CARERE/Seila - foundation stone for a new Cambodia compiled

In the early 1990s Cambodia was fractured by violence, lack of social trust and deep suspicion of government. From origins as a post-conflict refugee/IDP repatriation and resettlement scheme, the CARERE programme has evolved into a unique development process which mobilises local and international actors to tackle poverty and promote good governance in rural Cambodia.

Following an internationally-brokered peace agreement in 1991, UNDP launched the Cambodian Resettlement and Reintegration initiative to provide immediate post-crisis reconstruction and livelihoods support to rural communities of Cambodia where 85% of the population lives.(1) CARERE initially sought to produce visible post-crisis improvements in communities most directly affected by the influx of IDPs and returnees from refugee camps in Thailand. Quick Impact Projects (QIPs) provided tangible benefits for local communities - roads, schools and wells - built, wherever possible, using local contractors and local labour who received on-the-job skills training. Though UNDP was aware of the need for a longer-term strategic vision, the first phase of CARERE aimed to maximise short-term benefits for the rural poor during the 'settling in' period following the peace accord.

Initial targeting of villages was determined through a joint UNDP/UNHCR planning process based on returnee data, indicating that a substantial majority would return to the four north-western provinces where CARERE opened operations. While access to some areas was limited by the continued presence of the Khmer Rouge, in later years opportunities for expansion of CARERE reconciliation and reconstruction efforts arose as their cadres defected to the government.

Following Cambodia's first democratic elections in 1993, CARERE began to confront the inherent unsustainability of a short-term, external implementation approach to rebuilding rural Cambodia. CARERE was reconceived as an experiment in decentralised, participatory planning, financing and implementation of local development with a marked shift toward local capacity building and a transfer of responsibilities to Cambodians themselves.

CARERE was renamed the Cambodia Area Rehabilitation and Regeneration programme and eventually became a support programme to the Cambodian government decentralisation initiative entitled Seila ('foundation stone' in Khmer Sanskrit). Seila was an outgrowth of CARERE2 and has been an expression of national ownership of the principles and processes inherent in a decentralised approach to rural development. Seila, a collective undertaking of seven national ministries, has worked to break down barriers between ministries and promote joined-up, integrated government.

CARERE2 provides intensive capacity building and investment resources but it is the national and local Seila structure, backed by government decree, that plans and implements development activities. Whereas CARERE1 was short-term, materially oriented, flexibly responsive to local emergency needs but time-limited in vision, CARERE2 involved planning and financing for the long term, oriented towards local human resource development and good local governance.

Learning by doing and designing by using

CARERE/Seila is a complex and ambitious attempt to bridge the divide between emergency assistance and development. The key operational device has been to constantly assess, learn, reflect, revise and adapt to the issues and challenges that emerge. Guided by four principles - dialogue, clarity, agreement and respect - Seila aims to create partnership between government and civil society. The idea is that decentralised and participatory rural development will lay a foundation for peace and socio-economic improvement. The goal of strengthening the coping capacities of the rural poor has been advanced by mobilising critical local economic inputs (including contractors and labour) and bottom-up integration of local priorities into national planning and resource allocation processes.

Seila has prioritised both non-material and material aspects of poverty. During the first five years of implementation, 1996-2000, Seila piloted and strengthened new systems for decentralised and deconcentrated planning, financing and implementation in a third of the country's provinces and communes. The north-eastern province of Ratanakiri was added to the original four (Banteay Meanchey, Battambang, Siem Reap and Pursat) because of its high concentration of under-served ethnic minorities, over-exploited natural resources and acute levels of poverty and isolation. Considerable emphasis was placed on the election of Village Development Committees (VDCs) to take the place of the top-down village leadership to which Cambodians were accustomed. Special attention was given to ensuring the inclusion of women through a quota system.

Seila has worked on many fronts: institutional and community capacity building, delivery of services and investments, promoting democratic participation, fostering peace and reconciliation, poverty alleviation, support to the private sector, provision of non-formal education and training in gender awareness. The initial programme document was visionary in character and rather vague on practical details. Being experimental and adaptive, however, has necessitated a reflexive approach to changing strategies and policies. Seila was not based on a formal analysis setting out the relationship between various activities and development objectives. It broke with both the management culture within international development organisations and with the perception of typical management and political practices of Cambodian state authorities.

Under Seila's second five-year phase (2001-05), the programme has continued to support the design and implementation of the decentralisation policies. Villages, communes and districts have been assisted to develop locally-owned development plans. The Partnership for Local Governance has become an important component. In addition, CARERE/Seila has been tasked with mobilising and coordinating external development assistance in support of the national decentralisation and deconcentration policies.

External advisors have gradually withdrawn from involvement at district and provincial level and by 2003 the government was managing Seila through appointed government committees at national, provincial and district levels and through elected members at commune level.

Proving the critics wrong

Central level politics that have slowed down many other development projects were avoided by CARERE by the initial focus on province and lower administrative levels. However, at the time this approach was controversial. The vision expected commune officials to change their behaviour from oppressors to agents of participatory development at a time when conventional wisdom among donors was that Cambodia's officials were irredeemably corrupt and lazy and that the entire administration was in moral decline following the breakdown of the previous planned economy and command structures. While this had a certain truth to it, local administrations were nevertheless extended enough trust to be nurtured into being the 'driver' of the new CARERE2/Seila programme. Most observers have been surprised by the readiness of many commune and province officials to embrace opportunities provided by training, to change working practices and to take pride in becoming efficient and accountable managers.

For a while, the degree of genuineness of the CARERE2/Seila-established VDCs was one of the most debated development issues in Cambodia - and in general development literature. Critics have argued that the participatory process has at times been shallow. Women have been poorly represented in the VDCs, the selection of candidates has often been steered, voter turnout low and projects were sometimes of relatively little importance for the village as a whole.

In general, however, although promotion of participation was not perfect and required high initial investment in technical assistance, it has had desirable long-term spin-offs. For the most part VDCs have been popularly elected, have been (semi-) independent from political manoeuvring and generally managed to perform their tasks. Their establishment and introduction of bottom-up decision-making has not caused major turmoil to either government officials or to the social structures of Khmer villages.

CARERE/Seila was also criticised for not taking enough account of maintenance of infrastructure and establishing designs and standards without coordination with other agencies. A more controversial and serious critique relates to the accusation that it has created parallel structures and has not been working through the existing governmental institutions. However, although new horizontal accountability and reporting structures were created they were at the same time under the control of the government authorities. These structures have now become the core in the provincial structural reforms which need to be pushed to completion.

Jealousies flourish in any donor community and the success of CARERE/Seila has prompted much sniping. Many point out that the programme is nowhere near sustainability - at least not when using the traditional definition of 'sustainability' as capacity for a national initiative to continue after the exit of the external assistance, or capacity to generate internal capital resources to replace the investment funds for local development coming from outside sources. However, in reality, hardly any major development projects in chronically poor Cambodia are sustainable in that sense. If sustainability is to be measured by the degree to which it has acquired a vital momentum of its own and is driven by Cambodian authorities then it is highly sustainable. The degree of long-term commitment shown by CARERE's major external donors is a testament to the sustaining power of a shared experiment that has produced sustained changes in local and national governance.

Sustainable reintegration: never a quick fix

After more than a decade of working with Cambodians to rebuild their ravaged country, CARERE has demonstrated how reflection, recognition of mistakes and willingness to reformulate objectives are required in order to constantly refocus activities to make reconciliation and reintegration sustainable. Time has been required not just to establish local confidence in the process but to convince a sceptical donor community to go on funding a costly experimental venture which lacked a detailed plan, could not produce quantifiable evidence of poverty reduction, nor for many years produce any material evidence of developmental 'results'.

A history of this major, well-funded and well-documented international initiative shows the importance of open-mindedness and change. Managers recognised that:

  • Initial failure to engage with local populations made it necessary to reformulate CARERE2 objectives to emphasise the importance of democratically electing commune councils in local areas.
  • Earlier preoccupation with poverty alleviation had to give way to acknowledging the equal importance of promoting reconciliation, especially in areas long controlled by the Khmer Rouge.
  • Sustainable poverty alleviation cannot be achieved without the improvement of local governance.
  • The kind of staff most successful at forging relations with provincial and commune counterparts were not regular UN experts but a combination of international, national and Cambodian expatriate staff with a broad range of development experience, strong knowledge of local customs and history, fluency in the local language and commitment to recognising and building the capacities of Cambodians.


CARERE/Seila's departures from prevailing paradigms can be summarised. The programme:

  • did not include a detailed and concrete implementation plan in the original project document and did not use the logical framework format
  • was in constant transition, operated in a policy void and yet was able to influence the development of policy
  • used monitoring and evaluation systems which enabled input from national, local and community actors
  • dared to trust people who did not have a credible record for this kind of operation - within the Cambodian state machinery as well as among expatriate staff
  • assumed that deep-seated cultural traits could be partly reversed, or at least managed
  • operated in the fuzzy midfield between politics and development activities
  • has made the difficult transition from an emergency to a development approach
  • worked with invisibles or softwares as development objectives
  • spent resources on initial activities which many stakeholders, donors and observers judged to be unnecessary.


s CARERE/Seila has successfully straddled the divide between the emergency and development phases of a complex post-conflict environment, its key achievements have been to:

  • create the conditions for broad-based participation among local and subnational authorities in a way which has not posed threats to the central government
  • reduce funding uncertainties by ensuring that predictable development investment funds are available, for which province, commune and village administrations are held responsible
  • ensure that rural roads, schools, water supply schemes and irrigation have been designed on the basis of local assessments and that benefits have been spread to marginalised socio-economic groups
  • foster attitudinal change among local officials: the civil administration has been transformed from one uninterested in development and plagued by inefficiency to one staffed by decently-educated technocrats concerned with administering bottom-up processes and good governance
  • confound those who believed that values of democracy and development are alien to Cambodian society.


Seila is scheduled to end in 2005. It has done much to promote empowerment, transparency and accountability and local democratic practices are slowly emerging at all levels of administration throughout Cambodia. However, major problems remain in what is still a desperately poor country:

  • Chronic under-employment and unemployment in rural areas are exacerbated by cuts in civil service employment.
  • High infant and maternal mortality rates, deaths from preventable diseases, the prevalence of water-borne diseases, malaria and TB and the spread of HIV/AIDS make Cambodia one of the unhealthiest countries in the developing world.
  • Despite remarkable achievements in the education system - upon which the success of sustainable reintegration depends - remains in crisis: qualified teachers are in short supply, the low quality and the lack of relevance of education lead to high repetition and dropout rates and to inadequate levels of achievement and many rural schools are in an advanced state of disrepair.
  • Seila was set up before the legislative framework for decentralisation and many aspects of deconcentration of government responsibilities remain unclear.
  • Commune councils - which were democratically elected in 2002 - control few resources and have had very few service delivery functions devolved to them.
  • Institutional arrangements for achieving more effective aid coordination for local government reform have not been finalised.
  • Local officials are still required to account upwards to the higher levels in the ruling political party and administration, rather than downwards to the communes and the public.
  • In the aftermath of genocide and selective emigration of men, it is estimated that women comprise 56% of Cambodia's population, yet they remain grossly under-represented in decision-making fora.
  • Donor support to local government is often inconsistent and conflicting.
  • Capacity building of commune councils is not necessarily based on an in-depth assessment of the effectiveness and impact of past training: thorough training in good governance principles is required to enable councillors to understand roles, functions, requirements and visions for development.


Is CARERE/Seila replicable?

The achievements of CARERE/Seila stand out against a background of many failed reconstruction initiatives elsewhere. Regular evaluations of the programme and retrospective analysis of its evolution suggest the need for reintegration programmes in other fragmented failed states to:

  • be driven by a central vision that only harmonious and respectful relations between the state and civil society can promote prospects for alleviating poverty
  • ensure a strong sense of ownership of reintegration and development processes at sub-national and local levels
  • deliver a material product: a programme developing and establishing concepts and systems without an operational content making material changes to peoples' daily lives will have much less potential to make changes and impact
  • question familiar procedures and reject ready-made models: CARERE2/Seila has rarely used preconceived models, systems and structures but rather developed the models with the stakeholder to fit local realities
  • encourage donors to stay the course, commit themselves for ten years and explicitly recognise the difficulties of making predictions and prognoses: agencies should not overstress the need for specific statements of the impact and expected outcomes of a project
  • ensure a high degree of communication between stakeholders: without this, tensions between conservatively-inclined donors and CARERE2/Seila management could have led to destructive tension.


In post-conflict countries such as Cambodia - with a legacy of weak political and administrative structure, repressive central government and inability to generate funds for reintegration and reconstruction - the role of donors and their influence on the scope and approach to development has necessarily been strong. The reflexive approach to changing strategies and policies has at times led to 'change fatigue' but CARERE/Seila's experimental approach would not have been credible - and probably not successful - without constant change. CARERE's achievements challenge the belief that decentralisation can only work in an environment where central government is already strong. A pure technical assistance approach to decentralisation can have limitations, but when coupled with capital injections at local level and a 'learning by doing' mentality tangible benefits can result as local capacity is built.


Assistance in writing this article has been provided by Scott Leiper (Programme Coordinator for CARERE's Partnership for Local Governance Project. Email: and Judith Karl (Senior Adviser, UNDP's Bureau for Crisis Prevention and Recovery, Geneva. Email: Neither is responsible for the views expressed.

For more information, see the Seila and CARERE sites at and and evaluation reports at and


  1. CARERE has been funded through UNDP core resources, international donors (particularly Sweden, Norway and the Netherlands), and benefited from the close partnership of the UN Capital Development Fund (UNCDF) through it's innovative Local Development Fund facility. The programme has been executed by the UN Office for Project Services (UNOPS) with related components executed by the International Labour Organisation (ILO).


Opinions in FMR do not necessarily reflect the views of the Editors, the Refugee Studies Centre or the University of Oxford.
FMR is an Open Access publication. Users are free to read, download, copy, distribute, print or link to the full texts of articles published in FMR and on the FMR website, as long as the use is for non-commercial purposes and the author and FMR are attributed. Unless otherwise indicated, all articles published in FMR in print and online, and FMR itself, are licensed under a Creative Commons Attribution-NonCommercial-NoDerivs (CC BY-NC-ND) licence. Details at