Rehabilitating health services in Cambodia: the challenge of coordination in chronic political emergencies

Transcription

1 HEALTH POLICY AND PLANNING; 14(3): Oxford University Press 1999 Rehabilitating health services in Cambodia: the challenge of coordination in chronic political emergencies STEVEN LANJOUW, 1 JOANNA MACRAE 2 AND ANTHONY B ZWI 1 1 Health Policy Unit, London School of Hygiene and Tropical Medicine, UK and 2 Humanitarian Policy Group, Overseas Development Institute, London, UK The end of the Cold War brought with it opportunities to resolve a number of conflicts around the world, including those in Angola, Cambodia, El Salvador and Mozambique. International political efforts to negotiate peace in these countries were accompanied by significant aid programmes ostensibly designed to redress the worst effects of conflict and to contribute to the consolidation of peace. Such periods of political transition, and associated aid inflows, constitute an opportunity to improve health services in countries whose health indicators have been among the worst in the world and where access to basic health services is significantly diminished by war. This paper analyzes the particular constraints to effective coordination of health sector aid in situations of post -conflict transition. These include: the uncertain legitimacy and competence of state structures; donor choice of implementing channels; and actions by national and international political actors which served to undermine coordination mechanisms in order to further their respective agendas. These obstacles hindered efforts by health professionals to establish an effective coordination regime, for example, through NGO mapping and the establishment of aid coordinating committees at national and provincial levels. These technical measures were unable to address the basic constitutional question of who had the authority to determine the distribution of scarce resources during a period of transition in political authority. The peculiar difficulties of establishing effective coordination mechanisms are important to address if the long-term effectiveness of rehabilitation aid is to be enhanced. Introduction In the wake of peace settlements in countries from Bosnia to Mozambique, and from Cambodia to the Lebanon, the 1990s has seen a growing interest internationally in the concept of post -conflict rehabilitation. This interest has been marked by, for example, the creation of a new Post-Conflict Unit within the World Bank responsible for monitoring and influencing the policy response in situations of war peace transition (World Bank, 1997a; Holtzman, 1997), 1 and within the United Nations, where an Executive Committee on Peace and Security (ECPS) has been established and is responsible for coordinating the efforts of specialized agencies in post - conflict situations (United Nations, 1997). Debates regarding the role of aid in post -conflict situations can be seen as part of a wider concern within the aid community to enhance the links between relief and development aid (see, for example, European Commission, 1996; Development Assistance Committee, 1997). Countries emerging from periods of violent political conflict typically face a number of substantive health and health system challenges. The negative effects of conflict on health and health systems have been documented elsewhere (see Zwi & Ugalde, 1989; Zwi 1996; Levy & Sidel, 1997). The effects of injury and increased exposure to disease as a result of increased poverty and displacement are not quickly reversed. Addressing the direct and indirect effects of war on health and health systems therefore represents a major challenge for those concerned with post -conflict rehabilitation. Alongside these challenges are also potential opportunities. International political support for negotiated settlements, such as that in Cambodia, is typically accompanied by pledges of significant financial support. However, while this may be of great value in the aftermath of so much destruction and economic disruption, this aid is disbursed in a highly charged and unstable political environment. In such settings, attempts to coordinate aid for the health sector, a complex issue at the best of times (see Buse & Walt 1997), are simultaneously more pressing than is usual and confront particular difficulties. They are especially urgent and necessary given that the needs are particularly acute, the total volume of resources small (although they often represent a substantial increase over prior resource flows) and, as described below, the channels for implementing aid are more diffuse and complex. They are more difficult to manage because of the competing political agendas and the lack of a mandated authority to mediate these political decisions and enforce policy coherence. This paper examines the context and process of health aid coordination in Cambodia between It focuses on the relationship between national political actors and the many international agencies who came to work in the country

2 230 Steven Lanjouw et al. during this period. It complements other analyses, including those from key actors, regarding the experiences faced by Cambodia during this period (see Heng & Key, 1995, for example). The case study highlights two key issues. First, political factors influenced greatly the type of aid provided to the country during the period , and the way in which it was managed. These structural, political factors constrained the scope for health professionals to establish effective coordination mechanisms. Second, the limited capacity of technical coordination mechanisms to influence resource allocation contributed to the emergence of a pattern of rehabilitation investment which was highly fragmented, did not correspond fully with national health priorities, and has proved difficult to sustain. Overcoming such difficulties in other similar situations will depend upon donor organizations, in particular, adopting more coherent strategies and appropriate aid instruments to support rehabilitation measures. The remainder of the paper is divided into three main sections: a brief historical overview of Cambodia and the transition period; a description of the main mechanisms of aid coordination operating in the Cambodian health sector during the transition period; and, finally, an assessment of the implications of these findings for other post -conflict settings. The context: a political history of aid Historical overview Cambodia s turbulent and destructive history over the past few decades is summarized in Figure 1. The Kingdom of Cambodia ( ) under Prince Sihanouk was replaced by that of Lon Nol following a coup with the support of the USA. Throughout this period, Cambodia was dragged inexorably into the Vietnam war, suffering widespread carpet-bombing, which led to massive displacement and contributed to the country s own political instability. The Khmer Rouge overthrew the Lon Nol regime and introduced a radical attempt to return Cambodia to its early forms of agriculture and traditional society. During this infamous period of the killing fields, intellectuals and minority groups were systematically annihilated. Many others died as a result of disease, malnutrition and summary violence. At least half a million people, up to as many as two million, are estimated to have died during this period. The Khmer Rouge was overthrown in 1979 with the support of the Vietnamese who established the People s Republic of Kampuchea (PRK), which lasted for 10 years. During this period Vietnamese style governance and development were introduced and substantive support from the Eastern Bloc was obtained. War continued with the Khmer Rouge, the Royalists and other factions opposed to the Vietnamese-backed regime, seeking to destabilize the country through low-intensity incursions and guerilla warfare. These factions, based in the refugee camps along the Thai-Cambodian border, were actively supported by the USA, China and Thailand, among others. It was these factions, under the heading of the Coalition Government of Democratic Kampuchea, who occupied Cambodia s seat in the United Nations from 1982 onwards, constituting an effective government in exile. The Vietnamese-backed regime in Phnom Penh, which controlled the vast amount of territory and redeveloped the machinery of state (such as the public administration), was recognized by very few governments, mostly those associated with the Eastern Bloc from whom it drew financial and political support. Otherwise, the PRK, as the country was known, was subject to international sanctions, blocking aid and trade relations with the country (Mysliwiec, 1988). In 1990, marking an internal process of economic and political liberalization, the PRK became the State of Cambodia (SOC). With the backing of the Permanent Members of the United Nations Security Council, negotiations were initiated to end the conflict and facilitate Cambodia s re-entry into the international community. These negotiations culminated in the signing of the Paris Peace Accords on 23 October Aid and war Since independence in 1954, successive Cambodian states had relied upon aid for their survival. Reflecting its political history, the source of aid has swung from the West (particularly from the US) to the East (predominantly Vietnam and the USSR) and back to the West again, in different periods of its political economy. In the wake of the genocidal Khmer Rouge regime, a major emergency operation was initiated by the West in September In early 1982, the United Nations declared that the emergency was over, effectively suspending Western aid flows as development aid was subject to the embargo, pending the withdrawal of the Vietnamese from the country. The only exception to this was a small group of Western NGOs and UNICEF which continued their advocacy work and some development work in the country, funded from private and, in the case of UNICEF, assessed contributions to the UN (Mysliwiec, 1994). Between 1981 and 1989, Vietnam and the USSR were the major financial backers of the Cambodian state. As the peace negotiations progressed, and the process of political and economic liberalization continued within Cambodia, so the scope for international aid action increased. Externally, donor agencies such as the European Commission and USAID started to fund NGO-led work inside Cambodia. Internally, the Phnom Penh Government allowed NGOs to move out of the capital and in to the regions. International organizations, such as UNHCR, UNDP and WHO, sent numerous fact-finding missions to the country in the lead up to the deployment of the United Nations Transitional Authority. In search of legitimate coordination: UNTAC, SNC, TAC and other good intentions The legitimacy dilemma The concept of public policy implies the presence of an authority with the authority and the legitimacy to represent (or at least claim to represent) the interests of a population, and

3 Coordination in post -conflict Cambodia 231 Figure 1. Cambodia s political evolution since independence with the technical and administrative competence to implement that policy. In terms of national health policy, this implies control over territory, usually defined in terms of internationally accepted borders. In other words, national policy, specifically national public health policy, implies the existence of a competent and legitimate state, if not to provide services at least to finance and regulate them (Macrae, 1998; Cassels, 1992). This idealized strong state, with its functioning Ministry of Health firmly in control of policy-making, resource allocation and regulation of the sector, is increasingly absent in many parts of the Third World. Instead, many developing countries, particularly in Africa, are characterized by states which lack the capacity, and in some cases the political will, to function as sovereign states (Clapham, 1996). The weakness of public policy in these countries, together with the current preference of official aid organizations for policy-based lending (LaFond, 1995), means that the locus of health policy-making is increasingly internationalized with decisions regarding major elements of the content of health policy in recipient countries frequently being made in Washington, Copenhagen and London, rather than in national capitals (Duffield, 1991; Clapham, 1996; Okounzi & Macrae, 1995).

4 232 Steven Lanjouw et al. While the capacity of national governments to manage international pressures on health policy may be diminishing in many countries, the fact that they are sovereign and recognized internationally matters (Clapham, 1996; Reno, 1999). At least in theory, states remain the final arbiter of public policy, since they alone have the legal power to determine who operates within the confines of their national borders. In situations of ongoing conflict, and in situations of post - conflict transition, the legitimacy and authority of the state may be contested by national and international actors. This presents particular problems for aid actors who wish to neither support nor reinforce existing state structures which are also seen as belligerent parties to a conflict. For this reason, it is rare that development aid is provided by the international community; rather, relief aid, delivered outside state structures, is the primary tool of engagement in countries such as Sudan, Somalia and Afghanistan. In situations of post -conflict transition, such as that in Cambodia between , the problem of finding a legitimate interlocuter for aid is somewhat different. Here, there was a political process, marked by the signing of a peace accord, demobilization of troops and culminating in an election. Until the elected government was in place there was a vacuum in terms of the political authority recognized by the international powers. Coinciding with the political continuum from war to peace was an aid continuum. Prior to the signing of the Paris Accords, relief had been the only form of aid not subject to the embargo, and this had been virtually suspended after During the interim period of the transition, rehabilitation aid was provided. It was only after the elections that normal development-aid relations could be resumed, marked most importantly by the establishment of World Bank and IMF lending. As indicated above, there is currently considerable debate internationally regarding the need to link relief and development aid, and in particular to ensure that relief measures achieve developmental objectives and contribute to longterm sustainability. A problematic dimension to effecting a transition from relief to rehabilitation to development, however, is that one key condition for the resumption of development aid is an engagement by the aid community with the very state whose legitimacy is in question. When the legitimacy and authority of the state remains contested nationally and internationally, it is unclear who has the mandate and power to set priorities and to distribute scarce resources. billions of riels - constant 1989 prices Figure 2. Cambodian Ministry of Health recurrent expenditure in billions of riels in 1989 constant prices 1991, levels of government public-health expenditure were less than half their value in 1989 (see Figure 2). No reliable figures exist to report accurately international aid expenditure on health during the period , however, data from provide an indication of the absolute and relative value to the Cambodian health sector. Figure 3 shows that aid expenditures exceeded government expenditures nearly four-fold in 1992 and more than seven-fold in How these funds were invested had important implications for determining access to quality care by most of the population. UNTAC, SNC and the TAC: the tangled web for decisionmaking The establishment of the United Nations Transitional Authority in Cambodia (UNTAC) was designed to address the constitutional vacuum which existed between the signing of In Cambodia, the Paris Peace Accords established a framework which sought to respond to this dilemma of legitimacy, and to establish what Meier (1993) has called a constitution for decision-making, a mechanism for deciding how decisions should be made regarding the use of international rehabilitation assistance. Constructing such a mechanism was not simply a diplomatic nicety, but rather had important practical implications. At stake was how the US$800 million of aid pledged in June 1992 by the international community would be spent over the coming 18 months. US$ millions year Cambodian Govt Aid In the health sector in particular, the stakes were high. In Figure 3. Total health expenditures: government and aid

5 Coordination in post -conflict Cambodia 233 the peace accords and the formation of the elected government. Unique in international law, UNTAC drew its legitimacy from a body comprising the four Cambodian factions the Supreme National Council (SNC). The SNC comprised 12 members representing the four factions; Prince Sihanouk, Head of State , was an additional member and served as its President. The SNC embodied the legitimate sovereign state recognized by the United Nations. The SNC delegated all powers necessary to implement the Paris Peace accords to the United Nations, represented by the Special Representative of the Secretary General, Yasushi Akashi. The mandate of UNTAC drew on the peace accords, and comprised seven components: military (including disarming and demobilizing the factions), civilian policing, electoral, civil administration, repatriation (focused on facilitating the return of refugees from Thailand), human rights, economic affairs and rehabilitation (Charney, 1992). The constitutional arrangement set up in Cambodia effectively created a UN Trusteeship, allowing in principle for the United Nations to act as the government of the country. The Paris Peace Accords, included a Declaration on the Rehabilitation and Reconstruction of Cambodia (Box 1). The Rehabilitation Component of UNTAC, with its own Director, Declaration on The Rehabilitation and Reconstruction of Cambodia 1. The primary objective of the reconstruction of Cambodia should be the advancement of the Cambodian nation and people, without discrimination or prejudice, and the full respect for human rights and the fundamental freedom for all. The achievement of this objective requires the full implementation of the comprehensive political settlement. 2. The main responsibility for deciding Cambodia s reconstruction needs and plans should rest with the Cambodian people and the government formed after free and fair elections. No attempt should be made to impose a development strategy on Cambodia from any outside source or deter potential donors from contributing to the reconstruction of Cambodia. 3. International, regional and bilateral assistance to Cambodia should be coordinated as much as possible, complement and supplement local resources and be made available impartially with full regard for Cambodia s sovereignty, priorities, institutional means and absorptive capacity. 4. In the context of the reconstruction efforts economic aid should benefit all areas of Cambodia, especially the more disadvantaged, and reach all levels of society. 5. The implementation of an international aid effort would have to be phased in over a period that realistically acknowledges both political and technical imperatives. It would also necessitate a significant degree of cooperation between the future Cambodian Government and bilateral, regional and international contributors. 6. An important role will be played in rehabilitation and reconstruction by the United Nations system. The launching of an international reconstruction plan and an appeal for contributions should take place at an appropriate time, so as to ensure its success. 7. No effective programme of national reconstruction can be initiated without detailed assessments of Cambodia s human, natural and other economic assets. It will be necessary for a census to be conducted, developmental priorities identified, and the availability of resources, internal and external, determined. To this end there will be scope for sending to Cambodia fact-finding missions from the United Nations system, international financial institutions and other agencies, with the consent of the future Cambodian Government. 8. With the achievement of the comprehensive political settlement, it is now possible and desirable to initiate a process of rehabilitation, addressing immediate needs, and to lay the groundwork for the preparation of medium and long-term reconstruction plans. 9. For this period of rehabilitation, the United Nations Secretary-General is requested to help coordinate the programme guided by a person appointed for this purpose. 10. In this rehabilitation phase, particular attention will need to be given to food security, health, housing, training, education, the transportation network and the restoration of Cambodia s existing basic infrastructure and public utilities. 11. The implementation of a longer-term international development plan for reconstruction should await the formation of a government following the elections and the determination and adoption of its own policies and priorities. 12. This reconstruction phase should promote Cambodian entrepreneurship and make use of the private sector, among other sectors, to help advance self-sustaining economic growth. It would also benefit from regional approaches, involving, inter alia, institutions such as the Economic and Social Commission for Asia and the Pacific (ESCAP) and the Mekong Committee, and Governments within the region; and from participation by non-governmental organizations. 13. In order to harmonize and monitor the contributions that will be made by the international community to the reconstruction of Cambodia after the formation of a government following the elections, a consultative body, to be called the International Committee on the Reconstruction of Cambodia (ICORC), should be set up at an appropriate time and be open to potential donors and other relevant parties. The United Nations Secretary-General is requested to make special arrangements for the United Nations system to support ICORC in its work, notably in ensuring a smooth transition from the rehabilitation to reconstruction phases. Box 1. The Declaration on the Rehabilitation and Reconstruction of Cambodia

6 234 Steven Lanjouw et al. was responsible for this, and was to work according to three key principles laid out in the Declaration to guide the international assistance for the rehabilitation and reconstruction of Cambodia. These were: Sovereignty the Cambodian people and the government to be elected should be primarily responsible for delivering reconstruction assistance; Respect for local capacity external assistance should complement and supplement local resources; Balance assistance should benefit all areas, especially the most disadvantaged. The counterpart to the Rehabilitation Component of UNTAC in the SNC was the Technical Advisory Committee (TAC). The TAC, chaired by the UNTAC Director of Rehabilitation, solicited approval from the factions for all proposed projects and mission proposals, with the exception of those implemented by NGOs. The public administration, highly politicized under the pre- UNTAC regime, nominally fell under the control of UNTAC. The Paris Accords had stipulated that all existing ministries and departments would continue to function during the transition period. UNTAC had responsibility for overseeing the civil service. Its capacity to do so was limited, not only by the ratio of UNTAC personnel to Cambodian civil servants (there were 400 of the former and of the latter), but by the linguistic obstacles facing the international civil servants attempting to shadow and control a bureaucracy effectively run by one of the factions (CPP). This complex machinery provided the backdrop against which coordination in the health sector was to take place. In a context of continued hostility (including military action) and distrust between the factions and their respective political and financial backers, delegating responsibility to a third party (the United Nations) potentially provided a means of neutralizing decision-making. An impartial UN, acting according to the provisions and principles of the Paris Accords, offered a way for deciding how decisions could be made regarding rehabilitation priorities, among other things. UNTAC s capacity to do so in practice was limited, however. Coordination in practice UNTAC did not work as a transitional government, capable of setting and enforcing systemic policy. It did not do so for three main reasons. First, the leadership of the UN mission in Cambodia interpreted its mandate differently. Rather than acting as a temporary sovereign power, which would then hand over power to a newly established regime (rather as had the colonial powers), the Special Representative of the Secretary General (SRSG) sought a more consensual and pragmatic approach. The SRSG saw UNTAC s legitimacy as contingent upon achieving consensus in the SNC. This interpretation assumed both that the SNC could achieve such consensus and that the SNC itself was seen as legitimate by the political factions themselves and could claim to be representative of the interests of the Cambodian people. Second, the factions did not allow UNTAC to function effectively. The Khmer Rouge left the UNTAC/SNC framework early in the proceedings, refusing to comply with the processes of demobilization and preparations for the elections. Other factions, namely FUNCIPEC and the KPNLF were acutely sensitive to the risk that aid might serve to reinforce the advantage of the communist-led State of Cambodia, which remained as the de facto government in much of the country. For example, they blocked agreement on a World Bank loan, despite the fact that this was not to come into force until after the elections. Thus, the Technical Advisory Committee was unable to achieve consensus and thus to provide a practicable focus for aid decision-making during the transitional period. Finally, donor countries did not want to establish a rigorous coordination regime. In principle, UNTAC provided a mechanism for such coordination, and the Paris Accords established criteria for it to allocate resources. That both the UNTAC rehabilitation component and the TAC were weak reflected the (deliberate) under-investment by donors of the financial and political capital which could have provided a competent coordination regime. The questionable legitimacy of the interim administration in Cambodia, dominated as it was by the former government, led many donors to by-pass the existing public administration structures. In the health sector, UNTAC faced an awesome task of getting a sometimes hostile donor community, which wanted to identify its own priorities and the groups to whom aid would be distributed, to prioritize critical needs, prevent duplication, share information and avoid concentrating resources in certain geographic areas of the country. With close to 70 external organizations involved in the health sector by late 1993, some with little or no official recognition and no obligations to locally account for how resources were disbursed, this uncontrolled environment was, to take a common phrase, a free for all. Major donors, in particular the United States and to a lesser extent the European donors, including the European Commission, sought to maximize their room for manoeuvre by working outside the coordination regime they themselves had worked to establish. Specifically, they sought to direct aid to the non-communist factions and to avoid engaging with the State of Cambodia. They did so in two ways. First, they bypassed the United Nations by sub-contracting NGOs directly, rather than working through the UN. Importantly, NGO projects were not subject to scrutiny by the TAC. Donors argued that channelling aid directly through NGOs was necessary in a context where there was a critical lack of institutional capacity and skilled personnel (USAID, 1994). One of Cambodia s primary and perhaps most influential donors, USAID, articulated this approach in May 1994 by stating, Cambodia s unique circumstances dictate the need for an innovative USAID post-crisis transition strategy. The critical lack of institutional capacity, skilled personnel, and local financing dictates an interim USAID strategy

7 Coordination in post -conflict Cambodia 235 of relying largely on NGOs for much of the project implementation (USAID, 1994). Although USAID describes Cambodia s circumstances as unique, donor reliance on NGOs to disburse rehabilitation aid funds is common practice (see Macrae et al., 1995). In the Cambodian case, bilateral relations (or indeed relations with the international financial institutions) could not be resumed until the elected government was in place. At the same time, those multilateral agencies with a major involvement in rehabilitation, such as UNHCR, had limited operational capacity, sub-contracting NGOs to deliver particular services. This strategy led to a mushrooming of international NGOs (see Table 1). In 1991 there were approximately 60 international NGOs in Cambodia. This rose to 87 in 1992 and to 105 in 1993, with many nearly doubling their number of staff in those years. Much of the assistance from these agencies was directed towards Phnom Penh and the north-west provinces, to which the majority of the refugees sought to return. This expansion in the number and activity of NGOs heightened the coordination challenge, as each came with its own assessment of need, strategy for implementation and, in some cases, political agenda. In the health sector, many activities, such as malaria control and essential drugs, remained critically underfunded, while others, such as infrastructure rehabilitation, were funded generously, with pledges exceeding the amounts stated in the appeal. The time-frame for the rehabilitation period was designed to coincide with that of the political process some 18 months. The short project cycle placed pressures on NGOs to identify projects that could deliver visible outputs quickly. While many NGOs undertook training initiatives, these were developed in an ad hoc way. Thus one senior WHO official in Phnom Penh described the training approach adopted by many agencies as highly prescriptive (Lanjouw, 1997). Donors and agencies involved in training often determined the costs in advance of an assessment of training needs, as these training costs were inexorably linked to short-term monies and budgets. A second donor strategy to circumvent the UNTAC/SNC structure was to negotiate directly with individual factions for access, rather than through the TAC. In so doing, they jeopardized not only the principles of neutrality and impartiality of the Accords, but also the mandate of the United Nations to uphold such principles. In addition to the primary, statutory coordination mechanisms, a number of ad hoc coordination bodies were established. These included the Donor Consultative Group, a permanent donor consortium for Cambodia (ICORC) which met outside the country, and the Joint Support Unit (JSU). The latter body, chaired by UNHCR and UNDP, was established to coordinate overall humanitarian and rehabilitation assistance, and specifically repatriation and resettlement assistance. It included representatives of bilateral governments, UN specialized agencies and three NGO representatives. Supporting the JSU was a Joint Technical Management Unit, which provided the Secretariat for the JSU and a network of offices at the Provincial level, primarily in the north-west. These Provincial Support Units (PSUs) comprised technical and management staff able to launch, at district or community level, projects of simple and standard design having immediate and quick impact. PSUs aimed to raise the capacity of district and provincial authorities to respond to the needs of repatriated refugees. The first PSUs were financed by UNDP and UNHCR, and were to be the forerunners of the later UNICEF and WHO-financed provincial health advisors (see Hirono, 1992). The JSU and PSUs worked closely with the growing NGO community. The latter had its own coordination mechanism, the Cooperation Committee of Cambodia (CCC). Aiding recovery? Health coordination mechanisms, The mechanisms developed for health sector coordination largely mirrored that for overall aid coordination. Thus, WHO and UNICEF established replicas of the JSU and PSUs primarily in the north-west provinces. Similarly, international NGOs active in the Cambodian health sector had in August 1989 established an informal coordinating mechanism at national level called MEDICAM, which sought to forge a closer relationship with the Ministry of Health and constituted a sub-committee of the CCC. 2 MEDICAM gained momentum in 1991 when the MOH designated a representative to MEDICAM, and in April 1991 three NGO representatives were voted in to join CoCom (see below) for an initial six months. Subsequently MEDICAM, with substantial support from MSF, also gained representation and participated in the numerous technical sub-committees of CoCom. The UN, WHO and CoCom Of particular interest, however, is the health corollary of UNTAC itself. Until 1991, UNICEF had been the lead agency Table 1. Number of NGOs working in Cambodia, International NGOs Local NGOs Sources: Cooperation Committee for Cambodia, for international NGOs; Ponleu Khmer for local NGOs.

8 236 Steven Lanjouw et al. in the health sector in Cambodia. At least until the 1990s, UNICEF s mandate was unique within the UN family in its ability to work with non-state entities, and states such as Cambodia which were not internationally recognized (Richardson, 1995). The signing of the Paris Accords provided for the beginnings of a normalization of relations beween the UN specialized agencies and Cambodia. UNDP, WHO, UNFPA, although lacking permanent representation in Phnom Penh, started to provide technical support to the health sector in Such normalization was tentative and conditional, a fact which was to have implications for the UN s ability to provide a leadership role in the coordination of aid for health. During the transitional period, the WHO office was not fully accredited, and was run instead by a Special Envoy of the Organization s Secretary-General. Health development assistance provided by the Organization had to be provided through a humanitarian assistance programme, and to steer clear of the... usual WHO development terminology for the time being for political and mandate reasons (Kreysler, 1991). The implications of relief rather than development funding have been described elsewhere (Macrae, 1997; Macrae et al., 1995). The key distinction between these two forms of aid lies not in their particular content, but in the fact that in contrast to development aid, relief is politically unconditional and does not confer legitimacy on the incumbent regime. WHO s emphasis on the humanitarian nature of its intervention, reflected the essentially political distinction between relief and development aid. The ability to preserve the neutrality of a humanitarian actor, whilst also striving to provide a focus for a coordinated approach to rehabilitation inevitably placed strains on WHO, and the Coordinating Committee (CoCom) it helped to establish and maintain. As indicated earlier, the nature of working in a situation of contested legitimacy meant that the developmental approach it adopted required engaging with a state administration. WHO justified this engagement on humanitarian grounds, in other words, arguing that the longterm development of effective health services required working with professional staff in relevant ministries and provincial authorities. However, the majority of donors questioned the legitimacy of these bodies, and allocated resources outside of them. CoCom The capacity of the Ministry of Health to coordinate health sector aid was extremely limited. At precisely the time when more international assistance than ever before was coming in to the country, the public administration was in crisis. There were few officials with expertise or training in planning and coordinating humanitarian or basic health care assistance programmes and salaries were plummeting. Even if it had been functioning, the public administration was identified with the Communist Party and thus not seen as a legitimate partner by the majority of donors. As a result, coordination initiatives at central and provincial level were left largely to the international community itself. WHO, and to a lesser extent UNICEF, sought to fill the vacuum left by the erosion (de facto and de jure) of the state. In an attempt to build a coherent management framework for health services rehabilitation, in 1992 WHO started to actively bolster CoCom, which by then drew on senior-level representation from the existing SOC health administration and from other international agencies. The Committee was chaired by an Under-Secretary of Health, and its membership included representatives of the MOH, WHO, UNICEF, the International Committee of the Red Cross (ICRC), UNDP, bilateral agencies and three NGO representatives. CoCom was supported by a Secretariat and Executive Committee which was chaired by a Director of Cabinet and included the Director General of Health, the Director of Policy and Planning, and the Director of International Relations (see Figure 4). Technical assistance was provided by WHO, whose representatives, along with other senior officials of the MOH, and representatives of some of the bilaterals, participated in CoCom Secretariat meetings. The terms of reference and working practices of CoCom helped define the relationships between the transitional national authority and the range of actors in the health sector. CoCom s objectives were: to monitor and evaluate all health activities by international aid agencies working in the health sector; to provide advice; and to make recommendations to the MOH in order to support the planning, coordination and implementation of health sector activities in Cambodia. In addition, it was to ensure the continued generation and evaluation of health information and health activity mapping with a view to identifying priorities and constraints, and to recommend national policies and plans; review public health sector human and financial resource needs; undertake studies and reviews (through sub-committees and working groups) and facilitate the most appropriate use of technical advice to the Ministry; provide information on current and planned activities of all international, multilateral, bilateral and nongovernment organizations working in the health and health-related sectors; and to review all externally funded health proposals to ensure they were consistent with national health policy. The establishment of CoCom at national level was followed by a number of provincial coordination committees (ProCo- Coms), set up to coordinate international assistance at the province and district level. ProCoComs were chaired by provincial directors of health, supported by a secretariat which convened meetings, prepared agendas, and took minutes of meetings between the MOH authorities and international organizations working in the provincial health sector. Support for this process came from both UNICEF and WHO. Four ProCoComs were established in 1992, and two in By 1994 there were eight ProCoComs in the 21 provinces/ cities throughout the country. The overall objective of the ProCoComs was to facilitate the implementation of national health programmes and policies via provincial health plans

9 Coordination in post -conflict Cambodia 237 Figure 4. Coordination Committee of the Ministry of Health through the coordination of NGOs, international organizations and provincial health resources to avoid duplication and to maximize health impact. The ProCoCom mechanisms also sought to disseminate information from the CoCom level to the provinces but do not appear to have functioned well as information collecting institutions which fed insights and material up to the national level. Even in Battambang province, one of the better functioning and resourced, Pro- CoComs had real difficulties getting problems addressed and in Svey Rieng, a province along the Vietnamese border, for example, there was little movement of information in any direction. WHO and UNICEF also supported the establishment of a number of CoCom sub-committees and technical working groups to assist in advising the MOH on the implementation of specific health services (see Figure 4). Each sub-committee had its own specific terms of reference, which were reviewed and updated annually; they were supposed to meet regularly and to feed information to the MOH, through CoCom, on heath policy and planning issues. Membership reflected the composition of the CoCom, and included one or two NGO representatives with particular experience in the issue concerned, who were identified through MEDICAM. By 1994 there were eight sub-committees in existence. 3

10 238 Steven Lanjouw et al. Coordination of repatriation and reintegration activities A second important focus of health coordination during the transition period placed emphasis on efforts to repatriate refugees, notably those from the Thai border camps, and to reintegrate returnees into local structures. In relation to the health sector, a key element of reintegration was to ensure that health workers trained through different systems and, working with the different political factions, could be absorbed and utilized within the health system. Although final responsibility for the repatriation and reintegration of refugees lay with UNHCR, implementation of a health component for the integration of refugees into the existing health services was to be coordinated by WHO and UNICEF. Based on the findings of the UN interagency humanitarian needs assessment (United Nations, 1991) and the UN Secretary-General s Consolidated Appeal of June 1992 (UNTAC, 1992), it was agreed that WHO would act as the coordinator for the health component, with UNICEF responsible for implementation. Thus ahead of a newly mandated government in Cambodia in November 1993, WHO, along with the interim health administration, and in collaboration with senior personnel of the four political factions which together made up the SNC, undertook to develop and support the health services reintegration and unification of three of the four factions into the existing SOC (government) health services (Hun Chun et al., 1993). These activities were concentrated in the north-west provinces of the country to which the majority of refugees aimed to return, and as bilateral donor support for the health services of the former factions (FUNCIPEC and KPNLF) ended in October 1993, this undertaking paved the way for the emergence of a civilian health system. It was proposed that this would take place through the integration of factional health staff and health programmes into the national district and provincial health management system. It was supported by a technical planning and working group which considered policy recommendations regarding the strengthening of provincial and district level management capabilities, issues regarding the private sector health services and health facilities, and human resources development. In Banteay Meanchay and Siem Riep, staff from a variety of hospitals, infirmaries and MCH clinics were rationalized in order to reduce operating costs, reduce the number of health workers employed, especially those with questionable training, and reduce the number of hospital beds from 415 to 270. In addition, the Health Worker Qualification Equivalencies Paper was produced in February 1993 and agreed by the SNC factions, collapsing the number of different health worker diplomas nation-wide from 59 to 23. NGO mapping In May 1991 the MOH prepared a country-wide NGOmapping exercise to determine the exact type and quantities of aid flowing in from external sources to Cambodia in order to facilitate coordination of NGO medical assistance. For a number of reasons, including the fact that there was no mechanism to enforce compliance, this information was not collected and it was not until March 1992, with support of CoCom and WHO, that this initiative really got underway. Subsequently general information about each NGO operating in the Cambodian health sector, and some information about bilateral and multilateral support, was collected, listing projects by geographic area and activity, human resources, budgets, and the level of health system intervention. Thus by mid 1992 there were 55 NGOs (including the international Committee of the Red Cross and other Red Cross societies) working in the health sector, 15 of which worked with returnees, 14 with internally displaced. Thirty-nine were based at provincial level, 26 at district level and another 26 at central level in Phnom Penh. Total funds budgeted for NGOs in 1991 were US$16.6 million, rising to US$28.8 million in 1992 (Lanjouw et al., 1998). 4 Alongside the establishment of a data base informing the allocation and distribution of NGO health resources and activities, it was hoped that this information would help attract NGOs to underserved areas of the country, notably those to the north-east and south-west. Moreover, after national elections it was hoped that this exercise would be repeated regularly and that it would encourage organizations in the health sector to submit project proposals and reports on their activities to the proper departments in the MOH. Other coordination mechanisms Outside of governmental mechanisms, health coordination also took place in a variety of other forms. Notable among these were the Memoranda of Understanding (MoUs) between UN agencies, which established inter-un coordination mechanisms. UNDP and WHO, for example, established an MoU for a project on hospital infection control; another between UNICEF and WHO covered collaboration around the Expanded Programme on Immunization/Polio eradication, the pharmaceutical sector, strengthening health services, maternal and child health and nutrition; and one between UNHCR and WHO covered the joint appointment of provincial health advisors. Coordination in practice The UNTAC period ended in December 1993 with Cambodia still plagued by a low-intensity war. In particular, the Khmer Rouge continued its battle to undermine the emergent state, which remained weakened by political (and not occasionally, military) in-fighting. Some preparations for the post-election period had been made by the UNTAC Rehabilitation Directorate, including hastily designed efforts for demining, emergency and rehabilitation support such as food security for repatriated and resettled populations, investment planning, and a structure to continue the mobilization of donor assistance. It was hoped that the ICORC mechanism would assist the new government and its administrative bodies to coordinate international assistance, and to oversee the continuation of the reconstruction and development process. During the period the size of the international aid

11 Coordination in post -conflict Cambodia 239 programme in Cambodia increased eighteen-fold from US$17.3 million to US$317 million. No reliable figures exist that detail the precise allocations to the health sector during this period. That such figures are not readily available is worrying in terms of accountability and in terms of subsequent planning of health expenditure. What is clear, however, is that by 1993, international aid accounted for 88.6% of total public health expenditure in the country (CDC database, see also Lanjouw et al., 1998). Aid then was a critical resource for the Cambodian public health system. Yet, as described above, the capacity of the public administration to coordinate this aid was limited. Lacking a clear national counterpart, and reflecting the sophisticated political mechanisms put in place to provide an interim state, a strikingly large number of bodies emerged which sought to coordinate aid generally and health aid in particular. In addition to their number, what is also striking is that, with the exception of ICORC which met only rarely, none of the coordination mechanisms which emerged were linked to resource allocation. So, for example, the UNTAC rehabilitation component, itself underfunded, had no operational responsibility. Similarly, WHO, which had its own resources for technical assistance, had no operational budget. MEDICAM represented NGOs, the majority of whom relied on official funding for the bulk of their work. It is bilateral allocations to operational agencies which shape aid responses. It is they who decide the overall scale and pace of the response, the type and distribution of aid instruments and projects, and the timeframe of the project cycle. While some of the detail of planning may be delegated to an intermediary multilateral body, such as UNHCR for refugee repatriation, through the choices they make regarding which projects to fund, bilaterals clearly have a key role in shaping sectoral policy. Bilaterals lacked a coherent strategy for sectoral rehabilitation in Cambodia, and nominally delegated responsibility to the UN and its specialized agencies to guide policy in the absence of a functioning state. However, the ability of the UN to play this role was undermined by national and international actors who worked consistently outside the UNTAC framework, damaging its capacity to monitor and coordinate aid. In the process, its ability to ensure that the principles for rehabilitation laid out in the Charter (Box 1) were adhered to was also undermined, as was the ability of specialized agencies to promote technical standards. In the case of the health sector, efforts by health professionals national and international to establish effective mechanisms for coordination were undermined to a considerable extent by structural features of the aid system, and by the particular political problems it confronted working in a situation of contested legitimacy. Coordination is clearly a means, rather than an end in itself. It is a means by which the sum of collective efforts can be used to maximize health gain through the establishment of an effective and sustainable health system. Bilateral bodies were largely unwilling and unable to work with and through the public administration then in place. The international aid system could only become operational by channeling funds through NGOs. While NGOs offer important advantages, not least their capacity to act as competent interlocuters, they also suffer from a number of limitations. In particular, NGOs could not (nor would they claim to) substitute for the presence of an effective state, nor could they develop and implement policy through a national network of public health professionals. While bodies such as WHO sought to re-establish the capacity of central policy-making institutions, the pattern of resource allocation tended in the opposite direction, namely to invest in disparate project activities, implemented by many different NGOs. Such an approach was unable to address key structural problems facing the sector, in particular the problems of acute underfinancing and problems of very low public sector pay. In real terms, the value of government health spending in 1993 was half that of While salaries comprised a relatively large part of this expenditure, in absolute terms, health worker wages were below subsistence levels, forcing them to leave the sector, seek alternative employment or to unofficially charge their patients. In a context where the bulk of aid resources were channelled outside the public administration, donors could not adopt a coherent, policy-based approach to health sector development, relying rather on a project-based approach. It is not surprising that efforts to coordinate these initiatives were at best partially successful, since the mechanism for aid disbursements was inherently fragmented rather than coherent. While CoCom and MEDICAM provided important vehicles for information exchange, and for much vigorous discussion regarding options for the development of the Cambodian health system, these were not juridical bodies with the authority to determine the location, content or duration of projects. Rather, they relied upon establishing consensus between different agencies, each with their own mandate, organizational style and existing commitments to their donors. Delinked from resource allocation mechanisms, the valuable technical guidance they sought to provide was undermined by the effects of the wider political context, which in turn influenced the choice of aid instruments. In particular, rehabilitation was designed as a stop-gap measure between relief and development aid. While claiming developmental objectives, in contrast to development aid the project cycle was short, thus limiting the range of activities which could be undertaken. Also inherent was a donor preference to invest in infrastructure. This combination of factors resulted in strategies which were unlikely to prove sustainable in the medium term, and which arguably decapacitated the already weak public health service in the country. So, for example, under-financing of essential drugs probably increased the problem of underutilization of public facilities. Similarly, no funds were formally targeted towards a strategy to address the very low