WATER 4 T A R I . . S & S U B S I D I E S I N S O U T H A S I A APERP Do current water subsidies reach the poor? RE.ORM O. THE WATER AND SANITATION SECTOR is occurring in many countries, and offers the potential to improve services to all. Of particular concern, however, is the situation of the poor, and reform must be designed so that they receive increased access to affordable services. A key issue in this regard is water pricing, which is one of the main variables affecting the distribution of benefits between different stakeholders. However, experience shows that water pricing, and the subsidies which are often delivered through water tariffs, can be a source of major inefficiencies in the sector. Project While affordability has been one of the prime concerns of those setting tariffs and designing .IRE-D subsidies, there may be significant flaws in many common pricing strategies and subsidy delivery Indo-USAID mechanisms. Rather than providing affordable by water to the poor, these may in fact be leading Picture to financial unsustainability of utilities, lack of L access to services, and inequity. The reform arge scale subsidization of the water sector is typical in process provides the opportunity to rationalize south Asian countries. .or example, the federal and state and reconsider the design of tariff and subsidy governments of India together spend an estimated structures, and seek new ones which may provide Rs. 5,470.8 crore (US$ 1.1 billion) per year on subsidizing the better results. water sector; accounting for around 4% of all government subsidies in India and amounting to 0.5% of GDP. These subsidies are typically justified in terms of the need to ensure that essential This series of papers is designed to examine these issues water services remain affordable to the poorest in society. in South Asia. It is designed to present the basics of tariff and subsidy issues, to disseminate recent research findings, However, there are a number of reasons for questioning whether and to stimulate debate on the subject. these subsidies really benefit the poor. .or example, the tariff The preparation of these papers was funded by the Public- structures typically used in the water sector in South Asia do not Private Infrastructure Advisory .acility (PPIA.). Additional financing was provided by the World Bank, the World discriminate between rich and poor, so that everyone benefits Bank Institute and the Water and Sanitation Program. from the general subsidy to water consumption. .urthermore, a high proportion of poor people in South Asia do not have private Water and Sanitation Program DO CURRENT WATER SUBSIDIES REACH THE POOR? connections (the figure for India is 60%); as a result effective way to channel resources to the poor, these they are unable to benefit from the heavy subsidization also suffer from serious disadvantages. of this service. Even though this group may benefit from subsidies to public taps and tanker deliveries, these The Two City Studies services deliver such small amounts of water that the This paper is based on the results of two city-level overall value of the subsidy is quite small compared household surveys conducted in Kathmandu (Nepal) with that going to users of private taps. in April 20011 and Bangalore (India) in August 2001. Nonetheless, until now, there has been no hard empirical evidence to confirm or reject the hypothesis that water subsidies may be failing to reach the poor. This paper summarizes the results of some new survey-based research that estimates the value of the water subsidy received by individual households in two South Asian cities, and also estimates the income A high proportion of poor people in South Asia do not have private connections; as a result UNICE. by they are unable to benefit from the heavy Picture subsidization of this service. Both surveys collected a wide range of information including household water expenditure, type of water supply, physical characteristics of the dwelling, and levels of those households. Using both pieces of socioeconomic characteristics of the household information together, it is possible to evaluate (including overall income or expenditure). The the extent to which subsidies are actually reaching surveys interviewed a representative sample of 1,500 the poor. households in Kathmandu and 2,905 households The analysis shows that about three quarters of in Bangalore. the subsidies available to water utilities in the cities The two cities have a number of features in common. of Bangalore and Kathmandu are delivered in a way In both cities, service is provided by a large publicly- which completely fails to reach the poor. The vast owned utility (Bangalore Water Supply and Sewerage majority of subsidy resources are spent on lowering Board, BWSSB, in the case of Bangalore and Nepal tariffs for private connections through an Increasing Water and Sewerage Corporation, NWSC, in the case Block Tariff, even though many poor people do not of Kathmandu). In Bangalore, BWSSB reaches around 2 use water from these types of connections. While 80% of the population, of which 73% have private taps. subsidized public taps were found to be a more In Kathmandu, NWSC reaches 77% of the population, 1Carried out by Research Triangle Institute, with support from WSP, as part of research into the impact of proposed water and sanitation sector reform on the poor W A T E R T A R I . . S A N D S U B S I D I E S I N S O U T H A S I A 66% with private taps and a further 11% with public .igure 2: Water consumption patterns of poor and non-poor taps. As might be expected, the vast majority of non- poor customers are supplied through private taps (.igure 1). .or poor customers, the picture differs by city. In Kathmandu, there are more poor customers with private taps than those relying on public taps, whereas in Bangalore the opposite is true. Consumption by households with private taps in both cities is estimated to be of the order of 20 cubic meters per month. It is interesting to note that average .igure 1: Distribution of existing water sources Given the similarities in the consumption patterns and tariff structures, average monthly household expenditure on water falls in the range US$ 1.50-2.50 per month for both cities. In both cases, water at public taps is provided free of charge. .igure 3: Comparison of water tariffs consumption by non-poor households is only about 20% more than that by poor households (.igure 2). On the other hand, those relying on public taps consume little more than one cubic meter per month in Kathmandu but almost five cubic meters per month in Bangalore. Both utilities apply a relatively similar rising block tariff structure, although the lifeline block in Bangalore at 25 cubic meters per month is more than double that 3 applied in Kathmandu; moreover the cost of part of this Due to the intermittent nature of water supply, first block is levied as a minimum fixed charge for water as well as deficient water quality, households in both consumption. (.igure 3). cities face significant coping costs. These include W A T E R T A R I . . S A N D S U B S I D I E S I N S O U T H A S I A DO CURRENT WATER SUBSIDIES REACH THE POOR? constructing storage tanks to insure against service cuts, Three pieces of information were used from the queuing at public standposts for considerable periods survey to calculate the amount of subsidy that each waiting for the water to arrive, performing household household received: monthly household expenditure on treatment of drinking water. The survey indicates that water, monthly household consumption of water, and households in both cities spend an overall average of the average cost of producing water in the city where US$ 1.00-1.50 on coping costs. This represents about the household lives. two thirds of the average monthly water bill. Households were asked directly about their monthly .inally, there are also important differences between expenditure on water, which, in most cases, was reported the two cities. .or example, in Kathmandu only 42% of from memory. households with private taps are metered, while the Monthly water consumption was estimated according remainder pay a fixed monthly charge that varies to a variety of different techniques depending on the according to whether they are connected to a distribution type of customer. Those relying on public taps were main line or branch line. In Bangalore, on the other asked to report the number of containers that they filled hand, virtually all legal customers of the BWSSB are on an average each day, together with the volume of metered and only those supplied from alternative private those containers. .or those with metered private taps, networks lack meters. Another key difference is that in consumption can be inferred by applying the known Kathmandu, most households tend to rely on a wide tariff structure to the reported expenditure. .or those portfolio of water sources, so that for example even those with unmetered private taps, consumption was imputed with private taps continue to make regular use of public based on other household characteristics, using a taps, and those who use public taps also collect water statistical water consumption model developed with data from other non-utility sources. In Bangalore, on the other from metered households. hand, most households tend to rely exclusively on either It is important to note that, due to the intermittent public or private taps. nature of water supply in South Asia, meters can often get damaged, either under-recording true consumption How were Water Subsidies Calculated? or breaking down altogether. This clearly makes it The subsidy received by each household can be thought difficult to achieve reliable estimates of water of as the difference between the full economic cost of consumption from water expenditure, particularly since producing the water that the household actually there are no reliable estimates of the extent to which consumes and the monthly charge that the household meters malfunction or under-record. The likely effect of actually pays to the utility (Box 1). The larger this gap, this phenomenon is to dampen the variation in water the larger the subsidy that the household implicitly consumption estimates between households. However, receives from government. as long as there is no systematic relationship between Box 1: Subsidy formula meter malfunctioning and income level, the phenomenon should not necessarily affect measurements Household subsidy of the distribution of water consumption across income 4 = .ull economic cost of producing the households monthly groups, which is the primary focus of this analysis. consumption The average cost of producing water was derived from - engineering consultants studies for each city that estimated Household monthly utility charge the operating and maintenance cost of the service. These W A T E R T A R I . . S A N D S U B S I D I E S I N S O U T H A S I A of inclusion (or proportion of total subsidy beneficiaries who are not poor), and the errors of exclusion (or proportion of the poor who are not subsidy beneficiaries). More precise definitions are provided in Box 2. What are the Key .indings? Project The results indicate that the average monthly subsidy .IRE-D per household with a private tap is US$ 8.88 in Kathmandu and US$ 15.74 in Bangalore. Moreover, the Indo-USAID average non-poor household receives 44% more subsidy by than the average poor household in Kathmandu, and Picture 15% more in Bangalore. Given much lower levels of yielded estimates of US$ 0.17/m3 in Kathmandu and US$ water consumption, the subsidies received by users of 0.34/m3 in Bangalore. Consultations with a number of public standposts are correspondingly smaller, with an water engineers suggested that, based on typical ratios of capital to operating expenditure, the full financial costs of water production (including all investment costs) The average non-poor household receives could be three times higher than the operating and 44% more subsidy than the average maintenance costs. poor household in Kathmandu, and .inally, for the purposes of evaluating whether subsidies reach the poor, the people falling into the 15% more in Bangalore. bottom 40% of the income distribution in each city are considered to be living in poverty. The use of official average value of US$ 0.71 in Kathmandu and US$ 3.74 poverty lines was deliberately avoided, since applying in Bangalore. However, poor households receive a larger a relative rather than absolute concept of poverty has subsidy via public taps than do non-poor households, the advantage of ensuring that the definition is consistent owing to the fact that the latter use this only as a backup across the two city cases. Using this definition of poverty source of water. the equity characteristics of existing subsidies are Looking at the overall distribution, the average subsidy evaluated based on a number of standard indicators. received by the richest 10% (decile) of the population is These include the leakage rate (or percentage of total 2-3 times as high as the average subsidy received by the subsidy resources captured by the non-poor), the errors bottom 10% of the population. (This can be seen in Table 1: Average monthly subsidies (US$) Private Tap Public Tap Kathmandu Bangalore Kathmandu Bangalore Poor 6.75 14.22 0.69 4.19 5 Non-poor 9.76 16.38 0.72 2.30 All 8.88 15.74 0.71 3.74 Note: Private and public tap subsidies are not mutually exclusive, since some households may be receiving subsidies from both sources. W A T E R T A R I . . S A N D S U B S I D I E S I N S O U T H A S I A DO CURRENT WATER SUBSIDIES REACH THE POOR? Box 2: Understanding Errors of Exclusion and Inclusion A key consideration in evaluating the efficacy of subsidies of inclusion, since they indicate that the subsidy is failing to is measuring to what extent they succeed in reaching the meet its primary objective of helping the poor. poor. Two standard indicators are commonly used for this Errors of inclusion (EI) arise when people who are not purpose; known as errors of exclusion and inclusion. poor benefit from the subsidy. The error is defined as the The following diagram helps to illustrate the meaning percentage of subsidy beneficiaries who are not poor. In of these terms. The diagram divides the general population terms of the areas drawn in the diagram, this can be into two groups: poor (P) and non-poor (NP). A subset of expressed as: EI=BNP /B. Errors of inclusion are essentially a this overall population are beneficiaries of a subsidy program form of inefficiency, because they represent a leakage of (B). Since it is never possible to direct subsidies perfectly subsidy funds towards a subset of the population that doesnt towards the target population, some of these beneficiaries really need them. are poor (BP), while others are non-poor (BNP). Related to the error of inclusion, it is also interesting to Errors of exclusion (EE) arise when people who are calculate the leakage rate, which is the percentage of subsidy genuinely poor fail to receive the subsidy. The error is defined resources that are captured by the non-poor. The leakage as the percentage of the poor who do not receive any rate tends to follow a similar pattern to the errors of inclusion; subsidy. In terms of the areas drawn in the diagram, this however if non-poor beneficiaries consume relatively large can be expressed as: EE=1-(BP /P). Errors of exclusion could amounts of water compared to poor beneficiaries, then the be regarded as an even more serious problem than errors leakage rate may be even higher than the error of inclusion. .inally, it is interesting to note that errors of inclusion and exclusion tend to move in opposite directions. Thus, a subsidy with very high errors of inclusion will typically have relatively low errors of exclusion and vice versa. This happens because it is difficult to identify the poor, and to be sure of reaching most of them, very broad eligibility criteria are required; which in turn brings in a large number of non-poor too. The implication is that to be sure of reaching the majority of the poor, it is often inevitable that there is a considerable amount of subsidy resources wasted on the not so poor. .igure 4, which shows that the subsidy received increases .igure 4: Average monthly subsidy received by income decile as the income deciles increase, from the poorest 10% at the left of the graph to the richest 10% at the right.) The key factor underlying these distributional results is the targeting mechanism used to distribute the subsidy. In both Bangalore and Kathmandu, two different targeting mechanisms are at work. The targeting mechanisms can be evaluated in terms of leakage rates, and errors of inclusion and exclusion. A full definition of these technical terms can be found in Box 2. 6 The first targeting mechanism is the principle of self- selection, which suggests that poor households will be the predominant users of the more highly-subsidized public tap service, since richer households will be able W A T E R T A R I . . S A N D S U B S I D I E S I N S O U T H A S I A to afford private taps and hence will prefer to have that non-poor households (recall .igure 2), the IBT does not higher quality of service. Self-selection appears to work perform very well in targeting terms (Table 2). In relatively well, since only 14 to 39% of the resources particular, as much as 70 to 80% of subsidy resources assigned to public taps fail to reach poor households fail to reach the poor. Errors of inclusion are (Table 2). The errors of inclusion are correspondingly correspondingly high at 71%, given that the IBT provides low at 20 to 40%, reflecting the fact that public taps are subsidies to all consumers with private taps, the majority unattractive to those who can afford a better service. of whom are not poor. In both cities, the errors of The performance regarding errors of exclusion differs exclusion are just over 50% given that more than half substantially across the two cities. In Bangalore, only of the poor lack connections to the piped water network. 23% of the unconnected poor fail to benefit from the In other words, subsidies to public taps perform much network of public taps, whereas in Kathmandu as many better in distributional terms than subsidies to private as 61% of the unconnected poor lack access to public taps. The reason is that the willingness to accept a low taps (or dont use them because other non-utility sources, quality public tap service because it is free of charge is such as shallow tubewells, are more attractive). a more reliable indicator of poverty than the amount of water a family consumes. The absolute value of subsidies to public taps However the absolute value of subsidies to public taps is very small compared with subsidies to private is very small compared with subsidies to taps. This means that 90 to 95% of the total subsidies private taps, absorbing only 5 to 10% of allocated to the water utilities of Kathmandu and overall subsidy resources. Bangalore are spent on subsidizing private taps, and barely 5% on public taps (Table 2). Thus, the share of overall subsidies captured by the poor is still only 20 to The second targeting mechanism at work in these 30%, while the errors of inclusion are 60 to 65%. two cities is the Increasing Block Tariff (IBT). The IBT Looking at the overall picture, however, errors of directs the distribution of subsidies towards smaller exclusion are somewhat lower, at 12 to 33%, because consumers, on the assumption that these are more likely most of the poor benefit from one or other of the two to be poor. However, due to the relatively small services provided by the utilities, whether it be a public differences in consumption patterns between poor and or a private tap. Table 2: Performance comparison of subsidy schemes Leakage Rate or Errors of Inclusion or Errors of Exclusion or percentage of subsidy percentage of subsidy percentage of poor resources captured by beneficiaries that are who do not benefit the non-poor non-poor from the subsidy Kathmandu Bangalore Kathmandu Bangalore Kathmandu Bangalore 7 Public taps 39% 14% 38% 23% 61% 23% Private taps 78% 73% 71% 71% 53% 51% Overall 77% 69% 64% 60% 33% 12% W A T E R T A R I . . S A N D S U B S I D I E S I N S O U T H A S I A DO CURRENT WATER SUBSIDIES REACH THE POOR? Moreover, looking beyond distributional issues, .inally, it is interesting to see whether water subsidies public taps present various disadvantages to users. soften or accentuate existing inequalities in the Collecting water from public taps is onerous both in distribution of income. This can be done by calculating terms of the time spent (users often spend hours queuing) quasi-Gini coefficients, which measure the extent to and drudgery (water used at home has to be carried which the distribution of the subsidy is skewed towards there). Water quality is jeopardized as the water is so the rich (positive Gini) or the poor (negative Gini). (A often brought into contact with hands and implements fuller explanation of this concept is provided in Box 3.) during collection, transfer to storage vessels in the By comparing these indices with the income distribution house, and storage. .inally, the burden of water Gini it is possible to say whether subsidies are more or collection often predictably falls disproportionately onto less equitable than the economy as a whole. The results women, but also in many cases children, who then have suggest that subsidies to private taps, although skewed less time for school and recreation. towards the rich, are more equitably distributed than While the distributional performance of water income (.igure 5). To that extent at least, they reduce subsidies in Kathmandu and Bangalore is not all that the degree of inequality in society, though this in itself is not an argument for preserving them in their current form. Subsidies to public taps are heavily skewed Barely a quarter of the subsidies provided by towards the poor, and can thus be regarded as equitable State governments and distributed by water both in a relative and an absolute sense. utilities in the cities of Bangalore and Conclusions Kathmandu end up benefiting the poor. To summarize, barely a quarter of the subsidies provided by State governments and distributed by water utilities .igure 5: Comparison of Gini coefficients good in absolute terms, it is important to put the results into broader perspective. The problem of effectively targeting subsidies is neither unique to South Asia, nor unique to the water sector. A recent study of water subsidies in Chile and Colombia found errors of inclusion in the 60 to 80% range2, while a number of studies of electricity subsidies around the world have found that barely 10 to 35% of these subsidy resources reach poor households. Moreover, a study by the Indian National Institute for Public .inance and Policy (NIP.P) concluded that food subsidies allocated via the Public Distribution System (PDS) suffered from errors of inclusion between 34 and 52%, and errors of exclusion between 25 8 and 98%. 2Gomez-Lobo, Andres, Contreras, Dante, Subsidy Policy for the Utility Industries: A Comparison of the Chilean and Colombian Water Subsidy Schemes, University of Chile, August 2000. W A T E R T A R I . . S A N D S U B S I D I E S I N S O U T H A S I A Box 3: An introduction to Gini coefficients In addition to looking at errors of inclusion and exclusion, Gini coefficient is defined as: QGC=A1/T. When the Lorenz it is important to understand the overall pattern of subsidy curve bows up above the 450 line, the area between the incidence across the full spectrum of rich and poor. A simple Lorenz curve and the 450 line is deemed to be negative. way of doing this is to rank the total population from richest Hence, for the second Lorenz curve (LC2) the quasi-Gini to poorest, and then plot a Lorenz curve which shows the coefficient is defined as: QGC=A2/T. percentage of subsidy that is captured by the poorest X The quasi-Gini coefficient is bounded between 1 and percent of the population. +1, with an intermediate value of zero. A quasi-Gini of Equal distribution. If the distribution of the subsidy was zero essentially indicates that the Lorenz curve lies right on equal across the population, then the poorest 20% of the top of the 450 line. A quasi-Gini close to +1 means that the population would receive 20% of the subsidy, the poorest distribution of the subsidy is very pro-rich, so that the Lorenz 50% would receive 50% of the subsidy, and so on, so that curve is bowing out almost to the edges of the triangle and the Lorenz curve would essentially be equivalent to the 450 almost 100% of the subsidy is going to the richest few people line shown in the diagram. Such a distribution is neither in the society. On the other hand, a quasi-Gini close to 1 pro-poor nor pro-rich, since everyone essentially gets the means that the distribution of the subsidy is very pro-poor, same amount. so that the Lorenz curve is bowing out to make almost a Regressive (or pro-rich) distribution. The first Lorenz curve triangle above the 450line and almost 100% of the subsidy plotted in the diagram (LC1) represents a situation where the is going to the poorest few people in the society. poorest 20% of the population receives only 5% of the total Gini coefficients are also commonly used to measure subsidy, while the poorest 50% receives only 15% of the the distribution of income in a society, and in this case they total subsidy. As a result, the Lorenz curve bows down below can only take values between 0 and +1, since it is (by the 450 line, indicating that the distribution is regressive, definition) impossible for the poorest 20% of the population or pro-rich. to have more than 20% of the income. Progressive (or pro-poor) distribution. The second Lorenz curve plotted in the diagram (LC2) represents a situation where the poorest 20% of the population receives 60% of the total subsidy, while the poorest 50% receives 90% of the total subsidy. As a result, the Lorenz curve moves up above the 450 line, indicating that the distribution is progressive, or pro-poor. .or convenience, it is typical to summarize the shape of the Lorenz curve in a single indicator known as a quasi- Gini coefficient (QGC). The quasi-Gini coefficient is defined as the area underneath the 450line down as far as the Lorenz curve, divided by the whole area of the triangle under the 450 line. Thus, for the first Lorenz curve (LC1) the quasi- in the cities of Bangalore and Kathmandu end up poverty line, around 70 to 80% of these resources benefiting the poor. fail to reach the poor. There are two underlying Around 90 to 95% of these resources are used to reasons for this. .irst, barely half of the poor have keep tariffs faced by households with private taps private taps, hence most are excluded altogether from 9 low, with each of these households receiving an this type of subsidy. Second, the IBT structure used implicit subsidy of US$ 10 to 15 per month. However, in both cities tends to skew subsidies towards small- since 70% of those with private taps live above the volume consumers, on the assumption that they are W A T E R T A R I . . S A N D S U B S I D I E S I N S O U T H A S I A DO CURRENT WATER SUBSIDIES REACH THE POOR? Prepared by: Vivien .oster, Subhrendu Pattanayak, and Linda Stalker Prokopy. The paper is based on the following consultants reports: Prokopy, L. 2002 Distributional Incidence of Current and Potential Water Tariffs and Subsidies in Bangalore, India, University of North Carolina; and Pattanayak, S.K. and Yang, J.C. 2002 Distributional Incidence of Water Tariffs and Subsidies in Kathmandu, Nepal, Research Triangle Institute, North Carolina more likely to be poor. However, the evidence suggests that although poor customers consume less Series Editor: Clarissa Brocklehurst water on average than non-poor customers, the This series of papers was prepared as part of the Capacity Building difference is not all that large. and Learning initiative undertaken by the South Asia Energy and Infrastructure Unit of the World Bank in collaboration with the The remaining 5 to 10% of subsidy resources are Government of India, the World Bank Institute and the Water used to finance free public taps in poor and Sanitation Program, supported by a grant from the Public- neighborhoods. Due to the low volumes of water Private Infrastructure Advisory .acility (PPIA.), a multi-donor delivered by the public tap network, the implicit technical assistance facility aimed at helping developing countries improve the quality of their infrastructure through private sector subsidy received by each household is no more than involvement (for more information see http://www.ppiaf.org). The US$ 1 to 4 per month. Although modest in value, initiative is designed to support the implementation of sector these subsidies are comparatively well targeted reforms and public-private partnerships in the provision and towards the poor. Since 60 to 80% of public tap users financing of water supply and sanitation services in India. are poor, as much as 60 to 85% of subsidy resources April 2003 channeled through this route reach the poor. The Task Manager: Midori Makino reason that public taps seem to be more successful Design and Production Coordinator: Vandana Mehra than private taps in targeting subsidy resources to Thanks to: Lee Travers, Deepak Sanan and Kirsten Hommann the poor is because they offer a level of service that proves unattractive to anyone who could afford Designed by: Roots Advertising Services Pvt. Ltd. Printed by: PS Press Services Pvt. Ltd. anything better. In Bangalore, public taps manage to reach the vast majority of the unconnected poor, but this is not the case in Kathmandu, where most of the unconnected poor are still forced to rely on traditional PPIAF sources of water. Notwithstanding the relatively 1818 H Street, N.W. favorable distributional performance of public taps, Washington, D.C. 20433 Phone: +1(202) 458-5588 they carry a number of other disadvantages which Fax: +1(202) 522-7466 may outweigh these benefits. Email: info@ppiaf.org Although the distributional performance of water Website: www.ppiaf.org subsidies in the two cities is not good, they are nonetheless more equitably distributed than income and hence make some small contribution to reducing inequality. Notwithstanding this, the substantial Water and leakage of resources away from the intended Sanitation Program beneficiaries raise the question of whether a more 1818 H Street, N.W. sophisticated approach to targeting would prove any Washington, D.C. 20433 more effective. The next paper in this series will take- Phone: +1(202) 473-9785 10 up this challenge. Fax: +1(202) 522-3313, 522-3228 Email: info@wsp.org Website: www.wsp.org The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors and should not be attributed in any manner to the Public-Private Infrastructure Advisory .acility (PPIA.) or to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. Neither PPIA. nor the World Bank guarantees the accuracy of the data included in this publication or accepts responsibility for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this report do not imply on the part of PPIA. or the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. W A T E R T A R I . . S A N D S U B S I D I E S I N S O U T H A S I A