Independent Evaluation - Home > Search

Implementation Completion Report (ICR) Review - Rural Investment And Local Governance Project

1. Project Data:   
ICR Review Date Posted:
Project Name:
Rural Investment And Local Governance Project
Project Costs(US $M)
 62.20  175.60
L/C Number:
C3747, CH326
Loan/Credit (US $M)
 22.00  60.35
Sector Board:
Agriculture and Rural Development
Cofinancing (US $M)
 32.90  28.03
Board Approval Date
Closing Date
12/31/2007 12/31/2010
Sub-national government administration (30%), Central government administration (20%), General agriculture fishing and forestry sector (20%), Roads and highways (15%), General water sanitation and flood protection sector (15%)
Other rural development (33% - P) Other public sector governance (33% - P) Other environment and natural resources management (17% - S) Gender (17% - S)
Prepared by: Reviewed by: ICR Review Coordinator: Group:
Keith Robert A. Oblitas
George T. K. Pitman Soniya Carvalho IEGPS1

2. Project Objectives and Components:

a. Objectives:
The project development objectives stated in the Project Appraisal Document (page 2) were:

    “to contribute to rural development and poverty reduction through supporting provision of priority public goods and services at the communal level, as well as to promote good local governance through support of decentralized and deconcentrated participatory local governance systems at the commune and provincial levels.”

The statement of objectives in the Development Credit Agreement (page 19) is almost identical:
    “to assist the Borrower in its rural development and poverty reduction efforts through supporting provision of priority public goods and infrastructure at the commune level, as well as promoting good local governance through support of decentralized and deconcentrated participatory local governance systems at the commune and provincial levels.”

This Review uses the development objective described in the Project Appraisal Document.

b. Were the project objectives/key associated outcome targets revised during implementation?

c. Components:

Local Planning and Investment (Cost of $37.3 million estimated at appraisal. Actual costs at closure, including additional financing were $141.3 million.) This had two subcomponents:

    (i) Establishment of decentralized planning processes at village communes including development of five-year Commune Development Plans, three-year rolling Commune Investment Programs and annual commune budgets for implementing the investments; and
    (ii) provision of grants for small-scale commune level sub-projects for investment in infrastructure, identified and prioritized through local participatory planning processes in 15 provinces.
Policy Support and Project Management (Cost of $24.8 million estimated at appraisal. Actual costs at closure, including additional financing, were $34.2 million.) This had two subcomponents:
    (i) Strengthening - through funding technical assistance, training, logistical and operational support, and incremental operating costs - the capacity of commune and government institutions to implement the first project component; and
    (ii) provision of consultancies for strategic studies related to decentralization, and for monitoring of project implementation and performance.

d. Comments on Project Cost, Financing, Borrower Contribution, and Dates
Project Costs:

  • Appraised at US$62.16 million, final project costs grew to US$175.57 million. This was because: project coverage was expanded from 15 to 23 provinces to achieve a nation-wide scale-up (excluding Phnom Penh) and a larger investment program; a major typhoon (Ketsana) damaged many commune sub-projects which required rehabilitation; and the price of fuel and materials doubled due to the financial crisis. Financing:
    • The original credit was US$22.00 million and this increased due to appreciation of SDR against the US$ to US$24.34 million. At project closing US$23.98 million had been disbursed and US$0.10 million was cancelled. In mid-2007 Additional Financing as a grant of US$36.25 million equivalent was added to the project. By project closing the value of the grant had increased to US$37.83 million due to exchange rate appreciation. After cancellation of US$0.26 million, net disbursement of the grant was US$36.37 million.
    • There were three cofinanciers: UNDP, DFID and CIDA. First they contributed grants of US$13.31 million (against US$15.57 million planned) under the Partnership for Local Governance; second they contributed grants of US$14.72 million (against US$17.22 planned) under the Project to Support Democratic Development through Decentralization and Deconcentration.
    Borrower Contribution:
    • Initially the Borrower contributed US$31.46 million (against US$24.39 million planned). After additional financing was approved, the Borrower contributed another US$55.73 million (against US$74.73 million planned). Overall, the Borrower contributed 88% of its planned contribution.
    • At the time of additional financing in 2007 the project closing date was extended by three years to enable the expansion to the whole country.

  • 3. Relevance of Objectives & Design:

    a. Relevance of Objectives:
    High. Project preparation in the early 2000's was in the aftermath of 30 years of conflict that had resulted in a hardly functional governance structure at all levels, the need for reconciliation between the war-divided peoples, dilapidated infrastructure, and very high poverty, especially in rural areas where 90 percent of the population were classified as poor. The "Seila" program introduced by Government in the late 1990's was the premier program to build decentralized institutions and finance locally managed infrastructure and public services. It was considered by Government as the foundation stone to build democracy and improve welfare from the bottom up. The second phase of Seila, 2001-2005 provided technical and financial resources for: (a) refining decentralized planning, financing, and management systems for service delivery and local development; (b) providing discretionary budget support to province and commune authorities for investment in services and infrastructure; and (c) providing practically tested lessons for national level policy and regulations for decentralization, deconcentration and poverty alleviation.The Government's 2001 "Governance Action Plan" viewed decentralization, deconcentration and local governance as a means to further democratize the country and improve service delivery in the regions. The Cambodia National Poverty Reduction Strategy highlighted the cross-cutting linkages between decentralized governance, rural development and poverty alleviation.

    The project's objective to contribute to rural development and poverty reduction, and to promote local governance, are thus highly relevant to the "Seila" program and Government's National Strategic Development Plan for 2006-2011, now extended to 2013. The project's objectives were relevant also to Cambodia's Rectangular Development Strategy (2004-2008) that was based on good governance that included public administration, decentralization and partnerships with the private sector and civil society.

    The project remains relevant to the Bank's Country Assistance Strategies and sector work for Cambodia both at entry and at completion. The 2000 Strategy covered the FY00-FY03 period and had four key thrusts: (a) supporting good governance, through measures articulated in the Governance Action Plan, including decentralization of political authority and deconcentration of administrative responsibilities; (b) building physical infrastructure; (c) rebuilding human capital; and (d) facilitating private sector development. The World Bank report to the December 2004 Cambodia Consultative Group meeting, Cambodia at the Crossroads, took governance as its central theme. The 2008 Strategy covered the period FY05-FY08 and is still current in 2012. The 2008 Strategy emphasizes improved and more decentralized governance, rural poverty alleviation, rural development through participatory grass-roots processes, and investment in rural infrastructure.

    b. Relevance of Design:

    Project design presented a results chain that related project inputs and outputs to the expected outcomes of improved rural development and promotion of local governance. The causal links between the poverty alleviation objective and the project's design features was less clear. Benefits from new public facilities would affect all villagers and in that sense the project would contribute to poverty reduction. But there is little indication in the PAD of particular design features that would help enable the poor and women to be as well or better served by the sub-projects than the non-poor. Even so, construction of small-scale infrastructure chosen by the communes, whether transport, water supply, school classrooms or other investments, would be responsive to community needs and could be expected to improve rural welfare.

    Governance improvements paralleled the rural development activities, and the commune councils were expected to provide a base for decentralization at the village level, complementing the governance reforms and decentralization of activities taking place at provincial levels. Processes designed to tighten accountability and financial controls were also appropriate given Cambodia’s acute fiduciary management problems.The community processes by which infrastructure was selected and monitored during construction was expected to bring citizens together and improve investment quality. The inclusion of studies to evaluate the Seila program and formulate future policy and strategies for community-led rural development recognized the need to adapt and adjust the program as experience was gained.

    A shortcoming was that the project's exit strategy was not addressed at appraisal - specifically, how communities would maintain their infrastructure, and the measures and institutional arrangements needed to backstop this need.

    4. Achievement of Objectives (Efficacy) :

    To contribute to rural development and poverty reduction through supporting provision of priority public goods and services at the communal level, as well as to promote good local governance through support of decentralized and deconcentrated participatory local governance systems at the commune and provincial levels.”

    The project's efficacy is reviewed below in terms of the three objectives: (a) to contribute to rural development; (b) to contribute to poverty reduction; and (c) to promote good local governance at the commune and provincial levels.

    (a) To contribute to rural development: Substantial


    • Based on the Commune/Sangkat Administration Law passed in 2001, local Councils were established by direct election in 2002. Subsequently, the project provided capacity-building and technical backup to facilitate these councils to identify local needs and to invest annually in priority sub-projects, building skills and capacity through repetition. According to the ICR (page 7) the project only funded about about half of all the investments made by the local Councils, the balance coming from either government or other donor assistance.
    • 100 to 125 full-time locally recruited UNDP consultants worked to build the capacity of local subnational administrations to manage the subprojects and ensure that eligible projects were funded. In addition over 30,000 staff, elected councils and citizens were trained in various aspects of project planning and implementation.
    • All 1,545 communes participating in the project and subnational support staff were fully trained in the first year of implementation and received refresher training at the beginning of each new planning cycle.
    • Technical support officers made 10-13 visits to each sub-project during each project cycle, with the number of visits depending on the complexity and remoteness of the sub-projects.
    • 11 Executive Committee Building were constructed/renovated by 2008 and two were completed in late 2009.
    • Physical achievements were well above expectations at appraisal. Originally, about 1,110 communes in 15 provinces representing some 1.5 million households, were targeted under the project. With additional financing the number of communes was expected to increase to 1,473 communes in 23 provinces with around 2 million households receiving benefits. At the end of the project there were actually 1,545 communes in 23 provinces which benefitted from the project. These communes implemented 8,327 subprojects.
    • Subproject investments reflected the public goods demanded by the communes. The relative share of the number of subprojects by category was a follows:
          • 77% was for feeder roads. Transportation structures included 10.8 km of concrete/bitumen roads, 9,770 km of laterite/gravel roads, 4,965 km of earth roads,17,830 culverts and 511 bridges.
          • 15% was for irrigation, drainage and flood protection. Irrigation infrastructure included 1,343 km of canals, 234 dams/reservoirs, 233 drainage structures and 128 irrigation structures.
          • 4.5% was for water supply. Water supply infrastructure comprised 13,383 wells (all kinds), 4,905 piped water systems, and 367 ponds.
          • 2.5% was for education facilities. Some 500 school classrooms were constructed and 704 were refurbished. Health investments were nearly all for latrines (257 latrines).
    • Lack of proper maintenance of the investments is a general problem of the Seila program and has also manifested itself under this project. The ICR considers that, while minor maintenance can be done by the communities, larger maintenance should be handled by a "systematic maintenance program" supported by government. Further details on this key issue are not provided in the ICR. The Task Team subsequently stated that in 2012, road maintenance committees at District and Commune level are being rolled out as policy, and they have a mandate to seek the most appropriate local solutions. This initiative builds on the Commune Sangkat Fund Technical Manual (2009) which includes appropriate design standards for rural roads and well-designed materials explaining how to apply these standards to practical road design.

    • Communes increased their capacity to plan rural development, carry out procurement, manage and account for funds, report on progress of small, local development projects, and to begin to cooperate with local district and municipal government agencies. Even so, local preferences still provide a minority of local development projects. In 2008, for example, 60,000 commune priorities were submitted to District Integration Workshops and only 40% of those approved for further consideration represented commune-identified projects (ICR page 18).
    • Improved transport assisted rural development. Agricultural-related travel makes up two-thirds or more of total travel on village roads (ICR, page 24). The ICR reports that a 2009 Asian Development Bank socio-economic survey of project roads found that the frequency of road travel in the wet season for agricultural purposes increased from the baseline by 82%. The study also found that the average changes in primary income from selling rice doubled and for non-rice products tripled (though it should be noted that the project would not have been the only influence on these increases – the improvement in the general economy would also have provided a stimulus as market demand and general governance improved).
    • The Task Team stated the average number of motorbikes owned by rural households increased by 122% over the period 2004-2010. In the bottom quintile the increase was 400%, and the number of motorbikes per household in the second quintile was over 300%. This is an indicator that increased household incomes among the poorest quintiles increased use of local commune roads built by the project.
    • 90% of villagers confirmed that travel time was less dues to improved roads. 37% utilized saved time for farming and 17% for other activities.
    • Beyond the references above to frequency of travel and income from selling rice, there is little further data in the ICR on the outcomes of the other infrastructure investments. A study by Aruna in 2006 found, for example, that only 43% of irrigation projects financed by the project (7% of total project expenditures) were operational three years after their completion (ICR page 28).

    (b) Contributing to Poverty Alleviation: Substantial


    • There is minimal information in the ICR regarding the project’s contribution to poverty. Information that is available is of a proxy nature and is based on assuming an impact expected from a project output or other indication. For instance, it would be reasonable to expect that greater access by road to markets and urbanized areas would benefit the poor as well as the better-off. And women, who do much of the small-scale market transactions, might be expected to benefit at least as much as men. Similarly, the boost in incomes noted above would likely have been at least partly shared by women and the poor, although if land ownership is primarily in the hands of men and the wealthier populace, an often found situation, first round benefits would go primarily to these better-off groups.
    • There is no reference in the ICR to any specific project feature to facilitate better social inclusion for women and the poor.
    • The Task Team subsequently reported, based on the Cambodia Poverty Assessment 2004-2010 (World Bank 2010), that:
          1. Assuming only the poorest two quintiles (800,000 households) received the benefits from improved access, the total household investment through the project was 5.1% of the $582 average increase in household income (as measured by consumption) for the poorest two quintiles.
          2. The volume of rice production over the period 2004-2010 doubled and value of production increased 2.5 times during the period 2004-2010. The poorest two quintiles increased production the most (70%) compared with the wealthiest quintile (35%).
          3. Based on household consumption data, rural poverty in Cambodia declined from 53.2% in 2004 to 22.1% in 2009. In addition, the Task Team, utilizing empirical evidence from the Micro Panel Survey Data 2003/2008 and 2008/2010 (publication pending), stated that Commune/Sangkat Fund investment results in approximately 20% increase in welfare for Cambodian households in villages that received the benefits of the Fund compared to households in villages that did not receive the benefits. Second, disaggregated data shows that Fund investments in rural roads produced a greater percentage increase in the welfare index than irrigation and water supply – although this was differentiated to some degree according to provincial conditions (e.g. existing access to roads and ability to benefit from irrigation). Third, the impact was pro-poor: the welfare of the poorest quartile welfare benefited 36% more than the highest quartile.
          4. The average number of motorbikes owned by rural households increased by 122%. In the bottom quintile the increase was 400%, and the number of motorbikes per household in the second quintile was over 300%. This is not only an indicator of increased household incomes among the poorest quintiles, but of the increased use of local commune roads built by the project.
    (c) Contributing to Good Local Governance: Substantial


    • A total of 11,242 commune councilors and relevant government staff were trained in procurement.
    • Internal audit capacity was strengthened through the training of 150 provincial officers. A Project Information Database was established to support eligibility screening of subprojects.
    • All key information for the project's subprojects and implementation was posted on National Committee for Sub-National Democratic Development Secretariat's website with regular updates. This was supplemented by detail information of commune project information and progress accessible via a web-based database.
    • All bidding information is posted on the Secretariat website and on the notice boards of the relevant province and commune; however, there is better compliance in urban areas than in remote communes.
    • Accountability Working Groups to review complaints and provide remedies were established at national and subnational levels.
    • The Code of Conduct for staff was established with sanctions for fraudulent and corrupt activity by Secretariat staff at district and commune level, as well as for contractors, suppliers and consultants. The names of black-listed contractors engaged in collusive practices are posted on the official website.
    • Participatory management processes were introduced for all communes that were trained in the process.

    • The actions taken under the project to build a legal framework for decentralized government and community investment and management could be expected to strengthen local governance. The ICR reports that a study in 2010 on local governance found that the benchmark "local governance index" had improved during the project period by 11 percent, and that the community development approach had contributed to this improvement. However, the ICR comments (page 11) on a need for greater community participation - only 50% of commune residents participated in commune meetings - and that broader participation and more active oversight by the community of the commune councils would be desirable. Even so, about 30% of the heads of communes or commune councils were women (measured in the last year of the project).
    • Overall, however, the project has helped establish the base for local governance and community institutions that in most respects are contributing towards local development. Local governments are an active part of this and capacity improved through implementation of flow of funds, procurement and field supervision.The ICR also reports (page 11), based on a study in 2010, that local governance was strengthened through their handling of the commune fund. All communes carried out the new and participatory processes introduced by the preparation of a five-year Commune Development Plan; a three year rolling Commune Investment Program; and annual commune budgets. The Project Implementation Manual added more rigorous processes for procurement, and financing, and implementing the sub-projects enabled learning by doing.
    • Perhaps the best indication that the communities are functioning is the percentage of sub-projects that after completion were found eligible for project funding - 87% under fairly rigorous technical standards.

    5. Efficiency:

    At appraisal the economic rates of return (ERRs) by main types of investments were: 84% for roads, 98% for irrigation, and 14% for water supply. The project's overall ERR was estimated at between 64% and 74%. These latter ERRs depended on two different weighting assumptions: a weighting based on actual expenditures under the Seila program which yielded an ERR of of 74%, and a weighting based on expressions of consumer preferences based on a household survey which resulted in the 64% estimate.
    The ICR's overall ex-post economic rate of return is estimated as 18%, with rates of return by the project's main sub-project types of 25% for roads, 20% for irrigation and 10% for water supply. However, the ICR’s economic analysis estimates that, without maintenance, the ERR for roads would fall to 6% (page 28). The ICR also comments on maintenance issues with irrigation systems. While the worst case maintenance scenarios for these two investments (all of the investments not maintained) is unlikely, the ERRs would be lower if the reality of weak maintenance is factored in. These estimates are far lower than the ex-ante rates of return because the PAD calculations did not include implementation costs when aggregating sub-project ERRs.

    The Task Team updated the ex-post economic calculations and assumptions. noting in particular that the Abrams 2004 model only included a growth factor of a 2% annual increase in volume of all journeys to allow for population growth. However the Cambodian economy has grown overall at around 10% per year since 2003. The most important parameter is the value of time. Savings of wages account for 26% of benefits in the original model and savings in passenger time account for a further 19% of benefits. Value of wages and passenger time were derived from an approximate value of the agricultural unskilled daily wage which, in 2003, was $1 per day. The equivalent in 2012 prices is at least $4 per day, or allowing for overall price inflation of nearly 58% in the intervening years, and the present agricultural wage is about $2.50 per day at 2003 prices (increase by 150%). The updated economic rates or return are:

    • Leaving all other parameters in the model unchanged, and increasing the value of wages and time by this amount, say 150%, results in an ERR of 51% for roads with routine maintenance, and an ERR of 31% for roads without routine maintenance. A substantially lower 40% increase in value of wages and time improves the calculated ERR on rural roads with no routine maintenance to 12%.
    • Leaving all other parameters unchanged, increasing motorcycle traffic by 122% results in an ERR of 31% for roads with routine maintenance and an ERR of 11% for roads without routine maintenance. A 130% increase in volume of motorcycle traffic improves the calculated ERR rural roads with no routine maintenance to 12%.
    • Utilizing the same model but with 2012 data for both wages and time (increase of 150%) and motorcycle traffic (increase of 122%) the results are an ERR of 60% for roads with routine maintenance, and an ERR of 39% for roads with periodic rehabilitation.

    Procurement throughout most of the project period was difficult because stringent clearance processes and prior reviews (considered excessive by government officials and other stake-holders) created bottlenecks and delayed implementation. Despite this, about 95% of sub-project designs and 80% of sub-project construction were considered technically satisfactory (those that did not meet these standards were not financed, in effect excluding them from the project).

    Overall, efficiency is rated Substantial.

    a. If available, enter the Economic Rate of Return (ERR)/Financial Rate of Return at appraisal and the re-estimated value at evaluation:

    Rate Available?
    Point Value
    ICR estimate:

    * Refers to percent of total project cost for which ERR/FRR was calculated

    6. Outcome:

    Project objectives were highly relevant to Cambodia's post-conflict needs and the project's design was substantially relevant to achieving these objectives. The efficacy of the first objective to contribute to rural development was substantially achieved primarily through increased road usage that facilitated agricultural marketing. The efficacy of the project's second poverty alleviation objective objective is rated substantial. Strengthening the Seila program, the Government's principal program for development of rural communities, had strategic relevance well beyond the project's specific investment program and the the project, through its integration within the existing government institutions, helped strengthen governance of the national and local institutions involved. Thus the efficacy of the third objective is also rated substantial. Overall efficiency is rated modest.

    a. Outcome Rating: Satisfactory

    7. Rationale for Risk to Development Outcome Rating:

    • The project's main risk is the sustainability of the subproject investments. Inadequate maintenance of dirt roads is a problem, and unless the maintenance problem is resolved, development outcome is significantly at risk. A study cited in the ICR found that about only half of irrigation sub-projects were fully operational three years after the commune's initial investment.
    • With so many communes established, various anomalies can be expected at local levels, such as political capture and other vested interests, but the best counter to this is likely to be the power of commune members, with evident sway on local politicians. For this to be the general case, higher and more active community participation is an evident need.
    • The government needs to resolve the conflict of interest on which community projects they will finance. The 2008 Organic Law's intent is to ensure that lower capacity communes are not disadvantaged. A consequence of this is that completed community projects, rejected as not meeting the project's standard, are then financed by the government directly. There is clearly a need for additional technical support to these communities to ensure compliance with the new standards and guidelines.

      a. Risk to Development Outcome Rating: Significant

    8. Assessment of Bank Performance:

    a. Quality at entry:
    The project was effectively set up to pilot innovations which could be applied nation-wide. Amongst the innovations were tailor-made fiduciary management and community procurement procedures which are reported in the ICR (page 8) to have reduced malfeasance to minimum levels. Institutional and implementation arrangements were clear (e.g. PAD pages 10 to 13) and provided an adequate base for launching the project. Aiming to improve both local governance and local economic conditions at the same time as trying to establish a political consensus on full fiscal decentralization was ambitious. Additionally, there was no results chain or indicators to assess progress towards poverty reduction objectives, and monitorable indicators were primarily focussed on outputs, not outcomes (this was late remedied in 2009). More attention should have been given to strengthening institutions to ensure O&M, a problem well-known at appraisal.

    Quality-at-Entry Rating: Moderately Unsatisfactory

    b. Quality of supervision:
    Particularly noteworthy was the adaptable approach used by the supervision team - as experience was gained, project processes and features were pro-actively adjusted. However, some of the Bank procedures, though well intentioned, became roadblocks. There were strident complaints by stakeholders concerning Bank prior reviews. Each province had to submit documents at four stages in procurement for review and no other communes could progress until the first commune(s) received clearances. In many cases sub-projects were only cleared for implementation at the start of the rainy season and then implementation had to be delayed until the dry season, the end of the government's fiscal year, resulting in delays of 3 to 4 months. While this was clearly not efficient, this reflected the Bank team's appropriate concern regarding fiduciary management. However, a better way to safeguard against procurement irregularities might have been found (prior reviews were reduced later in the project period.) Appropriate to Cambodia's fiduciary weaknesses, financial management and procurement processes were closely followed, and a specialist Fiduciary Review was undertaken in the project's first year.

    The decision for Additional Financing to expand the project nation-wide was appropriate given that the project was proceeding positively and could be readily scaled-up. The supervision team improved monitoring capacity by agreeing with Government that project funds could be used to hire additional oversight consultants. This benefitted the project and also provided a source of information and oversight for the Bank. However, the opportunity was missed at this time to focus efforts on capacity-building to ensure operation and maintenance of project-financed infrastructure.

    Quality of Supervision Rating: Moderately Satisfactory

    Overall Bank Performance Rating: Moderately Satisfactory

    9. Assessment of Borrower Performance:

    a. Government Performance:
    Government was consistently committed to Seila and the project, carrying through a major program and making adjustments to the program as experience was gained. Government maintained good coordination among a number of ministries and between the various donor agencies. Government was also responsive to Bank recommendations during project preparation and supervision. A major Government achievement was the passing of legislation and strategy papers to backstop the project, and Seila generally; amongst these - the 2001 law establishing local councils, the Governance Action Plan, the National Poverty Reduction Strategy, and at the end of the project the National Program for Sub-national Democratic Development, which helped set the stage for a continued decentralization program. Overall, Government performance was generally good, except for one year, when provision of counterpart funds was a significant shortcoming, albeit relatively short-lived.

    Government Performance Rating: Moderately Satisfactory

    b. Implementing Agency Performance:
    The coordinating Seila Task Force and National Committee for Sub-national Democratic Development did well in steering implementation of the project - the sheer magnitude of the task (involving 23 provinces, 1,500 communes and training of 30,000 government officials and advisors) attests to the energy and oversight that they brought to the project. Processes for more citizen participation in council deliberations, commune planning and investment decisions could have been promoted more intensively. In a notably weak fiscal environment the Committee closely monitored contractors for fraud and collusive practices, and ensured that sub-projects involving corruption. Over the life of the project, collusion was found in 7 cases, with 15 cases of procedural deviation (e.g. bidding and administrative) and 3 cases of physical non-compliance (e.g. road not built to specification) or procedural deviation. For these 25 cases, Bank financing was returned. Initial weaknesses in accountability at province and commune levels were addressed vigorously (including use of accountability working groups, complaints boxes, and publically available information). Accountability improved, especially over the last year of the project. However, some actions of the implementing agency - financing rejected projects using government funds - undermined the overall impact and commitment to the procedures to ensure investment quality put in place by the project.

    Implementing Agency Performance Rating: Moderately Satisfactory

    Overall Borrower Performance Rating: Moderately Satisfactory

    10. M&E Design, Implementation, & Utilization:

    a. M&E Design:
    At appraisal, while monitoring was seen as primarily a tool for measuring project processes and outputs (procurement, status of sub-projects, community participation, training held, etc.) some impact assessment was also envisaged, through a baseline socio-economic survey with follow-up annual surveys. None of the monitorable indicators related to poverty alleviation, and there was almost no attention to outcomes for any of the project objectives.

    b. M&E Implementation:
    The M&E program covered all 23 provinces, and some 125 consultants were involved. This provided a strong base for the monitoring program, but information was primarily geared to project progress, the quality of project implementation, and outputs. This provided a good information system for project management, but findings on rural development and poverty alleviation received much less attention.

    a. M&E Utilization:
    The M&E system’s limited attention to project outcomes notwithstanding, project data reflecting progress, quality of construction work and efficiency of processes was a useful management information system which was utilized to improve project implementation.

    M&E Quality Rating: Modest

    11. Other Issues:

    a. Safeguards:
    The project was categorized “B” under OP4.01 Environmental Assessment. OP4.20 Indigenous Peoples and OP4.12 Involuntary Resettlement were also invoked at appraisal. The ICR (page 7) reports that the screening process used during sub-project preparation was satisfactory and well implemented, and provided a workable tool for identifying and mitigating environmental and social impacts. Impact studies indicated no significant problems with implementation in provinces with populations of indigenous people. A 'watch-list' was set up of communes that were likely sensitive with respect to environment and indigenous people. In 2004, the project was awarded the Bank’s "Green Award” for innovative mainstreaming of environmental and social safeguards. According to the ICR (page 7-8), "Although no serious problems were identified related to compliance with the safeguard policies triggered by the project, during the preparation of the additional financing, some concern was expressed regarding land acquisition and resettlement, which had been problematic in Cambodia in general. As such, an additional level of caution was deemed prudent. Three regional safeguard advisors were recruited to improve review and clearances, and an NGO was engaged to pilot a campaign for improving land awareness issues. The primary issue concerned land right-of-way and land acquisition (by donation in most cases), and the main concern was with documentation, ensuring that all affected households donated land voluntarily without coercion or disadvantage. The improved processes are now in place."

    Community participation could have been better. On average about 50 percent of commune members attended committee meetings, and attendees tended to be more passive than proactive participants. There were also cases where council and local government officials excessively influenced community decisions. Nevertheless, the ICR (page 11) comments that the NGO community was generally satisfied with community processes, suggesting that community influences and results were mostly satisfactory. Women appear to have been well involved in the sub-projects, although they tended to be less proactive participants than men. Nevertheless, attendance rates at commune meetings were higher for women than for men, perhaps auguring well for future gender involvement.

    b. Fiduciary Compliance:
    Fiduciary: According to the ICR (page 8): The overall risks for procurement and financial management were rated "high" due to the perceived fiduciary risks in Cambodia in general and because the project was implemented at the local level in the main through new local councils across the country. In a risky and collusive environment, with upward of 4000 officials involved, the process introduced for the management of the Fund, involved rigorous reviews, and the project’s disbursement mechanism ensured that project funds were used for intended purposes and safeguards were complied with. Misprocurement occurred in 2005, for a few sub-projects and the Bank funds were refunded. When policies and procedures were not complied with, sanctions were levied and adherence to procedures improved over time. Project procedures and enforcement mitigated the high risks for Bank-financed sub-projects, and consequently few serious problems were encountered during implementation.
    The annual cycle of prior and post review allowed the National Committee for Sub-National Democratic Development-Secretariat to manage procurement problems effectively; it was not necessary to suspend disbursements as problems were addressed immediately. Project monitoring and supervision, special studies such as the study on budget execution in 2008, and Integrated Fiduciary Supervision Assessment carried out by the Bank in 2009, allowed improvements to be made to manuals, training and practices. Post reviews showed significant improvement in practices and commitment to change was illustrated when the Secretariat commissioned a Small Contractor Study in 2010 to better understand the processes of collusion in awarding project contracts. The Secretariat has signaled their intention to keep these improved, rigorous processes in place. However while these intensive procedures are applauded by some, others request simplification. The start of the 3-year Implementation Plan provides an excellent opportunity for a review of procedures and their implementation without the Bank's direct involvement, allowing an assessment of capacity built over the last seven years and consistency in compliance.

    There is no comment in the ICR on whether audits were delivered and whether they were qualified or not. The Task Team subsequently indicated that the Audits for the period 2005-2007 were qualified. There was a need to improve the internal control of Local Contribution system; the documentation system at provincial level; and introduce a time-bound action plan to rectify these problems. This was done and the Audits for the period 2009-2011 were unqualified.

    c. Unintended Impacts (positive or negative):

    d. Other:

    12. Ratings:

    IEG Review
    Reason for Disagreement/Comments
    Moderately Satisfactory
    New evidence provided by the Task Team after the ICR based on the Cambodia Poverty Assessment, 2004-2010, indicates substantial achievement on all four objectives and substantial efficiency. 
    Risk to Development Outcome:
    The problem of poor maintenance of the some road and irrigation infrastructure has not yet been resolved. Similarly, commune construction standards are not fully harmonized with the most up-to-date techniques and standards. 
    Bank Performance:
    Moderately Satisfactory
    Inadequate attention to arrangements for operation and maintenance of project-financed infrastructure.  
    Borrower Performance:
    Moderately Satisfactory
    Moderately Satisfactory
    Quality of ICR:
    - When insufficient information is provided by the Bank for IEG to arrive at a clear rating, IEG will downgrade the relevant ratings as warranted beginning July 1, 2006.
    - The "Reason for Disagreement/Comments" column could cross-reference other sections of the ICR Review, as appropriate.

    13. Lessons:
    The ICR draws the following lesson:

    1. Integrating with the existing government institutions and processes can increase political commitment and help lever the project's influence on improving a national program. This well recognized principle is reinforced by the project's experience. The project chose to operate integrally as part of the government institutional system and the improvements introduced through the project are being gradually adopted in the national program.

    In addition, IEG derives the following lessons from the ICR:

    2. With effective training of government staff and community leaders, and technical back-up during implementation, communities can assume significant responsibilities in rural investment. The project vested a greater role for communities than often applied in community-driven development. The community planned and implemented the investments, did its own procurement following Bank procedures, had to follow stringent fiduciary management guidelines, and, facing the project's ex-post financing system, shouldered the risk that unless their investment was at completion assessed as good quality, they would not receive financial assistance. The project achieved this from a modest initial capacity. At the beginning of the project, there was little community investment experience in government and in the rural populace, and major training was required to surmount this.

    3. Construction of rural infrastructure needs to be accompanied by arrangements and incentives for subsequent maintenance. While it is difficult to find workable arrangements and incentives after the sub-project investment period is over, there is little indication that an effort was made to find practical ways of encouraging community interest in maintenance. Yet maintenance of dirt roads and irrigation, the most prevalent investments, was declining.

    4. Project features to encourage broad-based participation in community sub-projects should be sought. About half of the rural households participated in sub-project decisions, and most of the participants were passive rather than active contributors. Greater participation would likely have increased the ownership in the project, and perhaps resulted in a greater commitment to maintenance.

    5. Assessment of a project’s achievements needs to focus on achievement of objectives rather than be restricted to project outputs; and project design, monitorable indicators and M&E need to reflect this. Monitoring was output based – there is no information on the projects second objective (poverty alleviation); and very little discussion at design stage and at project completion on how the project would contribute to this objective.

    14. Assessment Recommended?

    To gauge the effectiveness and replicability of the project's innovations in community financing (post-investment provision of funds for quality investments, procurement by communities, fiduciary controls); and to examine the poor maintenance and other implementation issues as a learning vehicle for the scale-up of Cambodia's program; and as a contribution to the Bank's overall experience in community driven development.

    15. Comments on Quality of ICR:

    The ICR is a concise, well written report and the structure is consistent with the ICR guidelines. The discussion and analysis is evidence based and of good quality - although good outcome evidence is missing (a fault of the project's M&E systems). The lessons are thoughtful, and the commentary related to each is informative and convincing. The analysis in earlier parts of the report provides factual discussion supporting the lessons. Discussion in the report is results-oriented when the whole report is considered, but would have been significantly improved if a short discussion at the beginning of Section 3.2 (Achievement of Project Development Objectives), assessing outcome by each objective, had been included. The report is light in the discussion of poverty reduction and O&M issues which have been a problem area under this project. This objective, even though lack of data would have constrained quantification, could have been discussed in a qualitative assessment, with recourse to proxy indicators as available. The ICR does not adequately discuss the project's compliance with Bank fiduciary policies and does not also state if the audits were qualified or unqualified.

    a. Quality of ICR Rating: Satisfactory

    © 2012 The World Bank Group, All Rights Reserved. Terms and Conditions