R264.1 02co 17373
R264.1 02co 17373
R264.1 02co 17373
1 02C0
. , 5 0 YEARS
P OF TOUR
MANAGEMENT*
ERM
L. • ;RC
PO Be* §?• • •'-; ft ' rtD THE HAGUE
; . ;i ' ;• O 30 689 80
! :JX -:yi 70 35 899 64
LO:
CO
DRAM IINAL POSITION REPORT
/) /C
1 INTRODUCTION 1
2.1 INTRODUCTION 6
2.2 FINANCIAL AND ECONOMIC INTERPRETATIONS OF COST RECOVERY 6
2.3 RECOVERING THE ECONOMIC VALUE OF WATER 7
2.4 RECOVERING THE FINANCIAL COST OF WATER 9
2.5 UNDERSTANDING DEMAND H
2.6 CONCLUSION 18
3.1 INTRODUCTION 20
3.2 THE KEY POINTS 20
3.3 THE LITERATURE REVIEW 25
3.4 THE COST RECOVERY SURVEY 35
3.5 THE META ANALYSIS 40
3.6 THE FIELD TRIP TO INDIA 42
3.7 THE FIELD TRIP TO SOUTH AFRICA 50
4.1 INTRODUCTION 57
4.2 WHAT'S IMPORTANT? 57
4.3 EXAMPLES FROM THE FIELD 61
4.4 WHAT DOESN'T WORK? 62
4.5 WHAT DOES WORK? 66
4.6 WHO CAN PROVIDE SUSTAINABLE FINANCING? 69
4.7 STEPS FORWARD 71
4.8 THE CHRONICALLY POOR 77
5 COMMENTS 79
#
EXECUTIVE SUMMARY
Background
This is tlie draft Final Position Report for the Department for International
Development's Knowledge and Research Project R7384, entitled Cost Recovery in
Water and Sanitation Projects.
The project has been managed by Environmental Resources Management (ERM) Ltd.
The key research associate in tlie project was Dr Dominic Moran at the Scottish
Agricultural College, in Edinburgh. Vie project lasted September 1999 - March 2002.
The purpose oftlw project was "to review and interpret cost recovery medwnisms
across a wide range of water and sanitation schemes in order to recommend a portfolio
of best practices for achieving financial sustainability and maximising the potential for
private sector partnerships".
The investigation focused particularly on cost recovery for water and sanitation
schemes in rural and peri-urban areas. Tliis is because it is here where people are
generally the most poor and cost recovery/private sector involvement is presumed to
be most difficult. Geographically, tlie research focused on the experiences and
challenges of South Africa and India. Key research activities included:
• A literature review
• An analysis of data obtained from a survey on cost recovery in WATSAN
• A meta-analysis of willingness to pay papers
• Field visits to projects and policy makers in South Africa and India
Our key findings are presented and practical and strategic recommendations are made
for designing cost recovering WATSAN projects, programmes and policies for poor
people in rural or peri-urban areas. Viey are based upon an assimilation and
comparison of our findings drawn from across our investigations.
Key Findings
Financial Sustainability ,
Most practitioners focus on the financial costs of projects and how to recover tliem.
Few focus on economic costs, or values (opportunity or environmental costs), though
these can be extremely important to society as a whole.
In terms of financial costs, not many water and sanitation projects or programmes in
rural and pen-urban areas recover more titan operation, maintenance and a portion of
their capital costs. This means that long run financial sustainability in the rural and
peri-urban WATSAN sector is rare.
Cross subsidisation doesn't really work and is extremely difficult to reconcile with
first-best cost recovery principles. Cross-subsidisation or "free water" policies can
make it difficult for the service provider to bill or collect their dues from many of their
poorer users, aggravating cashflow and financial sustainability problems. Social
tariffs can also create disincentives for urban -providers to expand their services into
low-income areas, and are difficult to implement in non-networked rural areas.
Furtlwrmore, Donor Agencies themselves are not very good at achieving long run
financial sustainability in their WATSAN investments, especially in rural and peri-
urban areas. They tend to aim to recover operation & maintenance plus a portion of
capital costs. Private sector agencies (IFls and the private sector), however, tend to
aim for operation and maintenance costs plus replacement, interest, capital costs, and
costs offuture expansion. Vie Donor Agency aims have a more public sector focus -
they reflect a desire to meet basic needs, to ensure an equitable use oftlie subsidy and
to break even. The private sector aims reflect a desire to obtain an operating surplus or
a return on the assets. Private sector projects, however, tend to work with slightly
wealthier consumers, mostly in urban areas. Nevertheless, the private sector aims are
financially sustainable; the public sector aims are generally not.
Financial sustainability is important because there is simply not enough public sector
money available to either fund all oftlie water and sanitation projects and programmes
required for poor people in rural and peri-urban areas, or to continue to fund the
existing investments when their assets need replacing. Somebody lias to pay for water
and sanitation projects, in the long run, and this will have to ultimately be their users.
The challenge is how to retain the positive components of the public sector WATSAN
programmes (equity, poverty-focus, participatory etc), whilst making them as good as
the private sector projects at achieving long run financial sustainability.
In fact, a truly demand responsive approach and the willingness to pay (WTP)
estimates it derives from potential users, sits more closely with economic assessments
of value than financial assessments of costs. Within a WTP bid, evidence suggests that
respondents will place a value not only the technical option on offer and its cost to
them, but also the institutional arrangements, the management set-up, the suggested
pricing schedule and other attributes oftlie water and sanitation "product" on offer.
Hence, the water and sanitation technology may stay the same (constrained by
logistical or engineering limits), but ways and frequencies of paying, collecting fees,
managing, owning or using the product can be worked upon, causing people's WTP to
change, depending on their preferences.
What Works
The projects that are financially sustainable seem to have tapped into poor people's
notional demands then found ways to make the project derive a return from its assets
that allows it to finance itself. Viey tend to be locally run, easily replicable, but
dependant on a wider decentralisation policy within which to work. They have mostly
benefited from a "seedcom" grant and parallel community financing mechanism. The
poverty of their participants does not seem to have been a constraint to their success.
Unfriendly under policy environments have.
Importantly, success at the project level will be measured by staying power and local
replication, with minimal resort to grant financing. If people replicate the WATSAN
project, with no external grant funding, they obviously like it, want it and are willing
to pay for it. If tlie project remains, longer run costs are de facto being recovered
At the Programme Level
Think small.
Many smaller schemes should be encouraged that suit their particular users best. This
could be on a milage by village or on a slum-by-slum community basis. Smaller
schemes mean lower exposure for the users to the risk of financial failure. Micro-
networks, or non-networked systems, with up to about 30 households per group seem
to be a feasible size for success.
Think many.
Pepper a neighbourhood, or rural area with several small schemes, each having been
iteratively designed to suit very local needs. Local information exchanges will create
the most effective hybrids.
Think unsystematically.
There will be many different ideas and approaches as to what may work. By taking an
output based approach, it doesn't matter so much what the design of each local project
in the programme is (as technical options are limited, people may choose differences in
the institutional, payment, management options), so long as minimum (technical,
social, environmental) criteria are met and the project delivers the output required.
With a stake in their own project, people will also choose their options carefully to
match their local conditions.
In turn, the Partner Organisation should be evaluated based on how many local
WATSAN projects are created and then sustained, and how many are replicated;
success should not be equated to how much money was disbursed (in fact, the reverse)
or what technologies were used. Innovations in terms of local management, payment
and financial structures should be looked for, successful ideas encouraged and
knowledge transferred between users. Key indicators such as for social development,
environmental and liealth should be monitored and evaluated at key -points in contract
renewal, and the contract refocused or re-tendered if needs be.
The Agency should place the overall WATSAN Programme Management in the liands
of an international private sector company, with the responsibility for purpose level,
output delivery being theirs. The Company will be able to interact with tlie Agency,
the Partner Organisation and the State Partner in terms oftlie provision of training,
technical advice, strengthening of financial sustainability issues, monitoring and
evaluation and other key performance indictors of interest to the Donor Agency. With
overall output achievement resting with the Company, not the Donor Agency, tlie
Agency can replace tlie Company and tlie integrity of tlie programme with tlie
recipient State Government and its population will remain. Disbursements to the
Company could be sequential, and based upon outputs being achieved.
The most prevalent obstacle to achieving cost recovery was found to be political
interference, inappropriate public policy or a lack of political willingness to institute
cost recovery mechanisms.
A clear policy that encourages the sustainable financing of WATSAN services in rural
and peri-urban areas is therefore important. The objectives of the policy should be as
follows.
Agreement by donors and government agencies on the core objectives of this policy is
also essential.
There is a wide range of private sector roles at all levels within the WATSAN sector
for delivering financially sustainable projects, manly as thy are best at doing so. The
most viable private sector participation options at project level for rural and peri-
urban communities occur at the local level. They include the role of grassroots or
community-based organisations and small-scale contractors and local suppliers.
Some of these organisations already run their local WATSAN project in a "private
sector" manner; they employ local people or sub contractors; they manage their water
and sanitation diarges to provide a net surplus on tiieir short run costs; and they
invest the surplus in wider community development issues, or back into the
WATSAN scheme. Several projects in India show how this can be the case. Local level
NGO, or community decision-making forums can play a critical facilitation and
regulatory role in this process. Furthermore, the local private sector can also
contribute to much of the supply chain for inputs to these local projects, striking
competitive sourcing deals with local users.
At the programme level, an interesting role for tlie private sector within country is in
the potential formalisation of relationships between these small-scale private sector
suppliers and a larger WATSAN supplier is in the form of locally established
partnerships. Sector reforms in South Africa demonstrate a range of permutations for
such management arrangements.
A further role for the private sector at programmatic level is to manage the programme
on behalf of the Agency in an output-focused manner, which puts financial
sustainabilihj issues at the heart of its operations, drawing on its experience and skills
of managing WATSAN projects for profit in developing countries. This role may
attract international water utilities, but the role of engineering or consultancy firms
who liave a broader set of "developmental" skills in project and programme
management, may also be appropriate.
Accompanying Guidelines
Anotlier output of this project is the development of some practical guidelines, focused
at the project level. The guidelines will include:
Comments
While we feel that some key steps have been made through this research in the
development of practical recommendations for designing financially sustainable
WATSAN projects for the poor in rural and peri-urban areas, there are a number of
exciting opportunities emerging for further work, whose outputs could feed usefully
into the policy agenda on this issue. These opportunities can be split into more blue
skies research and more action-research orientated exercises.
This is the draft Final Position Report for the Department for International
Development's Knowledge and Research (KaR) Project R7384, entitled Cost
Recovery in Water and Sanitation Projects.
"To review and interpret cost recovery mechanisms across a wide range of water and
sanitation schemes in order to recommend a portfolio of best practices for achieving
financial sustainability and maximising the potential for private sector partnerships".
The guidelines, which form a sister publication to this report, reflect the
second of these aims. They form a separate report.
Thus, finding out how best to design and implement WATSAN services that
are financially sustainable for poor people continues to be a real and practical
challenge facing the international donor community, especially in rural and
peri-urban areas.
For DFID finding out more about cost recovery in WATSAN projects and
programmes is a key issue, for the following reasons.
• Within the context of DFID's overall strategy for aid delivery through
budgetary assistance initiatives that seek to improve the efficiency of
existing state programmes, it is of central importance to provide practical
guidance to partner governments on how to design financially sustainable
WATSAN projects for the poorest.
It is within this context and rationale that this research has been undertaken.
We hope that our findings can feed directly into policy and strategy
development for rural and peri-urban water and sanitation provision in many
of the poorest regions of the world, both for DFID and for other development
agencies and private suppliers.
During the inception phase of this project, four key research challenges were
identified:
Over the two years of the project, key research activities to address these four
challenges have included the following:
Liaison and discussion with, and surveys of, key stakeholders to draw
together and analyse a comprehensive set of case study material on cost
recovery initiatives
Analyses of those key WATSAN initiatives with the potential for long-
term private sector partnerships.
Field visits to projects and policy makers in South Africa and India
• Section 5 presents some brief comments on the directions this work could
be taken next.
2.1 INTRODUCTION
This section provides a definition and context to the term cost recovery.
1
The distinction between these two interpretations has been further complicated by the simultaneous and
related debates about treating water as an economic good and the meaning of a rights based approaches to
WATSAN supply.
"Water has an economic value in all its competing uses and should be recognized as an
economic good. Within this principle, it is vital to recognize first the basic right of all
human beings to have access to clean water and sanitation at an affordable price. Past
failure to recognize the economic value of water led to wasteful and damaging uses of
the resource. Managing water as an economic good is an important way of achieving
efficient and equitable use, and of encouraging conservation and protection of water
resource".
Note again, the use of the term value. The term may seem somewhat
subjective and normative. However, if the sum total of value "created" is
important in decisions to allocate resources to water then this value in
competing uses must be identified and, as far as possible, measured. It is
important to point out that value does not necessarily equate to price, and that
the value of many water uses may be difficult to detect or substantiate with
reference to existing markets or what households can spend on water.
Instead, value might be considered as something that goes beyond the
household's immediate income. It can encompass intangibles, such as those
deriving from things such as security, reliability, through to spiritual belief
and traditional rights. This point will become important when we try to
assign values to water resources. The discussion of water valuation leads to
the distinction between willingness to pay and ability to pay.
On the other hand, however, were water allocated to domestic use over
agricultural use, and were the agricultural value of water higher than
domestic willingness to pay, we would say that each unit allocated to
domestic use incurs an opportunity cost. That is, the loss in value incurred as
a result of allocating the last unit to one use rather than another. Opportunity
cost ultimately dictates whether a society allocated its water to its highest
value uses. However, the implementation of opportunity cost pricing is
difficult in practice. It is useful, only to mention it in the context of national
resource allocation.
2
Note that these objectives are related- A technically correct economic appraisal of a water scheme will use economic
values for the main output (water). A financial appraisal for a village scheme on the other hand may be based on a tariff
that does not necessarily have to relate to the correct economic price.
Water does not always have a monetary price and, if it does, the prevailing
price is not the same as its economic value. The absence of value
information aggravates the process of assessing the return to competing
uses and, by extension, to improving our understanding of the economic
viability of water schemes. Much energy has therefore gone into the
estimation of water value using willingness to pay methods. This agenda
has dovetailed conveniently with a related approach to water supply
planning that emphasizes the need to measure demand and for WATSAN
interventions to be demand responsive.
The task therefore is to consider the steps necessary to move towards both
the value of water supply (for appraisal) and the correct pricing of water
for charging consistent with its status as an economic good. This requires
us to consider the costs of supplying water including methods to assess
the value of water in competing uses. Once we have seen how the costs of
supplying water can be added up, we are then in a position to consider the
likelihood of cost recovery of some or all of the costs and the payment
mechanisms and devices that help to optimise recovery. Note that while
accurate economic pricing represents what economists call a theoretical
first best, pricing in practice is likely to be an approximation.
So how much of the past, current and future financial supply costs should be
recovered? As the following list makes clear, there are several levels of
recovery that can be targets for project viability. The extent of costs incurred
by the WATSAN initiative, and therefore to be recovered depends largely on
whether
It is helpful to benchmark cost categories while noting that much of the cost
recovery debate assumes a level of accountancy understanding that may be
beyond some communities or local WATSAN implementers.
• A zero cost tariff: free un-metered water subsidised through other means
with ad hoc recovery.
<5>This is commonly defined as the investor's cost of capital; the rate of return they must seek in order
to exceed the notional interest rates payable in turn to their own creditors.
10
• An increase in some existing arbitrary tariff modestly in line with
inflation.
• Operation and maintenance costs plus the depreciation costs of the assets:
this can be seen as a 'replacement fund' or performance fund to finance
future capital investment.
Note that the concept of the long run marginal cost (or its short hand
approximation of Average Incremental Cost) sets the tariff decision in the
context of the whole project life. This means that the calculation has to
undertake some form of time discounting so that current prices can account
for future costs. Factoring the future into current payment decisions is quite
difficult to conceptualise.4
Of course, supply costs will vary with the multiplicity of technical options,
which in turn will depend on the size of the community to be served. Cost
recovery can be defined to various levels of sophistication.
4
Calculation of the Long Run Marginal Cost is controversial in water economics. The measure is often
substituted by a more convenient approximation know as the Average Incremental Cost. This defines the
unit cost of water service $/m3 = NPV of annual costs to implement project /change in quantity water
consumption due to project (m3). NPV is the Net Present Value.
11
Once the categories of financial (and economic) costs that could be recovered
are understood, the questions become
The last two of these stages constitute our understanding of what has come to
be known as the demand responsive approach. The approach contrasts with
traditional supply focussed approaches to WATSAN, which, having found
some technical solutions, have simply implemented a project.
12
expensive to maintain. Or, in some cases failure may be due to a local culture
of non-payment, possibly exacerbated by the refusal to accept an
inappropriate technology. The problem of supply infrastructure not
matching the preferences of the host community has been a common
conclusion on the shortcomings of the supply side approach to WATSAN in
the past.
In the last decade or so, however, the engineering literature has recognised the
limitations in a supply-orientated approach to WATSAN services, and has
emphasized the role of more appropriate technology. More recently still, there
has been an emphasis on matching technology characteristics with community
preferences for supply. "Assessing demand" has then consisted of
undertaking more sympathetic interactions with the user community to
explain the technical options on offer and which they may prefer.
However, these approaches and their supporting literature are still rather poor
in their understanding of the importance of demand in the wider project
appraisal process. In other words, they have tended to de-emphasise the link
between project design (meaning more than simply the technical solutions on
offer) and user preferences. Since the mid 1990's, however, this discrepancy
has been addressed by the demand responsive approach (DRA).
The more recent emphasis on DRA suggests the need to match supply options
more closely with consumer preferences, including preferences not only about
the specific technology, but also about its management and how much and
how to pay. DRA is important, since in many ways it actually precedes the
consideration of costs in the sense that costs will largely be driven by the
supply option that most closely meets user preferences.
A central theme is that the process is likely to be iterative in the way options
are presented and modified between host communities and service
authorities. This process is also likely to take into account a wider number of
options and characteristics about the project than just technology, including,
the supply type payment schedule, water volume allowances, the price and
local management and institutional arrangements.
13
The importance of the price here is in informing the financial viability
question and the chances of cost recovery early in the design and project cycle
process. Note that, however, elicitation of household willingness to pay
provides guidance as to the overall welfare5 derived from the project and all
its component parts, and by extension the likely tariff that can be set.
Importantly, it is taken for granted that the service provider has made an
accurate estimate of the capacity and future water demand and that the
amounts used in a contingent valuation survey match the requisite costs. In
other words one is not looking to elicit a value or willingness to pay for an
inappropriately sized scheme that imposes a disproportionate cost burden, as
a result of a poor water demand forecast. Demand forecasting is therefore an
important stage of the DRA process.
"the degree to which consumers have choices over their level of service with the idea being that
consumer preferences, including willingness to pay, should determine the level of service
(LOS) to be provided and charges set accordingly to recover actual costs (Garn, 1998)
5
In economics jargon this measure of stated welfare provides the total value of a resource including consumer surplus
14
The economic concept of demand lends much to the understanding of what
has come to be termed DRA in WATSAN circles. While demand has a very
specific interpretation by economists, it is clear the DRA is not universally
built on this interpretation6.
If one can observe a downward sloping demand curve (relating WTP and
quantity) for a good, then at once there is some understanding of the
magnitudes of value associated with larger quantities of the good. Accurate
demand information allows an understanding of how much of something to
supply and allows price to be determined7. Thus, the supply can be got just
right if one knows what the demand is. Equally, in a fully functioning market,
an efficient supply will be forthcoming if costs are fully recovered - ie if
people face the economically correct price8 for the good.
6
As an example of the "alternative" interpretations of demand, see Parry-Jones (1999) which shows an
engineering interpretation of demand that is, in fact, very close to a definition of supply,
7
Price can be set somewhere between the cost of supply and the WTP.
8
A price that fully represents the scarcity value, or opportunity cost, of the good
15
and selecting one of these options with communities, can be termed, broadly,
as taking a demand responsive approach. Concerns can be raised about the
affordability or suitability of technologies to local conditions, and cost
recovery rates still appear to hover, at best, around the operation and
maintenance level. There are many WATSAN projects like this, and they are
mostly discussed as involving DRA to a lesser or greater degree (depending
on the number of focus groups held, the amount of options offered etc). It
could be argued, however, that few of these initiatives still actually tackle the
price/ quantity balance properly.
The challenge remains: how can just the right balance of a WATSAN product
and its price be found to design the right intervention?
Injecting the economic interpretation of demand into DRA for new WATSAN
schemes or upgrades, amounts effectively to the design of a "new good" that
the community either accepts or rejects. This is essentially a marketing
approach to new product design. Rejection of a scheme is not tantamount to
saying that the technology is not appropriate. Instead, it should be taken as a
signal that more information on one or more attributes (nature of supply,
price, tariff structure etc), needs to be elicited from the target community in
order to get the design bundle right.
It is well known that peoples' WTP can vary, depending on, for example, the
nature of the question asked, or the bid vehicle that is used in the survey
about paying for WATSAN services. Indeed, some have highlighted this
variance as a sign of weakness in stated preference studies, and as support for
reverting to more "robust" indicators of affordability, such as the "5%" rule.
16
for WATSAN services like this?/ do you prefer this one or this one and how
much would you pay?) seeks to elicit only their perceived affordability of the
option, or just an expression of their WTP. By not discussing the
price/product balance explicitly, the approach focuses on the immediacy of
cash in the household to pay for/ to afford the discrete service offered.9
Yet many studies have shown that in rural /peri-urban areas, the marginal
utility of money is high (people value hard cash highly), that incomes tend to
be low or insecure, and that often people pursue a wide range of livelihoods
whereby their wealth is tied up in a number of assets, and not just in
disposable income. This complexity of livelihoods means that households
frequently have diverse strategies for accessing income to satisfy basic needs
and that observed affordability (how would you spend the cash you have)
does not provide a good indication of preferences over appropriate service
delivery.
Further, many studies have also shown that water is an inelastic good in
relation to income or price (ie demand - or WTP- for water doesn't really
change significantly in relation to its price or people's income) and that
innovative ways of paying for water such as through savings groups,
revolving funds etc can be useful - the way the payment vehicle is designed
can also influence how much people will pay (see section 3 to this report).
These studies seem to be contradictory. Rural /peri urban people are poor and
value their cash greatly. They have to pursue many strategies to survive. And
yet they will still pay a lot for water in relation to their income, or will be
influenced to pay by project designs, which may be contrary to what an
affordability study may have suggested.
A key area of the demand literature, which may help explain this
contradiction, is an identification of a difference between notional and
effective demand. Sen (1981) illustrates effective demand by considering a
shop selling food in a famine area, where many people cannot afford to buy
the food. The need for food is great, but only a few people can buy it,
therefore, effective demand for food is small. Pearce (1981) defines effective
demand as the "aggregate demand for goods and services which is backed up with
the resources to pay for them distinguished from 'notional demand' which refers
to a desire for goods and services"
Credit or other financial assistance and the type of payment structure on offer
in a WATSAN product, can help to translate a person's notional into effective
demand, rendering affordability or rule of thumb analyses less useful. These
9
The analogue may be to ask someone in the UK how much they would be willing to pay per month for
their child's education at a specified, good quality, fee paying school. I can't afford it even though I would
like to, may well be the common response, unless you offer me some form of savings scheme or other
specific credit facility to help pay for it. Anyhow, the preference may not be for that particular school, but
instead for another (and even then boarding for some, day school for others etc etc).
17
differences may be of central importance to the WATSAN debate on financial
sustainability.
Hence in theory, the effective WTP of the consumer can be fixed, but if they
have access to other assets (labour, livestock, credit, savings schemes) and if
the overall design of the WATSAN "product" particularly suits them, then
their notional demand may be quite high. If the right mechanism can be
designed and implemented to translate their notional into effective demand
for WATSAN services therefore, and it is part of the WATSAN product on
offer, then they may be willing and able to pay beyond their observed and
immediate income constraint.
2.6 CONCLUSION
18
can be recovered by a WATSAN project or programme can be devised. Most
current initiatives recover only a few of these financial costs, and hence most
WATSAN projects are not financially, let alone economically, sustainable.
19
COST RECOVERY - WHAT SEEMS TO BE IMPORTANT?
3.1 INTRODUCTION
From across the range of our research investigations, the main recurring
findings are as follows.
Most practitioners and policy makers don't equate a cost recovery objective I
WATSAN service delivery to either the issue of financial sustainability, in the
long run, or to economic issues, such as opportunity or environmental costs.
3.2.2 On Demand
20
drier areas. The way they are asked to pay for it, however, affects how
much they are willing to pay.
• At the moment, less accurate, but more feasible methods for assessing
demand, such as focus groups, take prevalence in the field, over more
intensive approaches such as participatory rural appraisal or contingent
valuation studies. Affordability studies tend to dominate the demand
assessment domain.
• The use of credit or other methods of financial assistance and the type of
payment structure on offer are also important variables that influence
demand. They can help poor people to "buy into" a WATSAN project and
pay for it.
• The use of long run cost pricing for water plus a cash subsidy or a form of
credit given to the user, could be an alternative to subsidising water tariffs
in peri-urban area. It could also be a way of helping to pay back a loan for
a WATSAN connection or investment.
21
3.2.4 On Charges and tariffs
• Designing and charging for a WATSAN scheme on the basis of the 3-5%
affordability rule of thumb is misleading, though in practice many
practitioners use it.
• For connections, findings suggest that the poor should pay for their
connection charge, either via a flat rate fee or a fee that relates to their
income.
• An efficient tariff policy should be one that reflects the marginal economic
costs of supply.
• In practice, users are most often charged in relation to the volume of water
that they use. Lifeline tariffs for the lowest income households and
increasing block tariffs were the most popular charging structures. Free
blocks in the tariff are the least popular.
• For cost recovery, public sector agencies (bilateral agencies and NGOs)
tend to aim for operation & maintenance plus replacement costs. Private
sector agencies (development banks, IFIs, the private sector) tend to aim
for operation and maintenance costs plus replacement, interest, capital
costs, and costs of future expansion. The public sector aims reflect a desire
to meet basic needs, ensure an equitable use of the subsidy and to break
even. The private sector aims reflect a desire to obtain an operating
surplus or a return on the assets. The private sector aims are financially
sustainable; the public sector aims are not.
• In practice the most prevalent obstacle to achieving cost recovery was said
to be political interference, inappropriate public policy or a lack of political
willingness to institute cost recovery mechanisms.
• Often the only sustainable manner of managing rural water supply and
sanitation is for a community management model to take charge. However,
this can be difficult in cases where the legal and regulatory frameworks
cannot formalise the status of such organisations
22
• Peoples' willingness to pay for investment costs has been found to increase
dramatically when communities, rather than government agencies, have
control over how funds are spent.
• The most viable private sector participation options for rural and peri-
urban communities are local. They are seen as the following:
o Community based organisations.
o Small-scale contractors and local suppliers..
• The projects that are financially sustainable (and therefore cost recovering)
seem to have tapped into people's notional demands then found ways to
make the project derive a return from its assets that allows it to finance
itself. They tend to be locally run, easily replicable, but dependant on a
wider decentralisation policy within which to work. They have mostly
benefited from a "seedcorn" grant and parallel financing mechanisms. The
poverty of their participants does not seem to have been a constraint to
their success. Unfriendly wider policy environments have.
• In India large donor led projects with a project management unit at state
level, a budget of US$ tens of millions and an NGO implementing agency
to assist, seem to be less financially sustainable than smaller, more locally
23
focused community run WATSAN projects, with budgets of US$
hundreds or thousands.
o The larger projects aim for ballparks of 10-15% capital cost
payments from users, with upfront cash contributions varying
from 1% to 10%; 100% operation and maintenance cost
recovery; and about 60% cost sharing for individual latrines.
The smaller projects achieve cost recovery rates of 50 - 100%
capital costs, 100% operation and maintenance, replacement
costs and operate their local networks at a surplus, using extra
funds for wider investments. Both are working with the poor.
o The larger projects tend to follow a financial framework for
implementation, which doesn't necessarily change at the local
level. Some beneficiary groups could pay more towards the
capital cost of the schemes than the 15% they are currently
asked for. Other user groups do not necessarily easily or
willingly replicate the project without another grant arriving to
help them, given the low level of beneficiary contribution
required to start with.
o The smaller projects, however, tend to look for organic
replication following a first "seedcorn" investment. They use
and support local suppliers, and develop local rules. A local
body becomes a facilitator. The local owners tend to use the
asset to generate a surplus (like the private sector) and use local
employment to run it. These projects have tended to develop
out of crisis, or in spite of the wider WATSAN policy
environment. Parallel financing mechanisms and support
networks also seem to be important.
o A much more decentralised and community-focused model of
implementation, especially since constitutional amendments in
the early 1990's has created the policy framework within which
these smaller projects can take place.
24
Nevertheless, the new model structures for WATSAN service delivery in
South Africa at work in some locations may yet give rise to innovative
contracting and collection mechanism, particularly in cases where
community-based organisations are taking the role of service providers
and where they have financial autonomy (as opposed to simply paying
revenues to the municipal service authority).
Other NGOs in South Africa such as the Mvula Trust, claim that many
stand alone rural WATSAN projects do exist (kick-started through an
NGO facilitation process), with local level community management and
good rates of cost recovery. Interestingly, it seems that these projects will
probably not avail themselves of the free water policy, as the NGOs who
coordinate them believe that it is not consistent with their schemes
working.
3.3.1 Introduction
• A review of work since the mid 1990's, particularly from the Work Bank,
on issues relating to investments in infrastructure services, regulatory
polices, private sector participation and their combined effect on the poor,
especially in peri-urban and rural areas. These sorts of financing,
investment and regulatory policy (management model) themes are now
often at the forefront of the WATSAN debate.
There has also been a review of some of the output and follow up to the World
Water Summit in the Hague, September 2000, which heralded the beginning
of a return to a discussion about social entitlements and subsidy in the
WATSAN provision debate. Other issues, such as water supply within a
livelihoods context, and the emergence of the concept of output-based aid,
and its relevance to WATSAN delivery in rural areas, have also emerged
during the research period.
Each of these issues is reflected in the literature review, which can be found in
Annex B to this report. It should be stressed, however, that we do not consider
our literature review to be comprehensive or the last word, but simply to be
25
viewed as a useful text that highlights the main directions where the literature
suggests the cost recovery debate in WATSAN projects should be heading.
Four core issues emerged from the range of interpretations on cost recovery
we looked at.
3.3.3 DRA
In terms of assessing demand, "DRA", has been a key concept. The term first
appeared subsequent to the Dublin Declarations, in 1998, in a World Bank
note entitled "Managing Water as an Economic Good".
In this note, the concept of a Demand Responsive approach (DRA) was put
forward. DRA was an approach to WATSAN that attempted to respond to
consumer demands and which aimed at making projects more sustainable
than supply-driven approaches. The following were listed as key
characteristics of DRA:
26
Another important aspect of the DRA approach was the suggestion that
communities compete for funds for WATSAN projects, in order to help the
agency decide whether to provide support to a particular community and
what type of system and level of service to provide. This was a controversial
idea.
The research conducted by The World Bank between 1987 and 1990 was used
to back up the DRA model.10 This study found that household income,
though often important, was not the overriding determinant of demand for
improved WATSAN services. The following three factors (in no order of
importance) were found to be significant:
For many NGOs, however, the DRA debate was fundamentally flawed - how
could poor communities compete and pay for WATSAN projects? The DRA
model also became, in their eyes, contiguous with the problems of the
contingent valuation survey - how could you realistically ask a poor
community what they would be willing to pay for a WATSAN project? - and
with the Dublin Declarations about water as an economic good in general.
Partly as a result of some of these problems with the DRA model, there was a
shift away from the DRA agenda in the statements from The Hague, toward a
more rights based approach for WATSAN delivery.
A log linear functional form is often assumed for the demand curve for water.
This implies that the price elasticity of demand for water is a constant at all
levels of water use within households. Although this is at best a crude
representation of household water demand behaviour, it allows economists to
provide a reasonable approximation of people's actual demand behaviour.
This is an important assumption, but it may be a misleading
oversimplification, as many other studies indicate that income, though
important, is not the overriding determinant of demand for WATSAN
services.
10
Demand for Water in Rural Areas: Determinants and Policy Implications. Briscoe et al, World Bank Research
Observer, January 1993.
27
Another important issues from the literature is the identification of a
difference between notional and effective demand. Understanding this
difference helps to throw light on the DRA debate, and may help to show how
improving the understanding and use on demand assessments may help to
design more sustainable WATS AN projects and programmes.
Hence, if a demand assessment can be used to find out about both peoples'
notional demands for WATSAN (hypothetically, what sort of WATSAN
services would you like?), and also how they would like to translate their
notional into effective demand (and how would you best like to pay for
them?), then a key obstacle to the use of DRA is resolved (notwithstanding the
issue of competition for funds).
The use of credit or other methods of financial assistance and the type of
payment structure on offer can help to translate notional into effective
demand, rendering affordability analyses less useful. The constraint of cash
becomes less of a problem, if parallel to identifying (notional) demands,
structures for the preferred financing mechanisms that allow these demands
to be realised, are also discussed.
28
• Contingent valuation studies (for informal strategic decisions on level of
service; cost recovery policy and large investment programmes).
• The 'Benefits transfer' approach, under which results in one location are
used to estimate benefits in a 'similar' location is not recommended.
• How and where in the project cycle demand assessment techniques should
be used.
3.3.6 Affordability
A range of references suggests that the 3-5% rule of thumb is misleading. For
example, the World Bank, the UNDP- World Bank Water an Sanitation
Programme, the Asian Development Bank and DFID recognise that the
assumption that people would be willing to pay 3 to 5 per cent of their income
on water is arbitrary. They generally recommend that affordability studies be
are put more in the context of how consumers currently cope and what they
actually pay for water.
In general, the literature shows that willingness to pay (WTP) has proven to be
a more successful indicator of demand than ability to pay. Some studies have
shown higher WTP for poorer households, compared to a standard of around
2 to 3% spent on water services among high-income households. These results
indicate that rural communities may prefer service levels above the minimum
often prescribed and that rural or peri-urban water and sanitation projects
may not actually require extensive government subsidies.
29
DFID's 1998 WATSAN Guidance Manual is slightly contradictory in relation
to its encouragement, on the one hand, to assess demand and its approach, on
the other, to affordability. Although the manual agrees that affordability
"rules of thumb" have been shown to be a poor guide to WTP for WATSAN
service improvements, the guidelines do suggest that affordability should be
interpreted on a case-by-case basis. Hence, the identification of notional
demand for WATSAN services is potentially subsumed by the
recommendation to explore cheaper, more affordable (and therefore supply
orientated) options for the community.
This seems to be a centrally important point that potentially links the issue of
capturing notional demand to the design of a reasonable standard of
WATSAN service supply that can become financially sustainable.
DFID's position on this area is relatively weak however. The 1998 Guidance
Manual suggests that demand assessment studies can help in the design of
payment mechanisms that are appropriate for poor people. They can identify,
for example, people's preferences for weekly as against monthly payments, or
for credit arrangements to spread over time the capital casts of connection
fees. However, it doesn't make explicit the link between design a WATSAN
project and designing a complementary financing mechanism alongside.
The literature agrees that an efficient tariff policy should be one that reflects
the marginal economic costs of supply. Since most water is consumed by a
relatively small number of consumer groups (large residential, commercial
and industrial users) a marginal cost based price for all consumption other
than required for basic health purposes would generally be an efficient,
equitable and financially viable policy.
30
The literature agrees, however, that tariffs can also be used as instruments of
social policy, such as for the redistribution of income based on equity
concerns. A commonly observed approach is to use a tariff schedule that
consists of a low subsidised "lifeline" rate for the first 6-8 m3 used per month,
and a higher rate for all additional consumption. The lifeline rationale
underlies the Free Water Policy introduced in South Africa. However, it is
important to understand how such policies can compromise cost recovery
when the 'free' block is at or near the average household use for smaller
schemes.
3.3.8 Subsidies
In peri-urban areas, the review found that cross-subsidisation policies can fail
to provide incentives to the service provider to bill or collect from many poor
users, thus aggravating cash flow problems. In fact, social tariffs may create
disincentives for these providers to expand their services to low-income
areas. Unrealistically high rates to industrial and commercial users can force
them to look for alternative supplies, thus reducing market share and, in the
process, exacerbating the cross subsidy problem.
DFID's approach to subsidies for water supply schemes is that they should be
justified on income redistribution grounds, and not on direct health benefits.
DFID suggests that subsidies for sanitation may be needed to correct for
'market failures', which arise because inherent demand does not lead to the
level of investment in those sanitation services which would be most efficient
for the economy and society.
Policy, legal and regulatory frameworks that mandate and support both the
decentralisation of, and increased participation in, WATSAN service delivery
by the private sector and community groups, are seen to be vital for the
success of financially sustainable projects.
31
There is an increasing realisation that governments must make the transition
from being a 'provider' of services to becoming a 'facilitator'. National
agencies should be responsible for finance, long term planning, standard
setting and technical assistance. Private sector participation can assist in the
provision of demand driven services, but specific attention should be given to
the needs of demand driven services for poor people from the outset of
contract and regulatory framework design. If this does not happen, then there
is a risk that the resulting regime actually creates further barriers to the
provision of such services.
For rural areas, many water sector studies recommend that the only
sustainable manner of managing water supply and sanitation is for
community management models to manage their own water and sanitation
supply systems. However, this can be difficult in cases where the legal and
regulatory frameworks have not formalised the status of such organisations.
This can create very practical difficulties for such organisations, especially if
they are tasked with collection of tariffs and financing of maintenance,
without a clear legal entity allowing them, for example, to open a bank
account.
The particular management model adopted for water and sanitation projects
seems to impact significantly on the level of demand. Communities may
distrust a national government's capacity to manage the operation and
maintenance of water and sanitation projects efficiently, and consequently
their demand for such projects decreases. Willingness to pay for investment
costs has been found to increase dramatically when communities, rather than
government agencies, have control over how funds are spent. Here again,
however, community cohesion around a community-based service authority
requires a degree of transparency in relation to funds management. Distrust
and non-compliance with otherwise successful projects can be minimised by
the addition of training support to community managers.
The review found that the main types of support needed by communities to
help sustain their own WATSAN project management services are:
The literature is generally agreed that public utilities do not have the financial
or political autonomy to set tariffs at levels that recover costs. Private sector
participation in WATSAN projects and programmes is therefore seen as the
most efficient way of ensuring financially sustainable pricing policies.
32
Current research on how regulatory and legal frameworks can be formed so
that large-scale private providers of water and sanitation services can
effectively deliver water services to the poor, whilst ensuring that the sector's
financial viability is assured, points to the following key issues:-
• The needs of the poor should be central to any private sector reform
process.
• The assumption that poor customers are high risk, low return customers
should be challenged, which may require a significant information
gathering exercise.
• The distance between the provider and poor customers should be reduced
Many commentators feel that small-scale private sector (or currently informal)
inputs can play a role in resolving some of these issues. If these operators are
responding to expressed and identified demand, then in doing so, cost
recovery will be de facto. For example, recent studies suggest that
entrepreneurs in water and sanitation, responding to local conditions and
competing for market niches, can offer a wide and flexible range of water
supply options - residential re-sales, for example.
In general, the most viable private sector participation options for rural and
peri-urban communities are seen as the following:
• Small scale contractors and suppliers. These agents can help which develop
private enterprise in the water sector.
• Water and sanitation related NGOS. These organisations can provide social
development assistance and the small-scale management contracting of
systems. With decentralisation of service authority, the former role is
likely to be more important than the latter.
However, it may be difficult to combine the short time scales within which the
private sector operator may be required to work, with the longer time-scales
that demand led approaches take. One solution is to formalise the
relationships between these small-scale suppliers and the main supplier in the
form of partnership, a solution that the World Bank Business Partners for
Development has been investigating. In practice, partnerships have shown to
offer significant potential, but a range of problems are still to be resolved if
33
such partnerships are to meet their potential. Instead, locally established
partnerships might offer alternatives.
It is suggested that output based aid approaches can provide clearer thinking
about the use of the subsidy, perhaps directing the funds to the one-time costs
of service connection - typically the main impediment to expanded access to
services - rather than the ongoing costs of consumption. This may be
particularly pertinent to WATSAN service delivery, cost recovery, and how
donor agencies implement their projects, especially as small-scale local
entrepreneurs, community groups, or local NGOs, could be important
suppliers.
34
3.4 THE COST RECOVERY SURVEY
A cost recovery survey was designed by this project11. Water Aid and DFID
helped to review it. The survey was emailed during August- October 2000 to
722 actors in the water and sanitation sector globally, and also disseminated
via several WATS AN e-discussion groups.
Over 100 responses were received as well as accompanying notes and papers.
55 survey forms were completed in enough numeric detail to be inputted to a
database, allowing a comparative analysis to take place.
The Development Agencies and NGOs who replied were more involved in
WATSAN projects with households in rural /peri-urban areas at the bottom
of the income range (less than US$200 pp pa) when compared with those replies
from Development Banks and the Private Sector. This perhaps indicates
evidence for the "cherry picking" tendency, whereby private finance in
WATSAN tends towards the lower risk urban consumer.
Also of interest was the fact that Development Agencies tended to incorporate
WATSAN as part of a wider, or more integrated, development project more so
than the NGOs, Development Banks and the Private Sector who replied.
u
A copy of the survey is presented in Annex C.
35
3.4.2 Assessing Affordability
Nearly a third of those who responded (across all types of group) still use the
3-5% rule of thumb for assessing affordability, commonly stating that as an
initial tool it can be useful (the rule of thumb indicating that a household can
afford a WATAN service if it costs between 3 and 5% of their income).
36
(notional) demand of, the consumer for the service, sits above concerns about
affbrdability.
I Accurate
I Feasible
/ / /
*
It seems that less accurate, but more feasible methods for assessing demand,
such as focus groups, take prevalence in the field. More technical or resource
intensive types of demand assessment studies such as Contingent Valuation
are used less often. Responses to the use of PRA were interesting. Though it is
considered accurate by many respondents, its feasibility, though strong, is less
than that of focus groups. Again, perhaps time and resource constraints are an
issue here.
Costs
Capital cost recovery was less prevalent, but some private sector organisations
and NGOs who replied also included the recovery of capital costs, in their
definition of cost recovery, and said they used this to judge the financial
sustainability of a project (through a measure of recipient "buy in").
However, it was clear that few WATSAN actors outside of the private sector
saw the recouping of long run costs as the only yardstick of cost recovery,
Inherent in most thinking was the necessity of subsidy at some point.
Charges
Respondents felt that people should pay for their water connections, even
poor people in rural areas. Free connections to water services were not
considered to be the best form of sustainable water and sanitation provision.
In fact, 42% of respondents felt that users should pay equal amounts for their
water connections, while 37% considered a connection charge, which reflected
the user's level of income, was more appropriate.
For both water and sanitation provision, once the connection charge is
established, the majority of organisations who replied believed that users
should be charged in relation to the volume of water that they use. Lifeline
tariffs for the lowest income households and increasing block tariffs were
considered to be the most useful charging structures. Free blocks in the tariff
were the least popular option.
Although many organisations that replied have the intention to make their
WATSAN projects financially sustainable through adequate cost recovery
(according to their particular definition of cost recovery), there are often
38
factors that were said to complicate this process. The most prevalent of these
tend to be political factors, such as inappropriate public policy or a lack of
political willingness to institute cost recovery mechanisms. For sanitation,
failure of other agencies to recover costs, are also importantTabZe 3.1 ranks the
complications to cost recovery which respondents chose.
Table 3.1 Complications to Cost Recovery in WATSAN Projects (Ranked In order of the
issues most often stated by organisations working on WATSAN projects.)
Water Sanitation
1 Political Interference 1 Political Interference
2 Low/ Variable Incomes 2 Insufficient Willingness to Pay
3 Distrust of Cost Collection System 3 Low/ Variable Incomes
4 Insufficient Willingness to Pay 4 Lack of Management Transparency
5 Lack of Management Transparency 5 Distrust of Cost Collection Systems
6 Inappropriate Project Design 6 Failure of Other Agencies to Cost Recover
7 Failure of Other Agencies to Recover Cost 7 Inappropriate Project Design
8 Expense of Project 8 Cultural/ Religious Reasons
9 Social Exclusion Issues 9 Social Exclusion Issues
10 Land Tenure Issues 10 Land Tenure Issues
11 Cultural/ Religious Issues 11 Expense of Project
12 Flux of Population Size 12 Flux of Population Size
Finally, WATSAN practitioners were asked simply what tended to work best
to ensure cost recovery and financial sustainability among poor WATSAN
users. The most common answers tended to follow the following themes:
39
3.5 THE META ANALYSIS
The main purpose of the meta-analysis was to assess the relevance of different
variables in determining demand for water services; a secondary purpose was
to evaluate the potential of meta-analysis as a research tool to inform policy-
making and water project planning through the use of benefit transfer.
A more detailed account of the findings is given in Annex D to this report, but
the main findings are presented below.
• At low income levels, people pay to meet basic needs and will pay relatively
high amounts for this: The meta analysis found that someone with a zero
income will pay US$0.5 (or go into debt) for each extra unit of consumption or
level of satisfaction associated with an improved quality of water.
Water demand becomes more elastic as income increases and t)w use of water extends
beyond that requiredfor'basic needs'
• High-level income groups may be using water for less essential purposes, and
consumption above their level deemed 'adequate' for a reasonable standard of
living is therefore more likely to be more sensitive to price.
(12) Meta analysis is a means of providing a statistical review of existing studies. It permits a systematic investigation of
the magnitude and direction that selected variables across a range of studies may have on demand. A f u ^ o , , w a s
derived in order to find a generic model that can explain the relationship between WTP and key variables and that can
assess the likely WTP levels needed for cost recovery.
40
to base cost-recovery factors on general assumptions about what people can
afford to pay.
The most significant factor influencing WTP is the location of the study site
• Households in urban areas have a higher WTP than those in rural areas.
• Households in arid climates may be willing to pay more for their water
supplies because of the greater amount of time needed to collect water. Or it
may be that households in low rainfall regions will already have invested
large amounts of money in storage tanks and may have evolved sophisticated
coping strategies and may therefore be unwilling to allocate more income
towards piped services
The type of charging system is an important determinant of the median WTP for
connection
• Those asked to pay upfront are willing to pay less. This may be explained by
limited liquidity and a high marginal utility attached to the cash they do have.
• Households are prepared to pay more for a flat rate than volumetric rate. This
may be because they do not understand how the volumetric system works or
if they are opposed to paying the additional capital costs associated with the
installation of an effective metering and charging system.
3.5.2 Implications
Poorer people see water as a relatively income inelastic good and are willing
to pay more for it, especially in drier areas. The way they are asked to pay for
it, also affects how much they are willing to pay.
Given the importance of variables other than income, there is a also a need for
local investigations and understandings of what people can actually afford to
pay for their water services and what determines their willingness to pay, in
WATSAN project design. Affordability rules of thumb are not accurate and
could be misleading.
41
As it is impracticable to conduct detailed micro-level surveys for the purposes
of evaluating the costs and benefits of every WATSAN project or programme,
there could be a strong argument for the development of methods such as this,
that allow for the transferability of research findings within key regions.
3.6.1 Introduction
In November 2001 as part of this KAR project, ERM met with key WATSAN
stakeholders in India to discuss the research themes of the project and the
issue of cost recovery in water and sanitation projects, more generally, in
India. A wide range of stakeholders was met, from both a policy and project
perspective. These included representatives from the World Bank Water and
Sanitation Programme, South Asia, DFID India and Water Aid India, as well
as community representatives from projects in Kerela and Tamil Nadu.
Project visits were also made to DFID, World Bank, Water Aid and
community run projects in Andhra Pradesh, Kerela and Tamil Nadu.
There is a long and complex policy history to rural water supply and
sanitation in India, and an attempt has been made to summarise the key
policy and institutional issues in Annex E. In short, there has been a (arguably
unsuccessful) history of State level provision of WATSAN services since
independence, backed up with extra Federal funding for rural areas since the
1970's. This is now giving way now to a much more decentralised and
community-focused model of implementation, especially since constitutional
amendments in the early 1990's created the policy framework within which
decentralisation could take place. Cost recovery, partly of capital costs and
mostly of operation and maintenance costs, now plays a significant part in this
decentralisation process. The private sector, however, is less prevalent as a
service deliverer, especially in rural areas.
The UNDP/ World Bank Water and Sanitation Programme (WSP) South Asia
have usefully encapsulated the payment and cost recovery debate on
WATSAN in India, which remains problematic.
42
The WSP suggest that, politically, water supply has remained a state
responsibility since independence and that political promises on state
supplied WATSAN centre on the following four conceptual points:
• The poor cannot afford safe water and sanitation and should not have to
pay for it.
• Public subsidies are provided to help the poor pay for water and
sanitation.
• The Government of India can solve the problem of water supply by
running water and sanitation programmes itself.
• The Government of India can raise the financial resources needed for
water supply and sanitation.
However, the WSP suggest that the reality is more akin to the following:
• The poor in India actually do pay for water supply and sanitation, often
far more than their fare share.
• The subsidies for water supply and sanitation benefit mainly those who
are not poor.
• Public provision on water supply is inefficient and ineffective.
• The investment requirements for water supply and sanitation in India are
far too great for the Government of India to afford.
Hence the situation in India is of great pertinence to this research study, both
in terms of the size of the problem and because the key issues identified by the
WSP in India reflect some of the findings of this study, namely:
• The key issue that seems to makes cost recovery of WATSAN investments
less straightforward was found to be political interference. Below are
summaries of the six case study projects examined.
The Swajal Rural Water Supply and Sanitation Project is a US$63 million
project covering about 1,200 villages in 19 districts in the Hill and
Bundelkhand regions of Uttar Pradesh. Central to the project's design have
been two major polices:
• Partial capital cost recovery (10%, with upfront cash contributions varying
from 1% to 5%); 100% operation and maintenance cost recovery from user
communities; and about 60% cost sharing for individual latrines.
43
• The creation of an alternative service delivery mechanism for rural water
supply and sanitation, involving a partnership between the village water
and sanitation committee, NGOs and a Project Management Unit (PMU).
The Swajal project has undoubtedly been very influential in India's rural
water supply sector. The cost recovery objectives it set are now part of key
recommendations for India's national water supply policy.
Olevanna is a Gram Panchayat (GP) with twenty wards located on the eastern
side of Calicut,North Kerela. The GP covers 21.5 km2, with a total population
of about 50,000 people or 10,100 households (as of 1998).
Since 1987, and resulting from a crisis situation of both drought and a lack of
State funding, a network of privately run water supply schemes in Olevanna
has been established. Between 1987 and 1995, 26 privately registered schemes
arose across the Olevanna GP. In general, each household put in between
4,500-12,000 rupees for the capital costs of the scheme and then paid 5-10
rupees per month for operation and maintenance costs.
The GP played a facilitatory and regulatory role in this process, rather than
acting as a supplier. It also provided a review and audit procedure for each
co-operative's accounts, and it developed some rules of supply - limiting
supply to 400 litres a day and encouraging water meter installation to ensure
these limits were adhered to, or that extra costs for the additional supply (20
rupees per 1,000 litres) were paid.
The official state water supplier - the Kerela Water Authority (KWA) - was
not helpful in the early days, as the schemes challenged their remit. However,
since the decentralisation process for water supply and sanitation began in
earnest in Kerela in 1995, the KWA have been much more supportive.
Since 1995 a further 34 schemes have been created in Olevanna (which means
a total of about 10,000 households now benefit from these private schemes),
but the approach has changed slightly. Now, the GP can provide 50% of the
capital cost of the scheme and uses this, if required, to help with payments, for
instance if the households cannot pay the full 100% of capital costs. User
charges have increased to 30-40 rupees per month, due to increased electricity
costs for pumping. Many schemes are now operating at a surplus.
Of the participating households, it was estimated that 25% were below the
urban poverty line of 22,000 rupees per annum (Olevanna can be counted as a
peri-urban rather than rural GP). The start up costs to join the scheme could
represent 3-4% of household income (and in some cases up to 20%). To help
the poorest with the repayment process and monthly charges, monies are
collected "little and often" (a rupee a day) for those who prefer.
44
However, there have been some problems. Due to the level of water extraction
on the GP, issues of water resource management (and encouraging perhaps
water harvesting as a substitute to pumping) are now important - price
signals are seen by the GP as the best way to do this. Also, the GP admits that
support in developing its pricing policy to cover long run costs would have
been extremely useful.
Nevertheless, the Olevanna schemes show that full cost recovery can be
achieved in peri-urban areas (with reasonable levels of poverty) for water
supply systems. People did pay the full cost, relying on little external help.
Furthermore, these schemes have included effective metering and volumetric-
based charges and, importantly, have involved an institution that used to be
responsible for supplying water, re-emerging as an effective facilitator and
regulator of the system. Importantly too, the official water supply agency was
not a supporter of the system until a policy change toward decentralisation
took place, which legitimised the existence of the schemes in the state water
supply providers eyes.
Hi. The Kerela Rural Water Supply and Environmental Sanitation Project (KRWESP)
IBRD loan funds are passed on from the State Water Authority to District level
PMUs. Funds from the DPMU, the GP and the beneficiaries then flow into the
same bank account, to support each scheme within the district. The KRWSESP
is expected to directly benefit about 1.5 million people or 5% of the state
population. On average, capital costs for water supply are 6930 rupees per
household and annual operation and maintenance costs are 180 rupees per
household. In cost recovery terms, the aim is that 15% of the capital cost of
each scheme is to be paid by the beneficiaries (of which at least 7.5% of
payment must be in cash). The beneficiaries should meet 100% of the schemes
operation and maintenance costs.
For the water supply component, the project will aim for 70 litres pp pd, 8
hours pumping pd, with a 4 hour am and a 4 hour pm slot to pump water. For
the environmental sanitation programme, low cost latrines are envisaged for
45
those households below the poverty line only, and these households are
expected to meet additional costs over and above the 2,000 rupees provided
by the Project.
As per the project design, the Kunnummel Panchayat will contribute 10% of
the capital costs of the project, the beneficiary's will contribute 15% (7.5% in
cash and up to 7.5% in kind) and the remainder will come via the GoK (5%)
and the DPMU for the project (70%). NASRAD, a local NGO is the support
organisation for the project within this District and will help the DPMU to
provide community facilitation and participation expertise.
The project is proceeding well, although the DPMU thinks beneficiary groups
could pay a lot more towards the capital cost of the schemes than the 15% they
are currently asked for. Also, the project is not necessarily easily replicable by
other user groups, without another loan to the DPMU, given the low level of
beneficiary contribution required. The payback of the loan for the project by
the GOK is another issue that may question the replicability of the project.
iv. The Andhra Pradesh Urban Services for the Poor Project (APUSP)
The project adopts a process-based approach and so did not have any fixed
ideas about reasonable levels of cost recovery to aim for at the outset. Instead,
the aim has been to make the Municipalities more aware of the recurrent costs
of their WATSAN services and how best to both reduce the O&M costs and
fund these services sustainably. With regards to O&M charges, the project
46
follows state tariffs for water supply, which were only 30-40 rupees per
month, but have been increased to 60. The project would like to see them
increased to 70. At present, slum dwellers are obliged to pay only 50% of the
tariff.
Tiruchirappalli City is in the heart of Tamil Nadu, southern India and has a
population of 668,000.The city has 155 slum areas containing about 115,000
people. The Municipal Water Corporation had built community latrines in the
slums in the mid 1980's, but these latrines fell into in a state of decay and were
abandoned by local residents. The key problem was a lack of a sense of
ownership over the latrine blocks by local residents.
The NGO Gramalaya sat with local Women's Self Help Groups from within
these slum areas and discussed possibilities for sanitation improvements.
There was a clear local need for improved sanitation and washing facilities in
each neighbourhood and plenty of thought as to how such a facility could be
managed. One group came up with the idea of paying to use the toilet - 50
paisa a time, which would include the purchase of cleaning materials and the
wages for cleaners, and a watchman cum ticket issuer.
Water Aid provided Gramalya with a grant of 380,000 rupees to build the
"pay and use" latrine block and a child friendly toilet block for those less than
6 years old. The charging system was then introduced in one scheme, the first
of its kind.
A women's group looks after the upkeep and maintenance of the community
toilet block constructed in their slum. Every user is issued a 50 paise token
allowing him/ her to use the toilet, from a ticket issuer, employed bythe
women's group. A woman from the self-help group collects the money from
the ticket issuer and closes the account every 12 hours. The accounts (a ledger)
contain details on the number of users and money collected. During nights a
watchman cum ticket issuer is appointed.
The money collected is deposited in a Bank Account each week. At the end of
every month the particular Women's self help group from that slum convenes
47
a meeting and details to members the income and expenditure for a particular
community toilet. The balance, after meeting various expenses such as salaries
and purchases of cleaning materials is kept in the bank account. It is pooled as
a common fund for that particular slum towards health and sanitation related
promotional activities including extension of street taps, construction of
community halls, renovation of other damaged community toilets,
construction of domestic drains, rubbish bins and provision of street lights. As
the knowledge of these latrines has spread, loans have been made from the
surplus and advice given to other slum communities to help construct their
own "pay and use" latrines.
Visitors to the schemes have included state level decision makers and some
high level WATSAN officials from elsewhere, such as Chennai. However,
without an official policy change at State level, the local Municipal
Corporation tasked with providing WATSAN services in the slums won't
approve or promote the pay and use system.
In summary, the pay and use latrines are being utilised by very poor slum
dwellers and although the first schemes relied on 100% subsidy for the capital
cost, the schemes are more than 100% recovering their recurrent costs. The
capital cost could be paid back over a number of years. However, the
surpluses these schemes have built up as a result, are instead being used to
help wider WATSAN developments in the neighbourhood, or as loans to part-
pay for other slum self help groups to renovate their systems. As the schemes
are starting to self-replicate, less and less grant funding will be required over
time. No PMU is involved - a local NGO is the implementing agency. An
INGO coordinates the QC of the NGOs delivery. The "seeing is believing"
impact of these schemes is strong, both within and beyond the slum
communities. However, policy changes are required to allow the State
supplier of WATSAN to also "buy into" this approach to service delivery.
vi. Scope/Water Aid Rural Water and Sanitation Programme, Tamil Nadu
Scope is an NGO based in Tamil Nadu, which focuses on the conservation and
best use of water supplies. It is a local partner of Water Aid India.
Although SCOPE gets 60% of its funds from the State Government, in terms of
WATSAN issue in rural areas, it recognises that NGO's are a "drop in the
ocean", but seeks to obtain replication and uptake of its successes from the
Tamil Nadu Water Authority. A key achievement of the SCOPE- Water Aid
partnership has been to encourage the first village in Tamil Nadu (and
possibly in India) to develop 100% sanitation coverage for themselves.
48
pump on each. The community were asked for a cash contribution towards
the capital cost of these tube-wells of 10% in each case. Each household was
asked to pay 2 rupees per month for O&M. Hence 220 rupees per month were
collected (2,640 per year). A surplus was built up (operation and maintenance
costs were low - 60 rupees a year - and replacement/ extension costs - a new
set of piping - cost 3,900 rupees and have only occurred once).
At the same time as the first tube well, SCOPE also set up self-help and
savings groups for women. 25,000 rupees were put into each group by SCOPE
to kick start them. Each member paid 25 rupees a month as savings into the
group and 2 rupees subscriptions. Then, on a revolving basis, women could
draw loans from the fund at a 2% interest rate per month (24% per year).
Once the first 10 model toilets were built, they became very popular,
especially among the women (due to their convenience) and more households
requested a latrine. By the end of 1996, SCOPE had helped to build 68 toilets
for 96 families. By the end of 1997, every household had built a toilet. SCOPE
continued to offer loans of up to 650 rupees to those who could not afford to
meet all of their construction contributions. With a 650 loan and a 650 subsidy,
only 200 rupees maximum would be required to build the latrine. 39 out of the
110 households took on these loans. The self-help groups also offered further
financial assistance.
The income from the kitchen gardens helped to pay back the loan for the
toilets, producing on average 30 -90 kilos of fruit and vegetables a year, net of
personal consumption, which was sold in local markets for up to 400 rupees in
total.
By 1997, 75-80% of the women in the village were using the toilets, and by
1998,85% of women were and up to 72% of the men were. By 1999 almost
everyone was using a toilet. In 1999 a follow up participatory survey found
that no single case of diarrhoea had lasted longer than 2 days, and very few
people were spending any time or money in the clinics. As a result the saved
money was being spent on replacing roof tiles and buying more animals
49
among other things. By the end of 2001, latrine uptake was also occurring in
up to 40 neighbouring villages, with people willing to pay, or take loans, for
the full costs of the latrines. Households in other villages had seen the
economic benefits the latrines were bringing to their owners in Kar Hikulam.
3.7.1 Introduction
DFID has indeed adapted its approach to an overall context for WATSAN
projects that is substantially different from that in India. South Africa is the
richest country in Africa, enjoying considerable wealth from natural resources,
even though populations with very different income levels and access to
infrastructure services are cohabiting more or less successfully. As a result of
this overall level of resources, donors' involvement has been relatively limited
50
in the water and sanitation sectors, the projects being financed and
implemented directly by the central or local administration. This explains
why out of the projects visited, and in sharp contrast to the situation described
in India, none of the projects were predominantly donor-led with the NGO
role being redefined predominantly in terms of training and capacity building
One key element for ensuring long-term sustainability of water and sanitation
services was the introduction of radical institutional reforms, which involved
devolution of powers to local governments and modification of their
boundaries, in an attempt to integrate affluent white areas with black
townships and deprived rural areas. Institutional reforms also introduced a
clear demarcation between water service authorities (local government) and
water service providers. Water service authorities can choose between
providing services by themselves or through agreement with other types of
operators, including public and private operators, or even NGOs.
• The challenges of providing water and sanitation to all are huge, even though
considerable investment efforts have been accomplished in the years since the
end of Apartheid. These new investments are currently mostly financed
through national government subsidies, but this situation does not appear to
be sustainable, particularly with economic difficulties looming on the horizon.
51
Based on these circumstances, South Africa has developed innovative
approaches to providing services to poor and marginalized consumers that
generate interesting lessons for cost-recovery of WATSAN projects:
At national level, DWAF has displayed a great capacity for developing new
and innovative contractual forms for involving the private sector. In
particular, the Build Operate Train and Transfer (BoTT) programme has been
efficient for developing water services in rural areas in a speedy and
integrated manner, despite its own limitations.
Many water sector actors, including development engineers and private sector
companies, use community and demand-led approaches as they are very well
aware of the high risks of failure of WATSAN projects if local communities are
not involved.
Given the ability for local governments to contract with various types of
operators, ERM selected projects for visit in order to cover a broad spectrum
of potential contractual arrangements. Many of these projects are carried out
in a peri-urban setting rather than in a purely rural one, because this is where
many of the acute problems with cost-recovery have materialised, particularly
in township areas. Despite all their good efforts, however, few of the projects
visited had reached acceptable of cost recovery and the consolidation of any of
these small successes was made very difficult by the introduction of the Free
Water Policy. Projects visited included:
• ODI retail: where responsibilities for retail water services have been entrusted
to a public Water Board, Rand Water, which has mainly focused its actions on
the use of pre-payment metering devices.
52
society sectors) in order to solve cost-recovery issues in peri-urban township
areas.
• Kzva Zulu Natal rural services: where a development engineer has adopted a
community-based approach for developing rural water projects, with
interesting lessons for how to run and manage those projects in an efficient
manner.
• Peddie South within the Eastern Cape BoTT project: where DWAF, the South
African Ministry for Water, has adopted an innovative contractual approach
(the BoTT) in order to harness the capacities of the private sector in order to
transfer responsibilities for water and sanitation services to local
municipalities as quickly as possible.
• The Mvula Trust: a specialist NGO working on the provision of water services
in rural areas, which recently expanded its activities to the provision of
sanitation services but still mostly in rural areas.
Below are brief summaries of the projects examined during the field visits.
More detailed information about the overall institutional framework and the
projects visited is presented in Annex F to this report.
Rand Water is the bulk water supplier for the region of Johannesburg.
Following DWAF's initiative, it has signed a contract with municipalities in
the ODI region for provision of retail services in former township areas.
Given that the cost-recovery record is particularly poor in these areas, Rand
Water has put considerable effort into the development and installation of
pre-payment meters, for both individual and standpipe connections. Pre-paid
meters are seen as a way of keeping down costs through reducing losses from
uncontrolled use.
53
However, management quality appears to be significantly higher than in other
municipalities: it is akin to the success story of municipal water and sanitation
service management. Kruegersdorp municipality has also been experimenting
on a large scale with pre-paid metering devices, in a variety of socio-economic
environments. Its payment record has significantly improved since the start
of these experimentations. This has been matched with particularly good
budgeting practices. The municipality is currently considering a possible
corporatisation of the municipal services, which should be made possible
thanks to the relatively good payment records.
In 1998, Durban Metro water (and the city of Pietermaritzburg) initiated the
creation of two partnerships between themselves (public providers of water
and sanitation services in Durban and Pietermaritzburg respectively), a
private operator (Vivendi), a local NGO (the Mvula Trust), the Water Research
Commission and the local bulk water supplier (Umgeni Water). The objective
of these partnerships was to improve water services for the poor in Durban
and Pietermaritzburg, particularly through the development of innovative
approaches to water services provision. The partnerships were created
through two co-operation agreements, with one for each city, which outline
roles, responsibilities and financial commitment for these partnerships.
Powers in the partnership are equally shared, and financial contributions do
not imply more weight in decision-making. <13)
But Durban Water, one of the most efficient water utility in the country, is a
special case, as it can easily recoup the costs of this policy through cross-
subsidisation from a relatively rich household and industrial customer base.
(13) Note that these partnerships received considerable support and attention from the Business Partners for Development
initiative, for which the Kwa Zulu Natal projects are one out of eight pilot projects.
54
iv. Kwa Zulu Natal rural services: Nkwambase
ERM visited the Nkwambase project, in a rural area North of Durban in the
Kwa Zulu Natal province. The project serves a total of 800 households. For
this project, the development engineering firm Dynacon works as
sub-contractor to Umgeni Water, which is the bulk water supplier in the
region of Durban.
In 1997, DWAF signed BoTT (Build, operate, Train, Transfer) contracts with
"Project Implementing Agencies" (PIA) in each of South Africa's four poorest
provinces. Each PIA is a consortium between private sector actors and NGOs:
for example, both the Northern Province (Metsico) and the Eastern Cape
(Amanz'abantu) consortia are led by WASS (a subsidiary of Northumbrian
Water) and have the Mvula Trust, a nation-wide NGO specialising in
providing water services to the rural poor as partner for institutional and
social development aspects.
The rationale for designing this type of contracts was to accompany the
municipal service reform and to allow a quick transfer of responsibilities for
water and sanitation from central government (DWAF) to local governments.
Given that in many poor rural areas, local governments did not have the
capacity or the financial resources to take on such responsibilities in a short
period of time, DWAF developed the BoTT concept for accompanying and
overseeing this transfer.
55
The contracts require the consortia to offer a "one-stop shop approach",
covering five areas of skill: design, construction, operation and maintenance,
on-site sanitation and institutional and social development. Community
representatives work as partners on the project, as members of the Project
Steering Committees and Village Water Committees.
These contracts are now coming to an end, as they were initially signed for
two-years and renewed once. Both self-criticism (by DWAF, due to the high
cost of the programme) and external criticism (by NGO partners, due to their
limited ability to strategically orient the projects) are running strong.
However, it is generally recognised that these project management structures
have been successful at delivering rapid improvements in service delivery
with appropriate attention paid to the long-term sustainability of the schemes
established in such a way. Private sector operators have been quick in
recognising the interest of such approach, and are trying to win more projects
outside the BoTT structure directly with local governments. Indeed, following
that experience, DWAF decided to develop an "improved" BoTT model
contract, which could be used directly by local governments wanting to call on
the private sector to develop their water services on the basis of this "one-stop
shop" approach.
56
4 COST RECOVERY-HOW TO MAKE JT HAPPEN?
4.1 INTRODUCTION
Recall the four key cost recovery research challenges this project set itself.
• Does a role for the private sector enable (financial) cost-recovery in rural
and peri-urban WATSAN projects to be done better?
By drawing upon the theory and findings of the previous two sections, we
present an analysis of the factors that seem to encourage financial
sustainability in WATSAN projects in rural and peri-urban areas. We identify
a series of practical "silver bullet" issues that, if addressed, would help
maximise the chances of cost recovery happening.
• Financial sustainability is vital in the long run as there are not enough
public or grant based resources to finance all the existing or new
WATSAN projects required in rural and peri-urban areas.
57
• Many practitioners misunderstand the terms cost recovery, demand,
economics and financial sustainability.
4.2.2 On income
• The most prevalent and feasible method for assessing the demand of
users for WATSAN services is via focus groups.
• They are not the same thing, though both are often called, inter-
changeably, "affordability", "willingness to pay", or "demand
assessment" studies. The difference can be seen as to whether the study
assesses either a households' effective demand (their affordability of a
service on offer), or their notional demand (their willingness to pay for a
service they want).
58
respondents will value not only the technical option on offer, but also
their preferences for many other attributes - the institutional
arrangements, the suggested pricing schedule, the credit options on offer,
for example.
The weakness of the DRA debate was not to focus on the fact that poor
users will place a value on a broad range of the project's attributes, and not
just on the technical options. One key attribute in this wider range are the
set of options through which users can be helped to pay for the WATSAN
services they want. We have identified, therefore, a gap in the current
development of DRA that fails to uncover the relative weights that can be
put on service attributes to facilitate pro-poor design14.
The way poor households are asked to pay for WATSAN services affects
how much they are willing to pay. Discussing the use of credit or other
methods of financial assistance, and the type of payment structure on
offer, are important. They can help shape a poor households' willingness
to pay.
14
Much recent attention has focussed on the role of social marketing to improve WATSAN buy in among communities.
Little attention, however, has been given to a related marketing technique - conjoint analysis - that could be useful for
uncovering preference weights for service delivery.
59
4.2.5 On charges and tariffs
• With the right project, poor people will pay for their connection charge to
a WATSAN project. Some my need a loan or subsidy to help them.
• Alternatives to subsidising water tariffs for poor people, can be the use of
marginal/ AIC cost pricing for WATSAN services, plus a cash subsidy or
a form of credit given to the user, to help pay back a loan for a WATSAN
connection or investment.
• With these charging principles in place, poor people can derive a return
on their WATSAN asset and manage to run them in the long run.
Volumetric charges and other forms of targeted tariff policy can also be
introduced within the group. Surpluses can be lent to other groups to
start a similar WATSAN project. Without these principles in place, the
project will not be financially sustainable.
• The most viable private sector participation options for WATSAN service
provision for rural and peri-urban communities are local options. They
can include:
60
• Small scale contractors and suppliers. These agents can help develop local
private enterprise in the water sector.
• Water and sanitation related NGOS. These organisations can provide social
development and technical assistance.
To find out "what works" (or does not work) in the field, we looked at
outputs, not processes. We took a (non) cost-recovering project and then
disaggregated it to find out the process, which helped to create it. We
hypothesised that the field processes, and the commonalities between them
that created successful projects, when combined with some of the important
underpinning issues derived from the desk study, hold the key as to what
makes a successful cost recovering WATSAN project for poor people. Sections
2 and 3 and Annexes B though F contain the details of our investigations and
findings. We present here the conclusions.
• They are financially sustainable - they recover their (long run) costs.
• They do not rely on large external grants to help them out now, or in the
future.
• Other groups of poor people can find ways to replicate the project,
without a large reliance on external grants, and are interested in doing so.
61
In contrast, WATSAN projects that don't work do none of the above. They
don't recover their long run costs, they will need grants to keep them going
(say, within every 20 years), and there doesn't seem to be much sporadic
uptake of their approach by others, without the offer of a large external grant.
It is important to note that many WATSAN projects have wider aims and
objectives than the financial sustainability ones outlined above. Accordingly,
our analysis is by no means comprehensive in terms of defining success and
failure - it concentrates solely on the aspects of cost recovery that induce
financial sustainability.
These cost recovery objectives mostly reflect a desire for the project to meet
basic needs, to ensure an equitable use of the subsidy they offer, or at best to
break even on budget.
The answer to both is equally simple - another grant will be required. It is the
same answer for a public agency WATSAN project with a budget of millions
of dollars as it is for small grant project with a budget of hundreds. The grant
cannot continue in perpetuity, and the grant cannot meet everyone's needs.
There is simply not enough public sector or NGO money (portions of people's
tax, or their bequests) to fund these sorts of "public sector" WATSAN projects
and programmes for poor people, especially when considering the many other
worthwhile demands on fiscal resources. Somebody has to pay for the
WATSAN projects, in the long run.
The evidence suggests that those WATSAN projects that work best for poor
people have asked them, usually via focus groups, what they want from their
services and what mechanisms would help them best pay for it. Where people
62
have been unsure that they want, for example, sanitation, social marketing
techniques have been employed to show the tangible financial and economic
benefits of the option.
Clearly, the technical options on offer can be limited, by physical and logistical
constraints. Poor people can be consulted on these technical options (type of
pump, tapstand or private connection etc), and their preferred choices
selected. But an assessment of demands should not stop there. Empowering
poor people in the wider institutional design process, by discussing their
preferences over how these WATSAN options are initially paid for,
maintained, operated, charged for, owned and developed over time can
provide a much stronger foundation for a cost recovering project.
Indeed, when engaged in this manner, poor people start to consider the
financing issue for the WATSAN intervention in the wider context of their
own poverty and their livelihoods. The offer of credit can then help to both
increase their sense of livelihood security and bind them to repayment. It can
also lead to growth and sustainability of the WATSAN system, as ways of
using the surplus or supporting finance derived from the scheme can be
devised which help people both to move out from poverty and/or to expand
or replicate the scheme for others.
However, finding innovative ways for the users to make a return on their
assets, or focusing on the output - a financially sustainable WATSAN
intervention - whilst leaving the process of implementation up to the
community and local support agency (usually a good quality NGO, or an
innovative locally based private sector agent) are much less prevalent models
of demand-based design.
63
4.4.3 Projects that don't create sustainable institutions
While the PMU's can be useful to the public agency to account for their funds,
they are often not sustainable institutions and may be counter-productive to
the objective of creating or supporting financially sustainable WATSAN
projects that are managed at the lowest level of decision-making possible.
PMU's can be counter productive in two ways:
• They will tend to draw the better employees away from the Water Service
Agency to work with international experts on the programme or project in
the PMU. Yet it is within the Water Service Agency where innovative
thinking and leadership will be required, in order to change from a
provider to a facilitator of locally managed financially sustainable
WATSAN projects in the long run.
• They will focus on "capacity building" or "training the Agency staff" to re-
orientate their approach to WATSAN delivery, but based upon an
unsustainable source of funds (the programme budget and the grants it
provides) and a lack of cost recovering, or locally flexible projects, on the
ground that can be seen to financially sustainable in the long run.
Hence, when the PMU closes at programme end - after four years for example
- ex PMU staff will have to reposition themselves back in the Agency with
new goals, a changed remit and higher aspirations. However, they will be
unlikely to have a sustainable source of funds upon which to make these
changes the project recommended, or to sustain the investments the project
did make. Further, they may be left with a systematic approach to community
co-financing (maybe now enshrined in State policy) that focuses less on being
flexible to local conditions, on empowerment and on outputs and more on a
set of target cost recovery benchmarks (such as implement WATSAN projects
that recover X% of financial costs from users), which, again, maintains a
reliance on external funds in the long run.
4.4.4 Examples
Most of the projects that fit the above descriptions are large scale, public
(donor) agency driven projects. As such they are more prevalent in India than
n South Africa. The case study project and programmes in India that fit this
description include:
64
• The World Bank Swajal Project in India (US$63 million). A PMU based in
Uttar Pradesh overseas community contracting and target cost recovery
objectives for water supply projects of 10% capital costs and 100%
operation and maintenance; 60% cost sharing for sanitation investments.
There are many other examples of large development agency projects that are
not financially sustainable (in fact the majority are not). These case studies are
simply illustrations of the projects that staff members were kind enough to
introduce us to, during field trips to India and South Africa.
However, a counter argument to the criticism that these sorts of projects aren't
financially sustainable (or really why they should be as public sector projects)
is the premise that these projects can influence policy. This is an important
point. However, if WATSAN financing policy is to be influenced, it surely
should be influenced properly. However:
• If the policy is to focus the Water Service Agency, not on the objective or
output - support the development of WATSAN projects that make a
return on their assets- but on a part of the process - recover these targets
costs from users - then it is not financially sustainable.
65
4.5 WHAT DOES WORK?
The WATSAN projects for rural and peri-urban users that do work have
focused firstly on empowerment in the design process
4.5.1 Projects that follow an iterative process to design a demand driven "product"
• They discuss the social or financial costs that the potential beneficiaries
currently pay or incur with respect to their existing WATSAN services.
4.5.2 Projects that use a facilitating organisation who knows the poor well
Often these schemes flourish in spite of, not because of national policies, local
government, WATSAN agencies and sometimes the Donor Community. And
because they are often local, small-scale initiatives they tend to get less
exposure if they are successful than the larger Donor Agency project and
programmes.
66
4.5.3 Projects that use small grants and supporting finance to encourage
replication
Without this external seed-corn provision, the community may provide 100%
of capital costs, borrow money to do so, or rely on an entrepreneur, or a
wealthy benefactor to help them. In these cases, crisis often dictates.
With limited finance available the onus has immediately been on the local
institution to run the scheme they have designed as sustainably as possible.
The associated costs of doing so, and hence the revenues required, quickly
become transparent to all users.
In general the successful WATSAN projects have focused on outputs and not
particularly on systematic or specific cost recovery targets. This may be the
case because a crisis situation required a community to focus on what was
needed to be done, and the lack of finance available meant that it had to be
done as cost effectively as possible, and last in the long run. Or, it may be the
case that a facilitator encouraged the community to think about what they
wanted from their WATSAN services and what they would have to do or pay
to make it keep going. Working back from these outputs (financially
sustainable WATSAN projects), then innovative methods and processes for
local management and cost recovery are often devised by the users, such as
pay and use latrines.
67
focuses on the supply of money coming in and how it might best, or most
fairly, be spent.
4.5.5 Examples
Durban Water Metro whereby locally based flexible service levels were
introduced.
The Eastern Cape BoTT where partnerships between local government, the
private sector and NGOs have helped to deliver some financially
sustainable WATSAN schemes.
4.5.6 Conclusion
• Develop partnerships between the users, local NGOs or CBOs and local
private sector water providers.
68
• Undertake an iterative process to design a demand driven WATSAN
"product".
• Focus on the output of the project required and work back to create the
"product".
• Get the users to find a way of managing costs and making the initiative
pay for itself in the long term.
• Find ways to replicate those successful projects that don't rely on heavily
on external grants.
The last two of these four points, which relate to sustainable financing, can be
achieved by focusing on the (economic) benefits to be gained from the project.
Through working at the local level, innovative ways of managing or paying
for the project can be designed, often with the help of a supporting
community finance mechanism. If the output is set such that an operating
surplus needs to occur, then specific cost recovery targets need not be made
systematic, unless the cost of capital is to be included. However, in the first
instance, a seed-corn grant may kick start this process. Costs of capital may
emerge as an issue for second or third generation projects, which draw on the
surplus of the first to finance the capital for theirs.
There is clearly a market for WATSAN projects for the rural and urban poor,
which exists, for someone to enter. The number of customers is huge, their
demand for services cannot be questioned and the financial returns on the
right products can be impressive.
69
However, as discussed this market is focused on very small scale, local level
innovations, with poor people as the key decision makers. Their WATSAN
"products" differ in different locations and circumstances, and their
institutions, management structures and payment schedules consequently
differ as well.
This is not a market, therefore, that the Donor Agencies, the international
private sector or the International Finance Institutions (IFIs) operate well
within. Limited Donor Agency success in financially sustainable WATSAN
projects is an illustration on this. For the international private sector, it is
generally too risky a venture to invest time and resources in such a large and
diverse number of rural and peri-urban schemes with such a heterogeneous
set of consumers, management options and cash flows.
The international private sector needs an economy of scale - large sunk costs -
for their finance to work and a minimal risk, medium-term time horizon
within which to accrue its profit from the investment. Costs can be minimised
through standardising and systemising collections, maintenance and other
activities. And these attributes - large amounts of sunk cost, systemised
design, repair and collections mechanisms - are exactly not what seems to
work for the successful, financially sustainable WATSAN projects in peri-
urban and rural areas.
It seems therefore that the international private sector could operate as a bulk
supplier or a wastewater treater, but not as a financial underwrite and not as a
lone actor delivering WATSAN services for all of the rural and urban poor.
Furthermore, neither the international private sector nor IFIs work very well
with the poor. In fact, they often hire NGOs to do this work for them or, more
recently, are being encouraged to work in partnership with NGOs.
The NGOs of course, even the INGOs, don't have access to the amounts of
"kick-start" funds necessary to cover all the small-scale WATSAN projects
that are required. Further, many NGOs can also suffer from a public-sector
approach to project design, whereby long run costs are not recovered.
This leaves just one group of stakeholders who have the money in the long
run, the incentive and the ability to develop these types of WATSAN projects,
identified as successful in the long run for the rural and urban poor.
70
4.6.2 The poor
Only the poor really have the money and the incentive, in the long run to
design, manage, and run these WATSAN projects as they so desire.
If one were to look to see how much the poor, in total, are spending, or willing
to spend if their nominal demands can be made effective, on WATSAN
supplies, then this may be the only (and cumulatively the largest?) source of
finance really available for most of the long run expenditure required for
WATSAN services in rural and peri urban areas. And who better to scrutinise
how the limited finance available from other sources is spent and what kind of
WATSAN "product" can best be got for the money, than those who need that
it the most?
Thus, it seems that donors/ IFI/ state/ private sector money can only help, at
best, to provide a small kick-start to the WATSAN project process, which
seems sensible from an economic point of view. The design, management
structure, institutions, payment schedule and other arrangements of the
WATSAN product should be left to the poor to decide upon, with help to
make informed choices.
Design, build and operate (and transfer if appropriate) a WATSAN system the
poor want, using most of their money to finance it in the long run, with a
supporting community financing system that can help people to pay. Ensure
that the long run costs of these schemes are covered. Give the users a small
amount as a loan or grant to kick the project off, and ask for it back over a
longer time period. Allow any operating surplus that is created to be invested
in further WATSAN or other community development initiatives. Perhaps
find ways to minimise costs through upfront charges, or to maximise intra-
community choice through offering a range of flexible service options at
different prices.
If the poor designed it, own it, and like it, they'll sustain it. If others like it,
they'll copy it. The others may even ask to borrow some money from the
project's surplus to help finance their own version.
71
approach to the Donor Agencies themselves. Based on the evidence of what
seems to work, however, the following practical steps are seen as critical.
4.7.1 A Water Supply and Sanitation Finance Policy for the Poor
A clear and uniform national policy for the investment and longer term
financing of WATSAN services in rural and peri-urban areas is therefore
required within the country that assistance is to take place (a "WATSAN
Finance Policy for the Poor").
• To promote the need for iterative, demand focused design processes that
can strengthen key parts of the WATSAN "product" to suit local
conditions, for example particular institutional or payment arrangements,
or certain technical components, such as sanitation.
72
4.7.2 No Easy Money.
Often examples of people's own payments for 100% of the WATSAN schemes
have arisen through a crisis in supply. The aim is to avoid crisis, while
capturing the desire and ability to pay and take a stake.
New projects should not grant invest more than 50% of capital costs.
Supporting financial mechanisms are important to help people pay for the
remainder, and should be kick started through a one-off grant. Replication,
however, should not require any further capital investment grants to the same
level. New users should seek to draw mostly upon the financing mechanism
of the original project.
Rural and peri urban WATSAN schemes are often non networked, or fall
outside of the regular networked tariff system. Prior to the project,
successfully networked coverage is usually very low. Hence, many new
initiatives will have to be designed and created. Furthermore, rural and peri-
urban communities are not homogenous either within or between themselves.
So many smaller schemes should be encouraged that suit their particular users
best. This could be on a village by village or on a slum-by-slum community
basis. Smaller schemes mean lower exposure to the risk of financial failure.
Micro-networks, or non-networked systems, with up to about 30 households
per group seem to be a feasible size for success.
There will be many different ideas and approaches as to what may work. By
taking an output based approach, it doesn't matter so much what the design
of each project in the programme is (as technical options are limited, people
may choose differences in the institutional, payment, management options), so
long as minimum (technical, social, environmental) criteria are met and the
ENVIRONMENTAL RESOURCES MANAGEMENT DEPARTMENT FOR INTERNATIONAL DEVELOPMENT
73
project delivers the output required - a WATSAN project that people want,
which sustains itself financially in the long run. Some ideas and designs will
work and some will fail. The ones that work will quickly be replicated and the
failures will be forgotten, or learned from. With a stake in their own project,
people will also choose their options carefully to match their local conditions.
4.7.6 Others will know whom the poorest are. Use them to design and deliver local
projects.
There are usually better placed organisations, grassroots or local NGOs, CBOs,
community groups or local vendors, private sector actors or informal service
providers, who know who the poor are, what they want and what they will
pay for in terms of WATSAN services; or who can find out quickly and
efficiently if not.
Via the INGO or Partner Organisation, therefore, the Donor Agency should
encourage these grassroots organisations to identify the initial desires of the
community, rather than conducting its own research. In using grassroots
organisations for this initial information exercise, their potential to be local
service coordinators can be ascertained by the Partner Organisation.
Once a network of local service coordinators has been identified from among
the grassroots organisations, the Agency should use the INGO or Partner
Organisation to maintain this network (which may include local water user
groups, local private sector operators, CBOs, NGOs or informal service
providers), with an aim to help them form local partnerships to develop and
manage a portfolio of small-scale WATSAN projects with local users. Training
in long term financial planning, social development and technical issues can
be provided via the Partner Organisation.
74
To start with, some seed-corn money, disbursed by the Partner Organisation
to the local service provider, may be required either for capital investment or
to kick start savings schemes to help pay for the local WATSAN projects.
Following this initial injection of funds, all long term costs for the project
should be sought from its users.
Quite simply, if the local project works it works, and its customers will
becomes its best advocates for replication. If it doesn't work it will fold, but
the local contractor or user group will face the loss of her stake. If the local
contractor knows her market, the project should not fold, but should
organically grow.
Performance satisfaction about each local scheme would come via feedback
from the community themselves. Success at this level will be measured by
local replication and uptake; capital and longer run costs will de facto be
recovered, especially with good tutalage from the Support Organisation. If
people replicate the WATSAN project, with no external grant funding, they
obviously like it, want it and are willing to pay for it.
The Partner Organisation should have its own performance evaluated based
on how many smaller WATSAN projects are created and then sustained, how
many are replicated and by asking local customers what they think of the
project; success is not equated to how much money was disbursed (in fact, the
reverse) or what technologies were used. Innovations in terms of local
management, payment and financial structures should be looked for and the
successful ideas encouraged and knowledge transferred.
Key indicators such as for social development, environmental and health can
to a greater or lesser extent be focused upon depending on whom the Partner
Organisation chooses to contract to deliver services at the local level, or who
they choose to help provide additional technical assistance. However, these
indicators should be monitored and evaluated at key points in contract
renewal, and the contract refocused or re-tendered if needs be.
75
4.7.9 Don't use rules of thumb in steering design
Communities and hence the WATSAN projects they like, will differ. The
Partner Organisation, in collaboration with the Donor Agency should
concentrate on the outputs and indicators that they want to see achieved and
the grassroots contractors should be left to find the right process that people
will buy into.
4.7.10 The State cannot be ignored, but a mindset change will be required.
The cornerstone of success will be to firstly unpick the key financial and
economic drivers (cost savings, time savings, revenue generation, less medical
bills, other positive attributes) that will enable the WATSAN investment to be
easily marketed to potential users; and then to create a payment mechanism
and related financial support structures (community funds, savings clubs,
revenue streams from the intervention etc) that can help to turn nominal into
effective (paying) demand.
76
training, technical advice, strengthening of financial sustainability issues,
monitoring and evaluation. With overall output achievement resting with the
Company, not the Donor Agency, the Agency can replace either the Company
or the Partner Organisation and the integrity of the programme with the
recipient State Government and its population will remain. Again,
disbursements to the Company could be sequential, and based upon outputs
being achieved. Financial incentives to deliver could be based upon key
performance indicators at the programme level being met - financially
sustainable and replicable local schemes, which are equitable and
institutionally robust, with feedback coming from the users, the Partner
Organisation and the State Partner.
An overview of how to take these silver bullet issues and design a cost
recovering WATSAN programme upon them, is presented in the
accompanying Guidelines to this position paper.
Much of the previous discussion, however, focuses on the poor in rural and
urban areas who are not in a crisis position. They have access to social
networks, credit and information, however limited these may be.
However, there is one other group of WATSAN users not yet discussed. This
is the chronically poor, usually located in rural or more remote areas. This is
the target group, which humanitarian or emergency assistance organisations
cite as those who really can't afford to pay for WATSAN services, or who can't
assemble the resources to mobilise themselves.
Whether they are suffering from economic collapse, military conflict, post-
conflict stagnation, extreme environmental stress, political/cultural/ethnic
biases in policy or society, or combinations of these issues, there will be certain
groups of people in certain locations who simply will not be able to engage in,
or pay for, a demand driven WATSAN project.
They own no assets (they may have lost them all, or never had any), they have
no social networks to speak of (there may be social fissures within as well as
between communities), and they have absolutely no cash or other savings, or
access to formal or informal credit networks.
Yet WATSAN services are often close to the top of the list of their demands.
And the need to create financially sustainable WATSAN investments for this
group, is no less pertinent than for others, and is perhaps even more so if an
attempt to build this group away from dependency on grant based aid is to be
propagated.
77
• Firstly examine why people are so chronically poor (perhaps by using a
PPA process); then
• Assess which assets, if invested in, could best help in assisting to build
them out of poverty. In many cases (especially in semi-arid areas) the key
asset will be water supply and, less obviously, sanitation.
In this way, water and sanitation interventions can be viewed strategically (as
income generators, money savers, livestock waterers, land investments,
dwelling/ lifestyle improvements etc); as things that can add value to people's
livelihoods, help to alleviate poverty and facilitate longer term planning.
This could be step one in the process towards the kinds of initiatives
mentioned in the earlier section. WATSAN can thus be seen as a turning
point, or a catalyst to improving these people's livelihoods
78
COMMENTS
While we feel that some key steps have been made through this research in
the development of practical recommendations for designing financially
sustainable WATSAN projects for the poor in rural and peri-urban areas, there
are still are number of clear opportunities for further work, whose outputs
could feed usefully into the policy agenda on this issue.
These opportunities can be split into more blue skies research and more
action-research orientated exercises.
79
Annex A
Logical Framework
Al Logical Framework
W4 Raise the well being of the rural and urban poor through cost-
effective improved water supply and sanitation
Purpose: (Purpose to goal)
To review and interpret cost recovery mechanisms across a wide Strategies are available for developing a range of context- 1. Water and sanitation schemes in There is a recognition of a wider
range of WSS schemes in order to recommend a portfolio of best specific forms of cost recovery and institutional poorer areas are seen to be affordable potential for cost recovery for
practices for achieving financial sustainability and maximising the arrangements able to produce financial sustainability within and potentially successful - more water and sanitation schemes
potential for private sector partnerships. The investigation will those WSS sectors currently regarded as "more risky" - eg investment, expansion and involving a wider range of
focus particularly on those sectors where the client base is poorest rural and peri urban areas. replication takes place within and ownership and operation
and cost recovery is presumed to be most difficult. across countries. structures than the current
2. Take up of connections and supply models of private sector
options by the lowest income participation which focus on the
households. "lower risk" urban consumer.
Outputs: (Output to purpose)
1. A report on what seems to make a successful rural /peri-urban 1. A final draft report and draft best practice guidelines are 1.Mailing lists and key informant
water and sanitation project, in terms of financial sustainability, produced for DFIDand WEDC respectively. addresses. WSS project managers, NGC s,
user "buy in", tariff structuring and private sector involvement. The and water agencies provide
report will contain: 2. Through combining with WEDC's outputs under R7386, 2. Inception and progress reports to information relating to strategy,
dissemination occurs through the web and the production of DF1D targets for cost recovery and
• The results of a survey and literature review indicating the a suite of hard copy guidelines. exposure to risk.
current understanding about cost recovery in WSS projects, 3. Paper manuscripts for review and
levels that are currently being attained, an analysis of what 3. Agencies, NGO's and water service operators in the key publication
aspects of project design make it work, and the extent to which project regions are actively involved in the research process.
demand assessment exercises play a role and are useful. 4. Referees outputs from journals
4. Publications reviewing the limits of current service
• An analysis of the key attributes of case study VV5S projects in provision and the potential for cost recovery in more risky 5. Publications, draft final report and
India and South Africa that maximise private sector sectors are produced draft text for a manual.
involvement and cost recovery
Al
Narrative summary Measurable indicators Means of verification Important assumptions
other existing frameworks or models for WSS design, and will
maintain a focus on user affordabiiity and welfare.
2. Draft texts for a Guidance Manual and web site on the theory and
best practice guidelines for achieving cost recovery and private
sector involvement for water and sanitation schemes in peri-urban
and rural areas. These will include a stepwise version of the
developed methodology for achieving cost recovery.
3. Undertake a case study review of the evidence and experience (or Disbursement records
lack) of demand assessment and cost recovery initiatives for
selected water and sanitation projects in peri-urban or rural areas of
South Africa and India . Projects where the private sector has
played a role, will be particularly sought out as case studies.
. • . • • • •
5. Develop the text into a project report for DFI D, a draft "best
practices" manual for WEDC to use in the production of a suite of
guidelines for DF1D, and a draft web page, again for WEDC to
develop and launch in conjunction with their own research..
• • m•
Annex B
Literature Review
Bl LITERATURE REVIEW
1.1 INTRODUCTION
There has also been output and follow up to the World Water Summit in the
Hague, September 2000, which heralded the beginning of a return to a
discussion about social entitlements and subsidy in the WATSAN provision
debate. Other issues, such as water supply within a livelihoods context, and
the emergence of the concept of output-based aid, and its relevance to
WATSAN delivery in rural areas, have also emerged during the research
period.
Firstly, an overview of the key issues in relation to the debate about cost
recovery in WATSAN projects has been undertaken. This looks to review
some of the definitions of (financial) sustainability in the WATSAN sector and
to summarise the thoughts and statements presented on issues of cost,
payment and financing.
Lastly, the review looks at the approaches for recovering costs in WATSAN
ENVIRONMENTAL RESOURCES MANAGEMENT DEPARTMENT FOR INTERNATIONAL DEVELOPMENT
projects and programmes. It is divided into a review of tools (tariffs, subsidies,
micro-finance initiatives) and models (decentralisation, private sector
participation, community-based organisation, pro-poor regulation, output
based approaches and livelihoods approaches).
1.2.1 Interpretations
c> Business Partners for Development November 1991: Cost Recovery - Partnership Frameworks for
Financially Sustaianble Water and Sanitation Projects; Linda Stalker Prokopy and Krisitn Komives.
<2> Seven Key Factors for Sustainable Cost Recovery in the context of community - managed water supply.
Francois Brikke and Johnny Rojas. IRC Occasional Paper. December 2000, IRC International Water and
Sanitation Centre, Delft, The Netherlands
P> Understanding the Cost Recovery Challenge: A Study of Ten Rural Communities North West Province
and Northern Province, July 1998. A DANIDA Report for the Department of Water Affairs and Forestry
South Africa. Hilary Syme and Frank Snijder
ENVIRONMENTAL RESOURCES MANAGEMENT DEPARTMENT FOR INTERNATIONAL DEVELOPMENT
B2
other resources, not only for the initial infrastructure, but also for recurrent
costs such as operation, maintenance, repairs and replacement.
Three core issues emerge from the range of interpretations on cost recovery
we looked at.
Until the 1990's and UNCED, the review found that need, especially in
relation to health, tended to be the chief criterion for matching communities to
WATSAN projects. Historically, technologies were supplied by donor
agencies on the basis of assumptions made about consumer's preferences.
However, from the mid 1990's onwards, project implementation in the water
and sanitation sector started to reflect the principles established at the
International Conference on Water and the Environment held in Dublin in 1992 and
subsequently endorsed at the Rio de Janeiro UNCED Conference in 1993.
<4> Assessing Sustainability - The Sustainability Snap Shot, Steven Sugden, Water-Aid Malawi.
<5> Managing Water as an Economic Good - the transition from supply orientated to demand responsive
services, Mike Gam, World Bank 1998.
ENVIRONMENTAL RESOURCES MANAGEMENT DEPARTMENT FOR INTERNATIONAL DEVELOPMENT
B3
These core "Dublin Principles include the following.
In this note, the concept of a Demand Responsive approach (DRA) was put
forward. DRA was an approach to WATSAN that attempted to respond to
consumer demands and which aimed at making projects more sustainable
than supply-driven approaches. Garn listed the following as key
characteristics of DRA:
(6) See International Conference on Water and the Environment, 1992; Briscoe 1997
(?) McNeUi, D (1997) Water as an Economic Good, Natural Resources Forum, vol. 22,4, pp 253-261; Perry, C.,
Rock, M. and D. Seckler (1997) Water as an Economic Good: A Solution, or Problem? In Key, M. Kay., T.
Franks and L. Smith (eds.) (1997) Water: Economics, Management and Demand, E&FN Spon.
(8) J. Winpenny, 1997; Carter et al, 1998
0 For example, J. Winpenny (1994) Managing Water as an Economic Resource; Routledge; M Kay, T Franks
and L Smith (eds) (1997) Water: Economics, Management and Demand, E&FN Spon.
ENVIRONMENTAL RESOURCES MANAGEMENT DEPARTMENT FOR INTERNATIONAL DEVELOPMENT
Garn also suggested that in order for DRA to work effectively there needed to
be competition among communities for funding, in order to decide whether to
provide support to a particular community and what type of system and level
of service to provide.
For many NGOs, however, this competitive aspect of the DRA debate was
fundamentally flawed - how could poor communities compete and pay for
WATSAN projects? To these stakeholders, DRA represented the problems that
seemed inherent to them both in declaration to "treat water as an economic
good", and in the methods for finding out how much people were willing to
pay, such as contingent valuation. For proponents of the DRA approach,
however, entering into this debate was seen as critical in order to move the
sector away from the minimum level of service, supply orientated approach
reliant on affordability rules of thumb, which Garn criticised.
The extensive empirical research conducted by The World Bank between 1987
and 1990 was used to back up the DRA argument.10 This study found that
household income, though often important, was not the overriding
determinant of demand for improved WATSAN services. The following three
factors (in no order of importance) were found to be significant:
However, despite this debate, there remained little practical agreement (or
conclusive evidence) indicating how sustainable cost recovery strategies could
be ensured over time for WATSAN projects, or how they were to be defined.
Conceptual camps tended to emerge between the DRA/ World Bank
approach to WATSAN and the more "rights-based" approach to WATSAN.
The second World Water Forum was held in The Hague in March 2000. Partly
as a result of the debate between the DRA and rights-based camps on
WATSAN, among many other issues, the debate and the subsequent
Ministerial Declaration on Water for the 21st Century was slightly different to
that of the Dublin Declaration. It stated, on valuing water:
"to manage water in a way that reflects its economic, social, environmental and
cultural values far all its uses, and to move towards pricing water services to reflect
<lq> Demand for Water in Rural Areas: Determinants and Policy Implications. Briscoe et al. World Bank
Research Observer, January 1993.
ENVIRONMENTAL RESOURCES MANAGEMENT DEPARTMENT FOR INTERNATIONAL DEVELOPMENT
B5
the cost of their provision. This approach should take account of the need far equity
and the basic needs of the poor and the vulnerable."
There is a clear shift away from the DRA agenda in the statements from The
Hague, toward a more rights based approach for WATSAN delivery.
The third world water forum will be held in March 2003 in Japan, where again
this issue will no doubt be re-visited, however, it does seem that the findings
and arguments that the DRA agenda put forward in the late 1990's remain
pertinent to the cost recovery debate and could provide part of the solution,
when one is trying to identify "what works" vis-a-vis financially sustainable
WATSAN projects and programmes.
1.2.3 Demand
There are many sources of information on the range of water demand curves
that exist for consumptive uses. The literature agrees that, in reality, several
different demand curves could exist for different water uses within and
between households. To derive a demand function (or series of functions)
from observed data for all of these uses would be a very difficult exercise to
undertake given the time and resource limitations for any standard project
feasibility study. Instead, in general, a log linear functional form is assumed for
the demand curveforwater. This implies that the price elasticity of demand for
water is a constant at all levels of water use within households. Although this
is at best a crude representation of household water demand behaviour, it
allows economists to provide a reasonable approximation of people's actual
demand behaviour."
Conversely, there are many, many references to both the willingness of poor
people to pay for WATSAN services, and the often-costly coping strategies
that they endure to access WATSAN services, as evidence of the fact that they
di) Dale Whittington and Venkateswarlu Swarna "The Economic Benefits of Potable Water Supply Projects
to Households In developing Countries. Asian Development Bank, Economics and Development Resource
Centre. January 1994.
ENVIRONMENTAL RESOURCES MANAGEMENT DEPARTMENT FOR INTERNATIONAL DEVELOPMENT
B6
often pay high prices. Literature from the UNDP-World Bank Water and
Sanitation Programme provides a good overview of both these issues.12
Attention has also turned to looking at peoples productive uses of water as well.
These uses can range from vegetable gardens, cattle farming activities to
traditional beer making activities. Surveys conducted in villages in South
Africa on the uses of water, indicate that there is a high economic value for
water (up to 40 litres beyond the basic consumption level), which is accounted
for mostly by productive, or economic activities. Other work has found that
the daily per capita amount of water used for productive activities is more
than double in villages with better access to water supplies. And importantly,
the sustainability of the system depended on these demands being met. 13
In general, the most highly rated benefits are water reliability and water
security.14 These benefits can be of particularly high value if they are seen in
relation to the scheme's perceived usefulness to assist in other income
generating activities within the household.15 Combined with this, is the
ability to provide capacity for all consumption uses into the future - ie to meet
future as well as present demand. Often more people will return to a village or
a region, if the water sources become more reliable. Conversely it has been
found that water shortages or unreliability of service will, not surprisingly,
decrease the rate of payment for water services.16
The literature generally sees demand assessment as crucial for the correct
determination of consumer preferences.17 If it is assumed that consumer
preferences are a function of costs, other variables and varying qualities and
quantities/ levels of water and sanitation provision, then it is important for
demand to be assessed accurately as unsatisfied demands for water and
sanitation provision can make a project unsustainable.18 Importantly the
literature also suggests that demand assessment may highlight the need for a
mixed level of service, to reflect both users differences in demand, and seasonal
variations in geophysical conditions.19
<12> Willing to pay, but Unwilling to Charge. UNDP-World Bank Water and Sanitation Programme South
Asia, 1999; Water Resources Policies and the Urban Poor; Innovative Approaches and Policy Imperatives,
Bhatia, Ramesh, Falkenmark and Malin, Water and Sanitation Currents UNDP-World Bank Water and
Sanitation Programme, 1993.
(13)j.CFerez de Mendigueren, WEDC Conference, 1998
<"> Garn, M. (1997) Managing Water as an Economic Good, paper posted at:
http://www.oneworld.org/ think tank/ water/index.html.; World Bank Water Demand Research Team,
World Bank Research Observer, 1993 pp47-70
(1S) (Waughray et al, 1999; B. Rihardjo and D. O'Brien, 1994
(17) J, Briscoe, 1996; Carter et al, 1999; World Bank-UNDP Community Water Supply and Sanitation
Conference, 1998.
(18) J.CPerez de Mendigueren, WEDC Conference, 1998
(19)M. Webster, WEDC, 1998
( 20 )B.Copeland, WEDC Conference, 1995; J.C. Perez de Mendigueren, WEDC Conference, 1998
ENVIRONMENTAL RESOURCES MANAGEMENT DEPARTMENT FOR INTERNATIONAL DEVELOPMENT
The World Bank Global Study on Water and Sanitation found that, although
technology choices should be linked to prices so that communities understand
how their payment is determined, most projects, which were reviewed, did not
make this link explicit. The result was that many household members
perceived their contribution as a tax. This World Bank study also revealed that
it is important to give communities follow up information on project
management, especially operation and maintenance so that they can
understand the level of tariffs. In many cases, therefore, it seems that
communities do not perceive that they face an economic trade-off for a higher
level of service. And it is exactly this issue that demand assessment aims to
resolve.
Pearce (1981) defines effective demand as the "aggregate demand for goods
and services which is backed up with the resources to pay for
them distinguished from 'notional demand' which refers to a desire for
goods and services" (21>
Credit or other financial assistance and the type of payment structure on offer,
can help to translate notional into effective demand, rendering affordability or
rule of thumb analyses less useful. These differences may be of central
importance to the WATSAN debate on financial sustainability.
For example, for poor people in rural /peri-urban areas, the marginal utility of
money is high (people value hard cash highly), incomes tend to be low or
insecure, and often people pursue a wide range of livelihoods whereby their
wealth is tied up in a number of assets, and not just in disposable income. A
large part of the literature (especially Government policies) has therefore
focused on the fact that poor people cannot or should not pay for WATSAN
services. However, many studies have also shown that poor people do in fact,
often pay a high price to access WATSAN services, just to cope. A range of
other studies have also suggested poor people are willing to pay for improved
WATSAN services. And there are also many examples of rights-based
WATSAN projects, which have found that innovative ways of paying for
water such as through savings groups, revolving funds etc can be useful - the
way the payments are designed can also influence how much people will pay.
Hence in theory, the effective WTP of the consumer can be fixed, but if they
have access to other assets (labour, livestock, credit, savings schemes) and if
the design of the project particularly suits them, then their notional demand
may be quite high. If the right mechanism can be designed and implemented
to translate their notional into effective demand for WATSAN services,
therefore, then they will be willing and able to pay beyond their observed and
immediate income constraint.
«2> Dale Whittington and Venkateswarlu Swama "The Economic Benefits of Potable Water Supply Projects
to Households In developing Countries. Asian Development Bank, Economics and Development Resource
Centre. January 1994.
ENVIRONMENTAL RESOURCES MANAGEMENT DEPARTMENT FOR INTERNATIONAL DEVELOPMENT
The literature suggests that the advantage of community meeting techniques
is that the additional discussion generated by groups of individuals should
increase the quantity and quality of information available to those
participating. This additional information should reduce the hypothetical bias
associated with CVM. On the other hand, there is also a debate that the threat
of strategic behaviour might increase during community meetings if
participants realise that they might benefit from exaggerating or mis-stating
their preferences.
IRC carried out some recent research to identify key factors that contribute to
sustainable cost recovery in the context of community managed water supply
services. The seven key factors for sustainable cost recovery are briefly
highlighted below (Brikke & Rojas, 2000):
• The community should be assisted to set a clear strategy for cost recovery
that articulates the concepts of equity, efficiency, participatory process
and capacity building.
These seven steps broadly concur with earlier suggestions as to the key
variables that determine whether demand assessment leads to sustainable
financial projects:
And DFID's guidelines do not differ significantly from these same general
viewpoints concerning critical factors for financial sustainability of WATSAN
projects. DFID criteria include the following:
• Ensuring that the correct financial and technical training is given to the
community;
At the earliest stage in the project cycle, most agencies identify their WATSAN
projects from country sectoral strategy and country programmes.
The review then found that many development agencies agree that the
identification of specific communities for water and sanitation provision may
be the most important step to project design. ^However, the issue of
identifying demand to ensure financial sustainability tends to be side-stepped
slightly, and instead appropriate communities the agency or organisation can
In this instance, demand assessment does not play a role in identifying the
demand for agencies projects; instead the onus is on finding communities the
agency can effectively "work with", or finding implementing agencies (often
NGOs), who can liaise between the agency and the community.
• Current joint policy in Vietnam between UNICEF and CERWAS for water
and sanitation projects (a Vietnamese government body for water and
sanitation) stipulates that communities must contribute 20 per cent of the
capital costs of their water systems. On a pilot basis communes have
repaid 100 per cent of capital costs in regular instalments over a number of
years.
(24) ibid.
t25) World Bank-UNDP Water Supply and Sanitation Conference
ENVIRONMENTAL RESOURCES MANAGEMENT DEPARTMENT FOR INTERNATIONAL DEVELOPMENT
agreement on how and where in the project cycle demand assessment
techniques should be used.
DFID states that demand assessment is important at both the policy stage of
the project cycle and at the project identification, preparation and appraisal
stages. In larger more complex systems, there will be a need to develop
institutional capacity for continuing demand responsiveness. Demand
assessment should be carried out in the early stages to establish broad
parameters of demand to inform strategic design decisions. In later stages of
the projects cycle, a more detailed costing will inform tariff structure design
decisions.
1.2.8 Affordability
Some agencies such as the Asian Development Bank 27 and DFID recognise
that the 5 per cent affordability indicator (the assumption that people would
be willing to pay 3 to 5 per cent of their income on water) is arbitrary and
recommend that affordability studies be put in the context of how consumers
currently cope and what they pay for water.
In general, the literature shows that willingness to pay (WTP) has proven to be
a more successful indicator of demand than ability to pay. For example, a
WTP study in Nigeria reports WTP values that are 18 per cent of incomes for
rural low-income households.28 Some studies have shown higher WTP for
poorer households, compared to a standard of around 2 to 3 per cent of spent
on water services among high-income households. These results indicate that
rural communities may prefer service levels above the minimum often
prescribed and that rural or peri-urban water and sanitation projects may not
actually require extensive government subsidies.
The literature suggests that the reliance on ability to pay studies derives from
the fact (or belief) that WTP analyses, with their preference for using
hypothetical markets, ignore the actual financial constraints the household
faces. Further, WTP by women in particular may not be even close to their
However, the debate about the need to identify notional demand through the
use of WTP or other community-based studies remains a persuasive one, in
order to break the cycle of low levels of supply orientated service, and low
expectations. For, although it remains common for subsidies to be given to
cover the difference between notional demand and affordability (if notional
demand has been identified), other more sustainable approaches to help meet
effective demand are emerging in the literature, such as the linking of income
generation activities or the use of micro credit, as components that can create
more demand orientated and sustainable water and sanitation projects.
This section of the review is divided into a review of tools (tariffs, subsidies,
micro-finance initiatives) and models (decentralisation, private sector
participation, community-based organisation, pro-poor regulation, output
based approaches and livelihoods approaches)
1.3.1 Tariffs
In general, the literature agrees that an efficient tariff policy should be one that
reflects the marginal economic costs of supply.
The literature also agrees, however, that tariffs can also commonly be used as
instruments of social policy, such as the redistribution of income based on
equity concerns. For example, connection fees can be kept low to prevent
entry being deterred and often charges can be subsidised for some period.31
A commonly observed approach in urban environments in developing
countries is to use a tariff schedule that consists of a low subsidised "lifeline"
rate for the first 6-8 m 3 per month, and a higher rate for all additional
consumption. More complex tariff schedules with multiple increasing blocks,
sometimes in proportion to the recorded income distribution of the country,
The problems associated with pricing water of rural, agricultural35 and peri-
urban areas are highlighted by an IRC-UNDP project that assessed how far the
Dublin principles had been followed. The study showed that progress with
tariffs in line with the Dublin principles is greatest in urban areas where
metering is more common, compared to other contexts, where the consumer
base is more difficult to define. 36
Where metering is possible, the literature seems to concur that a two part tariff
is an improvement on increasing block tariffs as it can achieve the intended
income distribution element required and eliminate the within-block
regressivity of an increasing block tariff. The first part is a capacity charge that
determines a user's maximum usage during periods of excess demand plus
any other fixed costs The second part is a usage price equal to marginal cost.
The two part tariff confronts most households with the full marginal cost of
supply, f38)*39)
Optimal designs for water tariffs will of course vary with the varying nature
of water resources. Where the opportunity cost of water resources is low
(demand is below supply, currently and in the future) the fixed costs of water
system are high compared to the variable costs. The costs of metering and
usage based billing are thus high compared to the cost of water and meter
installations may not be worth the cost. However, for those areas where the
opportunity costs of water resources are high, metering is more likely to be
worth installing. 40
Some commentators argue that a more holistic approach to water supply and
sanitation - using an integrated water resource management policy for
example - may be a more productive way to build truly sustainable financing
strategies into peoples water tariffs, because user charges for water often do
not currently include a pollution cost element.42
DFID suggests that the use of cross subsidies, lifeline tariffs and rising block
tariffs are all part of the same strategy to include equity objectives into cost
recovery. In addition, water connection costs can be eased for low-income
households by subsidising the costs or allowing the connection fee to spread
over a longer period.
The review found that, although there is widespread agreement that users
must be responsible for operations and maintenance,43 there might be
circumstances where subsidies are acceptable. For example, there may be
conventional economic reasons for subsidising water or sanitation services,
such as the presence of external health benefits, which may be compounded
by a lack of awareness of health issues.
In fact, the literature suggests that there are many reasons why water and
sanitation suffer from price distortions. Commonly, treating water and
sanitation as social instruments for the wider objective of rural development
or alleviating macro-economic inefficiencies may lead to the second best
problem. u If inefficient conditions prevail in other areas of the economy, for
example, the price of labour is below its economic cost, then consumers are
not able to pay for water. National priorities on food security may lead to
inefficient water pricing practices for the agricultural sector. In most
developing countries, for example, charges for irrigated agriculture have been
much lower than those required even to pay for operation and maintenance.45
Furthermore, the review found that the subsidy ceiling can be the most critical
element of a sustainable financial policy.47 The per capita ceiling on capital
grants has been instrumental in ensuring sustainable financial rules in the
Mvula programme. Conversely, perverse incentives may occur with
expressing subsidies as a percentage of the capital cost of each option. The
option with the highest subsidy may be chosen on the basis of it being the
biggest capital contribution to the community. *
In peri-urban areas, the review found that cross-subsidisation policies can fail
to provide incentives to the utility to bill or collect from many users at the low
end of the tariff spectrum, thus aggravating cash flow problems. In fact, social
tariffs may create disincentives to expand services to low-income areas. 49
Unrealistically high rates to industrial and commercial users can force them to
look for alternative supplies, thus reducing market share and, in the process,
exacerbating the cross subsidy problem. 50
In relation to lifeline or rising block tariffs, the poorest often have to obtain
water through shared connections, from neighbours with connections or from
water vendors. If several households use the connection the group quickly
exceeds the volume in the initial block, pushing water use into the higher
priced blocks. This problem occurs also with water re-sales from households
with private meter connections from water vendors. Under such pricing
policies, the poor may pay as much as 30 per cent of their income for water,
while wealthier households pay less than 2 per cent. 51
This ties in with other current research on subsidies, which indicates that
where a government decides that a subsidy is appropriate, then it is usually
preferable to subsidise access to a service rather than consumption.55
DFID's approach to subsidies for water supply schemes is that they should be
justified on income redistribution grounds, and not on direct health benefits.
DFID suggests that subsidies for sanitation may be needed to correct for
'market failures', which arise because inherent demand does not lead to the
level of investment in those sanitation services which would be most efficient
for the economy and society. This policy, in fact, reflects the argument stated
earlier for identifying the notional demand for WATSAN services, and then
finding a viable selection of financing instruments to help the community pay
for them. However, in this case, DFID advocates the use of subsidies at
present.
Sanda and Oya suggests that better access to micro-credit should reduce the
gap between service affordability and consumer's WTP and help low-income
households afford the longer term options. However, designing an
appropriate incentive compatible form of credit scheme seems to be
instrumental in stimulating disadvantaged groups to initiate income
generation activities and to enhance their group solidarity.56
Some practitioners in the water and sanitation sector believe that sustainable
financing strategies for WATSAN projects actually require the complementary
development of micro-credit and savings mechanisms to build a source of
financing to pay for improved levels of services and operation and
maintenance - that the financing issue should be considered in the wider
context of poverty eradication, for both sanitation and water services.57
These practitioners feel that credit binds people to repayment and leads to
growth and sustainability of the system. &
This seems like a centrally important point that can link the issue of capturing
notional demand to the design of a reasonable standard of WATSAN service
supply, which can become financially sustainable.
Although the survey found that private sector participation could also assist
in the provision of demand driven services61, it now appears clear that specific
attention to the needs of demand driven services for poor people should be
explored from the outset of contract and regulatory framework design. If this
does not happen, then there is a risk that the resulting regime actually creates
further barriers to the provision of such services.62
For rural areas, many water sector studies recommend that the only
sustainable manner of managing water supply and sanitation is for
community management models to manage their own water and sanitation
supply systems.63 However, this can be difficult in cases where the legal and
regulatory frameworks have not formalised the status of such organisations.
This can create very practical difficulties for such organisations, especially if
they are tasked with collection of tariffs and financing of maintenance,
without a clear legal entity allowing them, for example, to open a bank-
account.64
In general, the survey found that the particular management model adopted
for WATSAN projects seems to impact significantly on the level of demand.
Communities may distrust a national government's capacity to manage the
Also, the more homogenous the community the greater are the chances of a
successful common property regime developing.68 This point was highlighted
by a water supply project in Uganda where the size of catchment area (and the
heterogeneity of its population) for the project was seen as too large to be
sustainable. 69
<»> World Bank Rural Water Supply Global Study, 1998; WEDC Conference 1995; Wittington et al, 1990.
'"'World Bank Private Sector Note No 188,19991
im
(Upcoming publication) The interface between Regulatory Frameworks and Partnership Business
Partners for Development Water and Sanitation Cluster, January 2002
The literature is generally agreed that public utilities do not have the financial
or political autonomy to set tariffs at levels that recover costs.72 Private sector
participation in WATSAN projects and programmes is therefore seen as the
most efficient way of ensuring financially sustainable pricing policies73.
Many commentators feel that small-scale private sector (or currently informal)
inputs can play a role in resolving some of these issues. If these operators are
responding to expressed and identified demand, then in doing so, cost
In general, the most viable private sector participation options for rural and
peri-urban communities are seen as the following: 75
• Small scale contractors and suppliers. These agents can help which develop
private enterprise in the water sector.
• Water and sanitation related NGOS. These organisations can provide social
development assistance and the small scale management contracting of
systems.
However, it may be difficult to combine the short time scales within which the
private sector operator may be required to work, with the longer time-scales
that demand lead approaches take. One solution is to formalise the
relationships between these small-scale suppliers and the main supplier in the
form of partnership, a solution that the World Bank Business Partners for
Development (http://ivioiv.bpd-ivaterandsanitation.org) has been investigating. In
practice, partnerships have shown to offer significant potential, but a whole
myriad of problems are still to be resolved if such partnerships are to meet
their potential: for example, NGOs and other organisations may simply be
unwilling to enter such partnerships with the formality that the private sector
may demand, in order to minimise their operational risk.76
<*> ibid.
76
. (Upcoming publication) The interface between Regulatory Frameworks and Partnership Business
Partners for Development Water and Sanitation Cluster, January 2002
TO Carter et al, 1997
(78) World Bank-UNDP Community Water Supply and Sanitation Conference
ENVIRONMENTAL RESOURCES MANAGEMENT DEPARTMENT FOR INTERNATIONAL DEVELOPMENT
solutions, such as more locally established partnerships may offer an
alternative, however, practical issues concerning friction between
geographically close communities, may cause difficulties if private-public
partnerships are let at a local level.79
It is suggested by the authors that output based aid approaches can provide
sharper thinking about the use of the subsidy, perhaps directing the funds to
the one-time costs of service connection - typically the main impediment to
expanded access to services - rather than the ongoing costs of consumption.
This may be particularly pertinent to WATSAN service delivery, cost
recovery, and how donor agencies implement their projects, especially as the
(") ERM research as part of preparation for the World Bank publication "Lessons in Private Provision of Rural
Infrastructure Services" financed by tlie World Bank Public-Private Infrastructure Advisory Facility (PPIAF)
C0) "Contracting for public services: output based aid and its applications." P Brooke and SM Smith (eds)
World Bank/ IFC 2001 http://rru.worldbank.org/Hot_Topics_Smith_Brook.asp
• Nicol, ODI Working Paper 133: Adapting The Sustainable Livelihoods Approach
to Water Projects, ODI April 2000
• the potential hydrological resource base - the quality and quantity of the
water resource available;
• the existing and potential patterns of use, including which groups use
which particular water resources and the contribution they make to their
livelihoods;
• the legal, political and institutional context for water resource use; and
• the trajectories of change for all of these factors, facilitate the positive
changes and mitigate the negative changes
• Water and food security. Is water the only limiting factor to livestock?
What about constraints on financial, physical and human capital assets? To
what extent are local water markets captured by a local elite, thus
constraining rights over, or access to, water for the poorest?
(n) The livelihoods approach typically focuses on the analysis of a set of five interlinked assets, Nicol
provides the following description of the asset pentagon: physical capital (household/individual labour,
health), financial capital (access to money), social capital (community mobilisation), natural capital
(environmental resource stock), human capital (education and awareness). He also suggests that political
capital be included in this analysis of assets.
ENVIRONMENTAL RESOURCES MANAGEMENT DEPARTMENT FOR INTERNATIONAL DEVELOPMENT
525
• Water and household maintenance. The need to mitigate the morbidity-
threat of direct water borne diseases and water related diseases, is central
to the establishment of sustainable rural livelihoods.
1A.3 Scarcity
• understanding water supply and access conflicts which may be present and
the institutional processes through which they can be mitigated;
• the fact that water scarcity (in a social as well as a physical sense) cannot be
purely understood at the local scale alone;
• people are not passive victims of water scarcity, but respond in different
ways. These responses can provide indicators on how to develop water
based livelihoods or mediate water based conflicts.
• Decreased stress on the four water livelihood uses for water (food,
households, livelihoods and ecosystems)
For example, there may be a more tangible return to people's financial asset
from using water for livestock (money, savings), then on their physical asset
(their health). The way people choose to use the water supplied to them may
well be a rational reaction to which of their asset bases is currently weakest.
Further, a demand focused approach to water supply, taking on board a
livelihoods perspective, would therefore design a water project around the
interventions which people feel would provide them with the largest (or most
immediate) returns, and thus which they'd be most willing to pay for.
"an end goal..(may) not be the achievement of health benefits, but seeks to
increase the overall robustness of the household asset bundle, in order to
strengthen livelihood strategies (and) assist in creating pro-poor outcomes".
Briscoe, J. (1996) Water as an Economic Good: The idea and what it means in
practice. Presented at the World Congress of the International Commission on
Irrigation and Drainage (ICID).
Carter, R.C., Tyrrel, S.F., Howsam, P. (1999) The Impact and Sustainability of
Community Water Supply and Sanitation Programmes in Developing
Countries. J. CIWEM, 13.
Evans B (ed) (2002) Tariffs and Subsidies: Chapter 5: New Designs for Water
and Sanitation Transactions, upcoming World Bank publication
Kariuki, M. (1997) WSS Services for the Urban Poor Water Supply and
Sanitation Collaborative Council: Vision 21 - Water for the people.
Kay, M., Franks, T. and Smith, L (1997) Water: Economics, Management and
Demand. E &FN Spon.
Noll, R., Shirley, M.M., Cowan, S. (1999) Reforming Urban Water Systems:
Theory and Evidence from Developing Countries World Bank
Perez de Mendiguren, J.C. (1998) Water Policy hits the Rural Poor 24th WEDC
Conference
Sanda, K., Oya, K. (1998) Community Based Water Supply and Sanitation
Improvement 24th WEDC Conference.
Warford, J.J. (1997) Marginal Opportunity Cost Pricing for Municipal Water
Supply International Development Research Centre.
ENVIRONMENTAL RESOURCES MANAGEMENT DEPARTMENT FOR INTERNATIONAL DEVELOPMENT
B30
Waughray, D., Moran, D., Lovell, C. (1999) Using Contingent Valuation to
Help Evaluate the Benefits of Dryland Water Supply Project: A Case Study
from South East Zimbabwe. World Development. Vol. 26, No. 10.
Whittington, D., Lauria, D. T., Choe, K., Hughes, J.A., Swarna, V. Wright, A.M.
(1993) Households Sanitation in Kumasi, Ghana: A Description of Current
Practices, Attitudes and Perceptions. World Development, Vol 21, No 5.
Wittington, D., Okorafor, A., Okore. A., McPhail. A. (1990) Strategy for cost
Recovery in the Rural Water Sector: A Case Study of Nsukka District,
Anabra State, Nigeria. Water Resources Research, Vol 26, No 9.
Whittington, D., Lauria, D. T., Okun, D.A., MU, Xinming (1989) Water
Vending Activities in Developing Countries: A case study of Ukundu,
Kenya. Water Resources Development, Vol 5, No 3.
World Bank Water Demand Research Team (1993). The Demand for Water in
Rural Areas; Determinants and Policy Implications. World Bank Research
Observer 8(1) pp 47-70.
World Bank (1999) Competition in Water and Sanitation Public Policy for the
Private Sector Note No 188
World Bank (1999) Expanding Water and Sanitation Services to Low income
Households: The Case of the La Paz-El Alto concession. Public Policy for the
Private Sector Note No 178
World Bank (1999) Reaching the Urban Poor with Private Infrastructure
Public Policy for the Private Sector Note No 165
l.l INTRODUCTION
As part of the research investigation, a cost recovery survey was designed and
launched1. It was emailed during August- October 2000 to 800 actors in the
water and sanitation sector globally, and also disseminated via on several
WSS e-discussion groups. Water Aid and DFID helped to review the survey
instrument.
78 emails failed to deliver, creating a sample size of 722. Over 100 completed
surveys and a selection of written notes were received (a return rate of
approximately 10%), as well as a number of reports and background papers
which people also sent in.
In general, feedback was very positive to the survey, with 56 survey forms
completed to an adequate degree allowing comparative analysis. A database
was developed and the survey data inputted and subsequently analysed
between September 2000 - March 2001. An undergraduate dissertation for the
Scottish Agricultural College was produced from the results.
• What WATSAN organisations are doing about cost recovery in rural and
peri-urban areas?
1
A copy of the survey is presented and the end of this Annex.
Over a third of the WATSAN projects surveyed, are being carried out in
Southern and Eastern Africa. However, many WATSAN projects are also
being carried out in South East Asia, North Africa and the Middle East. The
majority of these projects are directed at 'low-income' households (between
$0-500/ year) in rural and peri-urban areas. The Development Agencies and
NGOs who replied serve more households at the bottom of this range
(,$200/year) when compared with replies from Development Banks and the
Private Sector (Figure 1.1).
Figure hi Income groups that organisations mostly serve through their WATSAN
projects (based on annual household income)
Income/'$< 200
Income/ $201 -500
Income/$501 -1000
• Income/$ 1001 - 4000
The scope of WATSAN projects varies for each type of organisation (Figure
1.2). However most organisations have a clear part of their programme
portfolio that concentrates on WATSAN projects, more than for water or
sanitation only projects. Interestingly, those Development Agencies who
replied, undertake slightly more of their WATSAN projects as part of wider
rural development than other organisations (Figure 1.3).
Figure 1.3 Organisations carrying out WATSAN projects as part of wider rural
development.
1.3.1 Meaning
The difference in outlook on this issue between the NGOs and Development
Agencies, and the Private Sector and Development Banks/ IFIs that replied is
noticeable. A larger number of private sector organisations and Development
Banks than other organisations consider that full recovery of costs is required
for a project to be financial sustainable.
Some Development Agencies and NGOs have begun to think along these
lines.
" Cost recovery is ensuring there is a system in place pr the collection of funds so that
communities can maintain the project" (NGO)
mo&M
MO&M + replacement costs
90&M + replacement costs of parts + interest
OO&M + replacement costs + interest + initial capital costs of project
• O&M + replacement costs of parts + initial capital costs of project + cost of expanding population
I O&M + replacement costs of parts + initial capital costs of project + cost of an expanding population
These differences between the private and public sector approaches to cost
recovery may differ for the following reasons.
For the public sector (NGO/ development agency), the financial remit for a
WATSAN project is:
• Meeting basic needs as stipulated in other government policies;
• Equitable use of subsidies; and
• Break-even, i.e., revenue generated equals the cost of supply.
"Ability to pay is critical, but providing affordable services to meet peoples' needs is
even more important."
"An appropriate level of service must be provided according to the demands of the
community. Willingness and ability to pay will play a role in this."
Historical practices were also said to play a role in the degree of cost recovery.
For example if other agencies have provided the service free of charge, a
history of non-payment and a subsequent lack of willingness to pay may
result.
(2) This list is by no means complete. Many other factors were also mentioned: - inappropriate project design, bad examples
created when other agencies fail to recover costs, social exclusion issues, land tenure issues and project expense
ENVIRONMENTAL RESOURCES MANAGEMENT DEPARTMENT FOR INTERNATIONAL DEVELOPMENT
For both water and sanitation projects, the flux of the population size is
deemed to be, the least important factor resulting in complications for cost
recovery.
The 3-5% rule has often been used as a measure of affordability for WATSAN
services. It has been considered a simple yet relatively accurate measure of
affordability that can assist decision-making. The survey revealed 28% of
those who responded still use the 3-5% rule, stating that as an initial tool it can
be useful.
However, there was also a realisation from other respondents that this rule is
too simplistic and the issue is far too complex to simply provide a blanket rule
for all situations.
In trying to find the most suitable way to measure affordability, some of the
organisations that replied have found interesting and somewhat unique
approaches. For example, a private sector organisation in South Africa, used
the amount of consumption of beer as an indicator of actual (effective)
affordability for a WATSAN service.
In the move away from using the 3-5% rule, most respondents saw demand
assessment techniques, of one form or another, as a useful tool in helping to
achieve financial sustainability in a WATSAN project.
"Demand assessment improves the chances of the project being financially sustainable
as compared to a supply driven approach". (NGO)
1.3.1 Meaning
'Demand responsive approaches need to reflect the present need and interest and
affordability of consumers or communities. Demand is dependent on health, social
and environmental values.' (NGO)
"Demand assessment involves assessing peoples needs and wants and comparing
these with the willingness to pay and ability to pay for the expected services"
(Private sector/consultancy)
However, many other factors to take into account were considered important,
such as:
1.3.3 Approaches
I Accurate
! I Feasible
I
y' / / /
When asked about charging for WATSAN, the respondents were asked to
consider what the connection charges should be, and how tariffs should be
structured. The survey revealed that free connection to water services were
not considered to be the best form of water and sanitation provision. In fact,
42% of respondents felt that users should pay equal amounts for their water
connections, while 37% considered a connection charge which reflected the
user's level of income was more appropriate as shown in Figure 1.7.
For both water and sanitation provision, once the connection charge is
established, the majority of organisations who replied believed that users
should be charged in relation to the volume of water that they use. Lifeline
tariffs for the lowest income households and increasing block tariffs were
considered to be useful charging structures (Figure 1.8).
B Water services
Sanitation services
Voumetric charge for Lifeline tariff for Increasing block tariff Fixed charge for time free Block Tariff for
all lowest income period for all all
households
Tariff Options
These factors within Table 1.2 are mainly associated with aspects of the supply,
but concerning the process a number of organisations stress the importance of
public involvement. For example, an NGO in India considered the payment
collection mechanism to be a success due to the people-based system of fund
collection, management, utilisation and maintenance. Along the same lines a
private sector company in South Africa stated that success was due to
transparency and a high level of public awareness and availability of relevant
information.
Others mentioned the financial arrangements for the projects, and one
interesting response, by a private sector company, was that availability of
flexible payment systems, helped to ensure a successful project due to successful
payment collection.
1. Name
jrft*
a
• • : * • •
g • rural communities;
contact detail* • peri urban/ slum communities; a
• small urban communities; a
3. Profession: (please circle) • larger urban communities a
• small businesses; a
Engineer Economist Social Anthropologist Accountant Other (specify) » industry a
responsibility • other (please specify):
5. Typ* of organisation (piease tick): .1}-Project Size"'•[; Please indicate in descending order [1«, 2-* «c) the most comn
••steesofWATSANprojecmyouworkon. ...
« Development Agency •
. NGO • Serving fewer than 50 households 'a
• Research/ Academic Body D 50-100 households a
• Private Sector/ Consultancy D
D
100 -500 households •
• Development Bank. more than 500 households D
Other (please specify): Do you tend to work with different numbers of households depending on the type ofWATSAN project customer you are
invoived with? (e.g. urban customers = larger project popukjtions, rural customers ~ smaller project populations}
Regional distribution of four WATSAN (if you operate in several regions, please indicate distribution in
projects descending order [1*, 2™1 etc]of your investment schedule or level
of activity
7. U.Wmtipes of project partner (if anyi do yoa generally tcortc with on four WATSAN projects?
• NGOs; •
a) Water only b) Sanitation only c) Water *t Sanitation • Donor agency; •
• Community organisations O
d) Pollution control and waste water projects; e) Water and sanitation as part of wider rural • Government parastatal. O
development projects.
• other (please specify):
Wbatistbctwk*Uncomep*rj>entmperyearUuUS$} that your WATSAN projects work within amuaUf
<pla>sectotUi?lfyo*optmtt trfffeftt tntmHtiemm *w»fc depending on different WATSAN projects, pkase
imliaiU in descending order the mmtctmmouimomt bands you work with.
20. WrichttilHamtasseismrnt techniques do you use? ' '"'' ; . ... " " : 7
IS: In general, what it the duration of the iletigu period tf the project* you art involvei with?: If you use more (iiUi cue approach, please indicate in descending order ( K 3"* most used etc} the demand assessment
• Under 1 year D
• Under a 2 year design period CD
« Over a 2 year design and implementation period CD a) focus groups using key informants/ stakeholders; CD
• other {please specify): b) FRA/community survey; CD
c) Revealed preference surveys - i.e. finding out about use of vendors / watercoping •
strategies;
16. What do you understand by the following terms? d) willingness to pay/contingent valuation surveys; D
e) data on national or regional household income distribution.
21. WMcb method, referred to In Q. 20 do you rUdt el&its peoples fimacial preferences most accurately?(pt?**:
circle) '":'~ •• , :
^ t ' '• ^v-.
17. What in your view arc the critical elements of a sustainable, water/sanitation project? If you think that there
are several elements that are critical, please indicate them in descending order H", 2»* most critical etcl c)
Water Sanitation
• Adequate availability of water to the user community over the CD Other:
project lifetime
• Fronimity of resources (water and sanitation) to user CD D
community;
• Developing education and awareness about project issues and CD a
equipment; f
referred to in Q.20, to youiKlieve it the iHOttfea'iblt we? of eliciting pfopmfinittctolt
a preferetuw,tfBen time atulresot»-cecomtraims?(plea& circle} f
• Encou raging c o m m u n ity management of the project; CD
a
• A stable political/ legislative environment
• Building something that people want?
LJ
a b) c)
• Building something that people will pay for?; D
• Incorporating community financing mechanisms; a a Other
• A stable community population; •
other (please specify)
23. What is your general opinion on trying to elicit people'sfinancialpreferences for water supply/sanitation
projects in mart depth? ;; \ : :^
a) Operation and maintenance costs (O & M); EZ] 30. If your answer to Q29 tens Yes, how is rfrfs done? (please tick or number in descending order)
b) O A M + replacement costs of parts; JZ3 • Population projections •
c) O t M * interest + replacement of parts; D
• Population and income elasticity information fll •
d) O 4 M + interest + replacement of parts +
initial capita! cost of project;
•
• Per head consumption from different types of water intervention •
e) O b M + interest + replacement of parts + •
initial capital cost of project + cost of Q
• Other:
expanding the population served by the
project in the future.
f) An approximation of long run costs (e.g.
average incremental cost/ m3)
37,
projects? (please cirtk)
25. Which type of water/sanitation cost recovery schedule, referred to in Q. It, do you most often toe/ would you
feel most comfortable with? ¥« No
a) b) c) d) . f)
32. Wlmt is this benchmark for water projects (litrcs/person/day)? Please circle
Further comments
10 20 40 100 120 +
33. What is the equivalent benchmark for the sanitation projects you are involved with?
26, Is Hit J-5W (of annual income spent on water and sanitation services) rule, a useful eo for gauging the
affordability of water supply and sanitation projects? (please circle)
Yes No
27. Do you use any other measures of affordability for your WATSAN projects? 34. Do tfouuse micro- financing/credit mechanisms to help with the payments of water?
Yes No
With Tuhfft degree of success ?
28, How do you determine anv Uvtl of sabtidy that is given as part ojSfour projects?
For eampite (piease tick)
• per capita grant amount for each project participant
• block grants as a percentage of capita] cost
• via cross subsidisation systems • (1)Th»s meana Ihe use at a facUw thai riinvi there4aliortthipbetween what h»ppen» lo ronawnf^inn as innwiK irwira$e» Fr
elaatldty O.i means that if income doubled, consumption would increase hy W V
35. Do you use micro- financing/credit mechanisms to help with the payments of sanitation?
Yes No
39. Which ontof the fallowingataientettsis closest to youVpietm about household water supply in then
36, What factors would you sag * t o at cmmmmtcbimlmsiR WATSAN projects are projects yoa are mefHiaolveJuHift?
successful?
Please tick one
« Everyone has a basic right to water which m e a n s that people should receive water O
for free;
• It costs money to supply water and payment for each unit of water consumed ED
should reflect some of the financial costs of supplying each unit of water;
• Everyone has a basic right to a minimum level of free supply, thereafter users Cl
should pay something to help towards the costs of supply
• Higher income households should pay more for water consumed than lower (13
income households to help towards the costs of supply
• It costs money to supply water and payment for each unit of water consumed D
should reflect ail of the financial costs of supplying each unit of water
• Water is a scarce resource and as such prices should reflect the availability of water. Q
'•37. Is your view, which tfpe of ittstitutiOM support WATSAN project user paftt Other (please specify):
Yes No
38. What do you feel are the key institutional requirements at both the municipal and the macro level for •
WATSAN project/programme success, particularly in relation to financial sustatnabiltttfi Please expand on your msuier below
Sanitation:
42. What in your optmtnt is the mott criticalfatior determining whether cost recovery is feasible?
For water?
A water supply f, sanitation project/prop •HUM can include nuter installation and tit provision of
access to a wider water and sewage nehco *. In these circumstances a volumetric tariff out be
calculated, ' 5:
For sanitation?
45. Which of the folloumig statements is cloiest tojfour organisation's views about paging for
connection to the nete network? ' ''.''" • '
Furtlier comments:
44. What, in your opinion or experience, are the difficulties associated wish the principle? of cost recovery for <K. Which tf the following ttatmentt is clotat to your organisatloit't views about ttrvctiaittg* water
^WATSAN projects? Phase rank in denuding order of most to teasi difficult issue. service* tariff in relation to comimtpHonl (please rankfrcrnmost to least damblt tariff structure)
Water Sanitation
• Free block tariff for all;
• a
Why should people pay? D •
•
Lifel i ne ta rifHor lowest income households;
Increasing block tariff; • a
D
Most people cannot afford to pay a realistic amount that will help towards
• • a
recovering costs.
If projects are only built for those who can pay, this is an tin-equitable approach for
•
•
Fixed charge per time period for all;
Volumetric charge for all; •a a
WATSAN development. • • Other
Though they might be willing, people often don't have the cash to pay a regular
water charge •
It is difficult to put money values on the non cash work and time contributed by
households or communities as part of the construction or O&M of the project. How
do we avoid double charging these people? D
Getting people to pay in principle is not the problem - engendering trust and
organising the routine collection and the secure management of revenues is a much
bigger obstacle; D *7. From your htowleigt/experience, ioet taking a demand responsive approach to WATSANprOfect
It is sometimes difficult to identify and explain exactly what is being charged for d e e l g n r ^ U i h ^ h ^
and to whom?
D important?
There can be an inconsistency in promoting a system of cost recovery for the
scheme, when compared to the availability of other grant sources and/or central
government approaches;
D
Households have to see an improvement in service first, before they start paying for
rt. This takes time and delays the recovery of any costs.
a
It is unrealistic to expect communities to pay on the basis of opportunity costs for
4$. Do JOB thittk that all We agenda tmelred tiiWAtSAN projects funm common approach to (or
d ^ ^ d ^ d f ^ i W A T S M f t r S U } 4 t h i il Thank you for your time and patience in filling out this questionnaire
49.How is any jktm»& assessntcfit itformotivn <m WATSAN projects that ts gatltend bf fonr • Name of Project
« Agencies Involved
• Location
• Brief Description
SO. <)f the following, wUdfii^fff consider the four most relevant attributes oj successful rural Water-'
ami sanitation projects? PLEAStMARKl 'MOST IMPORTANTTO*" LEAST IKW60^J[.S. \ , '•
Water'' j '. " Sak&fa* ;J<,;. "^SSfc.
Distance to Household (coiweitiemt) Q • Convenience '•%•. ' .• Q •'••—• Would you be happy for this research project to contact the project and find out more?
D .: • Ability to upgrnie ':• %•. Q
D .•• • • Cleanliness :?,' D
Admmisiering agency Q a
(operation, maintenance mid charging) • Contact Details of Project Manager
Community managed
Q
:
t
OTHER (please specify)
Given what you have said in this survey, what are four opinions OH your organisation's approach to cost
recm^, for exampbtlx difficulties intvlveJiu achievingfinat^lmtirimim
improvements in how to So so: What would you Itke to see from research in thit area?
Annex D
1.1 BACKGROUND
A paradox in the cost recovery debate is the observation that poor households
will frequently outlay a considerable percentage of monthly income to secure
water from alternative sources when regular supplies are unavailable or
substandard. These observations, arising from revealed value studies of time
spent collecting at informal sources, and household spending on water
vendors, give rise to an elementary question. If households are prepared to
make significant economic sacrifices in pursuit of alternative services, how
might it be possible to present recipients with a the service they actually want
and are prepared to pay for? The same revealed value surveys substantiate
the view that the 5% rule is inappropriate for planning and that appropriately
design and delivery could unlock similar household commitment in terms of
willingness to pay.
Dl
seems appropriate at this point to investigate this information for general
lessons about the drivers of household preferences and to see whether it is
possible to say anything about the key relationship between income and
willingness to pay.
1.2.1 Method
But by stacking the data provided in the individual valuation studies used in
this meta analysis, we investigate the relevance of the statistical function that
emerges from the analysis of the analysis in each individual study. In other
words we attempt to derive a meta function that explains the levels of WTP
across countries, and which can be borrowed to assess the likely WTP levels
needed for cost recovery.
i Both are measures of central tendency, but the median is the amount of money that just over 50% of a sample population
would be WTP.
What is estimated in this statistical analysis are the relationship between the
variables them as described by the coefficients (f, b, Q, fc). These coefficients
explain the causality and magnitude of the influence of these other variables
on WTP. Effectively they tell the story of how the data collected explains the
WTP responses collected in the CV section of the survey.
To proceed from a single study result to a meta analysis, we then take from
each individual study the estimated mean (or median) WTP for that study and
the coefficients from the specific study bid function to compile a new data set
that can be used to estimate mean or meta coefficients. The idea then is that
this reanalysis of this data set leads us to a mean function that can be used in a
generic predictive way to make general inferences about the relationship
between WTP and key variables. For example, assuming a meta function of
the generic form of equation (1) above. All we would have to do to predict the
level of WTP for any project in a particular country and site, is to take the
estimated mean coefficients and plug in the mean variable values for the
variables
HQ, Y, QP, and SE. Thee data will generally be observable for the site of
interest.
1.2.2 Data
From the exiting body of valuation studies, Table 1 was drawn up to identify
the variables that partly explain WTP. The table details the direction of
assumed causality
The average sample size was 180, with the smallest sample comprising 13
households and the largest, 968. The wide range of inter- and intra-study
sample sizes is likely to cause clustering and generate a significant amount of
apparent variability in the WTP bids (Deaton, 1997). hi surveys of rural areas
where the clusters are often villages, so that households in a single cluster live
near one another and are interviewed for the same study, data may be
produced where observations from the same cluster are more like one another
than are observations from different clusters. This may be as a result of
'neighbourhood effects' (Deaton, 1997) where those who live near one another
and thus become more or less uniform within a village copy the local
eccentricities. Sample villages are often widely spatially separated, their
inhabitants may belong to different castes or tribes and they may have distinct
occupational structures1.
1
The main problem with clustering is that the sample variance is inflated above what it would have been in the
independent case but where multiple observations are obtained from the same environment, random effects can sometimes
be controlled for in ways that would otherwise not be possible (Deaton, 1997). The cluster design of the data should not be
ignored as the standard formulae for variances of estimated means will be too small. This is as much an efficiency issue as
an accuracy issue. Since the error terms in the in regressions are correlated across observations, OLS regression is not
efficient within the class of linear estimators and some other linear estimator may be more appropriate (Deaton, 1997, p74).
1
It is assumed that bidding procedures can only be performed when CVM is used.
1
Classifications used here are based on the World Bank World Development Report of 1995.
2
Sweden has an index of 25, the United States, 40 and the United Kingdom, 36.1 (World Bank, 2000).
Figure. 1 Histogram showing the distribution of GNP per household across the 14
countries represented in the studies included in the meta-dataset
30 —
Mean = 316.72
Median = 242.25
Std. Dev= 206.922
n=13
r
10 —
0 —
500 1000
Monthly GNP per household
A simple plot of log WTP for water services against GNP per capita further
illustrates this finding (Figure I). It shows the positive relationship between
aggregate income and WTP but the regression correlation co-efficient
indicates that the model does not fit the data very well (the R-squared is only
29%). Most of the observations display relatively low levels of GDP but there
is wide variability about the best fit. The interpretation of the regression
equation is that a ten dollar increase in monthly GNP per household will raise
the WTP bid by 0.032 percent1. Thus a household currently paying US$1 per
cubic metre of water would be willing to pay 32c more if GNP was raised to
US$100 per capita. From the plot, one can also see that at low incomes some
households appear to be prepared to pay relatively high amounts for
improved water.
1
Note that GNP has not been logged therefore the interpretation is one of a semi-log relationship (see
Santos, 1999)
Y = -1.19560 + 3.29E-03X
R-Sq = 27.6% (F= 33.93, p = 0)
3 —
Z —
1 —
•2 — • ; • • •
.3 -
.4 _
I 1 i i i i i
0 200 400 900 600 1000 12K
Almost one third of the studies were conducted in sites located in arid
regions1 and only twenty per cent of the observations assessed the demand for
water of households living in urban areas. The difference between the average
WTP amounts across different features of the study site is statistically
significant for all the tested features of the study site at a 1 % (Table 9)
The range of monthly household incomes across the studies spans from
US$8.17 per month (constant 1995 prices) to US$212.86 with a standard
deviation of US$60.13. Reported monthly household incomes of household
1
Regions where rainfall is below 500 nun per year (Times, 1997).
The range in income may account for some of the variability in WTP bids. The
mean monthly WTP for water services is US$2.1, with some samples willing to
contribute as little as 3c per month towards the proposed service and others as
much as US$21.62, more than twice the mean income levels of some samples.
From an analytical viewpoint, variation in WTP estimates is desirable if it can
be associated with variation in some independent variables. However, as
mentioned earlier, this large variation could be an indicator of unreliable
water benefit estimates. The WTP amount for the households in Ecuador was
over US$21 per month and suggests that the estimate of the benefits of water
to households in this rural community has been highly overstated, especially
since it is mentioned (Hardner, 1996) that the community supports a non-cash
based economy. Mean WTP as a percentage of mean monthly household
income is around 4% which would appear to substantiate the five-percent
affordability rule. However, when one examines the shape of the distribution
of reported sample income (Figure 3) and compares it to the distribution of
stated WTP (Figure 4), it is once again shown that this rule-of-thumb may not
be generally applicable as the shape of WTP does not closely mirror that of
income.
1
It is possible that some of this divergence arises as a result of measurement error. Some of the studies
reported income levels based on proxy indicators such as the number of visible assets (such as radios,
refrigerators and television) observed in the homestead. In rural areas, wealth is not necessarily measured
in this way especially where the infrastructure necessary to operate consumer durables is not accessible.
Households may also have given incorrect estimates of their income levels where these fluctuate with
seasons (being highest immediately after the harvest season ceteris paribus).
20 -
10 -
0 - Wzm
I I I
0 100 200
Mean monthly household income (1995 US$)
50 -
£•
V
40
30 —
20 —
10 —
0 —
~
ii
1
0
1
10
Monthly Household WTP (1995 US$)
20
1
D9
Figure.5 WTP as a function of monthly household income
Y = 0.537461 + 2.08E-02X
R-Sq = 12.8 % F = 13.03 (p = 0.01)
33
100 200
Mean monthly household income (1995 US$)
Prom Figure 5 it can been seen that there is a positive relationship between
WTP and income but the relationship is poor (R2 is 12.8% and this fit is
significant at the 5% level). Moreover, one can clearly see the variability in
stated WTP about the best fit line and this variability appears to increase with
income suggesting that demand for water is elastic at higher levels of income.
Intuitively this would make sense as where income is low, households are
likely to use water primarily for meeting their 'basic need' requirements and
thus will pay a relatively high price for water below that margin. At higher
levels of income, households may be using water for non-essential purposes
(such as for the operation of consumer durables or for recreational purposes).
Consumption above the level deemed 'adequate' for a reasonable standard of
living is therefore more likely to be sensitive to price. Water consumption thus
displays the properties of a normal good with diminishing marginal returns at
higher levels of consumption. Since price and income are related (as the price
of one good increases, the amount of income available to spend on other
goods and service decreases, ceteris paribus) it can be postulated that water
demand becomes more elastic as income increases and the use of water
extends beyond that required for basic needs.
With every ten dollar increase in income, households are then willing to pay
20c more for each unit of consumption. Logging the relationship between
WTP and household income allows one determine the income elasticity of
D10
demand for water (the rate of change of WTP with respect to income) The log
specification of the relationship yields the equation:
It shows that for every one percent increase in income, WTP increases by 0.54
percent. Thus, if income were to increase from US$10 per month to US$20 per
month (that is, a 100% increase), then WTP would increase by 54 percent,
ceteris paribus, suggesting that water is a relatively inelastic normal good.
Over eighty percent of the households in the meta-dataset were offered a new,
as opposed to an improved, service and 56% of the bids were for private
connections. Just over one third of the respondents were told that the
recurrent costs for water would be worked out on the basis of their
consumption levels and only two thirds were told that an upfront installation
and connection charge would be expected from them. The median WTP was
surprisingly higher among respondents offered an improved rather than a
new water delivery system (Table 11). This may reflect current attitudes
towards water provision particularly in countries such as South Africa where
those currently using unimproved or traditional sources are often population
groups who have been politically marginalised in the past and believe that it
is now up to their now more representative governments to provide them
with new services at no charge. It could also indicate a lack of awareness
among respondents about the added benefits that a new service would bring
over an upgrading of the systems that are currently in place. The results of the
Kruskal -Wallis test (Table 12) do not, however, substantiate at any
conventional significance level the difference in median WTP values between
respondents offered new over improved systems.
Dll
that they have accounted for the opportunity costs associated with collecting
water from distant sources. Suprisingly too, households were prepared to
pay, on average, more for a flat rate than a volumetric rate. This may have
come about if they did not understand how the volumetric system works or if
they were opposed to paying the additional capital costs associated with the
installation of an effective metering and charging system which requires
efficient administration and management (Hazelton, 1997; Barrett and Sinclair,
1999). Households that were told that they would have to pay an upfront
connection charge also provided lower median WTP values than those who
were not. This may reflect the seasonal nature of income levels among
agriculturally-based rural households or households relying on income from
migrant labour. Often the cash will not be available as a lump sum, but if the
payments are spread then households can factor this into their monthly
budgets. The differences in WTP are significant across these factors at a ten
percent or better level of significance (Table 12)
Regression Results
The structure of the data used in the meta-analysis is quite complex. WTP
values are generated by different studies, carried out in different geographical
locations and using different valuation formats. Following the lead of several
earlier meta-analyses of environmental evaluation studies (Boyle et al, 1994;
Espey et al, 1997; Alston et al, 2000 and Poole et al, 1998), the basic relationship
between WTP and the explanatories can be specified as:
WTP = bo + Xb + v
1 A more accurate method of testing for the necessity of transformations is to use the Box-Cox test. This
involves scaling the observations on WTP so that the residual sum of squares (RSS) in the linear and log
models are rendered directly comparable (Dougherty, 1992).
2
The log specification of the model implies that the dependent variables should be interpreted as (1) an
elasticity when the specification is log-linear (income) and (2) as a percentage effect on WTP of a small
change in the independent variable when the specification is semi-log (all continuous variables except
income). The parameter estimates for dummy variables are interpreted in terms of percentage effects on
D12
The logged WTP values were then regressed onto the explanatory variables
using ordinary least squares (OLS). The regression output is displayed in Table
13.
From the results in Table 13 it can be seen that over half of the data is
accounted for by the model (the R-squared adjusted for degrees of freedom is
51.3%) and the probability that the goodness-of-fit arose purely by chance is
very low (this is confirmed by the F-statisric which is significant below the one
per cent level. Six of the explanatory variables are significant at the five
percent level and a further two at the ten percent level. It is interesting note
that not all the variables behaved as anticipated and to observe that the study
site and design features are significant factors in explaining stated WTP. The
results of the regression are explained more fully below.
The most significant factor influencing WTP is the location of the study site -
that is, whether the sampled household was located in an urban or in a rural
area. It was expected that bids would be higher in urban than in rural areas
because of correlation with income and education levels and households
living in close proximity may have learned of the benefits of an improved
supply from their neighbours. They are thus more likely to understand the
nature of the services being offered to them and are therefore more likely and
willing to pay a higher amount for a new service. Also, households living in
urban areas are more likely to be forced to pay for most of their water
consumption or, where unimproved sources are available, they are likely to be
heavily polluted.
The second most significant factor is climate, although the negative co-
efficient is contrary to expectations. Intuitively, one would expect households
living in arid climates to be willing to pay more for their water supplies
because of the greater amount of time needed to collect water from naturally
occurring sources. It can, however, be counter-argued that households living
in low rainfall regions will already have invested large amounts of money in
storage tanks and may have evolved sophisticated coping strategies (Nyong
and Kanaroglou, 1999). Where large investments have already been made,
households may be unwilling to allocate more income towards piped services.
Studies carried out after the publication of the NOAA guidelines also appear
to have a significant influence on the WTP bids. This may be because
respondents were provided with enough information (which they
understood) to make an informed decision about the amount that they were
willing to pay for the proposed service. The NOAA guidelines also
recommend the use of closed-ended procedures which have consistently
shown to produce significantly greater mean values than open-ended ones,
possibly as a result of respondents' uncertainty in answering the latter
(Bateman et al, cited in Brouwer et al, 1997).
WTP of a ceteris paribus change from 0 to 1 in each particular variable (Halvorsen and Palmquist, 1980; cited
in Santos, 1999).
D13
Neither the elicitation procedure nor the conduct method were significant at a
ten percent level although the co-efficients suggest that using a stated
preference approach would raise the WTP estimates by 66 percent with
respect to revealed preference approaches and using an iterative procedure
would lower estimates by 55 percent compared to using an open-ended
format. It is unclear what effect an iterative bidding procedure would have on
the responses a priori as using a bidding game provides respondents with
benchmarks upon which to base their value estimates. Respondents who are
faced with the interviewer might be tempted to inflate their responses in an
attempt to 'comply' with the perceived expectations of the enumerator but
using an iterative bidding game gives respondents an idea of the true
expected value of the WTP bid thus reducing the likelihood of 'wild'
estimates. The dichotomous choice format was developed to increase the
incentive-compatibility of the valuation question (Brouwer et al, 1997) and
matches the way consumers make choices in the market, giving respondents
the proper incentives to reveal their 'true' preferences. The dichotomous
choice format does, however, may induce what has been called 'starting-point
bias', that is, the initial bid amount offered to respondents may influence their
response. Since there is no consensus on which format (open-ended or closed-
ended) is less biased, the divergence between formats is interpreted as a
reliability problem.
Perhaps the most surprising result was the negative coefficient for education
levels (significant at the five- percent level). One would expect that
households with more years of schooling would have learned of the private
and social benefits of improved water consumption. Furthermore, more-
educated households are more likely to have members engaged in productive
employment or to have a higher probability of finding employment, implying
that the opportunity costs associated with collecting water from communal
sources are also higher. It may, however, be the case that more educated
households who are aware of the economic and social benefits of improved
water services, will already have invested in higher quality, more convenient
supplies and thus do not see the need to spend more money on improving
supplies when they are satisfied with their existing arrangements. Whittington
et al (1992b), based on an analysis by Hoehn and Randall (1987) note that one
could expect that applying CVM to rural households with limited education
would increase the frequency of incomplete responses and therefore
potentially increase the importance of strategic considerations linking their
informed values to their stated values.
In order to assess whether there was any correlation between income and
education, education and location and income and location, the Pearsons
correlation co-efficient was calculated for each of these relationships. The
results are presented in Table 14.
From the results it can be seen that there is no significant correlation between
locality and education although it is interesting to note the negative co-
efficient which suggests that education tends to be higher in rural areas than
D14
in urban ones. The small number of studies that were conducted in urban
areas may be a cause of this counterintuitive result. The high significance of
the other tested relationships suggests that there is a positive relationship
between income and education and also between locality as income. This is as
one would expect.
Although the goodness-of-fit of the full model was reasonably high (R-sq. is
59.4%), a stepwise procedure was run to see whether the fit could be
improved any more by sequentially removing those variables having the least
explanatory power and then finding the combination of variables that yields
the best fit. Two procedures were run, one in which all the variables that were
significant at the ten per cent level or better were forced into the model and
then another using all variables significant at the five percent level or better.
The results of these procedures are displayed in Appendix ??. Neither of the
procedures improved the goodness-of-fit of the model and the best
combinations of explanatory variables included all the variables that were
used in the full model. Thus all tested variables add some explanatory power
to the model even if only through second-order interactions.1.
1
Omitting variables can also introduce bias. If explanatory variables that should have been present in the
regression are omitted, and if these omitted variables are correlated with any of those that are included,
then the co-efficients on the included variables will be incorrect (Koop, 2000). Nevertheless, the counter-
argument can be made for using as few explanatory variables as possible. It can be shown that the inclusion
of irrelevant variables decreases the accuracy of the estimation of all the co-efficients (Koop, 2000).
D15
deviations of the errors differ across observations. There are three main
reasons why this should occur:
(1) variation within studies due to measurement error (inaccurate recording
of information by enumerators and mistakes in the calculation of mean
sample values);
(2) inter-study variation (as a result of the diverse methodologies employed);
and
(3) sample size discrepancies (mean WTP bids are likely to be more
representative of the total population where the sample size is greater. The
observation sample sizes across the studies in the meta-dataset ranged
from 13 to 968 and thus it could be argued that greater weight should be
attached to the values obtained from the larger sample sizes.
Poe et al (2000) suggest that instead of treating each unique value as an equally
weighted observation, each study could be given equal weight so that a study
with twelve observations would have equal weight as a study that provides
only one or two observations. Adopting an equal weight approach across
studies dilutes the effects of additional within-study variations (Poe et al,
2000). Since each observation was treated as an independent sub-sample it
was not considered necessary to adopt such an approach.
Commentary
Amid growing evidence that many urban and rural communities are willing
to pay more than the prevailing rates for water to ensure a better or more
reliable service, it is clear that there is great potential for cost recovery in the
water services sector. In order for full cost recovery to be a viable target in
water project planning, however, it is important for planners to understand
the multiple factors that influence demand for each level of service and for
communities to be provided with the services that meet their particular
preferences and for which they are willing to pay.
In this light, project planners and policymakers recognised the need for the
development of research tools to assess households' WTP. The CVM is by far
the most widespread in the water sector as it is believed to offer the most
reliable estimates for new services when it is carefully designed and
1
It was mentioned before that many studies did not indicate the 'scope' of their studies and so it is difficult
to assess the representativeness of the sample, ceteris paribus.
D16
administered. In the last decade, a substantial body of work on WTP for water
has accumulated and a wide range of WTP estimates has emerged. Traditional
narrative reviews often based on a core selection of studies have attempted to
account for the variability by noting that the different values attached to water
by communities and households reflects their environmental, socio-economic
and cultural circumstances. Allowing for these differences among
respondents, the conventional wisdom became that households would be
willing to pay for new service as long as the costs did not exceed 5 five per
cent of household income.
In this paper, an attempt has been made to deconstruct the CVM results
obtained from the wide-ranging body of literature on WTP for water in
developing countries by complementing a qualitative descriptive analysis
with a quantitative meta-analysis. An integral part of the meta-analysis is the
assembly of the studies and the creation of a database using the information
provided in the studies. In order for a full analysis of the reasons for
variability in the estimates to be made, it is important that results are
incorporated from as wide a range of studies as possible, as the results of both
the qualitative and quantitative analyses here reveal how the quality and
characteristics of each particular study, in terms of design, elicitation and
study site, have significant first-order effects on respondents' stated WTP. A
comprehensive analysis of the literature can thus provide a basis for
understanding why WTP values differ among studies. By accounting for
systematic differences arising because of different study methodologies,
enumerator eccentricities and recording techniques, and random calculation
error, closer estimate of actual WTP can be obtained.
The results of the meta-analysis show that demand assessment for cost-
recovery is a more complex issue than commonly believed. For water policy
and planning to proceed soundly, actual policy benefits need to be weighted
against costs as part of the process of evaluating this policy. As far as public
preferences for water services are recognised as a policy- or project-evaluation
criterion, formal cost-benefit analysis provides a useful tool to make these
cost-benefit comparisons in an explicit and straightforward way. In cost-
benefit analysis, the benefits from water services are measured as WTP for
water service improvements.
These findings underscore the need for a greater understanding of the nature
and determinants of demand. WTP studies clearly provide information which
D17
could help in technical planning and link investment decisions with pricing
policies. It is recognised that it is impracticable to carry out a new study for
every single policy but benefit transfer offer a potentially practical solution.
However, the transfers need to be informed by a sound set of input data. In
compiling the data for the meta-analysis, it became clear that, despite the
NOAA guidelines and the influence of experienced research teams in
conducting household surveys, there are still wide discrepancies in the quality
and level of detail that is recorded in the report.
This argument is strengthened when one sees that a significant degree of the
variability in WTP estimates across the studies is attributable to the particular
methodologies that were used to solicit responses. Nevertheless, the wide-
ranging nature of income levels both within and between studies suggests that
it would be unreasonable to expect households to pay for water services that
they either do not want (because they have already 'solved' their water
problems and are satisfied with their current supplies) or that they cannot
afford.
The high failure rate of so many water supply projects in the last decade
demonstrates the drawbacks of making general assumptions about what
people are willing or able to pay for their resources. Despite the incomplete
nature of the meta-analysis that was conducted here, it is clear that actual WTP
for water is a function of the locality of the household, the natural availability
of the resource, knowledge about the benefits that higher service levels will
bring and the amount of money that has already been invested in improving
supplies. It could reasonably be argued that each of these factors is related to
income in some way and therefore that it is ultimately income that drives the
demand for water.
Once further analysis has been made, the potential for benefit transfer can be
more reliably assessed, but what is clear from the studies collected for the
meta-analysis, is that there needs to be some standardisation in the way in
which CVM surveys are carried out or, at least, researchers need to make
explicit their assumptions and inferences. Although it is understood that CVM
studies are undertaken to serve particular research agendas and thus results
are often presented selectively, if water project planning policy-making is to
D18
be based on the outcomes of these surveys then a higher level and quality of
reporting needs to be encouraged. One way of improving the reliability and
validity of the meta-analysis outcome, would be to access the full range of
information collected by the authors, where this information is still available.
The database that was assembled here is ripe for further analysis and
suggestions have been made about where further progress in the analysis
could start. The use of a generalised linear modelling procedure that accounts
for heteroskedasticity and can be implemented on a multi-level scale to allow
for different study and within-study sample sizes and for the non-
independence of all observations within a study, would permit a more
rigorous analysis of variation and a more reliable estimate of the benefits of
improved water to the respondents themselves.
A contingent market defines the good itself, the institutional context in which
it would be provided, and the way it would be financed. The situation the
respondent is asked to value is hypothetical (hence, contingent7), although
respondents are assumed to behave as if they were in a real market.
Structured questions and various forms of bidding game' can be devised
involving yes/no' answers to questions regarding maximum willingness to
pay. Econometric techniques are then used on the survey results to find the
mean bid values of willingness to pay.
Over the last two decades interest in CVM has increased for a number of
reasons (see Carson 2000). First, a stated preference approach is the only
means available for valuing 'non-use' (or passive use) values, such as people's
existence values for a unique natural habitat or wilderness area. Second, the
evidence available suggests that estimates obtained from careful and
well-designed, properly executed surveys appear to be as good as estimates
obtained from other methods. Thirdly, the design, analysis and interpretation
of surveys have improved greatly with advances in scientific sampling theory,
benefit estimation theory, computerised data management and public opinion
polling.
For a detailed review of the Contingent Valuation Method, see Mitchell and Carson (1989) and
Carson (2000).
D19
In developing countries CVM has been applied widely in the field of WSS
planning. The complex range of WSS benefits means that the method is an
appropriate form of inquiry for eliciting community preferences for as yet
hypothetical supply options. But the hypothetical format of the approach is
also the root of existing criticisms of the method. Interestingly the method is
convergent with the DRA agenda, which requires the elicitation of household
willingness to pay.
D20
Table 1: The determinants of demand for new and improved water supplies
D21
CROSS-REFERENCE FACTORS AND FACTOR LEVELS
Expected
Variable Description Hypothesis
Sign
D22
CROSS-REFERENCE FACTORS AND FACTOR LEVELS
Expected
Variable Description Hypothesis
Sign
Water Supply Characteristics
An index to express the
degree of satisfaction of
Where households are satisfied with their
the household with its
current quality of supply (physical attributes,
existing services in terms
Satisfaction with convenience and reliability) they will value
of reliability, physical
existing source new or improved supplies less highly than a
quality of water and
household that is dissatisfied with current
convenience. Sometimes
service levels, ceteris paribw,
recorded as a referendum
dummy (ie yes/ no)
A continuous variable to the price of the new/improved scheme and
measure the amount of perceived opportunity costs saved with
Existing money that the household respect to added convenience and reliability
expenditure on is currently spending on will largely determine the household's
water obtaining water preferences for a new level of service.
Usually expressed as a
dichotomous variable to
show the ability of the
The greater amount of storage space the
household to collect and
house has to collect water and store it for long
Storage Capacity store water in times of
periods of time, the less valued a convenient
relative plentiful supply
source of water is likely to be, ceteris paribus.
(during the rainy season)
(1 = capacity to store >
500 gallons of water)
A dummy variable to The presence of vending activities has an
indicate whether or not ambiguous effect on WTP. Where proposed
the household purchases services can match the convenience and
water from vending quality of vended supplies then and where
distributors or kiosks to the price of the proposed project water is
supplement other similar to that of purchased water, then WTP
possible supplies (1 = is likely to be higher. WTP is this likely to
Vending vendor is primary source) relatively high for private connections, ceteris
paribus.
Where the proposed service offers no added
convenience and is priced higher than current
supplies, then WTP is likely to be lower,
ceteris paribus.
Variously expressed as
either a dichotomous or a
continuous variable to
The further away from the water source the
indicate the distance in
Distance from household is, the more likely a higher value
metres of the household
Source will be placed on more convenient and
from its primary source of
reliable services, ceteris paribus.
water (1 - distance is >
200m where a dummy
variable)
A continuous variable to
measure the time (in
minutes) that the One would expect WTP for more-
household spends in conveniently located and reliable supplies to
Time spent
collecting water each day be greater, ceteris paribus, where the amount
collecting water
of time spent collecting water is currently
large
D23
A dummy variable
indicating whether the
Higher values are likely to be placed on
service offered was for a
Characteristics of piped in-house private as opposed communal supplies
service offered because of the added convenience and
connection or a public
quality.
service (1 - in-house
connection)
Attitude to water provision
Dummy variable to
indicate whether the
respondent believes that
• Households believing that water is a basic
the provision of water
Government human right and that its supply is the
responsibility responsibility of government are less likely to
responsibility of the
invest in improving supplies, ceteris paribus.
public sector (1 =
government should
provide services)
Dummy variable to
show whether the
household agrees with • Where households agree with the concept of
the concept of being being charged for water, their stated WTP for
Metering charged for water or + water is likely to be higher than those who
whether it should be believe that water should be provided free of
provided free of charge charge, ceteris paribus.
(1 = agree with charging
principles)
D24
Year of Number of
Author Publication Details Study Site
Study Observations
McPhail Land Economics, 1994 1990 Tunisia 1
Singh,
Ramasubban, Water Resources
1988 Kerala, India 3
Bhatia, Briscoe, Research, 1993
Griffin and Kim
Griffin, Briscoe,
Singh, World Bank Economic
1988 Kerala, India 3
Ramasubban Review, 1995
and Bhatia
DFID/RWSG-SA report
on a Workshop on
Ramasubban Willingness to Pay for 1988 Kerala, India 3
Drinking Water Supply
and Sanitation, 1997
Development and Change, Rajasthan,
Reddy 1993 6
1999 India
Anand and Journal of International
1
1996 Madras, India
Perman Development, 1999
DFID/RWSG-SA report
on Willingness to Pay
for Drinking Water
Supply and Sanitation,
1997; originally part of
a study on WTP for
Improved Water
Asian
Supply, Sanitary
Information
Latrines and Sewage 1996 Punjab, India 1
Marketing and
Social Research Systems for Rural
Households
commissioned by the
Punjab Public Health
and Engineering
Department, Patiala
and the World Bank,
1996
DFID/RWSG-SA report
on Willingness to Pay
Vaidya for Drinking Water 1996 Baroda, India 1
Supply and Sanitation,
1997
Altaf, Water Resources Punjab,
Whittington, 1988 12
Research, 1993 Pakistan
Jamal and Smith
Altaf, UNDP-World Bank
Punjab,
Whittington, Water and Sanitation 1988 12
i (lTvlOVU-ll
Jamal and Smith Report, 1992
Third World Planning Punjab,
Altaf 1989 3
Review, 1994 Pakistan
Habitat International, Gujranwala,
Altaf 1990 2
1994 Pakistan
North and Water Resources 1978 Bicol, 3
D25
Griffin Research, 1993 Philippines
Year of Number of
Author Publication Details Study Site
Study Observations
Bohm,
Water Resources
Essenburg and 1987 Philippines 5
Research, 1993
Fox
Briscoe, de
World Bank Economic
Castro, Griffin, 1987 Brazil 9
Review, 1990
North and Olsen
Whittington, Economic Development
Briscoe, Mu and and CulturalChange, 1986 Southern Haiti 2
Barron 1990
Paper prepared for the
Limon, Costa
Unit for Environmental
Aguilar and Rica 2
Economics, 1993
Sterner Guanacaste,
Gothenburg, Sweden,
Costa Rica
1995
Water Resources Bulletin,
Hardner 1995 Ecuador 1
1996
Table 4
Mean
Variable Variable Description (Standard Minimum Maximum
Deviation)
Study Design and Site
Characteristics
1 • study was carried out 0.2747
Date
after 1993 (0.4488)
1 = Whittington was on
research team or the author
0.5934
Authors of the study had worked
(0.4939)
with him on an earlier WTP
study
Sample Size Number of households 179.9 13 968
D26
Mean
Variable Variable Description (Standard Minimum Maximum
Deviation)
interviewed in each (193.9)
observation
1 = middle-income; 0 = 0.3187
Region 0
lower income (0.4685)
1 = arid climate; 0 = humid 0.3187
Climate 0
climate (0.4685)
0.2143
Locality 1 = urban; 0 = rural 0
(0.4024)
1 * contingent valuation
0.8956
Conduct Method methodology/stated
(0.3029)
preference
0.7363
Elicitation Procedure 1 = bidding game was used
(0.4431)
1 =• respondent was not 0.9754
Time-to-Think
given time to think (0.1492)
Household and
Respondent
Characteristics
Mean number of people 6.848
Household Size 3.16 12.010
living in the household (2.102)
Mean number of years of 6.135
2.128 9.000
Education Level schooling of the household (1.985)
Mean monthly household 73.88
8.17 212.86
Household Income income in 1995 US$ (60.13)
Log of monthly household 3.837
2.1 5.361
Log Income income (1.072)
Water Supply
Characteristics
0.8131
Service Type Offered 1 = new; 0 = improved 0 1
(0.3766)
Level of Service bid 0.5636
1 = private connection 0 1
for (0.4908)
1 = respondents were told
Recurrent Cost
that the charging system 0.3516
Charging
would be based on volume (0.4801)
Arrangement
of water used per month
1 =* respondents were told
that an installation charge
Connection Cost 0.6703
would be levied and that it
Arrangement (0.4727)
would be required as an
upfront payment
Dependent Variable
(WTP) Characteristics
WTPbid Mean monthly bid of all 2.077 0.030 21.620
households in the (3.507)
observation in 1995 US$
Log WTP -0.154 -3.507
3.074
(1.296)
WTP as a % of income Mean WTP as a % of mean 3.969 0.086 27.958
monthly household income (5.850)
Log WTP as a % of 0.614 -2.453 3.331
income (0.605)
D27
Table 5 Summary statistics by study characteristics
Table 6: Results of the Kruskal-Wallis test for study design and elicitation factors
Time-
Conduct Elicitation
Hypothesis Date Author to
Method* Procedure
think*
H p HP H p H H
Ho: equality of
average WTP 0.65 0.418 4.53 0.033 15.67 0 0.65 0.42 4.76 0.093
in all groups
*One or more small samples
D28
Table 7: Studies by income band and number of observations
1
Source: World Bank World Development Report, 2000. The Gini co-efficient is an indicator of the degree of income
inequality based on the Lorenz curve which related income to the percentage of income recipients in different income
brackets. If income is equally distributed then the Lorenz curve is a 45 degree line such that 100% of the population is
earning 100% of the income. Where this is not the case, the Lorenz curve will lie below the 45 degree line. The coefficient is
the quotient of the area between the 45 degree (perfect equality) line and the Lorenz curve and the total area under the
perfect equality line (Todaro, 1994). Thus the higher the quotient the more unequal the distribution of income.
D29
Table 11 Summary statistics of the proposed water supply characteristics
Table 13
Co-efficient
Explanatory Variable Description (Standard T-Statistic P value
Deviation in
Parentheses)
-2.577
Constant -2.16 0.034
(1.195)
Dummy: 1= after 0.8332
Date 2.52 0.014
1993; 0 = otherwise (0.3305)
Dummy: 1 =
Whittington 0.3333
Authors 1.17 0.247
involved in study; (0.2855)
0 = otherwise
Dummy: 1 =
middle-income 0.7689
Region 1.75 0.084
country; 0 = low- (0.4391)
income country
Dummy: 1 "= urban; 1.2184
Locality 3.94 0.000
0 * rural (0.3093)
Dummy: 1 =
-1.3040
Climate arid/semi-arid; 0 = -2.92 0.005
tropical/ temperate (U.4461)
Dummy: 1 = CVM; 0.6646
Conduct Method 1.13 0.261
0 = otherwise (0.5872)
Dummy: 1 =
-0.5098
Elicitation Procedure Closed ended; 0 = -1.62 0.109
open-ended (0.3142)
Dummy: 1 =
0.6505
Time-to-think respondent was not 0.90 0.370
(0.7214)
given time to think
Continuous: 0.12446
Household Size 1.65 0.103
number of people (0.07538)
D30
Co-efficient
Explanatory Variable Description (Standard T-Statistic F value
Deviation in
Parentheses)
in household
Continuous: mean
number of years -0.14675
Education -2.24 0.028
schooling of (0.06555)
household
Dummy: 1 = new; -0.1879
Service level offered -0.65 0.517
0* improved (0.2882)
Dummy: 1 =
-0.0604
Proposed System private; 0 = -0.21 0.831
(0.2814)
communal
Dummy: 1 =
Proposed Charging -0.5206
volumetric: 0 * flat -1.71 0.092
Arrangement (0.3048)
Dummy. 1 •
charges amortised
into recurrent cost
Connection charge tariff; 0 = 0.1960
0.70 0.484
Information respondent was (0.2788)
told that up-front
payment would be
required
Continuous: log of
0.3488
Log of income income measured 2.08 0.041
(0.1673)
in 1995 US$
N = 91 R-Sq. =
59.4% R-Sq. (adj.) •
51.3%
Analysis of Variance
Source DF
SS MS
F P
Regression 15
89.7459 5.9831
7.32 0.000
Residual Error 75
61.3334 0.8178
Total 90
151.0793
D31
Annex E
1.1 INTRODUCTION
In November 2001 as part of the KAR project, ERM met with key
WATSAN stakeholders in India to discuss the research themes of the
project and the issue of cost recovery in water and sanitation projects,
more generally, in India.
• The poor cannot afford safe water and sanitation and should not have
to pay for it;
• Public subsidies are provided to help the poor pay for water and
sanitation;
• The Government of India can solve the problem of water supply by
running water and sanitation programmes itself; and
• The Government of India can raise the financial resources needed for
water supply and sanitation.
The WSP, however in these publications, suggest that reality is more akin
to the following:
• The poor in India actually do pay for water supply and sanitation,
often far more than their fare share;
• The subsidies for water supply and sanitation benefit mainly those
who are not poor;
• Public provision on water supply is inefficient and ineffective; and
• The investment requirements for water supply and sanitation in India
are far too great for the Government of India to afford.
El
The WSP have also gathered together a large amount of evidence from a
number of studies and practical experience across India, which indicates
that many urban and rural communities are wiling to pay for water and
sanitation services. Many people are often paying more, in fact, than the
official tariff through informal channels and coping mechanisms and
would be happy to pay for a better, more reliable service. However, the
WSP suggests that there is sluggishness within policy makers to respond
to this evidence. Political commitment to tariff reform is seen as critical.
There is a long and complex policy history to rural water supply and
sanitation in India, and an attempt has been made to summarise the key
policy and institutional issues in the two boxes below. In short, there has
been a (arguably unsuccessful) history of State level provision of services
since independence, which is giving way now to a more decentralised and
community-focused model of implementation. Cost recovery, partly of
capital costs and mostly of O&M costs, now plays a significant part in this
decentralisation process.
Box 1 A Brief History of Water Supply In India, with respect to cost recovery
Chronology
• A national water supply and sanitation programme was introduced in the social welfare
sector in 1954. The basic premise continues to be that the provision of safe drinking water is
the responsibility of the Government.
• Under this programme Funds are provided from the State budgets for drinking water
supply as a part of every five-year plan.
• The Federal Government funded Accelerated Rural Water Supply Programme (ARWSP)
was introduced in 1972-3 and again in 1977-8. Implementation continues. The aim of
ARWSP is to supplement the efforts of State Governments in providing rural water
supplies, through 100% aid grants from Central Government.
• A Mission approach to the water supply programme was created in 1986; this became the
• Prepared with the support of the World Bank, these Guidelines help to revise the rural
water supply programme and also suggest that:
• Panchayati Raj institutions should be encouraged to adopt the concept of community
participation in the development of their rural water supply programmes.
• Institutionally there is a State Water and Sanitation Mission; a District Water and Sanitation
Mission; a Gram Panchayat Water and Sanitation Committee and a Village Water and
Sanitation Committee. The State Mission should provide policy guidance, auditing and
evaluation; the District Mission should coordinate the receipt of central funds and the
selection of water supply agencies to deliver services (private sector, NGOs); and the Gram
Pnchayat and Village Committees should ensure community participation and manage user
contributions to capital and O&M costs
• Levels of service should, if possible, be in accordance with user preferences and expressed
demands
• For externally aided projects, the following elements should be adhered to:
Based upon this background of policy history and change and the
thoughts of some key WATSAN actors, the KaR research aimed to
investigate some specific projects, to see if generic lessons could be learned
from them and the policy contexts in which they were operating.
The Swajal Rural Water Supply and Sanitation Project is an oft-cited story
of World Bank success with regards to rural water supply and sanitation
in India. It is a US$63 million project covering about 1,200 villages in 19
districts in the Hill and Bundelkhand regions of Uttar Pradesh. It is a six-
year project lasting from 1996 - 2002. Central to the project's design have
been two major polices:
The Swajal project has undoubtedly been very influential in India's rural
water supply sector. The cost recovery objectives it set are now part of
key recommendations for India's national water supply policy (see above).
The use of the community contracting system too, has been successful,
with the supply chain for goods and services coming almost entirely from
the local or regional private sector and with village councils having much
more control over procurement, design and spend. Training villagers in
contracting and project management has also been extremely useful.
E4
to see how financially and institutionally sustainable the interventions,
which the project supported, have been.
With few or erratic official schemes in the area, households would tend to
collect water from streams or wells between 200-1000 meters away. This
could take up to 5 hours a day and result in just 100 litres per household
per day.
With the success of this scheme, and the continued support of the GP, 54
other households in Kamniliparamba Ward got together to meet their own
E5
water supply needs. They formed a registered co-operative society to
organise their finances and make their operations official. From this
beginning a network of privately run water supply schemes in Olevanna
grew. Between 1987 and 1995, 26 such privately registered schemes arose
across the Olevanna GP. In general, each household per scheme put in
between 4,500-12,000 rupees for the capital costs of the scheme and then
paid 5-10 rupees per month for O&M costs.
A key issue that the GP highlights has been the changing role of the KWA.
From 1987 - 1995 the Olevanna schemes worked despite of, rather than
because of, the KWA. The official state water supplier was not helpful in
the early days, as the schemes offered a threat to the KWA's own duties
and reason for being. However, since the decentralisation process for
water supply and sanitation began in earnest in Kerela in 1995, the KWA
have been much more supportive. Indeed, the KWA has now accepted the
Olevanna model as a legitimate and successful approach, but only as a
result of a statewide policy change.
E6
Tablet The monthly operating budget for October 2001 for one scheme in
Olevanna
This particular scheme was set up in 1997 and covers 643 households, with 25 more joining
in October 2001; the budget forms part of the monthly report to the GP.
Income:
25 new connections - 91,520 rupees
Monthly tariff (25 rupees per household) 16,700
Previous balance -1,693 rupees
Total income -109, 913 rupees
Expenditure
To reconstruct a well - 40,354
To replace a pipe - 8,805
To repair pipes - 8,458
Salary to the scheme operator/ caretaker - 9,900
Electricity charges -11,993
Total expenditure - 79,510
Operating Surplus for October - 30,403 rupees
E7
1.2.3 The Kerela Rural Water Supply and Environmental Sanitation Project -
Implementation within the Kunnummel Panchayat
The flow of funds is as follows. The GOI provides the IBRD finance to the
GOK in the form of additional central assistance. The GOK authorises the
KWA to draw new funds for the project through a separate bank account.
These funds are passed on to DPMUs operating at the districts. Funds
from the DPMU, the GP and the beneficiaries then flow into the same
bank account, to support each scheme within the district.
Key benefits and costs of the project were identified in the World Bank's
ex ante cost benefit analyses.
The key costs of the project (in constant 2000 prices net of taxes) were
identified as follows:
E8
households via co-financing and 4,000 rupees for 20% of the
households to co-finance a latrine.
• Annual O&M costs per household (weighted average) of 180 rupees
• Watershed development costs, taken as 10% of the WSS costs (690
rupees per hh)
• Software costs estimated at 1070 rupees per household
• Institutional strengthening costs (including capacity building, project
management etc) estimated as 2,080 rupees per household
Consequently, the ERR was estimated to be 25% for the whole project.
This ERR, of course, understated the benefits relating to improved health,
environment and community/ institutional strengthening activities.
Sensitivity analyses suggested that the project could sustain a 12% ERR,
despite substantial increases in costs and/ or decreases in benefits; and a
50% reduction in estimated time savings.
In cost recovery terms, the conception developed by the World Bank was
that 15% of the capital cost of each scheme was to be paid by the
beneficiaries (of which at least 7.5% of payment must be in cash). The
beneficiaries should meet 100% of the schemes O&M costs - about 180
rupees a year on average.
Kunnummel Panchayat
The project is covering 970 out of the 3868 households in the GP across 9
of the 10 wards. There are 27 beneficiary groups in total, with about 36
households per group. Each group has a proposal for household latrines,
schemes under drainage, waste disposal and women's development
initiatives including micro-enterprises.
E9
For the WS component, the project will aim for 70 litres pp pd, 8 hours
pumping pd, with a 4 hour am and a 4 hour pm slot to pump water.
As per the project design the Kunnummel Panchayat will contribute 10%
of the capital costs of the project, the beneficiary's will contribute 15%
(7.5% in cash and up to 7.5% in kind) and the remainder will come via the
GoK (5%) and the DPMU for the project (70%).
NASRAD, a local NGO is the support organisation for the project within
this District and will help the DPMU to provide community facilitation
and participation expertise.
• The WATSAN priorities within key wards in the GP are then short-
listed.
• Via more formal group meetings in each of the short-listed Wards, the
possibilities for project assistance are outlined and the potential roles
of the different stakeholders are explained to the communities,
including their role and responsibilities. If communities are interested
in participating further, they are asked to form themselves into
informal beneficiary committees and nominate representatives.
• Following this formal recognition, steps are taken for the formal
election of an executive committee for the Group and for subsequent
formalisation under relevant byelaws etc., of the application fees,
monthly subscriptions and so on. Bank accounts for the Beneficiary
E10
Group are established and the first 7.5% down payment are then
collected from the Group members in three instalments.
(So far, 26 out of 27 Beneficiary Groups have selected the open well
and house connection option).
• The 7.5% cash contribution from the Beneficiary Group is then asked
for (labour rates are calculated on local digging costs). The DPMU
provide 40% of their contribution.
Ell
• Operation and maintenance costs are collected from the Beneficiary
Group, averaging up to 1.5 rupees per hh pd
Observations
Thus far into the project, and at a DPMU level, there are a number of
observations about the KRWSESP:
• The DPMU thinks beneficiary groups can pay a lot more towards the
capital cost of the schemes than the 15% they are currently asked for;
• The design of the networks maybe a little top heavy - people may
want more than 10m of pipe towards their house or more than 70
litres pp pd for more than 4 hours a day and be willing to pay for it,
but they can't.
• The project has a top heavy institutional network of PMU and
DPMU's, rather than locally organised institutions disbursing funds
• The DPMU's are still deliverers rather than facilitators/ regulators of
water supply and sanitation services. The facilitation role is falling
mainly to the local support organisation, NASRAD.
• NASRAD, is used to provide the (neutral) community participation
and negotiation skills the KWSA and the GP doesn't have.
• The project is expensive - US$ 72 million - and not necessarily
replicable without another loan to the PMU, given the low level of
beneficiary contribution required. The payback of the loan for the
project by the GOK is another issue that may question the reliability
of the project. Comparing the cost: benefit ratio of Olevanna and
KRWSESP for its recipients (and for the state government) would be
interesting.
1.2.4 The DFID Supported Andhra Pradesh Urban Services for the Poor Project
(APUSP)
E12
organisations and communities to improve the performance of the
Municipalities to deliver and maintain the services, which poor people
need. APUSP comprises three linked and complementary components.
To administer the project, two new units have been established in the State
Government Department of Municipal Administration & Urban
Development (MAUD) - the Municipal Strengthening Unit (MSU) and the
Appraisal and Monitoring Unit (AMU). MSU supports municipalities and
provides assistance in and arranges their access to funds. AMU appraises
the MAPPs and monitors their implementation. The AP Urban Finance
and Infrastructure Development Corporation (APUFIDC) will channel
project funding to the municipalities
The two units and the international consultancy team, which provides
Technical Assistance to these units and the project in general, are working
very closely with state policy makers on urban infrastructure for those
towns. As a result, they experiences at first hand the difficulties decision
makers have in decentralising revenues or to raise tariffs for municipal
services such as water supplies and sanitation, particularly when state
funds are scarce.
E13
• Other issues to focus on should include improved management,
increasing revenue and reducing losses, training and public
awareness, and more community and private sector involvement.
An improved Operation and Maintenance report for the project (not just
for WATSAN) showed that:
In short the project feels it will be difficult to charge people much more
than currently for these services, but it is possible to collect much more of
what is owed and supply the services more cheaply, hence improving the
financial sustainability of the schemes it helps to implement.
Using the MAPP approach for designing its municipal projects, APUSP
can provide seed corn money for connection charges to water in main
urban areas, for instance, and seed corn money for network development
in the slum areas.
The project adopts a process-based approach and so did not have any
fixed ideas about reasonable levels of cost recovery to aim for at the
outset. Instead, the aim has been to make the Municipalities more aware of
the recurrent costs of their WATSAN services, what is being built new,
and how best to both reduce the O&M costs and sustainably fund them.
With regards to O&M charges, the project follows state tariffs for water
supply, which were only 30-40 rupees per month, but have been increased
to 60. The project would like to see them increased to 70. At present, slum
dwellers are obliged to pay only 50% of the tariff.
E14
Observations
Thus far into the project, there are a number of observations about the
APUSP:
Tiruchirappalli City is in the heart of Tamil Nadu, southern India and has
a population of 668,000.The city has 155 slum areas containing about
115,000 people. Gramalya NGO, which was the target of the KaR visit
with Water Aid, works in 8 slums in the East of the city. Under the
Programme, it was proposed to construct, inter alia, pour flush water seal
community latrines by demolishing the existing, state supplied, dry earth
community latrines, in each of the 8 slums. For maintenance of these new
assets, women members of self help groups from the respective slums set
up SHE teams (Sanitation and Hygiene Education teams).
E15
and had been totally abandoned by local residents. Drainage pipes were
broken and the sceptic tanks were damaged. This had meant that people
were defecating in, around and nearby to the latrines and their environs,
rather than using them as they had been meant. Defecation on the banks
of the local river also took place.
The key problem was a lack of a sense of ownership over the latrine
blocks. The blocks had been built by the state, but were not properly
maintained, cleaned or repaired. Local people felt no sense of ownership
and the infrastructure fell into decay. There was little interest in the
construction of new community latrine blocks as people felt the same
thing would happen again in time.
Gramalaya sat with local Women's Self Help Groups from within these
slum areas and discussed possibilities. Could a new sanitation block be
constructed? Was there a need for one? If so how could it best be run in
the long term? There was a clear local need for improved sanitation and
washing facilities in each neighbourhood and there was plenty of thought
as to how such a facility could be managed.
One group came up with the idea of paying to use the toilet - 25 paisa a
time to keep it clean. Once the purchase of cleaning materials and the
wages for cleaners, and a watchman cum ticket issuer were incorporated,
this was raised to 50 paisa a time.
Water Aid provided Gramalya with a grant of 380,000 rupees to build the
latrine block, usually consisting of spaces for 10 ladies; 10 men and a child
friendly toilet block for those less than 6 years old. The blocks were
connected to soak pits which need to be emptied on a regular basis (if no
soak pits were present Water Aid provided a grant for a further 160,000
rupees to construct them). With local inputs for labour coordinated by the
SHE teams, the site was decorated and finished in an attractive fashion.
The charging system was then introduced in one scheme, the first of its
kind. Some in the community were against the idea at first.
The SHE team looks after the upkeep and maintenance of the community
toilet block constructed in their slum. Every user is issued a 50 paise token
allowing him/ her to use the toilet. A woman from the self-help group
collects the money from the paid ticket issuer and closes the account every
12 hours. The accounts (a ledger) contain details on the number of users
and money collected. During nights a watchman cum ticket issuer is
appointed.
E16
compared to other Municipal Corporation blocks or other options
available for defecation.
The first pay and use latrine, collected 168,500 rupees gross over the past
16 months. Table 1 charts the performance of the five schemes as of
30.09.01
Table.2 Income and expenditure relating to pay and use community toilets
Table.3 Income and expenditures for one particular pay and use system up to end
September 2001
Income and expenditures for one particular pay and use system up to end September 2001.
1. Income - averaged 9,965 per month over past 16 months: 159,450 rupees
Community hall - gained a loan of 10,470 rupees to pay for 50% of costs, based upon the income from
the toilets. Paid for other 50% of costs out of toilet income.
Drainage channels -10,000
Extension of corporation pipeline and water tank - 3,000
Garbage collection -1500
Well digging and motor pump - 27,000
Loan of 5,000 rupees to neighbouring slum group at 12% interest p.a - received back 5,500 rupees
The money collected is deposited in a Bank Account under the SHE team's
name each week. At the end of every month the SHE-team and the
particular Women's self help group from that slum convene a meeting and
detail to members the income and expenditure for a particular community
toilet. The balance, after meeting various expenses such as salaries and
E17
purchases of cleaning materials is kept in the bank account. The amount
saved each month is pooled as a common fund for that particular slum
towards health and sanitation related promotional activities including
extension of street taps, construction for community halls, renovation of
other damaged community toilets, construction of domestic drains,
rubbish bins and provision of street lights.
Future investments from the toilets include plans for bathroom facilities, a
urinal complex, a sewing machine centre and other investments
depending on needs. As the knowledge of these latrines has spread, loans
have been made from the surplus and advice given to other slum
communities to help construct their own pay as you go latrines.
A visitor's book for the toilet blocks has recorded the impressions of state
level decision makers and WATSAN decision-makers from elsewhere,
such as Chennai. However, without an official policy change at State level,
the local Municipal Corporation tasked with providing WATSAN services
in the slums won't approve or promote the pay as you use system.
Observations
• The pay and use latrines are being utilised by very poor slum
dwellers
• The first schemes relied on 100% subsidy for the capital cost.
• The schemes are more than 100% cost recovering recurrent costs.
• The capital cost could be paid back over a number of years. However,
the surpluses these schemes have built up as a result, are instead
being used to help wider WATSAN developments in the
neighbourhood, or as loans to part-pay for other slum self help
groups to renovate their systems. As the schemes are starting to self-
replicate, less and less grant funding will be required over time. Is this
a more efficient way of paying back capital costs, in terms of the
outputs it creates?
• No PMU is involved - a local NGO is the implementing agency. An
INGO coordinates the QC of the NGOs delivery.
• With more financial training and co-financing from other sources, the
impact of these schemes is growing.
• The "seeing is believing" impact of these schemes is strong, both
within and beyond the slum communities. However, policy changes
are required to allow the State supplier of WATSAN to also "buy
into" this approach to service delivery.
1.2.6 Scope/ Water Aid Rural Water and Sanitation Programme, Tamil Nadu
E18
Scope is an NGO based in Tamil Nadu, formed in 1986. It operates in just
one rural block in Tamil Nadu and in some of the urban areas of
Tiruchirappalli. In the rural areas, SCOPE works independently, but with
the support of the local block development office.
SCOPE focuses on the conservation and best use of water supplies and
was an early local partner of Water Aid India. SCOPE is currently the
Water Net Convenor for some 36 NGOS working on WATSAN issues
across Tamil Nadu. However, since its formation, SCOPE has also
branched out into rural production centres and savings schemes,
supporting micro enterprises and businesses for rural people.
SCOPE has received funding and support from organisations such as the
Netherlands Embassy through their Sanitation Partnership, Lutheran
World Relief (Canada); Christian Aid, UNICEF, Solidarite Francais and a
range of other international agencies and NGOs. However, SCOPE gets
60% of its funds from local government in Tamil Nadu (the social welfare
board, the small wasteland development board and the forestry board). In
1987 SCOPE had a 30,000-rupee budget; it had in excess of 10 million
rupees for FY 00/01, all of which was targeted on promoting self-
sustaining, not for profit activities for local beneficiaries.
Water Aid has given SCOPE 600,000 rupees per year for the past 5 or 6
years to assist with their rural programme. As a result, Water Aid can
monitor SCOPE'S activities. In terms of the WATSAN issue in rural areas,
SCOPE recognises that NGO's are a "drop in the ocean", but seeks to
obtain replication and uptake of its successes on an equal partner basis
with the Tamil Nadu Water Authority.
SCOPE and Water Aid started work in Kar Hikulam in 1996. Their first
participatory surveys of the village found that diarrhoea was a key issue,
especially among women and the young. People were spending up to
2000 rupees a year on travel and medicine costs on the condition (the
nearest town with a clinic is about 40 kilometres away). There was also
just one pump and bore-well, and one water tank built by the state rural
water supplier in the village. Water supplies were intermittent, available
E19
in the morning and evenings only. To obtain reliable water supplies,
women were rising early in the morning (3am or so, in order to avoid
queues) and spending up to 2 hours a day travelling over a kilometre to
draw water from a 70 foot well by rope. Only about 15 litres of water were
obtained each trip.
At the same time as the first tube well, SCOPE also set up self-help and
savings groups for women - 4 women's groups of 20 members. 25,000
rupees were put into each group by SCOPE to kick them off. Each member
paid 25 rupees a month as savings into the group and 2 rupees
subscriptions. Then, on a revolving basis, women could draw loans from
the fund at a 2% interest rate per month (24% per year). Prior to this,
moneylenders were asking for 10% interest per month. From 1996 - 2001,
the women's savings groups have increased in the village to each hold
about 70,000 rupees.
Two men's savings groups of 14 and 28 members were started in 1997 and
these now have about 98,000 rupees in them.
Each volunteer was given a subsidy of 650 rupees per toilet to help build
the latrine pit safely up to a plinth level (the total cost being up to 1500
rupees to build the whole toilet structure). There was also technical
guidance to help build the superstructure, and the possibility of a 650
rupees loan from SCOPE to help build the structure if needed.
E20
The toilets were constructed with a bathroom and attached to a kitchen
garden in order to use the wastewater. Out of a range of standard designs,
different people chose different toilets.
Once the first 10 were built, they became very popular, especially among
the women (due to their convenience) and more households requested a
latrine. By the end of 1996, SCOPE had helped to build 68 toilets for 96
families. By the end of 1997, every household had built a toilet.
The income from the kitchen gardens helped to pay back the loan for the
toilets, producing on average 30 -90 kgs of fruit and vegetables a year, net
of personal consumption, which was sold in local markets for up to 400
rupees in total.
By 1997, 75-80% of the women in the village were using the toilets, and by
1998, 85% of women were and up to 72% of the men were. By 1999 almost
everyone was using a toilet.
E21
Both the goods the shops sells and its staff have provided jobs for local
rural people. Furthermore, each village is encouraged to take on and pay
for a water and sanitation caretaker, from a scheduled (lower) caste.
Although the wage is not significant, the strong role in the community,
which this provides for these people, is often of great personal
importance.
As households in these rural areas now complain much less about their
water supply and sanitation services, the Tamil Nadu Water Authority is
quite happy with SCOPE' s work. SCOPE and Water Aid, however, are
keen to continue with the capacity building of the Authority so that in
time these kinds of approaches become part and parcel of what the TNWA
does.
Observations
A wide range of stakeholders was met, from both a policy and project
perspective. These included the following:
E22
1.3.1 World Bank Water and Sanitation Programme, South Asia
2.3.3 DFID Andhra Pradesh Urban Services for the Poor (APUSP) Project Team
The current and previous Grama Panchyath project team (including the ex
and current Presidents of the Grama Panchayath, and Ward
representatives) and some recipient villagers.
1.3.5 World Bank Kerela Rural Water Supply and Sanitation Project
E23
1.3.6 Water Aid, Tamil Nadu
1.3.7 General
1.4 SUMMARY
The following table summarises some of the key cost recovery information
from the range of WATSAN projects we visited in India.
E24
Table.4 Summary of Cost Recovery
Project Beneficiary Capital Cost Amount & Replace Water Toilets Size of % Below Cost of getting Remarks
% Recovery Cost Supply Constructed Scheme BPL water/ toilets
O & M Cost before
Olavanna. 1987- Cram Panchayat 100% 4500/- to 100%Rs.5/-to 0% 400 It per day No 26nos 24% (Rs. 3 to 5 hrs 200 spent to No scope for PSP. Great
88 to 1997. Peri 50,000 12000/- per HH 10/- per month per HH + More 22,000/- p.a. foi fetch 100 lpd water horn Leadership. Crisis created start.
Urban. population ~ perHH at a price Peri Urban more than 300 m Demonstration effect.
10,000 HH areas) distance. Cases of
Coverage 80% diahorreah prevalent
Olavanna. After Gram Panchayat 50% by 100% Rs. 30/- to 0% 400 It per day No 60 Nos (lncl.26) 24% (Rs. People selling gold for Meters installed. Penalties for
1997. Peri 50,000 beneficiary 50/- p.m. per per HH + More 22,000/-p.a. for capital cost. No technical overuse. Subsidies for functions
Urban. population - HH at a price Peri Urban support from KWA. such as marriage, etc. WRM is
10,000 HH areas) Difficult relationship becoming an issue now. Festive
Coverage 80% between 1987-1997 with mood and media for
KWA. Convincing was inauguration. Chlorination
difficult and required done.
repeated meetings.
Decisions taken by
Executive Committee.
High Literacy existed.
KRWSA. 3868 HH (Total) 7.5% in cash and 100%. Rs.1.00 to Included in the 70 lpcd Yes. 3679 27 Beneficiary 35% (Total) (Rs. KWA scheme cost Rs. Support Organisation is
(Kunnunvel) GP 970 HH 7.50% in cash / 1.50 per day per 0&M cost (Already Groups. 970 HH 18,000/- p.a. for 22/-p.m. perHH. It NASRAD. Pumping hrs = 4+4
-Water Supply. (Beneficiary) kind. 10% by HH existing) covered (25% to Rural areas) involved 3 to 4 hrs of hrs. Only 10 m pipeline is
Rural. GP. 75% by total). 579/970 walking to get < 70 lpcd. supplied under the scheme to
KRWSA DPMU are BPL each HH. Willingness to pay is
(Finance by WB Rs 1.50 p.m. per HH. Project
project fund of conceived by W Bank after
$80 million) visiting Olavanna. Aim to build
capacity of GP.
E25
Project Beneficiary Capital Cost Amount & Replace Water Toilets Size of % Below Cost of getting Remarks
% Recovery Cost Supply Constructed Scheme BPL water/ toilets
O & M Cost before
KRWSA. School School children 30% to be 100% 100% N/A Poor sanitation facility 40 Self Help Groups created to
sanitation contributed by before. Unhygienic help poor in financing capital
program the school from condition. Health costs. Costs: Rs 30,000 for bore
their budget. Impacts. wells; 35000/- to 40,000/- for
open wells; 3,000/- for HH
connections; and 1,500/- for
public taps. 5% of the total
project cost can be used for
land purchase. Land cost varies
between Rs. 10,000/- (hilly) to
Rs. 25,000/- (good area) per
cent (=40 sqm).
KRWSA. Other Objective is to Beneficiary 100% 100% N/A Poor sanitation facility DPMU says community can
sanitation convert deep pit Contribution is before. Unhygienic pay more. Under People's
program. latrines to 2 pit Rs 3,470/- for condition. Health Planning Program
latrines for new toilets. Impacts. (Decentralization) Community
arresting GW KRWSA contribution is > 90%. Under
contamination. Contribution is Rajiv Gandhi National
Rs 2,000/- for Drinking Water Mission comm.
new toilets. Contribution is >10%. No.
KRWSA regard to WRM. NGO's wants
Contribution govt. to do WRM. High
used to be Rs Literacy.
500/- for toilets
earlier.
Women's Dev. Women Funding under WDI (Microenterprises) gets
Initiative WDI; 30% from Rs. 1500000/- out of allocated
(Microenterprise beneficiaries per Rs. 30000000/- per Panchayat.
s). BG and 70% Rs 1.0 million out of the Rs 1.5
from KRWSA. million goes for starting the
microenterprise. The balance
Rs 0.5 million goes in training
E26
Project Beneficiary Capital Cost Amount & Replace Water Toilets Size of % Below Cost of getting Remarks
% Recovery Cost Supply Constructed Scheme BPL water/ toilets
O & M Cost before
APUSP (Urban) 32 Class! towns 0%for Cost recovered 0% Political reiuctance for
(100,000 to 1.0 infrastructure through tariff. increased tariff. Government
million provision except tax, cess, etc. Rs order stating slums should pay
population) Rs 6,000/- as 60/- per month less than 50% of others. Process
connection perHH. of community negotiators
charges. called MAPP on how to spend
the grant money. Funds
required for seed money as
APUSP would reimburse the
expenses incurred. Current
tariff can meet only 40% to 50%
of the 0&M costs. CR by
reducing O&M cost and
increasing collection revenues.
Rule of thumb being used for
setting tariff (2% to 3%).
Water Aid 25 Slums in 0%.Rs 100%. (i) Use & 100% No Yes Poor Sanitation facilities Rs 4,000/- surplus p.m. for Pay
Slums (Urban) Tiruchirapalli 380,000/= borne Pay @Rs 0.50 + Go, used for wider
(Urban) - 9 in by Water Aid + per use; (ii) Rs development activities.
north zone; 8 in Labour provided 50/- per HH Federation of Committees
east lone; and 8 by community + p.m.; (iii) (FoC) at Panchayat level.
in west zone. Rs 160,000/- Interests from Sanitation fund of Rs 100/- per
borneby funds of SHG HH. Rs 0.5 million to Rs 1.0
Water Aid where used for other million collected by FoC per
disposal system projects @ high Panchayat. Sum total of al!
was not existing. rate of interest Panchayat is more than 0.5
PLUS Rs. 500 million, which is being used for
per quarter per development works.
community for
cleaning of
soakpits.
E27
Project Beneficiary Capital Cost Amount & Replace Water Toilets Size of % Below Cost of getting Remarks
% Recovery Cost Supply Constructed Scheme BPL water/toilets
O & M Cost before
Scope + 110 HH. 612 Capital Cost of 100%. Rs 2/-per Included in 150 Ipd N.A 100% 90% below Before Water Supply was Already each HH pays Rs 2/-
WaterAid people three borewells month perHH O&M costs. R.P.L. approximately 60 Ipd. p.m. (approx) for water tax.
Kattukulum Rs 24,000/-, Rs covers all 3 Collection time per day which is a past of house tax, for
{Water Supply). 27,000/- and Hand Pumps + was over 2 hours. the unreliable government
Rural. Rs.30,000/- caretaker + Distance of water source supply. Capital cost recovery is
respectively. Rs replacement of was more than 1 km and possible but NGOs believe in
2000/- pipes. ropes, etc. was required better utilisation is other
contributed by for fetching 80 feet deep development works. TWAD
beneficiaries for water. Women had to Board is very happy and so are
each borewell. start at 03.00 am to 05.00 the pump mechanics who have
Rest by Scope. am for water. been trained.
Scope + 110 HH. 612 Total Cost Rs 100% by 100% by 100% coverage 67 HH by 1996. 90% below rural PRA done in 1996 7 SHGs formed. Savings of
Water Aid people 1,500/-. Rs 650/- beneficiary beneficiary 110 HH by 2001. BPL indicated Rs 1,500/- to Women SHG is Rs 71,000/-.
Kattukulum borne by Scope. With attached Rs 2,000/- pa. per HH Saving of Men SHG is Rs
(Sanitation). Rs850/-by bathroom. was being spent on 98,000/-. Commitment shown
beneficiary. medication. This fact was by Scope by providing 1st
Interest free ased as lever. borewell and handpump.
Loans of 650/- Hygiene awareness
by Scope to be programmes. Kitchen gardens -
paid back in 10 sold net of consumptin Rs
equal 400/- pm for 3 months. High
Installments. 39 costs, inconvenience, time,
HH availed illness now reversed. Sanitary
loans from Mart, etc. provided. No serious
Scope. Other case of diahorreah. Men had a
HH availed longer take up. WaterAid in
loans from general in rural area. Go!
SHGs. New provides capita] Cost for
toilets coming borewell. WaterAid encourages
up on their own community through training to
now. meet O&M costs.
Scope+ 7 SHGs formed
WaterAid that are listed
Kattukulum hereunder:
(Self Help
Group - SHG).
E28
Project Beneficiary Capital Cost Amount & Replace Water Toilets Size of % Below Cost of getting Remarks
% Recovery Cost Supply Constructed Scheme BPL water/ toilets
O & M Cost before
1. Women SHG - Rs. 2/= Rs. 25/- saving Disburses loans ® 2% p.m. as
4 Nos X 20 Subscription per p.m. Total Rs against 10% p.m. charged by
members -1996. month 71,000/- savings local money lenders.
as on date.
2. Men SHG-2 Rs. 15/- saving
nos(28+14 p.m. Total Rs
members) -1997. 98,000/- savings
as on date.
3. Elders SHG
4. Youth SHG
5. Mother SHG
6. Children SHG
7 Hand Pump
Maintenance
SHG
E29
#
Annex F
Finally, despite a broad policy decision to supply free basic sanitation, neither
a definition of 'free basic sanitation' nor a detailed policy framework is yet in
place. The same departments often deal with water and sanitation at the local
level and financial viability of one service may affect the other. The free basic
water policy therefore may also have negative impacts on sanitation provision
and local authorities will have to consider the implications at the local level.
Fl
1.2 INSTITUTIONAL REFORMS AND INSTITUTIONAL MODELS
During the Apartheid era, homelands were managing their own affairs
independently from the rest of the country and, as a result, a large part of the
country was largely neglected, with poor or non-existent infrastructure and
very little information on the state of the services. In addition, there was no
common institutional framework for organising water services in black and
white areas. In 1994, a White Paper on Community Water Supply and
Sanitation was passed and set out key objectives for the reforms. Since then,
key pieces of legislation affecting the water sector have been enacted to
provide institutional, legal and regulatory frameworks to underpin the
provision of water services.
The Water Services Act (No. 108 of 1997) set out the common legislative
framework for water services for all municipalities in South Africa. It replaced
the Water Act of 1956, redefining the institutional roles and responsibilities
and bringing the water services delivery framework in line with the
Constitution and principles of local governance. The responsibilities of central
government and the national Ministry (DWAF) were refocused mostly onto
regulatory functions, (with the definition of methodologies for setting tariffs
and service levels for example, and monitoring of the water service activities
of local governments). Responsibilities as Water Service Authorities were
granted to the district municipalities and the metropolitan municipalities (for
the large cities). Water Service Authorities are in charge of monitoring and
regulating service provision by the Water Service Providers at a local level.
Those Water Service Providers can be the Water Service Authority itself (in
the case of direct management) or any other type of provider, including
private sector or possibly, civil society organisation. The Water Service
F2
Authority would need to sign a contract with the provider in order to ensure
adequate provision of those services.
DWAF, the Minister of Water Affairs, was assigned a regulatory role, with
publicly-owned water boards being assigned the role of bulk water service
providers to municipalities within their area of supply. This was in sharp
contrast to DWAF's role up to the Act (and still to date, due to the difficulties
of carrying out the transition), which was closer to one of water service
authority and provider, especially with the running of large supply schemes
in rural areas.
The 1997 Act assigned the role of Water Services Authority (WSA) to local
government, entrusting it with the responsibility of providing access to
affordable, sustainable water services to all existing and potential consumers
in their areas. According to the Act, WSA can act as service providers within
their areas or enter into a water services contract with a service provider to
perform these functions on its behalf, or enter into a joint-venture agreement
to jointly perform this function, with the condition that it has to consider all
the public providers capable of providing this function before entering into an
agreement with a private provider1. Specifically, WSAs are in charge of the
following.
11
Note that these reforms essentially alter role of other stakeholders such as NGOs - In the new South African strucutre
DWAF is the main conduit for funds that may formerlly have been channeled through NNGOs at specific projects. NGOs
must now function and compete as service providers.
F3
In addition, the Act defined the roles and responsibilities of Water Service
Providers (WSP), water service intermediaries and water service committees.
WSPs can be a public, private or mixed entity, or the municipal government
itself. In practice, WSPs are divided between bulk supply providers and retail
supply providers. Water Boards are usually 100% publicly owned: they
produce and treat water which they then sell-on to municipalities. There are
about 8 Water Boards providing this kind of services in the main urbanised
areas of the country - in between, these services are under DWAF's
responsibility.
District municipalities are one level up from local municipalities in the sense
that their area of responsibility may include more than one local municipality,
particularly in rural areas. The objective of this reform was to minimise the
risk of a situation whereby municipalities are too small to provide services in a
competent manner. One issue with this reform, however, is that "district
councils" do not have any real legitimacy and structures attached to them,
since there were relatively recent creations, and are more akin to just an
"empty shell". In addition, according to current arrangements, Category B
municipalities (local municipalities) receive the equitable share from the
central government, i.e. a nominal subsidy per person below the poverty
threshold in the municipality's area, whereas category C municipalities
(district) are responsible for the provision of water services. As a result, these
funds cannot be directly allocated to the water sector and a number of
problems with fund transfers, including timing and overall allocation, means
that the water sector is lacking in necessary funds to develop as fast as needed.
F4
the Apartheid regime, with formerly predominantly urban areas were made
responsible for a vast rural hinterland. Following a transitional change in
demarcation, the permanent demarcation was done in December 2000, just
prior to new local elections. Despite being there on paper, this transition has
yet to be completely carried out, due to mental barriers but also, due to limits
affecting human and financial resources. But in terms of water service
provision, it raises an enormous challenge for water service authorities, whose
scope of services was considerably expanded to include areas with very low
service and cost-recovery levels.
The clear demarcation between Water Service Authority function (WSA) and
Water Service Provider function (WSP) has introduced clearer responsibilities
for water service management and can give rise to any type of permutations
for performing water service provider functions. This research looked at many
of these combinations to assess their merits in terms of service delivery and
cost recovery potential.
South Africans have been very creative at inventing new models for water
service provision (including the private sector and the civil society sector, in
either bilateral or tri-sector partnerships) and sometimes, a little bit too quick
at dumping them, on the basis of assumed failure. For example, the BoTT
model (see below) has been recognised as a useful innovation in a crisis
environment. Despite criticism of the cost of the BoTT approach DWAF is
F5
currently developing thinking about what the continuation of the BoTT
should be. Various alternatives are being considered.
However, the basic non-payment culture is prevalent for all service providers
irrespective of location. For example, the issue of revenue collection being
separated from operation and management functions was raised in several
contexts.
In the case of Jowam in Johannesburg, only revenue collection from the largest
customers has been transferred, and Jowarn has no control over revenue
collection of those who don't pay at present. In some community-based
organisations, there is a similar problem. Under the new decentralised
arrangements if the district municipality has control over fund management,
these communities are obliged to channel funds upwards thereby depriving
themselves of a very important management tool.
Given the very poor record at cost-recovery, the water sector in South Africa
has so far subsisted and developed thanks to the provision of subsidies from
the central government.
F6
National Subsidy Arrangements
The primary source of financing for local government remains local taxes and
other revenues levied and collected by municipalities themselves. However,
the national government provides part of the finance through the equitable
share grant, which is an unconditional grant that goes direct from national
government to the District Municipalities to be spent according to their
budget priorities.
The allocation of the equitable share has risen rapidly in 2000/1 and is
projected to increase further in 2003/4. The equitable share grant will
eventually equate to about R20 per rural family per month, which is
considered to be insufficient for the cost recovery of free water (see below on
the free water policy). Local government will also receive conditional grant
funding. The implications for local authorities of the increases in the equitable
share will be a general increase of the average grant per poor household.
At present substantial subsidies to the water sector are occurring through the
support by national government of the operating costs of DWAF water supply
schemes. These schemes are in the process of being transferred to local
government. In general a re-allocation of the current operating subsidies
should support a free basic water policy. At present these subsidies are not
well targeted, however, and there are low levels of cost recovery in many of
these schemes. There are even indications that since the free basic water
scheme, cost recovery has declined.
The most important means of financing services at the local level remains
locally raised revenue through cross-subsidisation. In the case of more
populous municipalities with large urban cores, this system can work well, as
the tariff can be adjusted upwards for those that use more than 6kl of water.
However, in the more rural municipalities, the number of consumers who can
be taxed with a higher levy is very small, therefore this system doesn't work
in these areas.
Thabo Mbeki announced the free basic water policy in December 2000 as an
essential aspect of poverty reduction in South Africa. But the imitative has
been criticised in some sections of the WATSAN community The policy is to
provide 6000 litres of water per month to a household of eight people, in
accordance with the World Health Organisation basic level of water supply.
The status of implementation of Free Basic Water by local authorities on 1 July
F7
2001 (which was set as the deadline for implementation) revealed that
approximately 23 million people were being served with Free Basic Water, in
about 50% of the district municipalities of South Africa.
A Free Basic Water Strategy Task Team chaired by SALGA and including as
members DWAF, DPLG, National Treasury, Mvula Trust, Development Bank
of South Africa and the South African Association of Water Utilities was
established to spear head the 3-stage implementation programme. Free basic
water is to be funded using a combination of the equitable share of revenue of
local government (see above) and internal cross-subsidies from appropriately
structured water tariffs in a manner that best reflects the specific situation in
the respective local government area.
All municipalities were expected to start implementing the free basic water
policy from July 2001. However, it was recognised that some municipalities
may not have the capacity to implement the policy to a full extent
immediately. Therefore, it was recommended that policy implementation be
approached by developing orders of strategy.
In theory, there might be potential advantages of the free basic water policy
given that there are a few successful examples. In particular, Durban Water
Metro (see case study below) claims that such policy is actually cheaper than
paying for administrative costs to collect from low-volume and low-income
users and higher income areas are able to subsidise the scheme. Indeed,
whoever consumes more than 6m3 has to pay the full cost of its consumption.
However, the situation in Durban is quite unique when compared to the rest
of the country: the Metro is relatively industrial and has relatively affluent
domestic customers, so that both categories can afford to cross-subsidise poor
consumers. Quite apart from any practical advantage, the main advantage is
also for politicians to increase their popularity. But such advantage is by far
too short-term orientated, compared to the potential benefits that could be
extracted from sustainable WATSAN projects over the long-term.
The Free Basic Water Policy was basically a political promise, building on a
long history of promises of free services. Politicians introduced it before
administration had actually thought about implementation, which means that
a number of problems have emerged. As a result, many municipalities are still
F8
grappling with the idea and how they can possibly implement it in practice.
Although the FBWP might affect the need for pre-paid meters, the conclusion
from most operators is that they are most useful as a way to control
consumption than as a way of increasing revenues. Therefore, the FBWP does
not affect their technological choice, provided the smart cards (or tokens) can
be pre-programmed to allow for the free water allocation.
For less affluent municipalities and especially rural areas, this policy will be
costly and therefore, might be impossible to implement. Consumption is well
below the recommended lifeline allowance of 6 cubic meters, which means
that there would be no revenue collection at all. With limited potential for
cross-subsidisation and insufficient transfers from central governments, many
municipalities and village water committees have to find ways round the free
water policy in order to avoid bankruptcy. And in fact, many poor customers
have understood that long-term financial viability of water systems is in their
interest, and simply continue to pay.
Strategies currently used by the municipalities for the free water policy
include a combination of rising block tariffs, often with a low rate for the first
block, and targeted rebates to poor households are being used to provide pro-
poor subsidies. DWAF's consultants (PDG) are actually recommending a
number of "ways around" the free water policy, such as the payment of
membership fees or other forms of payment.
F9
1.4 CASE STUDIES
Rand Water is a publicly owned water board and the bulk water supplier for
the region of Johannesburg (Gauteng). According to the Water Services Act
(1997), Water Boards are permitted to undertake other activities (other than
their primary functions of supplying bulk water), provided that this does not
limit the Boards' ability to undertake their primary activity or place a financial
strain on the entire operation. As a result, water boards can for example
provide water services on contract or in joint venture with water authorities or
provide other types of services, such as management, training and support
services or catchment management services.
In 1998, Durban Metro water (and the city of Pietermaritzburg) initiated the
creation of two partnerships between themselves (public providers of water
and sanitation services in Durban and Pietermaritzburg respectively), a
private operator (Vivendi), a local NGO (the Mvula Trust), the Water Research
Commission and the local bulk water supplier (Umgeni Water). The objective
of these partnerships was to improve water services for the poor in Durban
and Pietermaritzburg, particularly through the development of innovative
approaches to water services provision. The partnerships were created
ENVIRONMENTAL RESOURCES MANAGEMENT DEPARTMENT FOR INTERNATIONAL DEVELOPMENT
F10
through two co-operation agreements, with one for each city, which outline
roles, responsibilities and financial commitment for these partnerships.
Powers in the partnership are equally shared, and financial contributions do
not imply more weight in decision-making. P)
But Durban Water, one of the most efficient water utility in the country, is a
special case, as it can easily recoup the costs of this policy through cross-
subsidisation from a relatively rich household and industrial customer base.
ERM visited the Nkwambase project, in a rural area North of Durban in the
Kwa Zulu Natal province. The project serves a total of 800 households. For
this project, the development engineering firm Dynacon works as
sub-contractor to Umgeni Water, which is the bulk water supplier in the
region of Durban.
(2) Note that these partnerships received considerable support and attention from the Business Partners for Development
initiative, for which the Kwa Zulu Natal projects are one out of eight pilot projects.
Fll
settle the accumulated debt. The cashier, representative of the task team,
displayed some relatively good knowledge of the business. He said that they
could not connect anyone who wanted to because costs can be particularly
high, so they would need to increase the connection fee in order to cover those
costs. In addition, he believed that they would need to increase the tariffs
given that two increases in bulk water prices had not been reflected into their
retail tariffs. However, they are dependent on the local municipality for
agreeing tariff increases.
South Africa has created a new model for linking business, government and
civil society stakeholders in the water systems development process. Rather
than following the traditional buyer-seller relationship of the commercial
contract, there is continued emphasis on "co-development", as exemplified by
the BOTT contracts.
In 1997, DWAF signed BoTT (Build, operate, Train, Transfer) contracts with
"Project Implementing Agencies" (PIA) in each of South Africa's four poorest
provinces. Each PIA is a consortium between private sector actors and NGOs:
for example, both the Northern Province (Metsico) and the Eastern Cape
(Amanz'abantu) consortia are led by WASS (a subsidiary of Northumbrian
Water) and have the Mvula Trust, a nation-wide NGO specialising in
providing water services to the rural poor as partner for institutional and
social development aspects. The four BoTT programmes were first
established with the objective of carrying out 100 small-scale projects.
The rationale for designing this type of contracts was to accompany the
municipal service reform and to allow a quick transfer of responsibilities for
water and sanitation from central government (DWAF) to local governments.
Given that in many poor rural areas, local governments did not have the
capacity or the financial resources to take on such responsibilities in a short
period of time, DWAF developed the BoTT concept for accompanying and
overseeing this transfer.
F12
establishment of a project management structure (the "Project Implementation
Agency"). The PIA is in charge of overseeing the services provided by the
Lead Service Providers as well as any other company or community
representatives that the latter supervise. Community representatives work as
partners on the project, as members of the Project Steering Committees and
Village Water Committees.
These contracts are now coming to an end, as they were initially signed for
two-years and renewed once. Both self-criticism (by DWAF, due to the
relatively high cost of the programme) and external criticism (by NGO
partners, due to their limited ability to strategically orient the projects) are
running strong. However, it is generally recognised that these project
management structures have been successful at delivering rapid
improvements in service delivery with appropriate attention paid to the long-
term sustainability of the schemes established in such a way. Private sector
operators have been quick in recognising the interest of such approach, and
are trying to win more projects outside the BoTT structure directly with local
governments. Indeed, following that experience, DWAF decided to develop
an "improved" BoTT model contract, which could be used directly by local
governments wanting to call on the private sector to develop their water
services on the basis of this "one-stop shop" approach.
F13
Table 2 Advantages and disadvantages of the BoTT approach
Advantages Disadvantages
• Allows to make considerable public • No direct investment by the private sector
investment in a short time: speedy and • Costs are higher than if projects were
efficient implemented in a traditional way
• Provides a good overall project (although those added costs may be the
management structure for running many price to pay for a speedier implementation)
small schemes • Local governments were not sufficiently
• Integrated approach to project cycle implicated into the process, which was
• Draws on expertise from various decided and implemented before the local
disciplines for various tasks (business government reform
planning, construction, O&M, ISD, • Insufficient emphasis on training of future
mentorship and transfer) water service providers
• Moves away from a strictly engineering- • Few schemes have been transferred to date;
driven approach to one where institutional difficult to assess the relative success
and social development specialists can • The impact on local service providers not
have a say about the project structure included in the PIA might be severe, given
• Strong emphasis on building capacity, both that they can not compete for those projects
at water service authority level and at during quite a lengthy period of time
village level
• Helped the establishment of consortia
centred on private operators who can now
work on small-scale projects originating
from local government on a more
competitive basis
F14
• Graham Simpson, VKE Engineers Durban
• Lansan Marah, Sigodi, Marah Martin, Development Consultants
• Rod Alence, Sigodi, Marah Martin, Development Consultants
Public Institutions
F15