Regional Environmental Technical Assistance 5771
Poverty Reduction & Environmental Management in Remote Greater Mekong Subregion (GMS) Watersheds Project (Phase I)

Review of Policies and Institutions related to
Management of Upper Watershed Catchments

Cambodia

By

Esa Puustjarvi

CONTENTS

1. Introduction

The review of policies and institutions related to watershed management in Cambodia relies on the general analysis framework developed at the sub-regional level, and presented in the Regional Report of this study. Country reports are presented in a form, which allows for reading them as stand-alone documents. However, a general discussion on the issues and conclusions can be found in the Regional Report. As the structure of country reports and that of the regional report is largely the same, and the discussion relevant to various issues can be referred to by looking up the corresponding sub-heading.

Topics such as health, education, and gender have been excluded from the scope of this report, as they are covered under other assignments.

2. Socio-economic status

General

The population of Cambodia has expanded from 6.9 million in 1970 to the level of 11.4 million in 1998. The annual rate of population growth is estimated at 2.8 %/a, which is high in sub-regional terms. Population density, estimated at 64 persons per sq. km, is average for the sub-region. Large-scale urbanisation has not yet taken place, and it is estimated that 81 % of Cambodians still live in the rural areas in 1997. Ethnic minorities inhabiting mainly the northern and eastern provinces represent approximately 3 % of total population. However, locally they account for a larger share. In Mondolkiri and Rattanakiri provinces ethnic minorities represent 93 % and 67 % of total population, respectively.

In general, the socio-economic status of Cambodian population is modest. Indicators show that despite the positive achievements recorded during the last few years, Cambodia remains one of the most disadvantaged countries in the sub-region. For instance, life expectancy has risen from 42 years in 1960 to only 53 years in 1995. Adult literacy rate stands at 68 % in 1997. Infant mortality has dropped to 104 per thousand live births in 1995. In 1995, GDP per capita in Cambodia was only USD 280, and purchasing power parity USD 1110, which is the lowest level recorded in the sub-region.

The development of national economic has been satisfactory, but in sub-regional terms the performance is below average. Between 1992-96 the average rate of GDP growth was 5.8 %/a (Table 1). The fastest growing sector was industries at 17.2 %/a, followed by services, which posted a rate of 5.0 %/a. Growth in agriculture has been recorded at 3.3 %/a, which is a low level in a sub-regional comparison as well as in international terms. In 1996, agriculture represented 43.1 % of GDP, which is among the highest proportions in the subregion. The industries and service sector accounted for 16.8 % and 40.1 % of GDP, respectively.

Table 1: GDP Growth in selected sectors and overall in Cambodia in 1981-1995

Source: Asian Development Bank

In 1993-94, measured by headcount index the number of people living under national poverty line was 39.0 %, which at the time was above the average level in the sub-region. By 1997 poverty had declined modestly to 36.1 %. Distribution of income is relatively even. Measured by Gini coefficient (0.36) Cambodia ranks in the middle group among GMS countries.

Measured by the human development index compiled annually by the UNDP* Cambodia takes the position of 140 (out of 174 countries), which is lowest rank in the sub-region. In global terms, Cambodia is classified as a country of "low human development".

Upland/rural areas

The upland areas of Cambodia are located in the northern, eastern and southwestern provinces (Ratanakiri, Mondolkiri, Stuang Traeng, Preah Vihear, Otdar Mean Chey, Kaoh Kong). Parts of other provinces such as Pousat can also be considered upland areas. However, as statistical data at provincial level is largely lacking, data breakdown is usually limited to separation of urban and rural areas.

The upland areas of Cambodia are very sparsely populated in absolute terms and in comparison with the rest of the country. The average population density is only 8 persons per sq. km (Table 2). The uplands account for 4.6 % of total population, but in terms of land area they represent 37.2 %.

Table 2 Population in Cambodia in 1997

Source: National Institute of Statistics 1998

As expected, poverty is most prevalent in rural areas, where 43.1 % of the total population is estimated to live under the national poverty line (Table 3). The discrepancy between the urban and rural areas is also shown by disparities in household expenditure, which in the rural areas is less than one third of expenditure level in Phnom Penh.

Table 3 Poverty and household expenditure in Cambodia by regions

Source: National Institute of Statistics 1998a

Information on infrastructure and health services is available at village level. Road access is satisfactory. Even in rural areas more than 70 % of villages have road access (Table 4). However, in some other regards difference between urban and rural surroundings is striking. Electricity is available only in 15 % of rural villages as opposed to 87 % and 55 % in Phnom Penh and other urban areas respectively. Access to services provided by doctors and nurses are also much more limited in rural areas than in urban settings.

Table 4: Proportion of (sampled) villages that have the stated facilities within the village itself

Source: National Institute of Statistics 1998b

Social indicators show also that rural areas lag behind Phnom Penh and other urban areas, even if the difference is not dramatic. Almost two thirds of the rural population is literate, but secondary education is held only by one fifth. The difference in using health services is not dramatic, even though rural people use them somewhat less frequently (Table 5).

Table 5 Selected social indicators in Cambodia

Source: National Institute of Statistics 1998b

Remote watersheds

Information on remote watersheds is scant and it has not been collected systematically. However, it is commonly known that their economy is based almost entirely on subsistence or semi-subsistence agriculture, and that cash economy is making only first inroads into these communities. From available documentation it may be inferred that people living in these area often suffer from poor health, and, yet, have extremely limited access to health care. For instance, it is reported that in Mondolkiri province 99% of the people have had malaria, and that in some areas infant mortality for children below three years of age would reach 80 %. Access to education is limited, and basic amenities such as water wells, proper sanitation and simple hygiene are most often lacking.

3. Environmental status

A number of studies have been conducted in order to assess the forest cover in Cambodia. Unfortunately the study methodologies used in the investigations differ, which makes comparisons rather unreliable. The most recent estimates on the present forest cover dates to year 1997, when the Forest Cover Monitoring Project (FCMP) established it at 58.2% (Table 6). The most recent available estimate on the rate of deforestation is provided by FAO based on the period 1990-1995. During study provides also an estimate on rate of deforestation, which is put at 32 800 ha/a corresponding to a rate of 1.0%/a. Comparison with the estimate provided for the period 1980-89 suggests that deforestation would have increased. FAO also provides comparable data on countries in the GMS region, and their analysis shows that Cambodia has the highest forest cover in the region after Cambodia. Regarding the rate of deforestation Cambodia is in the middle group in the region (cf. Regional Report).

Table 6: Estimates for forest cover in Lao PDR

Source: FAO 1997
FCMP MRC/GTZ 1999

Detailed breakdown of forest cover indicates that Cambodian forests (within the Mekong basin) are still in rather good condition. Continuous forest cover is found on 54.5 % of land area (Table 7). However, the available data is not able to fully capture the effect of logging as an agent of forest degradation. Further classification of non-forest areas based on satellite imagery remains approximate, as the methodology is not able to make a clear distinction between wood and shrubland and areas under shifting cultivation. The assessment put the area of wood and shrublands at 12.1 % of total land area, and that of mosaic of cropping (incl. shifting cultivation) were assessed at 2.3 % respectively. In sub-regional terms these are rather modest proportions. The available data did not permit an assessment of development trends.

Table 7: Actual land use in Cambodia in 1997 (Mekong River Basin)

Source: FCMP MRC/GTZ 1999

Based on estimates calculated from available FCMP data (1993) the extent of forest cover is high in the upland regions, 87-90 % (Table 8). Forests are also in good condition, as almost all forested areas have continuous cover. The proportion of secondary forests is very low. However, as mentioned earlier the impact of logging on forest degradation is not fully captured by these data.

Table 8: Regional estimates for forest cover in Lao PDR in 1993 (whole country)

Source: FCMP MRC/GTZ 1999

The reasons for deforestation have not been subject to a systematic study in Cambodia. Expert opinion suggests that commercial logging, shifting cultivation and wood harvesting for fuelwood and charcoal production are the primary causes. Shrublands are apparently the primary form of land cover to succeed forests after clearance. The high level of wood harvesting in recent years suggests that logging has been a major agent of forest degradation. However, as logs are usually removed through selective cutting, the impact of harvesting and road construction on deforestation is unclear. As population density in the upland areas is low, construction of forest roads does not necessarily lead to clearing of forest for agriculture.

4. Government Policies

4.1 Rural Development Policy

The Government’s policy on rural development is focused on addressing two principal issues: (i) poverty reduction and (ii) natural resource management. About 80 % of the poor live in rural areas, and progress in relieving their plight has been modest. Natural resources are the principal base for economic development, but it is being been severely eroded as a result of short-term financial pressures and the public sector’s weakened management capacity.

The Government’s strategy for rural development is built on sound principles. It aims to spread the benefits more widely through the country by expanding the support to rural areas and fostering broad-based economic development. One of the cornerstones is the principle of decentralised decision-making. The Government’s role is restricted to facilitating a process, where the principal actors are the intended beneficiaries of rural development.

However, while the strategy is clear it suffers from unusually difficult implementation constraints and human resource limitations. Despite recent initiatives, the actual implementation of rural development programs still suffers from too centralised and -top-down planning and implementation. This is partly reflecting a persistence of thinking and practice from earlier periods, but it is also a result of political pressures at local and regional levels. It is clear that if the strategy is to be successful planning and implementation of rural development strategies has to be brought much closer to intended beneficiaries in the villages. Current initiatives such as the establishment of SEILA Task Force for supporting the decentralisation process are an encouraging sign. However, the main problem is a lack of transparency and accountability in Government administration, which has left bodies such as SEILA without real influence. Much depends on the Government’s capacity to reform itself and to bring decision-making back to a more transparent framework.

4.2 Funding

The Cambodian Government is committed to reducing socio-economic disparities between different parts of the country. However, the dilemma faced by the Government is that poor regions often suffer from low rates of growth, and investments made there may not generate as much tax revenue as spending in richer regions would do. From an equity stand transfers from richer regions to poorer ones would be justified, but the problem lies in the reasoning "wealth can be redistributed only after it has been generated". In Cambodia, this is a severe problem as the Government’s revenue base is limited, and even distribution of resources over the whole country may shrink the resources available to all regions below a critical level, which is necessary to have any significant impact.

Still, poverty reduction and giving greater priority to rural areas in the allocation of public investment funds, are the two overarching objectives of the Government's Social and Economic Development Plan (SEDP) for 1996-2000. It set a target of a 65/35 ratio between rural and urban projects to be achieved by the year 2000. However, even if the second rolling Public Investment Plan of 1997-99 accords increased emphasis to rural areas, only one third of the proposed projects are exclusively for rural investments. Another 45 % straddle the urban/rural divide but the distribution of benefits between the rural and urban areas is somewhat unclear. The remote watersheds are probably the areas most susceptible to forego Government inputs as investments made there do not necessarily carry high economic returns.

Investment support to remote watersheds can be considered justified on two grounds. One of them is equity. Most disadvantaged groups need public support in order to overcome the constraints to self-help and self-sufficiency. Second, Government investments may be considered a compensation for use of natural resources. Remote watersheds hold abundant natural resources, especially timber, which are being utilised by external interest groups. While the local people have borne the negative consequences of forest harvesting, the benefits accruing to them have been few.

On the other hand, the discussion on Government investments should not divert attention from the fact that providing local population with access to productive assets is often more effective in catalysing development. In Cambodia, the huge amount of resources needed to rebuild of the country tends to dwarf the impact of public investment even under most optimistic scenarios of fund mobilisation. In rural areas, the most important productive asset is land, and therefore granting land tenure is the most urgent measure needed to mobilise the local resources (cf. 9.1).

4.3 Upland strategy

There is no Government-endorsed strategy for upland development. However, the Inter-Ministerial Committee for Ethnic Minorities Development has drafted an extensive Highland Policy (September 1997) which has been forwarded to the Council of Ministers to be presented to the next Government. The paper puts forward inter alia the following policy statements and strategies:

• Highland Peoples have the right to practice their own culture, adhere to their own belief systems and traditions, and use their own languages
• Procedures shall be established in order to recognise and protect traditional land and forest use rights, land use rights will be granted to areas delineated based on traditional village boundaries
• Highland People will be able to participate in all levels of policy formulation, development planning and decision-making affecting their lives
• The Government recognises swidden systems of agriculture and appropriate forest products gathering as valid forms of land use, improvement of agricultural systems shall be undertaken, but based on traditional practices
• Highland Peoples shall not be removed from the lands they occupy except in extreme cases, where significant national interests are at stake

The proposed policy is comprehensive and well designed. Most importantly it would recognise the traditional land use rights, and a participatory approach to development. However, it is obviously in direct conflict with some of the present practices, especially the granting of concessions for timber harvesting and plantation agriculture. The economic stakes in these undertakings are high, and there will undoubtedly be attempts to dilute some of the contents of the policy. The outcome is by no means certain, and it is to be seen, whether the proposed policy will be approved in its present form.

5. Agriculture

Agricultural production in the uplands consists mainly of shifting cultivation on the hillsides and rainfed rice production in the valley bottoms. The area of upland rice is about 50 000 ha. Some groups combine swidden systems with lowland paddy rice.

The agricultural policy the Government aims to promote sedentary agriculture as the principal farming system. On the other hand, large-scale plantation agriculture has also been promoted by awarding concessions to suitable land areas. Agricultural concessions account for 4 % of total land, and many of them are located in the upland areas.

Owing to a lack of implementation capacity, the Government’s policy towards sedentarisation has had little effect in the uplands. The abundance of land makes shifting cultivation a sustainable land use, and in most upland areas there has been little incentive to transform the present system. However, in some areas the concession policy as well as natural population growth are gradually eroding the sustainability of indigenous land uses by increasing land scarcity. The most immediate threat comes from the expansion of concession areas, but there are also examples of sites, where swidden fallows have been reduced to unsustainable levels (3-4 years) mainly due to natural population growth.

Expansion of concession areas is a serious threat to the sustainability of upland agriculture, and it will have to be contained by securing local tenure. However, even if this were achieved, the intensification of swidden practices would remain an appropriate objective for upland agriculture. A settlement of the land issue would most probably lead to a situation, where village boundaries are fixed, and local people’s agricultural land use would be confined to the area within the village. Given the high natural population growth in the upland areas increasing land scarcity is an obvious outcome and some measure of adjustment will be needed at least in the medium term.

On the other hand, as land scarcity is not a severe problem in most parts of the upland area, a gradual intensification of agriculture seems the most appropriate development path. Research results as well as observations in other GMS countries indicate that the adjustment can take place spontaneously provided that certain preconditions are in place. Some of the most important conditions are secure land tenure and access to markets.

Reliance on a spontaneous adjustment would be in line with the Government’s present approach to rural development, where Government adopts a role as a facilitator and decision-making is decentralised to the village level (cf. ch. 4.1). It would also be recognise the limitations imposed by the dearth of Government resources. By adopting a low-intensity strategy the resources could be spent on ensuring that the conditions conducive to change are in place (e.g. market access, market information, awareness of changes in regulatory framework, demonstrations on available technologies etc.). For technical innovations the strategy would rely largely on the villagers’ capacity to handle change.

By reinforcing the spontaneous adjustment of production methods the shift may not be as rapid as by using more radical approaches. However, it is likely to reduce the risk that farmer’s experience, which may, in turn, increase the attractiveness of transformation. When resources for technical support are limited, efforts to find alternatives to shifting cultivation are probably more successful, if a transition is gradual rather than abrupt.

6. Forest-based production

Current legislation and regulations governing forest management are inadequate in many respects. The forestry law provides no objective standards for forest protection or watershed protection and there are no integrated guidelines or standards for forest management. Volumes of timber extraction exceed by far sustainable levels. In principle, Concession Contracts require adherence to sound forest management practices, but circumventing or outright ignoring the regulations has been commonplace, as there has not been any real threat of legal sanctions.

Improvement of forest management and forest protection systems will require a complete overhaul of the institutional and regulatory framework for the private sector and local resource management. The issue was thoroughly analysed by the recently completed "Forest Policy Reform Project", which reached inter alia the following conclusions:

• Conduct a forest resource inventory with a view to determining annual allowable cut and reducing the logging volumes to this level
• Review mechanisms for land use planning and allocation
• Replace all prior forestry law
• Revise standards for forest management
• Improve management of protected areas
• Develop a mechanism for community involvement in forest management
• Review the system for collecting and distributing forest revenue
• Establish a system for monitoring and enforcement
• Strengthen relevant institutions

All these recommendations are still relevant and should be implemented as soon as possible. From the perspective of watershed management suffice to point out that formulation of management procedures and standards has little meaning in a situation, where the institutional and regulatory framework is highly subject to influence by interest groups (particularly military), and lacks any real means to implement activities in the field.

However, if improvements in this regard take place, one of the priority actions regarding watershed management is the establishment of a proper procedure for land use planning and land allocation including a mechanism to ensure that local people are able to participate in and benefit from this process. The fact that all protected areas are subject to either "high" or "low" intensity logging clearly indicates that restoring the regulatory function of the forest administration is another key task, if watershed protection is to be successful.

7. Rural industries

The industrial sector is the fastest developing sector of the Cambodian economy posting an average annual GDP growth figure of 17 % between 1992 and 1997. However, the expansion is partly due to the sector’s small size both in absolute and relative terms. Industrial production represents only 17 % of the GDP in Cambodia. Provincial figures are not available, but the industrial sector in the upland provinces consists of a handful of enterprises in the provincial capitals. In all rural areas only 3.3 % of the population i.e. 12 000 people are employed in the industrial sector, which suggests that their number in upland areas is negligible. A case study of four villages in Ratanakiri indicated that only 30 % of local people had experience of any kind of wage labour.

The two most obvious options for developing rural industrialisation in upland areas are forest industries and agro-processing. Regarding forest industries, the Government has promoted the development of capital-intensive wood processing sector through a log export ban, and grants of large concession areas at modest royalties. This has enabled the emergence of a few medium-scale wood-processing units. Probably owing to lacking infrastructure only few of these enterprises are located in the upland areas. The number of local small-scale sawmills is approximately 1 000. The operating capacity of all these mills goes far beyond the sustainable level of log supply. The implications are serious as the processing plants compete for the same resources, which makes expansion of the forest industry sector unlikely.

The development agro-industries is hindered by a lack of surplus production and the low quality of agricultural output. The sheer lack of production volume from local farmers is a major deterrent for investments in agro-processing. On the other hand, development of concession-based plantation agriculture is a potential basis for emergence of agro-industries and generation of employment in the upland provinces. From upland people’s perspective the main issue with such ventures is the distribution of benefits. It is unclear whether the benefits accruing to local communities from plantation development offset the negative effects caused by clearing and closing off the concession area (cf. Box 1 ).

In conclusion, it appears that while the development of an industrial sector in the upland areas may be a future option, the current focus of upland development programmes must remain on improving and increasing agriculture and forest production. An initial expansion of agricultural sector and surplus production beyond immediate needs are preconditions for diversification of economy. Generation of local capital and enhancement of local demand may eventually trigger investments into industrial development Provided that future arrangements for land tenure provide local communities with access to timber resources, the development of small-scale wood processing is one of the most promising options in the upland areas.

Box 1: Impact of O’Yadao Agricultural Concession on Local Community

8. Population movements

Large-scale urbanisation has not yet taken place in Cambodia. Approximately 30 % of rural population has moved from their place of birth. Two-thirds of all migrants move from one rural area to another often within the same district. Rural-urban migration accounts for 15 % of all migration.

In the upland areas, out migration is still modest, but some lowlanders have moved to areas previously occupied by uplanders pushing them deeper into the remote areas. However, the scale of such migration is apparently still modest. A more dramatic impact would be delivered, if the Government’s former plans to demobilise the army by providing soldiers land in Ratanakiri and Mondulkiri provinces were implemented.

In general, land scarcity in the upland areas is not a major problem, and providing sustainable livelihoods for the local population based on primary production seems a feasible option in Cambodia. The number of people migrating from the upland areas is likely to be limited making the need for supportive measures modest. Therefore, it appears that in Cambodia - as opposed to many other countries in the subregion - the development efforts regarding watershed management can be concentrated on on-site development.

On the other hand, in-migration from the lowland areas can increase or it may even become necessary and justified to make land available for landless, voluntary migrants. However, in order to prepare for this eventuality, and to ensure that the negative effects on people currently living in the upland areas are minimised, the significance of land and resource tenure cannot be overestimated.

9. Land tenure

9.1 Present situation

The basic policy of the Cambodian Government regarding land tenure is to allocate agricultural land to individuals or families, and to maintain forestland in state ownership. The current tenure of agricultural land is based on temporary permits. The programme for titling of agricultural land to households is in its initial stages, and it is estimated to take 8-10 years to complete. The upland provinces are likely to have low priority, when the target areas for programme implementation are selected.

Approximately 60 % of forestland has been awarded to timber concessions, 31 % is under protected areas and only 9 % is unallocated. The principal problem in the upland areas is that timber concessions give the concessionaires exclusive user rights even in areas, which are currently inhabited by local people. The situation is most extreme in Ratanakiri, where 130 % of the land area has been allocated to concessions. Logically, this implies that all forest and agricultural land used by local people fall within concession areas.

The status of these people is ambiguous. Some of the concession zones are "common property" areas to which communities have traditionally had access. The right of rural people to collect forest products in these areas is now unclear, as they are facing arbitrary rules that vary from concession to concession. These include denied access to forest roads, non-wood and wood forest products, and degradation of forest resource that they depend on. Concessionaires are faced with problems of wood poaching, encroachment and land clearing for agriculture. The absence of clearly defined rights and dispute resolution mechanisms is the fundamental shortcoming of the present arrangement, and it is the issue needing most urgent attention in the upland areas.

9.2 State vs. non-state tenure

Customary tenure of forestlands was mainly concerned with ensuring access to resources needed in subsistence economy. Therefore the argument that state ownership (implemented through timber concessions) is not in conflict with the rights of local people, carries some weight in Cambodia provided that subsistence use is secured. It should be noted that a mechanism for community involvement within existing forest concessions already exist in draft form in the Forest Concession Management Project’s Forest Management Planning Package. The proposal outlines a procedure for identifying community needs and developing joint management plans. However, even if the mechanism is there, the question is whether it can be successfully implemented in the present situation.

On the other hand, full non-state tenure can be motivated by the need to respect customary rights. It may be argued that forests held under customary tenure do not have such strategic value to the society as a whole that would justify superseding the traditional tenure arrangements. Another strong argument in favour of non-state tenure is simply equity. Since land is the main productive asset in rural areas, a balanced development can be achieved only if access to a reasonable land resource can be guaranteed. However, transferring all areas held under traditional tenure might prove quite difficult, as it would require revoking legally binding concession agreements. Therefore, a compromise solution may be a more feasible target than full cancellation of all concessions in controversial areas.

Despite the present degradation of protected areas by logging they appear suitable for state ownership. Lack of tangible benefits is likely to diminish the interest of local population in resource conservation, and they cannot be expected to have the means to protect the conservation areas against intrusion by illegal loggers. However, even if the problem with illegal logging could be contained, the Cambodian conservation area network is still very large compared to the available capacity for management and enforcement. In order to bridge this gap it is necessary to explore further the opportunities to maintain conservation values in production forests. The principal strategy would be to increase the interest of local communities in resource conservation by allowing the combination of conservation and sustainable use. This also implies that part of the present conservation areas would be transferred to non-state tenure.

There are three striking features to the present situation in the upland areas. First, the local people’s rights have been all but totally disregarded in the present situation suggesting that any lasting solution must take significant steps to improve their position. Second, conflict over land use in the uplands is actually not very severe, as upland people occupy only a small portion of the land area. For instance, in Ratanakiri province the area inhabited by upland people is estimated at only 15 % of the usable land area. This suggests that a solution accommodating the interests of all parties is within reach, even if finding it is subject to political will. The third key element is the fact that the workability of all tenure arrangements is dependent on the Government’s capacity to restore a functioning regulatory framework.

9.3 Community-based tenure vs. household tenure

Regarding the merits of community-based vs. household-based arrangements in areas that are transferred to non-state tenure, it is highly recommendable that both options including an element of flexibility for local variation were considered in the legal framework. The choice between community-based and household-based tenure can be used as policy instrument to guide land use towards different goals. Community-based tenure appears more suitable for promoting environmental objectives such as protection of existing forest, whereas household-tenure will orient the land use more towards solutions where restoration of some form of vegetative cover is sought.

In Cambodia, the model developed for community forestry in Per Ley Forest grants land use rights to degraded land in exchange for tree planting and protection (mixed with agriculture) (Table 9). The arrangement in the Ya Poey Forest restricts forest use to collection of NTFPs for commercial purposes, but restricts other use to subsistence gathering and hunting.

Table 9: Regulations concerning Community-based Forest Management in Cambodia

Source: Fox 1997

The experience gained in other GMS countries indicates that community-based arrangements seem best suited for protecting existing forest cover provided that the communities can derive some benefits from the forest. The model in Ya Poye Forest conforms with this principle in the sense that it restricts commercial timber use. The local population obtains benefits through collection of non-wood products. On the other hand, the experiment carried out in Lao PDR provides early indications that well-stocked forests handed over to local communities can be managed sustainably for timber production. The particular benefit of this model is that the cash returns are large, which reduces dependency on public financing. Given the dearth of public investment funds in Cambodia, exploring the relevance of this approach to Cambodian conditions may be warranted. The disadvantage is that traditional methods of forest management are not necessarily suited for commercial management, and considerable support is needed to familiarise the villagers with the methodology.

A particular advantage of community-based tenure is that it provides a safety net to the most disadvantaged members of the community. In absence of other mechanism for redistributing benefits, the solidarity of the community is seen as an important factor in ensuring the well being of all members of the community. This is highly relevant in Cambodian uplands where people living in subsistence economy are gradually being exposed to a market-based economy.

However, community-based tenure should not be considered the only option relevant to upland areas. Experience from other countries shows that reforestation with natural tree species is not very successful, even if it is undertaken by communities. The reason is simply that people prefer or absolutely need short-term benefits. The model applied in Per Ler Forest very appropriately attempts to overcome this constraint by allowing mixed production of agriculture and fast-growing trees. On the other hand, the feasibility of community-based tenure may be questioned in this case, since experience has shown since agricultural production does not easily lend itself to collective mode of production. Household-based tenure could be considered as an alternative option, as it has proved effective in encouraging productive activities.

On the other hand, granting existing forest to individual tenure may be risky, if the objective is to maintain forest cover. The opportunity to quickly gain short-term benefits by clearing the forest may be too attractive an option for farmers facing a daily struggle for livelihood. Community-based arrangements foster very effective social control in situations, where the benefits of common action are large and they are shared between the community members. As individual attempts to obtain personal gain would reduce the benefits available to everyone else, the community members have a high interest to ensure that everyone abides by the agreed rules.

9.4 Mechanisms for land allocation

A particular merit of community-based arrangements is that they provide an opportunity for accelerated land allocation. The basic strategy would be to concentrate on demarcating village boundaries and approximate land use zones. Once village boundaries have been delimited, the community and local organisations are made responsible for allocating fallow land to respective groupings and individual farmers according to the customary practice found in each respective village. More elaborate procedures of land titling and land allocation would be undertaken after more resources become available.

The approach would speed up the land allocation process considerably, as local people are usually well aware of the traditional village boundaries, and demarcation could be carried out swiftly through a participatory process. The state officials would not be burdened with land allocation within the villages, which is a very time consuming process. Speed in land allocation is vital, as the mere demarcation of boundaries is likely to have positive effects. For speed it would also be necessary that land allocation be implemented as an independent process implemented ahead of comprehensive and high-input development approaches with infrastructure investments and extension packages. The implementation of a more elaborate land allocation process should be restricted to priority sites.

10. Institutions

10.1 Government capacity

As elsewhere in the GMS region the number of Government agencies with a stake in watershed management is large. The principal Government agencies include (Table 10):

Table 10: Principal Government Agencies involved in watershed management

Source: Touch 1999

The key ministries involved in watershed management include

Ministry of Agriculture, Forestry and Fisheries (MAFF) has overall responsibility for policy, planning and procedures: however, it is administratively complex, overstaffed (Total, 11 375: HQ 5076: Provinces, 6229 in 1996), under-resourced (budget USD 25 million 1996: USD 39 million 1997) and staff receive low salaries:

Department of Forest and Wildlife (DF&F) has the technical responsibility for watershed management but it too is administratively complex, overstaffed (Total 1622: HQ 741: and Provinces 881 in 1997), under-resourced (budget USD 489746 in 1997) lacks equipment and vehicles (of 24 vehicles, 12 are operational). It has little direct authority, working through MAFF in all aspects of work at central and provincial levels

Ministry of Environment (MOE) has jurisdiction over protected areas. The Department of Nature Conservation and Protection is the executive arm of MOE as regards management of conservation areas. It is understaffed (Total 99: HQ 20: Provinces 79), and apparently under-resourced.

Apart from resource constraint, the principal capacity problem of the Government in the uplands is its very limited ability to support people-centred development. Given the resource constraint, it is clear that the Government cannot effectively handle more than a facilitating function to rural development. Therefore the thrust of developing line ministries should shy away from mere development of technical skills of staff to increasing their ability to facilitate the acquisition of these skills by the local communities. The development strategy of the Ministry of Rural Development is based on this approach (see ch. 10.3.1), and the development of line ministries should be geared towards supporting this effort.

10.2 Inter-sectoral Co-operation

The Ministry of Rural Development (MRD) was established in recognition of the need for intersectoral co-ordination in rural development. The MRD has been instrumental in formulating Rural Development Structures which comprise inter-sectoral committees at five levels of administration (Village, Commune, District, Province, and National). These co-ordinating bodies do not specifically address watershed management issues, but any larger development project would have to be channelled through its structure. For instance, the proposals put forward by the Inter-ministerial Committee for Ethnic Minorities Development are implemented through the Rural Development Structure.

At the watershed level co-ordination is very limited due to lack of staff and Government activities. The only agency with a major interest in watershed management is the provincial forestry department, which is monitoring logging and timber trade.

Since the structure of Rural Development Committees is quite new, it has yet to establish a track record. In principle, devolution of authority to provincial level should be conducive to inter-sectoral co-operation. It brings decision-making closer to the field level increasing flexibility and staff motivation. It makes problems become more tangible and the small size of units facilitates exchange of information. However, the experience in GMS countries shows that decentralisation does not automatically bring about inter-sectoral co-ordination, as administrative barriers at provincial level can remain high. On the other hand, the Cambodian arrangement attempts to avoid the most commonly found flaw in inter-sectoral bodies – a lack of authority. It is tru that the attempt may go only halfway - the Provincial Rural Development Committee are authorised to propose annual budgets to line ministries - but it is a step in the right direction.

10.3 Patterns of participation

10.3.1 Village level

The management structure on which MRD bases its work is a bottom up approach to development, starting with Village Development Committees (VDC). An elected village body, the VDC is recognised by the commune, district and provincial level committees as an autonomous committee. On the other hand, VDC recognises the authority of the Village chief as representing the Government. VDC is the focal point for interaction with all Government department, agencies and funding sources. Currently there are approximately 2000 VDCs organised and trained with a total target of 7500 VDCs operational by the end of year 2000 (60 % of all villages).

The principle of VDC is commendable in that it provides a forum for local democracy to develop. Bottom-up planning from the village level to higher levels of hierarchy is usually difficult to implement unless the institutional structures support it, but Cambodia has this facility. The strength of Village Development Committees vis-à-vis more informal arrangements for participation is the fact that their mandate is unambiguously derived from a popular vote.

However, two potential problems may be pointed out. First, there is a range of questions, which cannot be resolved effectively at the village level without considering the larger geographic context. The Village Committee System is effective in addressing issues at the grassroots level, but democratically elected represented are missing in higher levels of hierarchy, which leaves a major part of decision-making beyond democratic decision-making. However, while this is a serious problem, it should not dwarf the significance of the VDC system, which is a remarkable achievement considering the circumstances.

Another potential problem is the VDC program’s dependence on external funding. Even though the VDCs are a suitable mechanism for any type of decision-making within the village, the usefulness – as perceived by the villagers - is probably tied to the availability of external funds and the possibility to make small investments. This is understandable and appropriate in the sense that participatory and democratic procedures are time consuming and they should be applied only to issues that are "large" enough. Therefore it may be advisable to limit the expansion of the VDC system to areas, where the level of Government activities is sufficient to warrant their existence. In areas, where it is unlikely, traditional methods of decision-making could be depended upon. This may be a relevant viewpoint in the upland areas, especially in cases where the development process would rely largely on a spontaneous and gradual adjustment community processes rather than on an external "injection" of investment funds.

10.3.2 Non-Governmental Organisations

The first Socio-Economic Development Plan 1996-2000 recognises the role of NGOs in fostering socio-economic development, and their potentially positive contribution is stressed. However, drafting of a legal framework for the NGOs to operate has been marred with disputes between the NGO community and the Government. The latest sticking point is the extent to which Government is able to regulate the funding sources. The issue remains unresolved, and no legal foundation for the NGOs exists to date.

Currently, more than 200 international NGOs and between 300 to 1000 local NGOs are operating in Cambodia. As usual, the local NGO community is highly varied in terms of strength and the scope of their activities. The larger ones implement activities ranging from fieldwork to advocacy and policy development. However, the majority of local NGOs can be qualified as Village or Community-based Organisations operating at the local level.

The proliferation of NGOs may be considered a positive indication of the vigour of the civil society in Cambodia. Versatility and flexibility are inherent characteristics of NGOs and constitute a part of their strength. However, dispersal of activities into a large number of small projects makes co-ordination, management and monitoring difficult. It often results into lacking accountability, which is a serious hindrance to effective implementation, and NGO participation in large donor-funded development projects. The emergence of umbrella organisations is a positive development, but it is probably not sufficient to remove the constraint. As the international development projects will for some time to come remain the main players of development, establishment of closer ties between them and the NGOs might be conducive to increased co-ordination. However, as there is some apprehension among the NGOs about direct engagement to donor-driven activities, it may be necessary to explore new modes of collaboration (cf. Box 2)

Box 2: Options for modalities of collaboration between NGOs
and international development projects

11. Principal Conclusions and Recommendations

11.1 General

As pointed out by several observers, the present mode of governance in Cambodia effectively thwarts the impact of any development efforts. Substantial improvements in the institutional, legal and regulatory frameworks are needed, if external assistance is to have a meaningful impact. In upland areas, it is particularly important to contain illegal logging and bring concession management under effective control.

Provided that the situation improves, preparation of proper land use plans and implementing them through land allocation is a basic requirement, if watershed management is to be developed further. With respect to measures raising the socio-economic status of upland people, the most important one is to improve the local people’s access to resources they depend on. In respect of customary user rights as well as on equity grounds the option of granting full user rights to customarily held areas should be seriously considered. The minimum alternative is to grant the local people access to resources in the concession areas in a manner suggested in the Forest Concession Management Project’s Forest Management Planning Package.

It should be noted that granting full user rights to customarily held areas can be implemented rather quickly and with limited resources, if an appropriate approach is adopted. The key issue to focus on expanding community-based tenure. In order ensure swift implementation, the procedure of land allocation should be kept as simple as possible. In particular, the feasibility of restricting it in the first instance to demarcating village boundaries and approximate land use classes should be carefully studied. The villagers are usually able to indicate very accurately, where the traditional boundaries run. Owing to the low population density in the uplands, disputed boundaries would probably be few, and the conflict could probably be solved rather easily. Detailed land use planning within the villages would rely on customary arrangements until more resource become available for carrying out a full-fledged land titling process.

Intensification of agricultural production in the upland areas remains a valid objective considering the high rates of population growth among upland populations. However, as land scarcity is not yet a serious constraint, the transformation process can rely primarily on a spontaneous and gradual adjustment of current land use practices. Research results and experience from other GMS countries suggests that villagers may launch this process of their own accord, if necessary preconditions are in place. Much in the vein of the Government’s present approach to rural development, the Government’s role would be restricted to ensuring that the environment in which farmers operate is conducive to change (security of land tenure, market access, demonstration on technological options etc.). For adoption and adjustment of new technologies, one would rely on the farmers capacity to change.

Non-state tenure of forestland is based exclusively on household-based arrangements. This is likely to be the most appropriate arrangement in areas, where restoration of some form of vegetative cover is the principal objective or where community structures are weak or have been shattered by migration. However, household based tenure is not able to take full advantage of the social forces acting with communities, where internal solidarity can still be found. Experiments in the GMS sub-region indicate that community oversight of forest resource may be able to ensure its sustainable use. The relevance of this approach is not necessary limited to protection forest areas only, but the experiment carried out in Lao PDR suggests that is may have merit also in areas, where commercial forest use is allowed. The possibility of introducing community-based tenure in protection areas as well as in well-stocked forests should be explored.

The establishment of a rural development structure including the democratically elected Village Development Committees is a remarkable achievement, and providing support to their future development should be vigorously pursued.

11.2 RETA 5771

In recognition of the fundamental importance of land tenure in the development of upland areas, it is recommended that issues related to land tenure and land allocation are made the focus of policy development to be carried out within RETA 5771. In particular, the feasibility of implementing accelerated land allocation through demarcation of village boundaries should be studied further.

Given that many of the upland areas still have extensive forest cover the potential of forestry to trigger community development should be explored in areas, where commercial forest management is technically feasible, and where forests are well-stocked and have potential for commercial sales. The experience gained in Lao PDR in this regard should be carefully tapped.

Regarding agricultural development, it is recommended that in the remote areas the project adopt an approach building on a spontaneous transformation of swidden practices towards market-oriented agriculture. RETA 5771 would act primarily as a facilitator to a development process led by the farmers in view to ensure that the environment in which the farmers operate is conducive to change.

Increasing the staff capacity should be a primary concern for RETA 5771. Training should be undertaken in an extensive and systematic manner.

As regards the overall scope of technical interventions by RETA 5771, there are probably two feasible options, the appropriateness of which should be analysed in detail when designing the project. If the focus is set on development of productive activities, the experience shows that full -fledged integrated rural development projects have often had difficulty in delivering a sustainable impact. While technical issues can be tackled at the project level in a fairly comprehensive manner, institutional problems are usually more deeply rooted and require reforms beyond the project area. Attempts to address them simultaneously at all fronts are susceptible to be ineffective. While the problem analysis has to be holistic, the actual intervention should be limited to key issues. A suitable approach might be to identify 1-2 sectors (e.g. among the above-mentioned options), which have the highest potential to become "engines" of village development in the project area, and concentrate the efforts on their development. "Integrated activities" such as preparation of village development plans would be undertaken, but the thrust of the effort would be on the selected priority areas.

Second option is to support further development of the existing rural development structure. On one hand, the capacity of Village Development Committees would be supported in identifying, financing and implementing development activities within the communities. On the other hand, the local institutions from village up to the national level would be strengthened in terms of their capacity to facilitate and support the process at village level. In this approach the technical input of RETA 5771 would put prime emphasis on institutional development.