Access to infrastructure as For example, in the case of improved sanitation (table 3), as ex- opportunity pected, countries with the highest coverage (Maldives and Sri The main instrument for measuring the inequality of access Lanka) feature the lowest dissimilarity index, and therefore, to infrastructure across time is the Human Opportunity In- the HOI is very close to the coverage. In contrast, countries dex (HOI). It essentially measures how personal circumstanc- with low levels of coverage (such as Bangladesh) are associated es impact a child’s probability of accessing the services that with higher discount rates, and thus lower HOIs. However, if are necessary to succeed in life. This is critical because the op- two similar access rates, such as in the case of India (36 per- portunities a child gets throughout life are determined direct- cent) and Nepal (37 percent) are considered, there is a signifi- ly by the circumstances related to access to infrastructure ser- cant difference in how that access to sanitation is distributed, vices during their formative years—not necessarily to their with Nepal (0.14) being more even than in India (0.24)— personal decisions or level of effort. which results in India having a lower HOI (0.27) than Nepal The analysis calculates an HOI that is focused on basic (0.32). Also, a country with higher coverage, like Pakistan (44 infrastructure as opportunities and the importance of both percent), but featuring the same dissimilarity index as India improving overall access to it and ensuring its equitable alloca- (and thus higher than for Nepal), still ends up with a higher tion to achieve key socioeconomic outcomes—such as early HOI than India, but similar to that of Nepal. In the case of childhood development, education level, good health, and ac- access to improved water, rates are high enough to guarantee a cess to information. It can be interpreted as a composite indi- low dissimilarity index. Electricity also follows the pattern of cator of two elements: (i) the level of coverage of basic oppor- decreasing dissimilarity index as coverage is higher. tunities necessary for human development (such as access to primary education, water and sanitation, or electricity); and Pinning Down the “Infrastructure Gap” (ii) the degree to which the distribution of those opportuni- Over time, societies inherit man-made infrastructure stock ties is conditional on circumstances children are born into from previous generations. Yet different factors influence de- (such as gender, income, or household characteristics). This mand and supply, and, as countries grow, these needs—both study looks at four circumstances: (i) household size, (ii) loca- the type of infrastructure and the quality of service provi- tion (urban versus rural), (iii) education of household head, sion—are likely to evolve. The infrastructure gap can be as- and (iv) gender of household head.3 sessed using a framework with a five-step process (figure 1). Inequality of opportunity in the access to infrastructure The framework shows where a country is today (1); where a services country would like to be at a given point in time (2, the dif- South Asian countries with better infrastructure coverage ference between the two points is the infrastructure gap); typically also provide more equitable access for households how far business-as-usual scenarios (shown by the dotted with children under age 15, thereby producing higher HOIs. blue line) will take the country toward its goal (3); how far Table 3. Better Coverage, More Equitable Access, and Thus Higher HOIs (access to infrastructure services and HOI for households with children under age 15) Afghanistan Bangladesh Bhutan India Maldives Nepal Pakistan Sri Lanka Country/year 2008 2006 2007 2007 2009 2011 2006 2010 Improved sanitation Coverage — 45% 38% 36% 94% 37% 44% 90% Dissimilarity Index — 0.15 0.24 0.24 0.01 0.14 0.23 0.03 HOI — 0.38 0.29 0.27 0.93 0.32 0.34 0.87 Improved water Coverage 46% 98% 96% 83% 86% 88% 92% 88% Dissimilarity Index 0.11 0.00 0.01 0.03 0.09 0.02 0.01 0.02 HOI 0.41 0.97 0.95 0.80 0.78 0.86 0.91 0.86 Electricity Coverage 17% 50% 72% 68% 100% 75% . 85% Dissimilarity Index 0.58 0.20 0.12 0.12 0.00 0.08 . 0.05 HOI 0.07 0.40 0.63 0.60 1.00 0.69 . 0.81 Source: Authors’ calculations based on IRACSO (2008) for Afghanistan; BBS and UNICEF (2007); BNSB (2007) for Bhutan, Institute for Population Sciences (2007–8) for India; MOHF (2009) for Maldives; MOHP (2012) for Nepal; NIPS (2008) for Pakistan; and Department of Census and Statistics (2009–10) for Sri Lanka. All estimations represent the proportion of access in the sample with children below 15 years of age. 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 1. A Framework for Assessing Infrastructure Needs an increase of up to 3 percentage points from the 6.9 percent of GDP invested 2009 (table 5).5 coverage 2 target As expected, these overall patterns have been largely financial gap 5 driven by India, which accounts for the biggest share of total infrastructure infrastructure investment in South Asia during 1973– gap financial and policy options using 4 2009. In fact, infrastructure investment in India accounts existing resources for, on average, 79 percent of total investment in the region. 3 level of investments The second largest contributor—Pakistan—merely has an av- (% of GDP) 1 erage share of 12 percent, and is followed by Bangladesh with 7.9 percent, Nepal with 1.0 percent, and Bhutan with 0.2 percent. today time Policy Options to Provide Better Source: Authors’ illustration. Infrastructure Services Which policy options should SAR policy makers focus on to financial and policy options using existing resources (shown improve the level and quality of service provision for their di- by the dotted red line) could take the country toward its goal verse populations? (4); and the remaining financial gap (5) that will need to be First, rehabilitate and maintain existing assets. South bridged. Asian governments should invest in rehabilitating and main- At this point, most governments in SAR have some esti- mates of the investments required to reach certain targets, taining infrastructure assets to deliver services efficiently and such as for 24/7 electricity supply and the Millennium Devel- sustainably, moving away from the build, neglect, and rebuild opment Goals for water and sanitation, but those estimates mindset. Inadequate maintenance triggers progressive deteri- are not consistent across the region. That is why different oration in the quality of the infrastructure services, which methodologies were developed for different sectors. These hurts users (for example, higher costs because of imperfect models also allow analyses to calculate and compare the costs and costly substitutes and worse social outcomes in health of different sets of targets. and education) and development outcomes. In India, the Na- So how much money will be needed to close the infra- tional Transport Development Policy Committee estimates structure gap? SAR needs to invest approximately US$1.7 that poor road maintenance costs the country about Rs. 350 trillion to US$2.5 trillion in infrastructure until 2020— billion annually. While funds for new construction are some- equivalent to US$1.4 to US$2.1 trillion at 2010 prices (table times easier to obtain and implement, obtaining funding for 4).4 Going forward, a mix of investment in infrastructure maintenance is more difficult because it is needed on a regu- stock and implementation of supportive reforms will enable lar basis. However, there are mechanisms appropriate to each SAR to close its infrastructure gap. In GDP terms, if invest- sector that can be implemented to improve maintenance. In ments are spread evenly over the years until 2020, SAR needs the road sector, for example, some governments have adopted to invest between 6.6 and 9.9 percent of 2010 GDP per year— or considered a “road fund” type of arrangement to support Table 4. SAR’s Total Investment Could Reach US$2 Trillion (investment requirements 2011–20, total, in 2010 US$ billions)   India Bangladesh Pakistan Nepal Sri Lanka SAR (5) SAR (8)  Sector Low High Low High Low High Low High Low High Low High Low High Transport 340.0 595.0 36.0 45.0 17.2 21.5 3.7 5.5 10.8 18.0 408 685 411 691 Power 375.0 468.8 11.0 16.5 64.0 96.0 5.3 7.0 4.8 9.0 460 597 464 603 WSS 95.0 162.0 12.0 18.0 9.3 14.0 1.7 2.6 0.6 1.8 119 198 120 200 Solid waste 32.5 65.0 2.1 4.2 3.3 6.7 0.4 0.5 0.2 1.3 39 78 39 78 Telecommuni- 150.0 225.0 5.0 5.0 12.4 12.4 0.4 0.6 2.0 2.5 170 246 171 248 cations Irrigation 140.0 210.0 7.7 11.6 9.7 14.6 1.6 2.3 2.5 3.1 161 242 163 244 Total 1,133 1,726 74 100 116 165 13 18 21 36 1,356 2,045 1,369 2,064 Source: Andrés, Biller, and Herrera Dappe (2013). Note: Estimations based on the technical models as well as extrapolations from other sectors where the models were not run. WSS = Water and Sanitation Sector. 5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Table 5. Closing SAR’s Infrastructure Gap Requires Higher Share of GDP Investment (investment requirements 2011–20, % of GDP, per year)   India Bangladesh Pakistan Nepal Sri Lanka SAR (5)  Sector Low High Low High Low High Low High Low High Low High Transport 1.97 3.44 3.60 4.50 0.98 1.23 2.32 3.49 2.17 3.64 1.97 3.31 Power 2.17 2.71 1.10 1.65 3.66 5.49 3.34 4.46 0.97 1.82 2.22 2.89 WSS 0.55 0.94 1.20 1.80 0.53 0.80 1.08 1.62 0.12 0.37 0.57 0.96 Solid waste 0.19 0.38 0.21 0.42 0.19 0.38 0.24 0.30 0.04 0.27 0.19 0.38 Telecommunications 0.87 1.30 0.50 0.50 0.71 0.71 0.27 0.40 0.40 0.50 0.82 1.19 Irrigation 0.81 1.21 0.77 1.15 0.55 0.83 0.99 1.48 0.50 0.63 0.78 1.17 Total 6.55 9.98 7.38 10.02 6.63 9.44 8.24 11.75 4.21 7.23 6.55 9.89 Source: Andrés, Biller, and Herrera Dappe (2013). Note: These percentages are based on investment requirements at 2010 prices and on the technical models as well as extrapolations from the other sectors where the models were not run. WSS = Water and Sanitation Sector. maintenance. Under such arrangements, maintenance funds Fourth, decentralize service provision in an appropriate are assured from a mandated tax on gasoline and diesel and manner. SAR countries should be rethinking how much to are deposited into an assured and independently operated decentralize—that is, distribute the administrative powers or fund, overseen by a board that includes either the public or functions of a central authority over a less concentrated area— private sector, or both. as a means of improving service delivery for the smallest units Second, reform service providers and ensure financial/ of society (households and individuals). As the World Bank’s operational sustainability. Service providers should be finan- 2004 World Development Report puts it: cially viable, able to plan and implement sound investment Decentralization…can strengthen the links and strategies, and improve operational performance for the long accountability between policy makers and citi- term. This requires: (i) reliable, steady, and adequate revenue zens—local governments are potentially more streams to fund operations and investment; (ii) capacity and accountable to local demands. It can also strengthen them between policy makers and independence without threat of political interference; and providers—local governments are potentially (iii) appropriate incentives for becoming more efficient and more able to monitor providers. remaining so. In practice, the experience with trying to decentralize Third, establish solid legal, policy, and regulatory frame- infrastructure service delivery is mixed—the biggest prob- works. South Asian governments need to have solid legal and lem often being a mismatch (typically financial or fiscal in policy frameworks as well as a transparent and well-designed nature) between responsibilities in infrastructure service and implemented regulatory framework for both public and private operators to attract private investment in line with the delivery and the ability to execute such responsibilities. But best organizational form for each service. For example, gov- when decentralization succeeds, it is the result of: (i) a fully ernments across the region need to set the conditions for an democratic, transparent, and inclusive (of the beneficiaries) even bigger role for the private sector in a service, such as pow- local decision process; (ii) the cost of local decisions fully er generation, which is better suited for liberalization than borne by local government; and (iii) no spillover of benefits public-private partnerships (PPPs), and shift the public re- to other jurisdictions. sources and efforts toward other services where the public Policy Options to Help the Poorest Gain sector has the comparative advantage. Fortunately, when the Better Access private sector invests in infrastructure in SAR, it tends to choose the optimal organizational forms. These frameworks While South Asian governments are moving to improve the provide clarity to the private sector, increasing the attractive- overall level and quality of infrastructure services, they must ness of private participation in infrastructure projects. They also take deliberate steps to improve the access of the poor— also allow the public sector to clearly define responsibilities keeping in mind the following five principles. and manage the risks associated with private sector participa- Access is fundamental, but usage determines impact. That tion. A stable yet dynamic regulatory framework for infra- is why policy makers should complement access to infrastruc- structure services is particularly critical for: (i) attracting and ture with policies to incentivize the use of services, or make supporting desired levels of investment; (ii) ensuring service potential benefits more obvious or attainable. One way to do sustainability; (iii) protecting customers; and (iv) guarding this is to focus on subsidizing (implicitly or explicitly and the public interest. with sunset clauses) the infrastructures that provide the 6 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise greatest public benefit (public good) in contrast to those that ture also argues that infrastructure service expansion is close- provide large private benefits. This should be true across in- ly linked to rent seeking, since richer districts are better able frastructure sectors as well as within sectors. Another way is to lobby the government for infrastructure provision (Cadot, to focus on improving women’s access to services, because im- Röller, and Stephan 1999). provements in household outcomes are larger when women’s Although subsidies may improve affordability among un- access is increased. derprivileged groups, they can also increase income inequality. Ability-to-pay for access to infrastructure services cannot Subsidies tend to be captured by those who have political be the only instrument to determine provision. Infrastructure connections, which, at least among unconnected house- services have strong market failure characteristics, underscor- holds, tend to be the more middle class households. Ajwad ing the need for adequate regulation. Some infrastructure sec- and Wodon (2002) found that, in Bolivia and Paraguay, the tors are still close to natural monopolies (such as piped water marginal benefit of improved access to a service tended to be and off-site sanitation services). Many are associated with two to three times higher among the upper two quartiles. strong externalities (negative and positive) and public goods Thus, while all income quartiles benefited from decentral- (and bads) characteristics, as in the cases of lack of pollution ization, the richer 50 percent benefited more than the poor- treatment or lack of access to cooking gas. Information issues er 50 percent, a net effect that would tend to increase in- also abound. Moreover, since infrastructure can spur growth, come inequality. Estache (2005) points out that, in Latin relying on the ability-to-pay criterion might undercut efforts America, as much as 60 to 80 percent of cross-subsidy to reduce poverty and share prosperity. schemes “were aimed at households well above the poverty Some infrastructure programs are too costly to be sustain- threshold, while as much as 80 percent of poor households ably implemented without cost-recovery mechanisms that al- failed to benefit.” Thus, it is not surprising that even as abso- low them to be self-supporting. Balancing the trade-off be- lute levels of connection increase, regressivity in access to tween providing access to infrastructure services and fully infrastructure may still prevail. charging for these services is seldom easy. It involves under- Potential instruments standing the economic characteristics of particular infra- Keeping in mind the five principles, several instruments structure sectors and the technology available for provision have the potential to help policy makers improve infrastruc- under different physical, political, and socioeconomic con- ture access. ditions. Take the case of piped water provision, which is a Subsidies for connection rather than service consumption. private good. It has important market failures associated To avoid some of the drawbacks of subsidies, policy makers with it, but essentially individual households have clear in- can adopt measures that reduce the cost of providing network centives to pay for a superior service compared to other services or improve the ability of poor households to pay for forms of getting water in an urban environment. Yet, the service at a given cost (Komives et al. 2005). These would be fees for piped water are rarely enough to achieve full cost available only to unconnected households, reducing or elimi- recovery and often rely on direct or indirect subsidies that nating the price customers have to pay to connect to the sys- burden public budgets. Nonetheless, the expansion of tem. Alternatively, policy makers can subsidize lower service piped water provision is often part of political manifestos levels that the more affluent find less attractive, such as social during election campaigns. Now take the case of flood con- connections. trol, which is a public good. Direct cost-recovery mecha- Targeted interventions. Usually these instruments focus nisms like tariffs are difficult to design, but inadequate flood narrowly on a certain district or group that is perceived as un- control in a locality can cause households to incur large derserved. This approach has the advantage of fewer spill- costs through the loss of private assets and even lives. Bud- overs—that is, it is less likely that the intervention will end up getary allocation for flood control is often inadequate and benefitting unintended beneficiaries. In Mexico, this instru- the service is underprovided. ment is being piloted in a program that provides conditional Usually the aim of the policy maker is to balance infra- cash transfers to the poorest segments of the population, structure access (especially basic infrastructure), while allow- Oportunidades (Opportunities). Under this program, energy ing wealthier populations to shoulder most of the burden of im- subsidies are targeted using mechanisms that ensure they proving coverage for all. Given the equality achieved in reach the poorest residents. A note of caution here, however, improved water access in South Asia, one would be tempted is that because these interventions are operating only within to conclude that this objective is present as the service ex- impoverished and underserved areas, they may face issues pands. But this conclusion might conceal rent-seeking behav- such as inadequate staffing, funding, technical capacity, and ior, where the wealthier capture proportionally larger lack of political will (Menéndez 1991). amounts of rents that otherwise could be used for expansion Institutional groups. There are also a number of options and quality improvement for all, and not just a few. The litera- for designing programs to reduce elite capture and increase 7 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise the power of impoverished groups to allocate resources to- crease was mainly due to the region’s large investment in ward their priorities. These include: electricity generation. • Institutional re-centering. Organizations can be created The challenge of increasing access to infrastructure ser- that have a primary purpose of reducing poverty through vices across South Asia is compounded by the inequality in infrastructure provision. For example, Bolivia’s Emer- the distribution of existing access for households and indi- gency Social Fund was a temporary organization that was viduals. That is why providing some level of access is a start— created to finance infrastructure projects in underserved even if those services are not of the highest quality. At the communities. same time, policy makers should consider which types of ser- • Community participation. Incorporating transparent vices best fit each population’s needs (such as septic tanks for mechanisms for underserved people to easily provide in- a mountainous region, but sewerage lines for more accessible put into the design and decision-making process of infra- urban areas). This study sheds light on who has access in terms structure projects could potentially allow them to com- of space (a current framework) and over time (future genera- pete with the more informal mechanisms that richer tions). Countries with higher per capita income (like the Mal- populations use to influence decision making (Menén- dives and Sri Lanka) enjoy better access to infrastructure ser- dez 1991). vices both spatially (geographically within the country) and • Innovative mechanisms. Service delivery mechanisms based on income—even though conflict-ridden areas are clear- need to evolve to respond to the challenges of coverage, ly worse off. Among sectors in SAR countries, some, such as affordability, use, and sustainability. This is particularly water, tend to be more equally distributed than others, such important given that poor households tend to pay more as sanitation, energy, and phones—with the widespread use of for services when they have to obtain them through non- firewood for cooking, especially among the poor, somewhat network solutions. For example, community-based orga- surprising. Moreover, within SAR countries, some states and nizations and user groups can contribute to planning and districts have better access than others. Leading regions with- operations; nongovernmental organizations can help in a country typically have better access, but lagging regions with monitoring and evaluation, promoting social ac- do not necessarily have worse access. But if a poorer country countability and raising awareness; and the private sector or a poorer state has better access to a given infrastructure ser- can get involved with investment and delivery (Andrés vice than a richer country or a richer state, there is then hope and Naithani 2013). As they are tested and main- that policy makers can adopt measures that will improve ac- streamed, these context-specific alternative mechanisms cess in a manner that increases shared prosperity. are becoming part of the policy toolkit. There is no simple explanation for these inequalities, al- though certainly geography matters, policy intent matters, Conclusions and some household characteristics matter. At the country If South Asia hopes to meet its development goals and avoid level, household characteristics like location and education the risk of slowing—or even halting—growth, poverty allevia- are the main explanatory factors. Location seems obvious, but tion, and shared prosperity, it is essential to make closing its education does not—unless it is linked to income poverty and huge infrastructure gap a priority. Even though SAR’s eco- remoteness of household location (even among rural areas). nomic growth follows that of EAP, its access to infrastructure At the state level, education actually starts to become a bigger rates (sanitation, electricity, telecommunications, and trans- factor than at the country level. And while the contribution port) are closer to that of SSA—the one exception being water, of gender of the household head seems negligible at the coun- where SAR is comparable to EAP and the Latin America and try level, there are a number of states (and districts within Caribbean region. According to businesses in South Asia, in- them) where access is clearly biased toward male-headed frastructure is a major or severe hindrance to their growth, households. and electricity is the largest problem, with transport also an SAR also needs to rethink the infrastructure service para- obstacle for regional and international trade. The good news digm to bring in the private sector and decentralize adminis- is that policy makers do not have to choose between growth trative power functions. The sheer size of the gap and the and welfare, as there is enormous potential for them to be mu- macroeconomic situation in South Asia dictate that the re- tually supportive. gion taps other funding sources. However, this situation The cost to close this gap by 2020 is estimated at US$1.7 should also be seen as an opportunity to rethink and improve trillion to US$2.5 trillion. If investments are spread evenly how infrastructure services are delivered. One way to do this over these years, SAR needs to invest between 6.6 and 9.9 is by broadening service provision to give the private sector a percent of 2010 GDP per year—an increase of 3 percentage bigger role—whether through PPPs or regulated privatization points over the current 6.9 percent invested by SAR coun- and market liberalization. Another way is by giving greater tries in 2009, up from 4.7 percent in 1973. The past in- administrative powers and functions to lower levels of govern- 8 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise ments, although the degree to which such decentralization is 3. These circumstances reflect previous inequality of oppor- desirable will depend on the nature of the investment, the tunities studies and are in line with similar analyses that are reason it is being provided, how it is being financed, and part of Biller, Andres, and Herrera Dappe (2013). where it is located. 4. See Andrés, Biller, and Herrera Dappe (2013) for the de- Policy choices should be aimed at increased shared pros- scription on the methodology for computing these estimates. perity. Key principles to keep in mind are: (i) access is funda- 5. The Economist (2014a,b) reckons that the gap amounts to mental, but usage determines impact; (ii) ability-to-pay for US$57 trillion in 2010 prices worldwide for 2013—30. It also access cannot be the only factor to determine provision; (iii) estimates that annual infrastructure spending currently some infrastructure programs are too costly to be sustainably amounts to US$2.7 trillion, when US$3.7 trillion is needed. implemented without cost-recovery mechanisms that enable These estimates don’t include solid waste and irrigation. them to be self-supporting; (iv) market failures need to be cor- While not perfectly comparable, these global figures, when rected to avoid rent-seeking behavior; and (v) subsidies should compared with South Asian figures, provide a scale of the be designed so that they do not exacerbate income inequality. problem in the region. The choice of possible instruments includes subsidizing con- nections (rather than service consumption), adopting target- References ed interventions, creating organizations dedicated to reduc- Ajwad, M. I., and Q. Wodon. 2002. “Who Benefits from Increased ing poverty through infrastructure provision, and asking Access to Public Services at the Local Level? A Marginal Benefit NGOs to help with monitoring and evaluation, promoting Incidence Analysis for Education and Basic Infrastructure.” social accountability, and raising awareness. World Bank Economists’ Forum 2: 155–175, Washington, DC. If shared prosperity is one of the ultimate goals of poli- Andres, L., D. Biller, and M. Herrera Dappe. 2013. “Infrastructure Gap in South Asia: Infrastructure Needs, Prioritization, and cy makers, it is important to get accurate infrastructure Financing.” World Bank, Washington, DC. data. Existing data have enabled creation of a baseline to Andres, L., and S. Naithani. 2013 “Mechanisms and Approaches help track progress toward closing the infrastructure gap in Basic Service Delivery for Access and Affordability.” World and have answered some questions about who has access to Bank, Washington, DC. infrastructure services now and how that is likely to affect BBS (Bangladesh Bureau of Statistics), and UNICEF (United Na- future generations—and thus the equality of opportunities tions Children’s Fund). 2007. Multiple Indicator Cluster Survey across the region. This work can be expanded and improved (MICS) 2006. Dhaka, Bangladesh. by considering more circumstances, along with exploring Biller, D., L. Andres, and M. Herrera Dappe. 2013. “Inequality of alternative indicators of access, use, and quality of infra- Access to Infrastructure Services in South Asia.” World Bank, structure services. But to continue, better data are needed. Washington, DC. BNSB (Bhutan National Statistics Office). 2007. "Bhutan Living Without comprehensive data, private and public invest- Standard Survey 2007." ments may miss their targets of leveling the playing field and Cadot, O., L. H. Röller, and A. Stephan. 1999. “A Political Economy actually end up increasing the inequality of infrastructure Model of Infrastructure Allocation: An Empirical Assessment.” service provision. Discussion Paper No. FS IV 99–15, WZB, Wissenschafts- zentrum Berlin für Sozialforschung, Forschungsschwerpunkt Acknowledgments Marktprozeß und Unternehmensentwicklung. This paper is based on “Reducing Poverty by Closing South Department of Census and Statistics. 2009–10. “Household Asia’s Infrastructure Gap” (Andrés, Biller, and Herrera Dappe Income Expenditure Survey (HIES).” Ministry of Finance and Planning, Colombo, Sri Lanka. 2013). The authors are grateful to Rahul Kanakia for editorial Estache, A. 2005. “On Latin America’s Infrastructure Experience: assistance and to Juan A. Echenique for research assistance. Policy Gaps and the Poor.” In Reality Check: The Distributional About the Authors Impact of Privatization in Developing Countries, ed. J. Nellis and N. Birdsall, 281–96. Center for Global Development. Dan Biller is MIGA’s Economics Unit Sector Manager. Luis An- Institute for Population Sciences. 2007–8. “District Level House- drés is Lead Economist, Matias Herrera Dappe is Senior Econo- hold Survey (DLHS-3).” Ministry of Health and Family and mist, and Ashma Basnyat is a Consultant, all in South Asia Re- Welfare, Delhi. IRACSO (Islamic Republic of Afghanistan Central Statistics Orga- gion’s Sustainable Development Department. nization). 2008. National Risk and Vulnerability Assessment. Notes Komives, K., J. Halpern, V. Foster, Q. Wodon, and R. Abdullah. 2005. “Water, Electricity, and the Poor: Who Benefits from 1. See Biller, Andrés, and Herrera Dappe (2013) for a detailed Utility Subsidies?” World Bank, Directions in Development, discussion about the methodology in this section. Washington, DC. 2. See Biller, Andrés, and Herrera Dappe (2013) for further MOHF (Ministry of Health and Family, Maldives), and ICF Macro. discussion on this issue. 2010. “Maldives Demographic and Health Survey (DHS) 9 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise 2009.” Calverton, Maryland. The Economist. 2014a. “A long and winding road,” March 22. MOHP (Ministry of Health and Population, Nepal), New ERA, and ———. 2014b. “The trillion-dollar gap,” March 22. ICF International Inc. 2012. “Nepal Demographic and Health WHO (World Health Organization) UNICEF-JMP (United Nations Survey (DHS) 2011.” Kathmandu, Nepal; Calverton, MD. Children’s Fund Joint Monitoring Programme for Water and Menendez, A. 1991. “Access to Basic Infrastructure by the Urban Sanitation). 2013. http://www.wssinfo.org/. Poor.” Paper No. 28, World Bank, Washington, DC. World Bank. 2004. World Development Report 2004: Making NIPS (National Institute of Population Studies, Pakistan), and Services Work for Poor People. Washington, DC. Macro International Inc. 2008. “Pakistan Demographic and ———. 2013. World Development Indicators. http://data.worldbank. Health Survey (DHS) 2006–7.” Islamabad, Pakistan. org/country. Washington, DC. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. They are produced by the Poverty Reduction and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at: www.worldbank.org/economicpremise. 10 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise