Water 279

 WATER

12.1 GETTING WATER FROM PUBLIC–PRIVATE PARTNERSHIPS

Ashima Goyal

‘Dirty water cannot be washed’

West African Proverb

What should be the contractual type of PPP in order to get Experience with reform had clearly shown market failure
the maximum value from such partnerships in the Indian to exist along with government failure, especially in the
context? Well-designed and chosen contracts improve provision of public services. PPP could be a way to utilize
governance. Insights from contract theory and international the best of both worlds, while overcoming the specific
experience with PPP in water supply can suggest how services weaknesses of each. But for this to be feasible contracts and
from public utilities, specifically from Mumbai’s water supply, regulation have to be carefully designed. Theory is helpful
can be improved, with cost saving through the reduction in analysing these weaknesses and suggesting design elements.
of waste. They also need to be tailored to the specific context. Menard
As the literature has developed, PPP itself has come to and Saussier (2002) empirically establish that PPP contracts
be defined more broadly, to include not only a contract in France have endogenously adapted to suit local conditions.
between a government and a formal commercial business (Box 12.1.1).
firm to deliver a public good or service, but also wider
interactions between the government and informal small EXPERIENCE WITH WATER CONTRACTS IN OTHER
scale providers (SSPs) that can improve the efficiency of
COUNTRIES
delivery of public services. Civic society also has a major
role to play in making such partnerships work well. Experience in different parts of the world has tempered the
Transaction cost perspective tells us that contracts are enthusiasm of the 1990s for private provision of water,
essentially incomplete. Then many social institutions have which itself had been started as a response to severe municipal
to contribute to ensure effective coordination. failure. For a long time the World Bank was the dominant
international voice pushing for private participation,
An earlier version of this paper was presented at the IIR workshop, especially concessions and divestiture (World Bank 2000).
15 June 2003. The participants were knowledgeable and generous, But new, more nuanced, voices have emerged. A selective
and I thank them for very useful comments and discussion. It was survey of some of the experiences they recount is useful.
also presented at a symposium in honour of Jyoti Parikh at IGIDR, It turns out that the contractual forms gave too much
7 July 2003. Again I am grateful for very encouraging responses,
expost bargaining power to private parties. Major concessions
and a valuable discussion with A Vaidyanathan. I thank Sebastian
Morris for suggestions and encouragement, Anirban Acharya for given with fanfare in Manila in 1997 have exposed many
helpful research assistance including site visits, and T.S. Ananthi for potential problems. The companies made unrealistic bids
drawing the diagram. just to win the tender and then passed on large price increases
280 India Infrastructure Report 2004

Box 12.1.1
Contract Theory and Water Supply

Contract theory recognizes that there is much more to a market transaction than just price and quantitya. A contract is a bilateral
coordination arrangement (Brousseau and Glachant 2002). There are a number of difficulties associated with coordination. Laws,
regulations, incentives, and other mechanisms are required to achieve it among agents who may have asymmetric and incomplete
information.
Contracts tend to be incomplete, but even more so in developing countries, and difficult to enforce using only the courts. A
variety of institutional features that enforce commitments are needed. At the same time, unanticipated shocks are more frequent
in such countries. Therefore some expost flexibility in contracts is also essential. Renegotiations, sanctions, private conflict resolution
mechanisms, organizational forms, and norms, all have a role in providing the required mix of commitment and flexibility. Since
private conflict resolution mechanisms are imperfect, coordinating rules and sanctions are also required. These mechanisms have
been highlighted in the literature on transaction costs theory.
Transaction cost theory views contracts as inherently incomplete because it is not possible to costlessly include unforeseen or
unspecified contingencies in contracts. Therefore, the expost bargaining position is important. Residual rights of control, in
uncontracted for circumstances, determine agents’ incentives. For example, either residual control rights must be given to the party
who makes a relationship specific investment or that party will underinvest. When such allocation of control is not possible
internalizing the transaction can improve welfare. This is one of the justifications for the existence of a firm.
If complete contracts could be written public or private provision of public services would be equivalent. Public goods and
services are often experience rather than search goods, and their quality is revealed only by consuming. Contracts are incomplete
for such goods because all dimensions of quality of service cannot be fully specified. Under complete contracts agents are induced
to take appropriate discretionary actions and reveal their information. Control rights, however, cannot fully substitute for this
mechanism. If control rights are conferred on a single group it may result in self-serving actions. Therefore, control rights have
to be carefully divided and civic society and other democratic pressures used. Supervisory or coercive mechanisms may be required
to ensure that parties respect their commitments, but these should not restrict the flexibility required to ensure expost optimal
coordination. A wide variety of institutions that impinge on the transaction and have the potential to improve its efficiency should
be considered.
Different transaction structures solve the coordination problem to different degrees. A contractual structure can evolve endogenously
as a way of resolving the problem. Differing contract forms will be more appropriate in differing circumstances and will vary in
their efficiency properties. In considering the conscious design of contracts these aspects have to be kept in mind.
A choice of PPP in a particular public service delivery normally offers a menu of contract choices, ranging from outright
privatization or divestiture, to long-term concessions or management contracts where asset ownership remains in the public sector,
to outsourcing of specific jobs. Control, asset ownership and risk shared by the private sector falls in moving from the first to the
last. The general principle is that risk should stay with the agency most able to bear and best able to alleviate it. Service contracts
normally vary over a period of 6 months to 2 years. Private sector expertise is used or it may be developed and competition increased.
The responsibility for coordination and for investment rests with the public sector. Management contracts transfer responsibility
for operation and maintenence (O&M) also to the private sector. These run typically for 3 to 5 years. They may have a fixed fee
and include a performance target. In a lease contract the private party buys rights to the income stream so it assumes the commercial
risk. Incentives for reducing costs and quality standards can be written in. Concessions last for 25 to 30 years. Responsibility for
O&M as well as investment is transferred while asset ownership remains with the government. The contracts set out performance
targets, standards, investment amounts, mechanisms for adjusting tariffs, and arrangements for arbitrating disputes. Assets are to
be returned in good condition and consumers are to be protected from monopoly pricing. Under divestiture, asset ownership also
passes into private hands and the government is only left with the task of regulation (Suresh 1998). Contract theory suggests a
number of factors to consider in deciding the position taken on this range.
1. The more specific assets are to the particular service, the more ownership and control should stay with the public sector.
The private sector would have an incentive to underinvest, since it is liable to ‘holdup’ once it has invested in assets, which have
no use elsewhereb.This explains why there is so little private ownership of water sources and distribution networks since the assets
are highly specific. Well-written contracts may be able to ameliorate the problem to some extent but will not resolve it entirely.
2. A financially constrained public sector entity has a greater incentive to outsource, and a larger population or consumer density
will make outsourcing more feasiblec.The private sector may be able to provide funds that highly indebted governments lack. But
the former will be motivated to provide the funds only if it is possible to recover the investments through user charges. Therefore,

aThe discussion in this section largely draws upon Tirole (1994), Hart et al. (1997), Brousseau and Glachant (2002), and papers collected
in the latter.
b This is a basic premise of TCT, developed by Olivier Williamson (Brousseau and Glachant 2002).
c In French water utilities almost the entire range of PPP contracts is to be found. Menard and Saussier (2002) empirically test for the factors
that determine endogenous adoption of a particular contract type, and find size of population to be an important variable.
 Water 281

the key reform is one of pricing. Reasonable user charges can make either a public or a private project viable, and a larger customer
base will increase profitability.
3. Public sector provision should dominate, the greater the uncertainty affecting the quality and the more difficult it is to enforce
quality standards. If it is not possible to contract for quality, integration is to be expected, and provision of the good or service
would stay in the public sector. The literature generally expects quality to be higher in the public sector because incentives to reduce
costs and quality are too high for the private sector. With residual control rights or under a price cap from a regulator the private
sector would retain any profits made from a reduction in cost. If quality falls with a reduction in cost such high powered incentives
would reduce quality. Incentives, therefore, have to be low-powered, such as they are in a bureaucracy, or with cost-plus pricing
rules, in order to maintain quality (Tirole 1994).
In the public sector the bureaucrat-manager has poor incentives or control rights since he is transferable, but there may be a
mission, reputation, or career concerns, or political pressure to ensure quality service. Hart et al. (1997) point out that while private
provision will always have lower costs than public, quality in private provision may be lower or higher, if quality decreases with
costs but increases with innovation. The latter is normally higher under private provision and is likely to be more important in
some kinds of services. Private provision would be superior in cases where quality can be contracted for, or is insensitive to cost
reduction, or opportunities for cost reduction are small. Private provision will also dominate if incentives for quality improvement
are poor for public managers.
The poor quality associated with public sector provision of services in developing countries is a paradox in the context of these
arguments. The answer may lie in the use of poor quality as a rationing device in order to target the poor and serve them within
resource constraints. As we argue below, this is particularly relevant in the case of water where demand is highly elastic at low prices
so that quantity rationing becomes necessary at low user charges.
4. If there are multiple, non-measurable aspects of quality, public provision will do better. Incentives have to be low for bureaucrats
partly because they have to satisfy multiple objectives, some of which are not measurable. High-powered incentives or residual control
rights for measurable performance targets lead to a neglect of non-measurable aspects. Among non-incentivized tasks a public good
provider has to satisfy is inclusion of the poor, or geographically dispersed difficult to serve categories, and other universal service
obligations. In the water sector, it is ironic that the instruments used to target the poor, such as public standposts, or rationing
of piped water for limited time-periods, require a steep fall in quality.
5. Public goods provision is compatible with some competition. Many public goods and services are local monopolies. Even
so, a regulator can introduce elements of yardstick competition, where relative performance of one local body is compared to that
of another, which is similarly placed, and thus forced to reduce costs. Another example is allowing local water boards to compete
with each other through bids, for the award of management or concession contracts, in the supply of public services. Unbundling
makes competitive supply possible for parts of the public good or service. If the latter is a search good, where consumers can assess
quality for themselves, and suppliers are perfectly competitive at each level of quality, private supply would face socially-optimal
incentives (Hart et al. 1999). Citizen or user groups are another useful source of information for the regulator, to discipline public
monopolies. Such countervailing pressures make it feasible for incentives to be high powered even in the provision of public goods.
So far we assumed the public servant really serves the public. But there are a number of features contracts must cover to control
self-interested bureaucrats or politicians, or to prevent regulatory capture.
6. Transparent rules that minimize discretion may be necessary even at the cost of some flexibility. Parts of the public sector
may have more information compared to others. If they have discretion they could use the informational advantage to benefit interest
groups they favour. In general interest groups that benefit from withholding of information stand to gain. They have an incentive
to persuade the official not to release information that harms them. If the officials have to follow a rulebook, then interest groups
have less of a stake in the decision and therefore less incentive for regulatory capture. This is one reason officials are made to follow
rules. Thus low-powered cost plus rules are less subject to regulatory capture, since they leave low rents with the firm, and the
regulator has little discretion. If high-powered incentive schemes such as price caps are to be retained then alternative pressure groups
such as user groups are required to guard against regulatory capture by the supplier. Control rights need to be divided among
countervailing groups.
In government auctions there is more discretion in judging quality, and, therefore, more scope for favouritism. Therefore, tangible
variables such as price are given precedence (Tirole 1994). Klein (1996) points out that the large discretion French municipalities
have in the award of contracts for water supply are responsible for allegations of political funding received from firms.
7. Commitment for public service contracts should be short-term and they should be subject to rebidding. In many situations
social welfare is enhanced if a benevolent government can make an intertemporal commitment to a long-term complete contract—
this would remove fears of expropriation and encourage utilities to achieve optimal efficiency and investment, as is argued in point
1. But short-term commitments provide a check against decisions taken to favour interest groups. Thus, rebidding water contracts
may not only keep a check on costs but also allow removal of a corrupt or inefficient firm. Elections are another disciplinary device
and a change in the elected government makes it more feasible to force an exit of its corrupt or inefficient cronies (Tirole 1994).
Such considerations explain why governments cannot commit not to ‘hold-up’ sector specific investment, and, therefore, why private
investment would not be optimal.
282 India Infrastructure Report 2004

8. Checks and balances are required to prevent corrupt politicians making money from privatization or using public utilities
for patronage. If the first tendency dominates sale of government assets would not raise social welfare and if the second more
independent private sector management or sale of public utilities would be beneficial. The latter would prevent the non-public-
spirited politician from obliging special interest groups such as trade union vote banks (Hart et al. 1997). Control rights for any
particular decision must not devolve on a single group. In a democracy these are divided among the judiciary and executive as
well as the legislature; and the media, the user, and other independent groups must play a major role. Often advocacy by each
group reveals valuable information for a decision-maker (Tirole 1994).

using loopholes and weaknesses in regulation. Bidding low population left uncovered by large water utilities. In Paraguay
and hoping to renegotiate is regarded as an acceptable entry they have proved themselves to be more efficient, competitive,
tactic in the water business. Officials complain that once and cheaper in serving niches. These aguateros lay pipes
having won the concession, the companies make their own from ground/water wells they dig. They pay taxes and the
rules. There was some increase in coverage of the poor and government regulates them and tests their water quality
of slum areas, but the companies exaggerated it. For example, periodically. Thus, open entry and competition becomes a
they would provide a single water main to a group who possible model of water supply (Solo 1999).
would then have to set up a committee and hire a contractor. In the US, in 1850, the majority of water systems were
The company would charge the committee but count each privately owned, but this was reversed in the early 19th
household as a connection. Since the charge was higher for century because of a concern for quality and public hygiene.
higher water consumption, these poor households ended up Today 81 per cent of the population is served by public
paying more. But sometimes the companies were unable to utilities. Thus quality is a concern in private utilities. In
collect user charges. The companies had promised to reduce 2003 Atlanta Mayor, Shirley Franklin, cancelled a concession
leakage and theft, but were unable to do so. Non-revenue contract, which had been running for 4 years, on grounds
water increased slightly and accounted for between half and of non-performance (Hobbs 2003).
two-thirds of the water supplied. Since the companies installed In developing countries administrations are less resolute
meters near the mains and the leakage took place in the and water companies were often able to renegotiate contracts
distribution lines they were still able to bill the households on terms favourable to themselves, because of expost
for the water losses (Landingin 2003; Gutierrez 2001). monopoly power and the high cost of terminating a
Cost plus pricing, non-transparent accounting and padding concession once given. Moreover, transnationals need not
of bills, results in higher user charges for consumers even in bring in money themselves, but borrow from local resources,
developed countries like France. Companies tend to distribute and then recover the cost through user costs (Gutierrez
their costs across all municipalities they work for in a non- 2001). Their only contribution is management experience,
transparent way. In the Grenoble corruption case it was but even there they do not have experience of local conditions.
shown that this allowed them to give kickbacks to political An argument is sometimes made for their starting with
parties. The concession was investigated after a series of management or service contracts so that they can gain
letters from Grenoble citizens complaining about the sudden knowledge of the functioning of the utilities and local
jump in water prices. The contract even had the regressive conditions before moving on to concessions (World Bank
feature of discouraging conservation, since the company 2000). But building governance structures that allow public
could increase the price if consumption fell below a contracted utilities to run on more commercial lines could be an
amount (Godoy 2003). Often there are systematic incentives alternative solution. This is the strategy of strengthening the
to exploit new water sources rather than conserve water by capacity of the public sector.
reducing both leakage and consumption. A thorough review from Gleick et al. (2003) recognizes
Despite the wave of privatization in the 1990s, water that some privatization is required to improve efficiency,
utilities remain publicly owned and operated in most but warns that better regulatory controls are required to
countries of the world except the UK and France (Gutierrez ensure safe, affordable, high quality environmentally
2001). One of the key factors facilitating efficient public sustainable water supplies. It is necessary to have transparent
delivery is a reasonable user charge. Bogota, Columbia, is contracts with clear dispute resolution mechanisms and all
an example of a public water utility that transformed itself stakeholders involved in the negotiation. At the same time
in the 1990s after it was allowed to revise user charges it is necessary to strengthen the ability of governments to
(Ronderos 2003). meet water needs. An NGO, WaterAid, in their private
Competition can come not only from rebidding but also sector participation framework document emphasize benefits
from encouraging niche domestic informal firms or SSPs. for the poor, strengthening of the local community,
These are often the ones that reach large sections of the government capacity, and the domestic private sector to
 Water 283

deliver according to standards (despite support, usually from customer care would be removed (points 3 and 4 of Box
only a small and unstable financial sector) and give Value 12.1.1) and competition improved (point 5). Examples as
for Money (VFM) (Gutierrez 2001). in Box 12.1.3 show the beginnings of such low level service
Contextual analysis is also essential to decide on the contracts, suggesting that a wider adoption of such contracts
specific form of outsourcing, and to yield insights on the may be feasible.
ideal features of the contracts to be entered into. In a The key to these changes is price reform, and the
developing country, the main objectives in water supply establishment of independent regulators, who would set
provision must be, first, to maintain the quality and quantity tariffs on the basis of economic principles, free from political
essential to life. The poor must have access to adequate interference. But efficiency prices, with a lower dependence
quantities of affordable water1. Second, to conserve water on non-transparent fixed charges, and higher volume charges
or prevent its waste, given the pressures on world water require improved metering and billing. The latter are aspects
resources2. The lessons from the problems with both of improved quality of service, which will increase willingness
government and private sector provision of water suggest a to pay. The waste and slack in the system gives an opportunity
third objective—the importance of maintaining some for designing service contracts that will finance themselves
competition in water supply provision to ensure consumers by removing this waste. Institutionalized user groups have
a better deal. Finally, encouraging entrepreneurship is a a major role in monitoring contractors, ensuring quality,
fourth objective, to strengthen the local community. encouraging conservation, providing information to
Contracts should be designed with these objectives in mind. regulators, and preventing the misuse of political power in
These features of an ideal water service, together with the patronage or corruption (points 6, 7, and 8 of Box 12.1.1).
current structure and larger context, will be used to deduce Such user groups have leverage in aligning politicians’
appropriate contractual forms for PPP. incentives since they represent potential votes. Tax
concessions, reputation and recognition aspects can provide
incentives for these groups. We examine these aspects in
IMPROVING WATER SUPPLY greater detail below, and show why these interacting changes
Theory, international experience, and the current domestic are feasible, and why they would achieve the objectives of
context suggest that the Indian water supply system can a water supply system in the Indian context.
benefit from PPP in the form of service and management The major reason independent regulators are essential in
contracts, although assets should continue to be in public the Indian context is to free the pricing of water from
hands3. In Indian cities with large populations potential political influences. The Sukthankar (2000) Committee has
revenues from water supply are huge with price reform. (See recommended a state level independent regulator for
point 2 of Box 12.1.1.) Water assets will be developed Maharashtra. Since water is a state subject, the regulator
better, however, and their quality maintained, in public would have to be at the state level, but some general principles
hands because of disincentives for optimal private investment may be laid out to apply across states through a central body
(the asset specificity argument made in Box 12.1.1), such as the Finance Commission.
incentives to overuse scarce water resources, and the ability
of large foreign firms to renege on contractual commitments Regulation
in developing countries. Divestiture would give too much The general principles for ensuring efficient, independent
of residual control rights to one party, namely the private regulation, and preventing regulatory capture are well known,
firm. But strategic PPP in service contracts would strengthen and there is large international experience in the functioning
the public provision of water. Weaknesses in quality of of regulatory commissions, which can be drawn upon (Klein
1996). In the Indian context experience has to be built up,
1 but if the regulators have an understanding of basic economic
This is a major part of the millennium development goals
recently established by the United Nations. principles and the ability to withstand political pressures,
2 The World Water Development Report, subtitled ‘Water for technical expertise can be hired. The regulators must not
People, Water for Life’, from UNESCO, released ahead of the Third be from water utilities, retired government insiders, or be
World Water Forum in Japan in 2003, calculates that by 2050, elected. There must be staggered retirement, and the term
between 2 billion people in 48 countries and up to 7 billion people must not coincide with that of the elected government.
in 60 countries will be faced with scarcity of fresh water (source-
Elected officials tend to favour consumers too much, and
press reports).
3 A. Vaidyanathan made the observation that if rain harvesting officers earlier associated with water utilities would favour
techniques become widespread some water sources would effectively the latter.
be in private hands. We need to distinguish such private measures The pricing rule selected must provide good incentives
to augment water supply from development of large water sources. for conservation in water use, efficiency in water production,
284 India Infrastructure Report 2004

Box 12.1.2
Water Supply in Mumbai

We start with a brief description of the current state. Then some gaps are highlighted, where PPP can contribute. It turns out
that PPP may even be self-financing if it is successful in removing waste. Then it would truly be a partnership, with gains that
are potentially very high.
Since Mumbai’s population is around 15 million and is expected to grow substantially, its water requirements are huge, and
the water supply system is among the 8th largest in the world. The Brihanmumbai Municipal Corporation (BMC), which runs
the system, is justifiably proud of its assets. In 2002, the total supply was 3000 mld (million litres per day) compared to estimated
demand of 3975 MLD. Taking leakage as 20 per cent or 600 mld, the average consumption was 167 lpcd (litres per capita per
day), with total domestic consumption at 2160 mld, and non-domestic at 240 mld. Although the supply ratio to the two was 90:10,
their respective contribution to revenue was 50:50. The cost of production was Rs 5 per kilolitrea.
Organization Structure: Although in Mumbai the BMC directly runs the water supply system, in many municipalities water
boards run the system on behalf of the urban local bodies (ULBs). The water boards are state level organizations with more power
than municipalities so that the latter are unable to monitor their services. Moreover, their cost-plus administration charges of 18
to 22 per cent (Suresh 1998) give them no incentives to reduce costs.
A standing committee of elected representatives approves the setting and revision of tariffs in municipal corporations. For
municipal councils and ULBs state governments dictate tariffs. Under present governance structures politicians cannot credibly assure
coverage or quality but they can effect prices and tend to use that as a populist tool (Nanavathy 2001).
Pipe Network, Maintenance, and Leakage: The daily average water supply to the city is 2 hours, and to the suburbs 3.5 hours.
Trunk, feeder, and distributory water mains form the water supply network. The trunk and feeder mains carry water continuously.
They are of large diameters and made of mild steel which requires protective coatings. Among other materials used ductile iron,
and galvanized iron require internal mortar lining and external painting or plastic coating. Cast iron distributory mains are of smaller
diameter and carry water only at specific times. Because of intermittent water supply they are subject to alternate wet and dry
conditions which induces corrosion. This is a major cause of contamination and wastage of water. Corrosion is also aggravated from
aggressive soil conditions, vicinity of the sea, and currents induced by the railway’s direct current traction system. The pipes develop
internal scaling, pitting, tubercular and micro-bacterial growths. Periodic cleaning and scraping is required after internal inspection
through video equipment. GI water connections to customer premises corrode over time. They are the most neglected and a major
source of contamination and leakage. Periodic renewal or use of non-corrosive materials and leakproof joints is required. More recently
poly-ethylene pipelines are laid in particularly hostile conditions (MEA and BMLPA 2003). Domestic companies have started
manufacturing these, and accessories are now available.
Contamination arises from intermittent supply, low water pressure, leaking pipes, and inadequate wastewater collection systems.
In slums because of inadequate sewage disposal the soil around the water pipes is polluted. Moreover, since some pipes run close
to sewage lines there is great danger of contamination. The danger is compounded by slum dwellers who break the pipes, and then
cover the hole with cloth or other rough plugs, which are removed to draw water whenever neededb. During periods when water
supply is stopped the contaminated soil then gets sucked into water pipes (MEA and BMLPA 2003).
The Mumbai Water Engineers Association (MEA) reports that while water leakage in developed countries is between 8–12 per
cent, in India it is 30–60 per cent, and in Mumbai 20–25 per cent. Bringing this below 15 per cent is not cost effective. A National
Environment Engineering Research Institute (NEERI) study finds that 17 to 44 per cent of total flow in Indian distribution system
is lost through leakages, and about 82 per cent of this takes place in the housing service connection, and the rest is due to leakage
in pipelines.
The components prone to failure are generally pumping stations, pipelines, and appurtenances. Preventive maintenance is
required. But there is a poor stock of maintenance materials in the municipal store, and low availability of funds. The sanctioning
procedure is cumbersome. Repairs are undertaken during crises only. Maintenance staff report daily for attendance at the ward office,
then require 2–3 hours to reach the actual work site.The cross-section of pipes for household distribution varies from 0.5 to 2
inches. Metering is efficient at 1.5 to 2 inches pipe width in large housing societies, where water supply is 24 hours.
There have been major investments and improvements in the 1990s, but mainly in the water sources and plants. BMC has
drawn up a Rs 100 crore plan to replace old corroded pipes in a phased manner. Leak detection is possible through sounding and
a The source of these statistics is pamphlets from the Brihanmumbai Municipal Corporation (BMC). We are grateful to deputy municipal
commissioner R.K. Bhatia for a personal interview and for making these documents available. We also visited the Bhandup plant.
b da Cunha (2003a) reports one horrifying but not unusual case. The water supply line to Lokhande Marg in Mumbai passes through several
slums whose residents have punctured the line at 135 points. Suction pressure causes contamination to be sucked in and carried to the tanks of
the housing societies. The Deputy Municipal Commissioner who located the holes got them plugged but they were unplugged in 3 days. He ordered
installation of a new pipe bypassing the slums, but the order was not carried out. The Welfare Association and the housing societies wrote 51
letters of complaint, which were not answered. Two young people died of infection. The area pays Rs 3 lakh in municipal water taxes; for the
slums water is free. This illustrates the problems of maintenance, absence of prompt response to consumers, and intermittent supply. Privatization
and competition could address all of these issues.
 Water 285

electronic methods, or bypass meters can be used at the two ends of a water main and variation in flows recorded. A leak detection
cell has been set up since 1973, and there are 615 leak detection zones. Quality monitoring is undertaken with 6500 water samples
collected dailyc. Consumer satisfaction is reported at 73 per cent. But customer orientation is still poor and commercial principles
are not applied fully. Supply is released for various points in the city with almost no knowledge of demand, just rough estimates.
Tariff Structure and Use of Meters: For unmetered water connections water tax and sewerage tax is based on the rateable value
of the property (65 per cent for residential and 130 per cent for non-residential). Metered connections are charged tariffs on a
per kilolitre basis: stand-posts in slums Rs 2.25 per kilolitre, buildings and chawls Rs 3.5 per kilolitre (it is planned to raise this
to Rs 5.8), commercial and industrial establishments Rs 10.5-35 per kilolitre. A sewerage charge of 60 per cent of the rate is added.
In addition all premises are charged water and sewerage benefit tax based on the rateable value of the property (12.5 per cent for
residential and 25 per cent for non-residential). Thus, users pay charges or tax plus benefit tax. The revenue is sufficient to fund
40 per cent capital works in addition to operation and maintenance costs, 40 per cent is spent on the latter. Actual recovery works
out to 82 per cent. There are large outstandings from other government agencies, which often do not pay.
Till May 1997 water charges were only 60 paise per kilolitres (consumption of a family of 5). After that they were revised 3
times. They are expected to increase in future to cover higher capital costs of new assets. Only from 1997 were other households
required to pay higher charges compared to those paid by slums. Mumbai is doing well compared to Delhi where the lowest slab
for domestic connection is only 35 paise per kilolitre and the Delhi Jal Board loses Rs 260 crore annually on the revenue account
alone (Delhi Times, 21 April 2003).
Since April 2001 a decentralized computerized billing system has been introduced, with appropriate software. There are 2,46,000
metered connections, and 75,000 un-metered connections, of which municipal meters are 48,550, and private meters 1,97,450.
But only 20 per cent of domestic meters and 60 per cent of commercial meters work. Water meters are defective, and get clogged.
Problems arise with metering since water supply is intermittent, booster pumps are used, good quality non-magnetic water meters
are not available, and it is difficult to find locations for housing water meters in the old city and slums. Illegal tapping creates
more problems in metering for less-developed areas. Moreover, there is an acute staff problem. It is not possible to send meter
readers, house to house, and the reliability of the task performed is questionable (MEA 2003). Absence of full cost recovery, therefore,
is due not so much to absence of willingness to pay as to absence of infrastructure and personnel to collect dues (Suresh 1998).
Since it is difficult to bill the consumer correctly, disputes arise. A gap is assigned based on average monthly consumption for last
month/year. BMC argues that since domestic demand is almost constant it can be determined by size of the connection, number
of hours of supply, and pressure difference at key points on daily basis—or discharge capacity. For high consumption consumers,
meters are used, as demand is seasonal depending on business conditions.
c A study by Society for Clean Environment (SOCLEEN) and funded by Mumbai Metropolitan Development Authority (MMRDA) reported
in the Times of India, 10 July 2003, Mumbai edition, however, found levels of bacterial contamination hovering at about 20 per cent in many
wards. In some areas, especially those surrounded by slums, the level of fecal coliform (fc) touched 1600 fc per 100 ml of water. The safe norm
(Bureau of Indian Standards) is 10 fc per 100ml. The contamination was higher than that found by BMC perhaps because the tests were undertaken
at the household level. They were also found to be temporary suggesting that the BMC does respond to complaints.

investment, and financing, and efficient allocation of risks are necessary for the regulator to obtain information to be
among consumers, utilities, and taxpayers. A price cap that able to impose effective yardstick competition, and to limit
subtracts productivity improvements, and rises with cost regulatory discretion or regulatory capture under price caps
shocks that are out of the utilities’ control works best. But (point 6 of Box 12.1.1). Retired bureaucrats have captured
public reviews of proposals, and feedback from user groups the first few regulatory commissions set up in India showing

Box 12.1.3
Examples of Partial Privatizations

The Chennai Water Board has privatized water delivery to the poor through tankers. Tanker service is contracted to the private
sector with payment per trip, and annual contracts. The contractor, therefore, has incentives for regular trips and for the provision
of good servicea.
In Mumbai one private maintenance contract has been given in K-East ward. Operation of two pumping stations, one in Raoli
and the other in Sion, which were privatized, is successful. City and Industrial Development Corporation of Maharashtra (CIDCO),
Navi Mumbai, has privatized water pumps, meter reading and billing, maintenance of parks and gardens (Suresh 1998).
In Delhi complaints cells have been privatized in 2003 after complaints about rudeness of the water supply staff. Poor attendance
to water supply complaints is a sore point.
a Subir Gokarn made the point that since tankers are non-specific assets, there tends to be overuse of tankers in private water supply although
this mode of supply is more expensive compared to piped water. The explosion in credit to small road and water transport operators from commercial
banks validates this point. The figure rose from 2979 crores in 1994–5 to 4973 crores in 2000–1 (RBI, 2003, Table 54).
286 India Infrastructure Report 2004

the necessity of countervailing pressure by user and voter Tariffs

groups. Comparisons across similar local utilities can give
a regulator the necessary information to separate idiosyncratic Is a reform in pricing politically feasible? Box 12.1.4 suggests
risk from aggregate risk. The utility is then held accountable that even the poor would benefit from such reform. An
only for the former. For example, the monopoly can be independent regulator can free politicians from the status
allowed to pass on aggregate cost shocks, which are not in quo of low user charges they are caught in, partly because
its control. Thus, utilities should not have to bear the risk each party is afraid the other will make political capital from
of exogenous shocks. such subsidies. Political pressures have kept user charges low,
Presently local bodies revise water tariffs steeply once in with indirect and hidden costs, which are much larger as
5 years, which meets with consumer resistance. Suresh (1998) a result. These costs include the inefficiencies in water supply
suggests a 10 per cent increase every year. But this is cost and use, coping costs to compensate for poor quality and
of production pricing which does not pass on productivity coverage, and rack renting for industry, which just passes
gains to consumers or motivate utilities to lower costs. A on costs to consumers. Coping costs include time spent
price review could occur every year or two initially, and queueing for and purifying water, use of domestic tanks,
then, as learning proceeds could even be every 5 years, so pumps, use of private vendors and bottled water, which cost
as to give firms a longer period during which they can retain much more than piped water. There are also social costs such
profits from cost reductions. The regulator can also monitor as water hoarded due to irregular supply and then wasted.
compliance with other service standards, such as quality of User groups can make such indirect costs clear to voters, so
service and of water supplied. that the latter demand quality from politicians.

Box 12.1.4
Two-Part Tariffs and Benefits for the Poor

Water is a peculiar good in that it is essential to life. Therefore the utility it contributes at low levels of consumption is very high.
But it has many non-essential uses, and at high levels of consumption the contribution to utility of the marginal unit may be very
low. Demand is inelastic at low levels of consumption but very elastic at high levelsa. This has clear implications for the pricing
of water. Figure B12.1.4.1 develops a simple example with one rich (R) and one poor (P) consumer. The rich consumer has income
level YR and the poor YP. Y can be measured in terms of time available, with the assumption that time can be translated into incomeb.
Indifference curves are drawn for each type, showing the combination of water and consumption of other goods O, which would
keep the consumer at the same level of utility. The indifference UO is the reservation level of utility. It originates on the Y-axis
when other goods consumption O equals income. Initially the indifference curve is steep, the consumer is willing to give up large
amounts of other goods when water consumption is low. The curve also approaches the X-axis asymptotically since water
consumption can be very large at a low price in terms of other goods foregone.
Water is priced through a two-part tariff consisting of a fixed payment (F), which entails some free units of consumption, and
a price per unit of additional consumption or volume charge. Since normally the fixed payment is less than the maximum amount
the consumer would be willing to pay for his consumption of water, he is able to reach a level of utility higher than his reservation
level. Now suppose in period 1 there is no charge for the poor consumer, but he is subject to rationing of water. The implicit
cost of queueing, etc. he is subject to, FP1, drives him to his reservation utility UPO. This FP1 is a net welfare loss since it does
not go to the service provider but is a real coping cost for the poor consumer.
The rich consumer is charged a fixed cost FR1, which covers consumption of some water units along the horizontal section,
and pays a low price p1 per unit of additional water he consumes. The budget line gives the rate at which he can convert other
goods consumption of Y – F into extra units of water. The fixed cost is measured on his reservation utility curve and the price
line starts from the point on UOR giving free consumption at the fixed cost charged. His water consumption is at QR1 where his
budget line is tangent to his indifference curve. A fixed cost that does not extract all surplus and his ability to adjust his commodity
consumption at the margin take him above his reservation utility UOR. Since he is on the elastic portion of his demand curve,
he consumes a large amount of water but derives a low utility from the marginal unit.
a Samuelson in his famous textbook had used this paradox to illustrate the difference between total and marginal utility. He pointed out that
diamonds, which are not essential to life, have a very high price, but water whose contribution to total utility is much higher has a very low price.
The reason is that water is supplied to the point where its marginal utility is very low. Like air it is almost a free good. But in modern communities
delivering potable water to every doorstep is quite expensive, and costs have to be recovered. See Morris (2001) for an earlier treatment which
does not, however, analyse two part tariffs.
b T.C. Anant made the point that the poor may not be able to convert time into income. Indeed this belief is the reason that queueing is
used as a means of targeting the consumption of the poor. But the poor cannot afford to be unemployed and often have multi-activity going on
so that there is an income loss in queueing. Since water is essential to life high time costs of collecting it can have very large opportunity costs.
I know of one poor family who, since it took a large amount of time to collect their daily requirement of water, and the mother had a job, withdrew
their daughter from school to perform the task.
 Water 287

YR

Other goods
FR2
U 2R

Slope = –p2
1
FR

YP
FP2

0 Slope = –p1
Slope = –p2
1
FP

U 1R

U1P U 0R
U 0P
QP1 QP2 QR2 QR1 Water

Fig. B12.1.4.1 Price Reform and Welfare

Fig. 12.1.1 Price Reform and Welfare

Now consider a price reform in period 2 where both consumers are charged a higher unit volume price p2, but lower fixed
costs than they were paying in period 1. Thus the rich consumer pays FR2 > FP2. The consumption of the poor consumer goes
up QP2, but he remains on the inelastic part of his demand curve, and he reaches a higher level of utility, U1P. Although he is
paying a small fixed cost, lower than his earlier coping costs, his choice of additional units of consumption is voluntary, and takes
him to a higher indifference curve. A French student of BMC’s water supply reported that in the current rationing regime
consumption is taken to be 240 lpcd for the rich consumer and 45 lpcd for slumsc. The changes proposed would improve water
equity and the welfare of the poor.
The consumption of the rich consumer falls much more. But since he was on the elastic part of his demand curve, the total
utility loss from the marginal units of consumption is not very high. A high fixed and low volume charge turns all rich consumers
into high volume consumers. If the fixed cost he was paying earlier was high, he may continue to be on the same utility level
U1R. If the fixed cost he was paying earlier was lower at YRA (< FR1) his utility level may fall from U2R to U1R, after the change,
but he would continue to be above his reservation utility.
Moreover, social welfare will rise since the consumption of the rich falls about double the rise in the consumption of the poor.
Assume that there are constraints in expanding the supply of water, and that earlier some poor were uncovered. Now enough water
is released to cover two poor consumers. The move is revenue enhancing, since a higher price is paid per unit of water, and the
fall in fixed costs charged for the rich consumer is more than made up by the smaller fixed costs now actually paid by double
the number of poor consumers. Although in order to simplify the diagram we have kept p2 the same for rich and poor consumers,
the poor could be charged a lower price that varies with the consumption slab. Such differential pricing is common. If the coping
strategies of the poor earlier included puncturing pipes, and stealing water, these changes will reduce such wastage, and further
improve revenues for the water supplier.
c Presentation at IGIDR by a French engineer from Cerna in 2002.

Piped water supply has one peculiar feature—the easy after this is dispatched. Lump sum connection charges and
availability of quantity rationing. Managers can just turn taxes are easy to collect to cover costs. But a study in Uttar
off the water supply to whole areas, especially to selected Pradesh found a 50 per cent fall in domestic water
slums. Since this instrument allows them to cut demand to consumption upon full metering (Suresh 1998). An ADB
fit available supply, it is not surprising that they make so (1997) study showed 67 per cent metering in Mumbai
many excuses about the difficulties in extending metering compared to 0 per cent in Kolkata, and 94 per cent in
more fully. They say household demand tends to be constant, Colombo, Sri Lanka, while in Malaysia, Indonesia, and
so they can fix a daily amount and just shut off the valves China all the major cities had 100 per cent metering.
288 India Infrastructure Report 2004

Incentives for Service Contracts and Conservation utilities. Local high-level storage reservoirs, maintained by the
private contractor, may save the electricity costs of 24-hour
The Case of Water Supply in Mumbai (Box 12.1.2) has
water supply, and allow cheaper local treatment of water5.
made clear the enormous health and wastage costs of this
Another irony is that the municipality complains of staff
system of quantity rationing. It has also made clear that
shortage for maintenance and meter reading but its staff per
intermittent supply and poor maintenance affects mainly
1000 connections was 33.3, compared to 17.1 in Kolkata,
household distributory mains. About 82 per cent of the
and 7.3 in Colombo, (ADB 1997). Many low-skill and
leakage takes place in the household connections and the
low-education workers were recruited but even so there is
rest through deliberate puncturing of the distributory mains,
a large store of tacit knowledge and information in older
and household consumption accounts for 90 per cent of the
workers. For example, daily 757 valves are operated. The
water supplied. Irregular supplies lead to considerable
service contractors may be able to usefully employ some of
additional wastage through water hoarding and disposal. If
these workers.
service contracts are given at a pipe width of 150 mm4 in
Another group of potential employers for these workers
various localities, the contractors could be made responsible
could be user groups or water committees. Box 12.1.6 gives
for maintenance of the distributory pipes, attending to
examples of how well these have worked in some cases.
complaints, 24-hour water supply, metering, billing, and
Some such groups may perform monitoring of contracted
collecting. Uninterrupted water supply would reduce
quality, conservation, and information functions as by-
corrosion, leakage, contamination, and metering would
products of other club activities, others may usefully perform
reduce the demand for water and wastage at the household
meter reading and collection functions as sub-contractors,
level, thus keeping demand below available supply. Slum
or be motivated by tax exemption, or awards for water
households could be given one tap each, and the number
conservancy. Pilot groups and institutions aiding the
of standposts reduced. Households could be given a choice
widespread adoption of efficient structures are required.
of a high fixed cost and low or zero unit charge compared
Websites are a useful instrument to achieve these ends6.
to a low fixed cost and a water charge that increases with
These user groups would put pressure on politicians to force
consumption slabs. It may become worthwhile for peak
the municipality to coordinate service contracts, improve
consumption households to install meters themselves. Box
delivery, adopt transparent bidding and a customer-oriented
12.1.5 shows that even slum households are willing to pay,
service, and develop and guard precious core water resources.
and prefer a combination of volume and low connection
There are many ways in which citizens can reduce water
charge. A connection charge has some justification for
wastage, and water committees can play a major role in
recovering capital costs, and taking advantage of the
establishing these practices. With community pressure
willingness to pay, but non-transparent water taxes should
households would learn to replace leaking fissures, not throw
be avoided. Giving the contractor part of the reduction in
away stored water, install 5–10 litre slim dual flushing
water leakage as incentive payment would motivate him to
cisterns or control cocks, use flow control taps, avoid overflow
put in capital equipment unlike in the case of Manila’s
of storage tanks, replace rusted pipes and fittings, and inform
private water companies that could pass on the cost of
the municipality about leakage and theft. It is possible to
leakage to the consumer. Bypass meters could measure the
capture run-off from rooftops in special water harvesting
leakage he saves and give him some percentage of this. His
tanks.
profits then would come from improvement in quality,
On an average, a Mumbaiite uses 40 litres a day for
giving him strong incentives. If the volume charge makes
flushing, and 80 per cent of distributed water is discharged
a large part of payment contingent on delivery, delivery
as wastewater. If a part is reused it would help to cover the
would take place, and the tendency to rationing would be
huge projected demand deficiency, and reduce the necessity
lowered. The whole package could be self-financing.
of developing new sources.
Experience is slowly accumulating in designing PPP
Mumbai buildings constructed after 2001 have to put
contracts. The Andhra Pradesh Infrastructure Act passed
in water recycling procedures to get a no-objection certificate.
recently gives a useful prototype that may be used elsewhere.
Filters allow water used in the kitchen to be reused in the
Such service contracts could be tied with new initiatives
bathroom.
in privatizing metering of electricity consumption. Broad
synergy may be possible between service provision by different 5 Participants at the IIR writer’s workshop made the suggestions
mentioned in this paragraph. Joel Ruet suggested breaking up the
4 R.K. Bhatia, Deputy Municipal Commissioner, Mumbai, water supply according to its uses (only 8 per cent is used for
suggested in an interview with Anirban Acharya on 9 April 2003 drinking) with recycling and local treatment of other parts.
that, if retail privatization were to be considered, the most efficient 6 Chetan Vaidya points out that networking initiatives have
pipe width would be 150mm. started for utility managers.
 Water 289

Box 12.1.5
Willingness to Pay and Distribution between Fixed and Volume Charge

The Vijayawada Municipal Corporation is responsible for household connections. In 1999 it used to charge a connection fee of
Rs 4000 plus a monthly fee of Rs 40. Standposts were dysfunctional and leaking. A cut to Rs 2000 generated a huge response
from slum dwellers, and household connections were laid. (Communication from Arvind Kumar, Municipal Commissioner, VMC,
9 April 1999).
A non-refundable one time deposit or connection charge covered 37 per cent of project cost in Tirupur. Under these ‘own your
own tap’ schemes, which are becoming common, charges vary from Rs 4000–7000 per connection in large towns and Rs 2000–
Rs 3500 in small towns (Suresh 1998).
In Baroda upper- and middle-income households paid 20 per cent less than the actual cost of the water they consumed. Adding
opportunity costs, the average citywide cost of water was 1.7 per cent of household income. For poor households in income groups
less than Rs 1500 a month, the cost was 2 per cent of their household income. They were willing to pay Rs 275 per annum against
the prevailing Rs 43. Wealthier families in income groups Rs 4500–6000, were willing to pay Rs 440 against actual charges of
Rs 200 (Vaidya 1998).

Financial Viability, Entry, and Competition accountability. Foreign firms are also welcome to bid for the
service contracts, and their management experience will be
Reform of pricing, and transparent tariffs allowing price
valuable, but they must not be allowed special concessions
recovery is a precondition to attract private firms to water
supply. Foreign and domestic firms both lack experience in such as exchange cover. Foreign exchange risk can be hedged
Indian conditions; service and the occasional management in the market. On a larger scale, efficient water boards and
contracts are the best way to acquire these. Both will also large private firms should be allowed to bid for management
be looking for domestic capital to finance them. Service contracts in other states, thus improving yardstick
contracts will attract largely the small entrepreneur, well- competition.
structured contracts will make the enterprise viable allowing
financial closure. Thus, domestic enterprise will be developed CONCLUSION
and banks will find more safe avenues to deploy their excess
liquidity. Performance clauses, such as connection targets to While assets should continue to be in public hands, well-
be met, and annuities will ensure high standards and local designed service contracts would have to be incentive

Box 12.1.6
Examples of Community Initiatives

CIDCO, Navi Mumbai, has given the collection of service charges to Senior Citizens Clubs, to whom it pays 1 per cent as commission
(Suresh 1998).
SALMA, or senior advanced locality management (ALM) activists, has delivered promising results in solid waste management
in Bandra. Municipal citizen partnership has worked effectively through ALM and its senior wing. The ward office commits to
responding to the needs of a street committee and the households commit to follow segregation and composting. The ‘Marinisha
Award’ is given monthly to buildings that successfully handle their solid waste (Cunha 2003b).
In Delhi the Bhagidari or participative system of civic governance is working well. District councils have been established where
representatives of residents’ welfare associations (RWA) and the local government meet, discuss, and resolve problems relating to
water supply, energy, and environmental issues.
From Kolhapur, Maharashtra comes an example of sustainable community management of a multi-village water supply scheme.
Operating for more than 20 years as a self-help group, it has a revenue surplus even though charges are lower than in other
neighbouring water schemes. It operates 1723 household connections and 43 standposts. Transparency, committed leadership, and
importance given to consensus and discussions explain its success (Sukthankar 2000, chapter 2, pp. 22).
The Sukthankar Committee suggests an interacting multi-tiered institutional structure for water supply, with an independent
regulator at the state level, and ULBs and RLBs who would appoint service contractors through local Water and Sanitation
Committees, with open feedback from many groups including NGOs and customer councils. Women are to constitute the majority
in these committees. These autonomous entities would have the basic responsibility for the provision of water services. There would
be centrally sponsored ‘Urban basic services for the poor’ in slum areas with neighbourhood groups of 20 houses represented by
Women Resident Community Volunteers. Responsibilities of village level ‘water and sanitation committees (WATSAN)’ would
include formation of rules for use of water supply and sanitation. Thus, an institutional structure would be created for local
distribution of water by community groups.
290 India Infrastructure Report 2004

compatible. They would improve consumer orientation and contracts cannot be fully specified the latter two have a
low level accountability which is missing in public provision, vital role in enforcing contracts. Since citizen groups have
reduce wastage, encourage competition and entrepreneurship voting power, they can motivate politicians to change
in SSPs, achieve financial viability, and develop a thick network focus from direct subsidies and hidden costs to delivery
of low-risk credit-customers for banks and financial institutions. of quality services. Examples of initiatives from India and
They would, however, require support from price other developing countries suggest that changes are
reform, independent regulators, and user groups. Since feasible.

12.2 RESULTS OF AN NSSO SURVEY OF URBAN WATER ACCESS:

IMPLICATIONS FOR POLICY
Peeyush Bajpai and Laveesh Bhandari

Indian cities rate among the lowest quality of life in the world. levying user charges, however, would have to be preceded
The environment, infrastructure, land prices, and general by a good understanding of Indian households’ water
‘livability’—all leave much to be desired (UWSS 1997). This connectivity. The number of sole use connections and shared
chapter attempts to highlight the need for investment in connections need to be ascertained for any kind of tariff
water supply and sanitation infrastructure, duly recognizing plan. Lack of such information can cause reverse effects. For
the limitations and the potential of the households to support instance the increasing block tariff (IBT) model is
such investment. In the process it highlights the policy issues, inapplicable where the water connection is shared among
and the impediments in ensuring access to all. households, a characteristic that we show is common among
Urban local governments across India are generally the poor. The tariff rates under IBT are directly and positively
considered to be in extremely poor financial health (Srivastava related to the consumption beyond a threshold limit. Then
and Sen 1997; Khandwalla 1999; IPFS 1998–9). Their joint users end up paying a higher average price for water
revenues are low and, as a result, their investments and than the sole users. (Boland and Whittington 2001).
expenditures on urban services have suffered. As a result The bulk of the expenditures in augmenting water supply
urban India has poor infrastructure and even worse services. are front ended. Even if financial resources were available
Water supply is one such area where local governments have from external sources, they would typically require some
not been able to keep up with the increase in requirements. contribution from the local governments—which would
Indeed the quality of water supplied may have deteriorated not be easy for most of them. Most households would not
due to pollution of water bodies and groundwater. be able to pay for the up front costs. Thus ‘tatkal’ schemes
Pricing policies usually do not distinguish between access would hardly work. We study (a) the kind of demands of
and usage. Subsidies are generally on prices of water, but households with regard to the quantum and frequency of
access charges tend to be very high as when they are linked water supply, and (b) the quality of water supply.
to possession of built up space. As a result the subsidies are The data reveals that consumers do not expect much.
perverse and anti-poor (WSP 2001). This lack of expectations, we believe, is the strongest
Something that would be considered to be the very basic impediment in improving water supply in urban India. Any
of services—water on tap for 24 hours a day has been water supply improvement programme will only succeed if
unheard of for decades in most Indian towns (ADB 1993). society desires it, that is, it is willing to pay for it in the
A large part of the households depend on their own private
long run and rewards policymakers for delivering it.
tube wells and pumps for their daily water needs. As a result
Consequently, a public awareness programme that aims at
water tables are falling at dramatic rates. In areas of limited
increasing expectations would help facilitate access to water7.
groundwater, or where the heavy metal composition is high,
large costs are imposed on households, even those who can
7 Systematic denial and long years of poor quality supply make
make their own arrangements for pumping out groundwater.
Drinking water on tap for each household is desirable if people get used to the good as it is, so that they are unable to imagine
a significantly better quality good, or its enhanced availability. But
correct pricing of water is to be established. twenty-four-hour
the social benefits of adequate consumption of good quality water
water supply is desirable because there are large negative and its joint use good—sewerage—are so large that, it is important
externalities on account of health and wastage in intermittent to stimulate this consumption. Once critical levels of households in
supply. a particular strata of society consume the product, others will follow.
Maximizing social benefits would actually mean low or So the task may not be as difficult as it appears from the snapshot
very small connection charges and high usage charges. But picture that the survey results reported herein provide.
 Water 291

We also highlight the need for an evaluation of water About 30 per cent of the urban households do not obtain
supply requirements prior to any improvement programme. water from their municipality/local government. However,
Studies have shown that the absence of such assessment has even those households who do have some access to water
contributed to the failure of such programmes in other from the government, have to share it with their neighbours.
countries (Hardoy and Schusterman 2000). Almost 59 per cent of the households either share water
The paper is based on a data set from a survey conducted with their neighbours or the supply is for the community
by the National Sample Survey Organization (NSSO) in (Table 12.2.2).
1998 (the 54th round). More than 110,300 rural and urban Table 12.2.2
households were queried on the following aspects: Right of Use of Water
• Access to drinking water Number of Households Per cent
• Access to bathing water (millions)
• Access to bathroom/toilet Sole 19.6 41.3
• Latrine and drainage Shared 12.7 26.6
• Garbage removal Community 13.6 28.6
Others 1.6 3.5
Additional household level information on lifestyle, habits Not Available 0.0 0.0
such as on commuting, access to media, etc. were elicited Total 47.6 100.0
by the NSSO. This allows us to relate the occupation,
religion, and the geography of the availability of urban In other words, only 41 per cent have exclusive access
services. to their main source of water. Exclusivity in supply is
For the purposes of this paper, we limited ourselves to important, because that is a necessary precondition for
31,323 urban households. We focused on drinking water imposing any user charges/taxes on water. Many of the
because in our opinion it is the most indicative of the issues 70 per cent households who have access to taps have to share
related to water in general. Moreover, the type of queries water from their main source. Of the 33.4 million households
asked about drinking water had a wider scope. who have access to tap water about 54 per cent or 18 million
require some sharing (Table 12.2.3).
PRINCIPAL SOURCES OF WATER Table 12.2.3
Right of Use of Tap Water
The bulk of the households in urban India depend on the
municipal water supply for their daily needs—more than Number of Households Per cent
70 per cent depend on tap water and an insignificant number (millions)
on tankers. Access to underground water is the next most Sole 15.2 45.6
important—wells, tube wells, and hand pumps, together Shared 8.7 26.1
account for more than 27 per cent of the households’ main Community 8.4 25.2
Others 1.0 3.1
water supply. As would be expected, other sources that Missing 0.0 0.0
include tanks, ponds, springs, rivers, canals, etc. are Total 33.3 100.0
insignificant in urban areas (Table 12.2.1).
Table 12.2.1 Sharing is not the only issue. The majority of the
Main Sources of Water households do not have water within their dwellings
(approximately 61 per cent) and have to transport it from
Number of Households Per cent
(millions) the main source (Table 12.2.4).
It would only be natural that those who access water
Tap 33.3 70.1
Tubewells 10.2 21.4 from wells, tanks, tube wells, etc. would not obtain water
Wells 3.2 6.7 within the dwelling. But more importantly, a majority about
Tank/Pond Reserved 0.1 0.2 54 per cent of those who have access to tap water do not
Other Tanks/Ponds 0.0 0.1 have it within the dwelling but have to transport it from
for Drinking the source outside their place of stay (Table 12.2.5).
River, Canal, Lake 0.1 0.2
In sum, not only is the penetration of municipal water
Spring 0.0 0.1
Tanker 0.5 1.0 supply low (about 70 per cent of total households), it is also
Others 0.1 0.2 quite poor in terms of access. Most households that depend
Not Available 0.1 0.1 on tap water have to either share it with their neighbours,
Total 47.6 100.0 or themselves transport it to their dwelling, or both.
292 India Infrastructure Report 2004

Table 12.2.4 some improvements in access to water. But that is not all.
Distance from Principal Source of Drinking Another 14.3 million households do not have any access to
Water: All Urban Households
municipal water. These also need to be provided adequate
Number of Households Per cent access. Thus, of the total 47.6 million urban households at
(millions) least 32 million households require some improvements in
Dwelling 18.4 38.6 access to water within their dwellings.
Premises* 12.9 27.1 Providing access is likely to require major investments in
<0.2 km 15.0 31.5 water supply infrastructure. Consequently, we term the
0.2–0.5 0.8 1.7 15.1 million households that have sole access within their
0.5–1.0 0.4 0.7 dwellings or premises, as those who have a low requirement
1.0–1.6 0.1 0.3
>1.6 km 0.0 0.1
for such infrastructure improvements. The rest 32.5 million
Not Available 0.0 0.1 households are termed as high-requirement households, in
Total 47.6 100.0 that their adequate access requires infrastructure investment.
Note: *The term ‘premises’ means outside the dwelling but within In sum, the large majority of the households require some
the dwelling’s compound. investment in water supply infrastructure to satisfy their
requirements. These households tend to have a lower
Table 12.2.5 economic status as the discussions in the later sections will
Distance from Principal Source of Drinking Water: reveal.
Households with Taps As is apparent from Table 12.2.7, the requirement of tap
Number of Households Per cent water is a problem that spreads across all types of towns and
(millions) cities, irrespective of the size. However, from our perspective
Dwelling 15.3 45.8 what is more important is the economic status of the
Premises 8.4 25.3 households, since that would have a greater impact on the
Others 9.6 28.8 ability and the willingness to pay.
Not Available 0.0 0.0
Total 33.3 100.0
ECONOMIC STATUS AND ABILITY TO PAY
We argue that proper access requires two conditions to The data set does not contain any information on aspects
be fulfilled. One, the household should have the right to such as per capita monthly expenditures of households—
sole use. Only in such conditions would it be possible to the standard measure of economic status. Neither does it
levy any user charges on households. Two, the point of water provide any information on the households’ expenditure on
supply should be ideally within the dwelling. In general, the any products. However, some household- and individual-
closer water can be supplied to the dwelling, the better level information is present that allow us to extract the
would it be. Table 12.2.6 presents the distribution of the following information:
more than 33.4 million urban households who have access
1. TV ownership—The data provide details as to
to tap water.
whether a household has no television, or the type of television
Barely 15.1 million households, that is, 12.9 and 2.2
owned—black and white, or colour.
households, respectively, have sole access within the dwelling,
2. Newspaper subscription—Number of newspapers
or premises. These are the households where it is the easiest
subscribed to by the household.
to levy user charges. Arguably, all the remaining 18.2 million
3. Mode of regular commuting—Individual level
odd households with access to tap water require at least
Table 12.2.6
Distribution of Households across Right to Use and Distance from Source: Households with Taps (millions)

Right / Distance Dwell Premises Others Missing Total

Sole 12.9 2.2 0.1 0.0 15.2
Shared 2.3 5.1 1.4 0.0 8.7
Community 0.1 1.1 7.2 0.0 8.4
Others 0.0 0.1 0.9 0.0 1.0
Missing 0.0 0.0 0.0 0.0 0.0
Tapped Households 15.3 8.4 9.6 0.0 33.3
Untapped Households 3.1 4.4 6.7 0.0 14.3
Total Urban Households 18.4 12.9 16.3 0.0 47.6
 Water 293

Table 12.2.7
Distribution of Households across Towns and Requirement for Water Infrastructure Improvement (%)

Class of Town Low Requirement High Requirement Total Total

(Population as per (millions)
1991 Census)
Less than 50,000 23.3 76.7 100.0 13.3
50,000–2 lakhs 29.2 70.8 100.0 12.0
2 lakhs–10 lakhs 37.5 62.5 100.0 10.4
More than 10 lakhs 40.5 59.5 100.0 11.1
More than 10 lakhs 28.6 71.4 100.0 0.7
Total 31.7 68.1 100.0 47.6
Total (millions) 15.1 32.4 47.6

information on daily commute for work or education by: characteristics. Similar considerations were made for other
foot, cycle, animal drawn transport, public transport, taxi, categories.
autos, two and four wheelers. Table 12.2.9
Step 2—Final Rating for each Household
Based on this information we categorized households on
their economic status. This involved a two- step methodology. Consolidated Number of Per cent Economic
First, each household was given a value rating of 1, 2, or Rating Households of Total Status
3, for each of the lifestyle categories. Therefore, each (millions)
household now had a rating for each of the categories. 3 9.0 19.0 Low
4 10.3 21.6
Table 12.2.8 5 6.3 13.2 Medium
Step 1—Rating Each Household for Each Category 6 4.3 9.0 High
7 4.0 8.4
Lifestyle category Rating
8 2.3 4.8
TV 9 0.4 0.9
No television 1 Unavailable 11.0 23.1
Black and white 2 Total 47.6 100.0
Colour TV 3
Mode of Commuting Cross-checking the Results—Phone Ownership: As mentioned
Foot or bicycle 1
earlier, there are no standard measures for assessing the
Public transport, rickshaw, and animal transport 2
Owned car, two wheelers, taxi, and auto 3 income/expenditure capability of the households. The
Newspapers Subscribed to methodology thus ascertains the economic status, which in
None 1 a manner reflects the capability of the households (Bhandari
One 2 and Dubey 2001). The economic status as obtained after
Two or more 3 the aggregated rating was compared with the ownership of
phones. As expected, a majority of those having telephones
fell in the high economic status category. While majority
In the second stage a consolidated rating was given to of those not having a phone connection belonged to the low
each urban household by summing these value ratings. For and medium economic status. This result indicates that the
instance, a household that does not have a television, but variables used in classification do relate with other
where commuting occurs by a two-wheeler, and subscribes independent variables that highlight the economic status.
to a single newspaper obtains a consolidated rating of 5. We have, therefore, been able to stratify 37 of the
These consolidated value ratings were then used to classify 47 million urban households on the basis of their economic
households as high, medium, and low economic status. Table 12.2.10
The final economic status was, therefore, assigned on the Economic Status and Phone Ownership
basis as listed in the Table 12.2.9. For instance, low economic
Economic Status No Telephone Telephone Total
status households are considered to be those who do not
have a TV, do not subscribe to a newspaper, and commute Low 46.6 2.8 40.6
Medium 14.1 7.7 13.2
on foot or bicycle. Though, to account for certain
High 14.8 75.3 23.0
idiosyncrasies, we also include in this class the households Missing 24.5 14.3 23.1
who perform better than the minimum in any one of these Total 100.0 100.0 100.0
294 India Infrastructure Report 2004

Table 12.2.11 set. Having medium capability with a low need for
Urban Households and Water Supply: Economic Capability improvement, they can pay nominal user charges. They are
and Requirement
also less likely to support price increases to finance better
Infrastructure Requirement supply. Though they may not be as stringent protestors, as
Economic Status Low High Unavailable Total
those in group 1 above.
Low 3.3 16.0 0.1 19.3 Low Need–High Economic Status: These comprise about
Medium 2.3 4.0 0.0 6.3 6.6 million (14 per cent of the total) urban households.
High 6.6 4.4 0.0 11.0 These have the capability to pay for the provision of the
Unavailable 3.0 8.0 0.0 11.0
present service and also for little improvements, such as
Total 15.1 32.4 0.1 47.6
availability of tap water within dwelling from premises.
They are not likely to support improvements in access.
status. The bulk of the households, as expected, are in the However, as is true for higher economic status groups
economic strata who would have a low capability to pay for elsewhere, they have a high inclination to pay premiums for
any capital investment for improvement of water supply at better quality. In the case of water supply this would include
their end. cleaner water with low levels of dirt, minerals, and biological
matter, as well as 24-hour water supply. These groups would
POLICY IMPLICATIONS not support improvements in simple access (though they
may not be vociferous protestors either). However, they are
This section attempts to develop a policy strategy for ensuring likely to be strong supporters if better quality is assured.
good water supply to all. In doing that it takes into
High Need–Low Economic Status: The major chunk of
consideration various factors.
the total urban households, about 15.9 million, fall under
At one extreme we have a set of households who have
this category. Without government subsidy, their low
a high need but their economic status does not reflect their
economic status stands in the way of improvements. For this
ability to make investment for improvements. Such
set of households, budgetary provisions towards infrastructure
households account for almost one-third of the total urban
investment are essential. These households, a third of the
households. At the other extreme there is a set of high
total urban households, may have some ability to pay user
economic status households that already have adequate
charges. However, our data is limited and we cannot make
access—they are unlikely to pay though they have the ability
an unambiguous judgement on that front. For that purpose
to pay. Specific strategies have to be drawn up to cover all
a study that analyses their expenditure and income patterns
these sets of households (Figure 12.2.1).
would be required. This segment of the population has the
Low Need–Low Economic Status: These are less than most to gain by government support, and would be the
10 per cent of all the urban households. These households most vociferous supporters of water supply reform. However,
have their need nearly satisfied at present. They are also not care would have to be taken in designing the fee structure.
in a position to pay for improvements. Status quo needs to Some non-price constraints may have to be considered if
be maintained for these households. O&M expenditure charges are low.
would have to be on the part of the government. However,
High Need–Medium Economic Status: Slightly more than
if any across the board rate increases are made, these
4 million households (8 per cent of the total) fall in this
households are likely to be the most vociferous protestors.
category. Though they have a high need they are limited
Neighbourhoods that contain a large proportion of such
by their capability to generate funds. These households
households would need to be insulated from such increases.
would be less likely to have the ability to pay for
Low Need–Medium Economic Status: About 2.3 million improvements but more likely to be able to pay the user
(5 per cent of the total) urban households belong to this fees. Charges for infrastructure improvements, if imposed,

Requirement
Low High
Economic

Low Status Quo (3.3) Support (16.0)

Status

Medium Nominal Charges (2.3) Charges Spread over a period (4.0)

High Revised Charges (6.6) Investment & Charges (4.4)
Note: The figures in bracket indicate the number of households in each category in millions

Fig. 12.2.1 Strategy Chart for Various Categories

 Water 295

would need to be spread over a period of time. These that improvement programmes do not get much public
households would be willing supporters for improvements, support.
provided that the burden on them is not too high. The approach paper to the Ninth Plan estimates that
85 per cent of India’s urban population has access to water
High Need–High Capability: There are about 4.4 million
supplies. However, adequate details on quality of service
such households in urban India. These households can make
delivered are not generally available. While many schemes
one time capital investment for the required improvement
are designed for a 24-hour supply using 150/200 lpcd
and also pay any recurring charges towards maintenance of
demand, consumers experience regular shortages with only
the necessary services. These households would also be
few hours supply each day (UWSS 1997).
supporters of government initiative on water infrastructure
Take for example Delhi. Only 20 per cent of the capital’s
investment, provided credible initiatives are taken.
population receive 24 hours water supply; 60 per cent obtain
it for between 4–12 hours; and the rest for less than 4 hours.
VIEWS ON CONTRIBUTION A small survey indicated that the majority of consumers in
Delhi would be satisfied with a daily supply of 5–6 hours
The data also contains responses to queries on households’ per day (ADB 1993)9.
perceptions on improving sanitation in general. This also, Consequently, any improvement programme would first
to some extent, reflects their views on water supply in have to improve the consumers’ benchmarking. In other
particular. Note that about 40 per cent of the households words, for improving access, consumer support is essential.
are willing to contribute financially in some way. Another For that, urban consumers have to believe that a 24-hour
30 per cent are willing to put in their own labour for the supply, universal access, and clean water, are not
purpose. Thus, in total about 70 per cent of the households inconceivable but a likelihood. This we believe should be
state that they have some interest in improving their the first focus of water infrastructure improvement
conditions (Table 12.2.12). initiatives.
Table 12.2.12
Willingness to Contribute for Improvements in CONCLUSION
Sanitation in Neighbourhood

Contribution Number of Households Per cent

In almost all cities and towns in India many households do
(millions) not have access to water on tap. Of those that do, most have
to share it with others. Of those who do not have to share
Money 10.2 21.34
Labour 14.6 30.69 it, many have to transport it from outside their dwelling.
Both 8.6 18.18 It is well known that even those who have water on tap for
Neither 14.0 29.5 their sole consumption within their dwelling, the supply is
Missing 0.1 0.3 erratic and of poor quality.
Total 47.6 100 In relating customer’s income proxies to current levels
and quality of consumption, the felt need, and willingness
However queries on the ‘sufficiency’ of water reveal a to contribute to improvements, a strategy for enhancing
different picture—more than 80 per cent of the households water supply infrastructure that is realistic and feasible is
across different segments, consider that they have sufficient drawn out. The current low levels of supply at which
supply. However, the bulk of these households do not have satisfaction occurs, can be a barrier to enhanced supply that
even sufficient access as the data reveal. We argue that this requires significantly enhanced user charges. The size of the
is because the bulk of the population benchmarks support base, the opposition, and the indifference to water
sufficiency at very low levels8. Therefore, it is conceivable reforms are estimated.

9 Denial may be large. See S. Morris (2001) and A. Goyal in

this Report (Chapter 12.1). The low prices for water, in a situation
of limited supply, means that the elastic demand of the rich takes
away water from the inelastic demand of the poor creating denial,
8 See Footnote 7, earlier. even in a city where as much as 200 lpcd is supplied—Delhi.
296 India Infrastructure Report 2004

12.3 MORE CROP PER DROP: FINANCIAL VIABILITY OF DRIP IRRIGATION

A. Narayanamoorthy

The agricultural sector (irrigation), which currently consumes depending upon the nature of crops (wide or narrow
over 80 per cent of the available water in India, continues spaced) and the material to be used for the system. As
to be the major water consuming sector due to the farmers have been getting water at low cost for many years
intensification of agriculture (MOWR 1999; Iyer 2003). from the public irrigation system and also from well
Though India has the largest irrigated area in the world, the irrigation (because of the introduction of flat-rate electricity
coverage of irrigation is only about 38 per cent of the gross tariff ), there is no incentive for them to adopt this capital-
cropped area as of today. One of the main reasons for the intensive technology unless it is necessary. Moreover, since
low coverage of irrigation is poor water-use efficiency under it involves fixed investment farmers often ask questions
the flood (conventional) method of irrigation, which is like what will be the water saving and productivity gains?
predominantly practised in Indian agriculture. Available Is investment on drip irrigation economically viable? What
estimates indicate that water-use efficiency under flood will be the pay-back period of the drip investment? But
method of irrigation (FMI) is only about 35 to 40 per cent the matter needs a thorough study with use of field level
(Rosegrant 1997). Considering the water availability for rather than research station data. (Verma and Rao 1998;
future use and the increasing demand for water from different INCID 1994; Dhawan 2002). Some of the studies have
sectors, a number of demand and supply management shown that the results derived from research station data
strategies have been introduced in India to augment the are substantially different from that of survey data
supply as well as to control the demand for water. One of (Narayanamoorthy 2001).
the demand management strategies introduced recently to
control water consumption in Indian agriculture is drip
method of irrigation (DMI). Unlike flood method of COVERAGE OF DRIP IRRIGATION IN INDIA
irrigation, drip method supplies water directly to the root
zone of the crop through a network of pipes with the help Drip method of irrigation was initially introduced in the
of emitters. Since it supplies water directly to the crop early 1970s by the agricultural universities and other research
instead of to the land, as followed in the flood method of institutions in India with the aim of increasing the water-
irrigation, the water losses occurring through evaporation use efficiency in crop cultivation (INCID 1994). The
and distribution are completely absent (INCID 1994; development of drip irrigation was very slow in the initial
Narayanamoorthy 1996 and 1997a; Dhawan 2002). The years and only from early the 1990s significant development
on-farm irrigation efficiency of properly designed and has been achieved in certain states. In India, the area under
managed drip irrigation system is estimated to be about drip method of irrigation has increased from 1500 ha in
90 per cent, while the same is only about 40 per cent for 1985 to 70,589 ha in 1991–2 and further to 2,46,000 ha
surface method of irrigation (INCID 1994). in 1997–8 (INCID 1994; AFC 1998). Though India has
Though this technology is introduced primarily to increase enormous potential and prospects, the development of drip
the water-use efficiency in agriculture, it also delivers many irrigation does not match expectations in most of the states.
other economic and social benefits to society. Reduction in Table 12.3.1 presents state-wise area under drip method of
water consumption due to drip method of irrigation varies irrigation and its growth rate for 2 time points: 1991–2 and
from 30 to 70 per cent for different crops (INCID 1994; 1997–8.
Narayanamoorthy 1997; Postel 2001). Productivity gains In both time points, Maharashtra state alone accounted
due to drip method of irrigation is estimated to be in the for nearly 50 per cent of India’s total drip irrigated area
range of 20 to 90 per cent for different crops, as per research followed by Karnataka, Tamil Nadu, and Andhra Pradesh.
station data (INCID 1994). While increasing the Over the last 10 years, significant growth has been achieved
productivity of crops, it also reduces the cost of cultivation in the area under drip method of irrigation in many states.
especially in labour-intensive operations. The reduction in However, drip-irrigated area constitutes a very meagre
water consumption in drip method of irrigation also reduces percentage in relation to gross irrigated area in all the states
the energy use (electricity) that is required to lift the water of India. For instance, during 1997–8, the share of drip-
from irrigation wells (Narayanamoorthy 1995 2001). irrigated area to gross-irrigated area was just 0.34 per cent
Drip irrigation technology requires fixed investment and about 0.64 per cent in relation to total groundwater
that varies from Rs 20,000 to Rs 55,000 per hectare irrigated area of the country.
 Water 297

Table 12.3.1
Coverage of Drip Method of Irrigation by States

State Area (ha) Percent to Total Area ACGR

1991–2 1997–8 1991–2 1997–8 (Area)
Maharashtra 32,924 1,22,995a 44.64 50.00 24.56
Karnataka 11,412 40,800b 16.17 16.58 23.66
Tamil Nadu 5357 34,100 7.59 13.86 36.14
Andhra Pradesh 11,585 26,300 16.41 10.70 14.64
Gujarat 3560 7000 5.05 2.85 11.93
Kerala 3035 4865 4.30 1.98 8.18
Orissa 44 2696 0.06 1.10 98.55
Haryana 120 1900 0.17 0.77 58.46
Rajasthan 304 1600 0.43 0.65 31.89
Uttar Pradesh 111 1500 0.16 0.61 54.33
Punjab 20 1100 0.03 0.45 95.01
Other States 2117 1150 3.00 0.47 -9.67
Total 70,589 2,46,006 100.00 100.00 23.13
Note: a—includes state subsidy scheme area of 58,498 ha.
b—includes area under central and state schemes for development of oil palm and sugar cane.
ACGR—Annual compound growth rate percent per annum between 1991–2 and 1997–8.
Source: AFC (1998).

Although INCID (1994) report mentions that over cent), citrus group of crops (11.59 per cent), and
80 different crops are suitable for drip method of irrigation, pomegranate (10.27 per cent) together accounted for about
only a few crops have been dominating in the total area 75 per cent of the total area. Sugar cane is seen as one of
under drip irrigation so far. As of 1997–8, crops like coconut, the important crops in Maharashtra’s total area, which is
grapes, banana, citrus, mango, and pomegranate have not seen at the all-India level. This is because of the fact
together accounted for nearly 67 per cent of total drip that unlike central schemes, Maharashtra has been
irrigated area (Table 12.3.2). Of this, except for banana, providing subsidy for drip irrigation sugar cane, keeping
other crops are not heavy water consuming crops. More in view its water for potential saving.
importantly, out of 26,460 ha of the total area for banana,
Maharashtra alone accounted for as much as 93 per cent
A FIELD VIEW OF DRIP IRRIGATION
at the end of 1997–8.
The crop composition of Maharashtra, which accounts This study was carried out in Maharashtra, where DMI was
for over 50 per cent of India’s total drip irrigated area, is introduced during 1986 itself as a state-sponsored
different from the all-India picture. Of the total area of programme. The effort taken by the state government paid
1,60,281 ha (as March 2000), crops like banana (22.78 dividend and the area under DMI increased from
per cent), grapes (18.15 per cent), sugar cane (12.68 per

Table 12.3.2
Cropwise Area under Drip Method of Irrigation in India: 1997–8

Crop’s Name Area Leading states

Coconut 48,360 (19.66) Karnataka (24.00), Tamil Nadu (21.20)
Grapes 29,630 (12.04) Maharashtra (24.10), Andhra Pradesh (2.20),
Karnataka (3.00)
Banana 26,460 (10.76) Maharashtra (24.50)
Citrus 22,210 (9.03) Maharashtra (15.00), Andhra Pradesh (4.80)
Mango 21,860 (8.89) Andhra Pradesh (9.22), Maharashtra (5.00),
Karnataka (2.30), Tamil Nadu (4.00), Gujarat (1.20)
Pomegranate 15,250 (6.20) Maharashtra (11.40), Karnataka (2.00)
Total all crops* 2,46,006 (100.00)
Note: *—Total will not tally as all the crops are not included. Figures in brackets are percentage to total area.
Source: AFC (1998).
298 India Infrastructure Report 2004

236 ha in 1986–7 to 2,17,447 ha in 2000–2, an increase have compared the cost of each operation between the
of about 62 per cent per annum.10 farmers who adopt DMI and flood irrigation method.
Though more than 26 crops are being cultivated using Table 12.3.3 shows the operation-wise cost of cultivation
DMI in Maharashtra as of March 2000, we have selected for sugar cane and banana for both the adopters and the
2 important water-intensive crops, namely, sugar cane and non-adopters. It is clear from the table that drip irrigation
banana, for detailed analysis. Using the data compiled from reduces the total cost of cultivation by nearly 12 per cent
Commissionerate of Agriculture, Government of in sugar cane and nearly 3 per cent in banana over the
Maharashtra, Pune, two districts, namely, Pune for sugar method of flood irrigation. Though total cost saving in
cane and Jalgaon for banana, were selected for detailed field terms of percentage is not very high in the aggregate, it
survey. A sample of 50 farmers consisting of 25 drip adopters varies across different operations. Among the different
and 25 non-drip adopters have been selected for each crop operations, cost saving is very high in ploughing and
from each district. That is, altogether, a total of 100 sample preparatory works, weeding and interculture, fertilizers, as
farmers (50 drip and 50 non-drip adopters) have been well as irrigation. The reason for the cost reduction in
selected for this study. In order to avoid the differential fertilizers is that some of the adopters have used liquid
impact of sources of irrigation on productivity and other fertilizers and thus, the cost incurred on fertilizers is relatively
parameters, we have considered only those farmers who less. A few earlier studies have reported that drip irrigation
cultivated the 2 selected crops using groundwater under also reduces the cost of fertilizers enormously as it can be
both drip and flood irrigated condition. While the drip supplied with water (liquid fertilizers).12 Some of the farmers
adopters have been selected by random sampling method have argued that even without using liquid form of fertilizers,
using the list of names obtained from the Agricultural it can be reduced by avoiding wastage under DMI. Cost
Department of the respective districts, the farmers cultivating saving in irrigation is mainly due to 2 reasons. First, since
the same crops under flood method of irrigation nearest to water is supplied through pipe network under DMI, it does
the field of drip adopters have been selected purposively as not require more labour. But, in the case of surface method
non-drip sample farmers in order to reduce the variation in of irrigation, labour input is necessary to control water
agro-economic parameters between the 2 categories of supply (changing course of water from one field to other)
farmers. The field survey data from the sample farmers has and to govern leakage and seepage. Second, in addition to
been collected pertaining to the year 1998–9 for sugar cane saving in cost of labour, the cost incurred on electricity for
and 1993–4 for banana.11 While comparing the drip adopters operating pump-sets is also less as drip requires less amount
with the non-drip adopters in respect of various parameters of water as compared to flood method of irrigation. Since
underlined earlier, net present worth (NPW) and benefit– DMI does not warrant much ploughing as it supplies water
cost ratio (BCR) have been computed using the discounted only at the root zone of the crops, there is a cost saving
cash-flow technique to judge whether or not the investment in ploughing and preparatory operations as well. As shown
on DMI is economically viable. by earlier studies, the cost saving is also very high in weeding
operation especially in banana crop. Since DMI supplies
Operation-wise Cost of Cultivation water only at the root zone of the crop, it does not allow
weeds to come up in the non-crop space.
As mentioned earlier, DMI also helps to reduce the cost
of cultivation in certain operations besides increasing water
Water Saving
use efficiency and productivity of crops. Many studies based
on experimental data have confirmed that DMI reduces the One of the important advantages of DMI is that it saves
cost of cultivation, especially in labour-intensive operations substantial amount of water over the method of flood
like weeding, irrigation, ploughing, etc. (INCID 1994; irrigation. It is evident from Table 12.3.4 that despite higher
Verma and Rao 1998). When labour cost reduces, total number of irrigation used by farmers with DMI, the time
cost of cultivation also reduces, because labour cost utilized in hours per irrigation is significantly less for the
constitutes a considerable portion in the total cost of drip adopters. This is mainly because it supplies water only
cultivation. In order to understand the impact of drip to the crop and not to the land. In the case of surface
irrigation on cost of cultivation of different operations, we method of irrigation, there is a need to spend more for each
turn of irrigation because it supplies water not only to the
10 For achieving this impressive growth, the government of
Maharashtra has distributed a subsidy of Rs 332 crore to the drip
adopters through various schemes up to the end of March 2002 12 Application of fertilizers through water is called fertigation.
(GOM 2003). It substantially reduces consumption of fertilizers and wastage
11 Detailed methodologies followed for this study can be seen compared to methods like basal and top dressing. As a result,
from Narayanamoorthy (1996) and (2001). efficiency of input increases in DMI.
 Water 299

Table 12.3.3
Cost of Cultivation for the Adopters and the Non-Adopters of Drip Irrigation (Rs/ha)

Operations Sugar cane Banana

DMI FMI % change DMI FMI % change
over FMI over FMI
Ploughing and Preparation 5218 5790 –9.88 2633 3223 –18.30
Seed and Seed Sowing 6871 8502 –19.17 5331 5416 –1.56
Fertilizers (inorganic) 9587 11,042 –13.18 16,378 17,494 –6.38
Farmyard Manure 7817 8263 –5.39 9975 8316 19.95
Pesticides 1241 1042 19.11 10 – –
Weeding and Interculture 4869 4935 –1.32 1826 2123 –14.00
Irrigationa 5948 6822 –12.81 5757 6379 –9.75
Others 2178 3177 –31.43 2207 1895 16.47
Total Cost of Cultivationb 43729 49,573 –11.78 51,437 52,739 –5.47
Notes: a—Includes O&M costs of pump set and drip set.
b—Costs of harvesting, transport, and marketing are not included in sugar cane as these costs are incurred by the sugar factories in
Maharashtra.
Source: Calculated from field survey data.

crop zone but also to the non-crop zone. Uneven land be 1194 HP hours/hectare for sugar cane and about 3245
surface and water conveying channels also consume HP hours/hectare for banana crop.
considerable quantity of water in surface method of irrigation. Another way of judging the water use efficiency is to
Evaporation losses are also very high in open water conveying compare the water required to produce one unit (quintal)
channels, which increases the requirement of time used for of output between the two methods of irrigation. In order
each turn of irrigation. In spite of higher amount of water to estimate the per quintal requirement of water under drip
supply for each turn of irrigation under flood method, and non-drip irrigated condition, we have divided per hectare
moisture stress is common for those crops cultivated under consumption of water with per hectare of yield of the
FMI. respective crops. As seen in total water consumption, water
Water consumption is measured in terms of horse power use efficiency in terms of output is substantially higher for
(HP) hours of irrigation in the present study. HP hours of those crops cultivated under DMI as compared to the crops
water is estimated by multiplying HP of pumpset with cultivated under FMI. Estimate shows that sugar cane
hours of water used in each pumpset. Water consumption cultivated under DMI consumes only 1.23 HP hours to
by non-drip irrigated crops is substantially higher than the produce one quintal of sugar cane against the requirement
same crop cultivated under DMI (Table 12.3.4). Water of 2.51 HP hours under FMI. Similarly, for producing
saving due to DMI is about 47 per cent for sugar cane and one quintal of banana, drip irrigated crop consumed only
nearly 30 per cent for banana. In absolute term, water 11.60 HP hours of water, whereas the same is estimated to
saving over the method of flood irrigation is estimated to be 21.10 HP hours for non-drip irrigated crop.

Table 12.3.4
Water Use Pattern and Consumption by Drip and Flood Irrigated Crops

Particulars Crop’s Name Method of Irrigation Benefit over FMI

DMI FMI Per cent Value
No. of irrigation Banana 139.14 66.19 – –
per hectare Sugar cane 34.04 25.92 – –

Hours used per irrigation Banana 5.33 16.44 – –

Sugar cane 15.44 33.44 – –
HP of the pumpset used Banana 9.82 10.82 – –
Sugar cane 3.40 3.56 – –
Water Consumption Banana 7884.70 11,130.30 29.20 3245.60
(HP/hours/hectares) Sugar cane 1740.08 2934.08 40.69 1194.00
Notes: Data on banana and sugar cane are related to the year 1993–4 and 1998–9 respectively.
Source: Computed using field survey data.
300 India Infrastructure Report 2004

Table 12.3.5
Productivity of Drip and non-Drip Irrigated Crops

Particulars Crop’s Name Method of Irrigation Benefit over FMI

DMI FMI Percent Value
Productivity (quintal/hectare) Banana 679.50 526.35 29.10 153.20
Sugar cane 1413.20 1168.00 20.99 245.20
Cost of production (Rs/quintal) Banana 75.70 100.20 24.45 24.50
Sugar cane 30.94 42.44 27.09 24.50
Water use efficiency Banana 11.60 21.14 45.13 9.54
(HP hours/quintal) Sugar cane 1.23 2.51 50.99 1.28
Source: Same as in Table 12.3.4.

Productivity Gains productivity difference is possibly due to higher use of yield

increasing inputs by drip adopters. However, this argument
Another important advantage of DMI is increased productivity.
is not validated by our survey data. In both crops, the cost
Studies on DMI have confirmed that the problem of moisture
of cultivation (it is a proxy for level of input use) of drip
stress is completely reduced by providing water at required
adopters is marginally lower than that of the non-drip adopters.
quantity and regular interval, both of which help to increase
Therefore, one can firmly conclude that the increase in yield
the yield of crop to a considerable extent. Our survey results
is mainly because of DMI. Besides increasing the productivity
indicate a significant difference in the productivity of crops
of crops in absolute term the cost efficiency (cost required
cultivated under drip and non-drip irrigated condition (Table
to produce one unit of output) is also found to be higher
12.3.5). The productivity difference between drip and non-
among the crops cultivated under DMI.
drip irrigated crops comes to about 245.20 quintals per ha
for sugar cane and about 153.20 quintals per ha for banana.
That is, productivity gain due to DMI is about 21 per cent ECONOMIC VIABILITY OF DRIP IRRIGATION
in sugar cane and 29 per cent in banana. There are 3 main
reasons for the higher productivity of crops cultivated under Despite the gains, the costs too have to be considered. Drip
DMI over the method of FMI. First, the moisture stress is irrigation system requires relatively a larger amount of fixed
less in drip irrigated crops as water is supplied at a regular investment and, therefore, farmers often ask whether it is
interval and in required quantity (Figure 12.3.1). Second, economically viable or not. The farm business income (FBI),
weed problem is less in drip irrigated crops as it does not which is the difference between gross income and total cost
supply water to non-crop zone. Third, the efficiency of fertilizer of cultivation (our cost refers to A2), explains the relative
use is very high as losses through leaching and evaporation profit level of DMI and FMI. As expected, the FBI of drip
are very minimal while using DMI. One may argue that the adopters is significantly higher than non-drip adopters in

Drip Method
Moisture Content

Sprinkler
Method

Surface
Method
Wilting Point (15 atm)

20 15 10 5 0
Days
Fig. 12.3.1 Moisture Availability for Crops in Different Irrigation Methods
 Water 301

Table 12.3.6
Input and Output Pattern of Drip and Flood Irrigated Crops

Particulars Crop’s Name Method of Irrigation Benefit over FMI

DMI FMI Percent Value
Cost of Cultivation (Rs/ha) Banana 51,437 52,740 2.50 1303
Sugar cane 43,729 49,573 11.78 5844
Gross Income (Rs/ha) Banana 1,34,044 1,02,935 30.22 31,109
Sugar cane 1,14,536 91,871 24.67 22,665
Farm Business Income (Rs/ha) Banana 82,607 50,197 64.56 32,410
Sugar cane 70,861 42,298 67.53 28,563
Capital Cost of Drip Set (Rs/ha) Banana 33,595 – – –
(without subsidy) Sugar cane 54,470 – – –
Capita Cost of Drip Set (Rs/ha) Banana 22,236 – – –
(with subsidy) Sugar cane 35,313 – – –
Notes: Banana and Sugar cane are related to the year 1993–4 and 1998–9, respectively.
Source: Computed using field survey data.

both the crops considered for the analysis. The FBI of drip Table 12.3.6 presents the details of capital cost, subsidy,
adopters is Rs 28,563 per ha higher than that of the non- production cost, and gross value of production for sugar
drip adopters in sugarcane and the same comes to about Rs cane for both Pune and Ahmednagar districts. As mentioned
32,410 per ha for banana (Table 12.3.6). Since drip method earlier, since DMI is a capital-intensive technology, the state
of irrigation increases the productivity, and that too with government provides nearly 50 per cent of the capital cost
reduced cost of cultivation, the FBI of the drip adopters is as subsidy to encourage the adoption of drip irrigation in
substantially higher that of non-drip adopters. crop cultivation. The average capital subsidy comes to
We have next computed both NPW and in BCR using Rs 19,157 per ha for sugar cane and the same is about Rs
discounted cash-flow technique. Since drip irrigation involves 11,359 for banana. As a proportion of the total capital cost
fixed capital, it is necessary to take into account the income of the drip set, subsidy amount accounts for about 35 per
stream for the whole life span of drip investment. But, in cent for sugar cane and about 33.82 per cent for banana.
the absence of observed temporal information on benefits After deducting the subsidy, the fixed capital cost of the drip
and costs, the following realistic assumptions were used so set comes down to about Rs 35,313 per ha for sugar cane
as to estimate both the cash inflows and cash outflows: and about Rs 22,236 for banana. The estimated NPW and
BCR, with the assumption that there will not be any change
1. The life period of the drip set is considered as 5 years
in cost of production and gross income during the entire
for both sugar cane and banana as followed by the INCID
life period of the drip set, are presented in Table 12.3.7.
study (1994).
It comes clearly from the estimate that investment in
2. The cost of cultivation and income generated using
DMI is economically viable even without subsidy. The NPW
DMI is assumed constant during the entire life period of
estimated with subsidy is higher than that without subsidy,
the drip set.
indicating the potential role of subsidy in increasing the
3. Differential rates of discount (interest rates) are
viability of drip investment. For sugar cane, at 15 per cent
considered to undertake the sensitivity of investment. These
discount rate, the NPW of drip investment is about
are assumed at 10, 12, and 15 per cent as alternatives
Rs 2,06,692 per ha with subsidy and the same is about
representing the opportunity costs of capital.
Rs 1,90,025 per ha without subsidy. For banana, the NPW
4. The crop cultivation technology is assumed constant
of drip investment is about Rs 2,57,635 per ha with subsidy
for both sugar cane and banana during the entire life period
and Rs 2,47,753 without subsidy at 15 per cent discount
of the drip set.
rate. As subsidy reduces the density of capital, the NPW of
The investment on the drip set can be treated as drip investment increases considerably after excluding the
economically viable if the net present value of benefits is subsidy amount from the capital cost. Similarly, while
greater than the net present value of costs. The BCR is also reducing the discount rate from 15 per cent to 10 per cent,
related to NPW as it is obtained just by dividing the present a significant improvement is observed in the NPW of drip
worth of the benefit stream with that of the cost stream. investment under both with and without subsidy condition
Generally, if the BCR is more than one, then, the investment in both crops. This implies that farmers will gain more by
on that project can be considered as economically viable adopting DMI if the lending rate goes down, which is
(Gittinger 1984). possible in the new economic environment. The benefit–
302 India Infrastructure Report 2004

Table 12.3.7
Net Present Worth and Benefit Cost Ratio of Drip Irrigated Crops

Particulars Discount Rate % Without subsidy With subsidy

Sugar cane Banana Sugar cane Banana
Present Worth of Gross Income (Rs/ha) 15 3,84,039 4,49,449 3,84,039 4,49,449
12 4,12,902 4,83,228 4,12,902 4,83,228
10 4,34,206 5,08,026 4,34,206 5,08,026
Present Worth of Gross Cost (Rs/ha) 15 1,94,014 2,01,696 1,77,347 1,91,814
12 2,06,287 2,15,431 1,89,179 2,05,287
10 2,15,347 2,25,484 1,97,913 2,15,159
Net Present Worth (Rs/ha) 15 1,90,025 2,47,753 2,06,692 2,57,635
12 2,06,615 2,64,797 2,23,723 2,77,941
10 2,18,859 2,82,542 2,36,293 2,92,867
Benefit–Cost Ratio 15 1.98 2.29 2.16 2.34
12 2.00 2.24 2.18 2.35
10 2.02 2.25 2.19 2.36
Source: Computed using sample survey data.

cost ratio estimated for sugar cane varies from 2.16 to 2.19 irrigation pumpsets13. Appropriate pricing policies on these
with subsidy and from 1.98 to 2.02 without subsidy, while two inputs may encourage the farmers to adopt this
the same ranges from 2.34 to 2.36 with subsidy and from technology. Despite the fact that the investment on drip
2.23 to 2.25 without subsidy for banana crop. Interestingly, irrigation technology is economically viable, farmers are
our estimate of NPW also indicates that the adopters of often reluctant to adopt this mainly because of its relatively
drip-irrigation technology from both the crops would be capital-intensive nature. In order to increase the adoption of
able to recover the entire capital cost from their income in DMI by the cultivators including the resource-poor farmers
the very first year itself. (marginal and small categories), the capital cost required for
This analysis has considered only the explicit direct benefits drip irrigation system should be brought down. By recognizing
realized by the farmers in the form of productivity and does drip industry as an infrastructure industry as well as
not include the various benefits of drip method of irrigation announcing tax holiday for specific time periods to all drip
that the society ultimately expected to gain. For instance, set industries can the competition be increased, which will
by reducing the consumption of water, drip irrigation ultimately bring down the capital cost of the drip set. For
technology conserves water, reduces over-exploitation of widespread adoption of any new technology, quality extension
groundwater as well as degradation of soil fertility and network is essential. The present status of extension network
consumption of electricity required to lift water from the operated mainly by government agencies does not seem to
wells. If we quantify all these benefits and include them in be making significant impact on the adoption of this
the benefit–cost analysis, the BCR of drip investment would technology. Therefore, there is a need to revamp the whole
increase considerably. extension network by involving the drip set manufacturers
in order to increase the quality of extension service. Sustainable
agricultural development will not be possible unless the goal
POLICY IMPLICATIONS of more crop per drop of water is achieved.
Though our study confirms that DMI is economically viable
13 If we rework the data to incorporate the difference in true cost
for farmers even without subsidy in crops like sugar cane
of water reflected in the difference between the two technologies then
and banana, the subsidy cannot be stopped till the new
per hectare for banana the difference in HP hours is (11,130.30–
irrigation technology covers an area adequate to expand 7884.70)=3245.60 (Table 12.3.4 ), which is roughly 2550 units,
subsequently through demonstration effect. However, which today at a low estimate of the cost to serve of Rs 2, amounts
uniform level of subsidy schemes currently followed for to Rs 5100 in favour of drip irrigation. This is more than 10 per cent
water-scarce and water-abundant areas need to be changed of the gross private cost of cultivation. (In the case of sugar cane the
and higher subsidy should be provided for those regions savings is significantly less). This would substantially improve the
private return for the farmer and lead to the adoption of drip irrigation.
where the scarcity of water is acute.
When the good water table is deep down, the economics both social
One of the important reasons for the low spread of this and private would be in favour of drip, and the latter would, even
technology even in the water-scarce area is the availability without any subsidy (or with bundled input subsidies), have led to fast
of highly subsidised canal water as well as electricity for adoption of drip had there been no flat rate-based subsidies in electricity.
 Water 303

12.4 WATER IN URBAN INDIA: THE SCENARIO, ENERGY LINKAGE, AND

PRIVATE PARTICIPATION

Joël Ruet

This section adopts a long-term perspective on cost- it is possible to deliver reasonable services at a lower cost
sustainability. Assessing the ‘fuzzy economic characteristic’ using alternative approaches.
of water, with tariffs numbering in the hundreds and no What is the size of the problem in India? When the
guarantee on quality, it argues for greater scope for product treated wastewater is less than 10 per cent even in urban
as well as consumer differentiation, and a rethinking of the areas and 90,000 unprotected sites for industrial solid waste
entire technical-cum managerial organization of the sector. exist, the task left unfinished is simply humungous. This is
The idea is to reach differentiated targeted costs for specific despite the fact that the urban contribution to the national
uses of water and ways to integrate the management of GDP is more than 50 per cent and urban revenues should
pollution and sanitation, with water supply. The impli- have been used more to solve these issues (Table 12.4.1).
cations of technical choices in water are discussed
as well as the role for the civil society and alternative The Public Sector Response
systems.
Water is mostly publicly managed in India and this model
is in a critical shape today. As far as water access is concerned,
THE WATER SCENARIO: LIMITS OF THE PUBLIC, the average situation is quite poor for some states (Tables
LIMITS OF THE PRIVATE 12.4.2 and 12.4.3), even in official figures. For the total
annual demand for water for India, Sharma (2002) gives
The Johannesburg Summit estimated the world’s annual the figures of 470 billion cubic metres3 in 1985, and 770
public plus private capacity to invest in water and sanitation billion cubic metres3 in 2025, that is, an annual growth of
to be in a range from US$30 billion to US$60 billion. 1.65 per cent. Of this, groundwater meets 80 per cent of
However, the annual requirement has been assessed to be the demand in rural and 50 per cent of the demand in urban
of the order of US$100 billion (according to the pessimists, areas. However, about 70 per cent of available water is
a US$180 billion). Taking a more local perspective, if we polluted, about 80 per cent people live with inadequate
look at the different types of revenues in India, traditional sanitation facilities, about 70 per cent of all diseases are
financing modes for infrastructure may anyway only be water-borne, which materialize into about 73 million
relevant for a happy few for whom they are affordable. workdays being lost each year.
Beyond this group, it is imperative to reduce cost and In Class I cities, only 32 per cent receive more than
quantity, and some use of decentralized systems cannot be 145 lpcd. The situation becomes really worrying in Class
dismissed. In other words, the current estimates in billions II cities (population of more than 50,000 persons), where
of dollars have to be contested on the field to see whether only 20 per cent exceed this level. Further, even when cities
Table 12.4.1
Urban Population and Share of the Urban Sector in GDP

1951 …. 1981 1991 2001* 2011* 2025

Total population (million) 361 … 683 846 1012 1178
Urban population (million) 62 … 160 217 291 377
Per cent urban of total 17.3 … 23.3 25.7 28.8 32.0
Contribution of urban 29 … 47 … > 50
Sector to GDP (%)
Level of water demand (BCM1)
– domestic sector2 25 52
– industry and energy 67 228
Share of total withdrawals (%) 17 27
Notes: 1. BCM = Billion Cubic Metres.
2. Includes both rural and urban households.
* Estimates or Projections.
Sources: Ministry of Urban Development and Poverty Alleviation, New Delhi; World Bank (1998); Ninth Plan Document (2002–7); all
quoted by Meinzen-Dick and Appasamy (2002). Updated with Census 2001.
304 India Infrastructure Report 2004

Table 12.4.2
Access to Water Supply in Urban Areas (Status as on 1997)

Population provided with States

water supply (percentage)
High: More than 85 per cent Andhra Pradesh, Arunachal Pradesh, Delhi, Gujarat, Haryana, Himachal Pradesh,
Jammu and Kashmir, Karnataka, Madhya Pradesh, Maharashtra, Meghalaya, Nagaland,
Rajasthan, Uttar Pradesh, West Bengal
Medium: Between 75 and 85 per cent Bihar, Goa, Manipur, Punjab, Tamil Nadu, Tripura
Low: Less than 75 per cent Assam, Kerala, Mizoram, Orissa, Sikkim
Note: Table 12.4.2 shows that nearly half the states only get a medium to low water supply access. Even in Class I cities the quantity of
water is medium or low in a majority of states).
Source: GoI, 1999:136.
Table 12.4.3
Water Availability in Class I Cities (1988)

Per Capita water availability States

High: Over 160 lpcd Maharashtra, Orissa, Uttar Pradesh, Jammu and Kashmir, Delhi, Chandigarh, Pondicherry,
Medium: Between 120–160 lpcd Andhra Pradesh, Bihar, Gujarat, West Bengal
Low: Less than 120 lpcd Haryana, Karnataka, Punjab, Rajasthan, Tamil Nadu, Kerala, Madhya Pradesh, Manipur, Tripura
Source: Compiled by Saravanan (in Ruet, Saravanan, Zérah 2002, from TARU (1999:7) and MIDS (1995:7))

have supply, continuity of the service can be quite low, as problems of the sector. Chennai has been advanced as a case
Table 12.4.2 shows, even for the four metropolitan cities. in point (WSP 2001), especially because the Chennai Metro
Table 12.4.4 also shows that the only city that recovers Water Board has started buying water from farmers and
the cost of its public water service is Mumbai, which helps hiring transporters to deliver it, or even because, in a purely
it sustain its infrastructure in the long run. The figure for private transaction, lorry owners themselves buy and sell
Chennai relates to the (very limited) part of water that is water. However, one should not forget the context of the
charged. Actually, Chennai recovers only a limited part of scarcity scenario of water supply in Chennai that has led
its expenses, and this mostly through tax collection. Divided to a necessary complement by the private sector. However,
by the total water production, the de-facto water tariff was if one looks at the water supplied by the private sector and
only Rs 0.14 per kl in 2002 (Moench and Janakarajan 2002. compares the economic cost to its value it is no longer
See note to Table 12.4.5), while its total cost was Rs 15.11 certain that it is a panacea. Table 12.4.5 shows the various
per kl. The situation is unsustainable. ‘economic values’ of ‘the water good’. It gives an amazing
range of values, but what do they mean and what is the
A Private Alternative? economic value of water?
In a situation of scarcity, and limits to public sector expansion, First of all, let us look at water as a ‘raw product’, before
the private sector has been advocated as the solution to the treatment. Its extraction cost is difficult to assess, since it

Table 12.4.4
Service and Efficiency Indicators for the Major Metropolitan Cities

Service Indicators Mumbai Kolkata Delhi Chennai

1991 1997 1992 1997 1992 1997 1991 1997
Service Coverage (in %) N.A. 100 64 66 69 86 48 97
Water Availability (in hours per day) 5 5 10 10 7 3.5 3 4
Average Tariff (Rs per cubic metre) 2.1 2.7 1.5 0.5 1.4 1.6 N.A. 11.4*
Efficiency Indicators
Unaccounted Water (%) 24 18 36 50 30 26 N.A. 20
Unit Production Cost (Rs per cubic metre) 1.1 2.4 1 1.2 0.6 1.7 2.9 8.5
Accounts Receivable (months) 2.5 19.7 2 1.5 NA 4.5 9.5 5.8
Staff per 1000 connection 61 33.3 16.2 17.1 8.9 21.4 38.7 25.9
Average O&M cost per person (in Rs) 90.25 1358 84.30 25 98.80 355 85.9 675
Note: * The figure for Chennai relates to the charge for industrial consumers.
Source: MIDS (1995:4) and ADB (1997).
 Water 305

Table 12.4.5 Table 12.4.6

Value of Water for Different Economic Activities Revised Estimates for the Expenditures of Chennai Metrowater
Board (Rs in crores)
Rs/m3
Value of rice produced (per cubic metre of water) 6.00 Description R.E. 2001–2
Value of water at rate Chennai Metro Water Authority 0.14* Power 17.39
charges to customers Chemicals 2.21
Cost of supply for Chennai Metro Water Authority 15.11 Operation & maintenance 10.71
Rate at which water is sold by farmers to transporters 3.15 Drought 45.00
Water sale charges (bulk private) 33 Water tanker hire charges 30.00
Water sale as represented by metro bribes for 67 Payment & provision to employees 52.98
bulk deliveries, drought periods Office & administration expenses 3.99
Water sale in cans (at Rs 1.25/litre) 1250 Depreciation 45.00
Water sale rate in bottles (at Rs 10/litre) 10,000 Interest 116.73
Provision for bad debts 2.00
Note: * This figure is the average figure obtained by dividing the
Contribution to pension fund
total revenues by the total distributed water.
Renewal of asset fund —
Source: Moench and Janakarajan, 2002.
Total 326.01
Source: http://chennaimetrowater.com/archivemain.htm
would require a proper valuation of total electricity costs
(Nagaraj 1998). However, the tariff at which farmers sell
it to landowners might be a proxy, even if it includes the to the CMWSSB) shows that water extraction in rural areas
farmers’ ‘profits’. This gives a ‘value’ of Rs 3.15 per cubic for an urban use may lead to higher social segregation in
metre (kilolitre). Is this ‘the value of water’? Actually, if one concerned villages, as well as endanger the resource,
considers that the cost of treatment does not constitute a ultimately bringing a halt to the local rural economy. When
large share of expenses of the Chennai water board (Table the public sector fails to keep in view these externalities, it
12.4.6), the distinction between raw and treated water does is all the more unlikely that private lorries will be the better
not matter much if one is to seek for a ‘general economic solution.
value of water for urban use’. Then, if one looks at the total This discussion gives a more precise clue to the public
cost for water supply, of Rs 15.11 per kl for the CMWSSB failure, but suggests that the solution is rather to be looked
(Chennai Metro Water Supply & Sanitation Board), it at from the angle of usage segregation and adapting different
means that, one way or another, through public expenditure qualities than to different uses. But in that context, what
and subsidies, the economic system is ready to pay this rate. does the private sector do? From buying water at a Rs 3.15
This implies that the social value of water may be higher, per kl from the farmers, it sells it at Rs 33 per kl in bulk,
and the value is much higher than its mere extraction cost. after little (if any) treatment. This tariff partly reflects the
Compared to this the economic value generated through huge costs of transportation, which are socially inefficient
water-intensive agriculture like rice, is lower by 60 per cent (notwithstanding the congestion and pollution costs in the
(Rs 6.00 per kl). city), but it is also, taking into account the private profit,
In other words, and with appropriate caveats for the use unaffordable to most of the Chennai residents. This way of
of proxies, at the level of the metropolis, a resource with using the private model simply cannot be a long-term general
a high social value is used for an agricultural purpose worth solution. However, it has to be compared with the larger
60 per cent less. This suggests what more refined studies public sector’s failure: the unofficial bribing costs for access
confirm: that less water-intensive agricultural practices should to so-called freely distributed water during scarcity periods,
be thought of in densely populated water-scarce areas which Moench and Janakarajan (2002) estimate to be in the
(Nagaraj 1998). range of Rs. 67 per kl!
Ultimately, and even if the implicit market rationality of We therefore see that for a supposedly homogenous
the above reasoning cannot directly apply to a good whose product, the market price for a cubic metre of water varies
nature differs, whether it is locally extracted or transported from Rs 0.14 (with no guarantee either on quality or quality)
to the city, the rough evaluations give the scope of a public to Rs 33 and even Rs 67 (with no additional guarantees)
regulation of water usage at the metropolitan level. However, even ignoring the cost of bottled water. Our main line of
this has to be organized carefully. Indeed, a preliminary argument is that the sector offers scope for product as well
study (Gambiez Lacour 2003) on the so-called ‘tripartite as consumer differentiation, and calls for a rethinking of the
agreement’ (between the Tamil Nadu Electricity Board, the entire technical-cum managerial organization of the sector.
farmers, and the Chennai Metro Water Supply and Sanitation The advantage of considering separately the various uses
Board (CMWSSB), to extract rural groundwater to be sold of water allows a more refined analysis of the role of the
306 India Infrastructure Report 2004

Table 12.4.7
Regional Variations in the Urban Water Markets in Ahmedabad

Attributes Hard Rock Region Central Alluvial Southern Water Abundant

Public Sector
User perceptions of public 48% pressure adequate, 58% pressure adequate, quality 72% pressure and
supply quality quality generally good generally unsatisfactory quality good
Cost of supply to government Rs 7.81/m3 Rs 6.8/m3 Rs 5.7/m3
Tariff charged by government Rs 1.61/m3 Rs 1.23/m3 Rs 1.21/m3
Per cent tariff recovery 60 61 60
reported by government
Loss to government Rs 6.84/m3 Rs 6.04/m3 Rs 4.97/m3
Per cent willing to pay more 51 36 39
Private Sector
Per cent of users depending on 25 18 2
private sources due to inadequacy
of government supply
Cost of supply in private sector Rs 20/m3 Rs 26/m3 Rs 23/m3
Tariff charged by private sector Rs 45/m3 summer Rs 41/m3 summer Rs 44/m3 summer
Rs 34/m3 non-summer Rs 33/m3 non-summer Rs 34/m3 non-summer
Profit for private sector Rs 25/m3 summer Rs 15/m3 summer Rs 21/m3 summer
Rs 14/m3 non-summer Rs 7/m3 non-summer Rs 11/m3 non-summer
Per cent supporting privatization 25 25 22
of government water services
Source: Survey by VIKSAT, 1999–2000, Ahmedabad, quoted by Moench and Janakarajan (2002)

private sector. For instance, if one looks at the lessons on the long-term cost scenario with the pollution and treatment
the functioning of urban water markets in Ahmedabad costs.
(Table 12.4.7), one can see that there is a section who, in
order to receive better service conditions, are ready to pay
more and go for comprehensive privatization. In some
WATER AND POLLUTION ABATEMENT
zones, anyway, people have to de facto depend on private Though we will keep the focus on water, pollution must
sources. However, if one looks at the production costs, be discussed. Indeed, it links up with water quality as far
which measure the economic cost efficiency, they vary from as treatment of raw water to obtain potable water is
Rs 5.7 to Rs 7.8 per kl for the public sector, but from Rs concerned. Pollution ultimately leads to an increase in
20 to Rs 26 for the private sector14. It is true that these treatment costs, and, hence, water costs. Also, local pollution
higher private costs hide the fact that both the quality and directly impacts groundwater quality, which still meets
the service are possibly higher. But such tariffs, if possibly 50 per cent of urban needs. The question is as follows: at
understandable for potable water, cannot be extended in a the current pollution levels, are abatement costs affordable
sustainable way to the bulk of the consumers, or to the bulk and sustainable for India? Abatement costs are always difficult
of different water usages. This example suggests how, to assess ex ante, and feedback in the Indian context is
according to the water usages, the delineation of the public lacking. However, an interesting benchmark can be taken
role and the private role can be assessed. In the present from the European case (Box 12.4.1).
situation it is very likely that the Indian private sector is As far as India is concerned, the World Bank (1998) gives
able to take charge of the supply of relatively ‘higher’ and estimated costs of ‘inaction’ on health, soil degradation,
‘purer’ usages of water, but not so much the bulk supply costs of power subsidy, and industrial production. Health
for other uses. aspects were estimated to be in the tune of Rs 230 billion
Before examining possible solutions based on usage in 1991–2, (3 per cent of GNP). This gives only an indirect
segregation for some specific Indian cities, let us complete measure of the costs that should be borne to treat water
effluents; basically, if one assumes these social costs, this
14 We consider only the costs, and not the tariffs, since the tariff
means that a benevolent public state can spend up to that
level in addressing pollution. More direct assessment studies
issues are linked with regulation as well as with a trade-off between
affordability to the consumer and a reasonable profit, which is not such as those made by TERI (report ‘Green India 2047’)
the issue discussed here. details that, for water treatment with a full coverage by
 Water 307

Box 12.4.1
A Benchmark: The Costs of Removing Pollution in Europe

In Europe, cost of water has risen with higher quality norms, and Dangeard (2001) reports that ‘At the time the Directive on urban
wastewater treatment (91/271) was adopted (1991), the initial cost analysis estimated that additional investments of € 40 bn to
60 bn were required. In 1998, the Court of Auditors of the EU found that the costs were to reach € 201 bn for nine member
States. … An expert report has estimated the average incremental annual treatment costs per capita to be incurred by implementation
before 2005 in 15 member States: in 1995, at € 41 increasing to € 48 in 2010’. He further states that, from the new water Directive
(2000/60), in the UK, for example, ‘costs of implementation up to 2010 would range from £ 2 billion to £ 9.3 billion’. Tariffs
had to follow and sewerage and wastewater treatment represents a 50 per cent of the water tariffs. Dangeard (2001) indicates that
the charges come to € 302 per capita per year for an annual consumption of 120 cubic metres, approximately 1.2 per cent of
the GDP.
Source: Dangeard (2001)

sewage, the annual cost per capita varies between Rs 188 (Sharma 2002) from sectoral estimates, they are in the range
for small (Class II) cities, up to Rs 316 for ‘million plus’ of 16 to 25 per cent for domestic use, 20 per cent for
cities, for a total of US$99.1 billion over 50 years. Similarly, industries and workshops, 25 per cent in construction and
IEG (2000) estimates US$2.9 billion a year, as urban and public works, 10 per cent in commercial establishments, 15
industrial sector investment requirements between 1995 to 25 per cent in transportation including road, rail and air
and 2020, a sum which would be below 1 per cent of GDP. transport, and storage, and 10 to 25 per cent in public
Ultimately, and even if the costs and land requirement services like government offices, courts, police, etc. In rural
vary from one technique to another, capital as well as O&M irrigation systems, water losses amount to 15 per cent,
costs are far from negligible. Table 12.4.8 shows the estimates 25 per cent for field application losses, and 15 per cent for
from the Ministry of Environment and Forests (MoEF). For farm distribution losses. At the end of the day, there remains
instance, the estimate of O&M costs for an activated sludge only 45 per cent for effective use by the crop.
treatment plant is Rs 3 lakh per mld per year. A capacity Locally, the segment of the population which is the most
of 1 mld used over 365 days represents 3,65,000 kl. That affected by pollution, are slum dwellers, and especially those
is, wastewater treatment nearly adds a rupee per kl, only by from slums that are not notified. There, complete dependence
taking into account the O&M. If capital costs are to be on groundwater resources, in the absence of sanitation,
amortized over 20 years, this adds another Rs 0.5 per kl. leads to serious health hazards due to infiltration.

The Linkage between Pollution, Misuse of Water and

WATER AND ENERGY: CREATING VALUE THROUGH
Inefficient Management
DECENTRALIZED SYSTEMS AND DECENTRALIZED
In India, treatment is negligible, but the need for it would MANAGEMENT
be rendered easier by reducing the demand of water. The
latter is feasible if one looks at the ‘useful’ part of the Differences in technology including design, and modes of
produced/used water. In that respect, huge untapped savings organization for maintainence and monitoring imply
exist, and, if one takes some examples of quantity wasted different categories of stakeholders, with varying patterns of

Table 12.4.8
Costs of Domestic Wastewater Treatment

Technology Land Required Capital Costs Operation and Maintenance Costs

Hectares/mld Lakh rupees/mld Lakh rupees/mld/year
Activated Sludge Treatment Plant (ASTP) 0.4 35 to 40 3.0
Oxidation Ponds 1.0 12 to 15 0.5
Aerated Lagoons 0.6 15 to 20 2.75
Upward Anaerobic Sludge Blanke 0.2 23 to 28 1.5
Duckweed and fish culture technology 0.7 to 1.0 10 to 12 0.5 to 1.0
Karnal technology 1.0 to 1.5 0.6 to 0.8 0.25 to 0.3
Trickling filter 0.4 35 to 40 3.0
Source: Ministry of Environment and Forests, 1998
308 India Infrastructure Report 2004

consumption, to result in varying in the use of water and pollution sources, again focusing on the share of the
hence economic valuation ultimately, a different willingness wastewater that really needs thorough treatment. These
to pay, and different capacities to actually enforce willingness kinds of measures (whose financial viability needs further
to pay. This offers scope to specify and reach appropriately research and is insufficiently documented in India today),
modified levels of purification and treatment according to necessitates a managerial change (decentralization of
the desired quality of water which depends on the type of information and decision making) as well as technical
use that is expected. Table 12.4.9 shows the break-up of adaptations.
different domestic uses by supply standards. Flushing and
house cleaning, that is, the use of water that needs the Pollution Avoidance
lowest norms of potability and clarity of water represents
India traditionally has a workforce that is linked with
30 per cent of domestic usage15. Given the energy share in
alternative institutional arrangements (for example,
total supply costs, potential savings may be made here.
recycling). These can be taken advantage of and developed,
Indeed, this water could be produced locally, either by using
compared to capital and energy-based techniques.
groundwater sources that otherwise tend to be gradually
Dangeard mentions one way that has been relatively
eliminated by public organizations, or more likely by local
unexplored, viz., the (partial) local reuse of water for some
(area-wise, ward-wise) recycling of the wastewater coming
uses, to be segregated from the potable water. They would
from the other 70 per cent uses of water. This would, in
turn, save on transportation costs for water. Such a measure include bathing, washing, cleaning, toilet, or many industrial
is relevant not only for cities like Chennai (resources uses as discussed above. These ‘short circuits’ are being
constrained) or Bangalore (where water has to be pumped attempted in Europe and in the USA but are still
from a 150 km distant source and brought to the heights insufficiently considered in India, where no publicized
of the city), but even for a water-rich city like Mumbai, experiment exists.
which, offers 250 lpcd to its people, but takes it from distant Though this is difficult to appreciate in a growth situation
sources (Ruet, Saravanan, and Zérah 2002). where expansion of cities favours the development of
Table 12.4.9 also clearly shows that only a 15 per cent connections at the expense of pollution abatement, the
of water consumption is related to food, drinking, and latter is, in the long run, much more profitable and, therefore,
dishwashing, that is, only this share of the total consumption sustainable and the policy makers should focus their budgets
really requires treatment norms up to potable standards. on this issue. This should be done in a targeted way, as far
The problem is that potability norms are lost through as possible at the source level itself, and in that respect it
leaking pipes, catchment pollution, and insufficient treatment is far from obvious that even Common Effluent Treatment
or pipe level pollution. This treatment can be done more Plants (CETPs) can be a panacea. Indeed, industrial pollution
locally, (with a differentiated piped system) to actually reach is more efficiently treated if specific pollutants are directly
potability standards. reduced from the source, and not from a set of sources, with
Simultaneously, this segregation also means that mixed pollutions, as non-homogeneous industrial clusters
appropriate wastewater treatment can be very specific to the would lead to.
If one looks at the energy linkage, and especially on the
Table 12.4.9 energy component in the inputs for water supply, the figures
Break-up of Minimum Domestic Water Supply Standard from Chennai are illustrative of two aspects. First of all,
among technical operational costs on the centralized piped
Use/ Activity Amount (in lpcd) Percentage system (power, chemicals, O&M), power amounts to 57 per
Drinking 5 4 cent. Second, this share is actually much higher as far as
Cooking 5 4 long-term energy use is considered, since the insufficiencies
Bathing 55 41
of the centralized piped system have led to a ‘drought
Cleaning dishes 10 7
Cleaning house 10 7 scheme’, and hiring of water lorries. In both cases, the
Washing clothes 20 15 energy part represents the major part of the total burden
Flushing toilets 30 22 of Rs 75 crore! The amounts at stake imply that alternative
Total 135 100 technical approaches merit serious consideration.
Source: Sanjeev, 1997:15.
Institutional Structure in Metros: Different Kinds of
15 The part relating to flushing of the toilets may not necessarily Administrations and Administrative Contexts
represent ‘in-house’ toilets: it represents what, one way or another,
is needed to evacuate the excreta: that way, this estimate applies to Among cities, and independently from the water-resource
all categories. situation, the nature of the concerned administration in
 Water 309

charge of water can vary from a centralized agency with (expand the system: treatment, distribution), and as well to
principally a single output-oriented objective (water supply), optimize the present system are executed by the Public
strongly embedded in the state/municipal sphere, to a rather Health Engineering Department (PHED) of the state, which
autonomous agency, evenly focusing on different issues and is less favoured by the municipality. Public funds channelled
aiming at a managerial delegation, and making attempts for through the various agencies of the Indian public financing
decentralized techniques. In other words, and using the and development system: the PHED as an executing agency,
main metropolitan cities of India, this ranges from a Kolkata- and HUSA plus HUDCO as the financing agency. This
like situation, surprisingly both politically centralized and source of funding, by design, does not include any
unclear in terms of administrative responsibility, to a Chennai- conditionality regarding a tariff hike or any kind of tariff
like situation, where institutional and organizational reform and is, therefore, favoured by municipal corporations.
innovations exist, and are worth being replicated (Box 12.4.2). The second aspect, of optimizing and rationalizing the
operation, involves a more unusual set of internal studies,
Institutional Structure in Class I Cities designing of solutions not off-the-shelf, and a high dose of
Let us sum up the respective characteristics of the water social marketing to make the general public aware of the
scenarios in three Class I cities Hisar (Haryana), Vijayawada benefits they can derive from such measures.
(Andhra Pradesh), and Rajkot (Gujarat). The tariffs of these In contrast to the Hisar-type municipality, Vijayawada
cities are given in Table 12.4.10. shows a situation where the entire management of water is
In Hisar, where water is available, there are no typical integrated within a single organization—the Vijayawada
quantity constraints. Interventions to increase the capacity Municipal Corporation (VMC). The financing of the

Box 12.4.2
Proposed Institutional Evolutions for Chennai

Chennai illustrates the levels managerial decentralization can reach within the Indian administrative frame. While it relies on property
tax collection and government support (or government-routed support) for its schemes, the CMWSSB is still mostly structured
around procedures (technical and administrative), and not around services. As a matter of fact, providing full sanitation, in spite
of a tax collected for the purpose, is not looked up as a compulsion. Similarly, as a response to the water scarcity situation, the
Board has de facto given up on satisfying and supplying a part of the population, or rather a segment of the consumption. Whatever
the pragmatic reasons of impossibility to face the demand through a centralized technical system, the Board has had the courage
to promote some more decentralized resources.
CMWSSB treats 60 per cent of its wastewater, of which one-fourth (15 per cent) is sold to industry and power plants. O&M
of water-production wells, water-treatment plants, and sewage pumping stations have been outsourced since 1982 and almost half
of the 118 stations are operated by 24 private contractors. To supplement its piped delivery, it has signed contracts with 500 ownersa
of tankers, who are grouped in an association and has entered into tripartite agreements with TNEB and farmers to purchase water.
Exnora, an NGO with substantial experience in organizing informal solid waste collection, has formed committees among slum
dwellers to organize distribution of water supplied by tankers. CMWSSB has also tried to decentralize management. It has merged
billing and collection at the depot level and made the depot engineer responsible for collection targets. Financial powers have been
delegated down to the assistant engineer level. An information and facilitation cell has been established and monthly meetings
of ward committees are held.
Internally, this did mean that the headquarters have promoted a less centralized model over their technical and engineering
officers. But the Board has also gone beyond that, by promoting an internal delegation of power as well as a better service interface
as far as complaints monitoring is concerned. Whatever the contextual situation in terms of resource and perhaps human resource
at the top of the board, and in spite of the fact that the outsourcing has been mostly decided on a human resource policy basis,
the institutional aspects matter. It can be argued that, compared to the Corporation form, the Board form, with a more clear-cut
budget and allocation from resources to expenditures, allows a better isolation and identification of the tasks, rights, and duties
of the organization. Likewise, the integration of the supply activity and the planning and maintenance, brings some clarity in the
establishment and management of control rights over the assets. And, last but not the least, a sound definition of enforceable and
reasonably sustainable (at least in the short run) tariffs leads to soundness in decision making. Better definition and delineation
of the boundaries of the institution thus allows focusing on the margins of this perimeter, and consideration of suitable policies
in terms of outsourcing. Within that framework, marginal optimizing solutions are envisaged, but keeping this very integration
intact.
a There are other owners who remain outside the association but contract on a similar basis.
Source: Ruet, Saravanan, and Zérah 2002.
310 India Infrastructure Report 2004

Table 12.4.10
Tariffs and Costs in Three Class I Cities

Hisar Vijayawada Rajkot

Inhabitants (thousands) 257 845 850
Population covered (through piped system; rest is through 60 % 89 % 65 %
hand pumps or tanks)
Supply (lpcd) 135 150 70
Total municipal expenditure (in crores) Not relevant 155 79
Municipal budget for water (operating costs in crores) 4.6 7.5 10.8
Part recovered (in crores) 1.01 5.8 Less than energy costs
Note: Water supply (80%) + sewerage (20%).
Source: interviews.

infrastructure remains largely within the funding provided these few examples indicate that a decentralized organization
by the state, and again the kind of operations to be undertaken with powers delegated to the municipal level helps in
has to fit with traditional procedures and practices. However, promoting more original techniques and managerial schemes.
the newly-developing areas may be technically isolated,
though they are within the VMC. This zone, the eastern PRIVATE MANAGEMENT IN WATER
part of the city, used to be managed through panchayats
before it was incorporated within the VMC. Therefore, this Socially equitable, and economically efficient supply options
zone has its own supply through borewells and a stand-alone would be possible only if water is considered as an urban
distribution network. One option could, therefore, be a common property resource (CPR). Such a conceptual shift
more decentralized option taking into account the existing necessarily addresses (i) the definition of the property rights
system rather than building a new system. For this, and as and the CPR relationships at stake, (ii) the governance of
per the Vijayawada Municipal Corporation Act, there is no the sector, and (iii) the need for setting up coordinated de
formal request necessary per se for the sanction from the jure rights on the resource through both channels of local
state on designing such a decentralized project. Here, the (including civil society) and global actors. In both Chennai
fact that, compared to Hisar, the PHED does not have to and Kolkata the property rights are exercised by the
intervene, is clearly an asset. There is the issue of pollution municipality (state) and private individuals and enterpreneurs,
as far as the local water cycle is concerned, but here too, but lack of coordination between the two leads to higher cost
the VMC will be responsible for seeking the state clearance than necessary.
on the project. Since the city manages the waste disposal The reform of the governance can be based on specific
and has shown itself to be reasonably competent, it can be trade-offs between some amount of reform of the public
expected that the implementation of a local clean cycle will sphere (delegation of power within administrations, actual
be feasible and positively supported as far as environmental decentralization from state), and contracting out. The
regulations are concerned. Chennai example shows relatively better governance and a
In Rajkot, the water management is included in the technical-cum-managerial solidity of the system that allows
responsibilities of the municipality. Further, its experience capacity building through delegation, and contracting out
in PPP17 makes it possible to hope for a good response on (though the politicization of such deals is not negligible).
experiments of ‘differentiated use of water’. Besides, the With the recourse to local entrepreneurs for pumping
acute scarcity, with service durations of half-an-hour a day, stations, which the experience of Chennai has proved to be
is the strongest incentive for the general public to welcome profitable, as it reduces formal costs. The consequence is
original and decentralized attempts. Further, in the local that the informal sector gradually increases its part of the
context of resource scarcity, the traditional infrastructure- revenue, while the formal sector reduces its deficit for a
driven and quantity-driven public projects can no longer be given structure by aggregating local entrepreneurs. The
exclusively advocated. experience of Chennai shows that this is not enough to
From the example of a few Class I cities, one can again cover the structural financial situation, but that there is a
notice that a clear-cut allocation of responsibilities in favour high incremental profitability in such measures. The best
of a single administration is an asset in evolving projects; strategy, would be to support and coordinate with the
informal sector, and to start with local entrepreneurs. Then,
17 These have to be properly evaluated on field, through an ex- after some time, when all the (limited) possibilities given
post analysis of several projects and decision-making processes. to the local entrepreneurs and the informal sector have been
 Water 311

reached, or close to be fulfilled, one can still consider the the system that allows capacity building through delegation
recourse to a global service company. and technical decentralization.
Given the administrative structure of water agencies in The traditional question of giving water to the private
India, access to information is virtually impossible in the sphere as a concession versus leaving it within the public
sense that the information gathered is rarely compiled in an sphere can be examined in this very framework. There are
integrated manner. Since the management of physical no credible projects for a large management contract, or
quantities is done independently from the management of recourse to the large (international) private water companies.
financial sources, which is rather an administrative clearance On the contrary, and despite the doubts and hesitations on
process, investors face difficulties in investing in Indian its possible implementation, Mumbai seems to mostly focus
public agencies or negotiating a contract with them (Ruet on schemes supported by large international organizations.
2001 and 2003). Initial recourse to local entrepreneurs de Chennai at the moment is in the intermediary position, as
facto generates information and clarity about legal and it simultaneously considers local outsourcing for the water
property rights that help analysing the situation of the supply and sanitation operation, and large private companies
agency for the potential investor, that is, the global service as far as waste disposal is concerned. However, questions
company. In that context, negotiating such a contract after arise on both the environmental sustainability (without
the involvement of local entrepreneurs is also beneficial to recycling), and the economic sustainability (compared to
the service company, and leads to less uncertainty and, employment generation and recourse to the informal or
therefore, to a lower premium in financing investments, and NGO sector).
to more targeted investments. That way, the transitory period, So far, the debate is limited to the adherents of technical
that is, the period that the company needs for the benefits centralization, among them mostly engineers of water boards/
to cover the costs can be considered as substantially less as departments, and the others who hold that decentralized
compared directly bringing in a global company. options such as water tankers (but mostly rain water harvesting),
This strategy is thus to be preferred by the global company, and the use of local water sources through mini water supply
by the policy makers in the sense that it saves jobs and schemes hold the key to service improvement. Applied research
provides preliminary (visible) results before privatization, on the actual costs for the society (economic and environmental
and for the water agency itself in the sense that the strategy costs), and the potential development of such schemes
maximizes its revenue. (identification of actors that can be involved, financing tools
required, quality control and monitoring, etc.) could be the
CONCLUSION first step towards engaging in a fruitful debate.
Bringing in local entepreneurs in outsourced tasks and
Water scenarios are city-dependent. If there is an opportunity in smaller supply systems could be useful intermediate stage
to jointly promote decentralized techniques, decentralized before global companies are brought in. The clarity of
management, and delegation of management, even in limited information and diligence that this intermediate stage can
respects to the local private entrepreneurs and communities, bring are vital to reducing risks for both the public entity
one should consider the initial situation, and adopt a more and the private company. There is considerable savings in
dynamic approach. Indeed, the Chennai example shows water usage and in costs that are possible if the variegated
both governance and a technical-cum-managerial solidity of uses of water are recognized to exploit recycling opportunities.

REFERENCES

ADB (1997) Second water utilities data book—Asian and Pacific Boland, John J. and Dale Whittington (2001) ‘Water tariff design
region, edited by A.C. Mc Intosh, C.E, Yniguez ADB. in developing countries: Disadvantages of IBTs and advantages
—— (1993) Theme Paper—Service Levels and the Urban Poor, of UPR designs’, Draft.
Managing Water Resources to Meet Megacity Needs, Brousseau, Eric and Jean-Michel Glachant (2002). ‘The Economics
Proceedings of Regional Consultations, Manila. of Contracts and the Renewal of Economics’ in Eric, Brousseau
AFC (1998) Evaluation of Drip Irrigation System, Agricultural and Jean-Michel Glachant (eds.) 2002.
Finance Corporation (AFC) Limited, Mumbai, India. da Cunha, Garson, (2003a) ‘51 Letters—no Reply!’, Bombay
Bhandari, Laveesh and Amaresh Dubey (2001) ‘The affluent in Times, The Times of India, 21 March.
—— (2003b) ‘SALMA banishes garbage’, Bombay Times, The
India 2000’, mimeo, NCAER.
Times of India, 9 May.
Biswas, Asit K. (2001) ‘Water Policies in the Developing World’,
Dangeard, Alain-Louis (2002) Conditions for a new start in
Water Resources Development, Vol. 17, No. 4, pp. 489–
decentralised management of water, Note to ICRIER, Meed
99. SA, 22 August 2002, unpublished document, Paris.
312 India Infrastructure Report 2004

—— (2001) Measuring the costs of water pollution (micro and IPFS (1999) Indian Public Finance Statistics, 1998–99, Department
macro approaches) in the West and in India, Network Indo- of Economic Affairs, Ministry of Finance, GOI , New Delhi.
French Environmental economists: DEMATEDEE Paper No. Iyer, Ramaswamy R. (2003) Water: Perspectives, Issues, Concerns,
1, Paris, 29 June. Sage Publications, New Delhi.
Deshpande, R.S. and A. Narayanamoorthy (2001) ‘Issues Before Khandwalla, Pradip N. (1999) Revitalizing the State, Sage, New
Second Irrigation Commission of Maharashtra’, Economic Delhi.
and Political Weekly, Vol. 36, No. 12, pp. 1034–43. Klein, Michael (1996) ‘Economic Regulation of Water Companies’,
Dhawan, B.D. (2002) Technological Change in Indian Irrigated Policy Research Working Paper 1649, World Bank, Private
Agriculture: A Study of Water Saving Methods, Commonwealth Sector Development Department.
Publishers, New Delhi. Landingin, Roel (2003) ‘Loaves, Fishes and Dirty Dishes: Manila’s
Disha (1996) A rapid assessment survey of the health and Privatized Water Can’t Handle the Pressure’, in The Water
environmental impacts of solid waste recycling, final report, Barons project coordinated by William.
pp. 42, Direct Initiative for Social & Health Action, Kolkata. Lundqvist, Jan Paul Appasamy, and Prakash Nelliyat (2002),
Gambiez, Marie and Emilie Lacour (2003) Rural impact of farmers Dimensions and Approaches for Third World City Water
selling water to Chennai Metropolitan Water Board: A case Security, working document.
study of Magaral Panchayat, Centre de Sciences Humaines, Marsden, The Center for Public Integrity, Washington D.C.,
New Delhi. available at http://www.icij.org/dtaweb/water/default.aspx?
Gittinger J, Price (1984) Economic Analysis of Agricultural Projects, SECTION=CHAPTER&ID=5.
Second Edition, The Johns Hopkins University Press, London. Meinzen-Dick, R. and P.P. Appasamy (2002) Urbanization and
Glachant (eds.) (2002) The Economics of Contracts: Theories and Intersectoral Competition for Water, in Finding the Source:
Applications, Cambridge University Press, UK. The Linkages Between Population and Water, Woodrow Wilson
Gleick, Peter H., Gary Wolf, Elizabeth L. Chalecki, and Rachel International Center for Scholars, Environmental Change
Reyes (2003) ‘The New Economy of Water: The Risks and and Security Project, Washington D.C. pp. 27–51.
Benefits of Globalization and Privatization of Fresh Water’, Menard, Claude and Stephane Saussier (2002) ‘Contractual Choice
The Pacific Institute, Oakland, California, available at http:/ and Choice and Performance: The Case of Water Supply in
/www.pacinst.org/reports/new_economy.htm. France’ in Brousseau, Eric, and Jean-Michel Glachant (eds.)
Godoy, Julio (2003) ‘Water and Power: The French Connection’, The Economics of Contracts: Theories and Applications,
in The Water Barons project coordinated by William Marsden, Cambridge University Press, UK.
The Center for Public Integrity, Washington D.C., available MIDS (1995) Water allocation and management in the Chennai
at http://www.icij.org/dtaweb/water/default.aspx?SECTION= metropolitan area, by Paul Appasamy et al., pp. 63, Chennai
CHAPTER&ID=2. Institute of Developmental Studies.
GOM (2003), Economic Survey of Maharashtra: 2002–3, Moench, Marcus and S. Janakarajan (2002) Water Markets,
Government of Maharashtra, Mumbai. Commodity Chains and the Value of Water, presented on
—— (1999) Maharashtra Water and Irrigation Commission Report, CSH Seminar on Water Management, New Delhi, 17
Government of Maharashtra, Mumbai. December.
Government of India (1999). National Commission for Integrated Morris, Sebastian (2001) ‘How Low Usage Charges for Drinking
Water Resources Development Plan. Report of the Working Water Act against the Interests of the Poor’, Box 9.4.1 in
Group on Water management for Domestic, Industrial and Anish, Nanavathy, ‘Implications of the 74th Amendment on
other Uses, Ministry of Water Resources, New Delhi, the Water and Wastewater Sector’, Indian Infrastructure Report,
September. 2001, 3iNetwork, Oxford University Press, New Delhi.
Gutierrez, E. (2001), ‘A Survey of the Theoretical Issues on Private MOWR (1999) Report of the Working Group on Water Availability
Sector Participation in Water and Sanitation’, Framework for Use, National Commission for Integrated Water Resources
Document, PSP Project Team, available at http://www. Development Plan, Ministry of Water Resources, Government
wateraid.org.uk. of India, New Delhi.
Hardoy, Ana and Ricardo Schusterman (2000) ‘New Models for Municipal Engineers’ Association (MEA) and Brihan Mumbai
the privatization and sanitation for the Urban Poor’, Licensed Plumbers’ Association (BMLPA) (2003) Water,
Environment and Urbanization, Vol. 12, No. 2, October. published on World Water Day, 22 March, Mumbai.
Hart, Oliver, Andrei Shleifer, and Robert Vishny (1997) ‘The Nagaraj, Nareppa (1998) Institutional Management Regimes and
Proper Scope of Government: Theory and an Application to Economic Instruments for Sustainable Use of Water,
Prisons’, Quarterly Journal of Economics, 112, pp. 1127–61. Contributions CSH 99/2, Centre de Sciences Humaines,
Hobbs, Erika (2003) ‘Low Rates, Needed Repairs Attract ‘Big New Delhi.
Water’ to Uncle Sam’s Plumbing’, in The Water Barons project Nanavathy, Anish (2001) ‘Implications of the 74th Amendment
coordinated by William Marsden, The Center for Public on the Water and Wastewater Sector’, Indian Infrastructure
Integrity, Washington D.C., available at http://www.icij.org/ Report, 2001, 3iNetwork, Oxford University Press, New Delhi.
dtaweb/water/default.aspx?SECTION=CHAPTER&ID=8. Narayanamoorthy, A. (2001) Impact of Drip Method of Irrigation
IEG (2000) ‘The Case of Water Resources’, note. on Sugarcane Cultivation in Maharashtra, Agro-Economic
INCID (1994) Drip Irrigation in India, Indian National Research Centre, Gokhale Institute of Politics and Economics,
Committee on Irrigation and Drainage, New Delhi. Pune, India.
 Water 313

—— (1997a) ‘Drip Irrigation: A Viable Option for Future Sivanappan, R. K. (1994) ‘Prospects of Micro-Irrigation in India’,
Irrigation Development’, Productivity, Vol. 38, No. 3, pp. Irrigation and Drainage System, Vol. 8, No. 1, pp. 49–58.
504–11. Solo, Tova Maria (1999) ‘Competition in Water and Sanitation
—— (1997b) ‘Economic Viability of Drip Irrigation: An Empirical and the Role of Small-Scale Entrepreneurs’, in The Private
Analysis from Maharashtra’, Indian Journal of Agricultural Sector in Water Competition and Regulation, World Bank,
Economics, Vol. 52, No. 4, pp. 728–39. Washington D.C.
—— (1996) Evaluation of Drip Irrigation System in Maharashtra, Srivastava, D.K., Tapas K. Sen (1997) Government Subsidies in
Mimeograph Series No. 42, Gokhale Institute of Politics and India, National Institute of Public Finance and Policy, New
Economics, Pune, India. Delhi.
NSSO (1999) Drinking Water, Sanitation and Hygiene in India, Suresh, V. (1998), ‘Indian Experience in Urban Water Supply and
Report 449(54/31/1), National Sample Survey Organization, Sanitation’, HUDCO Bhawan, Updated version of country
GOI, New Delhi, July. presentation at ESCAP workshop, New Delhi.
Sukthankar, D.M. (2000) Draft Sukthankar Committee Report
Postel, Sandra, Paul Polak, Fernando Gonzales, and Jack Keller
on Operation, Maintenance and Management of Rural and
(2001) ‘Drip Irrigation for Small Farmers: A New Initiative
Urban Water Supply Schemes, Water Supply and Sanitation
to Alleviate Hunger and Poverty’, Water Resources Journal,
Department, Government of Maharashtra.
No. 211, December, pp. 87–102.
TARU (1999). Shillong Water and Urban Environmental Services
RBI (2003) Handbook of Statistics on the Indian Economy, RBI, (Meghalaya): A Situation Analysis, Prepared for AUSAID,
New Delhi. TARU Leading Edge, New Delhi.
Ronderos, Maria Teresa (2003) ‘A Tale of Two Cities’, in The Water Tirole, Jean (1994) ‘The Internal Organization of Government’,
Barons project coordinated by William Marsden, The Center Oxford Economic Papers, 46, pp. 1–29.
for Public Integrity, Washington D.C., available at http:// UWSS (1997) Consultants Report, Workshop on Urban Water
www.icij.org/dtaweb/water/default.aspx?SECTION= Supply and Sanitation, May.
CHAPTER&ID=7 Vaidya Chetan 1998, ‘Willingness to Pay for Water and Sanitation
Rosegrant, W. Mark (1997) Water Resources in the Twenty-First Services: A Case Study of Baroda’, HSMI, New Delhi.
Century: Challenges and Implications for Action, Food and Vaidyanathan, A. (1999) Water Resources Management: Institutions
Agriculture, and the Environment Discussion Paper 20, and Irrigation Development in India, Oxford University Press,
International Food Policy Research Institute, Washington New Delhi.
D.C., USA. Verma, C.V.J. and A.R.G. Rao (eds.) (1998) Proceedings: Workshop
Rosegrant, W. Mark, Ximing Cai, and Sarah A. Cline (2002) on Micro Irrigation and Sprinkler Irrigation Systems, Central
World Water and Food to 2020: Dealing with Scarcity, Board of Irrigation and Power, New Delhi, India.
International Food Policy Research Institute, Washington, VSI (1998) Proceedings of the National Seminar on Irrigation
D.C., USA. Water Management for Sugarcane, Vasantdada Sugar Institute,
Ruet, Joël (2003), Against the Current : restructuring the SEBs, Pune, India.
Manohar Publishers, New Delhi. World Bank (2002) Maharashtra: Reorienting Government to
——(2001) Winners and Losers of the SEB reform, Occasional Facilitate Growth and Reduce Poverty, Vol. I and II, Report
Paper No. 1, CSH, New Delhi, www.csh-delhi.com No. 25053-IN, Poverty Reduction and Economic
Ruet, Joël, V.S. Saravanan, and Marie-Helene Zérah (2002) ‘The Management Unit, South Asia Region.
Water & Sanitation Scenario in Indian Metropolitan Cities: —— (2000) ‘India: Country Framework Report for Private
Resources and Management in Delhi, Kolkata, Chennai, Participation in Infrastructure’, World Bank, Washington
Mumbai, Occasional Paper no. 6, CSH, New Delhi, www.csh- D.C.
—— (1998a) India—Water Resources Management Sector
delhi.com
Review: Report on intersectoral water allocation, planning
Saleth, R. Maria (1996) Water Institutions in India: Economics,
and manage-ment, Volume 1, Main report, Washington D.C.,
Law and Policy, Commonwealth Publishers, New Delhi.
World Bank.
Sanjeev, S. (1997) ‘Management of Urban Water Services in
—— (1998b) ‘Water Resources Management Sector review’, 29
Delhi: A conceptual framework and application, Masters September.
thesis submitted to Macquarie University, Sydney, Australia. WSP (2001) ‘The Buenos Aries Concession—The private sector
Sharma, Vinod K. (2002) Handbook of Environment, Bookwell serving the poor’, Water and Sanitation Programme, South
Publishers, New Delhi. Asia, January.