Completion Report. Project Number: Loan Number: 1531 October Pakistan: Dera Ghazi Khan Rural Development Project

Transcription

1 Completion Report Project Number: Loan Number: 1531 October 2007 Pakistan: Dera Ghazi Khan Rural Development Project

2 CURRENCY EQUIVALENTS Currency Unit Pakistani rupee/s (PRe/PRs) At Appraisal At Project Completion August June 2007 PRe1.00 = $ $ $1.00 = PRs PRs61 ABBREVIATIONS BME benefit monitoring and evaluation C&WD Communication and Works Department CD community development CO community organization CSP country strategy and program DG Khan Dera Ghazi Khan DIP Department of Irrigation and Power DOFWM Directorate of On-Farm Water Management DTW deep turbine tube well EA executing agency EIRR economic internal rate of return FS financial services ha hectare IA implementing agency km kilometer LGO local government ordinance NGO nongovernment organization NRSP National Rural Support Program O&M operation and maintenance PCR project completion review PERI Punjab Economic Research Institute PMU project management unit SCO savings and credit organization SDR special drawing rights WC watercourse WUA water users association NOTES (i) The fiscal year (FY) of the Government and its agencies ends on 30 June. FY before a calendar year denotes the year in which the fiscal year ends, e.g., FY ends on 30 June (ii) In this report, $ refers to US dollars.

3 Vice President Liqun Jin, Operations 1 Director General J. Miranda, Central and West Asia Department (CWRD) Director P. L. Fedon, Country Director, Pakistan Resident Mission (PRM), CWRD Team leader Team member M. S. Shafi, Project Implementation Officer, PRM, CWRD N. ul Islam, Associate Project Analyst, PRM, CWRD

4 CONTENTS Page BASIC DATA MAP I. PROJECT DESCRIPTION 1 II. EVALUATION OF DESIGN AND IMPLEMENTATION 1 A. Relevance of Design and Formulation 1 B. Project Outputs 3 C. Project Costs 5 D. Disbursements 5 E. Project Schedule 6 F. Implementation Arrangements 6 G. Conditions and Covenants 6 H. Consultant Recruitment and Procurement 7 I Performance of Consultants, Contractors, and Suppliers 8 J. Performance of the Borrower and the Executing Agency 8 K. Performance of the Asian Development Bank 9 III. EVALUATION OF PERFORMANCE 9 A. Relevance 9 B. Effectiveness in Achieving Outcome 10 C. Efficiency in Achieving Outcome and Outputs 11 D. Preliminary Assessment of Sustainability 11 E. Impact 12 IV. OVERALL ASSESSMENT AND RECOMMENDATIONS 13 A. Overall Assessment 13 B. Lessons 14 C. Recommendations 15 APPENDIXES 1. Planned and Achieved Outputs of the Project 2. Categories of Community Development Subproject 3. Completion Dates of Rural Roads under the Dera Ghazi Khan Rural Development Project 4. Achievement under the Financial Services Component 5. Appraised and Completed Costs of the Project 6. Contract Awards and Disbursements 7. Implementation Schedule 8. Status of Compliance with Loan Covenants 9. Economic and Financial Analyses 10. Project Framework ii vii

5

6 BASIC DATA A. Loan Identification 1. Country 2. Loan Number 3. Project Title 4. Borrower 5. Executing Agency 6. Amount of Loan 7. Project Completion Report Number B. Loan Data 1. Appraisal Date Started Date Completed 2. Loan Negotiations Date Started Date Completed 3. Date of Board Approval 4. Date of Loan Agreement 5. Date of Loan Effectiveness In Loan Agreement Actual Number of Extensions 6. Closing Date In Loan Agreement Actual Number of Extensions 7. Terms of Loan Interest Rate Maturity Grace Period 8. Terms of Relending (if any) Interest Rate Maturity Grace Period Second-Step Borrower Pakistan 1531 Dera Ghazi Khan Rural Development Islamic Republic of Pakistan Punjab Planning and Development Department SDR million PCR:PAK May June July July September October January January 1998 None 31 December May % 35 years 10 years 1.0% 35 years 10 years National Rural Support Program, $2.7 million at 10.0%, repayable 1 year after project completion 9. Disbursements a. Dates Initial Disbursement 9 October 1998 Final Disbursement 31 May 2007 Time Interval 104 months

7 iii Effective Date 12 January 1998 Original Closing Date 31 December 2004 Time Interval 83 months b. Amount (SDR million) Category or Subloan No. Original Allocation Net Amount Available Amount Disbursed Undisbursed Balance 01A B C A (0.137) 03B C (0.113) 03D Total Local Costs (Financed) - Amount ($ million) Percentage of Local Costs Percentage of Total Cost 42.1 C. Project Data 1. Project Cost ($ million) Cost Appraisal Estimate Actual Foreign Exchange Cost Local Currency Cost Total Financing Plan ($ million) Cost Appraisal Estimate Actual Implementation Costs Borrower-Financed ADB-Financed Beneficiary-Financed Total IDC Costs Borrower-Financed ADB-Financed Other External Financing Total ADB = Asian Development Bank, IDC = interest during construction.

8 iv 3. Cost Breakdown, by Project Component ($ million) Component Appraisal Estimate Actual Irrigation Development Community Development Rural Roads Rural Finance Institutional Support Physical Contingencies Price Contingencies Interest during Construction Total Project Schedule Item Appraisal Estimate Actual Date of Contract with Consultants a Contract with NGO 8 February 1999 Contract with Design Consultant 1 January 2000 Contract with Construction Supervision Consultant 2 October 2003 Contract with Baseline Survey Consultants 21 February 2000 Contract with Impact Evaluation Consultants 21 May 2005 Civil Works Contract Date of Award (first contract) 5 April 2003 Completion of Work (last contract) 30 November 2006 Equipment and Supplies Procurement Dates First Procurement 10 October 1998 Last Procurement 28 March 2006 Other Milestones Formation of First Men s SCO 1 February 1999 Formation of First Men s CO 12 February 1999 Formation of First Men s CTWG/WUA 23 February 1999 Formation of First Women s SCO 7 February 1999 Formation of First Women s CO 7 February 1999 CO = community organization, CTWG = community tubewell group, NGO = non-governmental organization, SCO = savings and credit organization, WUA = water users organization 5. Project Performance Report Ratings Implementation Period Development Objectives Ratings Implementation Progress 12 Jan Dec 2005 S S 1 Jan Feb 2006 S U 1 Mar Mar 2007 S S S = satisfactory, U = unsatisfactory

9 v D. Data on Asian Development Bank Missions Name of Mission Date No. of Persons No. of Person-Days Specialization of Members TA fact-finding 8 21 Mar a, m TA inception 7 11 Oct a TA review Nov a Loan fact-finding 27 Feb 14 Mar a, c, d, e, f, m Loan post-fact-finding 25 Apr 1 May a, c Loan appraisal 27 May 4 Jun a, c Loan inception 5 11 Jan g, e, l Special loan administration 19 Apr h, i Loan review Nov i, j Loan review Jan b, l Loan review Nov b, l Loan review Feb k, l Loan midterm review 25 Aug 8 Sep k, l Loan review Dec k, l Loan review Dec k Loan review Jun k Project completion review 12 Dec Feb k, l, m a = senior project engineer, b = senior project implementation officer, c = program officer, d = project specialist, e = project economist, f = social development specialist, g = rural development specialist, h = resident representative, i = senior program officer, j = senior control officer, k = project implementation officer, l = project analyst, m = staff consultant., TA = technical assistance

10 vi

11 vii

12

13 1 I. PROJECT DESCRIPTION 1. The Dera Ghazi (DG) Khan Rural Development Project (the Project) in Pakistan is in the DG Khan and Rajanpur districts of the central province of Punjab, along its southern fringe. At appraisal, rural poverty in the province was highest in the districts of DG Khan (with 24% of the rural poor in the province) and Bahawalpur (36%). The Project was designed particularly for the poorest rural populations in (i) the rain-fed plains, (ii) rain-fed hills, and (iii) canal-irrigated areas of these two districts. A survey done during appraisal revealed a rural poverty incidence of about 85% in the rain-fed plains and hills, and about 70% in the canal-irrigated areas. 2. The Project was intended to increase rural incomes and employment, and improve the quality of life of the population. Given the area s potentials, these objectives were to be achieved by (i) improving irrigation and road infrastructure, (ii) developing communities, (iii) providing financial services, and (iv) providing institutional support through organizational and skills training for beneficiaries in village communities. The Project had the following components and scope of work: (i) (ii) (iii) (iv) (v) Irrigation improvement:,(a) rehabilitating temporary diversions and lining works in perennial hill torrent networks, (b) developing 100 water users associations (WUAs) and improving 100 watercourses (WCs) in hill torrents, and (c) installing 100 tube wells and improving the related WCs in freshwater zones. Community development (CD): (a) mobilizing and strengthening 600 community organizations (COs); and (b) identifying, prioritizing, and implementing CD subprojects (to be done by the communities themselves). Rural roads: upgrading of about 175 km of rural roads; Financial services (FS): (a) mobilizing village savings and credit organizations (SCOs), (b) financing sub-loans, (c) mobilizing savings, and (d) providing training. Implementation support: consultancy services and incremental administration support for the purpose of (a) mobilizing and strengthening beneficiary WUAs, COs, and SCOs in the villages; and (b) designing and supervising the infrastructure construction and strengthening the project management unit (PMU). II. EVALUATION OF DESIGN AND IMPLEMENTATION A. Relevance of Design and Formulation 3. The objective of increasing rural incomes and employment, and improving the quality of life especially for women, was consistent with the new operation strategy prepared by the Asian Development Bank (ADB) for Pakistan in 1995, which was centered on human development and, above all, on poverty reduction and on the improvement of the status of women. In agriculture and rural development, ADB s strategy at appraisal was focused on rural development projects in more backward areas with high poverty, where better use of natural and human resources would lead to economic growth. The choice of the project area was dictated by poverty incidence and the geographic focus of ADB s sector strategy. The design was consistent with the strategic objectives of the Government of Pakistan economic and social development, greater private sector participation, improved agricultural production, more efficient use of increasingly scarce irrigation water, rural development, reduced poverty, wider coverage of basic infrastructure (including rural roads) for more accessible services, and environmental and social sustainability.

14 2 4. At completion, the Project s design and objectives were still relevant and consistent with ADB s country strategy 1, emphasizing sustainable pro-poor growth through rural development and infrastructure development. The CSP also stressed inclusive social development, particularly improvements in social service delivery systems. Accordingly, the Project was designed to encourage communities to participate in subproject implementation and operation. 5. The Government is now pursuing a well-designed, multi-pronged poverty reduction strategy to generate employment and build infrastructure in rural and low-income urban areas by increasing the access of the poor to basic services, microcredit, and social safety nets. The Project design was in line not only with this strategy but also with the Government s Medium-Term Development Framework , which is focused on the rural sector and specifically on enhancing the asset ownership of the poor, strengthening the nonfarm sector, promoting the participation of the rural population in development, improving rural financial markets, and developing the human resources of rural development institutions. 6. The components and activities of the Project, beneficiary selection, and stakeholder participation and consultation, as designed at appraisal, were all sound. A particular strength of the design was the emphasis on building the capacity of COs and increasing the participation of beneficiaries in project planning, design, and implementation. Decentralizing the management of operation and maintenance (O&M) and encouraging cost sharing by communities dramatically changed the development model for the area. The Project assisted in (i) surface and subsurface irrigation, (ii) WC lining, (iii) soil and water conservation, (iv) hill torrent management, and (v) livestock and agriculture productivity improvement. These activities helped develop and harness the area s agriculture, livestock, and water potential. The rural roads component gave farmers better access to markets and services. The FS component brought them economic empowerment through much-needed cash at affordable interest rates at their doorstep, with which they could buy good-quality inputs, sell their produce at the right time, and increase their profit margins. The implementation support component developed and strengthened COs and WUAs, and intensified networking, thus creating social capital and contributing to improved services by the line agencies in the communities. Capacity building and skills training of COs led to more jobs. The provision of drinking water, sanitation, and village infrastructure improved the quality of life of the village communities. Generally, all the project activities were in line with the needs of the area and served their purpose effectively. 7. The implementation arrangements at appraisal relied heavily on the ability of the Implementing Agencies (IAs) to implement the Project, without dedicated support or decentralized authority. During implementation it was realized that the IAs capacity had been overrated and their need for implementation support had been taken too lightly. The devolution of most IAs to the district level, under the Local Government Ordinance (LGO) of 2001, magnified the IAs lack of capacity. Similarly, the costs and size of certain subproject components had been miscalculated, and invalid assumptions had been made about the average size of the COs. To address these issues and to meet the beneficiaries targets as anticipated at appraisal, ADB made minor changes in the cost category allocations, implementation arrangements, and scope of the Project during implementation. Registering the COs mobilized for the Project as community citizen boards under LGO 2001 gave them access to more resources and ensured that the social capital of the Project was mainstreamed. 8. At completion, the design of almost all Project components remained relevant and consistent with current approaches to rural development. However, the success of the 1 ADB, 2002 Country Strategy and Program (CSP) , Manila

15 3 implementation was due primarily to the dedicated and relentless support of the Executing Agency (EA) and its PMU, assisted by a nongovernment organization (NGO), rather than the performance of the IAs (line departments). B. Project Outputs 9. The planned and achieved outputs of the Project are summarized in a table in Appendix 1 and discussed below. 10. Irrigation Improvement. This component comprised: (i) surface irrigation, and (ii) irrigation. The achievement of outputs under the surface irrigation subcomponent was only partly satisfactory. On the rain-fed plains, the Project was to support small-scale irrigation improvements, to increase the efficiency of water use at the Vehova and Kaha hill torrents, by (i) rehabilitating the irrigated command area of about 6,410 hectares (ha), and (ii) developing about 2,910 ha of unirrigated land. The appraised activities were: (i) minor improvements in temporary diversion headworks and structures at Vehova, and the strengthening of earthworks in the Kaha main channel; (ii) the partial lining of the main supply channel; and (iii) the formation of about 100 WUAs, and the improvement of about 100 WCs through the lining of one third (300 km) of their collective length. The improvement of the cross-drainage, diversions, and earthworks of the Kaha hill torrent brought an additional 4,854 ha of unirrigated land under the Kaha command. But only 10 WCs (30 km in total length) were lined, improving water flows to about 400 ha of land. Emerging water rights issues in Vehova, low demand for WCs, and lining works already carried out by the Punjab government on canals in Kaha before the Project led to a reduction in the scope of activities. The IA could not provide reasonable alternatives to Vehova, despite repeated consultations and requests by ADB. Project efforts at WC development were stopped in December 2004 at the request of the Punjab government to facilitate implementation through a national program, which followed a slightly different approach. Funds were allocated by the Punjab government for this program, and work on all the remaining 50 WCs in Kaha has been completed. 11. The achievement of outputs under the irrigation subcomponent was satisfactory. The support for irrigation in two major freshwater zones on the rain-fed plains was designed to involve (i) forming about 100 WUAs, (ii) installing 100 tube wells (DTWs), (iii) improving about 100 WCs, and (iv) monitoring the recharge. The achievements exceeded the original targets in several respects: 136 WUAs were formed, and they completed work on 136 DTWs and 40 WCs. Work on the remaining WCs has been completed with government funding, as explained in para. 10. Piezometers were installed in the tube wells to allow the monitoring of the groundwater level during implementation. No reduction in the water table was noted, and the Department of Irrigation and Power (DIP), the IA for this component, continues to monitor the water table. The irrigation component cost $3.7 million, compared with $5 million estimated at appraisal. 12. Community Development. Progress in achieving the outputs of the CD component was highly satisfactory. The Project was designed to respond to priority CD works identified and prioritized during the implementation by the communities themselves, with the help of a facilitating nongovernment organization (NGO). The activities in this component were grouped under five headings: (i) hill torrent management, (ii) soil and water conservation, (iii) agricultural production, (iv) livestock production, and (v) village and community infrastructure improvement. The activities, as designed, included mobilizing and developing the capacity of 600 COs averaging 25 members each (for a total of 15,000 members); 40% of these organizations were to be women s or mixed COs. At completion, more than 1,110 COs with an average

16 4 membership of 14 members each (for a total of 15,560 members) had been mobilized, and 25% were all-women COs. In total, 1,398 CD subprojects were completed at an average cost of $6,167, as against $20,000 envisaged at appraisal. As appraised, the CD subprojects and plans were prioritized and implemented by the communities concerned and facilitated by the NGO. Appendix 2 lists the activities in this component. About one third were categorized as productive investments and another third as social sector investments (provision of drinking water, sanitation, community halls, etc.). The cost of this component was $13.7 million, as against $17.2 million estimated at appraisal. 13. Rural Road Upgrading. Progress in achieving the outputs of the rural road component was highly satisfactory. The Project was designed to upgrade about 175 km of rural roads. At completion, 389 km of road had been upgraded (Appendix 3). The increase in scope was approved by ADB at the request of the Punjab government halfway through the Project, given the low road-to-population and road-to-area ratios in the target area, and the need to cover the appraised target population of 1.3 million. The construction specifications for the roads were upgraded under the Project to meet the higher specifications required for durability and minimum maintenance, as envisaged at appraisal. This subcomponent cost $23.4 million, compared with $14.3 million estimated at appraisal. The increase in cost was due mainly to the increase in the scope of the component. The cost per kilometer of road was significantly lower (by 28%) than the appraisal estimate. 14. Financial Services. Progress in achieving the FS component outputs was highly satisfactory. The Project envisaged support for the National Rural Support Program (NRSP) an NGO that was concerned mainly with the social mobilization of the poor, including women and the provision of FS through 1,000 village SCOs (half of which were women s SCOs) with a total membership of 25,000. The NRSP would be extended a credit line of $2.67 million, and more than 36,000 recycled sub-loans would be financed over a 5-year implementation period. The sub-loans were expected to average $150 each and to have a recovery rate of over 90%. Total outstanding savings would amount to $300,000 by the end of the Project. This component also included management training for about 2,000 SCO leaders (two per SCO) and skills training for about 4,000 other members. At completion, the NRSP had (i) mobilized more than 1,494 SCOs (34% of which were women s SCOs) with a total membership of about 25,000, and (ii) provided 110,000 sub-loans using a credit line of $2.65 million over a 7-year implementation period. The loans averaged $235 each and had a recovery rate of 99%; total outstanding savings amounted to $1.04 million. The NGO provided management training to more than 2,800 SCO members and skills training to another 12,500 members (Appendix 4). This subcomponent cost $2.8 million, compared with $3.1 million estimated at appraisal. 15. Institutional Support. The Project was designed to provide institutional support to the beneficiary WUAs, COs, and SCOs in the villages to enable them to prepare investment programs through participatory approaches, implement the Project, establish a monitoring and evaluation system, and update and maintain databases relevant to the system. The PMU envisaged hiring NGOs for about 800 person-months to set up and strengthen the COs, WUAs, and SCOs. Consultants to be hired in three packages for a total of about 60 person-months would also assist the PMU and the relevant IA in designing roads and supervising construction, establishing a benefit monitoring and evaluation (BME) system for the PMU, and preparing simple legal contracts between the IAs and beneficiary organizations. A total of 1,410 person-months of NGO services went into the formation and strengthening of 1,110 COs, 200 WUAs, and 1494 SCOs; 78 person-months of consulting services, into the design of roads and construction supervision; and 59 person-months, into the establishment of baseline data and the conduct of BME studies. A simple legal contract between the IAs and beneficiaries was

17 5 developed by the PMU without the help of consultants. The cost of this component was $4.6 million, as estimated at appraisal. Person-months had to be added because the scope of the services had been underestimated during appraisal and the CD and rural roads components had increased in scope. However, the overall cost of the subcomponent did not increase. 16. The above changes in the cost and the scope of individual components helped achieve the Project s objectives and cover the 1.3 million beneficiaries envisaged at appraisal. These changes had only a minor impact on the implementation schedule since effective measures were taken the number of PMU staff and logistical resources were increased to avoid delays in implementation. C. Project Costs 17. The actual project cost was $48.4 million, compared with $52.2 million estimated at appraisal. (The appraised and actual project costs are in Appendix 5.) The CD and irrigation components cost less than planned. The saving of $3.5 million in the cost of the CD component was a result of overestimation of the size and cost of the CD subprojects at appraisal, and insufficient community capacity to afford larger subprojects. The saving of $1.4 million in the cost of the irrigation improvement component was due mainly to a reduction in the scope of the surface irrigation subcomponent compelled by water rights issues, a partial overlap with other programs, and lack of demand for WCs. The rural roads component, however, cost more than anticipated at appraisal. The increase of $8.6 million in the cost of this component was due mainly to an increase in scope (from 175 km to 389 km of road upgraded) to benefit the Project s target population of 1.3 million. However, the cost per kilometer of road was significantly lower (by 28%) than estimated during appraisal. The institutional support and FS components, on the other hand, were completed without any significant variation in costs. Thus, in most cases (except for the irrigation component), there was an increase in scope and yet the Project was completed for less than the appraised cost. The changes in cost had a positive overall impact on the Project s economic and financial rates of return. D. Disbursements 18. No disbursement schedule was developed during appraisal. Disbursement was initially slow but picked up in 2002 and peaked in the last 2 years ( ) of the Project. The initial slow disbursement was due to a delayed start, without a full-time project director, and two separate investigations into allegations of mismanagement and financial malpractice. The overall impact of this delay was more than 3 years. Subsequently, minor delays were caused by the revision of the Project documents, (PC-I) 2, flaws in project design assumptions, and contractors unresponsiveness to the rural roads contracts in Delays were also caused by LGO 2001, in view of the lack of implementation capacity in the districts and, in some cases, the lack of interest on the part of the IAs in providing the envisaged design and implementation support for subprojects. Minor changes made by ADB in the scope of the Project and its implementation arrangements in September 2002, allowing the rural roads component to be divided into smaller procurement packages, improved implementation and disbursement. The actual disbursements and contract awards under the Project are listed in Appendix 6. 2 Planning Commission proforma I.

18 6 E. Project Schedule 19. The implementation of activities according to the Project s planned schedule was only partly satisfactory. (The appraised and actual implementation schedules are in Appendix 7.) Although the targets identified in the appraised schedule were realistic and practical, they were based on the (incorrect) assumption that the IAs could meet these targets on schedule (para. 8). The capacity of the IAs was further weakened after the introduction of the devolution plan in 2001, which transferred more responsibilities to IAs, in some cases without giving them the additional staff and resources required. Moreover, the project schedule, which had already been delayed by 19 months by management issues (para. 18), was delayed further by having to reestablish the Project s credibility in The situation started improving in 2002 and the Project showed major progress in 2003, after changes in the Project s implementation arrangements. The Project was completed in November 2006, within the revised implementation schedule, with an unavoidable delay of 23 months carried forward from F. Implementation Arrangements 20. The Punjab government s Planning and Development Department was the Project s EA, providing overall supervision and coordination through a PMU established in DG Khan. The project design included several IAs, in accordance with the capacity required for the various subcomponents. The IAs were (i) the Directorate of On-Farm Water Management (DOFWM), for the irrigation improvement component and the management of smaller hill torrents under the CD component; (ii) the DIP, for channel improvements under the irrigation improvement component and the management of larger hill torrents under the CD component; (iii) the Department of Agriculture, for the soil and water conservation and agricultural production activities under the CD component; (iv) the Department of Livestock, for the livestock production activities under the CD component; (v) the Local Government and Rural Development Department, for village and community infrastructure improvement activities under the CD component; (vi) the Communication and Works Department,(C&WD) for the rural roads component; (vii) NRSP, for the FS component; and (viii) the PMU, for the institutional support component. The implementation arrangements at appraisal were based on prior experience in several ADBfinanced area and rural development projects where having a relatively large number of IAs had proved workable and efficient as long as the components were designed in a straightforward manner and could be implemented independently. However, in the Project, where the capacity of district IAs varied from district to district, this approach did not deliver well. In remote districts with scarce human resources, IAs were hard put to retain high-quality staff without special incentives or administrative control over the staff by the district government. The Project design did not foresee the need for dedicated implementation staff for the IAs, except the DOFWM, and the consulting services were of insufficient volume to meet capacity gaps in the IAs. This issue was resolved during implementation by increasing the PMU s technical staff with the help of an incremental staff budget, to provide support to the IAs in design and technical supervision wherever gaps existed. However, such support also limited the IAs involvement, and hence defeated the objective of building project ownership and capacity in the IAs. G. Conditions and Covenants 21. The conditions of loan effectiveness were met and the loan was declared effective within 88 days of its signing. All the covenants were relevant and the Project s compliance with key

19 7 social, financial, environmental, economic, and other implementation loan covenants was deemed satisfactory (Appendix 8). Of the 49 loan covenants, 46 were met, 1 was partly met, 1 is still to be met, and 1 is not yet due (recommended follow-up actions are in para. 59). The loan covenant still to be met relates to the O&M of the roads completed under the Project; the IA has allocated enough funds in its annual O&M budget for this purpose. The covenant that was only partly met relates to the establishment and operations of the project coordination committee (PCC) at the divisional level, to review plans, address coordination issues, and resolve obstacles to implementation. The PCC was formed and met twice. After LGO 2001 was passed, the divisional tier of government was abolished and coordination and planning devolved to the districts. With the delay in the implementation of reforms under LGO 2001 and in the transfer of adequate numbers of staff to the districts, the project steering committee (PSC), with all the line departments, stakeholders, and district management represented, took over the role of the PCC. The PSC met regularly and performed this role efficiently. The PMU, for its part, reported on the progress and achievements of the Project to stakeholders every 15 days, to facilitate coordination. The PMU also submitted quarterly progress reports, audited financial statements, and the project completion report well within the time covenanted in the Loan Agreement, and the quality of these reports was generally acceptable to ADB. H. Consultant Recruitment and Procurement 22. Consultants were selected and goods and works were procured according to the Loan Agreement (schedules 4 and 5) and ADB s guidelines for the hiring of consultants and for procurement. In the first 2 years the EA had problems applying the ADB procedures and guidelines; however, after the first few contracts and with regular guidance from ADB, the hiring of consultants and the procurement of goods and works were undertaken efficiently for the rest of the project period. The consultants and the NGO were hired in four packages. The first package for 6 person-months of design services was awarded to an individual consultant since few firms had expressed interest, given the small size of the assignment and the remote location of the project area. The contract for 59 person-months to establish a BME system was awarded to the Punjab Economic Research Institute (PERI), through single-source selection (SSS) procedures approved by ADB at inception 3 again because of the lack of interest from private sector firms in providing this service in a remote area. For the third contract, a consulting firm was hired through quality- and cost-based selection to provide 72 person-months of construction supervision services for the rural roads component. The NGO s services for community mobilization and strengthening and the FS component were engaged through SSS procedures approved at appraisal. A contract for legal services was also awarded but later canceled as the standard contracts used for communities in similar projects were deemed sufficient reference models for the Project and there was no longer a need to outsource legal services. Variations were made in the NGO and construction supervision services contracts to accommodate the increase in scope and the delay in implementation. However, these changes did not have any cost implications: all the consulting services contracts were completed at a cost of $0.66 million, compared with $0.89 million envisaged at appraisal, and the NGO s contract was completed at a cost of $2.12 million, as against $1.96 million envisaged at appraisal. 23. The civil works contracts for rural roads were packaged into seven large road contracts at appraisal; another 35 km of road would be identified and tendered during implementation. 3 ADB Back-to-Office Report of the Inception Mission on the Dera Ghazi Khan Rural Development Project in Pakistan. Manila.

20 8 However, with the delayed start of the Project, most of the roads identified at appraisal (except for two packages) were taken up by the Punjab government for implementation. Under the changed scope approved by ADB, the PMU used the selection criteria approved at appraisal to identify a total length of 360 km of remaining and additional new roads. The roads were initially packaged into large contracts (around $1 million each). But in the surge of construction activities of national and provincial road authorities in the project area, few large contractors were available and fewer still responded to the advertisements for bids. The two larger contracts that were initially awarded were delayed by negligence and lack of interest from major contractors. In consultation with ADB, the Borrower divided the roads into smaller packages. A total of 96 small contracts were awarded, and most were completed within the contract period. I. Performance of Consultants, Contractors, and Suppliers 24. The performance of the design and construction supervision consultants was satisfactory. The consultants made staff available on-site and maintained their staff presence even under difficult conditions in remote areas. They could do so because they mobilized and trained local staff. The overall cost estimate for the services was $0.15 million, but $0.48 million was required to complete the work. The increase in cost was due mainly to the underestimation of costs at appraisal and an increase in project targets for the upgrading of rural roads. 25. PERI s BME performance was satisfactory. There were initial delays in collecting baseline data and producing interim reports on the impact assessment. However, after PERI made some changes in staff assignments, these issues were resolved and PERI was able to increase the size of the baseline sample according to the revised scope. PERI delivered an impact and completion report according to the revised schedule agreed on with ADB. This work was estimated at $0.71 million; it was completed for $0.13 million. 26. About 100 contracts were awarded. The overall performance of contractors was generally satisfactory. Small contractors tended to perform better than large contractors. Looking for a quick turnover and facing internal fund constraints, the small contractors were quick to mobilize and completed works on time. Almost all of the large contracts (above $0.5 million) were much delayed and produced a lower quality of construction than the small contracts. The awarding of the larger contracts itself delayed project implementation, and then after the contracts were awarded the large contractors were a major cause of delay in implementing works. Small local contractors also performed better because they understood the local culture and traditions better and built rapport with local communities. All additional land needed for widening roads and improving their alignment was provided voluntarily by stakeholder communities. Road contractors quoted highly competitive rates and rebuilt or improved 389 km of roads at an average cost of $59,000 per km against an appraisal estimate of $82,000 per km, for a substantial saving in the cost of construction. No major supply contracts were awarded. J. Performance of the Borrower and the Executing Agency 27. The performance of the Project s EA, the Planning and Development Department, was highly satisfactory. After an initial delay due to a tussle between PMU staff and the district administration, the EA was able to manage and supervise the Project very effectively. All policy decisions, compliance reviews, subproject processing, and approvals were made diligently and efficiently. The EA s PMU fulfilled its management and coordination targets, including the

21 9 procurement and management of contracts, implementation, and monitoring and evaluation of the Project s institutional support component. The performance of the IAs, on the other hand, was only partly satisfactory. Some IAs DOFWM (for irrigation), DIP (for WC development), the Department of Livestock (for livestock production), and the Local Government and Rural Development Department (for community development) lacked capacity and interest in the Project. The PMU had to mobilize its resources on most fronts for works that should been implemented by these IAs. (Implementation arrangements were changed during the Project s midterm review to strengthen the PMU s ability to take on the role of nonperforming IAs.) However, the NGO (NRSP) performed very well, effectively assisting the PMU and IAs in organizing and strengthening COs and facilitating the participation of COs and WUAs in the implementation of the Project. The NRSP proved flexible and provided stopgap support to many IAs with inadequate capacity. With its own resources, the NRSP was also able to continue the FS component during the period when Project activities were suspended ( ). Likewise, the performance of the C&WD, the IA for the rural roads component, was satisfactory. The C&WD managed to complete 389 km of road in the project area against the original target of 175 km within the extension period of 23 months, with very little external support. K. Performance of the Asian Development Bank 28. ADB s performance in administering the Project was satisfactory. No delays in the implementation can be attributed to ADB. Missions were conducted regularly, and most revisions, changes, and extensions were forestalled and approval was given well ahead of time. PMU management was suspended by the Government after major disputes between PMU staff and allegations of corruption at the start of the implementation. All project activities stopped. The delegation of the Project to the Pakistan Resident Mission in 2000 increased close and regular consultations between the Borrower and ADB. PMU staff acceptable to ADB were appointed in October 2001 and activities resumed. The Midterm Review Mission in 2004 identified gaps in the implementation arrangements related to the limited implementation capacity of the IAs and the need for a change in scope to cover the thinly spread population to be served by the Project. These issues were resolved during the mission through a change in scope and in implementation arrangements. The only other minor delays related to disbursement during the two extensions of the project (loan) closing dates, when funds could not be transferred to the imprest account. ADB acted promptly on both occasions to allow limited replenishments since both the COs managing contracts and many small private contractors (more than 80 with contracts of less than $0.1 million each) could not sustain any delays in payment by ADB. No disagreement with the Borrower on the terms of reference, bidding, or awards affected project implementation. 29. The Pakistan Resident Mission interacted regularly with the IA. Project staff were trained in project management and administration, including procurement, disbursement, and consultant hiring procedures. The EA and IAs received advice and guidance daily and during ADB missions a practice that was acknowledged and appreciated by the Borrower. III. EVALUATION OF PERFORMANCE A. Relevance 30. The Project remained relevant to and consistent with ADB s CSP and with the Government s long-term plans and strategies for rural development and the agriculture sector,

22 10 both at appraisal and at completion. Its overall impact shows that most of its appraised design aspects were relevant to and helped increase farm productivity, created employment, increased water use efficiency, and increased the incomes of small farmers (Appendix 9). 31. While the design of the Project was relevant, however, the notional targets for community schemes and road construction had been underestimated. The CO targets assumed an average CO membership of 25 per CD scheme. A membership of only 14 was observed at midterm not enough to meet the Project s target of benefiting 15,000 households. Similarly, the targets assumed at appraisal for improving rural roads were too low to serve the 1.3 million intended beneficiaries, given the project area s scattered settlements. At midterm, ADB increased the number of CD schemes and the targets for CO formation and rural roads to respond to the thin population spread, while still keeping within the overall cost of the Project and without compromising quality standards. These changes during implementation improved the Project s relevance and its capacity to reach the targeted population. B. Effectiveness in Achieving Outcome 32. The Project was effective in achieving its outcomes. At completion, the total value of production (net of production cost) from the additional hectares brought under cultivation in Kaha was estimated at $183,200, for an increase of $37.7 per ha in the yearly gross margin (Appendix 9, Table A9.6). The low gross margin was due mainly to seasonal variations in flow in the Kaha hill torrent and the related low cropping intensity in the area. The total annual value of production (net of production cost) from the 137 DTWs of about $5.1 million meant a yearly gross margin of $932 per ha in the irrigated plains, compared with the $300 per ha envisaged at appraisal. This significant increase in the estimated gross margin was due to the following: (i) the higher-than-expected cropping intensities achieved by farmers; (ii) the cultivation of high-value crops such as onions; (iii) the availability of loans through the lines of credit supported by the Project; and (iv) the Government s deregulation policy, which gave farmers higher compatible values for their outputs, and thus an incentive to grow high-value crops. 33. During the Project Completion Review (PCR) Mission, road users indicated that passenger fares had gone down by almost 50%, as envisaged at appraisal. The average passenger fare is equal to that charged to transport a 40 kilogram (kg) load. The PCR Mission estimated that, on the average, each household now saves $20 in passenger fares and $40 in cartage yearly (Appendix 9, para. 15), for a total of $60 almost three times the appraised estimate. The increase in savings is attributed to various factors: (i) the transfer of a larger share of the vehicle operating cost to consumers, (ii) competition in the transport sector made possible by liberal bank leasing facilities, (iii) a surge in economic activity from greater-than-expected development, and (iv) widespread demand for CD infrastructure. 34. The average increase in household income for all CD schemes was $21 per household, with the highest increase reported from soil conservation structures, which contributed an annual increase of $327 per household, against $25 estimated at appraisal for soil conservation measures alone. 35. In addition, the Project s investment in nonfarm community schemes and social services such as drinking water, street paving, sanitation, and community facilities has helped improve living standards. The provision of doorstep credit facilities has made it economically possible for farmers to grow high-value crops, buy better-quality inputs, and sell outputs on their own terms.

23 11 C. Efficiency in Achieving Outcome and Outputs 36. The Project proved efficient in achieving its planned outcomes and outputs. It was completed at 93% of the estimated cost and was able to achieve more than the appraised targets for all components (except the surface irrigation subcomponent). The delay of more than 23 months in implementation did not reduce the overall economic internal rate of return (EIRR), which was estimated at 33.5%, compared with 23.6% estimated at appraisal (Appendix 9). The methodology adopted during appraisal was also used in evaluating the project interventions at completion. At appraisal, EIRRs were estimated for the irrigation improvement and rural roads components but not for the community development, FS, and institutional support components. The EIRR for the surface irrigation subcomponent was evaluated at 23.0% against an appraisal estimate of 43.2%, despite the reduction in scope and the decision not to rehabilitate the head reach, where the additional area allocated for high-value crops (the main contributor to the high EIRR at appraisal) was located. The EIRR for the installation of DTWs was estimated at 39.0% at the design stage, and 57.8% at the end of the Project, when higher yields were assumed. The EIRR for the rural roads component was evaluated at 36.9% 3 percentage points higher than estimated at appraisal. The estimated combined EIRR for the above components, together with the cost of institutional strengthening, was estimated at 23.6% at appraisal, and a significantly higher 33.5% at completion. This suggests that the expenditure on institutional strengthening was cost-effective. 37. At the appraisal stage, no financial analysis, indicative or otherwise, was done for the interventions that were to be identified and managed by communities, such as income-generating physical infrastructure (cattle and poultry sheds, warehouses, small irrigation schemes, small wheat-flour mills and sawmills, and soil conservation structures). These were to be funded after their viability was assessed, during implementation. 4 Similarly, the Project also supported the development of community-managed social infrastructure such as community halls, dispensaries, nonformal school buildings, small drinking water supply structures and ponds, link roads and street soling, and culverts and bridges. Most of these interventions have proved beneficial. However, some interventions for which there was inconspicuous demand might take a longer time than anticipated to yield the desired returns. Cattle sheds, for instance, will require support services, including a network for milk collection and financial support for building herds, before they are fully used. Similarly, warehouses are currently underused because of the lack of marketable surpluses of grain, and of holding capacity for cash crops. Moreover, community organizations need to be strengthened further to manage collective inputs and market outputs. D. Preliminary Assessment of Sustainability 38. The Project s operations are likely to be sustainable. The Punjab government has shown strong ownership of the Project s interventions and has extended the Project s activities after completion, using its own funding and retaining all project staff and NGOs involved. 39. About 1,398 CD schemes were implemented by communities in the project area. These schemes were based on the priority needs of COs, which also contributed toward the cost of 4 National Rural Support Program (NRSP) Monitoring, Evaluation, and Research for the Dera Ghazi Khan Rural Development Project. Islamabad, NRSP

24 12 construction, showing their ownership. According to the completion report from PERI 5 and the observations of the PCR Mission, more than 95% of these schemes, including those that were completed more than 5 years ago, are fully operational. The COs were registered under LGO 2001 to ensure continuous utilization at capacity. 40. There are minor risks to the sustainability of the road infrastructure supported by the Project. The Punjab government has developed a formula under which all roads built under the Project receive an annual allocation of PRs33,000 per km for O&M (rather than the PRs20,000 provided in para. 35, schedule 6, of the Loan Agreement) to sustain the operation and economic benefits of these roads. The PCR Mission observed that, although the roads had been built to higher specifications and would require less maintenance, the districts had limited capacity to use O&M funds efficiently. To ensure the sustained O&M of the roads, regular follow-up by the district roads departments and reporting to ADB every 6 months was recommended. 41. The FS component is highly likely to be sustainable since the Punjab government has agreed to extend the services of the NGO for the next 5 years, after project completion. The rural finance market has also developed in the area and all mainstream microfinance institutions have opened branches to capitalize on economic opportunities. 42. The agriculture and irrigation infrastructure components are likely to be sustainable as well. The NGO has secured funding and will continue to assist farmers in networking and market connectivity in the medium term. Extension services in the project area, which started in 2006, will continue under ADB s Second Agriculture Sector Program Loan (ASPL-II) 6 through a provincial trust fund for the NGO and an outreach program for agriculture extension and research funded by ASPL-II. 7 In addition, policy reforms under ASPL-II are promoting greater private sector participation by introducing market-based mechanisms, and are also expected to encourage farmers to increase production and use farm and irrigation infrastructure effectively. The DIP has, moreover, instituted a system of allocations for O&M of irrigation systems in Punjab, including the Kaha perennial hill torrent system, thus ensuring funding support for the sustained maintenance of the system. E. Impact 43. As envisaged, the Project significantly reduced poverty in the project area. The impact evaluation studies carried out by PERI 8 show that the project interventions have successfully reduced poverty incidence by 16%, against the 10% target at appraisal. The PCR Mission estimated a $95 increase in the annual income of the average household with support from the Project, compared with $100 estimated at appraisal (Appendix 9, Table A9.10). 44. According to the PCR Mission, the improvements in enterprises and in farm and nonfarm activities have increased the demand for both permanent and seasonal labor. Agriculture and related activities alone employ about 2,112 more each year, compared with 1,000 anticipated at appraisal. The additional employment created during construction is estimated at 4,700 person-years (Appendix 9, para. 20). Annual employment due to the project activities is 5 Punjab Economic Research Institute (PERI) Economic Analysis of Various Components of the Dera Ghazi Khan Rural Development Project, Part 1. Pakistan, Lahore. 6 Loan No. 1877/1878-PAK[SF: Agriculture Sector Program Loan II (ASPL II), amounting to $350 million equivalent. 7 A condition for the release of the second tranche under ASPL-II. 8 PERI Economic Analysis of Various Components under the Dera Ghazi Khan Rural Development Project, Part 2. Pakistan, Lahore.