34379-013: Completion Report - ADB

Completion Report

Project Number: 34379 Loan Number: 2035-CAM (SF) June 2014

Cambodia: Northwest Irrigation Sector Project This document is being disclosed to the public in accordance with ADB's Public Communications Policy 2011.

CURRENCY EQUIVALENTS

Currency unit riels (KR) At Appraisal At Project Completion As of 31 October 2003 As of 30 September 2012 KR1.00 = $0.00026 $0.00024 $1.00 = KR3,835 KR4,044

ABBREVIATIONS ADB - Asian Development Bank AFD - Agence Française de Développement EIRR - economic internal rate of return FWUC - farmers’ water user community ha - hectare IMT - irrigation management transfer ISF - irrigation service fee IWRM - integrated water resource management MAFF - Ministry of Agriculture, Forestry and Fisheries MOWRAM - Ministry of Water Resources and Meteorology O&M - Operation and Maintenance PDWRAM - Provincial Department of Water Resources and Meteorology PIMD - participatory irrigation management and development RRP - report and recommendation of the President TA - technical assistance

NOTE

In this report, "$" refers to US dollars. Vice-President S. Groff, Operations 2 Director General J. Nugent, Southeast Asia Department (SERD) Country Director E. Sidgwick, Cambodia Resident Mission (CARM), SERD Team leader P. Long, Senior Project Officer, CARM, SERD Team member S. Chea, Project Analyst, CARM, SERD

In preparing any country program or strategy, financing any project, or by making any designation of or reference to a particular territory or geographic area in this document, the Asian Development Bank does not intend to make any judgments as to the legal or other status of any territory or area.

TABLE OF CONTENTS

BASIC DATA i

MAP iv

I. PROJECT DESCRIPTION 1

II. EVALUATION OF DESIGN AND IMPLEMENTATION 3

A. Relevance of Design and Formulation 3

B. Project Outputs 5

C. Project Costs 6

D. Disbursements 7

E. Project Schedule 7

F. Implementation Arrangements 8

G. Conditions and Covenants 8

H. Related Technical Assistance 9

I. Consultant Recruitment and Procurement 9

J. Performance of Consultants, Contractors, and Suppliers 9

K. Performance of the Borrower and the Executing Agency 9

L. Performance of the Asian Development Bank 10

III. EVALUATION OF PERFORMANCE 10

A. Relevance 10

B. Effectiveness in Achieving Outcome 11

C. Efficiency in Achieving Outcome and Outputs 11

D. Preliminary Assessment of Sustainability 12

E. Impact 12

IV. OVERALL ASSESSMENT AND RECOMMENDATIONS 13

A. Overall Assessment 13

B. Lessons Learned 14

C. Recommendations 14

APPENDIXES

1. Project Framework 16

2. Appraisal and Actual Project Costs 20

3. Physical Irrigation Infrastructure Provision 21

4. Farmers’ Water User Communities 30

5. Developing Irrigated Agriculture 39

6. Summary of Resettlement Practice and Issues 45

7. Project Monitoring and Evaluation 49

8. Financial and Economic Analysis 52

9. Project Management Structure 59

10. Status of Compliance with Loan Covenents 60

11. Summary of Gender Equality Results and Achievements 64

12. Environmental Safeguards 71

i

BASIC DATA A. Loan Identification

1. Country Cambodia 2. Loan number 2035-CAM 3. Project title Northwest Irrigation Sector Project 4. Borrower Government of Cambodia 5. Executing agency Ministry of Water Resources and Meteorology 6. Amount of loan SDR12.572 million 7. Project completion report number PCR: CAM 1454

B. Loan Data

1. Appraisal – Date started 15 September 2003 – Date completed 3 October 2003

2. Loan negotiations

– Date started 3 November 2003 – Date completed 5 November 2003

3. Date of Board approval 9 December 2003 4. Date of loan agreement 10 November 2004 5. Date of loan effectiveness

– In loan agreement 8 February 2005 – Actual 2 February 2005 – Number of extensions 0

6. Closing date – In loan agreement 31 December 2010 – Actual 13 February 2014 – Number of extensions 4

7. Terms of loan – Interest rate 1.0% during grace period; 1.5% thereafter – Maturity (number of years) 32 – Grace period (number of years) 8

8. Disbursements

a. Dates Initial Disbursement Final Disbursement Time Interval 31 May 2005 26 November 2013 103.4 months Effective Date Original Closing Date Time Interval 2 February 2005 31 December 2010 71.97 months

ii

b. Amount (SDR million) Cat No.

Category or Subloan Original Allocation

Amount Cancelled

Last Revised Allocation

Amount Disbursed

a

Undisbursed Balance

01A Civil works 6.315 – 6.965 6.965 0 02 Vehicles and equipment 0.630 – 0.351 0.351 0 03 Training and study tours 0.820 – 0.756 0.756 0 04 Service providers 2.838 – 3.130 3.130 0 05 O&M of infrastructure 0.307 – 0.000 0.000 0 06 Other costs 1.295 – 1.057 1.057 0 07 Interest charge 0.367 – 0.306 0.306 0 Total 12.572 – 12.565 12.565 0 – = not applicable. a The unused/undisbursed amount of SDR0.007 million was disposed at loan closing date.

Source: Asian Development Bank.

9. Local costs (financed by ADB) – Amount ($ million) $13.921b – Percent of local costs 99.7% – Percent of total cost 72.7%

ADB = Asian Development Bank b ADB financed 88% of total local costs and 51.7% of total project costs.

C. Project Data

1. Project cost ($ million) Cost Appraisal Estimate Actual Foreign exchange cost 10.481 11.115 Local currency cost 20.391 15.823 Total 30.872 26.938

2. Financing plan ($ million) Costs Appraisal Estimate Actual Implementation costs ADB financed 18.000 19.136 AFD financed 3.742 5.337

c

Borrower financed 6.880 2.465 Beneficiary 2.250 ... Total: 30.872 26.938 N = not available, ADB = Asian Development Bank, AFD = Agence Française de Développement. c AFD disbursed around €4 million (equivalent to $5.337 million, exchange rate on 31 December 2012: $1 = €0,75).

AFD provided parallel grant cofinancing for consulting services and irrigation infrastructure in Preks of Kandal.

3. Cost breakdown by project component ($ million) Component Appraisal Estimate Actual Civil works 13.940 13.539 Resettlement compensation 1.374 0.749 Demining 0.330 N Vehicles and equipment 1.754 0.537 Training and study tours 1.102 1.163 Implementation and facilitation 1.409 1.613 Service providers 7.611 9.606 O&M infrastructure 1.076 N Price contingency 1.751 N Interest during implementation 0.525 0.480 Total 30.872 26.938 N = not available, O&M = operation and maintenance. Note: Actual total cost was $26.938 million (Asian Development Bank: $19.136 million, Agence Française de Développement: $5.337 million, Government of Cambodia: $2.465 million). The differences are mainly due to exchange rate fluctuations.

iii

4. Project schedule Component Appraisal Estimate Actual Date of contract with consultants Consulting services (AFD financed) January 2004 20 January 2005 Consulting services (ADB financed) July 2004 31 January 2011 Completion of engineering designs September 2006 September 2009 Civil works contracts Date of contract award (first contract) December 2004 17 October 2008 Completion of work (last contract) September 2009 7 December 2010 Equipment and supplies June 2004 13 June 2005 Dates First procurement June 2004 13 June 2005 Last procurement June 2010 31 August 2010 Start of operations Completion of test and commissioning June 2009 December 2010 Beginning of start-up June 2009 December 2010 Other milestones

a

ADB = Asian Development Bank, AFD = Agence Française de Développement. a (i) 23 April 2010: Approval of first extension of loan closing date by 12 months, from 31 December 2010 to 31 December 2011. (ii) 8 October 2010: Approval to increase local cost financing by $2.58 million, from $11.33 million to $13.91 million. (iii) 11 May 2011: Approval of reallocation of loan proceeds. (iv) 11 January 2012: Approval of second extension of loan closing date by 4 months, from 31 December 2011 to 30 April

2012 (including reallocation of loan proceeds). (v) 29 May 2012: Approval of third extension of loan closing date by 3 months, from 30 April 2012 to 31 July 2012. (vi) 27 July 2012: Approval of fourth extension of loan closing date by 2 months, from 31 July 2012 to 30 September 2012.

5. Program performance report ratings Ratings Implementation Period Development

Objectives Implementation

Progress 2 February 2005 to 31 December 2005 Satisfactory Satisfactory 1 January 2006 to 31 December 2006 Satisfactory Satisfactory 1 January 2007 to 31 December 2007 Satisfactory Satisfactory 1 January 2008 to 31 December 2008 Satisfactory Satisfactory 1 January 2009 to 31 December 2009 Satisfactory Highly satisfactory 1 January 2010 to 31 December 2010 Satisfactory Highly satisfactory 1 January 2011 to 31 December 2011 Satisfactory Satisfactory 1 January 2012 to 31 December 2012 Satisfactory Satisfactory

D. Data on Asian Development Bank Missions Name of Mission Date No. of

Persons No. of

Person-days Specialization of

Members a

Fact finding 9–28 Mar 2003 5 40 a, b, c, e, g Appraisal 15 Sep–2 Oct 2003 5 35 a, d, e, f, g Inception 14–24 Jun 2004 4 24 a, h, g, j Review 1 6–17 Jun 2005 2 12 a, g, l Review 2 24 Feb–6 Mar 2006 2 20 a, g Review 3 26–29 Sep 2006 3 12 a, g, j Review 4 7–18 May 2007 2 12 a, g Review 5 7–8 Nov 2007 2 6 a, g Review 6 (midterm review) 16–24 Jun 2008 3 18 a, g, h Review 7 2–13 Feb 2009 1 5 g Review 8 9–22 Jul 2009 2 6 g, h Review 9 3–16 Dec 2009 3 9 g, h, i Review 10 5–15 Jul 2010 3 15 g, h, i Review 11 4–22 Feb 2011 3 12 g, i, k Review 12 18 Nov–1 Dec 2011 3 15 g, h, i Project completion review 4–8 Nov 2013 3 15 g, i, m a a = project implementation specialist, b = resettlement specialist, c = environment specialist, d = counsel, e = agro-economist, f = institutional specialist, g = senior project officer, h = social safeguard officer, i = project analyst, j = associate project analyst, k = operations assistant, l = poverty consultant, m = project completion review consultant.

iv

MAP

I. PROJECT DESCRIPTION

1. The Northwest Irrigation Sector Project was designed to reduce poverty in a relatively neglected and isolated area of Cambodia.1 It aimed to increase agricultural production, thereby reducing food insecurity and increasing household income. The design logic was predicated on better use of water resources to take advantage of the potential for irrigated agriculture.2 2. The total project cost estimates at Appraisal of $30.87 million equivalent was to be financed by the Asian Development Bank (ADB) with a loan of $18.00 million equivalent from Special Funds resources, Agence Française de Développement (AFD) with a grant of $3.74 million equivalent,3 the Government of Cambodia with $6.88 million equivalent (mainly for taxes, duties, demining and resettlement), and beneficiaries with $2.25 million equivalent as in-kind contribution. The ADB loan became effective on 2 February 2005; project completion was on 30 September 2012. The project had three components:

(i) Institutional Strengthening in Water Sector

a. Water Resources Policy and Strategic Framework

3. The project was to support the Ministry of Water Resources and Meteorology (MOWRAM) in further developing a policy and strategic framework and asserting its statutory leading role in the water sector. Better planning of investments and coordination between agencies would be founded on (i) a finalized water sector strategy prepared in a consultative and participatory process involving all stakeholders (i.e., government agencies, nongovernment organizations, and international agencies); (ii) regular government and agency forums to discuss ongoing and planned investments; and (iii) strategic development plans and annual investment plans.

b. Regulatory Framework for Irrigation Management Transfer

4. The project would promote the development of a policy, legal, financial, and administrative framework for irrigation management transfer (IMT), and especially help clarify the roles and responsibilities of the various stakeholders, such as farmers’ water user communities (FWUCs), for operation and maintenance (O&M) and for the calculation and collection of irrigation service fees (ISFs). The project was to support an interministerial irrigation working group by providing (i) international technical expertise, (ii) resources for organizing and conducting coordination workshops, (iii) resources to carry out case studies on FWUC experiences within Cambodia, and (iv) in-country study tours and field visits for the working group members (Appendix 2). c. Integrated Water Resource Management in River Basins

5. Adopting and applying the principles of integrated water resources management (IWRM) in river basins required (i) focused training for staff at MOWRAM and its provincial departments (PDWRAMs) in environmental and social impact assessment, integrated water resource

1 The project was located in the provinces of Banteay Meanchey, Battambang, Pursat, and Siem Reap. Until 1998,

much of the northwest had been controlled by the Khmer Rouge and was inaccessible for development agencies. At appraisal, the northwest provinces were among the poorest in the country; average provincial poverty levels ranged from 24% to 54%, and included pockets of poverty reaching up to 60% and more.

2 The project framework is in Appendix 1.

3 AFD provided parallel grant cofinancing to cover the costs of the consulting services and irrigation infrastructure in

Preks of Kandal.

2

planning, and river basin management; (ii) on-the-job training to local service providers in engineering, agronomy, economics, sociology, hydrology, and environmental impact assessment (during river basin studies, subproject feasibility studies, and detailed design); and (iii) formal training at international and local institutions in project management, English language, and computer skills, along with regional and domestic study tours. The project also would provide hydrological measuring equipment, upgrades or construction of office facilities, vehicles, equipment, and operational resources for national and provincial offices.

(ii) Developing and Managing Irrigation Infrastructure a. Irrigation Development 6. The project was to rehabilitate and improve 10 to 12 small to medium-sized—less than 3,000 hectares (ha) each—irrigation and/or water management subprojects (e.g., river diversions, reservoir upgrading, flood control dikes, and flood spreading infrastructure) in identified basins of the target provinces.4 This was expected to result in better water availability and management to irrigate nearly 35,000 ha,5 involving 120,000 beneficiaries or about 15% of the total potential irrigation command area for the four project provinces, equivalent to about 40% of the total area currently under irrigation. Most selected subprojects were expected to require varying degrees of land acquisition and compensation for houses and other assets (located along new or widened canal alignments and enlarged reservoirs), while rationalization and improvement of irrigation and drainage canals (especially for tertiary and quaternary canals in the command area) would require land rationalization and adjustment.6 A summary of the physical aspects of irrigation is in Appendix 3.

b. Farmers’ Water User Communities, Operation and Management 7. Consistent with work under the institutional strengthening component, the project was to (i) train PDWRAM staff, and FWUC management and members, in group establishment and organization; (ii) provide assistance to the negotiation of IMT agreements; and (iii) help facilitate resettlement and land adjustment processes. Nongovernment organizations with expertise in community organization (including gender mainstreaming), O&M management, and irrigated agriculture would be engaged to field irrigation extension teams to supplement PDWRAM staff. 7

(iii) Developing Irrigated Agriculture

a. Improved Agricultural Extension

8. Recognizing the limited capacity of provincial departments of agriculture, fisheries, and forestry and their district agriculture offices, the project was to assess training needs, develop an extension training program for the departments and district offices, and also support participatory farmer-led extension processes through the FWUCs.8

4 During preparation, 35 subprojects were found eligible, and 5 were studied to feasibility level. Of the eligible

subprojects, 17 were stand-alone schemes (i.e., drawing water from one discrete catchment). 5 Including an estimated 19,000 ha of flood spreading and flood recession (i.e., paddy grown in reservoirs as the

water level recedes) and 16,000 ha of irrigated command area, of which about 5,000 ha was to be double-cropped.

6 A participatory resolution process involving farmers and local authorities (i.e., commune, district, and village

authorities) was to achieve this, assisted by provincial and district staff of the PDWRAMs and the provincial departments of land management, urban planning and construction.

7 Issues relating to FWUCs and IMT operations are discussed in Appendix 4.

8 A summary of implementation history and issues in developing irrigated agriculture is in Appendix 5.

3

b. Improved Rice Seed Production 9. To increase availability of early-maturing paddy varieties and maximize the advantages of better water availability, selected farmers would be trained in seed production while others would be encouraged to multiply seed on their own farms (using small quantities of better-quality seed supplied by the project), and the project would train agronomists from district agriculture offices in testing seed for germination and purity. The project provided technical support for a seed certification procedure in each province so that farmers could achieve higher prices for their certified seed.

c. Improved Livestock and Fisheries

10. The project was to reactivate provincial livestock disease diagnostic laboratories, train village livestock assistants, and supply basic veterinary kits to improve animal health services. The project would also support fishery initiatives, exploiting opportunities afforded by better control of natural waterways, rehabilitated irrigation canals, and better water management in paddy fields (allowing diversification into agricultural activities, reducing pressure on natural stocks).

d. Access to Rural Credit

11. As part of the individual subproject feasibility studies, rural credit options were to be reviewed to help bridge gaps between formal credit institutions and FWUCs, or their subordinate user groups.

II. EVALUATION OF DESIGN AND IMPLEMENTATION

A. Relevance of Design and Formulation

12. In 2003, the government had just published the National Poverty Reduction Strategy 2003–20059 and the Second Socio-Economic Development Plan 2001–2005.10 These were incorporated in the project design 11 ; interventions to improve irrigation efficiency and sustainability as a means to reduce poverty were therefore relevant and consistent with government policies.

13. More specifically, recent work at MOWRAM—some with ADB support in the form of technical assistance (TA)12—to produce a range of plans and strategies designed to strengthen water and irrigation management had demonstrated strong and continued commitment to subsector reform, and provided a robust policy background for the project.13 By mid-2004, i.e.,

9 Government of Cambodia. 2002. National Poverty Reduction Strategy 2003–2005. Land and water resources are

fundamental to both economic growth and poverty reduction, especially in rural areas where people in agriculture rely on them.

10 Government of Cambodia. 2002. Second Socio-Economic Development Plan 2001–2005. This emphasized the

importance of agriculture and rural development. 11

ADB. 2003. Report and Recommendation of the President to the Board of Directors: Proposed Loan to the Kingdom of Cambodia for the Northwest Irrigation Sector Project. Manila.

12 ADB. 2002. Technical Assistance to the Kingdom of Cambodia for Capacity Building in the Ministry of Water

Resources and Meteorology. ADB. 2003. Technical Assistance to the Kingdom of Cambodia for Preparing the Northwest Irrigation Sector Project (Part A: Capacity Building in the Ministry of Water Resources and Meteorology).

13 In particular, (i) the National Water Sector Profile (2001); (ii) a draft Law on Water Resources Management (2001)

and supporting implementation subdecrees; (iii) a draft National Water Resources Policy (2001); (iv) a draft National Water Resources Strategy (2001); (v) a Strategic Framework for the Water Sector (2001); and (vi) a draft Investment Strategy for the Water Resources Sector (2002).

4

after the RRP had been prepared but before project implementation began, the new government had reiterated these objectives as part of the Rectangular Strategy For Growth, Employment, Equity and Efficiency in Cambodia, in which “side two of rectangle two” dealt with water resources and irrigation management along the same lines as those along which the project had been designed. 14. Project design was consistent with ADB objectives, manifest in its country analysis (i.e., the Poverty Reduction Partnership Agreement),14 sector-specific thinking (i.e., the Water Sector Roadmap for Cambodia,15 and ADB’s Water For All corporate policy16), and other ongoing investment initiatives at that time (e.g., the North Western Rural Development Project17 and the Tonle Sap Initiative18). The 2003 Cambodia Poverty Review had also identified investment in irrigation as a strategy to reduce poverty (limited access to water being a development constraint).19 ADB’s country strategy and program, 2005–2009 continued the previous support for the main design elements of the project. 20 The concept of rehabilitating various small and medium-sized irrigation schemes was also seen as complementary to the approach being taken at the much larger Stung Chinit irrigation scheme.21

15. During implementation, the relevance of the project was maintained at the highest level from the government’s perspective, as the objectives of the Rectangular Strategy (Phase 2) clearly stated22 and as MOWRAM–MAFF’s Strategy for Agriculture and Water embodied.23 By project closure, the production of rice generally—from increased wet season irrigation productivity to increased dry season cropping—had assumed an even greater role in government thinking through the adoption of the National Rice Policy.24

16. However, from other perspectives the relevance of irrigation investment to sector objectives was beginning to be questioned; based on a review of the implementation experience of this project and others in the sector, ADB’s Agriculture and Rural Development – Sector Assistance Program Evaluation in 2009 concluded that while enthusiasm for irrigation in Cambodia might be very high (and also seemed intuitively appropriate in a country that was so “water-wealthy”), the history of a large number of unworkable schemes developed in an extremely flat topography suggested that technical solutions were not always as straightforward as they might first appear.25 Other views of rice policy in Cambodia during implementation also suggested that yield and production increments of the type encouraged by the project were less

14

ADB. 2002. Poverty Reduction Partnership Agreement. Manila. 15

ADB. 2003. Water Sector Roadmap: Kingdom of Cambodia 16

ADB 2003. Water For All – The Water Policy of the Asian Development Bank. Manila. 17

ADB. 2001. Report and Recommendation of the President to the Board of Directors on a Proposed Loan to the Kingdom of Cambodia for the Northwestern Rural Development Project. Manila.

18 ADB. 2002. The Tonle Sap Initiative: Future Solutions Now. Manila.

19 2003. Cambodia Poverty Review; James C. Knowles.

20 ADB. 2005. Country Strategy and Program, 2005–2009: Kingdom of Cambodia. Manila. (p. 24)

21 ADB. 2000. Report and Recommendation of the President to the Board of Directors on a Proposed Loan to the

Kingdom of Cambodia for the Stung Chinit Irrigation and Rural Infrastructure Project. Manila. 22

Government of Cambodia. 2008. Rectangular Strategy for Growth, Employment, Equity and Efficiency in Cambodia – Phase 2.

23 MOWRAM–MAFF. 2007. Strategy for Agriculture and Water, 2006-2010. Phnom Penh.

24 Supreme National Economic Council. 2010. Policy Document on the Promotion of Paddy Rice Production and

Export of Milled Rice, Phnom Penh. 25

Independent Evaluation Department. 2009. Agriculture and Rural Development Sector in Cambodia. Manila: ADB.

5

important than cultivation of niche and specialized rice varieties, based on more extensive and lower-input regimes.26 17. Regarding project formulation, to include complementary institutional strengthening and improvements in irrigated agriculture systems alongside investments in physical rehabilitation of irrigation schemes was a coherent approach, and necessary for more sustainable subsector productivity. However, although institutional ownership of the investment logic appeared strong, there is limited evidence of noninstitutional stakeholder participation in the design. Furthermore, the overall adequacy of preparatory work may be questioned given the implementation issues encountered—(i) the lack of capacity of service providers, (ii) delays in procurement, (iii) underestimation of the costs of civil works, and (iii) complexities experienced in land titling and resettlement.27 Identifying a relatively large number of potential subproject sites (of which only about a third were expected to be funded) raised unnecessary expectations on the part of communities, which slowed the selection of subprojects and created unnecessary costs.28 18. The overall specification of the project objectives’ hierarchy, and of the risks and assumptions inherent at each stage, was inadequate. In the project framework, the “goal” statement conflates the various levels of objectives, the “purpose” statement concentrates mainly on one of the project’s three components, and “risks to achievement” are completely omitted. There is also confusion about the terminology used for the objectives, which differs in the RRP text and the project framework. 29 The project’s financial and economic analysis presented at design also had several methodology issues.30 B. Project Outputs

19. The project’s expected outputs are listed in the project framework.31 What follows is a broad assessment of the outputs. 20. Regarding institutional strengthening in the water sector, most of the tasks for the development of a water resource investment strategy and its business plans were transferred to the Technical Working Group on Agriculture and Water, which eventually produced the Strategy for Agriculture and Water; for the IMT regulatory framework, 4 subdecrees were drafted, of which one (regarding FWUCs) has been with the Council of Ministers since 2009,32 while the other three have been with MOWRAM for “final consultations” since 2011.33 Twelve FWUCs were established and operational by project completion, and MOWRAM’s capacity to adopt IWRM was mainly delivered through apex water committees. 34 As for the component of developing and managing irrigation infrastructure, all planned river basin studies were completed, and 12 subprojects had been rehabilitated by project completion—although the gross command area covered was only 10,761 ha, compared with the target of 16,000 ha (plus

26

Cambodia Development Resource Institute (CDRI), Phnom Penh. 2011. Future Development Options for the Rice Sector – A Policy Discussion Paper. Bingxin Yu, PhD, Xinshen Diao, PhD, International Food Policy Research Institute, Washington, DC.

27 Resettlement issues are summarized in Appendix 6.

28 Given the number of preliminary investigations undertaken, 28 feasibility studies were conducted during the

project. 29

For more comment on the project framework, see Appendix 7: Project Monitoring and Evaluation: An Overview 30

See Appendix 8: Financial and Economic Analysis 31

Not all were defined in quantitative terms (e.g., expected areas of agricultural diversification). 32

This is despite ADB employing two lawyers from the Council of Ministers to expedite subdecree approval. 33

These cover river basin management, water quality, and water licensing. 34

Formation of five apex committees (comprising adjacent and/or connected FWUCs) began in early 2010 and was completed. The committees cover 215 villages and 19 FWUCs (7 non-project) in all four targeted provinces.

6

recession areas of 19,000 ha). IMT agreements between the FWUCs and MOWRAM were in place by completion, and the FWUCs had prepared operational plans and budgets, although only one of them had successfully applied for the O&M subsidy from MOWRAM. Regarding the development of irrigated agriculture, provincial and district staff members had been trained in improved rice and livestock activities, and over 10,000 training days had been delivered to farmers. While there was evidence of significantly improved rice yields from the demonstration plots at completion,35 the FWUCs themselves were not institutionally established as vehicles for farmer lead extension by 2012, and the extent of diversification and intensified agricultural activities on farmers’ plots was uncertain (and likely modest). The planned improvement to rural credit access was not implemented due to the late completion of the irrigation schemes and because availability of funds from ACLEDA Bank had since improved. 21. Overall, delivery of project outputs can be described as partial, both in quantity and quality. Quantitatively, the most obvious shortfall was the smaller than targeted area covered by subproject rehabilitation; other things being equal, this would tend to reduce project benefits proportionately. Similarly, the regulatory framework for a refined IMT system (as constituted by the 4 subdecrees) was not in place at completion. While the FWUCs were certainly formally established before project completion, delays to the civil works program meant that most had had only 1 year of irrigation management experience before the project closed, and can best be described as nascent organizations with remaining technical, administrative, and financing needs.36 Survey evidence from late 2011 indicated increases in wet season rice yield in all but one of the subprojects, and some dry season cultivation being undertaken at all but two subprojects. However, crop mixes are not recorded, nor are there targets in the project framework to assess whether this is the extent of output delivery that was originally expected. To reflect the status of output delivery, a revised project framework was agreed in 2011.37 C. Project Costs

22. At appraisal, the project cost was estimated at $30.87 million; foreign exchange accounted for $10.48 million (about 34% of the estimated total), and the local currency cost was estimated at $20.39 million (66% of the estimated total), including $0.53 million for interest during construction. ADB provided a loan equivalent to $18.00 million from its Special Funds resources to finance 58.3% of the project cost. AFD cofinanced $3.74 million (12.1% of total project costs) for the consulting services. The government funded $6.88 million of domestic currency costs (22.3%), while beneficiaries were to fund $2.25 million in kind (7.3% of the total). 23. ADB and its cofinanciers fulfilled their undertakings. At project completion, the cost came to $26.94 million, of which ADB financed $19.14 million, or about 71%, for (i) civil works contracts, (ii) vehicles and equipment, (iii) training and study tours, (iv) service providers, (v) incremental administrative costs, and (vi) interest during implementation. AFD financed $5.34 million equivalent (19.8% of the cost) for consulting services and the rehabilitation of the Prek-Kandal irrigation scheme, while the government funded the remainder, equivalent to $2.46 million, or 9.2% of the project cost, for the civil works contracts and resettlement and land acquisition. The actual cost covered by the government was lower than estimated at appraisal, as the government waived taxes and duties, thereby reducing its contribution by $4.42 million. The $2.25 million from the beneficiaries was in kind. The actual total project cost was $26.94 million (excluding the government’s waiver on taxes and duties, and the beneficiary

35

Appendix 5 discusses demonstration plot performance. 36

Appendix 4 discusses the FWUCs’ capacity at project completion. 37

This is contained in Appendix 7.

7

contribution estimated at appraisal). All supply and works contracts came in under budget when compared in real terms. 24. The actual cost for the project components financed by ADB rose by $1.14 million to $19.14 million (from $18.00 million estimated at appraisal) due to US dollar inflation. The actual cost of the service provider category increased from $7.61 million to $8.19 million (an $0.58 million increase). At most sites, the hydrological conditions were more complex than expected, which greatly increased average unit costs of civil works and of work done by service providers. Size downscaling reduced the total costs associated with resettlement 38 and demining, and implementation delays resulted in fewer vehicles being bought and also avoided O&M expenditures for completed schemes during project implementation.39 Actual costs for other categories were more or less the same.40

D. Disbursements

25. To a large extent, disbursement was driven by the progress in procurement of civil works and service providers; it was very slow in the first few years of implementation, when the annual disbursement targets were not met, but then accelerated quickly. 26. By May 2006 (i.e., 15 months after loan effectiveness) only $286,824.78 (1.5% of the loan amount) had been disbursed. By the midterm review in September 2008 (66% of elapsed time), disbursements still only totaled $1,999,205.26 (10.5% of the loan amount). By December 2009, disbursements totaled $5,355,798.43 (28%); by July 2010, they stood at $10,637,707.12 (55.6%); by December 2011, during completion of the civil works, they totaled $17,992,735.31 (94%); and by loan closing (13 February 2014), all $19.14 million had been disbursed. E. Project Schedule

27. The project experienced start-up delays of more than 1 year. This was due to the Cambodian elections and the formation of a new government during 2004, and also because of some disagreement between ADB and MOWRAM about the staffing of the project director post. The original closing date was 31 December 2010, but was extended by 1 year to the end of 2011 to allow for completion of civil works and some support to FWUCs. A further extension of 9 months into 2012 was made to undertake repairs to civil works due to flood damage. 28. The overall delay in the project schedule during implementation was primarily due to lengthy procurement of the service contracts for the river basin and feasibility studies, and for the detailed subproject designs. On average, it took about 12 months per contract to complete the procurement process because the poor quality of service providers’ submissions required several iterations of the process and lengthy reviews by TA consultants. Furthermore, the internal procurement process of government agencies was highly centralized. Most of 2007 was almost entirely devoted to procurement of studies by service providers, and by the midterm review overall project progress was estimated at only 25%. Delays in finalizing selection and design of the subprojects to be rehabilitated or constructed necessarily caused further delays to the start of civil works and, in turn, to the originally scheduled programs in support of FWUC establishment and to the irrigated agriculture demonstration and training activities.

38

Appendix 6 discusses the resettlement issues. 39

Including the expected in-kind and cash contributions from beneficiaries. 40

Appendix 2 compares the actual detailed costs for each category with appraisal estimates.

8

F. Implementation Arrangements

29. The project’s management structure is in Appendix 8. During implementation it became apparent that, for one, the implementation arrangements reflected the need for central and provincial capability and functionality, and drew on existing structures in the country (i.e., the project management office and its resettlement and environment monitoring units, and also the project steering committee for other irrigation projects in Cambodia, plus provincial project steering committees). These institutional arrangements were broadly suitable and were not substantially changed during implementation. Second, the ability of the project management office to adequately deliver outputs was hampered until late 2008 by internal processes regarding the operations of its procurement review committee, difficulties in communicating with the project implementation units, and the part-time nature of the project manager’s engagement. Third, the capacity of private sector service providers to act as IETs was overestimated—they required significant technical support from the TA consultants, and IETs in practice sometimes included MOWRAM staff as members, even though poor public capacity had originally justified the use of private providers. Finally, some small adjustments were made to the implementation arrangements for resettlement.41 G. Conditions and Covenants

30. Appendix 9 summarizes the status of compliance with loan covenants. The covenants under article IV of the loan agreement regarding acceptability of consultants and contractors, as well as accounting and financial reporting, were all complied with. The covenants under schedule 6 of the loan agreement regarding project organization (paras. 1–7) were all complied with, although para. 7 took the form of apex committees rather than the originally envisaged multi-user management committees. Likewise, the covenants covering review and reporting, and counterpart funding (paras. 8–12) as well as project operations (paras. 15–19), covering environment, 42 mines and unexploded ordnance, resettlement activities, the processes for the selection of subprojects, and implementation of the project’s gender plan, were all complied with. The submission of audited project financial statements (article IV of the loan agreement) was timely and satisfactory. Withdrawals from the imprest account were slow in the early years but eventually accelerated, and overall compliance with the covenants improved alongside project implementation performance after the midterm review in 2008. 31. The covenant regarding the law on water resources and its subdecrees (loan agreement, schedule 6, para. 14) was not complied with. Compliance has now been encouraged by including finalization of the regulatory framework for IMT in the tranche release conditions for the ongoing Water Resources Management Sector Development Program loan.43 The covenant regarding the borrower’s provision of sufficient funds to ensure routine O&M of subprojects, and for periodic and emergency maintenance of infrastructure after project completion, is being partly complied with. 44 41

This includes land titling; the original plan of issuing registration certificates to commune councils was dropped in

favor of contracting the provincial departments of land management, urban planning and construction. 42

Appendix 7 discusses the environmental safeguards. 43

ADB. 2010. Report and Recommendation to the President to the Board of Director on Proposed Loans, Grant, Technical Assistance Grant, and Administration of Loan and Technical Assistance Grants to the Kingdom of Cambodia for the Water Resources Management Sector Development Program.

44 Funding for irrigation maintenance is still not systematic, predictable, or reliable. In aggregate, it is not adequate

for the national irrigation system (including subprojects of the Northwest Irrigation Sector Project), there is an absence of programmed planning, and the system remains largely emergency-oriented, supplemented by donor support. In late 2013, the Ministry of Economy and Finance reported that $2 million was available for this purpose, but MOWRAM is yet to make a detailed submission through the current budget process for the sector.

9

H. Related Technical Assistance

32. No technical assistance was attached to the project loan. The consulting services to implement this project were funded by AFD. I. Consultant Recruitment and Procurement

33. Recruitment of loan implementation consultants was done using AFD procedures with ADB’s endorsement. Consultants were mobilized in February 2005, with seven contract variations during implementation—notably, consulting inputs were rescheduled after the midterm review to focus on fewer core staff, accelerate preparation of the subprojects, and intensify support for the farmers’ use of irrigation facilities thereafter. 34. Service providers were recruited through 14 packages. 45 It meant preparing and managing quite a large number of contracts—grouping them into three or four packages would have been more manageable for the executing agency. Local firms often found it difficult to provide both engineering and agriculture extension expertise, and to comply with bidding processes. Four agriculture service agreements were signed with the provincial agriculture departments in the project area to provide Farmer Field School training; the service quality was good but the payment-processing capacity of PDWRAMs was modest. Civil works were provided through 13 contract packages; this process used the post-qualification method and was much simpler and faster.46

J. Performance of Consultants, Contractors, and Suppliers

35. Service providers recruited to undertake river basin studies, feasibility studies, and detailed designs struggled to produce outputs of the necessary quality. Most service providers had insufficient experience in the social and technical aspects (including hydraulics and irrigation design) of small—but often complex—schemes. They also found it difficult to recruit suitable international staff, given the short and fragmented inputs involved, and the limited bid budget. Although feasibility studies improved eventually, all submitted reports required revision and input by TA consultants,47 which consumed a lot of time. Delays occurred when the (part-time) international design specialist was not available. Some service providers did manage to engage international engineering expertise, which led to significantly better output quality.

36. Most of the civil works contractors did not have irrigation and hydraulics experience. This caused issues for some subprojects, where it appears that sandy soils were not always fully replaced with firmer materials, and subsidence of canals and dikes occurred.48 The quality of the works generally improved over time, and the major physical work items (e.g., spillways) appeared sufficient to largely withstand the floods of October 2010.49

K. Performance of the Borrower and the Executing Agency

37. The performance of the borrower and the executing agency, MOWRAM, was satisfactory. MOWRAM’s capacity to implement the project was probably overestimated at

45

Two for river basin studies, and 12 for feasibility studies and detailed designs. 46

Service provider recruitment took about 12 months on average (a major cause of implementation delay) whereas civil works recruitment took only about 4 months.

47 Resources for support to service providers were included in the TA contract.

48 Locally referred to as “dragon holes”.

49 A contract variation was issued in 2011 to allow further work at all subprojects to repair what damage had

occurred.

10

appraisal, and the use of a project management office and a project steering committee simultaneously engaged in implementing other irrigation projects—while attractive from a synergy point of view—meant that many staff (including the project director) were unable to work on the project full-time. This had notable impacts on the procurement processes, especially for the feasibility and detailed design work, which were not entirely transparent until after the midterm review. MOWRAM’s performance greatly improved after the midterm review, however. On the other hand, its technical capacity and authority to finalize the refinement of the IMT regulatory framework was also overestimated at design. At the provincial level, the ability to implement the Agriculture Service Agreements (ASAs) also became a performance constraint in terms of mobilizing provincial department of agriculture staff on time.

L. Performance of the Asian Development Bank

38. ADB’s performance was satisfactory: (i) the project was administered by the Cambodia Resident Mission, which allowed sufficient interaction with MOWRAM staff and TA team members; (ii) supervision missions were timely; (iii) procurement procedures were followed as planned; (iv) approvals for contract variations and revisions to the project framework were speedy; and (v) agreement to rescheduling and extensions was appropriate. ADB provided extra resources to investigate resettlement issues at Ponley subproject when this threatened to delay implementation, and also to support procurement and design work in 2009 (following the resignation of the TA team leader). ADB took a necessarily robust approach to the midterm review (being willing to confront MOWRAM about the slow implementation progress up to that point), and this marked a turning point in the intensity and nature of project administration. Coinciding with a change of ADB staff, closer engagement with MOWRAM in design and procurement processes (including interaction in Khmer) was critical in speeding up design finalization for the subprojects (and the completion of all designs by the end of 2009) and of the civil works construction by 2011.

III. EVALUATION OF PERFORMANCE

A. Relevance 39. The project is rated relevant. The overall concept and logic, i.e., contributing to poverty needs by improving irrigated agriculture productivity, was relevant before and during project implementation. The project’s relevance to national policy and sector objectives (e.g., national rice export policy) and ADB’s ongoing sector operations (as well as ongoing country partnership strategy development) has actually increased since then. The partnership with AFD (and its finance for consulting services) was complementary to ADB, although the later addition of a subproject outside the project area (Preks of Kandal, chosen to fit with wider AFD operations) inevitably diluted the original geographic focus.50 The economic rationale for the project was grounded in a thorough and substantial (sub) sector analysis of rice production, marketing, and consumption circumstances in Cambodia. As for internal design logic, all three project components were necessary to achieve more productive and diverse irrigated agriculture practices on a sustainable basis, and the scope and scale of activities identified at design would have been largely sufficient to deliver the expected outputs—had implementation been much as planned. 40. The caveat to this rating concerns the quality at entry, which was undermined by several factors: (i) very poor specification of objectives (Appendix 7), (ii) inadequate identification (and

50

It also absorbed some TA time and resources (in assisting MOWRAM in the bidding process for consultants).

11

quantification) of the risks and assumptions affecting project performance,51 (iii) underestimation of difficulties in implementing legislative reform, (iv) lack of awareness of the institutional complexities of land titling, (v) poor appreciation of the extent of potential resettlement needs, (vi) inadequate attention to gender mainstreaming, and (vii) overestimation of local service providers’ and contractors’ capacity for the tasks expected of them. To this list may be added an oversimplification of the problems of dealing with flat topography and sandy soils in many places. By the time of the midterm review, the “pilot nature” of many project activities was being explicitly recognized, but this had not been stated as such beforehand. B. Effectiveness in Achieving Outcome 41. The project is rated less effective since the expected outcome was only partly achieved. First, a sustainable basis for irrigated agriculture in the priority river basins was not established in the form of a refined IMT framework, because the necessary legislation (i.e., the four subdecrees) still did not exist at project completion; the FWUCs are functionally weak; and MOWRAM’s capacity to plan and undertake O&M is still very limited. Second, the original targets of 16,000 ha of command area and 19,000 ha of spreading or recession rice cultivation were not met; because of higher than expected construction costs per hectare52 and given the scheduling delays caused by design complexities and procurement processes, the total command area constructed was only about two-thirds of the target; and the dropping of Stung Chikraeng accounted for the loss of the spreading or recession areas. 53 Third, although the project framework did not quantity the degree of “diversification”, the command areas to date have almost no non-rice cultivation, and estimated rice productivity reportedly has not reached 3.5 tons per ha in all places. 42. The actual extent of outcome delivery was achieved through some reorientation and intensification of efforts on the part of MOWRAM and ADB after the midterm review, when project progress was estimated at only around 25%. Specifically, the adoption of time-bound procurement processes, pursuit of a streamlined and intensified approach to resettlement, and concentration of design work on those subprojects already investigated at that point led to civil works being completed before the end of 2010, and at least some infrastructure being operated by the FWUCs before project completion.54 C. Efficiency in Achieving Outcome and Outputs

43. The project is rated efficient. An estimate of the economic internal rate of return (EIRR) of the irrigation infrastructure and management component as a whole suggests a rate of 23%. EIRRs across the subprojects vary but are largely positive, and only 3 subprojects are estimated at less than the opportunity cost of capital (Appendix 8). Notwithstanding that (i) the composition of subprojects constructed is somewhat different than originally envisaged; (ii) the delivered command area has been considerably reduced (and spreading or recession rice cultivation is entirely absent); and (iii) delays to the implementation schedule were significant, the trebling of the economic price of rice since design has more than offset these factors. The subprojects’ EIRR estimates also do not factor in the substantial range of nonquantifiable social, economic, and environmental benefits at each site.

51

This is significant, given the changes that have occurred in the project area since design, e.g., in the form of (i) output prices (especially for rice) and (ii) climate change (incidence of floods and droughts).

52 Eventual costs averaged about $1,188/ha, compared with the design’s maximum criteria of $800/ha.

53 Following the revised feasibility study, done after project design, where it was realized that the command area

would be over 5,000 ha and exceed the subproject selection criteria of 3,000 ha. 54

The project framework was revised in 2011 based on these achievement areas (Appendix 7).

12

44. Two components (institutional strengthening of the water sector, and irrigated agriculture) have not delivered any outputs or have outputs not fully defined in quantitative terms, but there is no evidence from project processes or other sources to suggest that they could have been undertaken at lower cost. However, anecdotal concerns about resource efficiency do surround the quality of civil works at some subproject sites where damage to water management structures and erosion of banks and road surfaces have occurred; this may be due to subcontractors’ negligence and/or incapacity, and also to water user communities’ lack of ability or understanding in accepting responsibility for civil works as handed over to them.

D. Preliminary Assessment of Sustainability

45. The project is less likely to be sustainable. This assessment is based on an appreciation of current institutional circumstances, where the conditions surrounding PIMD and IMT essentially remain unclear or incomplete after well over 18 months since project completion. The absence of a legal framework—governing (i) the status of FWUCs; (ii) the relationship between FWUCs and government agencies in relation to O&M rights and responsibilities; and (iii) water quality and licensing—militates against the likelihood that the operational status quo at the subprojects will continue. The legal basis for river basin management (and for resolving conflict between competing users) and the technical capacity for applying IWRM practices are largely absent, as is MOWRAM’s ability to promulgate three of the four relevant subdecrees. The technical capacity of FWUCs remains modest, their collective experience with collecting ISFs is so far very limited, and MOWRAM’s ability to prioritize and fund irrigation maintenance systematically is weak. While measures to improve the likelihood of institutional sustainability now exist in the Water Resources Management Sector Development Program, it is not yet clear how effective these will be. 46. On the positive side, the demand for project outputs—in the form of paddy rice—is expected to remain very strong, and this means that financial affordability (i.e., for farmers to be able to pay their agreed share of O&M costs) is highly likely, and at least some stakeholders have a high incentive to continue to participate. Evidence garnered in late 2013 quite clearly suggests that where farmers’ production increases thanks to successful subproject management (e.g., at Po Pi Deum and Krouch Saeuch), FWUCs have no problem collecting the agreed ISFs. Conversely, in the economically weaker schemes (e.g., Kouch Nop, Ta Saom, Kork Thnong), ISF collection tends to be low (and falling). A general caveat exists, however, in that rice farming at the subprojects competes with the attractiveness of seasonal and temporary migration opportunities, which farm household members in the project area are increasingly able to access (Appendix 8). Environmental conditions for the subprojects are also positive (if O&M can be maintained)—by better managing local water resources and by using better agriculture practices, soil erosion is avoided, soil quality is improved, and forest loss is avoided. E. Impact

47. The project had several impacts. The most obvious is the aggregate increase in rice production—in the order of 18,000 tons of paddy annually, with a net worth of over $3 million in current financial prices. 55 Given the nature of supply chains in the project area, this flow will likely have contributed to national exports (to Thailand) and also to domestic supply and value addition (via rice mills, some very new, in Battambang and elsewhere). Further, and as reported by many farmers in interviews in 2013, the subprojects have made a fundamental difference to

55

Which excludes production costs to farmers and subprojects.

13

families—before they had to make do with subsistence rice production, now they have income-generating surpluses to sell. 48. The subproject sites also benefit from impacts that so far have not been quantified:

(i) fewer animals dying in the dry season due to lack of water (which can now be retained); (ii) some small areas of dry season vegetable cropping (using retained water); (iii) more commercial services (e.g., shops, transport, traders), many located along access

roads or at junctions;56 (iv) better rural roads (sometimes built as roads, sometimes as dikes) that facilitate access

to health, education, and other social services; (v) more fishing opportunities (since lakes and ponds, many quite shallow, have increased);

and (vi) enhanced tourism and recreational facilities in some places (e.g., lakeside restaurants,

modest sports facilities). 49. The project has also increased social and human capacity. Notwithstanding the earlier comments about the relative weakness of the FWUCs, they have had some technical and administrative training that has at least enabled the weakest of them to still be functioning (even in places where crop damage has been considerable and ISF collection extremely low); only one subproject has so far failed to collect any ISF.57 50. The impact on poverty in the project area will probably have been smaller than expected, due to (i) greater than expected persistence of relatively extensive techniques of broadcasting rice seeds (requiring less labor than does transplanting under the system of rice intensification regime) and (ii) the increasing mechanization now apparent in land preparation, crop harvesting, and transportation. 51. The gender impact of the project appears quite modest—fewer women participate as FWUC members or FWUC management committee members than targeted,58 and gender-specific indicators on irrigated agriculture and income generation apparently were not developed. In agriculture training, on the other hand, 43% of the wider participants were women, and the indicators about gender mainstreaming and inclusion of vulnerable households would have raised awareness.

IV. OVERALL ASSESSMENT AND RECOMMENDATIONS A. Overall Assessment

52. The project is rated partly successful based on the earlier assessments of relevance, effectiveness, efficiency, and likely sustainability.59

56

Some of these are seasonal, some are permanent (at harvest time and just after), but all represent impacts arising directly from the project investments.

57 Large numbers of subproject farmers (accounting for about 10,000 person-days) received agricultural and

livestock training, and partial adoption of recommended practices is widely reported. Large numbers of central government, commune council, and village development committee members (nearly 5,000) received irrigation management training.

58 The project framework did not set any gender-disaggregated targets, and the only target in the gender action plan

was for 30% of FWUC members to be female. 59

The project scored a weighted average of 1.5 against these criteria. For methodology see: ADB. 2006. Guidelines for Preparing Performance Evaluation Reports for Public Sector Operations. Manila.

14

53. The project was an appropriate response to the recognized problems of rural poverty in Cambodia’s northwest in the early part of the 21st century. Its technical composition was adequate, although if detailed formulation had been somewhat more rigorous, the implementation delays (arising from the difficult conditions on some subproject land, lack of service provider capacity, and MOWRAM’s complex procurement procedures) might have been avoided; in that event, the project might have been far more successful in building the strength of farmer institutions and achieving the adoption of more intense rice cultivation and more diverse cropping activities. Despite the problems early on, progress did pick up in response to remedial institutional arrangements, and a reasonable level of outputs was eventually delivered. Project design probably overestimated the domestic appetite for subsector reform in IMT, and took a level of technical engagement in IWRM and PIMD by MOWRAM to indicate a wider political commitment—which in the end was not forthcoming. 54. The project piloted a complementary set of interventions at subproject sites and nationally. The project was implemented in circumstances that have continued to evolve rapidly, yet retained its fundamental relevance. While some economic, social, and institutional successes are now obvious at the subproject level (in part due to the fortuitous rise in value of its primary quantifiable output), it is the lack of national achievement in delivering a refined framework for IMT that compromises the long-term sustainability of the investment. This is the most fundamental shortcoming of the project, particularly in conditions where environment- and climate-related pressures are growing. The fact that measures to support IWRM and IMT, among others, are included in current ADB operations in this subsector indicates that the project’s objectives were highly germane to Cambodia’s needs, but also that its agenda is somewhat unfinished. B. Lessons Learned

55. Probably the most obvious lesson is that it is easy to overestimate the capacity needed to implement a project that is largely undertaking pilot activities and which also involves significant reforms. At all levels of involvement—executing agency, service providers, contractors, FWUC members—capacity constraints affected project performance. In these circumstances, the project design was both overambitious (e.g., by not focusing on fewer subprojects from the outset) and not rigorous enough (in its analysis of the risks and assumptions inherent in many operations). While appropriate and necessary interventions to support capacity development and training were included in the project scope, the dependence of FWUCs gaining experience with improved physical works and for farmers undertaking improved and diversified agriculture upon the completion of civil works was not brought out fully as a project design characteristic—again, partly because the risks and assumptions were not fully thought through. Had the project implementation schedule been largely as intended, sustainability based on effective FWUC functioning would have been likely. 56. Also, a realistic understanding of local procurement practices and an ability to engage with executing agency staff on a daily, face-to-face basis can greatly facilitate implementation. C. Recommendations

1. Project Related

57. Absolutely fundamental to sustainability is the maintenance of the physical irrigation structures—without this, crop yields and production will not be maintained and nonquantified benefits will also deteriorate. Measures to solve this issue are largely contained in the Water

15

Resources Management Sector Development Program (para. 31), such as resources for FWUC operations and further training. 60 Resources are also included for improving MOWRAM’s capacity for IWRM and O&M planning. Successful implementation of that program will help ensure the sustainability of this investment; conversely, continuing delays in disbursement of training funds will have the opposite effect.61 58. ADB should continue to press the IMT reform agenda in Cambodia. Again, the project’s covenants as regards the outstanding subdecrees are embodied in the tranche release conditions of the Water Resources Management Sector Development Program; close management and performance monitoring will be essential. ADB may also consider intensifying the Public Financial Management program to improve local public procurement. 59. The project’s design and implementation experience has yielded significant technical and institutional lessons on the nature of conditions in Cambodia’s water sector; as it is likely that rehabilitation of small and medium-sized irrigation schemes will continue to feature in the sector development agenda, it is essential that these experiences are reflected in ADB’s future analytical work—specifically in the assessment, strategy, and roadmaps prepared as part of the country partnership strategy process. 60. To understand what is happening at subproject sites in terms of agricultural practices, farm incomes, and poverty impacts, it is essential to undertake field-based monitoring. Though costly, MOWRAM and MAFF, and their provincial departments, could collect periodically and systematically data on crop yields, farm budgets, crop sales, and family incomes, among others, (e.g., through a 2–3-year rolling survey program) to capture the true situation.62

2. General 61. Project frameworks (now design and monitoring frameworks) should be applied more rigorously than was the case here, so as to situate the project realistically in its institutional setting and to bring out risks and assumptions implicit in project design, which later may well be key determinants of performance (i.e., because temporal or spatial relationships were not fully understood). When combined with diagnostic material from the sector, the framework should properly be used as a design tool for situation analysis and project identification, rather than as an ex post means of reflecting the investment content.

62. Where the capacity of service providers for unfamiliar outputs (e.g., river basin studies) is weak, a higher level of technical support (e.g., through on-the-job training by an international technical assistance consultants) needs to be included to overcome this, and time must be allowed for the completion of activities.

63. Even quite severe implementation difficulties can be overcome if local staff who are administering the project have the capacity for effective engagement with procurement and other administrative processes, and can recognize problems and expedite solutions in the early years of project life. Where projects involve civil works and/or physical infrastructure, some level of technical support (e.g., from an irrigation engineer) is also vital during review missions.

60

These are largely the resources that the 2011 AFD review of FWUCs recommended—provision of some funds to “prime the pump” for FWUC functioning, for specific accounting training, and for a credit line of technical support from which they could draw down according to their needs (i.e., a demand-driven source for on-the-job training).

61 Realistically, FWUCs probably need 5 years of operational experience to become fully sustainable.

62 A start could be made by incorporating data on cropping practices (collected by the FWUCs) into the Benefit

Monitoring Evaluation module of the Cambodia Information System on Irrigation Schemes.

Ap

pe

ndix

1 1

6

PROJECT FRAMEWORK

Design Summary Performance Indicators Revised Indicators in 20111 Achievements at PCR

Goal Support Government’s effort to reduce poverty in selected northwest rural areas through enhanced agricultural production thereby alleviating food insecurity and improving farming household incomes

Average household incomes raised by at least 30% for 35,000 beneficiary households. Incidence of poverty reduced to 10% of beneficiaries.

Average household incomes raised by at least 30% for 35,000 beneficiary households Incidence of poverty reduced to 10% of beneficiaries

Increased household incomes and reduced poverty are likely achieved exceeding the targets. However, Impact Assessment to be carried out at least three years after beneficiaries have been able to benefit from the systems.

Purpose

Establish rehabilitated and sustainably operational small to medium-scale irrigation schemes and other water control infrastructure within the studied priority river basins

Four basin studies completed using an integrated water resource management (IWRM) approach. 35,000 hectares (ha) of irrigated (16,000 ha) and flood spreading/recession (19,000 ha) agriculture in 10–12 rehabilitated schemes. Increased rice productivity to 3.5 tons per hectare and diversified agricultural production where appropriate.

Four basin studies completed using an integrated water resource management (IWRM) approach. 16,000 hectares (ha) of irrigated areas in 10–12 rehabilitated schemes

2.

Increased rice productivity to 3.5 tons per hectare and diversified agricultural production where appropriate.

Six river basin studies completed. Total project areas: 10,771 ha including reservoir areas in 12 irrigation schemes

3.

Average rice yield was 3.5 tons per ha and agricultural production have diversified where appropriate.

Outputs

Component 1: Water Sector Institutional Strengthening

1.1 A water resources investment strategy with subsequent annual business plans.

Water resources investment strategy adopted by end of year 1. MOWRAM annual business plans generated and followed.

Water resources investment strategy adopted by end of year 1 MOWRAM annual business plans generated and followed

Most of these outputs were transferred Technical Working Group of Agriculture and Water that was established in 2005 and incorporated in Strategy for Agriculture and Water.

1.2 A regulatory framework for executing a refined irrigation management transfer (IMT) policy prepared and applied countrywide in irrigation development initiatives.

A refined IMT policy including sub-decrees on farmer water user communities (FWUCs) and IMT and operational guidelines for FWUC registration.

Operational guidelines distributed.

200 training activities delivered and 40 staff trained.

At least 15 aid agency meetings and 18 coordination initiatives conducted.

A refined O&M handover policy including sub-decrees on farmer water user communities (FWUCs) and IMT and operational guidelines for FWUC registration

Operational guidelines distributed

200 training activities delivered and 40 staff trained

Improved coordination among aid agencies

Four sub-decrees have been drafted of which one the Farmer Water User Community (FWUC) sub-decree is now being discussed at the Council of Ministers. The remaining 3 (River Basin Management, Water Licensing, and Water Quality) have been drafted by MOWRAM and reviewed at workshop late June 2011.

12 FWUCs established and registered with MOWRAM FWUC Department

17 A

ppe

nd

ix 1


1.3 The capacity within MOWRAM/ PDWRAM to adopt an IWRM approach in river basin context.

Improved planning and prioritization of water resources investments.

30 training courses delivered and 20 staff trained.

Topographic and hydrological equipment delivered and installed.

Improved planning and prioritization of water resources investments

Topographic and hydrological equipment delivered and installed in the priority basins of the target provinces

5 Apex Committee formation (adjacent/connected FWUCs) commenced in Q1 2010 and completed Q1 2011. Covers 215 villages and 19 FWUCs in all 4 targeted provinces.

Component 2: Irrigation Infrastructure Development and Management

2.1 Basin water use studies completed in selected basins

By middle of year 2, all four priority basin water use studies completed.

6–8 stand-alone candidate subprojects identified.

By middle of year 2, all four priority basin water use studies completed

6–8 stand-alone candidate subprojects identified by end of year 1

Six river basin studies completed: (i) Daun Try, (ii) Svay Don Keo, (iii) Boribo, (iv) Thlea Maam, (v) Mongol Borei, and (vi) Svay Chek.

2.2 A list of candidate subprojects (either stand-alone or within studied river basins) for feasibility study

Briefs prepared for subprojects based on Output (2.2).

By end of year 1 stand-alone noncore

subprojects selected for feasibility study.

By 3rd quarter of year 2, all remaining noncore subprojects selected for feasibility study.

Budgetary allocations made.

Briefs prepared for subprojects based on Output (iv)

By end of year 1 stand-alone noncore

subprojects selected for feasibility study

By 3rd quarter of year 2, all remaining noncore subprojects selected for feasibility study

Budgetary allocations made

Feasibility study and detailed designs of 12 irrigation schemes have been completed in 2009 and ready for construction works.

2.3 A total of 10–12 irrigation schemes rehabilitated

35,000 ha of irrigated (16,000 ha) and flood spreading/recession (19,000 ha) agricultural land in operation at project completion.

By end of year 1, incremental information on core subprojects collected.

By 1st quarter of year 3, all feasibility studies completed.

By 3rd quarter of year 3, all detailed engineering designs completed.

Resettlement plans submitted and no. of households compensated prior to start of the construction phase.

By end of year 1, one core subproject contract awarded.

By end of year 2, all core subproject contracts awarded.

16,000 ha of irrigated agricultural land in operation at project completion

By end of year 1, incremental information on core subprojects collected

By 1st quarter of year 3, all feasibility studies completed

By 3rd quarter of year 3, all detailed engineering designs completed

Resettlement plans submitted and no. of households compensated prior to start of the construction phase

By end of year 1, one core subproject contract awarded

By end of year 2, all core subproject contracts awarded and two other subproject contracts awarded

Total project areas 10,771 ha including reservoir areas

12 schemes rehabilitated: (i) Punley, (ii) Po Pi Deum, (iii) Kuch Noup, (iv) Krouch Seauch, (v) Kork Thnong, (vi) Ta Saom, (vii) Anlong Svay, (viii) Tram Mneash, (ix) Chork, (x) Don Aov, (xi) Canal #1, and (xii) Kandal Prek.

Ap

pe

ndix

1 1

8


By end of year 2, two other subproject contracts awarded.

By end of year 3, 2–3 other subproject contracts awarded.

By end of year 4, 2–3, two to three other subproject contracts awarded.

By end of year 3, 2–3 other subproject contracts awarded

By end of year 4, 2–3, two to three other subproject contracts awarded

2.4 Structures and canals in all rehabilitated schemes operated and maintained by stakeholders

200 training courses conducted for FWUCs.

72 training courses in IMP procedures conducted for PDWRAM and 20 staff trained.

15-20 IMT agreements signed between MOWRAM and each FWUC.

10-12 FWUC facilities constructed by project completion.

Work plans of irrigation extension teams

Operational maintenance plan implemented.

Irrigation service fees collected.

Budgets allocated for routine and periodic maintenance.

200 training courses conducted for FWUCs

72 training courses conducted for PDWRAM and 20 staff trained

15-20 IMT agreements signed between MOWRAM and each FWUC

FWUC facilities constructed in each subproject site by project completion

Operational maintenance plan implemented

Irrigation service fees collected

Budgets allocated for routine and periodic maintenance.

Total of 468 FWUC training courses/events conducted throughout 12 subprojects.

16 training courses for each PIU – total 64 events with total participation of 1,558 from the Government, 3,347 from Commune Councils and Village Development Committees, and also 22,833 male and 16,704 female beneficiaries.

Component 3: Support to Development of Irrigated Agriculture

3.1 Provincial and district extension staff and IETs with improved technical capabilities in rice and diversified agricultural production and with enhanced extension skills.

Improved capacity of district extension staff.

160 technical training courses and 32 district agriculture office (DAO) staff trained.

At least 12 district offices equipped and resourced by the end of year 3.

Improved capacity of district extension staff

160 technical training courses in irrigated agriculture, integrated farming systems, on-farm water managements practices, etc., and 32 district agriculture office (DAO) staff and all IETs trained

At least 12 district offices equipped and resourced by the end of year 3.

Four Agriculture Service Agreements were signed in June 2009 and agricultural training completed end March 2011

3.2 Diversified and intensified agricultural production activities being undertaken in subproject areas

Increased cropping intensity.

Area of irrigated crops (rice and others) and area where agriculture diversification takes place.

Number of livestock and fishery

Increased cropping intensity in the subproject areas

Area of irrigated crops (rice and others) and area where agriculture diversification takes place

Number of livestock and fishery

Improved agricultural production recorded. Significant dry season rice areas cropped in excess of targets.

19 A

ppe

nd

ix 1


enterprises.

Quantity of certified high-yielding variety seed produced.

Area of seed multiplication farms.

Commodity price databases analyzed and disseminated monthly.

4 provincial livestock disease diagnostic laboratories functioning by end of year 2.

240 village livestock assistants trained and kits distributed by end of year 3.

enterprises

Quantity of certified high-yielding variety seed produced

Area of seed multiplication farms

Commodity price databases analyzed and disseminated monthly

4 laboratories functioning by end of year 2

240 kits distributed by end of year 3

3.3 FWUCs being used as vehicles for farmer lead extension and other group initiatives

10–12 functioning FWUCs by end of year 3.

Conduct at least 200 farmer field schools and demonstrations.

Conduct at least 24 farmer cross visits

About 50 ha of demonstration blocks for each scheme.

No. of training courses on group organization, administration, and resource mobilization conducted and no. of people trained.

No. of FWUCs acting as credit intermediaries.

No. of FWUCs acting as marketing agents or input suppliers.

A functioning FWUC in each selected subproject s by end of year 3

Conduct at least 200 farmer field schools and demonstrations

Conduct at least 24 farmer cross visits

About 50 ha of demonstration blocks for each scheme

No. of training courses conducted and no. of people trained

No. of FWUCs acting as credit intermediaries

No. of FWUCs acting as marketing agents or input suppliers

12 FWUCs established by end of Year 3 but strengthening and training ongoing until end Year 6. Training included Exchange Visits by FWUC representatives to 2 other projects.

Farmer Field School and selection of farmer training groups based upon WUG and block layout.

120 demonstrations at each subproject.

Two training courses on group organization and administration for each FWUC conducted.

Total participation days total 10,000 comprised men: 4,588 and women: 5,412

Future roles not yet adopted as FWUCs activated upon completion of irrigation rehabilitation and by project closure had only just begun to O&M their schemes.

1 It was necessary to revise the project framework for subsequent realistic project monitoring during the review in February 2011. The proposed change in scope was discussed and documented in the aide-memoire. The change was approved by Country Director.

2 The reduction in irrigated areas was due to the exclusion of the Chi Kreng and Link Canals that contained large areas of flood spreading among the sub-projects to be studied when the Project commenced. It was found that these sub-projects did not fit the sub-project selection criteria.

3 The shortfall was due to (i) the engineer’s cost estimates during the design period were lower than the actual costs, and (ii) the price escalations from mid-2007 to mid-2008, which led to shortage of funds for civil works.

Appendix 2 20

APPRAISAL AND ACTUAL PROJECT COSTS

Component Appraisal Estimate Actual

Foreign Currency ($ million)

Local Currency ($ million)

Total Cost

($ million)

Foreign Currency ($ million)

Local Currency ($ million)

Total Cost

($ million)

Civil works 3.959 9.981 13.940 3.478 10.061 13.539

Resettlement compensation – 1.374 1.374 – – –

Demining – 0.330 0.330 – – –

Vehicles and equipment 1.118 0.636 1.754 0.528 0.009 0.537

Training and study tours 0.029 1.073 1.102 0.035 1.128 1.163

Implementation and facilitation 0.224 1.185 1.409 0.321 1.292 1.613

Service providers 4.204 3.407 7.611 6.273 3.333 9.606

O&M infrastructure – 1.076 1.076 – – –

Price contingency 0.421 1.330 1.751 – – –

Interest during implementation 0.525 – 0.525 0.480 – 0.480

Total 10.480 20.392 30.872 11.115 15.824 26.938

– = not applicable or not available, O&M = operation and maintenance. Source: Asian Development Bank.

21 Appendix 3

PHYSICAL IRRIGATION INFRASTRUCTURE PROVISION A. Introduction 1. This appendix provides an overview of the process and history of physical irrigation infrastructure provision under the project, and discusses some major issues that arose during implementation. B. Subproject Identification and Selection 2. At project preparation, a list of 35 eligible subprojects had already been drawn up, and five of them had been studied for feasibility: (i) Thlea Maom—rehabilitation of a river diversion scheme in Pursat;1 (ii) Link Canal—construction of a transfer canal to increase the water supply to an existing reservoir in Battambang; (iii) Po Pi Deum—redesigning an irrigation scheme and rehabilitating key water control infrastructure; (iv) Ponley—rehabilitating and/or upgrading a reservoir and its command area in Banteay Meanchey; and (v) Stung Chikraeng—rehabilitating and/or upgrading a reservoir and its command area in Siem Reap. Financial and economic analysis of four of these potential subprojects was undertaken at appraisal, and included in the report and recommendation of the President (RRP). 3. The original 35 eligible subprojects are shown in Table A3.1. A distinction was made between (i) stand-alone schemes, for which it was assumed that the required water resources were not dependent on any water user located upstream and that the planned use of water after rehabilitation had no consequences on any water user located downstream; and (ii) schemes within river basins (for which conflicts over water use might be assumed to exist). For the stand-alone candidate subprojects, river basin and water use studies were considered redundant. 4. One more subproject—Preks of Kandal—was later added to this list to fit the objectives of cofinancier Agence Française de Développement (AFD), and some other potential subprojects in Siem Reap were also considered but later dropped. Necessary reconfirmation of feasibility work in line with Asian Development Bank (ADB) requirements at Stung Chikraeng (funded by AFD) raised issues over potential resettlement costs that led to its being dropping from the project scope. This had the effect of removing the 19,000 hectares (ha) of spreading or recession agriculture from the project targets, since they would have been located at this subproject. The subprojects at Naktaror and Preks were also later dropped because they fell within the subsequent designation of the Tonle Sap Biosphere Reserve. Overall, 47 potential subprojects (21 in river basins, 26 stand-alone) were considered during project implementation. 5. The selection of subprojects involved a four-phased process: identification, river basin and water use studies, feasibility study, and detailed design. The anticipated processes of subproject selection and the criteria on which this process was based at different stages are summarized in Table A3.3. C. River Basin and Water Use Studies 6. The river basin and water use studies were to provide a major part of the technical basis for subsequent master planning of subprojects. The studies were conducted (in 2006) by competitively selected service providers in packages combining several river basins. Package 1

1 This subproject was rejected because of poor economic returns and engineering and environmental issues, and

did not make the list of eligible subprojects.

Appendix 3 22

covered the Mongkol Borei sub-basin in Battambang, Banteay Meanchey, and Pailin provinces and the Svay Chek sub-basin in Banteay Meanchey Province. Package 2 covered the Dauntri sub-basin in Battambang and Pursat provinces, and the Boribo sub-basin in Pursat and Kg Chhnang provinces (and with a small corner in Kampong Speu Province).

Table A3.1: Original List of Eligible Subprojects

Province River Basin Scheme Identification Code

(Halcrow Report)

Indicative Area (ha) a

Wet Season Supplementary

Irrigation

Dry Season Irrigation

Pursat Boribo-Thleamaorm Srang Bomnork KK3 1,750 50 Tram Mneash KK51 1,200 60 Dang Toekkleach N/A 500 30

Stand-alone schemes Kouch Nup Cluster KK8 to KK11 2,000 60 Angkonh KK6 505 40 Damnak Kranh KK13 450 30

Naktaror KK14 400 30 Kanseng KK7 600 30 Voat Rung N/A 500 30

Anlong Svay BK2 220 0 Raneam Prayal BK3 300 0 Krouch Seuch BK5 300 0

Battambang Stung Mongkol Borei O Don Pov Link Canal b BV51 4,450 2,850 Presh Khposh BV50 1,200 100 Canal No. 1 (Bovel) BV1 3,000 400

Don Aov (Bovel) BV50 1,800 300 Stung Sangker O Kroch Reservoir N/A 200 20 Beong Ampil Reservoir RM5 250 50

Daun Try-Svay Don Keo Prek Chik M51 3,000 200 Stand-alone schemes Koas Kralor KK50 & KK51 800 50 Chork Reservoir M53 1,500 100 Preks on Stung Sangker DP53 3,000 500

Banteay Meanchey

Stung Mongkol Borei Po Pi Deum b MBR1 1,400 350 Prolay Cha MBR51 820 50

Stung Svay Chek Pol Sena SCH50 1,530 10

Sre Laor TP52 1,200 20 Kork Prich TP51 1,200 20 Anlong Rot SCH1 1,200 70

Thmar Kdep TP50 2,100 50 Baray SCH54 1,800 50 Stand-alone schemes Kanseng OC54 2,450 150

Ta Meng TP53 2,400 30 Prasat Pram SCH3 550 20 Ponley b PS51 675 75

Siem Reap Stand-alone schemes Chikraeng b (phases 1 & 2) --- 1,240 1,240

ha = hectare. a These figures refer only to the estimated potential command areas for fully controlled irrigation practices (supplemental irrigation during wet season and full irrigation during dry season). The estimated potential command areas for flood spreading and/or recession practices is not shown as accurate detailed breakdown by scheme will only be known after the feasibility study of the noncore subprojects. b Core subproject studied to feasibility level during project preparation. Source: Report and recommendation of the President, Appendix 4.

D. Feasibility Studies 7. Feasibility studies for the stand-alone schemes (i.e., those not requiring river basin and water use studies) were conducted from early 2007. They were initially carried out for 12 subprojects in three packages—Pursat (4 candidate subprojects), Battambang (3), and Siem Reap (5).

23 Appendix 3

8. Another 12 river basin feasibility studies were also conducted: 1 in Banteay Meanchey (Po Pi Deum); 6 in Pursat (Bomnork, Tram Mneash, Kroch Saeuch, Aniong Svay, Roneam Prayol, and Thlea Maam), 5 in Battambang (Prek Chik, Prey Khpos, Canal No. 1, Don Aov, Link Canal).2 In total, 28 feasibility studies were conducted during project implementation. E. Detailed Design and Construction 9. On the basis of the stated selection criteria, 12 subprojects (6 stand-alone, 6 river basin) eventually proceeded to detailed design; this work was contractually included in the feasibility study work so as to overcome delays and difficulties due to the limited capacity of service providers (and also because of resettlement issues at the most advanced of the subprojects’ sites, Ponley). The Ministry of Water Resources and Meteorology (MOWRAM) and its provincial departments undertook detailed design at two sites themselves, although with technical assistance (TA). 10. All subprojects with detailed design work were prepared to a largely standard format—including an economic analysis along the lines of the one in the RRP 3 and eventually rehabilitated by the project. These subprojects were Ponley (command area 440 hectares [ha], reservoir 357 ha), Po Pi Deum (1,279 ha), Kuch Noup (790 ha), Krouch Seauch (1,000 ha), Kork Thnong (command area 154, reservoir 45 ha), Ta Saom (command area 196 ha, reservoir 84 ha), Anlong Svay (795 ha), Tram Mneash (1,212 ha), Chork (command area 1,372, reservoir 596 ha), Don Aov (1,100 ha), Link Canal (1,061 ha), and Preks of Kandal (98 ha of polder; not included in the original project scope). 11. The gross supplementally irrigable area (for wet season cropping) of all subprojects was 10,761 ha (including reservoir areas). Irrigated dry season cropping would be possible in some of these areas, but it would take time for farmers to be able to fully exploit this potential. 12. Procurement totaled 16 contract packages at a total cost of $12.68 million, covering all civil works for the 12 subprojects and employing six contractors. Contracts were signed between 2008 and 2010, and construction started in 2009 and was completed in 2010. The defect liability period extended through much of 2011 for most subprojects. 13. Because of extreme weather events—Typhoon Ketsana in 2009 and very heavy rains in mid-October 2010—that damaged some subprojects’ physical works, it was decided to issue a variation order (No. 2) to cover civil works to mitigate the flood damage. The cost of repairs to existing works (and some redesign to allow for future contingency) was $446,956 (and thus brought the total costs of civil works on physical irrigation infrastructure to $13.12 million). F. Issues during Project Implementation 14. Overall there was a significant evolution of practice and pace in subprojects’ selection, design, and construction during project implementation. Various issues contributed to the eventual nature and level of physical output delivery, among them:

(i) Underestimation of subprojects’ rehabilitation costs. At appraisal it had been estimated that the investment costs of irrigated areas would be less than $800/ha. However, it became clear at the feasibility stage of several subprojects

2 Feasibility studies for candidate subprojects within the river basins followed the completion of the river basin and

water use studies. 3 For details see Appendix 8.

Appendix 3 24

that the construction costs would be significantly higher. Two reasons essentially accounted for this: (a) The level of required rehabilitation had been somewhat underestimated. Originally assumed to be only minimal, it turned out that most subprojects had been constructed or rehabilitated under the Pol Pot regime with little regard for hydrology or engineering design and construction principles. Much of this infrastructure had also been subsequently damaged during the conflict years (by mortar and artillery fire, and by mines). Structures and earthworks thus required checking for integrity, which also added to subproject costs. Except for three subprojects, major diversion works were necessary, some even requiring expensive spillways. Major earthworks were typically required, and also minor hydraulic structures. The delineation of boundaries between individual basins occasionally was made more complex by the general flatness of the surrounding topography. (b) Appraisal cost estimates were prepared in 2002 when basic prices were considerably lower than when construction was due to start (i.e., over 5 years later). Prices for materials (e.g., steel, cement), fuel, and labor had in some cases as much as tripled. According to an analysis of costs for a typical irrigation scheme in both 2003 and 2008 (discussed with the midterm review mission in 2008), costs had approximately doubled in that period. The higher development costs inevitably affected the total project area that could be realistically covered by the agreed funding, and also affected loan allocations for the civil works. Eventually a reallocation of resources and a downscaling of the target area was required. The final costs of civil works and command areas realized meant that financial costs averaged $1,188/ha.

(ii) Capacity of local service providers. The project design envisaged the use of local service providers to undertake river basin and water use studies, feasibility studies, and detailed design. However, it became very clear during the early years of the project that the capacity of local service providers was very limited and acted as a major constraint on project implementation performance. The limited capacity to adequately deal with the demands of study and design terms of reference meant that delivered outputs were often inconclusive, incomplete, and failed to provide a clear indication of what was being proposed. In many instances, the quality of the outputs of the service providers necessarily required numerous revisions from the TA consultants—and consumed significantly more time inputs of the project consultants than was originally envisaged in achieving a reasonable standard of the service providers’ outputs. While the studies were eventually appropriate and in effect necessary for subprojects’ prioritization and selection—e.g., based on these studies, 7 of 21 candidate subprojects were dropped, and the other 14 were cleared for feasibility studies—interactions between the project implementation unit, TA consultants, and service providers were somewhat limited, which contributed to delays in reviewing and finalizing documents and plans. After the 2008 midterm review, processes for reviewing the service providers’ outputs were sped up, and it was agreed to focus on the likely most viable subprojects, as well as MOWRAM taking over the Ta Soum and Kouk Thnong subprojects in Siem Riep Province for further design (para. 9).

(iii) Delays in contracting and procurement. Procurement and recruitment of service providers was initially an extremely slow process. By the time of the midterm review, and despite earlier efforts to improve procurement of service contracts for the river basin and feasibility studies, and for detailed design, it took on average about 12 months to complete procurement. With two-thirds of elapsed implementation time gone, overall progress of the project was estimated at about

25 Appendix 3

25%. The year 2007 was in fact almost entirely devoted to procurement of studies by service providers. Despite diverging views of the joint ADB–AFD midterm review mission4 and the project management office (PMO) at the time, the recommendations of the mission regarding the composition of the procurement review committee, the role of project director in chairing the committee meetings, and stricter adherence to standard operating procedures (as well as eliminating unnecessary and overly centralized internal procedures) were largely followed and procurement processes picked up. A key meeting in December 2008 between ADB and the PMO also agreed to adopt time-bound processes for reviewing procurement documents. Later review missions (from 2009 onward) noted improved and accelerated procurement practices, which contributed to the eventual output delivery.

(iv) Land titling issues. The project design included mapping and inventory of landholdings within subprojects (necessary above all for resettlement plans, and also for identifying government rights-of-way and government-owned land). It was envisaged that this would result in a registry of landholdings and ownership records to be held at district land registration offices and that, with the involvement of Ministry of Land Management, Urban Planning and Construction (MLMUPC) (as the mandated agency), cadastral surveys and issuance of land titles would continue. Landholding maps were indeed produced by the service providers for the 12 subprojects completed, but it was not possible to get endorsement from the Provincial Departments of Land Management, Urban Planning and Construction (PDLMUPCs) to proceed to the issuance of land tenure certificates (through the commune councils).5 Thus, MOWRAM decided to discontinue attempts at providing land tenure certificates as planned, and instead have the provincial entities directly issue these. But this could not be done with project resources, and it was anticipated at project completion that other resources would be employed.

(v) Resettlement planning. This proved more complex than anticipated at design, and meant revising arrangements during implementation to avoid delaying construction works.6

4 The joint ADB–AFD mission understood that the general technical evaluation of any bid proposal was first carried

out by the management consultant team, which then submitted a draft evaluation report to the project office manager and evaluation committee of the PMO for review and follow-up with the Ministry of Economy and Finance (MEF). However, it remained unclear to the mission how the rest of the procurement process operated (how the draft technical evaluation report was reviewed by the PMO, how the procurement review committee (PRC) meetings were convened, who chaired the PRC, how and by whom its minutes were prepared, and what was the process for signing the minutes). Coming to an agreement and signing the PRC meeting minutes appeared to take a lot of time and to be the main bottleneck but, in the absence of a detailed tracking system, it was unclear which members of the PRC (MOWRAM and/or MEF representatives), if any, were to be held responsible for these delays. It was the view of the mission that the overall procurement process was further exacerbated by MOWRAM’s internal administrative procedures, which required formal endorsement by higher levels of management before submission to ADB for “no objection”. This practice did not follow the 26 February 2007 Sub-decree (No. 14 ANKR.BK) on Standard Operation Procedures for Implementing Projects Financed by the World Bank and ADB, which included the standard procedure for procurement. According to the PMO, a key reason for the delays, despite all its efforts to avoid them, was largely that about 6 months had been lost in recruiting service providers for the river basin studies and other studies because the time it took to comply with the MOWRAM’s and MEF’s administration procedures had been underestimated.

5 Apart from any other institutional or technical issues, the work involved for the project alone was considerable;

about 10,000 titles (plots surveyed and registered) would be needed. 6 Appendix 6 contains a fuller description of resettlement issues.

Appendix 3 26

G. Post-Project Completion Experience 15. Since civil works for the subprojects ended (in 2010), several factors have affected performance. Table A3.2 summarizes some of the issues reported and observable from fieldwork in late 2013. A few key points emerge:

(i) Factors outside the immediate environment of a subproject can greatly affect system performance. This particularly applies where major civil works for surrounding water resource infrastructure affects water supply; the effects can be negative (e.g., Kouch Nop, Chork) or positive (e.g., Krouch Saeuch), as Table A3.2 shows.

(ii) Competition for water resources is clearly observable in several places (e.g., Canal No. 1, Krouch Saeuch), reinforcing the logic for integrated water resources management (IWRM) approaches in general and for undertaking project-specific river basin studies in particular.

(iii) Floods appear to have intensified in frequency and severity in recent years, although the actual effects will vary considerably by location. Forest loss in recent years exacerbates the speed and intensity of floods.

(iv) Areas of sandy soils, general flatness, and shallowness of streams and reservoirs make irrigation planning and management especially complex, and prone to the effects of surrounding topography.

(v) The civil works themselves have created some issues due to contractor’s management regarding the quality of work (e.g., at Anlong Svay) and design (e.g., at Don Aov, where the reduction of system costs has now led to some operational problems).

(vi) Local traffic is heavier than anticipated in many places (partly due to an increase in farm mechanization and crop marketing), which has contributed to greater deterioration of access roads than expected.

H. Summary 16. Overall, project planning for the subprojects’ construction and/or rehabilitation was systematic and logical. The river basin studies, feasibility studies, and detailed designs formed a coherent analytical chain, although the number of potential subprojects considered was perhaps unmanageable, and the logic for inclusion sometimes arbitrary. Implementation was generally more complex than originally expected—both in practice and as regards internal project management processes. Some implementation problems encountered could have been better anticipated and possibly mitigated at design (most obviously regarding the capacity of service providers, for whom some kind of training and support could have been envisaged). Nevertheless, despite the delays up to the midterm review, improved service provider procurement, faster resettlement planning, and a focus of project efforts on a revised schedule with fewer implementable subprojects did result in timely construction of civil works within the project lifetime. 17. Factors external to the immediate environment of the subprojects now govern irrigation system performance just as much as the subproject management arrangements. Despite major flooding events in the years since construction, most systems are still holding up relatively well, although significant damage is apparent in several places and needs to be addressed as a matter of priority to sustain delivery of investment benefits.

27 Appendix 3

Table A3.2: Summary of Subprojects’ Physical Management Since Completion

Subproject Physical Management Experience

Kouch Nop • Upstream irrigation system rehabilitation for large-scale cassava cultivation (Chinese investment) means subproject is getting less water than planned at design

Tram Mneash

• Significant damage due to flood in 2102—main canal, access roads and gates. • Sandy soils in places.

Krouch Saeuch

• Access roads being damaged by local machinery use. • System water was released due to local political pressure for downstream use.

Floods in 2012 resulted in 40% crop damage. • Local canal rehabilitation made more water available than planned. • Less sandy soils.

Anlong Svay

• 3 years of floods since completion, and drought in mid-2012. • Sandy soils and shallow rivers on flat topography make irrigation difficult. Some

embankment collapses (“dragon holes”)—possibly a construction quality issue (FWUC did not understand its role?)

Chork • Road damage due to October 2103 flooding (twice); 5 days of continual rain. • Upstream canal rehabilitation under way, which blocked supply in 2012, but should

ultimately increase water supply by 30%–40%. • Flooding now more quick due to local forest losses.

Canal No. 1 • Demands for water from farmers adjacent to system in the dry season. Small flood in 2011, but no damage.

• Embankment overflows in 2013, and crop damage (up to 50%). • Some dragon holes, but system is still working. • Local traffic damages system access roads.

Don Aov • Floods in all 3 years since construction (Mongkul Bareay overflows)—most serious in 2013. “Try to do 3 crops, but get 2.”

• Part of the reason is inadequate drainage capacity in system—this was recognized as a possibility at design, but costs were cut.

Po Pi Deum • Floods in all 3 years since construction (water comes over raised dyke). Pattern is “early season crop, flooding in October, then another crop.”

• Some damage to main canal; community mobilized to repair it. Ponley • Floods in 2013, but not 2011 and 2012. Ta Saom • Sandy soils, “not enough water” in 2012, but good in 2013. Kork Thnong

• Sandy soils; marginal improvement is retaining some water in dry season. • Some flood damage in 2012, major flood damage in 2013 as dyke overflowed and

laterite washed away.

Ap

pe

ndix

3 2

8

Table A3.3: Subproject Selection Criteria and Process

29 A

ppe

nd

ix 3

Source: Report and recommendation of the President, Appendix 5.

Appendix 4 30

FARMERS’ WATER USER COMMUNITIES

A. Introduction 1. At design, the intention was for the project to support government policy to achieve sustainability in irrigation by strengthening beneficiary communities’ capacity for management and operation and maintenance (O&M) of irrigation infrastructure and gradually transferring to them the responsibilities for (and ultimately ownership of) the schemes. The policy and legislative background to this was established in June 20001 , when the Ministry of Water Resources and Meteorology (MOWRAM) issued Prakas 306, an edict that provided the framework for introducing an irrigation management transfer (IMT) process to increase participatory management and to rationalize the sharing of O&M costs between the government and beneficiary communities. Prakas 306 covers (i) the formation, registration, and strengthening of Farmer Water User Communities (FWUCs) and their subordinate water user groups (WUGs); (ii) their empowerment in managing irrigation schemes, particularly in collecting irrigation service fees to cover the cost of water delivery and O&M; and (iii) the gradual transfer of O&M responsibility to FWUCs and their eventual ownership of the irrigation infrastructure.2 2. The project aimed to strengthen the capacity of MOWRAM’s provincial departments (PDWRAMs) for community and group organization, for establishing WUGs and FWUCs (consistent with the operational rules and regulations to be developed under the institutional strengthening component of the project), and for promoting irrigation scheme management, water management, and irrigation scheduling. PDWRAM staff were to be trained to instruct FWUCs and WUGs in management and O&M procedures, as well as in accounting and financial planning and reporting, conflict resolution, and mainstreaming gender and vulnerability issues. 3. The project recognized the limited resources of the PDWRAMs in this area and planned to enlist the private sector to further support the establishment and strengthening of FWUCs. To ensure successful implementation of this subcomponent, irrigation extension workers were to assist PDWRAM staff in each subproject. These experts, contracted as service providers, were to (i) field an irrigation extension team (IET) for each subproject with expertise in community organization (including gender development and mainstreaming), irrigation water management and agricultural production, and irrigation scheme management and O&M; (ii) help establish FWUCs and train management and members in operational matters; (iii) help negotiate formal agreements on subproject implementation and IMT between MOWRAM, local authorities, and FWUCs; (iv) assist the project implementation unit (PIU) and provincial authorities in facilitating the resettlement and land adjustment or consolidation process; and (v) support the irrigated agriculture development component. 4. In collaboration with other ongoing irrigation interventions, MOWRAM had already developed initial guidelines for participatory irrigation management development (PIMD) and training modules, and these were to be used as a starting point for operations.

1 MOWRAM. 2000. Prakas (Edict) No. 306 on adoption of basic instruments for Establishment and Development of

FWUC. Phnom Penh 2 Actual implementation of Prakas 306 varied widely between provinces, districts, and even schemes. Only about

100 FWUCs had been formed nationwide at the start of the project, and most were functioning only to a very limited extent.

31 Appendix 4

B. Establishment and Organization 5. All FWUCs are organized along the same lines and have the same relationship with their constituent WUGs, local communities, and various levels of public administration. Farmers form WUGs according to the physical layout of irrigation blocks, and each WUG elects a leader and a deputy. The WUG leaders then elect the FWUC management committee from among their numbers (Figure 1), which means that some WUG leaders may have a dual leadership function.

Figure 1: Farmers’ Water User Community Structure

6. A FWUC will interact with the local government through an advisory group that typically comprises the FWUC’s general manager, representatives of the commune council(s), 3 a PDWRAM representative, and village heads. 7. Once FWUC elections have been held and approved by the district, and the statute and internal rules are finalized (and an advisory group established), the FWUC is ready for registration. Registration is a critical step because it creates a legal entity that may then enter into contractual arrangements, such as handover of the irrigation assets under a management transfer agreement with the state.4 C. Roles and Responsibilities 8. Generally, the FWUC assumes responsibility for maintaining and operating secondary irrigation assets, while PDWRAMs and MOWRAM remain responsible for primary assets. All the canal systems and annex structures in the subprojects remain the property of MOWRAM. The FWUC is empowered to operate, manage, and maintain the operation of lower-level (e.g., secondary, tertiary, or quaternary) distribution canals.5

3 Depending on whether a FWUC covers more than one commune.

4 To be registered, the FWUC would submit to the PDWRAM copies of: (i) draft organigram of each WUG and

FWUC level, with names and job descriptions (includes advisory group); (ii) approved FWUC statute and bylaws; (iii) internal rules and regulations; (iv) list of members; (v) landholding map; (vi) system design and canal layouts; vii) irrigation system inventory. The PDWRAM will check the documents and send the application to the responsible unit of MOWRAM. Registration is announced through a Prakas, and the FWUC is given a registration number. The Ministry of Interior will then authorize issuance (through MOWRAM) of the FWUC seal.

5 The FWUC may use any materials left on the site after construction of the irrigation subproject for operation, maintenance, or repair of the irrigation structures.

Head WUG1

Head WUG2

Head WUG3

Head WUG4

Head WUG5

Head WUG6

FWUC General Manager

Deputy O&M Manager

Deputy Irrigation Manager

Treasurer

Appendix 4 32

9. FWUC management is responsible for (i) encouraging members’ participation in, and ongoing development of, the FWUC; (ii) convening an annual general assembly, and other meetings with FWUC members as needed; (iii) holding FWUC management committee elections every 3 years and filling vacant committee positions; (iv) managing the FWUC by disseminating and enforcing its rules and regulations, and ensuring that the irrigation service fee (ISF) is collected and properly accounted for; (v) increasing the efficiency of land and water use by preparing seasonal irrigation service plans that ensure adequate water supply to the members; (vi) reviewing the ISF periodically and adjust as needed by agreement with members; (vii) facilitating support from the government and other development partners to assist in repair, maintenance, improvement of infrastructure, and agricultural development; (viii) mobilizing local resources of labour and finance to support routine and periodic O&M of the secondary and lower parts of the irrigation scheme; (ix) carrying out annual inspections of the secondary and tertiary parts of the irrigation scheme, and filing and submitting a report to the PDWRAM through the apex committee6 and advisory group, including estimates and timing of periodic repairs, and status of their implementation; (x) working with the apex committee and advisory group to coordinate irrigation service plans with other FWUCs that share the main system; and (xi) resolve disputes about water distribution between WUGs and members. 10. The FWUC is responsible for repair and maintenance of lower-level canals, drains, and annex structures. Immediate repair and pre-seasonal clearing and cleaning, and periodic maintenance, is required for (i) secondary and lower parts of the distribution system (secondary, tertiary, and quaternary, if existing), including distribution canal structures (check mechanism or regulator, draining structure at the distribution canal tips, side spillways, structures in the middle of the canals, inlet structure to the block, bridges, etc.); (ii) regulators within the secondary drainage canal, and bridges along the main drainage canal; (iii) block-level draining structures (tail escape structure at the end of the tertiary canal, regulators in the canals); (iv) block irrigation supply canal, block draining structures and structures along the canals (tail escape structure at the end of the canal, tail structure at the end of the block irrigation supply canal, regulator in the canals); and (v) access roads within and surrounding the distribution system. The FWUC is further tasked to manage and distribute water in the irrigation system by regulating the water flow or monitoring various annex structures: (ii) distribution canals (secondary, tertiary, and quaternary, if existing); (iii) regulators of secondary and tertiary canals and associated structures; (iv) regulator installed at the secondary drain; and (v) regulator installed at the block drain. D. Provincial Department’s Roles and Responsibilities 11. The PDWRAM’s role in PIMD is to (i) provide guidance and support to the FWUC and subsidiary WUGs at all times; (ii) ensure PDWRAM participation in and chairing of the apex committee,7 and inclusion in the advisory group, and take a lead role in coordinating irrigation service planning and water scheduling between FWUCs; (iii) convene regular apex committee and advisory group meetings to ensure that FWUCs prepare, submit, and implement agreed

6 Please see footnote 34 of the man text.

7 This is a coordination body comprised of secondary users of shared main system infrastructure and water source.

It comprises the heads or deputies of each participating FWUC and is chaired by the nominated PDWRAM officer. The committee essentially replaces the multi-user committee envisaged at project design, charged with supporting IWRM principles and practices. An apex committee thus facilitates system-specific irrigation planning, resolves potential conflicts between FWUCs, and supervises the provision of PDWRAM services.

33 Appendix 4

water schedules in response to service plans; (iv) ensure that operational commitments through the apex and advisory group are properly managed and resourced; (v) ensure that the main system supplies adequate water for irrigation; (vi) sustainably operate and maintain the main system (main canal, head works, primary drains) of the irrigation scheme, with funding provided by the government or, where operation and/or maintenance of components of the main system are delegated to the FWUC, provide adequate funding to the FWUC; (vii) play a lead role in resolving water supply conflicts between FWUCs; and (viii) monitor FWUC progress and management performance and report back to MOWRAM. 12. The PDWRAM is also responsible for (i) carrying out annual inspections of the main system of the irrigation scheme, and filing and submitting reports to MOWRAM (including estimates and timing of necessary repairs, and status of their implementation); (ii) through MOWRAM, ensuring the provision of cash O&M funds to the FWUC, which will gradually decrease over the first 5 years as set out in MOWRAM Circular No 1); and (iii) ensuring immediate repairs to, pre-seasonal canal clearing and cleaning, and periodical maintenance of water reservoirs and other structures on the dam of the reservoir (e.g., spillway, sluice gate, tailrace), main canals and sluice gates, main drainage canals, as well as surrounding roads. 13. The PDWRAM is also charged with monitoring and evaluating the condition of the main canals, main draining canals, roads surrounding the main canals, main draining canals, and annex structures. In the event of any damage to the irrigation subproject, as a result of any natural disaster, the PDWRAM shall be liable for restoration or reconstruction of the affected parts of the canals or structures. 14. The PDWRAM is further responsible for management and distribution of water from the main irrigation system to the irrigation subproject by regulating the water flow or monitoring certain structures, i.e., the main canal’s sluice gate, tailrace, spillway, and the environmental flow sluice gate. It also needs to attend weekly meetings with FWUC authorities to determine the water level during distribution for irrigation. E. Support and Training 15. Within this scope, each service provider was required to undertake at least 89 training events per subproject,8 as follows:

Table A4.1: Training – Subject Areas and Events Per Subproject

No. Subject Areas Includes No. of Events

1 PIMD, IMT and FWUC policy Central government policy initiatives, responsibilities of stakeholders

6

2 Community orientation, Organizing gender/ vulnerable Mainstreaming

Need for participation, sensitivity to vulnerable households. Representative and equitable management. Implementing gender action plan. Construction and statute subcommittees

6

3 FWUC formation and development

Roles and responsibilities, functions. Organization, selection, need for rules and regulations.

6

4 Statute subcommittee Formulation of statute, methodology, review, role and responsibilities

12

5 Construction subcommittee Role and responsibilities in survey, design and construction, 12

8 As per the IET’s terms of reference, the training covered various topics—e.g., policy background, establishment of

FWUCs, irrigation system O&M, accounting and financial management, impact evaluation, and exchange visits or study tours.

Appendix 4 34

coordination with PIU and contractors 6 FWUC elections, committee

selection Candidatures, profiles, tasks and roles of FWUC functionaries, elections and selection

18

7 Basic administrative management of FWUCs

Tasks required, record keeping, administrative procedures. Filing, reporting. Meetings, minutes. Annual general meeting—quorum.

2

8 Irrigation system management

The general system, reservoir operation, standard procedures: crops, planting dates, water, rain, scheduling, indents, control, rotational supplies, water management control station

5

9 O&M General introduction. Operation, maintenance, O&M standard procedures and special procedures.

5

10 Repairs and maintenance The general maintenance system, standard procedures: inspections, prioritizing, budgeting, implementation, reporting, planning.

5

11 Irrigation service planning Scheme operations, maintenance schedules, block and group plans, budgets, ISF charges review

3

12 FWUC accounting Irrigation service fee collection, bookkeeping, reporting, ISF setting and review

3

13 Conflict resolution Procedures and mechanisms for resolving disputes between members and between FWUCs.

1

14 Financial management Advanced training for FWUC management, financial planning, audit, credit, commercial development

1

15 Participatory impact evaluation Annual evaluation of FWUC activities by beneficiaries 3 16 FWUC exchange visits In-country 1 Total Minimum Number of Events by Each Service Provider 89

FWUC=Farmer Water User Community; ISF=Irrigation Service Fee; IMT=Irrigation Management Transfer; O&M=Operation and Maintenance;; PIMD=Participatory Irrigation Management and Development; Source: Ministry of Water Recourses and Meteorology, Project Completion Report

16. Participation in the FWUC training events was high (nearly 40,000 individual training attendances), and some 44% of trainees were female.

Table A4.2: Summary of Attendance at Training Events

Site RGC, PIU Staff

CC, VDC WUG Members % Female Total

Male Female Ponley 410 58 4,615 4,033 46.64% 9,116 Po Pi Deum 186 526 1,697 784 31.60% 3,193 Krouch Saeuch 95 271 1,244 796 39.02% 2,406 Canal No. 1 n/a 235 1,672 581 25.97% 2,488 Chork 27 235 1,472 581 28.30% 2,315 Anlong Svay 133 373 1,518 867 36.35% 2,891 Don Aov 248 614 3,027 1,620 34.86% 5,509

Kork Thnong n/a 290 622 804 56.38% 1,716 Ta Saom n/a 268 551 954 63.39% 1,773 Tram Mneash 268 277 2,120 2,542 54.53% 5,207 Kouch Nup 191 200 2,532 n/a 2,923 Kandal Preks n/a n/a 291 97 25.00% 388 Total 1,558 3,347 35,000 44.19% 39.925

* The training for Preks of Kandal was completed in 2011. CC=Commune Council; PIU=Project Implementation Unit; RGC=Royal Government of Cambodia; VDC=Village Development Committee; WUG=Water User Group; Source: Ministry of Water Recourses and Meteorology, Project Completion Report

17. A set of milestones around the key tasks of orientation, elections, planning, and management were developed to monitor the progress of FWUC training implementation (Table A4.4). Training was completed at all subprojects by 2010 (except for Preks of Kandal, which was completed in 2011), although the elapsed time varied substantially—e.g., in Ponley, work started in 2007, was paused while resettlement issues were reconsidered, and then continued; in Canal No. 1 and Don Aov, the whole program was completed within 1 year.

35 Appendix 4

18. In addition to the training provided by the IET, a technical assistance (TA) team drafted various manuals and documents to guide FWUC operations and subproject management: (i) an accounting manual (e.g., to track revenues and expenditures, manage and control the collection of irrigation service fees, prepare budgets and plans, distribute accounts to members); (ii) O&M manuals (procedures for planning and budgeting, irrigation service plan preparation, and processes to be followed when applying for O&M assistance from MOWRAM); and (iii) draft agreements between FWUCs and PDWRAMs to implement PIMD according to the roles and responsibilities outlined earlier (section C).9 F. Issues 19. Several issues arose regarding the establishment and operation of FWUCs:10

(i) By the end of project implementation, the FWUCs still showed weaknesses in institutional and operational terms. On average, they had been able to use the rehabilitated systems for only 1 year before project closure—equating to 1 wet and 1 dry season—which gave them limited experience. The extent of feasibility and detailed design required, and the procurement delays collectively reduced the time for actual subproject implementation and for learning-by-doing.

(ii) By the end of project implementation, only one subproject had gone through the process of applying for the O&M subsidy (at 80% of the costs) 11 from MOWRAM.12 This was Punley in mid-2011, and the total amount requested was about $7,300. As regards a wider dimension of IWRM, five apex committees were functioning—having begun to form in 2010—but these were generally recognized to be nascent institutions at the time of project completion.

(iii) Systemic tensions became evident throughout the project life between subprojects’ physical rehabilitation on the one hand, and community participation and FWUC formation on the other. Individuals and communities were sometimes reluctant (or uncomprehending) about agreeing to participate in revised O&M arrangements and wider IMT until hydrological boundaries and detailed subproject design (including block boundaries and thus WUG membership composition) were complete. Effectively, FWUC formation and registration could only take place at a later stage than the project design had anticipated. Although informal subcommittees were used to overcome this tension and drive things forward while physical design was still ongoing, they never involved the entire community. Had the construction subcommittees been given a formal role in signing off on eventual designs, the process might have worked better.

9 And specifically in relation to relevant legislation, i.e., Circular No. 1 of the Kingdom of Cambodia, dated 11

January 1999, on implementation of policy for sustainability of irrigation systems; Prakas 306 from MOWRAM, dated 20 July 2000, on adoption of basic instruments for establishment and development of FWUCs; Prakas 151 from MOWRAM, dated 30 July 2003, on adoption of seven basic documents for PIMD in the creation and development of FWUCs across the country.

10 Some issues also arose regarding gender (Appendix 11).

11 According to the agreements for irrigation management signed between FWUCs and PDWRAMs, FWUCs may

apply for subsidies for O&M from MOWRAM. These are triggered by collection of 75% of the ISF due from FWUC and WUG members factor in actually charged ISF rate, expected yield rates (and increases from pre-project production), and a declining proportion of total O&M costs over a 5-year period (so that communities take responsibility for the structures but in a gradual and sustainable way).

12 This was accompanied by a PDWRAM request for O&M assistance for the main system assets.

Appendix 4 36

(iv) Issues with the IETs also arose. Having one per subproject meant that project consultants and management had to deal with 12 separate entities, whereas three or four contract packages might have been more manageable. Also, the rationale for using service providers to overcome the limited capacity of MOWRAM in the first place was somewhat undermined by employing many MOWRAM and PDWRAM staff as IET members.

(v) The typhoon and rainfall events of 2009 and 2010 moved the level of importance and awareness of climate change up the agenda. The project responded well to rehabilitation and redesign needs, but this is obviously a factor that needs to be built into any further similar interventions in support of FWUCs.

20. In light of this, in mid-2011, Agence Française de Développement (AFD) funded a review of the FWUC experience to take stock of their institutional status and to prepare for the post-project situation.13 While the FWUCs’ capacity varied considerably at that time, the review raised the following points about the project as a whole:

(i) At the end of project implementation, the FWUCs had been established, but they were not really able to perform their intended tasks and still needed both practical advice and on-the-job training to become properly functional (including to promote proper communication between members);

(ii) Although initial budget plans had been made with the help of the IETs, FWUC members themselves did not always have a deep understanding of them, and thus were unable to communicate properly to members about the calculation and collection of ISFs.

(iii) Likewise, while some training on basic accounting and financial management had been provided, practical application of transparent procedures was very limited. Members had not fully started keeping invoices, records of expenditures, and other documentation, so further development of financial management tools was required.

(iv) Technical aspects of water management were generally well understood, but in some cases the subproject design had not allowed distribution to all blocks, and some further construction of (for example) tertiary canals was needed 14 —discussing the issues and raising sufficient fees to pay for works not carried out by members was proving difficult.

21. Two key recommendations of the review were: (i) to provide some financial resources to FWUCs in the first instance to cover the first year of operating costs—such “seed money” would “prime the pump” and motivate committee members in the first instance; and (ii) to establish a flexible and reactive technical support mechanism within which FWUCs would have a credit line of technical support to draw down according to their needs (in other words, a demand-driven source for on-the-job training). Participatory monitoring of FWUC capacity and operational performance across a range of standard criteria, and further technical support from service providers—e.g., from the Irrigation Service Centre in Kampong Thom, PDWRAMs, and local authorities—was also seen as highly desirable.

13

2011. AFD. Mainstreaming of actions in support to FWUCs under NWISP (Cambodia); Statement, lessons learnt and follow-up support. Jean-Marie Brun

14 Serious water management issues of various types were reported at five subprojects at this time. Most

subprojects had started constructing tertiary canals, but none had completed them.

37 Appendix 4

G. Status of Farmers’ Water User Communities in 2013 22. In late 2013, the status of the FWUCs was still rather similar to the situation observed in 2011, with considerable variations across subprojects but underscored by general systemic weakness. Table A4.3 summarizes their overall situation regarding membership composition and ISF collection in broad quantitative terms. Although women typically figure as accountants on the management committees (i.e., one out of four positions), they are vastly under-represented in general membership terms and also tend to resign from this function. ISF collection is highly variable due to increased production in the previous season, and has undoubtedly been affected by the impacts of floods during 2011–2013. The ongoing support from the Water Resources Management Sector Development Project15 in terms of both seed money for operational expenses and for on-the-job training (as recommended by the 2011 AFD review) have undoubtedly kept the institutions and the chance of sustainable IMT alive.

Table A4.3: Membership Composition and Fee Collection, 2013

Subproject

FWUC Status

ISF Collection

Kouch Nop 35 members (3 women, including accountant and 2 block heads)

KRs 200,000 collected in 2012 (about 10% of planned amount)

Tram Mneash 66 members (woman accountant) KRs 1.1 million in 2011 (50% of planned), nothing thereafter (due to system damage in year 2)

Krouch Saeuch 43 members (woman accountant) KRs 12 million in 2011 (90% of planned), Rs 5.2 million in 2012

Anlong Svay 27 members (no women) No ISF collected

Chork 68 members (6 women); committee has 2 women)

KRs 1.7 million (50% of planned) in 2012 (after 2011 season), nothing in 2013

Canal no. 1 30 members (1 woman, not on committee) KRs 2.5 million in 2011, 1.7 million in 2012 (60% and 20% of planned, respectively)

Don Aov 49 members (2 women; 1 on committee, but resigned). FWUC formed relatively late, little training, and reported to be “struggling”

KRs 8 million in 2011, Rs 4.6 million in 2012 – about 25% of planned target)

Po Pi Deum 40 members, 2 women (including accountant) KRs 1.6 million in 2011, Rs 1.6 million in 2012

Ponley 34 members, 3 women (including accountant – since resigned)

KRs 2011 1.7 million, 2012 2.7 million, 2013 3.1 million)

Ta Saom 12 members, 1 woman accountant No ISF collected Kork Thnong 10 members (“now 5”), of which 2 are women KRs 900,000 in 2011, nothing in 2012

FWUC=Farmer Water User Community; ISF=Irrigation Service Fee; KR=Khmer Riel. Source: Ministry of Water Recourses and Meteorology, Project Completion Report

15

ADB. 2010. Report and Recommendation of the President to the Board of Directors on Proposed Loans, Grant, Technical Assistance Grant, and Administration of Loan and Technical Assistance Grants to the Kingdom of Cambodia for the Water Resources Management Sector Development Program. Manila.

Ap

pe

ndix

4 38

Table A4.4: FWUC Training Program: Milestones and Completion Dates By Subproject

FWUC=Farmer Water User Community; MS=Milestone. Source: Ministry of Water Recourses and Meteorology, Project Completion Report

Milestone

Number

Task

No.

Subject Area/ FWUC

Milestones

Task %

WeighingPunley Po Pi Deum Chork Kouch Noup Kork Thnong Ta Saom Anlong Svay

Krouch

Saeuch

Tram

MneashCanal No. 1 Don Aov

Preks

Kandal

Actual %

Achieved

1 PMD, MT, and FWUC Policy 3Completed,

Jan 07

Completed,

Oct 08

Completed,

Jan 09

Completed,

Dec 08

Completed,

Mar 09

Completed,

Mar 09

Completed,

Dec 08

Completed,

Dec 08

Completed,

Dec 08

Completed,

Jun 09

Completed,

Jun 09

Completed,

Jun 08100%

2

Community orientation,

Organizing Gender/ Vulnerable

Mainstreaming

3Completed,

May 07

Completed,

Oct 08

Completed,

Jan 09

Completed,

Jan 09

Completed,

Apr 09

Completed,

Apr 09

Completed,

Dec 08

Completed,

Dec 08

Completed,

Dec 08

Completed,

Jun 09

Completed,

Jun 09

Completed,

Jun 08100%

3FWUC Formation and

Development3

Completed,

Sep 07

Completed,

Oct 08

Completed,

Jan 09

Completed,

Jan 09

Completed,

Jan 09

Completed,

May 09

Completed,

Dec 08

Completed,

Jan 09

Completed,

Jan 09

Completed,

Jun 09

Completed,

Jun 09

Completed,

Jun 08100%

4 Statue Subcommittee 5Completed,

Mar 08

Completed,

Oct 08

Completed,

Jan 09

Completed,

Jan 09

Completed,

May 09

Completed,

May 09

Completed,

Feb 09

Completed,

Mar 09

Completed,

Jan 09

Completed,

Aug 09

Completed,

Aug 09

Completed,

Sep 08100%

5 Construction Subcommittee 5Completed,

Sep 07

Completed,

Oct 08

Completed,

Jan 09

Completed,

Jan 09

Completed,

May 09

Completed,

May 09

Completed,

Dec 08

Completed,

Dec 08

Completed,

Dec 08

Completed,

Jun 09

Completed,

Jun 09

Completed,

Mar 08100%

MS 9 19 19 19 19 19 19 19 19 19 19 19 19 19 100%

6FWUC Elections, Committee

Elections12

Completed,

Jan 08

Completed,

Sep 09

Completed,

Aug 09

Completed,

Aug 09

Completed,

Oct 09

Completed,

Oct 09

Completed,

Oct 09

Completed,

Aug 09

Completed,

Aug 09

Completed,

Jan 10

Completed,

Dec 09

Completed,

Jun 08100%

7Basic Administrative

Management for FWUC10

Completed,

Sep 09

Completed,

Oct 09

Completed,

Sep 09

Completed,

Sep 09

Completed,

Nov 09

Completed,

Nov 09

Completed,

Oct 09

Completed,

Oct 09

Completed,

Sep 09

Completed,

Jan 10

Completed,

Dec 09

Completed,

Aug 09100%

MS 10 22 22 22 22 22 22 22 22 22 22 22 22 22 100%

8

Irrigation Service Plan, Water

Requirements, Supply and

Distribution

11Completed,

Mar 10

Completed,

Feb 10

Completed,

Jul 10

Completed,

Jul 10

Completed,

Dec 10

Completed,

Dec 10

Completed,

Jul 10

Completed,

Jul 10

Completed,

Jul 10

Completed,

Jun 10

Completed,

May 10

Completed,

Jun 10100%

9Community Development

Activity Planning3

Completed,

Sep 09

Completed,

Jun 10

Completed,

Jul 10

Completed,

Jul 10

Completed,

Jul 10

Completed,

Jul 10

Completed,

Jul 10

Completed,

Jul 10

Completed,

Jul 10

Completed,

Jul 10

Completed,

Jun 10

Completed,

Jun 08100%

10Water Distribution and Scheme

Management11

Completed,

Sep 09

Completed,

May 10

Completed,

Jul 10

Completed,

Jul 10

Completed,

Dec 10

Completed,

Dec 10

Completed,

Jul 10

Completed,

Jul 10

Completed,

Jul 10

Completed,

Jul 10

Completed,

Jul 10

Completed,

May 08100%

11 FWUC Accounting 11Completed,

Mar 10

Completed,

Nov 09

Completed,

Nov 09

Completed,

Nov 09

Completed,

Dec 09

Completed,

Dec 09

Completed,

Nov 09

Completed,

Nov 09

Completed,

Nov 09

Completed,

Nov 10

Completed,

Jan 10

Completed,

Oct 09100%

12 O&M 11Completed,

Mar 10

Completed,

Apr 10

Completed,

Nov 10

Completed,

Dec 10

Completed,

Dec 10

Completed,

Dec 10

Completed,

Sep 10

Completed,

Sep 10

Completed,

Sep 10

Completed,

Nov 10

Completed,

Oct 10

Completed,

Jun 10100%

MS 11 47 47 47 47 47 47 47 47 47 47 47 47 47 100%

13 Conflict Resolution 2Completed,

Jan 10

Completed,

Feb 10

Completed,

Jan 10

Completed,

Jan 10

Completed,

Mar 10

Completed,

Mar 10

Completed,

Dec 09

Completed,

Dec 09

Completed,

Jan 10

Completed,

Apr 10

Completed,

Feb 10

Completed,

Jan 11100%

14 Financial Management 3Completed,

Jan 10

Completed,

Feb 10

Completed,

Jan 10

Completed,

Jan 10

Completed,

Mar 10

Completed,

Mar 10

Completed,

Dec 09

Completed,

Dec 09

Completed,

Jan 10

Completed,

Apr 10

Completed,

Feb 10

Completed,

Jan 1197.2%

15 Participatory Impact Evaluaiton 2Completed,

Jan 10

Completed,

Jun 10

Completed,

Jan 10

Completed,

Jan 10

Completed,

May 10

Completed,

May 10

Completed,

Feb 10

Completed,

Feb 10

Completed,

Jan 10

Completed,

Apr 10

Completed,

Mar 10

Completed,

Jan 1196.8%

MS 12 7 7 7 7 7 7 7 7 7 7 7 7 7 97.6%

Stu

dy T

ou

r

16 FWUC Exchange Visits 5Completed,

Feb 10

Completed,

Apr 10

Completed,

Jan 10

Completed,

Jan 10

Completed,

Mar 10

Completed,

Mar 10

Completed,

Jan 10

Completed,

Jan 10

Completed,

Jan 10

Completed,

Mar 10

Completed,

Mar 10

Completed,

Jan 1198.3%

MS 13 5 5 5 5 5 5 5 5 5 5 5 5 5 98.3%

100 100 100 100 100 100 100 100 100 100 100 100 100 99.8%TOTAL OVERALL

Ori

en

tatio

nE

lectio

ns

Pla

nn

ing

Ma

inte

na

nce

39 Appendix 5

DEVELOPING IRRIGATED AGRICULTURE

A. Introduction 1. The overall objective of the irrigated agriculture component was to maximize project benefits by ensuring delivery of support to improve rice production and crop diversification through (i) assistance to and training of subproject-specific agricultural extension services and (ii) fielding of a cadre of irrigation extension workers to facilitate the delivery of these support services within the subproject areas, and to maintain direct and regular field contact with beneficiaries to guide them. The aim was thus to help ensure that maximum benefit would accrue from the project’s investment in irrigation infrastructure development. 2. The original project design included strengthening of the provincial departments of agriculture (PDAs) so that provincial (and particularly district) staff could implement a program of farmer training and irrigated agriculture extension in the subproject areas. Detailed terms of reference called for the development of field research and extension programs, and provision of training by setting up farmer field schools and on-farm demonstrations and implementation of seed production programs. 3. Largely because of a thrust toward decentralization and deconcentration, and in an attempt to make the provision of public services more transparent, the approach adopted at implementation aimed to (i) increase the capacity of the provincial departments of the Ministry of Water Resources and Meteorology (PDWRAMs), which were already more involved in project activities than the PDAs; and (ii) contract PDA services directly through agriculture service agreements (ASAs), whereby funding for field operations would be provided and reliance on national budget allocations alone would be obviated.1 The ASAs were agreed and in place by early 2009. 4. Beside the ASAs, the irrigation extension team (IET) at each subproject included an agricultural extension specialist. Their function was to assist provincial and district staff in establishing and supporting farmers’ water user communities (FWUCs), and to provide a link between the beneficiaries and the designers in a participatory design approach. The IET agriculture specialist was therefore a main link, and coordinator, between PDAs, PDWRAMs, and FWUCs. 5. The project’s training initiative largely followed the Farmers’ Field School (FFS) approach. FFSs are groups of 30 farmers each who receive weekly training on topics related to the progress in the three demonstration plots associated with each FFS. PDA master trainers are in charge of the FFSs (including the demonstration plots) and the various subjects, partly depending on the season. The FFSs cover one wet and one dry season per year. “Field days” are meetings of FFS members at the end of a season (wet or dry) to present results and share experiences with other members and participants, and where certificates are handed over to FFS members.

1 The original contracts with the PDAs totaled $892,540, comprising provincial allocations of $320,080.00 (Pursat),

$243,580.00 (Battambang), $167,080.00 (Banteay Meanchey), and $161,800.00 (Siem Reap). Contracts consisted of two parts. Part 1 provided allowances to support the PDA officer, master trainers, district agriculture office teams, and PDA office operation. Part 2 provided funding for agricultural training activities in the subprojects.

Appendix 5 40

B. Component Implementation 6. Implementation of the component was in practice largely determined by (i) standard base-cost estimates per demonstration and other expenses—allowances, staff travel, fuel, materials—that could be accommodated within the budget; and (ii) seasonality—i.e., 15 weeks in the dry season, 25 weeks in the (main) wet season. 7. The delays in the construction of physical irrigation infrastructure at the subprojects had inevitable knock-on consequences for the irrigated agriculture component. Originally (in 2008) proposed for implementation by a technical assistance (TA) team as a two-phase program, the operational stage of the component did not occur until 2009–2010, under the ASAs. Because some subprojects’ physical works were not completed until 2010 or 2011, some irrigated agriculture demonstrations were in effect being done on existing schemes, i.e., before construction and/or rehabilitation. Some vegetable demonstrations were located in home gardens and involved manual pumping of water. 8. The implementation schedule began with the farmer selection process, which followed basic criteria—e.g., FWUC membership, block location and/or affected person status, gender. Due to delays in subproject construction and finalization of ASAs, and seasonality, about two-thirds of demonstrations in the first year (2009) were devoted to animal husbandry (pigs and chickens). 9. A summary of the farmers trained by province and subproject, gender, and topic for all 3 years is in Table A5.3. Nearly 10,000 persons were trained (over 53% of them women) in 1,320 demonstrations. The largest number of demonstrations (42%) were for rice (wet and dry seasons). Innovative techniques for use of organic fertilizers and pesticides were part of the dry season work. In 2009, only a quarter of the wet season demonstrations concerned rice, but in the next wet season the focus was 100% on rice (396 demonstrations). 10. Rice production before the project largely involved traditional techniques such as broadcasting, use of local seed, minimal fertilizer application, and informal transfer of technology. Traditional rice seeds were of the long-term variety (180 days), or new varieties like Phka Rodual, Pkha Romdeng, Phka Romeat (medium term, 145–150 days). The project, through the PDAs, largely adopted the system of rice intensification (SRI). This involves practices that differ from traditional techniques: (i) early transplanting of younger seedlings (e.g., 8–10 days old) with wider and equal spacing, preferably a single plant per hill; (ii) draining of the field 2–3 times a season to allow for soil aeration in the upper root zone; (iii) adequate weed control (including the use of mechanical soil scrappers); and (iv) greater use of organic fertilizers (as opposed to chemical fertilizers). Other specific interventions were: (i) use of a short-growing (85 days) Vietnamese variety at the Chork subproject; (ii) seed production; (iii) improved crop rotation (i.e., green manure early in the wet season); and (iv) better post-harvest technologies. As a result of the project interventions and once the irrigation systems had been completed, farmers gained knowledge and skills for growing dry season rice and/or recession rice in command areas where this had not been done previously (i.e., Kouch Noup, Tram Mneash, and Ponley). Table A5.1 summarizes traditional rice cultivation practices versus recommended ones.

41 Appendix 5

Table A5.1: Traditional Farmer Practices vs. Recommended Practices Summary

Steps in the Cultivation Traditional Practices Recommendations

Varieties Choice among the 10 varieties promoted by MAFF. Seeds Keep seed from one year to

another. No particular selection of grains.

Renew seeds periodically. Take care of varietal purity. Check the germination rate. Select the grains used as seeds (use of less seeds if follow recommendations regarding transplanting).

Soil preparation Ploughing and harrowing Two to three ploughings, harrowing, better care for land levelling. Making bunds around the field.

Transplanting Transplanting of several seedlings per hill, age of seedling variable, but often around 45 days.

Transplanting 1 or 2 seedling(s) per hill, young seedlings (around 15–20 days).

Water management Transplant in flooded parcel. Transplant in muddy soil or low water level (2–3 cm) then increase the water level 2–3 weeks after transplanting.

Fertilization Mainly urea. Timing of fertilization not always well controlled.

More diversified fertilizers: organic fertilization (manure, compost) in priority. DAP as basal, urea afterward.

Pest management Treatment whenever the farmer sees insects. Not always good choice of pesticides and no precaution.

Integrated pest management principle: identify insects that are harmful before the treatment.

Harvesting Harvesting when 80%–85% of the grains are ripe. Source: Ministry of Water Recourses and Meteorology, Project Completion Report

C. Crop and Livestock Productivity Improvements at Demonstrations 11. Before the project, rice yields across all subprojects averaged around 2 tons per hectare (ha). SRI and integrated pest management (IPM) techniques were able to as much as double this. Table A5.2 summarizes rice yields from the demonstration plots across seasons and subprojects. The SRI approach in particular surprised many farmers at the demonstrations—they did not expect single-stem transplants to be so productive and had never seen multi-tilling—and the yield increases were evident. At Tram Mneash, in wet season 2009, traditional methods yielded 2.5 tons/ha whereas the new technology yielded 4.75 tons/ha.2

2 Rice yields were impacted during project implementation by extreme flood events in two successive years. In wet

season 2009 much of the project area suffered from flooding as a result of Typhoon Ketsana; many farmers lost crops and demonstration performance was affected on most subprojects. Rice yields for wet season 2010 would have been significantly higher at Po Pi Deum, Don Aov, and Canal 1 subprojects had it not been for the severe flooding in October 2010 during a critical growth period (which affected panicle formation and lowered yields).

Appendix 5 42

Table A5.2: Summary of Rice Yield Demonstration Results by Subproject, Season, and Year

Source: Ministry of Water Recourses and Meteorology, Project Completion Report

12. Demonstrations of catfish raising in both ponds and in plastic sheets were considered successful. Pig demonstrations were also successfully implemented, although problems with imports from Thailand and the prevalence of blue ear disease in Viet Nam imports caused depressed returns. Gross margins compiled by the TA team indicated enterprise profitability and improved returns when compared with traditional agricultural (non-livestock) output. 13. The impact of training on farmers’ capacity (based on training assessments in 2010 and 2011) were highly positive—technological capacity rose by over 40% on average. D. Issues 14. Some issues with the irrigated agriculture component arose in the course of project implementation, such as:

(i) Pressure in the first year to get started with any agricultural activities at all despite the overall delays to the project tended to result in a higher proportion of fisheries and livestock demonstrations than was consistent with the investment objectives to promote irrigated agriculture (i.e., wet and dry season rice, and dry season vegetables), pending finalization of the subprojects’ physical construction. Other issues concerned attendance by farmers who were not FWUC members, and some demonstration sites being at considerable distances from local villages.

(ii) The ASAs with the PDAs were administratively cumbersome, notably with regard to the flow of funds and advance liquidation (often taking 2–3 months), which tended to delay implementation even further.

(iii) Most significantly, the actual extent of yield and production improvements on the subproject sites was not clear at the end of the investment period. While the various demonstrations’ yields, budgets, and gross margin analyses had been

Traditional

Practice

(T/ha)

Improved

Technology

(IPM+SRI)

(T/ha)

Traditional

Practice

(T/ha)

Improved

Technology

(IPM+SRI)

(T/ha)

Traditional

Practice

(T/ha)

Improved

Technology

(IPM+SRI)

(T/ha)

Traditional

Practice

(T/ha)

Improved

Technology

(IPM+SRI)

(T/ha)

Siem Reap 1.94 3.42 2.37 3.86

Ta Saom 1.50 Bakan 1.92 1.92 3.68 n/a n/a 2.31 4.1 n/a n/a

Kork Thnoung 1.50 Prateah Lang 1.8 1.96 3.16 n/a n/a 2.42 3.61 n/a n/a

Banteay Meanchey 2.89 4.15 3.29 4.57 2.30 4.19 2.70 4.22

Punley 1.50 Bakan 2.00 2.53 3.53 3.07 4.50 2.29 4.37 2.40 4.08

Po Pi Deum 2.00 Thoulsamrong 2.25 3.25 4.76 3.50 4.64 2.30 4.00 2.99 4.36

Battambang 2.95 4.45 2.62 4.32 2.64 4.75 2.98 5.15

Canal No. 1 1.37 Thoulsamrong 2.80 - 4.15 n/a n/a 2.50 4.57 n/a n/a

Don Aov 1.50 Thoulsamrong 1.80 3.35 5.20 n/a n/a 2.79 4.77 n/a n/a

Chork 1.50 Thoulsamrong 1.80 2.55 4.00 2.62 4.32 2.64 4.90 2.98 5.15

Pursat 2.30 3.84 2.56 5.08 2.74 4.26

Tram Mneash Prey Khmer 1.35 2.20 3.62 n/a n/a 2.70 4.86 2.43 3.98

Kuck Noup 1.25 Prey Khmer 1.20 2.15 3.69 n/a n/a 2.81 4.28 2.54 3.86

Krouch Saech 2.68 Bakan 2.38 2.50 4.75 n/a n/a 2.60 5.62 2.45 3.76

Anlong Svay 1.50 Bakan 2.10 2.35 3.29 n/a n/a 2.12 5.55 3.54 5.42

Average

Land

Holding

Area (ha)

Province/

Subproject

Wet Season 2009 Dry Season 2010 Wet Season 2010 Dry Season 2011

Demonstration Rice Yield Data

Rice Yield

Before Project

(T/ha)

Soil Classification

43 Appendix 5

showing considerable production and profitability improvements, the evidence on which to assess actual project impacts was very limited (Appendix 6).

E. Irrigated Agriculture In 2013 15. Two years after project completion, fieldwork at all locations suggests the following conclusions about the impact of the irrigated agriculture component:

(i) Adoption of recommended practices in rice cultivation, such as SRI, has been notably limited and partial, despite the apparent success of demonstrations. Broadcasting of seed is still far more widespread than transplanting, and reported fertilizer application is typically below recommended rates (and, where tried, it has often been reduced because of perceived damage to the soil).

(ii) There is almost no cultivation of non-rice crops in the dry season and, where tried, it has competed with water needed for livestock.

(iii) Both early-season and dry season rice cultivation has increased in aggregate, although the effect of the former is far greater.3

(iv) The livestock and fisheries demonstrations seem to have inspired few adopters.

(v) The most dramatic changes to agriculture practices (partly linked to the overall aggregate demand for rice) were mechanization and replacement of labor for land preparation, planting, harvesting, and transportation.

3 The estimated extent and economic impact of these changes are discussed in Appendix 8.

Ap

pe

ndix

5 44

Table A5.3: Summary of Demonstrations By Province and Subproject, Gender of Participant, Topic, and Date Source: Ministry of Water Recourses and Meteorology, Project Completion Report

Rice Pig Chicken Vegetable Mushroom Fish Fruit Tree

Siem Reap

Ta Saom 30 900 586 120 45 18 21 26 7 3 0 120 5-Jan-09 30-Mar-11

Kork Thnoung 30 900 691 120 45 18 21 26 7 3 0 120 5-Jan-09 31-Mar-11

Subtotal 60 1800 1277 240 90 36 42 52 14 6 0 240

Banteay Meanchey 0

Po Pi Deum 30 900 446 120 57 21 17 21 0 4 0 120 5-Jan-09 25-Mar-11

Punley 30 900 478 120 57 21 21 21 0 0 0 120 5-Jan-09 3-Nov-11

Subtotal 60 1800 924 240 114 42 38 42 0 4 0 240

Pursat 0

Kuck Noup 30 900 512 120 51 17 21 17 0 13 1 120 15-Jun-09 26-Apr-11

Tram Mneash 30 900 530 120 51 18 18 15 0 18 0 120 15-Jun-09 19-Mar-11

Krouch Saech 30 900 433 120 51 21 21 15 0 12 0 120 15-Jun-09 28-Apr-11

Anlong Svay 30 900 415 120 47 20 21 21 0 11 0 120 15-Jun-09 31-Mar-11

Subtotal 120 3600 1890 480 200 76 81 68 0 54 1 480

Battambang 0

Chork 30 900 460 120 57 18 21 10 9 5 0 120 7-Jun-09 29-Apr-11

Canal No. 1 30 900 360 120 45 21 21 16 0 17 0 120 7-Jun-09 4-Dec-11

Don Aov 30 900 401 120 45 26 21 24 1 3 0 120 7-Jun-09 24-Mar-11

Subtotal 90 2700 1221 360 147 65 63 50 10 25 0 360

TOTAL 330 9900 5312 1320 551 219 224 212 24 89 1 1320

Total Start Date Finish

Date

Number

of Demos

Province/ Subproject Farmer

Groups

Total All

Trainees

Total

Trainees

(female)

Topics

45 Appendix 6

SUMMARY OF RESETTLEMENT PRACTICE AND ISSUES A. Introduction 1. The design of the project included a resettlement framework,1 which set out the policy, principles, procedures, and implementation mechanisms applicable to all resettlement taking place in subprojects. Impacts were expected on agricultural land in the corridors of main and secondary, or branch, canals and drains and in expanded reservoirs. Land losses in tertiary irrigated areas would generally be absorbed in voluntary land adjustment. There was to be no substantial relocation of communities, but only scattered and localized replacement and relocation of houses built on canal embankments or in reservoirs by relatively poor households that had migrated to these areas in recent years. 2. Resettlement policy and principles were based on (i) the development of a Cambodian national policy and regulatory structure; (ii) adherence to the Policy on Involuntary Resettlement (1995) and Handbook on Resettlement: A Guide to Good Practice (1998) of the Asian Development Bank (ADB); and (iii) enactment of the Land Law of 30 August 2001 and its subdecrees, which established rights of land ownership and of entitlement.2 3. A resettlement plan was to be prepared for each selected subproject. In preparing the plan, the executing agency—the Ministry of Water Resources and Meteorology (MOWRAM)—would work in consultation with affected persons, commune chiefs, and district and provincial resettlement officials, and would take into consideration the land grants that were to be available to the poor in the project districts under the Social Land Concession Program.3 The resettlement plans would follow ADB requirements and also draft a national policy to establish the following principles:

(i) Adequate steps to be taken to minimize the impact of project works on affected persons and their land and structures, and to mitigate any adverse effects (i.e., by choosing canals alignment and widths that minimize impact).

(ii) Affected persons and their representatives to participate fully in project design and in planning and implementing resettlement. They are to be consulted both over subproject works and timing and over resettlement proposals, and participate fully in subproject design and works, and in planning and implementing resettlement. For this purpose an information program would be conducted before project implementation; an information leaflet had already been prepared.

(iii) Specific provisions would be made in the resettlement plans to safeguard or improve the welfare of affected vulnerable groups, including women heads of households, the aged, the disabled, the very poor, landless, and minority groups.

(iv) The economic and social future of people would not be unfavorably affected, and they would not face a material reduction in income and living standards, or unnecessary social and cultural dislocation as a result of the project.

1 The full resettlement framework was in Supplementary Appendix G of the report and recommendation of the

President. 2 A Sub-Decree on Social Concessions gave the legal basis for allocating communal land for the purpose of

reducing landlessness and poverty, and was to apply to the replacement of land lost in the context of involuntary resettlement.

3 The Government’s program, but funded by the World Bank.

Appendix 6 46

4. Before any construction works, compensation was to be made in the presence of affected persons’ representatives and an external monitor. The entitlements4 were broadly as follows:

(i) Compensation at replacement cost for houses and other affected structures5 without deduction for depreciation or salvageable materials.

(ii) For agricultural land, compensation in the form of land for land of equal productive capacity, or compensation in cash permitting purchase by affected persons6 of land of equal quality and productivity to that lost.

(iii) Replacement of premises or residential land with land of equal size acceptable to the affected person.

(iv) Replacement of commercial land with land of equal size and commercial potential acceptable to the affected person(s).

5. In addition to compensation for land or any structures, each affected household would be entitled to a one-off disruption allowance to be paid in cash where the household is relocating away from the existing site.7 6. The institutional framework for preparing and implementing resettlement plans at the national level was that of an interministerial resettlement committee (IRC)—a committee of the Council of Ministers administered by the Ministry of Economy and Finance (MEF), MOWRAM, other implementing ministries, the municipality of Phnom Penh, and governors of the provinces involved in the project. A resettlement unit in MOWRAM would be responsible for conducting surveys and consultations, and for preparing resettlement plans. 8 The unit would conduct internal monitoring and evaluation and report three-monthly to the IRC and to ADB. An external monitor appointed by MEF would work under the supervision of the resettlement unit to evaluate the implementation and effectiveness of the resettlement plans and to assist affected persons in any grievance or dispute over resettlement and compensation.9 B. Resettlement Issues during Project Implementation 7. The first subproject for which resettlement planning was undertaken was Ponley. This work (presented to the IRC by MOWRAM’s resettlement unit in early 2008) initially indicated that resettlement costs would be dramatically different from appraisal estimates, and MEF advised MOWRAM to drop the subproject. However, because Ponley was considered likely to

4 A grievance process was established to allow affected persons not satisfied that the compensation, replacement

of lost assets, livelihoods, and allowances due to them had been fairly provided to seek redress and to make a formal complaint. A provincial resettlement grievance committee was established and similar district, commune, and village committees were created. Affected person could be assisted by any nongovernment organization and by the external monitor, and could take court action without cost to them.

5 Land, houses, and other structures and fixed assets such as trees, wells, and graves were to be compensated on

the basis of a survey agreed with the affected persons and at prices based on the market and on the need for replacement of livelihoods at least to the same level as before project actions.

6 A rate in riels, equivalent to $40 per household, was determined as the disruption allowance.

7 Affected persons losing 10% or more of their total productive agricultural land (i.e., severely affected persons)

were entitled to an agriculture training program. This program was included in the irrigated agriculture component. 8 For this purpose, the resettlement unit combined forces with provincial departments, headed by the Provincial

Department of Water Resources and Meteorology (PDWRAM), as a provincial resettlement working group answerable to a provincial resettlement subcommittee.

9 SBK Research and Development was awarded the contract to monitor resettlement activities of the project after a

competitive bidding process. The contract was signed with MEF on 31 July 2007. The monitoring service began on 8 August 2008 for a period of 2.5 years.

47 Appendix 6

be representative of many potential stand-alone schemes, and after a meeting between ADB, MEF, and MOWRAM, and a resettlement review mission by an ADB staff consultant, the issue was reexamined. One recommendation was to take into consideration operation of the reservoir versus the opportunity for cropping. The overall approach adopted by the resettlement unit and the IRC for assessing or measuring the social impact of the subproject, especially in the reservoir area, and for updating the subproject resettlement plan—while remaining in compliance with the resettlement framework agreed for the overall project—was revisited and led to a substantial reduction of the originally estimated costs. The implementation consultant fielded a national expert (in July 2008) to help the resettlement unit undertake the necessary revisions. The work was done in close consultation with specialists in operation and maintenance, water management, and agriculture. 8. Due to the time and effort taken to complete the resettlement plan for Ponley, which was ongoing throughout 2008, it was obvious that preparation of the remaining 10 subproject resettlement plans would be a major, if not impossible, task if the overall implementation was to remain within the planned schedule—given the enormous amount of detailed field and desk work involved. To solve the issue, the project management office (PMO) convened a meeting with the IRC on 16 February 2009 and reached this agreement: (i) only two more resettlement plans would be prepared (one covering the five subprojects for which detailed design was ongoing, and one for the remaining 5 subprojects for which detailed design was just starting); (ii) the preliminary inventory of losses need not be prepared (as was done for Ponley); and (iii) detailed measurement surveys should not start until designers were confident that there would be no major changes to alignments (90% or above), but with some flexibility to allow for minor changes, e.g., involving small canals or drains. Follow-up meetings between PMO, resettlement unit, TA team, and service providers, led to the development of a detailed, time-bound implementation schedule, which was used to monitor progress thereafter. 9. Strong commitment to fast-tracking the work was demonstrated by the agreement to assign three IRC teams full-time until the end of the year. The first of the two resettlement plans was to be submitted around the end of September 2009, and the second around the end of December 2009. After an IRC-led field visit in June 2009 for a comprehensive review of the plans, changes were made to adapt to conditions and the sequence of subprojects was revised. The outcome was positive, and helped keep this major task on schedule; resettlement plans 2 and 3 were submitted in October and November of 2009. 10. Under the resettlement plan for Ponley, compensation payment was completed in June 2009 for a total of 409 affected persons. (Of these, 168 were severely affected and entitled to agricultural training). Under resettlement plan 2, compensation was completed in June 2010 for a total of 1,029 affected persons (Po Pi Deum: 105, Anlong Svay: 273, Krouch Saeuch: 251, Kouch Noup: 52, and Tram Mneash: 348). Of these, 111 were severely affected (Anlong Svay: 24, Krouch Saeuch: 23, Tram Mneuch: 59, and Kouch Noup: 5) and entitled to agricultural training. An additional impact on farming land and trees belonging to 31 households arose when irrigation in the Tram Mneash subproject was expanded by 1,200 meters. The IRC compensated the 31 households in August 2010. Under resettlement plan 3, compensation was completed in June 2010 for a total of 1,006 affected persons (Ta Soam: 87, Kork Thnong: 50, Don Aov: 282, Canal No. 1: 373, and Chork Reservoir: 214). Of these, 70 were severely affected (Ta Saom: 2, Kork Thnong: 2, Don Aov: 44, and Canal No. 1: 22) and entitled to agricultural training. The IRC developed a relocation site for 12 affected persons (landless) under the Canal No. 1 subproject. The relocation site is about 1 kilometer from the existing site, in Knay Romeas village, Knay Romeas commune, Bovil district, Battambang Province. The IRC

Appendix 6 48

developed 12 plots (plot size: 11.5m x 25m or 12m x 24m) with one common pumping well for the 12 landless people. 11. Overall, the original delays connected with the Ponley resettlement plan coincided with other aspects of detailed design (done by service providers) and procurement, and did not in themselves delay project implementation. The resolution of resettlement issues for the other subprojects (through the procedures described above) successfully avoided delays to the project that might otherwise have occurred.

49 Appendix 7

PROJECT MONITORING AND EVALUATION A. Introduction 1. The project was required to adopt a project performance management system (PPMS).1 The loan agreement states that during project implementation the PPMS shall comprise: (i) the project logical framework,2 (ii) regular project performance reports to track progress from the baseline situation against identified indicators; and (iii) a project completion report. These activities are to include participatory baseline physical and socioeconomic surveys conducted in each of the subprojects before the interventions, and with follow-up surveys to evaluate the changes and to measure benefits generated, particularly the impact on the poor.

2. Prior to the project, Cambodia’s Ministry of Water Resources and Meteorology, through its department for farmers’ water user communities (FWUCs), had been trying to collect similar information at selected sites and with time-bound, and there had been previous assistance by Agence Française de Développement (AFD) in developing a Cambodia information system on irrigation schemes (CISIS)—although the focus of the latter was largely on recording and inventory assets rather than monitoring and evaluating benefits. 3. Against the institutional and capacity background of MOWRAM and its provincial departments (PDWRAMs) at the time it was decided to establish a relatively simple reporting format based on Excel software and located in MOWRAM’s planning department, whereby project progress (physical and financial indicators, oriented around measuring outputs achievements) was mainly a centralized (i.e., MOWRAM) function, but project impact monitoring was more decentralized (i.e., the PDWRAMs had a greater role in recording, compiling, and reporting, as well as concerning themselves with outcome achievements). B. The Project Framework

4. The first point of reference for project Monitoring and Evaluation (M&E) was of course the hierarchy of objectives as defined in its logical framework or, to go by Asian Development Bank (ADB) practice at the time of design, the project framework. 5. The project framework was contained in the report and recommendation of the President (RRP) and some relevant points in this regard include the following:

(i) The “Goal” statement (i.e., to “Support Government’s effort to reduce poverty in selected northwest rural areas through enhanced agricultural production thereby alleviating food insecurity and improving farming household incomes“) conflates various levels of objectives and its achievement is consequently impossible (in its own terms) to measure. Although the ideas contained herein may well be related, “reducing poverty,” “enhancing agriculture production,” and “improving farming household incomes” are not the same thing—agriculture production may be enhanced and yet farms’ incomes still fall and poverty increases (e.g., if output prices go down); and food insecurity may be increased if crop composition or seasonality changes. A clearer delineation between what was in the project to possibly deliver (e.g., “more productive irrigated agriculture on planned

1 As specified in the loan agreement (Schedule 6, para. 9) and mentioned in the report and recommendation of the

President (para. 51). 2 The project framework is in Appendix 1 of this report.

Appendix 7 50

schemes”) and what was not (i.e., poverty reduction and income increases) would have been highly desirable.

(ii) The space for “Assumptions and risks” was left blank in this item ("Goal”)—had it been completed, the relationship between output prices and changes of incomes and poverty would have necessarily become more explicit.

(iii) The “Purpose” statement is meant to anchor the project design and state what the project will achieve if implemented as designed. In this case, the “Purpose” refers only to the physical aspect of irrigation provision and completely omits the institutional strengthening and irrigated agriculture components of the project.

(iv) The terminology used for the objectives in the RRP main text and the project framework is misleading. The project framework uses the “Goal” and “Purpose” terminology that applied within ADB at the time of design and before implementation (2003–2005), yet the RRP main text talks of “overall objective” and “immediate objective” (Loan and Project Summary, and para. 12). It is not clear if these are the same respective levels of achievements, which likely contributed to later confusion (e.g., in MOWRAM’s draft project completion report, annex IV on PPMS) about project impact monitoring, defined in that document as “which assesses whether the intended outcomes (goals or results) are contributing to government’s efforts to reduce povertyN”.

6. In sum, the project framework appears cursory and incomplete. Shortly after the project was designed, more robust guidance—in the form of Guidelines for Preparing a Design and Monitoring Framework3—became available, and internal review processes would have required that a more thorough approach be employed. C. Project Progress Monitoring 7. The project developed a system of monitoring financial and physical progress. Regarding physical progress, this was based on a weighted estimate 4 of activities to be undertaken and completed across the three project components. The measure of the extent of actual implementation progress was compared (in an “S-curve” format) against what was expected. On the basis of the assumptions made, this system provided a clear and consistent measure of implementation performance. 8. However, this system was started only after the midterm review of 2008, and revised targets for project outputs (and an anticipated extension to the project life of 1 year) were agreed on shortly thereafter. Thus, reported levels of implementation progress (which were generally very positive) refer not to original targets in the project design, but rather to the achievement levels agreed for the second period of project implementation. D. Project Impact Monitoring 9. Measuring the impact of the project on agriculture production (volumes, yields, crop composition, seasonal patterns) is obviously dependent on some understanding of pre-project or baseline conditions. Accordingly, baseline surveys were incorporated in the work of the

3 ADB. 2006. Guidelines for Preparing a Design and Monitoring Framework. Manila.

4 In the weighting system developed, component 1 was 5%, component 2 was 82.5%, and component 3 was 12.5%

of the overall project. These weights were subjective, but thought to be reasonable.

51 Appendix 7

service providers at each subproject,5 and surveys covering commune-level and household-level indicators of land, agriculture activity, extension, water quality, labor usage, FWUC membership and operations, credit, among other things, were conducted throughout 2009. Clearly, the definition of “baseline” after several years of project life is less than ideal, but became inevitable given the knock-on effect of delays to the subprojects’ construction. 10. The capacity of service providers to undertake these tasks with the required level of consistency across all subprojects and with sufficient technical rigor depended on considerable support from technical assistance (TA). 11. Since completion of the baseline surveys, the data has been transferred from its original Excel format to an Access database, linked to the CSIS. 12. There is 1 year of impact data available from surveys undertaken after completion of the physical works, and still within the implementation life of the project, i.e., in early 2011.6 This data suggested (i) considerable increases in average wet season rice yields across the subprojects—varying from a decrease of 38% at Po Pi Deum to a trebling at Kork Thnong, with an overall average increase of 61%;7 and (ii) increases in dry season cultivation and/or cropping at 9 of the12 subprojects. Work done in preparation of the project completion report in late 2013 included an informal investigation of project impacts nearly 2 years after completion, and the findings of this work are discussed in Appendix 8 (Financial And Economic Analysis). E. Revised Project Framework 13. As a result of the revised implementation schedule as agreed after the midterm review of 2008, and of focusing on a smaller irrigated area and other outputs than had been originally anticipated, it was necessary to revise the project framework’s indicators for subsequent realistic project monitoring. 14. ADB and MOWRAM agreed on a revision of the project framework (Outputs) in 2011; the revised framework’s indicators are shown in Appendix 1.

5 The exceptions were the two subprojects in Siem Reap for which detailed design was done by MOWRAM–

PDWRAM’s design and implementation team. PDWRAM staff did the baseline surveys for these subprojects under TA guidance.

6 2011. MOWRAM. Project Completion Report; Annex IV

7 This estimate is unweighted by size of the 12 subprojects.

Appendix 8 52

FINANCIAL AND ECONOMIC ANALYSIS A. Introduction 1. This appendix summarizes the major issues connected with the financial and economic analysis of the project. It begins with some commentary on the financial and economic analysis that was done at design, and describes the current status of financial returns and economic viability of the irrigation subprojects based on data collected in late 2013. B. Financial and Economic Analysis at Project Design 2. The financial and economic analysis of the project that was done at design was typical for a sector investment at that time (around 2002), in that no overall financial or economic internal rate of return (FIRR or EIRR) was calculated for the project as a whole, but returns to expected (or representative subproject) investments were described and estimated, and non-quantified benefits discussed. 3. Again, for an irrigation project, the quantified benefits were largely typical in that they were essentially additions to rice (and other crops’) production based on area and cropping pattern changes “with” and “without” the various planned investments (e.g., in institutional strengthening, physical infrastructure, irrigated agriculture). The analysis covered investment costs during the implementation period, and also the implications of changes to operation and maintenance (O&M) arrangements (and levels of financial charges to farmers in relation to capacity and willingness to pay these) thereafter. 3. The financial and economic analysis at design raises several methodological issues:

(i) The economic pricing is done using the world numeraire. (Asian Development Bank [ADB] guidance 1 is that, in general, the domestic numeraire should be used, as this allows direct comparison of financial and economic flows [costs and benefits], although for small, open economies, where a large proportion of economic activity is traded, the world numeraire can be applied). However, an explanatory reason given for the choice made 2 seems to suggest a misunderstanding of the numeraire (or price level) concept, rather linking the choice of numeraire to the point of likely final rice consumption.

(ii) This impression of basic misunderstanding is confirmed by the description of the US dollar as the “unit of account” being appropriate “because this is commonly used by people in Cambodia’s northwest” – whereas in fact the “unit of account” in project economic analysis is both the chosen price level (i.e., world or domestic) and the particular currency (i.e., foreign or national) chosen (e.g., US dollar, Cambodian riel, Thai baht).

(iii) The border parity pricing basis for rice (as well as fertilizers) is implicitly an import-substituting one. The economic analysis could have made this assumption

1 ADB. 1997. Guidelines for the Economic Analysis of Projects. Manila—“guidelines” hereafter.

2 ADB. 2003. Report And Recommendation Of The President To The Board Of Directors On A Proposed Loan To

The Kingdom Of Cambodia For The Northwest Irrigation Sector Project.– Appendix I, and Supplementary Appendix I. “Domestic price numeraire would be equally appropriate in the northwest of Cambodia as many of the local villages still suffer extended food shortages for up to 6 months a year. It is likely that incremental production will be consumed within the local or neighboring villages. Only in certain producing areas of Battambang will incremental production need to compete on the international market@”

53 Appendix 8

clearer by referring to a “supply-based” and/or “nonincremental” basis for benefit valuation, as the guidelines suggest.

(iv) Given the extensive analysis of rice marketing in Supplementary Appendix A, and its acknowledgment that some rice from northwest Cambodia may end up in export markets (Thailand, Viet Nam), it is perhaps surprising that the economic valuation does not explicitly mention this point (and perhaps consider some type of weighted average import substitution or export-based price).

(v) The attempt to describe the impact on poverty, while welcome, concentrates on the estimation of financial benefits (from crop sales and farm labor opportunities) accruing to the poor—and is not the formal poverty impact ratio (PIR) methodology calculation, as described in the guidelines (Appendix 26).

4. The project’s economic analysis could also have been somewhat fuller by considering (for example) whether the without-project situation involved any deterioration from the then current conditions (environmental, social, economic), and could also have applied some quantitative risk analysis to (for example) commodity input and output prices. The latter consideration was being formally recommended as standard ADB practice by 2002,3 and the former was contained in the 1997 guidelines. 5. Economic and financial analysis of the subprojects was (re-)undertaken in each of the feasibility studies (and some crop budget descriptive data was collected in the baseline studies)—the methodology essentially followed that of the report and recommendation of the President (including the methodological problems described above), although it was obviously more closely related to the particular subprojects eventually constructed. C. Financial and Economic Analysis After Project Completion 6. In late 2013, data was collected on agricultural practices (i.e., cropping patterns and crop yields) before and after the subprojects’ construction, and on various types of economic impacts from the subprojects—collected mainly in meetings with members of farmers’ water user communities (FWUCs) at all 11 subprojects (i.e., excluding Preks of Kandal). This data, while not statistically rigorous, does nonetheless allow to form something of a plausible picture of the economic performance of this component 2–3 years after implementation. The main results of this data collection exercise are summarized in the table at the end of this appendix. 7. From the material in table A8.1, it can be seen that:

(i) In all subprojects bar one (Kork Throng) aggregate rice production has increased, although the nature and source of this increase varies between subprojects. In some places (e.g., Kouch Nop) there is a yield change in the wet season rice crop only, on others (e.g., Link Canal) the subproject now allows early-season rice cultivation (and thus two crops to be cultivated in the wet season—one is relatively shorter and effectively “forced,” the other more traditional in nature), while in other places (e.g., Ponley) dry season cropping (rice and other crops) becomes possible for some or all of the command area.

3 ADB. 2002. Handbook for Integrating Risk Analysis in the Economic Analysis of Projects. Manila. It is especially

unfortunate that specification of risks and assumptions and modelling of risk sources was not captured in either the project framework or the economic analysis, given the variety and magnitude of changes in production conditions that have occurred in Cambodia in the decade since the project was designed.

Appendix 8 54

(i) Some rice crop varietal development and product differentiation has occurred.

(ii) Non-rice cultivation (dry season) lower than expected at design. Competing needs (water for animals) is one stated reason, and major difficulties in non-rice crop commercialization and marketing are still obvious. Also, in several places (e.g., Kouch Nop) water availability has not been as expected for reasons external to the subproject itself.

8. Based on of the subprojects’ various command areas and the reported changes to cropping patterns and yields, an estimate of potential economic returns has been made (also shown in the table). The assumptions behind this estimation of economic returns are as follows:

(i) Estimates are based on actual financial costs of each subproject incurred 100% in 2010, adjusted to economic prices on the basis that 20% of civil works costs were labor (adjusted by a shadow wage rate factor of 0.75) and 50% were tradable items (adjusted to a world numeraire using a standard conversion factor of 0.9).

(ii) Increased rice (paddy) production is valued across all projects at an economic value of $150 per ton. This figure is an aggregate, approximate, and simplifying measure—in the absence of detailed farm budget data—based on reported 2013 farmgate prices of KR1,000–KR1,200 per kg (i.e., $250–$300 per ton), less all farmers’ costs of production (purchased inputs, hired and own labor) and all costs of scheme operation (e.g., annual O&M, routine and periodic maintenance), and deflated to a (comparator) 2010 price level. The resulting EIRRs based on this value and shown in the table should therefore be seen in this context.

(iii) Increased rice production is taken to be constant in all years of project life (20 years) and to start in year 1—except for Kork Thnong, where (modest) benefits start in 2014 only. Again, this is something of a simplifying approximation of reality, and variable returns from year to year are very much a feature of local rice production regimes.

(iv) No other costs or benefits are included in the EIRR estimates.

(v) Net present values (NPVs) are calculated at a 12% discount rate. 9. If the assumptions are varied in sensitivity testing such that (a) project costs are assumed to have been incurred over a 3-year period (2008–2010) rather than in 2010 alone, the EIRRs tend to fall overall by about 20% (proportionately greater for the more profitable schemes); (b) if the economic value of rice were taken to be only $100 per ton, the subprojects’ EIRRs tend to fall by up to a half (depending on their relative profitability)—e.g., at $100 per ton Kouch Noup’s EIRR is 5%, Link Canal’s is 18%, and Kork Throing’s is negative. If it takes 3 years for all rice yields and production to be achieved, EIRRs tend to fall by about 10%–25% (e.g., Kouch Noup’s becomes 9%, Po Pi Deum’s becomes 34%). Clearly, any combination of such effects will tend to depress the model estimates further. 10. Major features of the subprojects’ economic analysis (and in relation to design expectations) on the basis just described are as follows:

(i) The economic price of rice is about three times higher than expected at design (e.g., around $100/ton was estimated in 2010). The impact of this price rise will

55 Appendix 8

largely offset any other changes (e.g., higher construction costs, construction delays, reduced command areas, areas, absence of dry season or livestock and fisheries activities, floods and crop damage) that may have occurred during project implementation and afterwards.

(ii) The aggregate rice production increase across all subprojects is of the order of 18,000 tons annually, worth over $2.7 million in increased economic value (at 2010 prices; over $3 million in current financial prices). Given the nature of value chains in the project area, this flow will likely have contributed to national exports (as paddy to Thailand, and thus will be incremental in nature in economic terms) and also to domestic supply (via mills in Battambang and elsewhere) and thus be non-incremental in nature in economic terms.

(iii) There is considerable variation in economic performance across the subprojects, with some appearing highly viable (e.g., Po Pi Deum, Krouch Saeuch), while others are likely marginal (e.g., Anlong Svey) or negative (e.g., Ta Saom and Kork Thnong, which are both on sandy soils). At those subprojects with the highest returns, reported changes to land prices are notable (e.g., at Po Pi Deum);

(iv) Overall, across all subprojects, the component EIRR is 23%.4

(v) As reported by many farmers, the subprojects have made the fundamental difference for many households between making do with a subsistence level of production and having an income-generating surplus to sell.

(vi) The production of significant surpluses has led to major crop purchases by traders and intermediaries at the subproject sites, and has contributed to the development of local value chains (e.g., to supply newly constructed rice mills in Battambang and elsewhere).

(vii) The estimates of economic viability made in the detailed design studies (shown in parentheses in the table) were broadly in line with current estimates. Po Pi Deum was estimated to be quite viable, Ta Saom and Kork Thnong were only marginal. Krouch Nop seems to have been affected by the lack of water due to outside pressures. Seven of the 11 subprojects have performed worse than expected.

11. A broader and more qualitative consideration of the economic context of subprojects, given information about their physical characteristics, FWUC operations, resources for O&M from the Ministry of Water Resources and Meteorology (MOWRAM), changes to farming practices, and competing demands for labor, among others, also suggests the following points:

(i) Nonquantified benefits are various and likely to be very significant. They may mean more commercial services (e.g., shops, transport, traders) now evident at the subprojects’ sites; better access roads (e.g., for health and social services); more fishing opportunities (as lakes and ponds have increased); more tourist and recreational facilities in some places (e.g., lakeside restaurants, modest sports facilities); and an increase in social capital (e.g., as FWUC members acquire new administrative and technical skills).

4 If costs are assumed to be incurred over 3 years to 2010, all subprojects’ EIRR is 19%; if rice gross margin value

is $100, all subprojects’ EIRR is 15%; if it takes 3 years to realize final improved yield and production, the EIRR is 20%.

Appendix 8 56

(ii) The role of labor in on-farm production at these subproject sites (as elsewhere in the northwest of Cambodia) is changing rapidly. Most obvious is the increased use of machinery for land preparation, some transplanting, harvesting, and transport. Less obvious (but linked to this through a process of labor-substitution where possible) is the increase in migration to seasonal and temporary employment in Thailand. What is clearly happening in many places is that rice-based agriculture as a full-time occupation—even with improved water management at the subprojects—typically cannot compete with the migration opportunities that now exist.5 In this regard, at least some areas of Cambodia are becoming like Thailand’s northeast, where families continue rice farming on land they already own, but the labor and inputs regime adopted thereon is relatively extensive (so broadcasting remains preferable to transplanting), and is one that allows some family members to undertake migratory work in agriculture, factories, construction elsewhere but also to travel home for harvest (and the festivals that follow).6 A result of this development from the subprojects’ point of view is the fact that the envisaged creation of expanded labor opportunities through incremental rice production is very likely to be much less significant than anticipated. This will tend to reduce the project’s poverty impact. 7 Any future interventions in irrigation schemes would do well to closely consider labor availability at design stage.

(iii) The sustainability of estimated economic benefits from crop production thanks to better water management depends on the long-term institutional arrangements for irrigation management transfer (IMT). At present, the FWUCs are still nascent and relatively weak bodies, and MOWRAM’s ability to rationally plan for and fund O&M nationwide is still very limited. The project’s interventions in strengthening institutional capacity at subproject and national levels and in refining the regulatory framework for IMT remain an unfinished agenda, and continuing support through (for example) the Water Resources Management Sector Development Project for FWUC capacity building, planning based on integrated water resources management (IWRM), and O&M resources remains essential for the sustained delivery of irrigation-based benefits. 8

(iv) The realized extent of economic benefits in the longer term is also highly contingent on physical and social conditions surrounding the subprojects. The subprojects must be seen in the wider (i.e., typically river basin) context in which they are located. The impacts of floods (made worse where local forests may have been cut), intensifying climate change (which can intensify floods but might be offset, at least to some extent, if subprojects can store water to avoid mini-droughts during the wet season or prolonged dry seasons themselves), and the demands for water from competing users (often farmers adjacent to or downstream from the subprojects) will all be able to affect yields and production (and thus economic returns) both in seasonal and annual terms.

5 For example, at the Kork Thnong FWUC meeting, one member had just come back from harvesting pineapples in

Rayong, Thailand and he and his wife had saved B8,000 (about $260) in a week. 6 At various of the 11 subproject sites, construction of houses and the presence of vehicles funded from such

migratory activities are now widely apparent. 7 Although it is also the case that at some subproject sites (e.g., Ta Saom) labor is employed (for harvesting) from

elsewhere in Cambodia, and so at a national level the implied reduction in poverty impact may be somewhat offset—especially if these persons are extremely poor.

8 ADB. 2010. Report and Recommendation to the President to the Board of Director on Proposed Loans, Grant,

Technical Assistance Grant, and Administration of Loan and Technical Assistance Grants to the Kingdom of Cambodia for the Water Resources Management Sector Development Program.

57 Appendix 8

(v) Based on the reported yield and production figures, the affordability of the irrigation service fees (ISFs) agreed by the FWUCs is not an issue for most subproject farmers. Given current financial rice prices and likely net financial returns to farming—even with relatively extensive regimes—the planned ISFs are relatively modest. Not surprisingly, data collected at the FWUC interviews suggests that the more economically robust subprojects (e.g., Po Pi Deum, Krouch Saeuch) have relatively consistent and/or high ISF collection levels and percentage of targets met, while the economically weaker schemes (e.g., Kouch Nop, Ta Saom, Kork Thnong) tend to have low (and falling) ISF collection levels.

Table A8: Cropping Patterns and Yields, Other Impacts Summary

Subproject

Cropping Patterns And Yields Summary

Economic Impacts Summary

EIRR Estimate

(%)

NPV Estimate ($'000)

Kouch Noup

Wet season rice yields up to 2.5–3.0 t/ha from around 2.0 t/ha. Small areas of dry season vegetables (beans, salad, morning glory)

• Death of animals in dry season avoided • Vegetables for some families in dry season,

small areas (30 ha?) of dry season rice • Rice crop sales (previously only subsistence) • Access roads, and new roadside shops • Land transaction increase; prices going up • Increase in rice production: (670 ha @ average

of 0.75 t/ha of wet season) + (30 ha of dry season @ 3 t/ha) = 595.5 t/year

10%

(19%)

–72

Tram Mneash

Wet season rice yields up to about 3.0 t/ha from around 1.5 t/ha. About 1 third of area under dry season rice in 2011, but not since

• Year-round water availability for most people • Traders and trucks come to the village (access

roads) • Some local tourism (restaurant and cable sports)

and fishing activities • People do not want to sell land any longer • Increased rice production (1212 ha @ average of

1.5 t/ha wet season = 1,818 t/year)

16%

(18%)

350

Krouch Saeuch

Wet season yields up to 4.5 t/ha from 2.5 t/ha, dry season possible all scheme at average 4.0 t/ha

• Major dry season cropping activity • Traders and trucks arrive; seasonal local markets • Local fishing increases • Increased rice production (1000 ha @ average of

2 t/ha wet season + 50% of dry season @ 4 t/ha = 4,000 t/year)

46%

(18%)

2836

Anlong Svay

Wet season yields up to about 3.0–3.5 t/ha from 1.5–2.0 t/h (broadcasting vs. transplanting). 20–30 ha of dry season rice (with pumping)

• Traders and trucks arrive • District-wide fishing due to large shallow lake • Increased rice production (795 ha @ average of

1 t/ha wet season + 25 ha dry season @ 3 t/ha = 870 t/year)

14%

(26%)

87

Chork Wet season yields up to 2.5 t/ha from around 1.5 t/ha; dry season of 3–4 t/ha on one third of system

• Dry season cropping now possible • Animal deaths avoided • Increased rice production (1372 ha @ average of

1t wet season + 300 ha dry season @ 3 t/ha = 2,272 t/year)

21%

(23%)

874

Canal no. 1 No yield change (3–3.5 t/ha), but early water availability means two crops are possible (early and late wet seasons)

• Double wet season crops; varietal diversity (“soft” varieties now grown) and differentiation increases

• Increased sales to local traders and millers (Battambang); deepening local value chains

• Increased shops and farm services • Increased rice production (1015 ha * 90%

additional wet season @ 3 t/ha = 2,740.5 t/year)

28%

(27%)

1449

Appendix 8 58

Don Aov Wet season yields now 2.5–3.0 t/ha, up from 1.5–2.0 t/ha before subproject. Early-season crop on about two-thirds of subproject = 2.0–2.5 t/ha

• Early-season water availability (previously none)—animals kept alive

• Some intensification of cultivation—attempt to “force” crops

• Improved access • Increased rice production (1,031 ha * 60% wet

season @ 1 t/ha + 1,031 ha * 60% dry season @ 2 t/ha = 1,856 t/year)

18%

(27%)

541

Po Pi Deum

Wet season yields similar before and after (around 2.5 t/ha), but now early season water means 2 crops possible for increasing numbers of farmers

• Early season water availability, and increasing areas under two crops per year (now 75%)

• Increased intensification to “force” early crop • Increased mechanization • Sixfold increase in land values since subproject

($6,000/ha) • Rice sales to local intermediary and truckers,

largely baht-denominated • Increased rice production (1,156 ha * 75% of

area under early-season crop @ 3 t/ha = 2,601 t/year)

43%

(38%)

1798

Ponley Wet season yields up from 1.0–1.5 t/ha to around 2.5 t/ha (“if no flood damage”); about 25% of area (increasing) under dry season crop @ 2.8 t/ha

• Dry season water availability • No land sales (“people now want to keep land” —

previously there were sales) • Some shops due to better access • Increased rice production (480 ha @ 1 t/ha plus

200 ha of dry season @ 2.8 t/ha = 1,040 t/year)

17% *

239

Ta Saom Wet season yields up from less than 1.5 t/ha to 2.0–2.5 t/ha. No dry or early-season crops

• Low input regime remains • Considerable migration (seasonal) competition;

agriculture becomes part-time or for some family members only

• Increased rice production (187 ha @ 1 t/ha = 187 t/year)

6%

(16%)

-90

Kork Throng

No yield change yet—“expected to be better in future”

• No impacts as yet, apart from some small areas of dry season cropping (which will not be repeated as it took water away from animals)

• Full-time rice cannot compete with migration opportunities in Thailand

• Increased rice production (203 ha @ 0.5 t/ha = 101.5 t/year, from year 4 onward)

1%

(12%)

-142

( ) = figures are estimates from the feasibility studies, EIRR = economic internal rate of return, ha = hectare, NPV = net present value, t = ton, t/ha = tons per hectare.

* Ponley EIRR not contained in main design report.

59 A

ppe

nd

ix 9

PROJECT MANAGEMENT STRUCTURE

MINISTRY OF WATER RESOURCES AND

METEOROLOGY (MOWRAM)

Project Management Office (PMO)

National

Provincial

Irrigation Scheme Beneficiaries

Governor Office

PIU Provincial Project

Manager

Consulting Services

Administration Administration/Accountant Procurement Officer Implementation Officer Monitoring and Evaluation Officer

Technical Secondments from PDAFF, PDOE, and PDLMUPC (3) FWUC Support (1) Supervisory Engineer (1) Maintenance Team (1)

PDOE PDWRAM PDLMUPC PDAFF

etc...

PROJECT STEERING COMMITTEE (PSC)

MEF (Chairman), MOWRAM (Secretariat), MAFF, MLMUPC,

MOE, MRD, MPWT, CDC

Environmental Monitoring Unit

(EMU)

INTERMINISTERIAL RESETTLEMENT

COMMITTEE (IRC)

Irrigation Extension Teams

Resettlement Unit (RU)

PRDC Governor Directors all Departments

Resettlement Working Group

PPCC Deputy Governor (Chair) Director PDWRAM Director PDAFF Director PDOE Director PDLMUPC

District Agriculture Office

CDC = Council for Development of Cambodia, FWUC = farmer water user community, MAFF = Ministry of Agriculture, Forestry and Fisheries, MEF = Ministry of Economy and Finance, MLMUPC = Ministry of Land Management, Urban Planning and Construction, MOE = Ministry of Environment, MPWT = Ministry of Public Works and Transport, MRD = Ministry of Rural Development, PDAFF = Provincial Department of Agriculture, Forestry and Fisheries, PDLMUPC = Provincial Department of Land Management, Urban Planning and Construction, PDOE = Provincial Department of Environment, PRDC = Provincial Rural Development Committee, PDWRAM = Provincial Department of Water Resources and Meteorology, PPCC = provincial project coordination committee.

Ap

pe

ndix

10

60

STATUS OF COMPLIANCE WITH LOAN COVENENTS Reference in Loan Agreement

Covenant Responsibility Status of

Compliance

LA, Sch. 6, Para. 1

Project Steering Committee The existing Project Steering Committee (PSC), chaired by the Under Secretary of State of the Ministry of Economy and Finance (MEF) and consisting of representatives from the MEF, MOWRAM, Ministry of Agriculture, Forestry and Fisheries (MAFF), Ministry of Environment (MOE), Ministry of Land Management, Urban Planning and Construction (MLMUPC), Ministry of Rural Development (MRD), Ministry of Public Works and Transport (MPWT) and Council for Development of Cambodia (CDC), is responsible for overall direction of the Project and interagency coordination. PSC shall meet at least twice yearly.

MEF, MOWRAM,

MAFF, MOE, MLMUPC

Complied

LA, Sch. 6, Para. 2

Project Management Office. The existing Project Management Office (PMO) established within MOWRAM shall assume overall responsibility for implementation of the Project. Particularly for this Project, MOWRAM shall appoint a dedicated Project Manager within the PMO to head a team, which shall comprise technical and administrative staff. The PMO shall be responsible for day-to-day management issues including (i) coordinating training; (ii) tendering, negotiating, awarding and administering goods supply, civil works and consulting contracts; (iii) preparing consolidated annual work plans and budgets; (iv) coordinating and providing technical and management support to provincial Project Implementation Units (PIUs); (v) guiding and ensuring compliance with the Borrower’s policy for FWUCs/WUGs activities; (vi) monitoring resettlement and compensation processes; (vii) monitoring and evaluating project performance; and (viii) managing the project imprest accounts and reporting the liquidation of funds in this account to MEF, through MOWRAM.

MOWRAM Complied

LA, Sch. 6, Para. 3

Project Implementation Unit. A Project Implementation Unit (PIU) shall be established within PDWRAM in each of the Project Provinces and shall be headed by a dedicated Provincial Project Manager, whose qualifications are acceptable to the Bank and AFD. The PIUs shall be assisted by seconded staff from PDAFF, PDLMUPC and the provincial Department of Environment (PDE), and shall include technical and administrative staff. PIUs shall have access to PDWRAM’s technical specialists, if required. The PIUs shall be responsible for day-to-day management of project activities at the provincial level including: (i) participation in and support to surveys and studies commissioned by PMO; (ii) support in identifying and selecting subprojects; (iii) monitoring of civil works construction; (iv) coordination, monitoring and supervision of the activities contracted to NGOs and (v) undertaking portions of O&M not carried out by beneficiaries.

MOWRAM, PDAFF, PDE, PDMLMUPC

Complied

LA, Sch. 6, Para. 4

Provincial Project Coordination Committee. A Provincial Project Coordination Committee (PPCC) shall be established and chaired by a Provincial Deputy Governor and shall comprise representatives, of director level, of the provincial line departments directly concerned with the Project. All eligible subprojects proposed under the Project shall be referred by the PIU to PPCC for approval prior to final ratification by MOWRAM, ADB and AFD.

MOWRAM, Provincial

Government, Line Agencies

Complied

LA, Sch. 6, Para. 5

Environment Management Unit. An Environment Management Unit (EMU) shall be established within the PMO and shall consist of one environmental specialist, who is seconded from the MOE, and one sociologist. At provincial level, one environmental specialist will be seconded to each of the PIUs from PDE. The EMU shall be responsible for (i) guiding and monitoring the preparation of the initial environmental examination (IEE) required for each subproject at feasibility stage in accordance with the agreed environmental assessment and review procedures; (ii) the timely submission of IEEs to the concerned environmental agencies, the Bank and AFD for clearance prior to the commencement of detailed design and construction; (iii) ensuring the environmental mitigation measures recommended in the IEEs are incorporated into design and civil works contracts and in subsequent operations of subprojects; and (v) ensuring that environmental monitoring is adequately reported in project review and progress reporting. Seconded PDE staff in PIUs shall (i) assist in the preparation of each

MOWRAM, MOE

Complied

61 A

ppe

nd

ix 1

0

subproject IEEs; (ii) ensure adoption of recommended mitigation measures during construction; (iii) monitor water quality in all subproject locations; and (iv) report environmental monitoring results to the EMU.

LA, Sch. 6, Para. 6

Provincial Resettlement Working Groups and MOWRAM Resettlement Unit (RU). Resettlement Working Groups (RWGs) shall be established in each of the Project Provinces. The RWG shall consist of representatives from PDWRAM, PDAFF and the provincial Department of Land Management, Urban Planning and Construction (PDLMUPC). Implementation procedures, as detailed in the agreed Resettlement Framework (RF), shall be applied for the preparation of the Resettlement Plans (RPs) required for non-core subprojects. MOWRAM’s Resettlement Unit (RU) shall be responsible for (i) supervising the provincial working groups in the land holding identification surveys, socio-economic surveys, detailed measurements for the identification of affected people, and land adjustments/consolidations resolution process required for the RPs and (ii) preparing the RPs and submitting them to the Inter-ministerial Resettlement Committee (IRC), the Bank and AFD for endorsement and subsequent implementation through MEF.

MOWRAM, PDLMUPC,

PDAFF

Complied

LA, Sch. 6, Para. 7

Water Resources Multi-user Committees. Water Resources Multi-user Management Committees (WRMCs) shall be established at basin or sub-basin level where subprojects draw water from sources that are used by others. WRMCs shall consist of water users, and representatives of relevant local government and provincial departments. Overall policy relating to management of water, and its release from barrages and reservoirs shall be discussed with and agreed by WRMCs.

MOWRAM Complied

LA, Sch. 6, Para. 8

Accounting and Reporting. Without prejudice to the provision of Section 4.07 (b) of this Loan Agreement, the Borrower and MOWRAM shall ensure that each of the PIUs shall provide the PMO with a monthly financial report and quarterly progress reports detailing the physical and financial progress of activities against agreed targets and summarizes project performance. MOWRAM shall submit to the Bank and AFD, the consolidated progress reports, including environmental monitoring and comprehensive annual progress report. The Borrower shall further ensure that not later than six (6) months as of the Effective Date, MOWRAM shall establish a simple computerized management information system in the PMO for efficient monitoring, financial management, and reporting of component progress and contracts. The PMO and PIUs shall establish and maintain, separate records and accounts adequate to identify income and expenditures relating to the Project.

MOWRAM Complied

LA, Sch. 6, Para. 9

Project Performance Monitoring and Evaluation. The Borrower shall ensure that the Project shall adopt the Bank’s project performance management system (PPMS), including development and implementation of a computerized management information system, to monitor and evaluate implementation performance and development impacts at various stages of the Project. The PPMS shall build on the experience gained (i) in developing a computerized data base system under previous Bank-financed technical assistance, (ii) through the baseline surveys and data collection carried out during study of the core subprojects, and (iii) the benefit monitoring and evaluation activities carried out for other Bank projects. During project implementation, the PPMS shall comprise of (i) the project logical framework; (ii) regular project performance reports to track progress from the baseline situation against identified indicators; and (iii) a project completion report. PPMS activities shall include participatory baseline physical and socioeconomic surveys conducted in each of the subprojects during the feasibility studies, and follow-up surveys to evaluate changes and measure benefits generated, in particular, the impact on the poor.

MOWRAM Complied

LA, Sch. 6, Para. 10

Project Review. During the initial 3 years of Project implementation, the Bank, AFD and the Borrower shall jointly carry out semi-annual reviews to (i) determine if all proposed implementation arrangements are appropriate and in place; (ii) assess the progress of selection of the subprojects; (iii) review resettlement and land adjustment program; (iv) review the study, design and construction programs; and (vi) review the conduct and the outcomes of the capacity building and training activities.

MOWRAM Complied

Ap

pe

ndix

10

62


At the end of the third year of Project implementation, the Bank, AFD and the Borrower shall undertake a comprehensive mid-term review in consultation with the PSC. Prior to the mid-term review, the PMO shall prepare a comprehensive report identifying issues for consideration. The midterm review shall examine (i) project progress to date, including implementation of the subproject selection process, quality of designs and physical works, and initial benefits; (ii) future implementation plans and targets; (iii) allocation of funds for components and loan categories; (iv) effectiveness of the project management structure and implementation arrangements as well as any need for modification; (v) the status of institutional developments; and (vi) any major environmental, social, technical, and economic aspects of the Project. Based on the findings of this review, changes in project scope, financing, or implementation arrangements deemed necessary may be considered.

MOWRAM Complied


Counterpart Funds. The Borrower shall provide counterpart funds for the project implementation on time. The Borrower shall make timely submission of annual budgetary appropriation requests and ensure prompt disbursement of appropriated funds during each year of Project implementation.

MEF, MOWRAM

Complied


O&M. The Borrower shall ensure that following the Project completion, the infrastructure constructed and/or rehabilitated under the subprojects shall be maintained in good working condition. For this purpose, the Borrower shall continue to provide the required funds through budgetary allocations for the Borrower’s share of the routine O&M, and for the periodic and emergency maintenance for the infrastructure.

MOWRAM Partly complied


Law on Water Resources Management and its Sub-decrees. The Borrower shall (i) before the Law on Water Resources Management is approved by the National Assembly, commence work to prepare the required sub-decrees to define the status, rights and duties of FWUCs and the process for IMT; (ii) not later than 12 months following approval of the Law on Water Resources Management by the National Assembly, promulgate and declare the Law on Water Resources Management effective; and (iii) finalize and issue the sub-decrees not later than six months after the promulgation of the Law on Water Resources Management.

MOWRAM Not complied


Environment. The Borrower and MOWRAM shall ensure that (i) EMU be responsible for ensuring full compliance of the Project with the Bank’s Environmental Assessment Guidelines, the Borrower’s laws and regulations, and best environmental practices; and that the Project shall meet the mitigation and monitoring requirements described in the agreed Environmental Assessments and Review Procedures; and (ii) adequate budgetary provisions shall be made for environmental assessment and review of subprojects, implementation of mitigation measures, environmental monitoring, and institutional arrangements for coordination and supervision as identified in the subproject IEEs.

MOWRAM, MOE

Complied


Mines and Unexploded Ordnance. The Borrower and MOWRAM shall ensure removal, at its own cost, of all mines or unexploded ordnance suspected to exist on subproject sites. Further, the Borrower and MOWRAM shall engage the services of a reputable and experienced mine security organization, acceptable to the Bank and AFD, to confirm that the concerned area is clear of mines or unexploded ordnance. The Borrower shall further ensure that no subproject rehabilitation will commence until de-mining has been completed and certification to that effect has been obtained.

MOWRAM Complied


Resettlement. The Borrower and MOWRAM shall ensure that land acquisition and resettlement activities are implemented in accordance with all of the Borrower’s prevailing laws and regulations and the Bank’s policy on involuntary resettlement, as set out in the agreed RF. The Borrower shall provide unforeseen cost in excess of budget estimates to meet the requirements as outlined in the RPs. The Borrower shall further ensure that the land adjustment program at the tertiary and quaternary canals level will not adversely affect the villages and individual farmers.

MEF, IRC, MOWRAM

Complied

63 A

ppe

nd

ix 1

0

IRC = Inter-Ministerial Resettlement Committee MEF = Ministry of Economy and Finance MLMUPC = Ministry of Land Management, Urban Planning and Construction MOE = Ministry of Environment MOWRAM = Ministry of Water Resources and Meteorology PDAFF = Provincial department of Agriculture Forestry and Fisheries PDE = Provincial department of Environment PDLMUPC = Provincial department of Land Management, Urban Planning and Construction


Gender. The Borrower shall ensure that the Project will provide full opportunities for women participation, in the FWUCs and the IEWs, and shall implement the gender action plan, which has been agreed between the Borrower and the Bank, in order to reach the targets aimed at maximizing women’s participation in Project activities and benefits.

MOWRAM Complied


Selection of Subprojects. The Borrower and MOWRAM shall ensure that selection of subprojects shall be carried out in accordance with the selection process, which has been agreed between the Borrower and the Bank.

MOWRAM Complied

LA, Art. IV, Sect.4.03 (a)

Particular Covenants. In the carrying out of the Project, the Borrower shall cause competent and qualified consultants and contractors, acceptable to the Borrower and ADB, to be employed to an extent and upon terms and conditions satisfactory to the Borrower and ADB.

MOWRAM Complied

LA, Art. IV, Sect. 4.06 (b)

Particular Covenants. The Borrower shall (i) maintain, or cause to be maintained, separate accounts for the Project; (ii) have such accounts and related financial statements audited annually, in accordance with appropriate auditing standards consistently applied, by independent auditors whose qualifications, experience, and terms of reference are acceptable to ADB; (iii) furnish to ADB not later than nine (9) months after the end of each related fiscal year, certified copies of such audited accounts and financial statements and the report of the auditors relating thereto; and (iv) furnish to ADB such other information concerning such accounts and financial statements and the audit thereof as ADB shall from time to time reasonably request.

MEF, MOWRAM

Complied

LA, Art. IV, Sect. 4.07 (b)

Particular Covenants. The Borrower shall furnish, or cause to be furnished, to ADB quarterly reports on the carrying out of the Project and on the operation and management of the Project facilities.

MOWRAM Complied

LA, Art. IV, Sect. 4.07 (c)

Particular Covenants. Promptly after physical completion of the Project, but in any event not later than three (3) months thereafter or such date as may be agreed for this purpose between the Borrower and ADB, the Borrower shall prepare and furnish to ADB a report, in such form and in such detail as ADB shall reasonably request, on the execution and initial operation of the Project, including its cost, the performance by the Borrower of its obligations under the Loan Agreement, and the accomplishment of the purposes of the Loan.

MOWRAM Complied

Appendix 11 64

SUMMARY OF GENDER EQUALITY RESULTS AND ACHIEVEMENTS

A. Project Description 1. The project was designed1 to reduce poverty in selected northwest rural areas through enhanced agriculture production, thereby alleviating food insecurity and improving family household incomes. The project outcome was to establish rehabilitated and sustainably operational small to medium-sized irrigation schemes and other water control infrastructure within priority river basins. The project outputs were to be (i) water sector institutional strengthening; (ii) irrigation infrastructure development and management; and (iii) support to development of irrigated agriculture. The project covered four provinces: Battambang, Banteay Meanchey, Pursat, and Siem Reap.2 The intended impact included raising average household incomes by at least 30% for 35,000 households and reducing the incidence of poverty among beneficiaries to 10%. 2. The project was implemented from February 2005 to December 2011. It was classified as effective gender mainstreaming (EGM) and had a gender action plan (GAP). B. Gender Analysis and Project Design Features

a. Gender Issues and Gender Action Plan features 3. Cambodian women are extensively involved in agricultural production, but mostly as unpaid family workers. Households headed by women constituted 10%–25% of the households expected to benefit from the project—such households faced additional disadvantages related to poor access to land, lack of adult labor for income earning, higher than average dependency ratios, and vulnerability to sudden shocks (that may intensify poverty). In Khmer culture, women are discouraged from participating in the community’s decision-making processes, it is regarded as a male preserve. This reduces women’s ability to effectively access resources and improve their situation. 4. Issues emerging from women's group discussions during project preparation included (i) lack of food security; (ii) indebtedness due to crop losses; (iii) lack of access to institutional credit; (iv) lack of access to decision making; (iv) heavy workloads; and (v) poor health, including high maternal and child mortality. Women identified needs for (i) food security; (ii) secure irrigation and water management for rice production; (iii) prevention of crop losses on account of natural disasters; (iv) training and capacity building for water management; (v) alternative livelihoods development; (vi) access to institutional credit; (vii) better access to rural infrastructure (including roads, markets, health centers); (viii) access to extension and training for better crop production; and (ix) land consolidation. Many of these needs were generally addressed by the scope of project interventions, and a GAP was developed to target women’s needs specifically. 5. During design, participatory appraisals were carried out to identify the main concerns and priorities of the poor, and women's focus group meetings were organized to enable poor women to speak openly about their particular concerns.

1 Revised project framework of 2011 (Appendix 7).

2 One subproject, Preks in Kandal Province, was funded by a grant from Agence Française de Développement.

65 Appendix 11

b. Overall Assessment of Gender-Related Results/Achievements 6. All project implementation units and service providers for the subprojects received initial gender mainstreaming training through technical assistance (TA) before setting up farmers’ water user communities (FWUCs). Water users in all communities then also attended participatory irrigation management development (PIMD) dissemination training, which included gender mainstreaming as a core subject.

7. Women participated in training activities, accounting for 44% of 39,925 total participants; training included FWUC and water user group (WUG) management and O&M procedures, accounting and financial planning and reporting, conflict resolution, and mainstreaming gender and other issues of vulnerable groups into irrigation management.

8. Women also made up 54% of farmers (5,312 out of 9,900) selected to attend Farmer Field Schools (FFSs) between 2009 and 2011. The FFSs covered rice, vegetable, fruit tree, mushroom, chicken, fish, and pig production.

9. It was also reported that all target areas had been provided with community orientation, including on gender mainstreaming.

10. The implementation of GAP activities was not regularly monitored and no monitoring table was prepared during implementation, although an international gender specialist was recruited for 2 person-months at the beginning of the project. She developed a gender and development awareness training module in 2005, which has been used by the community development specialists for training. Sex-disaggregated data was collected for some of the activities.

c. Gender Equality Results

11. Participation, access to project resources, and practical benefits. About 5,312 women (54% of 9,900 participants) benefited from 1,320 training units or demonstrations on livelihood issues, including rice cultivation (wet and dry seasons); innovative techniques for use of organic fertilizers and pesticides; pig, chicken, and fish raising; vegetable, mushroom, and fruit tree growing. Women also made up 44% of the 35,020 water users in all communities who attended PIMD dissemination training. All in all, women’s participation in training events averaged 44% and in some cases reached over 60% of participants. 12. Despite their strong participation in training, women held only 65 of 483 FWUC positions in late 2013 when the Project Completion Report was prepared, and the average number of women members across all subprojects was only 13%. In addition, women tend to take the positions of cashier or accountant (Table A11.1).

Table A11.1: Composition of Farmers’ Water User Communities and Gender Status, late 2013

Subproject FWUC Composition

Total Female

Kouch Nop 35 1 accountant; 2 block heads

Tram Mneash 66 1 accountant

Krouch Saeuch 43 1 accountant

Anlong Svay 27 0

Appendix 11 66

Chork 68 2 committee members; 6 members

Canal no. 1 30 1 member

Don Aov 49 2 members; 1 on committee but resigned

Po Pi Deum 40 1 accountant; 1 member

Ponley 34 1 accountant, but resigned

Ta Saom 12 1 accountant

Kork Throng 10 (now 5) 2 members

13. According to women beneficiaries, they wanted to participate and benefit from project-sponsored training and also be up to date on what was happening in the community. But they did not generally feel that they had sufficient spare time to take on work for their FWUC. Lack of time due to farming activities, housekeeping chores, and parental and nursing responsibilities was quoted as the main reason, indicating a basic design issue (assuming elasticity of women’s time). If women were to spend time working on FWUC committees, then other time-saving mechanisms would need to be established. 14. Strategic changes in gender relations. Few changes of this nature were noted, at least not on available evidence about women’s participation in formal institutional structures. Data collected in late 2013 during discussions with FWUCs indicated that: (i) although participating in FWUCs at all subprojects, women remain substantially under-represented overall—both as general members and as committee members; (ii) where represented on management committees, women almost exclusively fill the role of cashier or accountant (and by doing so constitute at least 25% of committee membership); (iv) only at Chork are 50% of committee members women; and (v) women accountants seem to have a tendency to resign after a year or two citing difficulties in balancing work and family responsibilities. 15. Clearly, in terms of participation and access to project resources (e.g., through training), gender objectives were largely met. However, in terms of strategic changes in gender relations at home and within the community, results have been modest. At the late-2013 FWUC meetings, women voiced their difficulties in participating in FWUC operations, mainly because of conflicts with more traditional household duties and roles in FWUC, such as attending meetings and water management operations. In the long term, women’s participation in training and FWUC operations will strengthen their voice in local decision making and increase women farmers’ productivity. 16. Contribution of gender equality results to overall loan outcomes and effectiveness. On currently available evidence, the contribution of gender equality results to overall loan outcomes and effectiveness has been modest. However, women’s strong participation in irrigation management training and agriculture training may contribute, in the longer term, to strengthened management of irrigation systems and to the achievement of the project’s goal of enhancing agricultural productivity in the target areas. C. Lessons and Recommendations 17. The GAP, attached to the report and recommendation of the President as one of the supplementary appendixes, was descriptive, with only one performance indicator clearly established,3 and the GAP monitoring table was not prepared. This was probably due to unclear guidance during the project design period. In that regard, the project completion mission was

3 Women would form at least 30% of FWUCs.

67 Appendix 11

not able to prepare the formal GAP monitoring table with specific gender results.4 However, Table A11.2 attempts to summarize major gender activities with quantified performance indicators where available. In general, if the GAP had been well prepared, particularly to capture the participation of women farmers in agricultural extension activities, and with guidance, the gender reporting would have been much better throughout project implementation.

18. The delay to the civil works program and the short time left to the FWUCs for operating the rehabilitated systems minimized the potential for farmers, including women, to benefit more deeply from their involvement and experience before the project ended.

19. Women’s strong participation in training is notable. However, women’s membership in FWUCs could have been improved by project interventions to support women in overcoming barriers, such as family obligations, or traditional notions of women’s responsibilities. D. Conclusion and Recommendations:

(i) GAP design and gender targets. The GAP should be designed in a simple

format, and targets and indicators need to be proposed in line with the project framework (now the design and monitoring framework [DMF]). Furthermore, gender targets should be clearly determined and included or mainstreamed in the DMF. The project framework should have at least 50% of project performance indicators to ensure and monitor the project’s contribution to poverty reduction and gender equity results.

(ii) GAP implementation and monitoring. Some critical factors associated with the

GAP’s success and results are: (a) gender focal points should be clearly assigned at various project implementation levels, from project management down to those actually implementing the project, with specific tasks and roles adequately defined to ensure implementation of the GAP. In addition, gender specialist inputs should be greater than the allocated 2 person-months, to ensure that the expert has enough time to transfer knowledge and assist GAP implementation; (b) sex-disaggregated data need to be an integral part of all project monitoring and evaluation (M&E), including in the baseline survey and the project performance management system (PPMS). Better M&E training and support should be considered for measuring behavioral change, traditional practices, and perceptions related to gender roles; (c) orientation or guidance on GAP implementation and the project’s gender issues should be conducted at the beginning of project implementation; and (d) the GAP should be revisited during the midterm review, and the progress of its implementation should be recorded and included in the project progress reports.

(iii) The FWUCs’ standard statute should be revisited and revised to be more user

friendly, and it should enable participation of women who are not household heads and less educated. Although the FWUCs follow democratic principles—decisions and elections are based on majority votes—a clear quota for women should be considered and integrated in committee structures to enable women to participate in leadership and increase their decision-making power. Women

4 The guideline for preparing this table was only issued in 2011.

Appendix 11 68

should be encouraged to assume leadership positions and not just be limited to being cashiers or accountants, which requires a lot of time for collecting the fees.

(iv) The timing of project activities should be more flexible to ensure that more

women are able to participate and draw benefits from the project, such as knowledge or new initiatives.

Table A11.2: Summary of Gender Activities

Project GAP Features and Targets a Achievements at Project Completion

• FWUCs - women will form at least 30%. It was not to be “compulsory” but would depend “upon the capacity available to carry out the functions of the FWUC”.

• All participating communities would undergo gender training prior to starting the community training aimed at formation of the FWUC.

• In order to facilitate women's representation on the FWUC, the functions of each position would be explained in order to help identification of suitable representatives including women to specific positions on the FWUCs. Selection criteria should be flexible and should not exclude the disadvantaged sections and women.

• At least one woman member and one representative of disadvantaged sections must be included in each FWUC to begin with, along with women who may be elected to specific committee positions.

• In order to address women's lack of capacity to participate equally in community level decision-making, women's sub groups would be organized at village level for capacity building to support their increased participation in decision making, training and extension, as well as microfinance. Such groups should be

• Of 483 FWUC positions, women held only 65 (13.5%), and the average number of women members across all subprojects was only 12%. However, women occupied on average 25% of FWUC management committees, but were most likely to take the position of cashier or accountant. The main reasons for low committee representation: (i) general feeling among women that they do not have enough spare time to take on additional workloads; (ii) not enough time due to farming activities, household chores, parental and nursing responsibilities; (iii) gender stereotype as to what are suitable jobs for men and women in the community; and (iv) less self-confidence due to education background.

• Water users in all communities attended PIMD dissemination training (16 subject areas), which included gender issues related to water use and irrigation as a core subject. 35,020 water users attended (44% were women).

• The functions of each FWUC position were explained during PIU training at each site before civil works began. As a result, women mostly assumed the role of cashier or accountant in FWUCs. Also, women’s high participation in training (e.g., on FWUC and WUG management, O&M) suggests good appreciation of FWUCs’ roles early in the project life.

• Women were included in all FWUCs from the outset. Throughout project implementation they were disproportionately over-represented in the position of FWUC accountant.

• During the early life of the project, 2008–2009, village-specific women’s subgroups were organized and trained (one subgroup per village). The subgroups were later integrated in the statute and construction committees. It appears that the high participation in PIU and agriculture training obviated the need for specific women’s subgroups to be formed. The subgroups were consulted on the

69 Appendix 11

formed prior to construction in order that they may be consulted during survey and design.

• Monitoring indicators for women's membership and active participation would include membership lists, attendance at meetings, and participation in training programs. These indicators will be updated by the IEWs and will used for monitoring levels of women's participation.

• IEWs - The Project will attempt to deploy women Irrigation Extension Workers in teams (one woman per team of IEWs) to work with the FWUCs.

• All IEWs will undergo mandatory training in gender responsive working, participatory approaches, and will actively provide support to women members of FWUCs.

• IEWs will set up women's sub-groups, organize regular meetings, and training sessions.

• Records of all meetings, participation by women will be maintained.

• Domestic Water Use - The Project will make particular efforts to fully involve women members of the Subproject communities in deciding about washing points, drinking water supplies, in order to benefit them by reducing drudgery, and increasing time availability of women.

• Monitoring and Evaluation - Monitoring and evaluation will be gender sensitive and specifically monitor levels of women's participation, women's access to irrigated land, rice production improvements, opportunities for crop diversification, income improvements impacts on women's work among others.

• Consultant services - The community development specialist will have specific

formation of FWUCs and on the design of irrigation infrastructures.

• Membership of FWUC management committee and

other committees by gender were maintained and monitored. Women’s participation in PIU and agriculture training by course and site was monitored thoroughly.

• Out of 11 IEW teams (three IEWs per team per

subproject), only four teams included women. However, there were two women in two teams, Ponley and Krouch Saeurch. All in all, there were 33 IEWs, 6 of whom were women (18%).

• This training was conducted by the service providers at each subproject and continued to be guided by the female FWUC specialist during implementation.

• Women’s subgroups were organized for meeting and training at the start of the project, but these subgroups have been integrated in the statute and construction committees. The two committees later formed FWUCs via election.

• Not all meetings and participation by women were recorded, particularly not early in project implementation. However, participation in FWUCs and agriculture extension training has been summarized in Appendix 4, Table A4.2 and Appendix 5, Table A5.1.

• Design and construction subcommittees were actively engaged in detailed design of washing points, drinking water supplies, taps, and other features, but no detailed information on gender composition is available. The whole community, women included, benefits from the infrastructure built.

• Women’s participation in livelihood training was monitored—with a notable 5,312 women (54% of 9,900) selected to attend FFSs from 2009 to 2011. The FFSs included trainings and demonstrations on rice cultivation (wet and dry season), innovative techniques for use of organic fertilizers and pesticides; pig, chicken, and fish raising; vegetable, mushroom, and fruit tree growing. However, by the end of the project, the extent of diversification and intensified agricultural activities on farmers’ plots was uncertain (and likely modest). In addition, data on women’s participation in irrigated agriculture was not compiled at project completion due to lateness of the schemes’ completion.

• An international gender specialist was recruited for 2 person-months at the beginning of the project. She

Appendix 11 70

skills and experience in gender action plan. She will be responsible for mainstreaming gender issues in to the Project, through training and capacity building of staff of the Project Implementation Units. She will develop the gender related activity plans of the IEWs and monitor their performance on these. All training and extension activities will identify gender-based requirements and address women's specific requirements.

developed a gender and development awareness training module in 2005, which has been used by the community development and FWUC specialists for training.

b Training on mainstreaming gender and

other issues of disadvantaged and vulnerable groups was conducted by IETs and by the community development and FWUC specialists (one international and one national).

FFS = farmer field school, GAP = gender action plan, FWUCs = farmer water user communities, IET = irrigation extension team, IEW = irrigation extension worker, O&M = Operation and Maintenance, PIMD = participatory irrigation management development, PIU = project implementation Unit, WUG = water user group, a

The GAP monitoring table was not prepared for this project. b The mission activities report was prepared by the international gender specialist in 2005.

71 Appendix 12

ENVIRONMENTAL SAFEGUARDS A. Introduction 1. The approach to environmental management recommended in the project’s environmental management plan (EMP) took into account the weak institutional capacity in environmental management in Cambodia and aimed to use the flexibility inherent in a sector project to recognize and avoid any potential environmental hazards through a four-step process:

(i) Selection criteria for subprojects were designed to avoid any really obvious environmental concerns. (Thus no subprojects were to be undertaken within the Tonle Sap Biosphere Reserve (TSBR) or within any wildlife sanctuaries or forest reserves.)

(ii) Every subproject was to be subjected to an initial environmental examination (IEE) and, if necessary, an environmental impact assessment (EIA). This would allow any potential environmental hazard to be identified and contingencies and mitigation measures drawn up. (If it appeared that mitigation measures could not be satisfactorily designed to ensure a minimal environmental impact, the subproject would be dropped).

(iii) The detailed design of every subproject was to be checked before construction to ensure that mitigation measures had been incorporated and appropriate safeguards and required practices written into construction contracts.

(iv) Construction and operation of subproject schemes was to be monitored to ensure compliance with environmental guidelines and that no serious environmental problem would arise.

2. Steps 1–3 would be undertaken during project implementation, when both international and local consulting expertise would be present in the project management office (PMO) to help ensure compliance with the plan. Concurrently, and as part of implementing the project and the EMP, the capacity of the Ministry of Water Resources and Meteorology (MOWRAM) and the provincial departments of the environment (PDEs) in environmental planning and monitoring will be strengthened. 3. Step 4 was to be initiated during implementation but was expected to continue into the operational phase.

4. In summary, the approach to environmental management was to avoid major negative impacts in the first place through (i) the subproject selection process; (ii) recognition and mitigation of lesser hazards identified by the IEEs; (iii) ensuring of environmentally sound construction and operation, and monitoring of subprojects to ensure sound operational practices were being followed and mitigation measures were working as expected.

B. Project Performance During Implementation 5. The EMP contained in the report and recommendation of the President (RRP) specified various activities to be undertaken during subproject identification, design, and implementation.1

1 Asian Development Bank (ADB). 2003. Report and Recommendation of the President to the Board of

Directors on a Proposed Loan to the Kingdom of Cambodia for the Northwest Irrigation Sector Project. (Supplementary Appendix D; Environmental Assessment and Review Procedures.)

Appendix 12 72

6. These were inevitably affected by the revisions to expected processes, as outlined in Appendix 3, and in fact the selection of subprojects applied broadly involved a four-phased process of identification, river basin and water use studies, feasibility study, and detailed design. This complied with the original design logic of the RRP, and environmental aspects (especially of water resources management) remained integral to the entire process. 7. Monitoring of the environmental aspects of project implementation was the responsibility of the PMO’s environmental management unit, and of the project implementation unit (PIU) assisted by environmental specialists seconded from the PDEs. Agency performance was generally adequate, and institutional and technical requirements of the loan covenants were complied with. Specific activities undertaken included:

(i) detailed designs followed the environmental analysis or IEE as captured in the river basin studies and/or feasibility studies;

(ii) detailed designs were approved by relevant agencies (i.e., Ministry of Agriculture, Fisheries and Forestry, Ministry of the Environment);

(iii) environmental specialists in the PMO vetted individual contracts to ensure that any special or particular safeguard recommended by the IEE for the particular subproject was incorporated within that contract; and

(iv) 12 IEE or EMP reports were prepared—three for core subprojects done during project preparatory technical assistance (TA) in 2002 (Link Canal, Po Pi Deum, and Ponley) and nine prepared by service providers during project implementation between 2007 and 2009. However, Link Canal was later dropped from the project. This means that all 11 subprojects have been accompanied by the IEE and EMP. Implementation consultants assisted the executing agency, and local authorities monitored environmental aspects during the construction period, as indicated in para. 27 of the Environment Assessment Review Framework in Project Administration Manual document.

8. The characteristics and quality of environmental work at the subprojects generally was affected by the factors surrounding the subprojects’ construction: (i) the greater than expected complexity of many subproject sites identified at the RRP stage (which led to a necessary refocusing on a smaller number of physical sites); (ii) the limited technical capacity of local service providers, which led to a lack of appreciation of design criteria in some instances, poor quality of reporting (including on environmental aspects), and reliance of TA support for problem resolution; (iii) delays in contract letting and procurement2 (which added to time pressures to complete designs and to complete construction phases)—all of which meant that there was much less time than anticipated for environmental monitoring before, during, and after construction, and during operations. 9. As of late 2013, the experience with subproject construction and implementation also illustrates the following points in environmental terms:

(i) individual subprojects’ performance is affected by natural resource problems associated with complex topography and flatness, and the local prevalence of sandy soils;

(ii) subprojects’ operations can be significantly affected by major (often not anticipated) civil works development nearby—e.g, the effects of dams reducing water flows or flows being diverted; and

(iii) climate change effects are greater than anticipated, and the effects of floods are often compounded by local forest losses.

2 Partly due to resettlement and land titling issues.

34379-013: Completion Report - ADB

Documents