You are on page 1of 45

Analyzing the sustainability of community driven development in Afghanistan using the National Solidarity Program case study

Faheem Merchant MPhil in Engineering for Sustainable Development August 2010

Supervisor: Dr. Heather Cruickshank Dissertation Submission to Department of Engineering University of Cambridge

Page 2

Analyzing the sustainability of community driven development in Afghanistan using the National Solidarity Program case study Abstract

In the same way that microfinance initiatives provide resources to individuals and allows them the decision making freedom for the best projects to pursue, community driven development (CDD) provides small villages with funding to develop communities as free decision making bodies responsible for their own development. The National Solidarity Program (NSP) is Afghanistan‘s community driven development initiative. Although the NSP is recognized as a successful programme, the long term sustainability of NSP has not been understood and this research investigates the NSP using many different characteristics including cost, time, funding, maintenance, lifespan and resiliency, insurgency attacks, skills development, dependency on government, community skills development, transparency and social cohesion. The sustainability of the NSP is a key focus of future World Bank investment in Afghanistan and the results can potentially be used by other countries considering their own CDD implementation in the future.

Page 3

Introduction

COMMUNITY DRIVEN DEVELOPMENT

Community driven development (CDD) is a development initiative that provides control of the development process, resources and decision making authority directly to community groups. The underlying assumption of CDD projects are that communities are the best judges of how their lives and livelihoods can be improved and, if provided with adequate resources and information, they can organize themselves to provide for their immediate needs. Moreover, CDD programmes are motivated by their trust in people (Naidoo and Finn, 2001) and hence it advocates people changing their own environment as a powerful force for development. By treating poor people as assets and partners in the development process, previous studies have shown that CDD is responsive to local demands, inclusive, and more cost-effective compared to centrally-led NGO-based programmes. CDD can also be supported by strengthening and financing community groups, facilitating community access to information, and promoting an enabling environment through policy and institutional reform (Dongier, 2002). CDD projects work by providing poor communities with direct funding for development with the communities then deciding how to spend the money. Lastly, the community plans and builds the project and takes responsibility for monitoring its progress.

Following from this description, field practitioners at the World Bank have denoted five key characteristics of CDD projects. 1. A CDD operation primarily targets a community based organization or a representative local council of a community. This community focus means that the essential defining characteristic of a CDD project is that the beneficiaries or grantees of implementations are agents of the community. Since the focus on small communities is so large the CDD normally targets small scale subprojects in the community. 2. In CDD operations, community or locally based representation is responsible for designing and planning the subprojects in a participatory manner. Since the concentration on participatory planning is considerable in CDD operations, often the possible types of subproject investment options are very large with only a small list of subprojects that cannot be carried out. 3. The defining characteristic of CDD projects is that a transfer of resources to the community occurs and control of the resources is delegated to the community. The amount of transfer and control of resources will depend on the CDD implementations. Page 4

4. The community is directly involved in the implementation of the subproject. Often the participation of the community comes directly in the form of labour or funds. However, the community may also contribute to the subproject indirectly in the form of management and supervision of contractors or the operation and maintenance of the infrastructure when complete. 5. An element of community based monitoring and evaluation has become a characteristic of CDD subprojects. Most often it is social accountability tools such as participatory monitoring, community scorecards and grievance redress systems which allow for the community to ensure accountability of the CDD implementation.

CDD vs. CBD

Community driven development is derived from community based development (CBD) which can include a much broader range of projects. For example, CBD projects can include everything from simple information sharing to social, economic and political empowerment of community groups. However, CDD projects fit on the empowerment end of CBD by actively engaging beneficiaries in the design, management
Figure 1 – Community driven development is a subset of community based development.

and implementation of projects. The stress on actual control of decision making and project resources at nearly all stages of a subproject cycle distinguishes CDD from the previous generation of CBD projects. In this continuum of community participation covered by CBD, new-generation CDD projects are located at the extreme right of the axis as shown in Figure 1 (Tanaka, 2006).

Since community driven development has only recently diverged from the broad community based development there are a few contrasts visible in the five characteristics of CDD programmes. In essence, all five properties of CDD projects exist together only in the newer generation of CDD implementations. Nevertheless, the first attribute of community focus would apply to all CDD projects and CBD projects. In contrast, the second characteristic of participatory planning and design and the fourth property of community involvement are often visible among all CDD projects but very rarely in CBD projects. Moreover, community based monitoring and evaluation which is the fifth aspect of CDD projects is only found in the some of the newer projects. The fifth characteristic is what positions many of the newer CDD projects in the extreme right of the CDD cluster as diagrammatically demonstrated in Figure 1. As mentioned above, the third characteristic of Page 5

community control of resources seems to be the key factor to conceptually distinguish between CDD and CBD projects. However, many of the early NGOs implementing CDD projects did not always interpret this factor rigorously (Tanaka, 2006). Thus, the distinction between CDD projects and CBD projects with CDD components was not always clear; however, this would be expected since there was a gradual evolution of CDD out of CBD.

To alleviate the earlier problems of overreliance on central governments as the main service provider, CDD programs were launched by the World Bank to improve the accountability and services in key areas. However, NGOs quickly learned that well designed and implemented CDD programmes had ripple effects of promoting equity and inclusiveness, efficiency and good governance. By effectively targeting and including the vulnerable and excluded groups, as well as allowing communities to manage and control resources directly it was evident that CDD programs could allow poverty reduction projects to scale up quickly. Efficiency is gained through demand responsive allocation of resources, reduced corruption and misuse of resources, lower costs and better cost recovery, better quality and maintenance, greater utilization of resources, and the community‘s willingness to pay for goods and services. Good governance is promoted by greater transparency, accountability in allocation and use of resources because the community participates in project decision-making processes. Some of the principles of CDD—such as participation, empowerment, accountability, and nondiscrimination—are also worthy ends in themselves (Asian Development Bank, 2008).

It was as early as 1881 when T.H. Green who wrote about the maximum power for all members of human society alike to make the best of themselves (Zakaria, 1999). However, it was not until the 1970s with John Rawls’ book ―A Theory of Justice‖ and in the 1990s with Amartya Sen‘s book ―Development as Freedom‖ where the notions of substantive freedom and the multidimensional nature of poverty were made explicit to the multilateral development banks. This recognition of the multidimensional nature of poverty as well as the combined failures of both markets and governments and the socio-political complexity of ground level realities has made it clear that relying on traditional top-down, state-led, ―big development‖ strategies would not be effective to combat poverty. Moreover this resurgence in participatory development and bottom up approaches in the NGO and development sector has come in only the last two decades as explained above.

EXPANSION OF COMMUNITY DRIVEN DEVELOPMENT

Page 6

Since the mid-1990s, community driven development has emerged as one of the fastest growing investments by NGOs, aid organizations and multilateral developments banks. This continued investment in CDD has been driven mostly by a demand from donor agencies and developing countries for large-scale, bottom-up and demand-driven, poverty reduction subprojects that can increase the institutional capacity of small communities for self-development. The success and scale of some CDD projects in the World Bank are especially notable. The World Bank supported approximately 190 lending projects amounting to $9.3 billion in 2000–2005 (Tanaka, 2006). Initiated by the International Development Association (IDA) at the World Bank, CDD projects have been instrumental in harnessing the energy and capacity of communities for poverty reduction. Since the start of this decade, IDA lending for CDD has averaged annually just over 50 operations, for an average total of US$1.3 billion per year (International Development Association, 2009).

Even the Asian Development Bank (ADB) has funded 57 projects worth about $2.5 billion between 2001-2007 that included community driven development approaches to enhance deliver of inputs and beneficiary participation. They constituted 14% of the total loans approved by the Asian Development Bank during this period. Over one-third of the projects were in the agriculture and natural resources sector, followed by a smaller proportion of water supply and sanitation, waste management, education and health projects. The projects were primarily in Southeast Asia, South Asia, and Central and West Asia, where the developing country governments were investing in rural development programs (Asian Development Bank, 2008)

In the last few years the International Fund for Agricultural Development has been working with the Agence Française de Dévelopement (AFD), the African Development Bank (AfDB), the European Union (EU), the Food and Agriculture Organization of the United Nations (FAO), the UN Capital Development Fund (UNCDF) and the World Bank to create a platform for learning and sharing knowledge on community driven development (International Fund for Agricultural Development, 2010). Intensive forms of community participation have been attempted in projects of several donors for many years. Bilateral donors, such as the Department for International Development (DFID) of the United Kingdom and the Canadian International Development Agency (CIDA), have used CDD-type approaches for a long time as part of their sustainable livelihoods and integrated basic needs development assistance in developing countries. The Swedish International Development Agency (SIDA) and Danish International Development Agency have used CDD principles in the mandate of a rights-based approach to the development projects they fund (FAO, 2010). Page 7

As described earlier, most of the multilateral development banks have large community driven development programs, however, this paper will focus on the World Bank CDD programs. More than 80 countries have now implemented CDD projects. The breadth and activities funded by the CDD programs at the World Bank can be explained by providing a brief overview of a few of them.

The Second National Fadama Development Project II (NFDP-II) targets the development of small scale irrigation, especially in the low-lying alluvial floodplains or "Fadama‖. NFDP-II increased the productivity, living standards and development capacity of the economically active rural communities while increasing the efficiency in delivering implementation services to an estimated four million rural beneficiary households and raising the real incomes of households by 45 percent (African Development Bank, 2003). The Social Fund for Development in Yemen provided support 7 million people of which 49 percent were female and generated 8,000 permanent jobs. It also increased the number of girls‘ schools from 502 to 554 and basic education enrollment rates from 63 percent to 68 percent. The program focuses on helping the poor to help themselves through providing incomegenerating activities and building community infrastructure rather than making cash transfers (El-Gammal, 2004). The Social Investment Fund Project V in Honduras benefited 2.5 million people with the implementation of 2,888 projects (1,446 rehabilitated schools, about 700 new schools, 163 new health centers, 347 small water/sanitation systems, and 461 latrines) resulting in all children in the targeted areas attending primary school. In addition the project communities were provided with better access to health care assistance and access to running water (Perez de Castillo, 1998). The Andhra Pradesh Rural Poverty Reduction Project (APRPRP) in India has help to organize 10.1 million rural poor women into community based organizations that collectively save over US$770 million and leverage credit over $2.7 billion from commercial banks (World Bank, 2003). The Kecamatan Development Program (KDP) in Indonesia which is what the National Solidarity Program in Afghanistan is based on has benefitted 18 million people by providing better services which include more than 37,000 kilometers of local roads and 8,500 bridges, 9,200 clean water supply units, and 3,000 new or improved health posts. In addition, more than 1.3 million people obtained loans to start or complement local businesses through micro-financing (Guggenheim, 2004). Lastly, the National Solidarity Programme (NSP) in Afghanistan will be the focus of this research. In this implementation elected village-level community development councils, which include women, use grants and local labor to rebuild bridges and roads, fix schools and install water pumps to benefit 13 million people across Afghanistan thereby building state credibility and strengthening local democracy. Page 8

Literature Review of CDD Sustainability

While success stories are mounting, there are limitations, challenges, and risks that need to be considered carefully when developing and implementing CDD projects. Despite the success of CDD programs and rapid increases in support, the approach still has critics. A recent evaluation of community-based development (CBD) and CDD in the World Bank was particularly critical about the shortcomings of these approaches. Some of the main criticisms were that CBD and CDD projects undertaken by the World Bank lack sustainability, are prone to elite capture, create parallel institutional structures, require special guidance on fiduciary and safeguard compliance, lack effective evaluation of poverty impact, and are costly to prepare (World Bank, 2005).

Another major criticism of CDD projects has been the scalability aspect of the program. For example Indonesia‘s Kecamatan Development Program (KDP) encountered difficulties scaling up transparency, accountability and participation (Gillespie, 2004). In Indonesia, the KDP encountered several challenges when attempting to scale up empowerment and sustainability (Biswanger, 2003). Moreover scaling coverage and scaling organizations have been major concerns and discussed widely in the literature, but the sustainability of coverage and sustainability of organizations has not been addressed (Gillespie, 2004).

It is amply clear that there are many issues with the current models of CDD instituted by the World Bank as described above. However, this paper will be focused on the sustainability aspects of CDD. It will concentrate on the sustainable development of CDD once the program has been implemented. There are currently various definitions of sustainable development, however, this paper will utilize the widely accepted definition of ―sustainable development‖ put forward by the World Commission on Environment and Development (WCED) which is as follows: ―Sustainable development is development that meets the needs of the present without compromising the ability of future generations to meet their own needs‖ (WCED, 1987).

Although this definition would suffice for a general understanding of sustainable development, the question still outstanding is: what does sustainable development mean in the context of community driven development. Recent interviews conducted with staff at the World Bank revealed a wildly different understanding of sustainability in the context of CBD and CDD. For some it implies sustainability of community processes, for others CBD/CDD is simply a means to an end, and sustainability is related to infrastructure investments. For still Page 9

others, it is related to the overall resource allocation, including support for decentralization (World Bank, 2005).

Although this paper will look at sustainability from several perspectives, it will first outline the current research conducted and conclusions drawn regarding sustainability of CDD in governance, resources, social cohesion, maintenance and infrastructure. It will then go on to outline the gaps in sustainability research which have not been explored at all such as funding, time, disaster resiliency, insurgency and community pride. This paper will then classify the research papers that it will leverage and the objectives this research will accomplish. It will then conclude with a justification of why this work is an essential component of research in the field.

SUSTAINABILITY OF LOCAL GOVERNANCE

There have been several qualitative studies indicating that CDD initiatives are dependent on the accountability of local leaders to their community and success of projects is dependent on agents and facilitators. Although the community development councils play a major role in the local governance process, there is still no consensus on how they might relate in the long term to the constitutionally-mandated local representative structures. Research also claims that councils remain very dependent on their NGO partners. Moreover, the work conducted on understanding the council and local governance structures in Afghanistan proves that state buildingis quite extensively analyzed as well as demand driven governance and governance in post-conflict situations (World Bank, 2005).

In addition, there has been some research conducted on the sustainability of community development councils (Brick, 2008). From a local governance perspective of sustainability, this paper will focus on how community development councils can evolve to be self sustaining bodies with governmental functions. Although there has been some research on initiating participatory development and sustainability, this paper will mostly be focused on ensuring that the existing structure and processes can be sustained in the long run.

SUSTAINABILITY OF RESOURCES

In terms of the sustaining the resources that are available for community based projects several studies have shown that unless there is a continued lobbying effort by the community they will suffer from neglect by the local government. This is in spite of the fact Page 10

that community driven projects are considered more sustainable that top down projects. Since many communities are able to extract resources by depending on government support, investment for the inputs, maintenance and training of staff to sustain the project falters (Frances, 1999). Moreover the poor are not able to fund their own teachers, doctors, desks and medicine to sustain the initial benefits of the project (Kleeimer, 2000). To accommodate for this dependence, this paper will also analyze the ability to secure resources for the sustainability of CDD projects.

SUSTAINABILITY OF SOCIAL COHESION

Current research shows that the sustainability of projects is expected to improve by enhancing community cohesion or social capital. However, the actual intensity of social bonding and community effectiveness is not clear (World Bank, 2006). It is likely that CDD projects are more effective in a socially cohesive community with better networking and education, therefore benefiting the community. However, there is clearly no reliable evidence on community participation projects actually increasing a community‘s cohesiveness. This paper will seek to validate whether CDD projects improve the long term social cohesiveness of a community using perspectives of NGOs who work with communities implementing CDD.

SUSTAINABILITY OF MAINTENANCE

In a recent study conducted on sustainability 37 percent claimed it was unlikely and more than 30 percent rated it as non-evaluable or uncertain. The more interesting scenario of those that rate projects sustainable have raised concerns about maintenance not being sustainable (World Bank, 2005). One reason for this is that poor communities often find it hard to raise funds for continuous maintenance of projects (Mansuri and Rao, 2003) In fact, not only are communities unable to raise funds for maintenance they are also dependent on external agencies for maintenance of community infrastructure (World Bank, 2006). However, these are not conclusive results. There has been evidence to suggest that projects managed by communities are more sustainable than those managed by local governments (Asim Ijaz, 2001). Moreover, multiple researchers have also found a strong association between the participation of community members and long term sustainability of projects (Katz and Sara, 1997). In another study on community investments conducted by the World Bank, the impact evidence showed that ―the majority of infrastructure appeared to be well constructed and operating adequately and levels of maintenance were equivalent or better than comparators.‖ (Rawlings, 2004). This paper will leverage these results and analyze the

Page 11

long term maintenance of infrastructure projects and whether the maintenance of CDD projects in the Afghanistan is sustainable.

INFRASTRUCTURE SUSTAINABILITY

Additional studies conducted of CDD projects such as schools and health centres at the community level in many countries has found that it is often difficult to sustain the interventions beyond the initial work conducted by the Bank. Although ―the communities are initially successful at creating the project, they lack the material resources to sustain their building efforts‖ (Mansuri and Rao, 2003). Another question that has arisen is whether community participation actually improves the sustainable of infrastructure projects. This has not been fully answered since many evaluations have been conducted in the first three years of the infrastructure building and it is ―difficult to find a comparable counterfactual for them, it is difficult to tell if facilities assisted by the social fund‖ do produce more sustainable infrastructure (Ana Ibanez, 2002). This paper will also investigate the sustainability of infrastructure projects in the context of CDD as it appears that their sustainability has not been the subject of much analysis (CIDA, 2007). In fact the evaluation of infrastructure interventions has not been considered in most monitoring and evaluations (World Bank Annual Report, 2009).

RESEARCH FOCUS

The above literature review makes it abundantly clear that the analysis of CDD projects with respect to sustainability has been concentrated on local governance, community development councils, resources, social cohesion, maintenance and infrastructure. However, other aspects of sustainability such as funding, time, disaster resiliency, insurgency and community pride have not been explored as extensively. In a development context it is evident that ―advances in one sector cannot be sustained in the absence of progress across society‖ (Centre for Policy and Human Development, 2007). To provide this holistic evaluation of sustainability in the CDD context this paper will focus on investigating all the factors which require a long term commitment. This paper will therefore build on the work of Mansuri and Rao (2004), Wassenich and Whiteside (2004) and the Effectiveness of CDD evaluation conducted by the World Bank (2005). Although this paper will draw on the conclusions of the various CDD implementations across the world, the data and analysis presented will be in the setting of the National Solidarity Program in Afghanistan.

Page 12

In terms of sustainability there have been several studies evaluating the quality of infrastructure projects in the ―Bolivia Social Investment Fund (SIF), Honduras Social Investment Fund 2 (FHIS2) and Panama Rural Poverty & National Resources Project (RPNRP). With reports due out on the Pakistan National Rural Support Program (NRSP) – DEC study, Senegal National Rural Infrastructure Project (PNIR), Benin Social Fund (AGeFIB), Cambodia Rural Investment & Local Governance Project (RILG), India Andhra Pradesh District Poverty Initiatives Project (APDPIP), Indonesia Support for Conflict-Ridden Areas Project (SCRAP) and Malawi Social Action Fund (MASAF)‖ (World Bank, 2005). However, not much work has been conducted in terms of sustainability with infrastructure projects in the Afghanistan context, hence the need for this research.

Once complete this research will have two main objectives. Firstly, it will outline the sustainability of each of the different aspects of the National Solidarity Program. By looking at a wide set of variables this will allow for a more holistic evaluation of CDD sustainability in Afghanistan. Secondly, this paper will be able to produce a complete set of variables which could be utilized to evaluate the sustainability of future CDD programs in other countries. This set of variables will be essential for the monitoring and evaluation of CDD both at a local and national level of a country. The primary focus of this research therefore, is to evaluate the contributions of the National Solidarity Program on sustainable development.

Page 13

Methodology

This research will be concentrated on the sustainability of the National Solidarity Program in Afghanistan. This research distinguishes between the national sustainability of provinces and the local sustainability of communities.

NATIONAL SUSTAINABILITY

The national sustainability research will be focused on the sustainable of scaling up the NSP in the 34 Afghan provinces using the eight criteria below. 1. Number of field staff deployed 2. Number of communities facilitating partners are contracted in 3. Number of communities mobilized 4. Number of community development councils elected 5. Number of community development plans completed 6. Number of sub project proposals submitted 7. Number of sub project proposal accepted 8. Number of sub project proposals completed

The vast majority of the data used to analyze the national sustainability of scaling up the NSP is secondary data which has been sourced from the Ministry of Rural Rehabilitation and Development (MRRD) in Afghanistan. The MRRD is a core structure of ―nearly 600 national staff and two international staff in its headquarters, its six regional and 34 provincial management units‖ that are responsible for the implementation of the NSP. (World Bank Annual Report, 2009). This secondary data is referred to as the monthly status reports by the MRRD. This research uses the monthly status reports of the NSP expansion by provinces between January 2008 and September 2009. Although the MRRD data reports on the expansion by district, this research will aggregate the data by provinces and regions to avoid data overload and to minimize the individual fluctuations across districts in a province. The assumption placed on this secondary data is that it is objective factual information reported by the MRRD to external donors. In addition, this research is constrained by the 21 months of data that has been provided by the MRRD and the research will be drawing conclusions based on information which is nine months old as of this writing in June 2010.

To be able to normalize the above data, it is also necessary to use the population and area of each of the provinces in Afghanistan. This information has been retrieved from

Page 14

the MRRD‘s National Area-Based Development Program. This information will be included in the Annex.

LOCAL SUSTAINABILITY

The local sustainability of communities will also be analyzed as part of this research to provide a holistic analysis of CDD sustainability in Afghanistan. Ideally, this analysis would be conducted by speaking to community development councils and local members to identify the individual factors affecting the NSP sustainability in the communities. However, since this was not possible under the circumstances due to security and sanctioning of travel to individual CDCs, this research instead surveyed the facilitating partners remotely. It is understandable that there are bound to be differences between the conclusions drawn from CDC responses versus facilitating partners, however, every effort has been taken to keep this to a minimum. It is already known that ―beneficiaries and facilitators have an incentive to present the impression of a successful project to outsiders and may collude for this purpose. NGOs often avoid working in difficult communities, where quick results may be harder to demonstrate.‖ Hence to avoid a false impression of successful projects, interviews were not only conducted with current facilitating partner employees but also past employees implementing the NSP in Afghanistan. This provided a measure to control for inflated accounts of NSP implementations. Moreover, responses were sourced from all 34 provinces to avoid NGOs not providing data for the difficult provinces. Responses from the MRRD as well as the World Bank employees were also obtained to provide a legitimate picture from the central authorities managing the NSP.

As mentioned previously, this analysis used primary data which was collected by surveying employees at several facilitating partners based out of Afghanistan. Facilitating partners completed the surveys and provided phone interviews during the months of May and June 2010. Although there are only 29 facilitating partners contracted by the MRRD for the NSP in Afghanistan, it is possible for a single province to have several facilitating partners contracted in that province. This means that there are a total of 91 possible facilitating partner province combinations and each of the 91 possible facilitating partner provinces was approached for this survey. For example, three different provincial representatives for the facilitating partner Afghan Development Association (ADA) were contacted because the ADA is contracted by the MRRD in Oruzgan, Takhar and Zabul provinces. Of the 91 possible contacts, 60 have completed and submitted the survey which translates into a response rate of roughly 66%. The 60 completed surveys provide diversity in responses from the provinces, facilitating partners and employee positions. Page 15

The survey to collect primary data was focused on comparing the current community driven development approach versus the traditional NGO using such measures as costs for project, budget expectations, funding from the ministry, community skills development, maintenance of buildings, infrastructure lifespan, disaster resiliency, insurgency attacks, community pride, transparency of accounts and social cohesion. The CDD implementation was compared to traditional NGO projects because a baseline was required for comparison of the implementation. The majority of facilitating partners have already implemented traditional NGO projects either in Afghanistan or in other countries.

Page 16

Case Study

NSP Background

In 2001, after gaining power, Ashraf Ghani had envisioned the creation of several national development projects which would create public trust in governance. These programs included a National Emergency Employment Program to provide jobs across the country, a National Health and Education Program to get basic health packages to citizens and get children back in school, a National Transportation Program to make Afghanistan a land bridge for South and Central Asia and the Gulf, a National Telecommunications Program to set up a cell phone network across the country and attract private investment and a National Accountability Program to build good financial management (Ghani and Lockart, 2008) After the Taliban were ousted from power in Afghanistan, the transitional power also realized that for the people to support the state, the most crucial national development project had to be visible to the 80% of the population in the rural areas. In order to interact with these people, engage them in development, include them in the reconstruction process and provide a uniform approach across Afghanistan, the National Solidarity Program (NSP) was launched. As one of the Afghan government‘s National Priority Programs the NSP has been publicized as one of the most successful CDD programmes in the world to date. As a large scale rural reconstruction and development programme, the NSP had two primary goals: to strengthen local governance to foster rule of law and to lay the foundations of community managed sub-projects comprising reconstruction and development. The reconstruction efforts would be aimed at improving access of rural communities to social and productive infrastructure and services. Launched in 2003, the program was instrumental in generating employment and initiating the rehabilitation of rural infrastructure devastated by severe drought and two decades of conflict. The president of the World Bank estimates the economic rate of return on the NSP to be almost 20 percent (Zoellick, 2008). The NSP is funded by the International Development Association at the World Bank and the Afghanistan Reconstruction Trust Fund. Implemented by the Ministry of Rural Rehabilitation and Development (MRRD) and funded by the World Bank, the UK Department for International Development (DFID), the Danish International Development and Assistance Agency (DANIDA), the Canadian International Development Agency (CIDA) and the US Agency for International Development (USAID), the NSP is the Afghan government‘s flagship programme. Acting as the oversight consultants and responsible for strengthening Page 17

local capacity and programme management at the ministry are the consulting firms GTZ and DAI.

Implementing the NSP

NGOs act as facilitating partners contracted by the MRRD to initiate the NSP and provide technical and capacity building support to the communities. The MRRD has developed an operational manual and facilitating partners that are contracted utilize it for community implementation. The facilitating partners serve many different roles in NSP implementations including training social community organizers tasked with educating the community about NSP, engaging community facilitators, providing technical assistance in designing and building infrastructure, providing election experts to implement CDC elections and assisting with the monitoring and evaluation of projects.

An NSP project cycle for a community is broken down into 5 steps and usually takes 2 years to complete. First the NSP facilitating partner is assigned to the province by contracting through the MRRD. The facilitating partner then contacts the community to inform them of the NSP and commence the community mobilization process. In the second step, the facilitating partner is responsible for initiating a fair and transparent election process to elect members of the locally governing community development council (CDC). The CDCs are elected through a process of identifying the eligible voters in a community, creating a cluster of approximately 25 families and ensuring at least 80% of the cluster votes for representatives. The elected CDC members then decide on the CDC president, deputy, secretary and treasurer. Thirdly, the newly elected CDCs consult directly with members of the community to reach a consensus list of subproject ideas. This list called the community development plan (CDP) comprises of projects which can be carried out with funds from the NSP and independent of outside support. The subprojects which require NSP funding are then submitted by the CDC to the MRRD and the Oversight Consultant in a proposal. The facilitating partner is responsible for training the CDC members in project proposal writing, accounting and procurement. In the fourth step, if the proposal is approved, NSP block grants are disbursed to cover the purchase of materials. Arriving in installments the funds are also used for subproject implementation. The CDC continues to report to MRRD and to the community about the project‘s implementation process and budget. Lastly, the facilitating partners and CDCs undertake an evaluation of the technical quality of completed subprojects and documents the lessons learned. NSP Expansion Page 18

Since there is no formal census data available about Afghanistan it is hard to accurately identify the number of villages. Previously, it was estimated that approximately 20000 rural settlements or villages existed, however this was markedly increased to 42000 villages. Since field coordinators have reported that several of the villages comprise of less than the minimum of 25 families required to initiate a CDC, it is estimated by the MRRD that around 28500 NSP communities would be a reasonable equivalent to rural settlements estimate. This approximation thus average 1 NSP community = 1.474 rural settlements. However the current average used is 1 NSP community = 1.583 rural settlements. (National Solidarity Program Website, 2010). Once fully implemented the MRRD expects to reach all the communities across Afghanistan through the NSP. The NSP community expansion across Afghanistan has been divided into several phases. In Phase 1 which commenced in May 2003, 3 districts in each province were targeted in the first year reaching 6000 communities. In the second year this target was expanded reaching 4500 more communities which translated into half of Afghanistan being included in the program. In 2005 or the third year of the NSP an additional 6000 communities were targeted. The NSP Phase 1 concluded in March 2007 reaching approximately 17300 communities. In Phase 2 which ran from April 2007 until March 2010 the World Bank proposed covering an additional 4300 new communities bringing the total to 21600. This coverage meant that approximately 80-90% of Afghanistan had CDCs. On June 27, 2010 the World Bank announced that it had approved a $40 million USD grant to support Phase 3 of the NSP. The third phase will build on the achievements of the first two phases of the NSP and complete the expansion of CDCs to all the rural communities in Afghanistan. In Phase 3, the Afghan government has introduced several innovations to institute the CDCs as lasting sustainable bodies of local governance. The expansion will also support the disbursement of block grants to the remaining 10320 communities for full NSP coverage across Afghanistan. In addition, to supporting the initial development needs, a second round of grants will be provided to 17,400 CDCs that have successfully used their initial grant. Most importantly, NSP III will focus on improving the institutional quality, sustainability and governance of CDCs and enhance their ability to engage with other institutions. The sustainability aspect of the NSP will be the focus of this research.

Page 19

Analysis and Discussion The national sustainability will be based on the secondary data received from the MRRD. It will concentrate on measuring the sustainability of the five phases of the NSP as mentioned previously: community mobilization, CDC elections, preparing community development plans, submitting proposals and implementing subprojects.

REGIONAL SUSTAINABILITY

Prior to comparing the scalability of the NSP across provinces it would be valuable to examine the regional differences across Afghanistan. Comparing the population, number of CDCs, number of proposals submitted and number of subprojects completed across UN designated regions it is problematic to draw definite conclusions. By separating Afghanistan into the seven UN regions, the provinces making progress are split between many different regions. In Figure 2, the regional percentage of population, CDCs, proposals and subprojects is displayed. For example, the graph shows that although the East has only about 11% of the Afghanistan‘s population it has 12% of the CDCs, 14.5% of the proposals submitted and 14% of the total NSP subprojects completed. The West, however, has more that 20% of the Afghanistan‘s population but accounts for only 14% of all completed NSP subprojects.
25.00% 20.00% 15.00% 10.00% 5.00% 0.00% Popuation density Councils formed Proposals Submitted Subprojects Completed

Figure 2 – NSP expansion across UN designated regions of Afghanistan

However, from a development perspective and a security standpoint, the World Bank and NGOs have remarked that the northern and western regions have made more progress than the east and south of Afghanistan. This draws doubt to the conclusions above that the eastern region is developing faster than the western region of Afghanistan. One of the main reasons for this contradiction is the geography of Afghanistan. The United Nations Department of Safety and Security maps split Afghanistan into three main groups. By Page 20

drawing a belt from the southwest of Afghanistan to the northeast, two main regions are created, the Pashtuns in the south and the Tajik, Hazara and Uzbeki in the north. The diagonal belt running across the country is the third region with pockets of both development and safety. By splitting Afghanistan into seven UN regions the major development efforts of the North and West are diluted by the central provinces showing only a marginal development effort in the north and the west. Likewise, the provinces located in the middle of the belt reduce their NSP development achievements by getting grouped with the provinces in the insecure south.

To make the differences between regions more visible, the same data can be analyzed using only five regions. In Figure 3 below the differences between regions are much starker. Since Kabul is an anomaly because it has a very high population density it has been separated out into the central region. It is also known that since Kabul does not have many CDCs and since it is already developing rapidly, it does not depend only on the NSP for funding infrastructure and development projects. The graph shows that the North, West and East have embraced the community development council election process. Although the North has only 27% of Afghanistan‘s population, it has more than 35% of the Afghanistan‘s CDCs. Similarly the Western and Eastern regions each have only 15% of the population but approximately 20% of the CDCs. The expansion of CDCs in the Southern region has been problematic; however, Nimroz is the sole oddity. All other provinces in the South: Ghazni, Zabul, Helmand, Kandahar, Uruzgan, Farah and Herat have fewer CDCs than their share of the population.
40.00% 35.00% 30.00%

25.00%
20.00% 15.00% 10.00% 5.00% 0.00% Central East North South West

Popuation density Councils formed Proposals Submitted Subprojects Completed

Figure 3 – NSP expansion across four regions and Kabul in Central region

Page 21

Although the East, North and West have all embraced the CDC election process there are difference in the actual implementation abilities of these CDCs. The West is able to leverage the abilities of the CDCs to submit many more development proposals than the North and the East. On average each CDC in the West has submitted over 3 proposals whereas the North and East have submitted 2.26 and 2.05 proposals per CDC. When analyzing the implementation abilities of CDCs, submitting proposals are the first step to completing the subprojects. To ensure the long term sustainability of scaling up in NSP phase 3, the MRRD and NSP face several challenges in the East and South. The first is to ensure that CDCs are being installed in the provinces and districts of the East and South that lack them. For example the South and North have approximately the same population, but the South has half the number of CDCs implemented. This means that the South would have to add more than 3000 new CDC just to match the number of CDCs in the equally populated North.

The security situation should be accounted for when pursuing expansion of CDCs into these regions, however, CDCs that have already been implemented in the South have shown remarkable progress. When comparing the ability of CDCs to translate an approved proposal from the MRRD into a completed project, the South is just as good at as the North. In both the North and South about 66% of projects that were approved by the MRRD were completed by the CDC. This alone is not proof of a stabilizing situation in the South and reason for CDC expansion but it does prove that the safer districts in the South can make significant progress if given the opportunity.

The second issue is with sustaining the already implemented projects. The Western region has completed 10,101 of the 11,270 approved projects, translating to a completion rate of almost 90%. This shows that many districts in the Western region where significant progress has been achieved need funding to continue the development of additional projects. The recently announced $40 million dollar grant from the World Bank for the execution of NSP phase 3 is a first step. In the future, as funding for the NSP phases dries up as donor money is no longer the major source for funding the NSP, the provinces and districts will need to seek a more stable funding source to continue the development initiated by the World Bank.

The third issue for the long term sustainability of the NSP regionally is to ensure that the money already invested in empowering the community and setting up CDCs are being used effectively. As already mentioned the CDCs in the South and East have not been as productive in submitting proposals as CDCs in the West and North. However, the issue is Page 22

much more fundamental. In the West, every elected CDC has already completed a community development plan (CDP), which is the first step to submitting proposals to the MRRD but there are still about 130 districts in the South and East which have not yet completed a CDP. This must be pursued by the MRRD and the facilitating partners.

For the regional stability of the NSP, the three issues of increasing CDCs, sustaining funding for existing CDCs and empowering the underutilized CDCs must be actively managed by the MRRD and the provincial facilitating partners.

PROVINCIAL SUSTAINABILITY

Although the regional data provides a good basis for understanding the overall progress of the NSP across Afghanistan, the provincial data is useful to highlight provinces that do not follow their regional norms. The provincial data is also useful for implementing organizations as the facilitating partners are divided into provinces for executing the NSP.

The analysis of provinces will compare the density of the provinces to the density of the CDCs. In theory, there should be a correlation between the number of people in a province and the number of members per CDC in that province.

Page 23

800
mem/cdc

6000
density

700 600 500 400 300

5000

Population Density

3000

2000 200 1000 100 0 0

Figure 4 - Comparing population density versus CDC density across Afghanistan's provinces.

Page 2

Members per CDC

4000

As apparent from Figure 4, when the population per square kilometer decreases the population per CDC does not necessarily decrease. Although, it should be noted that there will be an absolute minimum number of people per CDC because a CDC must include 25 families and based on the average of 8 members per family in Afghanistan the population per CDC will never go below 200. It should also be noted that there are no provinces that encounter this situation.

The anomaly of Figure 4 is easily visible when comparing the provinces of Kapisa and Paktya. Kapisa has a population density of 194 and Paktya has a population density of 65 but they both have 619 people per CDC in their respective provinces. Although Kapisa is three times denser in population than Paktya they both have the same number of people per CDC. The limited expansion of CDCs in Paktya might be explained due to the security situation, however, this is not the case for the relative peaceful province of Balkh. Balkh, like Paktya is also limited in the number of CDCs that it should have relative to its population size. However, this is not all bad. Mazar-e-Sharif in Balkh has made significant progress even though there has been some resistance to CDCs which can be partially attributed to the province‘s confidence in the governor, Atta Muhammad Noor (Gall, 2010). There is little doubt that the governor has managed to bring development and security to Balkh province with strong public support. This is even more astounding because the province is divided by ethnic and political rivalries. Moreover, the support of Western donors to Mr. Noor‘s governing may also undermine some of the efforts of the NSP expansion in Balkh. Nevertheless, there have been issues in two districts of the province where insurgents remain active. However, this city in Balkh has become an investment haven so large that even the Finance Ministry categorizes it as the largest source of revenue for the country. Hence, the graph shows that for the majority of cases where correlation does not occur between the population density and members per CDC, in some cases there is still progress being made.

As expected, towards the right of the graph are the majority of the provinces in southern Afghanistan. Herat, Helmand, Paktika, Farah and Zabul are all provinces where CDCs expansions have not made major inroads. There are still some outliers such as Nangarhar which has become safer, increased development and reduced narcotics under the leadership of Gul Agha Sherzai (Sandstrom, 2003).

Once the CDCs are implemented it is useful to compare how diligent the CDCs are at completing proposals and following up on project proposals which are returned to the CDC or proposals which have been outright rejected by the MRRD. The MRRD does not provide Page 2

information on the number of subprojects that have been changed by the CDC, the number of subprojects that have failed implementation or where the proposal has been returned or rejected. The worst case scenario would be a subproject failure where the proposal is approved by the MRRD and the funds disbursed but the community discards the subproject after some or all of the funds have been utilized.

However, using the numbers provided by the MRRD some conclusions can be drawn on the subproject proposals which have been returned and those that are rejected.

Province Day Kundi Hirat Farah Nuristan Wardak Balkh Parwan Ghazni Zabul Takhar Samangan Badghis Kandahar Faryab Kapisa Panjshir Paktika Kabul Sari Pul Laghman Baghlan Jawzjan Khost Bamyan Kunduz Hilmand Ghor Paktya Nimroz Badakhshan Kunar Nangarhar Logar Uruzgan

January 2008 Approval Rate Completion Rate 100% 47% 98% 51% 100% 68% 95% 31% 97% 45% 98% 47% 96% 49% 99% 70% 100% 30% 95% 40% 99% 54% 95% 47% 96% 59% 96% 23% 100% 67% 99% 51% 100% 78% 98% 66% 95% 27% 97% 51% 89% 38% 100% 45% 95% 59% 95% 56% 96% 48% 100% 36% 96% 47% 95% 55% 94% 62% 96% 25% 98% 44% 95% 42% 92% 52% 100% 54%

September 2009 Approval Rate Completion Rate 100% 64% 100% 61% 100% 92% 100% 46% 100% 74% 100% 85% 100% 77% 100% 69% 99% 44% 99% 62% 99% 80% 99% 81% 99% 70% 99% 57% 99% 78% 99% 78% 99% 82% 99% 85% 98% 69% 98% 63% 98% 77% 98% 83% 98% 59% 98% 80% 98% 78% 97% 48% 97% 61% 97% 70% 97% 82% 97% 48% 97% 50% 96% 53% 95% 57% 93% 18%

Table 1 – Approval and completion rate of NSP subprojects across Afghanistan's provinces.

Page 3

The MRRD reports the number of proposals submitted, approved and completed. The approval rate is derived from the number of submitted proposals and those that are approved. In addition, the completion rate is simply the percentage of approved projects that have been completed. On average in January 2008, 96.8% of the projects submitted by CDCs across Afghanistan were approved by the MRRD, this number jumps slightly to 98.2% in September 2009. In January 2008, Baghlan was the only province that had less than 90% of its projects approved. However, Baghlan is a textbook example of figuring out what will and will not get approved. Between January 2008 and September 2009, Baghlan had a 10% jump in proposal approval rate. It is obvious from glancing at the approvals above that the majority of the projects submitted do get approved. The Deputy Director of Programs for the International Rescue Committee explains this increase using two main reasons. Firstly, the MRRD has turned into a very slick machine in the last two years, and Kabul has become very responsive to subproject proposals. Secondly, the facilitating partners have a much better understanding of what will get approved. For example, in the past solar projects used to be allowed but because they are easy to implement without the community driven process and that parts of solar panels might be used by insurgents they no longer are approved. Large micro hydro projects which are too big for NSP funding tend to seek funding using a different route. Facilitating partners are more attuned to these changes and are much quicker at adapting to these situations. In addition, it shows that there are very few projects that will not get approved.

In terms of completion rates there are provinces that may show problems but that can be explained using district level data. For example, Table 1 presents Uruzgan province with a completion rate of 54% in 2008 and 18% in 2009. In 2009, the province with the second lowest completion rate is Zabul at 44%, more than double Uruzgan‘s subproject completion rate. To comprehend this large discrepancy for Uruzgan, its districts need to be analyzed further. When comparing the approval rates of the individual districts, Tirin Kot and Khas Uruzgan both have approval rates of 100%. Dihrawud had 106 of its 129 projects approved and Chora had 66 of its 73 projects approved. Comparing the completion rates for these two districts not a single project has been completed in either of these two districts lowering the overall completion rate of the province of Uruzgan. The fact that the majority of these projects in Chora and Dihrawud were only approved in June 2009 explains why the completion rate is so low for these two districts. However, it does not explain why these two districts were such late arrivals to the NSP development process. Dihrawud district has been a key district linking the North-South and East-West routes of Uruzgan (Ariana, 2010). This district borders the volatile Helmand province and it symbolic for the Taliban. It is here that the Taliban leader Mullah Omar had strong family ties before their ouster. This can explain Page 4

why some volatile districts are only now receiving approvals for NSP subprojects. It is also a concrete example of security deliberately slowing down development in Afghanistan. Looking at the data more widely it is clear that the fate of Uruzgan is shared by several other provinces in the South and East. Table 2 below shows a list of the provinces under Uruzgan with the lowest changes in completion rates from 2008 to 2009. Ghazni for example had 1% lower completion rate in 2009 than in 2008.

Province Uruzgan Ghazni Khost Paktika Logar Kunar Hirat Kandahar Nangarhar Kapisa Hilmand

Completion Rate -36% -1% 0% 4% 5% 6% 10% 11% 11% 11% 12%

Table 2 – Provinces with the lowest changes in completion rates of NSP projects from 2008 to 2009.

However, in terms of completion rates not all is bad. Leaving aside Uruzgan and Ghazni provinces, the completion rate for NSP subprojects has increased in all other province. Once the NSP is more stabilized, the MRRD should look at how projects are approved more systematically. For example there are several provinces such as Daikundi, Herat and Nuristan which are over appreciated by the MRRD and there are provinces such as Jawzjan, Bamiyan and Nimroz that are under appreciated. When ranking the provinces by their approval rates and completion rates, Daikundi is ranked 1st in approval rates in 2009 but 19th in completion rate in 2008. Herat is ranked 2nd in approval rates but 15th in completion rates, Nuristan is ranked 4th in approval rates but 30th in completion rates. This is a disturbing trend that shows Wardak, Balkh, Parwan, Zabul and Takhar in the top 10 provinces with the highest approval rate but all of these provinces have very low completion rates. Conversely, Jawzjan, Bamiyan and Nimroz had very high completion rates until January 2008 but were not rewarded with high approval rates for new subprojects in 2009. In the early stages of the NSP, expansion of the program to all provinces was the primary purpose which is understandable. However, once the NSP becomes more established, the MRRD should pursue a meritocratic approach for program expansion. The MRRD should consider the long term sustainability of the NSP program being hinged on results on the ground and pursue projects based on either completion rates, impacts of the project or simply reaching out to the poorest of the poor. However, this overarching strategy is not visible from the current provincial expansions of the NSP. In the future, donors will be Page 5

seeking results from the NSP as a condition of additional funding and meritocracy in MRRD approval for subprojects will be important.

When comparing the NSP expansion across provinces, it is also useful to understand the CDC expansions over the last few years. One conclusion that can be drawn from the CDC expansions is that CDC expansion has been greatest in volatile provinces. In addition, the expansion in volatile provinces means that the NSP is trying to be a truly national project by first laying the initial governance structures in all areas. Of the top nine provinces with the highest increases in CDCs, seven have security issues as demonstrated in Table 3 below. Only Takhar and Nangarhar on the list can be considered safe provinces.
Province Uruzgan Kunar Kandahar Day Kundi Khost Nangarhar Ghazni Takhar Hirat Percent increase in CDC from 2008 to 2009 232.84% 144.06% 138.72% 136.73% 135.11% 131.15% 129.37% 122.62% 121.14%

Table 3 – Provinces with the highest percentage increases in CDCs between 2008 and 2009.

The second conclusion which can be drawn from the CDC expansion results is that provinces which have a high number of members per CDC are getting the majority of new CDCs. The figure for members per CDC is derived by dividing the population of the province by the number of CDCs in that province. The expectation is that provinces where the members per CDC are high, these are the same provinces where enough CDCs have been instituted. Uruzgan, Herat, Kandahar, Nangarhar, Daikundi, Ghazni and Balkh had the highest members per CDC and the greatest increases in CDC proving that the MRRD has been concentrating its efforts on getting all of Afghanistan onto the NSP. The first step to making the NSP a national project is getting CDCs in all communities and Table 4 shows that the MRRD has concentrated its efforts in the right direction by increasing CDCs in provinces with highest number of members per CDC.

Province Kabul Hilmand Uruzgan Herat Kandahar Nangarhar Daikundi

Ranking of highest members per CDC 1 2 3 4 5 6 7

Ranking of greatest increases in CDC 22 23 1 9 3 6 4

Page 6

Ghazni Balkh Khost

8 9 10

7 17 5

Table 4 – Provinces with the highest members per CDC and the greatest increases in CDC.

LOCAL SUSTAINABILITY

The local sustainability analysis is derived from the primary data retrieved from surveys of facilitating partners across Afghanistan. Each of the 60 respondents was asked to complete a 16 question survey to assess the sustainability of several factors with respect to the local implementation of the National Solidarity Program in communities. The data is used as an aggregate to draw conclusions about the whole of the programme and the surveys were not disaggregated by provinces or regions because there are not enough data points to draw valid conclusions. Cost and Funding – The cost of implementing the NSP in communities across Afghanistan has been researched quite extensively. The general conclusion is that the NSP has achieved concrete successes at a price tag considerably lower than having Western NGO implementing traditional initiatives (Nagl, 2009). Human Rights Watch has observed that NSP projects are on average 30 percent cheaper than those built by foreign nongovernmental organizations. The NSP has been funded through two phases and was recently approved for phase 3 with the $40 million approved by the World Bank. In the survey conducted for this research the facilitating partners were asked to compare the overall costs of implementing NSP projects and traditional NGO led projects. Of the 60 responses, 85% have noticed the costs to be lower in NSP projects, 3% expected similar costs between NSP and traditional NGO projects and 12% have noticed higher costs in implementing NSP projects in Afghanistan. Moreover, the respondents were also asked to compare whether the final costs of the project implementations have been under or over their initial approved budgets. 13% of facilitating partners were under their initial budget, 78% had matched their budgets and 9% of facilitating partners went over budget. The facilitating partner ZOA Refugee Care for Afghanistan (ZOA) has established that ―too many parties were involved providing for delays and consequently more costs‖. Hence, coordination of activities between several actors is a definite cause for increased costs; however, in the long run as the implementation of subprojects becomes second nature to the CDC it should be able to include the additional cost of engaging more parties. However, this diversity of groups also provides a diversity of resources. Although ZOA experienced a higher cost for including additional actors, these additional parties actually provided more Page 7

resources. Of the 80 projects implemented by ZOA only 1 was over budget, the majority were just below budget.

Although including additional parties in the consultation process can be costly, natural disasters can be another factor that increases the costs for NSP implementations. The Aga Khan Development Network (AKDN) has found that ―due to the unpredictability of the climate, floods, earthquakes, etc. there can be unforeseen costs‖ resulting in some projects being over budget. However, the AKDN has also noted that under such circumstances communities will find ways to save and volunteer more along the way resulting in most projects being completed within the allocated budget. The facilitating partner Bangladesh Rural Advancement Committee (BRAC) in Nangarhar has also observed similar instances of CDCs increasing their contribution above 10% of required costs when they observe costs going over the initial budget.

Several anecdotes were provided by the facilitating partners to compare the costs of NSP and traditional NGO projects. CONCERN in Takhar province noted that an 11 km road constructed by an NGO cost $1,000,000 USD, however, the local CDC completed a similar road for less than $60,000 USD. The facilitating partner Danish Committee for Aid to Afghan Refugees (DACAAR) in Laghman province has observed that ―the cost of NSP subprojects are cheaper compared to other similar traditional NGOs based projects. For example, an eight classroom school will cost around $100,000 USD under the NSP program but if it is to be built by a construction company it will cost around $150,000.‖ The facilitating partner Norwegian Project Office /Rural Rehabilitation Association for Afghanistan (NPO/RRAA) has explained why the NSP projects are able to complete projects for so much cheaper. NPO/RRAA states that ―NSP projects are able to implement the projects without a significant administration cost; however, traditional NGOs have to account for administration, program and operational costs‖.

The other significant expense in implementing subprojects is the cost of materials. The facilitating partner Ghazni Rural Support Program (GRSP) has found that projects run over initial budget because of the increasing cost of materials. However, GRSP also noted that locals were also able to negotiate much better prices for materials than traditional NGOs. BRAC has also remarked that ―CDCs bargain a lot during procuring‖. This brings up an interesting point about the block grants provided by the MRRD for subproject implementation. Although it is equitable for the MRRD to provide the same funding for similar subprojects across Afghanistan, the funding does not vary based on local situations. The facilitating partner Swedish Committee for Afghanistan (SCA) has stated that one of the Page 8

major problems with the NSP is that block grant funding does not change with different locations, road accessibilities, security situations and availability of professional staff. In several cases the SCA has noticed that with the allocated budget one community can implement a subproject in one location but with the same budget the same subproject cannot be implemented in another location. The expectation is that for a sustainable long term implementation the MRRD should make implementation costs more variable. However, the globalized nature of contracts will mean that contractors will expand to other parts of Afghanistan if a demand for their services exists from CDCs. It is expected that in future subproject implementations, CDCs will experience a greater competition amongst contractors and eventually a lower cost of implementing projects. Until this becomes reality, the MRRD could ensure that those communities which are not as easily accessible and those experiencing security concerns are provided with additional resources to complete NSP subprojects.

However, it would be naïve to assume that the NSP subproject costs are the only costs associated with the NSP implementation. The facilitating partner is contracted by the MRRD, the facilitating partner expends resources to engage and empower the community and the facilitating partner disburses funds to ensure a CDC is elected. By ignoring these additional costs and only tabulating the subproject implementation costs, the NSP subprojects are much cheaper. Although the initial setup costs of electing CDCs are onetime costs, to provide long term sustainability, these costs cannot be ignored. In the future, newer CDC members will be elected and additional committees will be instituted to ensure transparency of procurement, the maintenance of infrastructure and the monitoring and evaluation of subprojects. The facilitating partner Afghan Development Association (ADA) in Uruzgan has stated that these costs should be included otherwise ―the NSP projects will always look cheaper if we do not calculate the costs of FP, PMU, RRD and others‖. Moreover, additional research should be undertaken to understand the real cost of NSP subprojects by including the overhead costs of contracting facilitating partners and empowering communities.

Lastly, for sustainable development to occur, meritocracy of CDCs should be a driving force in block grant disbursement. Although the NSP is a national programme which aims to integrate the rural population of Afghanistan with the state, the MRRD must seek to reward those CDCs which have completed projects on time, under budget and with all requirements met. The MRRD could start by removing the cap on budget allocations regardless of the number of beneficiaries of the project and incentivize CDCs which are able to produce results. In addition, the MRRD must seek to streamline the funding approval process by Page 9

reducing the number of block grant installments and providing greater transparency of subproject funding across districts. Completion Time – The time it takes to complete NSP projects is by far the biggest concern amongst facilitating partners in Afghanistan. The AKDN has stated that ―community-based consensus and prioritization can often be a longer process than top-down descison-making‖. This long lead time is apparent from the results of the survey conducted. Only 10% of respondents believe that NSP projects take less time to implement than traditional NGO based projects, 22% believe it takes the same amount of time and 68% of facilitating partners stated that NSP projects take longer to complete than NGO led initiatives.

Generally, CDD projects do take longer to complete. There is a primary tradeoff between short term impacts and the need for impacts to be sustained, which requires a slower and more costly setup so that capacities can be developed. In Nangarhar province, GAA has experienced that ―learning is always more time consuming‖. In Kandahar province, SDO states two main reasons for the lengthy implementation. ―Firstly, proposals have to be approved by the PMU which takes a long time, and it takes even longer to get the physical transfer to the CDC account. In these projects, the second installment can also take a long time to be transferred. Secondly, as NSP projects are technical and implemented by non technical people (CDCs and ordinary people of a community) it is a bit difficult compare to other NGOs work. Though, facilitating partners on the ground do assist in the whole process and provide assistance for the smooth implementation of the projects proposed by communities.‖

For the long term sustainability of NSP projects the fixed time required to elect CDCs are deadweight costs, however, these are onetime costs. Although normal project cycles for first implementation can take approximately 2 years several NGOs have reported that additional subprojects in a community take less time, this decrease in completion time is expected and should continue as CDCs become familiar with the MRRD requirements. Of all the NGOs who completed the survey only NPO/RRAA and GAA have reported that NSP implementations take less time. In addition, for the facilitating partners that did report longer completion times, every single facilitating partner reported significantly longer time to complete subprojects instead of slightly longer time to complete subprojects. Further research is required to understand the completion time of subprojects for communities in their second implementations. The MRRD should consider facilitating the sharing of best practices amongst communities to reduce the time required to complete newly approved

Page 10

subprojects. Moreover, the MRRD should seek to reduce the block grant funding bottlenecks which are a consistent concern in completing projects in a timely manner. Infrastructure maintenance and lifespan – The facilitating partner International Rescue Committee (IRC) has stated that in the medium term additional support is provided for project sustainability and self reliance. Moreover, the IRC has reported that this has resulted in people ―caring for maintenance of projects more than with other NGO‖ initiatives. However, in Indonesia‘s Kecamatan Development Project communities have postponed maintenance until it is unavoidable —for example, the road is about to become impassable, or the bridge is about to collapse, and then to do the minimum that will get it back into usable shape. In Mali, similar infrastructure projects were constructed but very few arrangements were made for its maintenance. Hence, even though infrastructure may be standing in a village or community, it is often underused or not being used for the purpose that was originally intended. In Benin, although the schools were constructed there was no steady flow of available teachers to maintain the standard of teaching in the schools. In Eritrea, the same was true for doctors in hospitals. However, the World Bank did find that water supply schemes in the villages of Eritrea were maintained well and concluded that if the village was not dependent on resources from outside the village maintenance increased and with it the lifespan of the infrastructure. The surveys conducted with the facilitating partners in Afghanistan‘s NSP show that 82% believe that NSP subprojects are better maintained that NGO led initiatives and 18% considered the maintenance regime to be similar. However, the surprising aspect was that not a single facilitating partner observed maintenance to be worse in NSP subprojects compared to traditional NGO based projects. In Sarepol, the facilitating partners still believe that the results are not conclusive with respect to lifespan. AKDN believes that although it is still too early to answer the question on infrastructure lifespan, there is a clear trend that NSP projects are of better quality. With respect to infrastructure lifespan, 70% of respondents stated that the lifespan was longer for NSP projects, 25% mentioned that the lifespan was similar and only 5% believed that infrastructure projects had a longer life when built with a traditional NGO based approach. Using the historical context of other countries, the expectation it that maintenance regimes should falter over time with CDD implementations and many facilitating partners in Afghanistan cannot conclusively state whether the maintenance and lifespan of NSP subprojects is actually better. As the majority of the NSP subprojects are less than 5 years old, it is crucial that future studies are able to investigate this question in more detail.

Page 11

Community skills development – The CDD mechanism has been interpreted differently by officials at the central and local levels and the World Bank. The local level officials view the NSP as a means of increasing the skills of communities but World Bank practitioners often view it as a means of doing more with less, rather than actually putting communities in control. The AKDN has remarked that if the ―community members participate in other NGO projects, it is only through manual labour, rather than financial management, procurement, and decision-making‖. Therefore the NSP has been able to diversify the skill set of individuals in local villages. Approximately 36% of the respondents believe that communities provide more time and labour for NSP subprojects than other NGO efforts, 59% consider communities provide equal efforts for NSP and NGO initatives, and 5% believe that communities spend more of their own time and resources for NGO led initiatives than NSP subprojects. In addition to the additional time invested by the communities, more than 67% of facilitating partners agree that community pride in infrastructure subprojects is sustained well after project completion. Therefore, not only are communities investing more time, but they are happy with the results of investing this additional time in NSP subprojects.

In terms of skills development, 82% of facilitating partners agree that the communities gain more skills by implementing NSP projects rather than implementing a similar NGO led project, 17% believe that the skills gained between NSP and NGO projects are similar and a mere 2% believe that communities gain more skills by implementing projects using the traditional NGO led approach. Although, the community has gained many additional skills from their NSP projects, ZOA has found that ―some do not like to be laid off as the project comes to an end‖. This means that to continue to maintain the skills gained and provide useful employment the NSP must seek to engage these community members with additional training and work. The skills gains should be sustained by providing additional funding for these communities. Moreover, civil society associations should be encouraged to harness the skills gained and leverage them for more work. In India, it was common for field drains and connecting drains to not be repaired. The local CDCs should encourage and hire these community members with technical skills to maintain the infrastructure projects which they have built. Disaster resiliency and mitigation – It is already known that CDD has been useful in the initial response and mitigation of natural disasters. As community members are usually the first responders in disasters, their active participation is crucial in many World Bank financed disaster management projects. After the 2005 tsunami, for instance, CDD approaches in countries such as Indonesia, India, and Sri Lanka had initiated a front line response to Page 12

ensure that resources were used effectively so that the communities themselves were involved in assessing the needs and recovery programmes. The Second Pakistan Poverty Alleviation Fund too was quick to respond after the earthquakes of 2005. The communities were quick to responds and were instrumental in the facilitation of the disaster response efforts.

There are two parts to the disaster mitigation efforts with respect to the NSP. The first as mentioned above is the training of local community members to respond to disasters as first responders. The AKDN and several other NGOs have reported that training is being planned and implementation would occur soon in the communities already implementing infrastructure projects. The second is in regards to the structural resiliency of the subprojects already implemented. The ADA in Zabul has reported that ―the people don't know about the disaster and its relation with the subprojects, they only wants the quick benefits.‖ However, CONCERN in Takhar has noted that the project feasibility study is undertaken with ―engineers which is a factor on selection of project location. however in some places, resilience of projects to natural disaster is useless because the disaster is too powerful.‖ Hence, the resiliency of projects depends on the communities implementing the projects. The NSP Operation Manual or MRRD do not automatically disqualify approval of projects because the resiliency of projects is not considered in the design process. However, the MRRD might consider the benefits of this in future approvals of the infrastructure projects in the NSP Phase 3 implementation, as the sustainability of projects is a key concern in the phase 3 based on the World Bank funding. In Paktika province, AREP has found that the resiliency of materials are considered which indirectly improves the stability of subprojects in the case of natural disasters. Only 17% of facilitating partners believe that disaster resiliency is always considered in the design and implementation of NSP projects, 51% believe that it is not usually considered and more than 32% believe that it is never considered in NSP subproject implementations. A recommendation should be made to the MRRD that block grant funding should be made available for mitigation of disaster and terrorist attacks that destroy NSP projects. The Wall Street Journal has also requested the implementation of a Disaster Mitigation Fund for these purposes. Insurgency attacks and security – Human Rights Watch had conducted a survey of schools which were built by the NSP and noted that they had a lower chance of being destroyed by insurgents than schools built by other aid programmes. The reason was simply local ownership. "If you're the Taliban, you feel some comfort in attacking things built by foreigners," de Tray says. "But you don't want to create animosity among citizens you're trying to recruit to your side." Sarah Chayes, author of The Punishment of Virtue: Inside Page 13

Afghanistan After the Taliban and now a resident of the southern city of Kandahar. Moreover, since the NSP subprojects are less, villagers are more willing to maintain and defend them. "If people feel invested in a project, they'll protect it, at least within reason." Not only are the subproject less likely to be destroyed, but they are also less likely to be vandalized. ―The Afghans have this great saying — ‗If you sweat for it, you protect it‘ — and so getting highly localized development in the hands of communities is critical, Of the 40,000 projects built through the NSP only about 5-10% have been vandalized by local insurgents or criminals. Community ownership is crucial and other aid programmes have not done as well.

Although NSP infrastructure projects have been safe, the same cannot be said for council members on the CDCs. There is evidence that the Taliban are fighting the councils much as they have resisted other government initiatives. The Taliban have killed more than five councilors and wounded one since the four councils were formed in Helmand Province.

In the survey conducted for this research, the results of the Human Rights Watch survey were validated. 57% of facilitating partners have noticed that NSP projects are less likely to be attacked than other aid initiatives, 37% of facilitating partners claim both projects are equally likely targets and only 6% believed that NSP projects were a more likely target than other aid initiatives. According to facilitating partners, the majority of NSP projects therefore are less likely to be attacked by insurgents. The facilitating partners were also asked a follow up question of who they believed provided the greatest security for NSP infrastructure projects and CDC members: 64% believed the Afghan National Police (ANP) provided the best protection and 36% believed that the ANP reduced the security situation in comunities implementing the NSP. With respect to the Afghan National Army (ANA), only about 56% believed the army provided the required security for CDCs and infrastructure projects. As the results are not conclusive, additional research should investigate whether the ANP, ANA or Afghan National Auxiliary Police (ANAP) were best suited to provide the required security to civil society development. For the long term sustainability of community members, it is crucial to understand whether the ANA or ANP actually provide the best support for the development of the NSP. Although it is known that the army is less corrupt, the police force provides the required day to day security in communities and support for the local long term development of districts. Fundamentally, no foreign military force can guarantee a sustainable and peaceful solution, therefore Afghanistan will require a robust and well trained security force to safeguard the successes and achievements of the NSP. However, the NSP should continue in its central focus of promoting peace and not fighting

Page 14

terrorism because fighting terrorism, is based in fear. However the promotion of peace, is based in hope. Community project management – One perspective of future CDCs has been to see if they can be leveraged as civil society organizations that are able to work with external organizations for the benefit of the community. This research asked facilitating partner organizations about the long term sustainability of CDC collaborating with external organizations and other neighbouring CDCs. Exactly 80% of the facilitating partners believe that CDCs are able to play such a role in the future and only 20% believe that the current organizational structure of the CDCs does not make conducive for the CDC to manage additional responsibilities. In Wardak province for example, SCA believes that by using CDCs politically to coordinate with other CDCs and the national governments in other matters can actually ―kill the CDCs and NSP as a whole‖.

Some of the expected broader range of responsibilities proposed will provide CDCs with a significant role in community development planning and community governance. The new responsbilities could include: resolving internal community disputes (e.g. land, water rights); bargaining with traders; sharing lending risk; bargaining for better share cropping arrangements; bargaining between CDCs and developing district plans and projects, thus achieving economies of scale; and the registration of births, marriages and deaths. A more ambitious and longer term task, which is nonetheless worthy of consideration, is to use CDCs to collect local taxes on the basis that they would retain the large part of the levy for community development activities, while transferring the remaining part to government. The collection of some form of taxes is an increasingly contentious issue. It is already known that many civil society organizations have risen and fallen as the agencies have come and gone and have been dependent on external funding to keep them in business. The CDC formation and activity has been driven by considerable funding, the level of which cannot be expected to last into the future. Without that funding, CDCs will not necessarily continue to function, even if they exist formally. The CDCs will be critical in initiating this process from donor funded development to community funded development. To help with this transition process the CDCs should play a greater role in empowering communities and providing top up grants and matching funds, hence, increasing the onus on villagers to provide the funding for CDC led projects. Although communities already pay 10% of labour and material for infrastructure development, the top up grants and matching funds could provide a transition to communities providing 50% of the funds and eventually 100% of the funds, making the NSP a self sustaining initiative.

Page 15

In terms of community skills development, the results of this research have been overly positive. More than 82% of facilitating partners believe that NSP projects are better able to increase community skills in trades, building and back office activities that traditional NGO led initiatives. The task now will be for CDCs to utilize these existing skills and leverage the abilities of community members for infrastructure repair and maintenance.

The critical aspect that threatens community development councils will be its inactiveness. In Rio Grande do Norte in Brazil, once community association mobilized to get the water system the people stopped mobilizing and the community did not benefit with any other projects. Once the target of getting water was achieved the people stopped mobilizing (World Bank, 2005). This is not a remote scenario for Afghanistan and to prevent its occurence, the community development councils will need to continuous engage with the community to refine CDPs and proposals so that the community does not lose interest in its CDC but more importantly in future civil society organizations that seek to improve the lives of the community.

It has also been common practice for community development councils to be compared to each other, however, the circumstances of CDCs working in difficult areas should be accounted for. The interviews with facilitating partners has revealed that the management of CDCs in problematic areas of the countries is much more difficult.

Ideally, for the long term availability of competent CDC members, remuneration of some form should be considered. It is clear that many CDC members are simply seeking to improve the living conditions of their communities by engaging with civil society organizations such as CDCs and not for supplemental income. However, the CDC cannot effectively increase its currently responsibilities without continous access to competent members of the community for extended periods of time. For this reason, CDC members who are directly involved in either the administrative or project management tasks should be adequately rewarded. There are some indications that the World Bank believes that only those CDC members with project and site management responsibilities and in enabling roles should be paid and those CDC members in empowerment roles should not be rewarded (World Bank, 2005). However, many of the empowerment roles required literate members of society who are at a premium in Afghanistan and the CDCs should compensate these individuals for the educational backgrounds. The independence of CDCs are the only way that facilitating partners can be eventually relieved of their tasks in empowering and enabling the communities of Afghanistan.

Page 16

Dependency and Approval from the MRRD – It is almost unreasonable to assume that the community development councils can be a viable solution by completely ignoring the government structures in Afghanistan. Moreover, there has been signficant hostility of other ministeries to allow the creation of legitimate councils through the MRRD as the MRRD already already receives a dispropotinately large portion of international support compared to other ministeries. The long term sustainability of these councils could be jeopardized by the failure to expand the current efforts to the subnational governance and rmore generally, risk the development of the NSP. This research has found that more that 57% of facilitating partners believe that the current approval process of the MRRD is a significant drawback to the NSP. Moreover, this research has found that 68% of facilitating partners believe that the NSP is not sustainable by depending solely on the MRRD and 83% believe that cooperation needs to be improved for long term viability.

The three recommendations to improve this process are to improve the disbursement process, including the provincial administration in the process and improving the cooperation with government departments. Firstly, to improve the disbursement process the number of installments needs to be reduced and the CDCs should be provided with a fixed date for when the funding will arrive so that contractors and materials can be planned accordingly. Secondly, in the provincial administration a strategic plan must be devised to understand what role the provincial authorities can play in improving the infrastructure development. Thirdly, the varying ministries should be provided with concrete roles and responsibilities to improve their co-ordination across departments for and effective and efficient delivery of the NSP. Transparency – Already ranked one of the five most corrupt nations in the world by Transparency International, the MRRD has improved the image of the NSP by conducting social audits with the help of shuras and the CDCs. The facilitating partner BRAC in Nangarhar has reported that CDCs have regular meetings with community members and present their activities for the last month. Some CDCs have also lost their credibility in this process and have been voted out in subsequent elections.

This research has found that in terms of transparency the NSP has performed better than traditional NGO led initiatives. The surveys conducted by this research found that more than 85% of facilitating partners believe that the NSP approach implemented by CDCs is more transparent than NGO initiatives, 10% believe it is similar and only 5% believe that the NSP performs worse in terms of transparency. Both the cyclical elections and the public

Page 17

notice boards have acted as crucial instruments for social audits and must be enhanced in the future to promote transparency. Social Cohesion – In Takhar, the facilitating partner CONCERN has noted that one of the biggest differences between NSP and NGO projects is the targeted population. NGOs are normally mandated with more specific objective such as to help with poorest of the poor whereas the NSP is targeting the entire community. This allows the community to build social cohesion as they have collectively identified common needs. The facilitating partner UN Habitat has also noticed that this collective includes many different tribes. The facilitating partner CHA in Herat has noted that as the elections are conducted in community clusters, each member is selected from a different cluster and as ―the members come together to select a project for their village this can bring cohesion among the groups and tribes and can create harmony amongst different individuals.‖ This research has confirmed that the NSP is much more successful at building social cohesion in the community than NGO led initiatives. 83% of facilitating partners believe that the NSP does a better job than traditional NGO projects at building social cohesion and 17% believe that both forms of development perform equally. More importantly, not a single facilitating partner believed that NSP did a worse job than NGO projects at building social cohesion. Conclusions Tim Kessler in the World Bank‘s 2004 World Development Report noted that only about 24% of CDD water projects were sustainable. Hence, sustainability has a long way to go in community driven development initiatives. The Aga Khan has noted that three essential conditions must be achieved for human development initiatives to be sustainable. First, an environment that invests in, rather than seeks to stifle, pluralism and diversity should be a requirement. Second, an extensive and engaged civil society must be present. Thirdly, stable and competent democratic governance should be encouraged. These three conditions are mutually reinforcing. Taken together, they allow developing societies gradually to become masters of the process and make that process self sustainable. These are all three notions that the NSP seeks to achieve by encouraging social cohesion within communities, building community development council with supporting committees and encouraging local governance in villages across Afghanistan.

The research has provided several recommendations for aspects of the NSP which are currently not sustainable in the above analysis. The academic question still outstanding is whether development programmes should be initiated even if they are not considered Page 18

sustainable. For example, should the World Bank have considered other programmes for implementation if they knew at the initial launch of the NSP that it was not sustainable? Should the World Bank have ensured the sustainability of the NSP before launching it in 2003? Should development programmes be implemented if they are not sustainable? These are all important questions that the World Bank and other governments will be considering before launching community driven development initiatives in the future.

Although sustainability is quite important with the NSP and other CDD programmes the important question should be whether the NSP is meeting the objectives that were set out for it. By nature, development programmes are not sustainable and CDD is just scaffolding. Food rations in the United States for example are not sustainable but it is an acceptable part of doing business. Even some CDD programmes that have been set up in the past have been either unsustainable or major failures such as in East Timor where the CDD failed miserably. CDD is nothing more than a mechanism to get money out quickly and be effective in setting up a local government so that when the scaffolding is pulled away the government structure stays in that form. Afghanistan is not like Pakistan where the army can be used to setup these structures. The objectives of the NSP have been to develop infrastructure and instill local governance. If this was the criteria that the NSP was measured against, then it can be considered successful. Sustainability and scalability will have to be worked into the programme. Even the MRRD considers the biggest hurdle in the future of the programme to be its sustainability of funding. In the first 3 phases of the NSP less than $1 billion USD been required to implement the programme. The recent announcement that over $1 trillion USD of mineral wealth exists in the southern and eastern regions of Afghanistan provides another major source of funding for the long term sustainability of the NSP. Although it will take decades for Afghanistan to exploit its mineral wealth fully, the untapped wealth will likely provide more incentives for Afghans to work together.

Lastly, this research asked facilitating partners what percent of facilitating partner employees and community members were happy with the results of the NSP. Overall 85% or facilitating partner employees and 84% of community members have reported being happy with the results of the programme. These results are significantly positive and only reinforce the fact that the programme should be expanded to newer communities but also supported strongly where it is already in existence.

Page 19

Bibliography African Development Bank. 2003. Summary of the Environmental and Social Impact Assessment Study of the Second National Fadama Development Project. African Development Bank. Accessed on the World Wide Web from: http://www.afdb.org/fileadmin/uploads/afdb/Documents/Environmental-and-SocialAssessments/ADF-BD-IF-2003-99-EN-NIGERIA-FADAMA.PDF Ana María Ibáñez, 2002. The Social Impact of Social Funds in Jamaica: A Mixed-Methods Analysis. Ariana, 2010. Tougher Taliban in western Uruzgan. Accessed on the World Wide Web from: http://web.eariana.com/ariana/eariana.nsf/7eca40d3d348841987256b79007d1cb8/e8fd3e978e5f546987 25736400342a23!OpenDocument Ashley, C. and Carney, D. (1999), ―Sustainable Livelihoods: Lessons From Early Experience‖ (DfID), p.1 Asian Development Bank. 2008. Community-Driven Development and Strategy 2020. Accessed on the World Wide Web from: http://www.adb.org/Documents/Participation/Brochure-CDD.pdf Asim Ijaz, K. 2001. ―Can Good Projects succeed in Bad Communities? Collective Action in the Himalayas.‖ Harvard University. Department of Economics, Cambridge, Mass. Binswanger, H., and S. Aiyar. 2003. Scaling up community driven development: Theoretical underpinnings and program design implications. Draft. World Bank, Washington, D.C. Brick, J. 2008. ―Investigating the Sustainability of Community Development Councils in Afghanistan,‖. Accessed on the World Wide Web from: http://polisci.wisc.edu/users/brick/files/brick-cv.pdf, 10 (accessed 15 November 2008). This section draws heavily on the material in this report. CARE Workshop in Atlanta, August 2002, ―A Rights-Based Approach to Achieving Household Livelihoods Security‖ from Drinkwater, M. (2001), ―Programming Principles for a Rights-Based Approach‖ Centre for Policy and Human Development. 2007. Annual Report. Accessed on the World Wide Web from: http://planipolis.iiep.unesco.org/upload/Afghanistan/Afghanistan_hdr2007.pdf CIDA. 2007. Review of the Afghanistan Program Final Version. Accessed on the World Wide Web from: http://www.acdicida.gc.ca/inet/images.nsf/vLUImages/Afghanistan/$file/RapportAfghanistanSept07e.pdfhttp://www.acdicida.gc.ca/inet/images.nsf/vLUImages/Afghanistan/$file/RapportAfghanistanSept07-e.pdf Dixon, J., Gulliver, A., Gibbon, D. (2001), ―Farming Systems and Poverty‖, FAO Dongier, P., Domelen, J. V., Ostrom, E., Rizvi, A., Wakeman, W., Bebbington, A, Alkire, S., Esmail, T., and Polski, M. 2002. Chapter 9: Community Driven Devlopment. World Bank. Accessed on the World Whttp://siteresources.worldbank.org/INTPRS1/Resources/3836061205334112622/5805_chap9.pdf Page 20

El-Gammal, Y. 2004. Yemen: Social Fund for Development III Project. MfDR Principles in Action: Sourcebook on Emerging Good Practices. Accessed on the World Wide Web from: http://www.mfdr.org/sourcebook/1stEdition/6-6Yemen-SocialFund.pdf Ellis and Biggs 2001. Evolving themes in rural development. Development Policy Review 19 (4), pp. 437-448 FAO. 2010. Annex 1: Profile of the Main Development Approaches in the Study. Accessed on the World Wide Web from: http://www.fao.org/docrep/007/j3137e/j3137e0c.htm Frances, C. 1999. ―Paradoxes of Participation: Questioning Participation Approaches to Development.‖ Journal of International Development 11(4):597-612 Ghani, C. and Lockart, C. 2008. Fixing Failed States (New York: Oxford University Press, 2008), 205 Gall, C. 2010. In Afghanistan's North, Ex-Warlord Offers Security. Accessed on the World Wide Web from: http://www.nytimes.com/2010/05/18/world/asia/18mazar.htm Gillespie, S. 2004. SCALING UP COMMUNITY-DRIVEN DEVELOPMENT: A SYNTHESIS OF EXPERIENCE. The World Bank. Accessed on the World Wide Web from: http://www.expandnet.net/PDFs/Gillespie.pdf Guggenheim, S. 2004. Indonesia’ s Kecamatan Development Program: A Large-Scale Use of Community Development to Reduce Poverty. Accessed on the World Wide Web from: http://info.worldbank.org/etools/docs/reducingpoverty/case/88/fullcase/Indonesia%20KDP%2 0Full%20Case.pdf International Development Association. 2009. Community-Driven Development: Delivering the Results People Need. World Bank. Accessed on the World Wide Web from: http://siteresources.worldbank.org/IDA/Resources/IDA-CDD.pdf International Fund for Agriculture Development. 2010. Community Driven Development: IFAD's engagement in community-driven development. Accessed on the World Wide Web from: http://www.ifad.org/english/cdd/index.htm Katz, Travis and Jennifer Sara. 1997. ―Making Rural Water Supply Sustainable: Recommendations from a Global Study.‖ Water and Sanitation Program , Washington, D.C. Kleemier, E. 2000. ―The impact of Participation on Sustainability: An Analysis of the Malawi Rural Piped Scheme Program.‖ World Development 28(5):929-44 Mansuri, G., and Vijeyendra Rao. 2004. Community-Based and -Driven Development: A Critical Review. Policy Research Working Paper No. 3209. Washington, DC: World Bank. February. McNiel, M., and K. Keuhnast, with Anna O‘Donnel. 2004. Assessing Capacity for Community Based Development – A Pilot Study in Tajikistan. Washington, DC: World Bank Institute. December. Nagl, John A., Exum, A., Humayun, A. 2009. ―A Pathway to Success in Afghanistan: The National Solidarity Program‖. Center for New American Security. March 2009. Page 21

Naidoo, K. and Finn, V. 2001. From Impossibility to Reality: A reflection and position paper on the civicus index of civil society project 1999-2001. CIVICUS World Alliance for citizenship Participation. Accessed on the World Wide Web from: http://www.civicus.org/new/media/CIVICUSPositionPaper1.pdf Perez de Castillo, C. 1998. Economic Analysis of Social Investment Fund Projects. Accessed on the World Wide Web from: http://siteresources.worldbank.org/INTCDD/Resources/EA.pdf Rawlings, L.B., L. Sherburne-Benze, and J. Van Domelen. 2004. Evaluating Social Funds – A Cross-Country Analysis of Community Investments. Regional and Sector Studies. Washington, DC: World Bank. Sandstrom, E. 2003. Afghanistan and Central Asia after September 11 – the SecurityPolitical Development. Accessed on the World Wide Web from: http://www.asia.foi.se/pdf/afghanistan-and-central-asia-2003-03-18.pdf Tanaka, S. 2006. A review of community-driven development and its applications to the Asian Development Bank. Asian Development Bank. Accessed on the World Wide Web from: http://www.adb.org/Documents/Participation/Review-CDD-Application-ADB.pdf Toulmin, C. (1994) Gestion de Terroirs: Concept and Development World Bank. 2003. PROJECT APPRAISAL DOCUMENT ON A PROPOSED CREDIT IN THE AMOUNT OF SDR 114 MILLION (US$150 MILLION EQUIVALENT) TO INDIA FOR THE ANDHRA PRADESH RURAL POVERTY REDUCTION PROJECT. The World Bank. Accessed on the World Wide Web from: http://www.rd.ap.gov.in/velugu/pdf/velugupad.pdf World Bank OED (World Bank Operations Evaluation Department). 2005. The Effectiveness of World Bank Support for Community-Based and -Driven Development, An OED Evaluation. Washington, DC. Accessed on the World Wide Web from: http://siteresources.worldbank.org/EXTEFFWBSUPCOMDRIDEV/Resources/cbdcdd_compl ete_report.pdf World Bank. 2006. Institutional Enhancement and Sustainability. Accessed on the World Wide Web from: http://siteresources.worldbank.org/EXTEFFWBSUPCOMDRIDEV/Resources/cbdcdd_chapte r4.pdf World Bank Annual Report. 2009. Accessed from the World Wide Web from: http://issuu.com/world.bank.publications/docs/9780821379462 WCED (World Commission on Environment and Development) (1987). Our common future. New York: Oxford University Press. Zakaria, F. 1999. Beyond Money: Development as Freedom. The New York Times. Accessed on the World Wide Web from : http://www.nytimes.com/1999/11/28/books/beyondmoney.html Zoellick, R. ―The Key to Rebuilding Afghanistan‖ Washington Post (22 August 2008)

Page 22

Annex The population and geographical area of the 34 provinces in Afghanistan sourced from the Ministry of Rural Rehabilitation and Development.

Province Badakhshan Badghis Baghlan Balkh Bamiyan Daykundi Farah Faryab Ghazni Ghor Helmand Herat Jowzjan Kabul Kandahar Kapisa Khost Kunar Kunduz Laghman Logar Nangarhar Nimruz Nurestan Orūzgān Paktia Paktika Panjshir Parwan Samangan Sar-e Pol Takhar Wardak Zabul

Population 819,396 499,393 741,690 1,123,948 343,892 477,544 493,007 833,724 1,080,843 635,302 1,441,769 1,762,157 426,987 3,314,000 913,000 358,268 638,849 413,008 820,000 382,280 322,704 1,342,514 117,991 130,964 320,589 415,000 809,772 128,620 491,870 378,000 442,261 830,319 529,343 244,899

Area (km²) 44,059 20,591 21,118 17,249 14,175 8,088 48,471 20,293 22,915 36,479 58,584 54,778 11,798 4,462 54,022 1,842 4,152 4,942 8,040 3,843 3,880 7,727 41,005 9,225 22,696 6,432 19,482 3,610 5,974 11,262 15,999 12,333 9,934 17,343

Page 23