CONTENTS

ABSTRACT

I. INTRODUCTION

II. ADB OPERATIONS IN THE IRRIGATION AND DRAINAGE

SECTOR

i. ADB Strategy

ii. The IDS Portfolio

iii. Evaluation Operations

III. EVALUATION FINDINGS ON PROJECT IMPLEMENTATION AND

PERFORMANCE

i. Implementation Experience

ii. Performance Results

IV. KEY ISSUES AND LESSONS

i. Key Issues Lessons

V. CONCLUSIONS
I. INTRODUCTION

This sector synthesis report presents (i) evaluation findings in the irrigation and
drainage sector (IDS); (ii) identifies the key factors affecting project
performance; and (iii) highlights the major issues and lessons that are
relevant and significant to the operation of ongoing and formulation of
future projects/programs in the sector. This is the second such report in the
IDS, therefore it builds on the findings of the May 1995 sector synthesis
report, which had covered evaluation findings up to 1994. The synthesis
report is based on a review of the findings in evaluation reports of
irrigation and drainage projects prepared by the Independent Evaluation
Department (IED), primarily project performance audit/evaluation reports
(PPAR/PPER) and reevaluation studies circulated from 1995. It also
considers information and data stored in the evaluation information
system.

II. ADB OPERATIONS IN THE IRRIGATION AND DRAINAGE

SECTOR

ADB Strategy

The IDS continue to be an important sector in ADB's operations since the

agriculture sector still accounts for a major share of the economy of ADB's
developing member countries (DMCs). In 2008, the share of agriculture to
total value added in the DMCs ranged from a low of 5.4% in Kazakhstan to a
high of 35.6% in Papua New Guinea. The sector is seen as one of the means
for reducing poverty, which remains largely a rural phenomenon in Asia and
the Pacific. Therefore, agriculture was envisioned to be given higher priority
since interventions in the sector are more naturally targeted to benefit the poor.
ADB's investments in agriculture reached a high of 35.9% of total lending in
1982, with 27.1% of the aggregate loans in the 1980s channeled to the sector.
However, there has been a downtrend, with the share averaging 12.1% in the
1990s and 7.7% in the 2000s (until August 2009).

ADB's support for the IDS is shown in Water for All, its water policy approved in
2001. Water for All highlights the need to "move rapidly from an era of
disaggregated water sector investments aimed primarily at creating assets to
an era of holistic, integrated investments to promote efficient water use."
Water is a dwindling natural resource and competition for available water
resources is increasing (with the need to provide drinking water and
sanitation as the first priority). Thus, food production to meet the needs in
2025 will have to be accompanied by a dramatic increase in the overall
efficiency of irrigation water use. It is assumed that regional food production
needs in 2025 will be met by (i) expanding irrigated areas to 230 million
hectares (ha), and (ii) increasing productivity.

B. ADB operations in the IDS were generally directed toward developing surface
water and groundwater for agricultural productivity. The operations were
aimed largely at assisting DMCs increase food production (mainly rice), farm
income, and employment opportunities by providing necessary capital
investments. The operations included mostly rehabilitation or upgrading of
existing irrigation and drainage facilities, while some projects included
developing new water sources or drainage systems and repairing
infrastructure that had been damaged by natural events.
Evaluation of IDS Operations

Of the 139 loans in the IDS, 120 have been completed (and loan accounts closed),
while 92 project completion reports (PCRs) and 46 PPER had been prepared
as of 30 June 2009. A closer look at the 92 completed projects showed a
generally improving performance throughout the years. Based on the latest
performance rating of the completed projects, those approved in the 1960s
had a success rate of 33.3%. This record improved, with the 1970 approvals
registering 40.0% success, followed by 44.7% in the 1980s and 61.5% in the
1990s (Figure 2). The better success rates exhibited by the relatively more
recent projects may be attributed to the experience gained and lessons from
earlier projects, which could have helped enhance the quality at entry and
project implementation.

Of the 92 completed projects, 18, which were evaluated from 1995 to mid-June 2009,
are the subject of this synthesis (Appendix 2). They were approved from
1977 to 1993, with two thirds of them in the 1980s. Total ADB investments
in the 18 amounted to $927.5 million (Table 1), 11.0% and 2.7% of ADB’s
aggregate lending to the ANR sector and to all sectors, respectively. Group B
countries received 66.7% and 55.5% of the total number of projects and
amount, respectively. Bangladesh accounted for the most number (27.8%) of
the evaluated projects.

Table 1: Loan Projects in the Irrigation and Drainage Sector by

Country Group

Country No. of Amount % Share

Group/Country Projects ($M)
Group A 2 50.4 11.1 5.4
Nepal 2 50.4 11.1 5.4
Group B 12 514.9 66.7 55.5
Bangladesh 5 212.3 27.8 22.9
Pakistan 4 154.5 22.2 16.7
Sri Lanka 2 71.6 11.1 7.7
Viet Nam 1 76.5 5.6 8.2
Group C 4 362.2 22.2 39.0
Indonesia 4 362.2 22.2 39.0
Total 18 927.5 100.0 100.0
IED = Independent Evaluation Department, M = million, No. = number, PPER
= project performance evaluation report. For loan projects with PPERs
circulated from 1995 to 30 June 2009. Source of Basic Data: IED's
Evaluation Database.

The objectives of the evaluated IDS projects were consistent with the Millennium
Development Goals of reducing poverty and reducing the loss of
environmental resources. The major objectives were to increase agricultural
production, and improve the economic and social conditions of the rural
population. Increased farm incomes, improved living conditions, and
additional employment opportunities were to be achieved by providing and/or
improving irrigation and drainage facilities, which would result in increased
crop yields, allow higher cropping intensities and crop diversification, and
expand productive areas. All the projects had a combination of any or all of
the following irrigation, drainage, flood protection, and other components
(Supplementary Appendix A). Almost 90% of the 18 projects had an
irrigation component, while half had drainage.

III. EVALUATION FINDINGS ON PROJECT IMPLEMENTATION AND

PERFORMANCE

. This chapter reviews the implementation experience and performance results of

evaluated projects in the IDS. In particular, it discusses the impacts of
changes on project design and scope, institutional capabilities of
implementing agencies (IAs), broader physical achievements including
factors affecting performance, project cost, implementation period,
socioeconomic and environmental impacts, gender effects, and the overall
sustainability and performance classification of the evaluated projects.

A. Implementation Experience

1. Project Design

The evaluated projects were designed according to suitable water sources and
appropriate interventions in the project areas. Most of the projects (88.9%)
were designed with an irrigation component, 50% had a drainage component,
while 38.9% had a combination of irrigation and drainage components Of the
18 projects, 2 were for developing groundwater supply; 15 for rehabilitating
surface water development, and 1 for rehabilitating an irrigation system
damaged by floods. Three projects were designed to address drainage
concerns for agricultural production. Supplementary Appendix B summarizes
information on the project designs and reformulation experience of the
evaluated projects.

Project preparatory technical assistance (PPTA) preceded all the evaluated projects
(except the Flood Rehabilitation Project 8 in Bangladesh and the Irrigation and
Flood Protection Rehabilitation Project9 in Viet Nam), with project designs
based on feasibility studies funded by the PPTAs. The Bangladesh project
was implemented to restore the affected portions of an irrigation system to
preflood conditions, and, as such, the project design process did not undergo
the stepwise study and investigation that project preparation would normally
entail. Subprojects for the project were selected on the basis of a set of
criteria provided in the loan agreement. The design of the Viet Nam project
was based on French consultants’ studies of the irrigation schemes, followed
by local engineers’ feasibility studies. The evaluation report for the Viet Nam
Project stated that "…Project objectives were not well defined, partly
reflecting lack of detailed project preparation." The Project was prepared
without a PPTA on the assumption that the executing agency was ready to
undertake the Project. Also, "…the lack of a PPTA reflected the Ministry of
Agriculture and Rural Development's belief in the urgency of rehabilitation
combined with ADB’s wish to be the first multilateral agency to commence a
project in Viet Nam following the resumption of lending."

Changes in project design were common in the evaluated projects. Six projects
underwent reformulation, primarily to adapt them to site, environment, and
local conditions that were not foreseen or considered in the original design.
The East Rapti Irrigation Project10 (Nepal) was reformulated to make the
Project more cost-effective and to prevent possible adverse environmental
effects by replacing the weir and canal system with a program of
rehabilitation and support for existing farmer-managed irrigation schemes
(FMIS). The South Rohri Fresh Groundwater Irrigation Project’s11 (Pakistan)
reformulation involved a reduction in area due to a larger saline groundwater
presence than was considered at appraisal. The Second Hill Irrigation
Project12 (Nepal) was reformulated to remove bottlenecks that had prevented
the Project’s smooth progress and to increase the involvement of the farmers.
The Kirinda Oya Irrigation and Settlement Project13 (Sri Lanka) and the
Pabna Irrigation and Rural Development Project14 (Bangladesh) necessitated
reformulation mainly due to a significant cost overrun. The Irrigation
Package Project15 (Indonesia) was reformulated to reduce project components
because of the shortage of counterpart funds.

2. Physical Achievements

The evaluated projects were reported to have improved or rehabilitated agricultural

production areas ranging from 2,300 ha to 292,000 ha. Also evaluated was
the Flood Rehabilitation Project (Loan 882- BAN[SF]), which was projected
to protect up to 500,000 ha from annual floods. Supplementary Appendix C
summarizes the physical achievements of the projects.

The PCRs and PPERs showed that the projects influenced agricultural productivity
over an aggregate area of 1.53 million ha. Areas covered ranged from
100,000 ha to 500,000 ha for 5 projects and from 2,300 ha to 56,000 ha for
13. Seven projects had larger areas than the appraisal estimates, and 10 had
smaller areas at completion.

The larger-than-expected area covered by some projects compared with appraisal

estimates could be attributed to the following: (i) savings, realized from the
devaluation of the local currency, allocated for areas not previously included
in the project scope such as in agriculture area development operations
involving additional lands for clearing and leveling within the project area;
(ii) improved conveyance efficiency of canals, thereby increasing the volume
of irrigation water distributed and thus covering more area than envisaged;
(iii) areas added by the government after the start of project implementation;
and (iv) reformulation of the project by cancelling uneconomical or
environmentally risky subcomponents such that the intended funds were
allocated to cover additional areas with lower costing subcomponents.

Some benefited areas were smaller than the appraisal estimate because of the
following reasons: (i) the unanticipated large amount of saline groundwater
in a fresh groundwater project, which affected the number of tubewells that
could be constructed; (ii) project design that was too complicated for a given
terrain and locality, with terrain conditions hampering the needed monitoring
and supervision, e.g., in the Second Hill Irrigation Project; (iii) design and
survey errors at appraisal; (iv) delayed implementation, causing a related
project to implement its own measures, thus reducing the need for the
intended intervention; (v) reduction in scope because government budgetary
constraints prevent provision of timely counterpart funds; and (vi) a trade
credit facility that had no takers due to complicated loan requirements
compared with existing sources.

3. Project Cost

The actual costs of the evaluated projects ranged from $12.9 million to $250.1 million
(average, $64.3 million). The unit project cost per hectare varied widely. The
lowest project cost of $77 per ha was for the Flood Rehabilitation Project,
and the highest, for the Pabna Irrigation and Rural Development Project at
$19,383/ha. Although the Pabna Project was much scaled down due to the
reformulation in 1982, the cost was still around 37% more than planned,
mainly because of long delays and major increases in the cost of civil works,
land acquisition, administration, and consulting services. The average total
project unit cost per ha was $757. Supplementary Appendix C summarizes
the costs of the projects.

The actual costs for almost three fourths of the evaluated projects were lower than the
appraisal estimates. The cost underrun of 5.3% to 73.7% was due to (i)
reduction in project scope, (ii) devaluation of the local currency against the
United States dollar, (iii) cancellation or replacement of components with
those costing less, (iv) lower consultancy costs and lower contract rates for
civil works, and (v) high allowances in contingencies due to projected
changes in currency rates and international inflation.

The actual costs for five projects were higher than the appraisal estimates, with cost
overrun ranging from 35.3% to 161.6%. While the cost overrun for three
projects was less than 40%, that for the remaining two was over 140%.
Specifically, the Flood Rehabilitation Project exceeded expected total cost by
140.5% following the occurrence of another devastating flood a year after the
1987 flood, the effects of which were to be addressed by the Project. The
Walawe Irrigation Improvement Project1 also significantly overshot appraisal
estimates by 161.6% mainly because of delays in project implementation,
which in turn were a result of civil disturbances from mid-1987 to early 1990.
Other reasons for the cost overrun included large increases in the cost of civil
works, consulting services, land acquisition, and administration.

4. Implementation Period

The implementation period for the evaluated projects ranged from 4.2 to 17.1 years,
and all were completed after the expected dates. The time overrun ranged
from 22.2% to 153.8%. Physical completion of more than half of the projects
was delayed by more than 3 years, with group B countries responsible for
70% of these projects. Specifically, three of Bangladesh’s evaluated projects
experienced delays of more than 3 years. Supplementary Appendix C
summarizes information on the implementation periods of the projects.

The implementation period was affected by the following: (i) delays in appointing
project staff, fielding consultants, awarding civil works contracts, and
preparing surveys and detailed designs; (ii) reformulation processes; (iii)
unsatisfactory performance of contractors; (iv) acquisition of land or rights-
of-way; (v) lack of counterpart funds; (vi) changes in design or additions to
project scope; (vii) lack of coordination at the national and local government
levels; (viii) controversies concerning site selection; (ix) delays in releasing
funds; (x) political unrest; (xi) weather conditions; (xii) slow action of ADB
in issuing no-objection letters; (xiii) sequencing and coordination of
interrelated activities; and (xiv) government compliance with loan conditions.

B. Performance Results

1
ADB. 2000. Project Performance Audit Report: Walawe Irrigation Improvement Project in Sri Lanka. Manila.
1. Crop Production

. An increase in crop production is normally achieved when irrigation or drainage

systems are rehabilitated or upgraded. All the evaluated projects in the IDS
except the Flood Rehabilitation Project exhibited increases in yields,
cropping intensity, and agricultural land for areas that were definitely reached
by the irrigation or drainage system as it was constructed. Yields were
reported to have increased by 5%–40%, and some farms produced up to 6.0
tons of rice per ha from the combination of irrigation and use of high-yielding
varieties. Cropping intensities improved by 100%–200%. Six of the evaluated
projects had incremental irrigated areas from 14% to 98% of the appraisal
estimates, while five were able to irrigate from 2% to 63% more than their
appraisal estimates. The combined area targets of the projects were about
92% of the appraisal estimates, excluding the Flood Rehabilitation Project.

2. Organization and Management

The evaluated IDS projects had varying degrees of effectiveness in organization and
management. Six projects were satisfactory, eight were fairly satisfactory,
and four had unsatisfactory project management and organization.
Supplementary Appendix D summarizes information on the organization and
management performance of the projects.

Projects that had unsatisfactory project management and organization had the
following characteristics: (i) lack of government guidance; (ii) slow decision
making in the hiring of consultants, evaluating contract bids, and awarding
contracts; (iii) overcomplicated organizational setup at appraisal; (iv) project
office far from sites; (v) high turnover of project staff; (vi) changes in
government structures; (vii) lack of authority of the coordinating director;
and (viii) ADB's inadequate assessment of the capacity of the government
agencies involved in the project.

On the other hand, projects that were implemented with satisfactory organization and
management had (i) close and continuous coordination among local agencies,
communities, ADB, and consultants; (ii) executing agency/ies with prior
experience in implementing similar projects; (iii) project directors with
appropriate financial responsibility for the timely release of funds in the field
without the need to get the approval of the central office; and (iv) project
engineers with experience in implementing a wide range of project activities.
Managing and organizing some projects were difficult because of factors such as (i) a
wide-ranging scope with numerous agencies to coordinate, (ii) delays in
engaging field consultants, and (iii) the terrain and location of field offices
with respect to project sites.

Despite the deficient coordination among the involved agencies, implementation in

several projects was satisfactory since the concentration of responsibilities in
one agency facilitated implementation. In another project, the physical
accomplishments of each agency were implemented as distinct projects,
without the intended integrated approach.

ADB’s performance in the evaluated projects was mostly adequate except in the South
Rohri Fresh Groundwater Irrigation Project, where ADB should have played
a more decisive role in facilitating the early appointment of the project
coordinator; in establishing the project coordination unit, project units, and
steering committee; and in holding regular meetings of the steering
committee.

3. Institutional Development

Institutional development activities in IDS projects have become a regular

subcomponent and have been included in the project objectives in almost all
approved loans in the sector since 1995. However, the earlier evaluated
projects did not have institutional development objectives. In the later
projects, these activities were either tied to the loan commitments or
considered as definite subcomponents of a project. Supplementary Appendix
E summarizes information on institutional development in the projects.

Major activities in this subcomponent dealt with strengthening the capability of IAs,
and organizing project beneficiaries into associations and training them to
participate in implementing projects. Some projects, such as the East Rapti
Irrigation Project, involved the participation of beneficiaries in selecting the
project, and then in implementation and eventual operation and maintenance
(O&M) of the developed system. The approach and experience of the East
Rapti Project can help guide the development of future FMISs.

In the Integrated Irrigation Sector Project, WUAs were organized with leaders being
chosen by the project managers instead of by the members. This practice
affected the proper development of the management capacities of the WUAs
for irrigation O&M. Consequently, a large number of WUAs have become
inactive.

The level of beneficiary participation obtained with the payment of an irrigation

service fee (ISF) is linked to the concerns of institutional development
activities. The Integrated Irrigation Sector Project experienced a slow and
complex introduction of the ISF, and collections were problematic.
Government subsidies for O&M may have given the impression that there is
little need to be serious about collecting the ISF.

Of the 18 evaluated IDS projects, nine were reported to have developed farmer
beneficiaries into functional organizations (WUAs, water management
organizations or farmer organizations) involved in the O&M of the developed
systems.

Sustainability of WUAs in the management and O&M of the irrigation facility also
depends on how early in the phase of project development they become
involved. In the East Rapti Irrigation Project, the FMISs were in place before
the project was reformulated to rehabilitate existing water systems instead of
constructing an irrigation infrastructure. Members of the WUAs would be
better oriented if they were aware from the conceptualization phase that they
are the ”owners” and primary beneficiaries of the project; thus, they would
pay more attention to how the project is being designed, its scope, how much
it would cost to build, and how much the O&M cost would be. They might
even give valuable suggestions to the design consultants based on their long
years of experience in the agriculture sector. Since they would be well-
oriented to the costs and benefits of building and operating the project,
paying the ISF would be expected to be a natural part of their farm
production activities.

Among the evaluated projects, four had TAs supporting institutional development for
a government agency. The Irrigation and Flood Rehabilitation Project in Viet
Nam was accompanied by a TA for strengthening the resettlement
management capacity in the Ministry of Agriculture and Rural
Development.2 The TA was rated successful as it strengthened the Central
Project Office’s resettlement management capacity for both the Project and
Red River Delta Water Resources Sector Project.3 The TA made a significant

2
ADB. 1998. Technical Assistance to Viet Nam for Strengthening of Resettlement Management Capacity in the Ministry
of Agriculture and Rural Development. Manila.
3
ADB. 1994. Report and Recommendation of the President to the Board of Directors: Proposed Loan to the Socialist
Republic of Viet Nam for the Red River Delta Water Resources Sector Project. Manila.
contribution to the reasonably timely and effective resettlement of persons
affected by rehabilitation of the Hanoi dike. It also contributed to the
establishment of effective resettlement systems within the Government and to
the decentralization of resettlement activities to the district level, with
positive impacts on responsiveness and linkages. The Northeast Minor
Irrigation Project4 had a TA for strengthening the Department of Agricultural
Extension (DAE). When the project was implemented, however, the
responsibilities of DAE were transferred to the Ministry of Agriculture due to
slow progress during the first 2 years of the project.

4. Economic Results

The economic internal rates of return (EIRRs) of the evaluated projects were generally
lower than the estimates at appraisal, except for the East Rapti Irrigation
Project (Loan 867NEP), which had a recalculated EIRR of 16% against the
appraisal estimate of 15.9%. The common factors responsible for the lower
than expected EIRRs were the (i) deterioration in the international price of
rice at the time of the evaluation, (ii) considerable increase in the use of
fertilizers, and (iii) reduction in incremental irrigation area and lower
incremental rice yields per hectare than projected at appraisal. Supplementary
Appendix F summarizes information on the economic results of the projects.

Other reasons for the differences in estimates include cost considerations that were
excluded at the time of appraisal, such as (i) opportunity costs of water and
electric power for other economic uses; (ii) distortions in financial prices
against economic values of crops (resulting in farmers planting rice and
sugarcane because these were much more financially attractive than wheat
and cotton, which on the other hand, had higher economic value); and (iii)
attribution of proportionate benefits that may be derived from costs of road
developments giving access to areas irrigated by the project as against the
costs of roads constructed by the project but not within the irrigated areas.
Among the considerations in the economic reevaluation of the Northeast
Minor Irrigation Project was that the shallow tubewells developed were
concentrated in about half of the district compared with road improvement
works scattered over the whole area. Thus, in these cases, the economic
analysis cannot offset all the costs of the road improvement component
against increased agricultural production from groundwater development as
was assumed in the appraisal and the PCR (Supplementary Appendix F).

4
ADB. 2003. Project Performance Audit Report: Northeast Minor Irrigation Project in Bangladesh. Manila.
The Second On-Farm Water Management Project5 had an EIRR of about 27% at the
time of evaluation. The EIRR was estimated to be 30% at appraisal and
19.4% at completion. The project completed earlier investments that did not
realize their full potential benefits because of significant weaknesses in on-
farm facilities. Correcting those weaknesses represented only a marginal
increase in investment costs to make the entire investment more effective in
realizing agricultural benefits. The economic gains of these farm-level
initiatives illustrate the shortcomings common in the design of irrigation
projects that give insufficient attention to the need for on-farm water
management to realize agricultural productivity targets.

The Ganges-Kobadak Irrigation Rehabilitation Project in Bangladesh posted the

biggest fall in EIRR. From the appraisal estimate of 73.4%, the EIRR at
evaluation was estimated at 15%. The significantly lower EIRR was
attributed to long delays in project implementation, overoptimistic
expectations of incremental production at the time of appraisal, and a
substantial reduction in the price of rice.

5. Socioeconomic Impact

The positive socioeconomic impacts of the evaluated projects are seen in the increased
farm incomes and improved access to markets and social services. All
improvements in income were attributed to increases in agricultural
production except in the Second Hill Irrigation Project in Nepal where the
reported sources of income were largely nonfarm. Supplementary Appendix
G summarizes information on the socioeconomic impacts of the projects.

Key informant interviews reported improved incomes, but the improvement usually
could not be quantified for lack of available benefit monitoring data,
especially in projects in which proper baseline data were never generated.

6. Impact on Women

Only seven of the evaluated projects had aspects concerning women’s involvement in
development. The East Rapti Irrigation Project had provisions regarding
women and WUA. As a result, all the FMISs had at least two female
members, as did the WUA executive committees, the decision-making
bodies. In this project, the Evaluation Mission confirmed that women were
active in irrigation system repair and maintenance even after project
completion. In the Integrated Irrigation Sector Project, women were included
5
ADB. 1998. Project Performance Audit Report: Second On-Farm Water Management Project in Pakistan. Manila.
on a pilot basis in technology transfer and institutional development
programs for household heads. Supplementary Appendix H summarizes
information on the impact of the projects on women.

The other projects did not have specific activities for women. In general the project
designs were not gender-specific, as they were intended to provide benefits
for the population at large. The Evaluation Mission’s common opinion is that
the increased farm incomes have benefited women, but the lack of benefit
monitoring and evaluation (BME) data makes it difficult to confirm the
nature or extent of impacts specific to women. The rapid appraisal conducted
by the Mission for the Irrigated Command Area Development Project6
revealed increased participation of women in farm production and marketing
activities in the project areas. This result was attributed mainly to the training
activities conducted by the project.

7. Environmental Impact

An investigation of the incremental use of agrochemicals in agricultural production in

the projects showed that it did not add to the levels of pollution in five of six
projects. The use of pesticides was also observed to be under control, with
some governments banning specific pesticides. A reduction in overall
pesticide usage was observed in some project areas. The Khulna Project
noted that an integrated pest management program of the Department of
Agricultural Extension contributed (although only to a limited extent) to
lessening the use of farm chemicals on selected crops in the project areas.
Also, the higher market prices of fertilizers and farm chemicals seem to have
squeezed the profit margins of farmers, resulting in no substantial increase in
the use of chemicals. Supplementary Appendix I summarizes information on
the environmental impacts of the projects.

Drainage projects have positive effects on the environment. Aside from improving the
soil environment for plant growth, they also help improve human health
conditions by removing stagnant water in depressions.

In the Irrigated Command Area Development Project, the conversion of erosion-prone

grassland and sparse secondary growth forest into irrigated rice fields helped
stabilize the environment and deterred farmers from further practicing
shifting cultivation in the project area.

6
ADB. 1998. Project Performance Audit Report: Irrigated Command Area Development Project in Indonesia. Manila.
The reformulated East Rapti Irrigation Project averted the threat in the original
project’s design of diverting most of the Rapti River water during the dry
winter months, which would have had severe negative effects on Royal
Chitwan National Park downstream.

8. Performance Rating

About 50% of the evaluated projects were rated partly successful; 44.4%, successful;
and 5.6%, unsuccessful (Appendix 3). It is noteworthy that BME was
planned for several projects but was not carried out for all, especially the
baseline conditions. Thus, the extent of the impacts of project interventions,
though generally delivered, could not be quantified.

Around two thirds of the evaluated projects were located in group B countries five in
Bangladesh, four in Pakistan, two in Sri Lanka, and one in Viet Nam.
Bangladesh and Pakistan, combined, contributed four successful projects
(two each), accounting for half of the projects rated successful. Indonesia's
three partly successful projects accounted for 33.3% of the number of
projects in that category. Each country had its own share of successful and
partly successful projects. The two projects in Nepal showed diverse
performance, with the Second Hill Irrigation Project fraught with design,
implementation, and environmental problems, while the East Rapti Irrigation
Project had effective experience in the participatory approach. PPER ratings
generally confirmed the PCR ratings, except for two. IED upgraded the
overall rating for the Irrigation and Flood Protection Rehabilitation Project
from partly successful to successful mainly because of the improved
economic performance of the irrigation scheme, based on O&M time and
cost savings identified in post completion surveys (Box 1). On the other
hand, IED downgraded the overall rating for the Khulna-Jessore Drainage
Rehabilitation Project as the evaluation found the Project to be partly
relevant, less effective, inefficient, and less likely to be sustainable.

There was no noticeable relationship between area size and level of success. Level of
success may be related to how well beneficiary participation was developed
in planning, implementing, and operating the project.

Box 1: Irrigation and Flood Protection Rehabilitation Project

9. Sustainability

Among the evaluated projects, seven were found to have low sustainability or to have
components that were not sustained, and six had high levels of sustainability.
Five projects may be sustainable, depending on the availability of future
funds for items such as O&M, WUA formation, and erosion control.
Sustainability of the evaluated projects has generally depended on timely
availability of funds for adequate maintenance. For instance, the FMISs in the
East Rapti Irrigation Project were all operating and had functional WUAs that
demonstrate the basic ability to carry out O&M. The WUAs have adequate
incentives in the high levels of cropping intensity and crop production to
continue to satisfactorily perform O&M. In the meantime, two projects
appeared to have doubtful sustainability: the Second Hill Irrigation Project
and some subprojects of the Integrated Irrigation Sector Project were subject
to the detrimental effects of erosion and unprotected watersheds linked to the
systems. Supplementary Appendix J summarizes information on the
sustainability of the projects.

The low sustainability of some of the projects was a result of any or a combination of
the following factors: (i) occurrence of structural failure or existing risks due
to the location of important structures, (ii) lack of funds for O&M, (iii) lack
of beneficiary participation or formation of WUAs and commitment to pay
ISFs, and (iv) inadequate routine and periodic maintenance works.

Projects assessed as highly sustainable exhibited a combination of the following

attributes: (i) availability of O&M funds either from the government or the
WUAs, (ii) well-performing WUAs with a strong sense of ownership of
tertiary and farm-level irrigation facilities, (iii) beneficiaries' realization of
financial benefits and belief that they come from well-maintained irrigation
facilities, and (iv) owner-operated equipment or facilities such as tubewells.

A closer look at the WUAs7 revealed that they played a major role in the sustainability
of most of the projects. The sustainability of 16 of the 178 evaluated projects
was influenced by the performance of the WUAs. The more sustainable
projects had WUAs that were (i) involved in planning, designing,
constructing, and supervising rehabilitation and improvement works, and had
7
WUAs are also called water management associations, water management organizations, water users cooperatives,
or water users groups. Farmer organization was a more general term used in some projects.
8
This excludes the Flood Rehabilitation Project (Loan 882-BAN[SF]), which is an emergency loan. The irrigation
schemes under the Irrigation and Flood Protection Rehabilitation Project were managed by irrigation management
companies, fully funded by irrigation service fees. The transfer of at least tertiary canal management to WUAs was
expected to have a positive impact.
adequate incentive (in sustaining high levels of cropping intensity and crop
production) to continue to satisfactorily perform O&M as in the case of the
East Rapti Irrigation Project in Nepal; and (ii) strengthened through a follow-
up TA implemented by a committed agency, as in the case of the Walawe
Irrigation Improvement Project in Sri Lanka (footnote 18). Experience from
the rest of the evaluated projects showed the importance of the following
factors, among others, in helping ensure well-performing WUAs: (i) clear
policies and guidelines for people's participation to avoid confusion and
uncertainty in institutionalizing WUAs; (ii) adequate representation from the
different people in the project area, not just the large landowners and
influential people, and the origin of the members (who should come from the
project area); (iii) use of a bottom-up approach in organizing WUAs, and
advocating active farmer participation at the different stages of a project; (iv)
authority of WUAs in determining the use of ISF they collect; and (v) a
change in corporate culture at all levels, ensuring full support for WUAs from
all levels of the government (Appendix 4).

IV. KEY ISSUES AND LESSONS

A. Key Issues

Key issues that have continuing relevance for ongoing and future ADB operations in
the IDS concern the following aspects: design of IDS projects, agricultural
support services, strengthening of institutional capacities, improving O&M
through participatory management, developing and managing water
resources, BME, and ISFs. The summary information on the key issues are in
supplementary appendixes K to Q.

1. Design of IDS Projects

The design of an IDS project will always be a major factor in its success or failure.
The evaluated projects show that the following considerations are important
in the design phase of IDS projects:

(i) thorough preparation of pre-engineering technical data such as surveys,

geotechnical investigations, and dependable water flow at source based on
years of monitoring data;
(ii) soundness of technical design that is appropriate to local conditions and
capabilities, and within economical unit cost limits;
(iii) rehabilitating and upgrading an IDS project would generally experience
better economic results than a new development project;
(iv) scope for developing on-farm distribution systems that would increase
economic returns by completing a previous irrigation system that extended
only up to the main distribution canals;
(v) availability of benchmark description of beneficiaries or stakeholders on
which the designer can base the appropriate design for sustainable O&M;
(vi) involvement of the beneficiaries and stakeholders right from the early stage
of conceptualization and design of the proposed project;
(vii) limiting the scope to manageable geographical boundaries, homogeneous
economic conditions, and uniformity of approach to subprojects; and
(viii) opportunity cost of water for irrigation against other competing economic
uses.
Measures to further improve the design of IDS projects would lie in the selection
criteria and preparation works. Project selection is influenced by the changing
relevance of activities in time. The projects in Indonesia saw the country’s
transition from a rice-importing country to a self-sufficient one and a nascent
exporter. This occurred at a time when the world market price of rice was
deteriorating. Under these circumstances, there appeared to be no clear
justification for routinely undertaking large-scale irrigation development
schemes in Indonesia except for the specific purpose of poverty reduction
through small-scale communal irrigation facilities in the relatively backward
regions with limited alternative opportunities for increasing employment and
income (as shown by the experience from the Second Irrigation Package
Project).

Rehabilitating and improving existing irrigation systems can have a low unit cost per
area, considering that the cost of the old system is a sunk cost. Thus, such
projects have better chances of being selected than new developments.
However, as pointed out in the evaluation report of the Irrigation Package
Project, this raises the question of why irrigation systems are left to
deteriorate to such an extent that large sums of money have to be invested in
their rehabilitation (Supplementary Appendix K). It is therefore important to
review the cause of deterioration of the system being proposed for
rehabilitation. Among others, the pricing of water induces farmers to invest to
improve on farm irrigation.

Development of on-farm irrigation systems and on-farm water management practices

would be worthwhile projects in the IDS for completing an aspect that may
have been left out of a previous project that was designed to be constructed
up to the main and secondary canals only. An irrigation project is justified by
the economic returns to the irrigated area that it will make productive.
However, after water is delivered to a farm entrance, it still has to be
distributed properly and efficiently throughout the farm to reach the intended
production area. The efficiency of the whole system therefore depends on
each individual farmer user's performance in his/her on-farm irrigation
system. In many cases, on-farm distribution systems still require additional
investments from the farmer for, e.g., construction of canals with proper
lining or installation of a piping network throughout the farm.

There may be areas with existing systems where efficiencies in on-farm distribution
systems can still be improved after assessing soil-plant-water relationships.
Such assessment may lead to recommendations to improve farmer irrigation
practices or to install or construct improvements that would enhance water
management. Implementation of such may then lead to higher efficiency in
water use and thus result in excess water for irrigating other areas.

For projects with multiple subprojects, it would be expedient if procedures for the
subprojects are standardized using basic design models.

Circumstances may change from the time of project design, and may necessitate
corresponding revisions in project design. Some evaluated projects showed
how they can be greatly improved with timely reformulation. This option
should always be open so that it can be seriously considered in case the
project design proves to be inappropriate during implementation. Due
diligence in preparing technical, government, and community aspects will
promote an effective and sustainable project design (Supplementary
Appendix K).

2. Participatory Approach

Stakeholders' ownership is one of the factors that spell the difference between a
project's success or failure. For projects using the participatory approach,
beneficiaries or intended users have to be involved as early as possible.
Beneficiary participation is not merely to obtain their endorsement of a
development concept, but also to determine what their needs are, i.e., the type
of development that is needed, considering that they will later be paying for
the cost of the investment. Participation of the beneficiaries during the
feasibility and design stages can minimize the kind of errors experienced in
previous projects. In addition, it will help develop in the beneficiaries a sense
of ownership and responsibility for a project. Benchmark data with a detailed
assessment of community capabilities, cropping practices, production
systems, irrigation practices, and efficiencies may be utilized in designing a
project. During implementation, a continuing BME is needed, and a working
management information system should be in place early on.

A participatory approach for O&M and sustainability should also be considered in the
project design. The beneficiaries’ awareness of the costs and how they were
derived will promote efficient project implementation and help them install in
their WUAs procedures that may be important for the system’s O&M and
sustainability. Topics such as how the O&M funds will be generated by the
WUA or how payments for services will be charged to individual members
should be considered.

3. Pricing Policies and Agricultural Support Services

Providing agricultural support services continues to be a necessary complement to

IDS interventions, especially when the target areas are far from cities or
major production centers. In such areas, technology transfer, farmer
organizations, and farm-to-market roads would still be of economic value
combined with the irrigation or drainage system (Supplementary Appendix
L).

4. Strengthening Institutional Capacity

Responsibilities for monitoring project planning and implementation, and continuing

postproject administration by the government and beneficiary institutions
require corresponding capacity and authority. Thus, for strengthening
institutional capacities, it is necessary to involve not only technology transfer
and training but, more importantly, the granting of authority to effectively
carry out the responsibilities given. Some projects operated well with
financial responsibility being given to the field unit, as it could respond in a
timely manner to project situations. Another project experienced a 2-year
delay due to lack of authority of the IA.

Another crucial factor in project implementation is coordination among agencies

involved in a project. Pooling of responsibilities can be efficient. However, if
implementation would be hampered by lack of cooperation, the lead agency
may have to take the initiative of pushing the project through (Supplementary
Appendix M).
5. Improving O&M through Participatory Management

With the focus of IDS projects moving toward development of communal irrigation
rather than large-scale systems, O&M by participatory management is
becoming more necessary. Such participation will have to be in place at the
early stages of project design, and appropriate resources should be allotted
for that activity.

As pointed out in one evaluation report, there may be the risk of a monopoly by larger
landowners or dominant ethnic groups, and the attendant need for a
regulatory and enforcement framework and, possibly, for moderate subsidies
to assist low income or ethnic groups less able to afford the privately
managed services. To safeguard the interests of these groups of water users, it
is necessary to define the precise role of the government agencies in the
sector (Supplementary Appendix N).

6. Irrigation Technology

The project experience in Bangladesh raised key concerns on ADB's approach to

flood protection and irrigation in that country. There was a debate on the
advisability of large embankments and pump surface irrigation compared
with private tubewell irrigation.9 Pump surface irrigation schemes protected
from floods behind embankments are expensive to build, operate, and
maintain, and are typically public sector undertakings. One project in
Bangladesh showed that beneficiaries pay neither for such investments nor
for operation and upkeep. The result is (i) economic inefficiency, as large
Government financial subsidies have proportionate economic costs; and (ii)
significant shortfalls in local funds resulting in delayed completion,
inappropriate maintenance, and inefficient operation.

7. Development and Management of Water Resources

In many agricultural areas, a proper inventory of the quality and quantity of available
groundwater or surface water is still lacking. An evaluated project included
the drilling of test tubewells to determine and monitor the depth of the
groundwater table in the command area. The data would rationalize the
efforts of farmers in the community who wish to set up their own deep
tubewell or shallow tube wells. It would greatly help in planning irrigation
systems if municipal or provincial governments were able to generate data on

9
A potential drawback is arsenic contamination of groundwater in parts of southern Bangladesh.
the available water sources in their areas and constantly update them for use
in future projects.

Water management technology on the proper use of water, and also on the use of
distribution equipment such as pumps and pipes for on-farm distribution,
continues to have a large potential for growth. The spreading of practical
knowledge on soil-plant-water relationships, such as water-holding capacities
of different types of soil, and moisture and aeration requirements of roots for
optimum plant growth, may improve the use of irrigation, lessen wastage,
and increase planting areas without the need to invest in costly new water
source systems. It can be assumed that a considerable percentage of farms
can still improve their efficiencies in using water, and thus increase
production while reducing water consumption (Supplementary Appendix O).

8. Benefit Monitoring and Evaluation

Baseline data collection and continued monitoring and evaluation are important
aspects of a project. However, they are not always done/done properly.
Evaluation of an IDS project10 pointed out that it is desirable that a
benchmark or socioeconomic survey be conducted at the outset of the
prefeasibility or feasibility study, that a follow-up project BME survey be
undertaken before starting engineering design if either the time that elapsed
or other events necessitate it, and that at least another survey be timed with
the completion of the project. Unless the information from the surveys is used
in determining feasibility and engineering design and is integrated with the
management information system, project BME is not only costly but also
pointless.

Some standard survey procedures that would help in project design and community
organization may have to be developed. Some projects had BME reports that
were incomplete in scope and nonrepresentative of the project area. The data
were not comprehensive enough to allow proper assessment of the project
impact on beneficiary households and communities.

Improved BME methods can help enhance project economic performance especially
in confirming the production area actually developed by a project. EIRRs are
computed on areas being developed, and therefore such confirmation is
critical in ascertaining the actual gains of irrigation projects.
Acknowledgment by the beneficiary of the actual area being served will also
10
ADB. 1995. Project Performance Audit Report: Second Irrigation Package Project. Manila. (Loan 627-INO, approved
on 19 May 1983, for $52 million)
be useful in area-based determination of ISFs and can therefore act as a check
on claims by the agencies on service areas developed.

BME can also tackle questions on pricing and how it is determined at the farmgate
(Supplementary Appendix P). A study on how the products are sold and who
buys them could explain why the farmer implements certain production or
marketing practices. As noted in one evaluation report, with proper price
incentives, farmers are able to diversify to cash crops without much extension
support.

9. Irrigation Service Fees

Proper O&M is crucial to the sustainability of a project, and collection of ISFs can
facilitate proper O&M. A number of issues are related to ISFs, including the
(i) extent of cost recovery, (ii) adequacy and timeliness of funds for O&M,
(iii) role of the WUAs, and (iv) role of the government. To have a firm
ground on the key issues on ISFs, however, the purposes for which they are
collected have to be clearly defined. As shown in Supplementary Appendix
D, the varying origins and purposes of the ISF will form the basis for its
amount and sources. Furthermore, it would be an advantage if initial amounts
are agreed upon at the start of the project rather than set toward the
completion of the project.

ISF for cost recovery may be more feasible for WUAs that have complete control over
the water source and the delivery of service, e.g., in some tube well systems.
Furthermore, O&M may be within the capacity of the associations to handle,
and the system of collecting from the members may be simple enough and
may even be demand-oriented.

For irrigation mains that are shared by several communities or associations,

memoranda of agreement (MOAs) among the stakeholders and the local
governments would have to be prepared in such a way that they are also
enforceable. Local customs and procedures would have to be considered, i.e.,
whether the ISF will be collected from individual farmers through
associations or directly by government agencies. The MOAs should also
consider sanctions and penalties for delinquent farmers.

Priority on the use of the ISF also has to be defined and agreed upon by the
stakeholders from the start of project conceptualization. O&M expenses will
have to be considered more important than cost recovery, but a definite
contribution from the WUA to cost recovery should be in place in the project
MOAs.

IDS projects can consider improving the payment schemes for the use of irrigation
water (Supplementary Appendix Q). WUAs may also need to develop
schemes on how the association can collect payments per volume of use.
There are several modalities for payment schemes, and WUAs may choose or
develop one that would suit the general membership. Payments based on
volume used encourage the practice of sound water management at the farm
level. Metering systems appropriate for particular projects would have to be
selected with due diligence and the cost of owning and operating the
metering equipment or systems should be considered in the economic
justification of the proposed project.

10. Crosscutting Issues

Though crosscutting issues such as women in development, environmental protection,

and poverty reduction were not the major targets of the evaluated projects,
they have to be considered when designing and implementing IDS projects.
The projects’ objective of increased agricultural production benefits generally
all in the community and the increase in income somehow helps reduce
poverty. The irrigation systems, on the other hand, are to be operated with
due consideration for the environment; otherwise, they will not be
sustainable.

B. Lessons

The highlights of major lessons in the IDS are summarized in the following paras. A
list of selected evaluation findings and lessons in the IDS is in Appendix 5,
which includes lessons given in the earlier sector synthesis.

1. Project Preparation

(i) Selecting a project after thorough diligence, preparing a conceptual

framework, and implementing the design continue to be major considerations
in the success and sustainability of IDS projects. Alternatives that may be
more economical should be studied. Reformulations help improve the
economic effectiveness of some projects.
(ii) Taking a holistic approach to upgrading and developing irrigation and
drainage schemes is desirable.
(iii) A BME system needs to be standardized and integrated with the management
information system. Socioeconomic surveys should become a routine part of
a feasibility study, not only to more accurately ascertain the present
agroeconomic conditions and household incomes, but also to involve
potential beneficiaries in the planning of projects and choice of crops. The
BME has to be defined in consultation with potential data users, and reports
translated and made widely available.
(iv) The optimal approach to project formulation is a gradual and progressive
process that focuses on institutional upgrading and participation by all
stakeholders.
(v) Participatory approaches are essential if system deterioration is to be
prevented or minimized. Mechanisms for beneficiary participation in the
design of IDS projects should be carefully planned. Beneficiaries need to be
fully involved in identifying, designing, and implementing projects if
ownership is to be promoted.
(vi) Establishing sustainable WUAs requires time and great effort. Such an
activity needs to be specified and funded by the project if it is to succeed.
High-level support is likely required for about 2 years, with some support
thereafter.
(vii) Because of the dominant effect of the price of rice on the viability of
irrigation projects and the volatility of rice prices in the world market, it is
necessary to exercise caution in designing irrigation projects that may lead to
producing an exportable surplus of rice.
(viii) Many problems and inefficiencies of irrigation stem from the view that water
retains the properties of a public good after it is derived from a source for
growing crops, whereas, by and large, water then becomes in actuality a
private tradable commodity. Establishing systems for collecting payments for
volumes of water used also increases the efficiency of its use at the farm
level.
(ix) In the design of rehabilitation projects, the optimum level of investment
should be studied carefully because the impact of rehabilitation on
agricultural production may not be as large as is often thought. Rehabilitation
per se does not bring about appreciable increases in crop yield unless such
rehabilitation makes a real difference in the availability of water. Therefore,
potential benefits of rehabilitation schemes need to be carefully researched.
(x) Developing on-farm irrigation distribution systems and a water management
scheme at the farm level may represent a small incremental investment cost,
but they are necessary to ensure the benefits of broader irrigation
development.
They should be included as part of any irrigation project.
(xi) Developing the command area must be coordinated with dam construction.
Watercourses should be constructed to permit irrigation soon after the dam is
filled to avoid excessive delay in generating benefits.
(xii) The terms of reference for consultants should place more emphasis on
building the capacity of local institutions rather than simply on preparing
reports.

2. Project Implementation, Monitoring, and Supervision

(i) Project supervision from an ADB resident mission rather than from
headquarters is often more effective, particularly when it results in a better
understanding of local problems and more judicious and timelier reallocation
of loan proceeds.
(ii) Continuous O&M is a necessary precondition to keep river and drainage
channels open, and adequate provision for that must be built into national
budgets, if not funded from other sources. Leasing disputed government land
for O&M is not a viable option.
(iii) ADB and the EA need to have a more proactive role in complex public sector
projects─supported by the active involvement of the local government or its
representatives, and establishment of a more cohesive basis for stakeholder
organization and participation─by paying close attention to and giving advice
on the policy and institutional development process in the EA.
(iv) Supervision missions need to more closely monitor the quality of project
planning and construction work.
(v) Ideally, a project should be accomplished within the periods envisaged at
appraisal. Implementation delays are sometimes due to justifiable reasons
such as reformulations so as to adapt implementation to actual local
conditions or to correct a defect that may have been overlooked at the
planning stage. Some delays, however, could have been avoided, such as
those due to administrative causes. These include delays in selecting and
hiring consultants, contractors, and project staff; slow government processing
or approval of particular loan requirements; and deficient project organization
and management. These latter causes could be minimized by an appropriate
level of activity by ADB as an external participant to check on the various
agencies and levels involved in the documents and funds flow for the project.
V. CONCLUSIONS

The IDS undergo constant development to adjust to the changing needs of

agricultural communities and as it learns from experience with past projects. The
approach of projects in the sector has gradually changed from top-down planning
to beneficiary participation in the planning process. The objectives of recently
approved project loans in the sector have also considered assisting communal
irrigation system needs with support for institutional development, unlike
previous projects, which called for the construction of large multipurpose
infrastructure with little participation from the targeted beneficiaries.
Increasing the participation of beneficiaries in the projects enhances the
relevance of these endeavors and adds to the cost-effectiveness of the
interventions being introduced by IDS projects. Furthermore, their participation
ensures the continued O&M of the projects and thus better chances of attaining
full cost recovery and sustainability.

Irrigation projects differ from other types of development projects in that specific
farmers or farmer groups become the end users of the water being distributed.
Though meant to add to the general productivity of the country, the investment
should actually directly benefit specific farmers with a concomitant increase in
income.

Thus, unlike a rural road project that can be used by most individuals in a
community, irrigation systems have a limited number of direct beneficiaries,
usually rice farmers. Furthermore, a road project is more easily understood and
measured, say, in kilometers of road length constructed, but the area actually
covered by an irrigation system cannot be quantified in the same simple manner.

These peculiarities of an essential aspect of rural development such as IDS

projects therefore, would rely heavily on the diligence of an enlightened end
user for its success and sustainability. It is important therefore for the
beneficiaries to be informed of what the system will deliver, how much they
will pay for it, and the benefits that they can expect, such as the probable market
expansion due to a diversified farming system.

The history of ADB’s activities in IDS shows a continuously improving process

that augurs well for more involvement in this sector and with greater confidence
for success in achieving rural development goals. The global shortage of
specific food items during the past year supports further the case for IDS and
related development assistance.