Policy Brief Customary Land Titling in Vietnam East Asia and Pacific Region Rural Development and Natural Resources 1818 H Street NW, Washington, D.C. 20433 July 26, 2004 This report was prepared for the World Bank Group. Customary Land Titling in Vietnam Customary land titling has made been possible for the first time in Vietnam under the country's newly revised Land Law1. This policy brief highlights the key findings and recommendations arising from Government-donor discussions on the implementation of customary titling ­ an approach to land tenure and management which finds strong and growing support in many other countries2. In Vietnam, as elsewhere, customary titling largely concerns areas of forest land held or claimed by ethnic minority groups, and th erefore needs to be approached with sensitivity. Customary titling should first be carried out for those communities already seen to be managing forest or "unused" land in practice, before being attempted more widely. This needs to involve: first, drawing relevant lessons from ongoing pilot activities, based on a careful assessment of their impacts. Second, the necessary legal framework requires more complete elaboration, which may be addressed through the current revision of the Forest Protection and Development Law3. And third, the institutional framework for implementation needs to be developed, along with `how to' guidelines and appropriate funding arrangements. Context In the Vietnamese context, customary titling refers to the provision of long-term, statutory land-use rights to community groups, taking into account long-standing principles (enshrined in "customary law") and traditional land-administration institutions that often operate outside or in parallel with the formal legal system, and define how rights are ruled, allocated, and preserved within a community4. Customary titles may be issued with respect to areas of forest and "unused" land dominated by ethnic groups with largely intact traditional social structures. Forests and forest land play an important role in the Government of Vietnam's socio- economic development strategy for 2001-20105. As a result of strong progress in forest-sector reform, forest cover has steadily increased from 8.3 million ha of natural forest and one million ha of plantation forest in 1995 to around 10 million ha and two million ha, respectively, by 2002. However, some 6.8 million ha of officially classified "forest land" has no current forest cover. The officially recorded contribution of forestry to GDP ­ 1.2% ­ is also much lower than the value of the benefits forestry brings and falls far short of estimated production potential. The failure fully to utilize forest production potential may be attributed largely to unclear and sub-optimal land-tenure arrangements and forest-management regulations. The solutions proposed in the Forest Development Strategy 2001­2010 are consistent with overall Party and Government policy priorities, focusing on poverty reduction and livelihood improvement, reducing conflict and fostering social harmony, and ensuring environmental sustainability. Clarification of land rights and forest- management regulations is paramount in realizingthese goals. Communities in Vietnam have a long tradition of sustainable land management, particularly with respect to communally held forest land and so-called "unused land". Such practices are officially acknowledged6, and are cited in justification of pilot schemes being carried out in several provinces including Dak Lak and SonLa to 1 formalize land and forest management by local communities.8 Results so far show that communities are able to manage their forest well following allocation. The allocation process itself is welcomed by co mmunities as it serves to clarify land use in a socially acceptable manner, increases investment in and productivity of the allocated areas, and legalizes people's forest-based sources of income, as well as securing the environmental functions of the forest. For the state, it is much less costly to allocate forest land to communities than to individuals. The corresponding community forest management regulations clarify , at a reasonable administrative cost, the income communities receive from transfer payments and taxes they pay to the state. Outlook The ultimate objective of customary land titling is to bring the forest sector more fully into the mainstream of socio-economic development. That is, it ensures that greater benefits to wider society are derived from the forest, while at the same time enhancing the livelihoods of people in local communities through expanded employment and income-earningopportunities. According to the guidance so far given by the Communist Party and existing legislation ­ notably the Democratization Decree, the Land Law, and the draft revised Forest Law ­ the main principlesof customary titling are as follows: · Benefits to the people: Land allocation, including community titling, should be people-oriented, focusing on the 1.5 million or more households living in forest areas with the highest poverty incidence and most limited means to overcome poverty; · Benefits to the state: The public-good environmental functions of forests must be safeguarded on a sustainable basis, with respect both to local and national interests (e.g. water supply, flood control) and to international agreements. Compensation payments from the state to those ensuring thesepublic goods, and payments to the state for private benefits derived, both need to be clearly specified; · Setting development incentives: People take local development into their own hands, and the state provides the framework for them to do so (including access to land, planting materials, credit, extension and markets), within a peacefuloverall social setting; · Local adaptation: Recognizing that geographical context and development priorities vary between regions and over time, divergent social, economic, and environmental concerns need to be harmonized at the local level through the informed participation of concerned stakeholders. The legal framework needs to provide the necessary flexibility to allow for local and temporary solutions and for learning by doing. Given this objective and these principles, a range of land-use and forest-management options needs to be considered, encompassing private households; village communities; local organisations; local, sub-national and national-level administrative units; and commercial entities. Some examples drawn from 2 international experience with community tenure and manag ement of land and forests are provided in Box 1. Box 1. International experience in community forest tenure andmanagement At least 7%, or some 246 million ha, of the world's forest are now recognized as fully owned by indigenous and other communities and at least a further 4%, or an additional 131 million ha, has been legally set aside for these groups by governments. These averages rise to at least 14% "owned" and 8% "set-aside" when only developing countries are included. While these numbers may appear small, community ownership and access has roughly doubled in the last 15 years. Communities own, or have primary access to, a majority of forests in Papua Ne w Guinea, Mexico and China , and they either own or manage 10 million ha or more in Brazil, the United States, Peru, Bolivia and India. However, a far greater proportion of the world's forest than these figures suggest is actively claimed and/or managed by communities. Several countries, including the Philippines, Panama, Mexico and Colombia, have recognized community land claims as full corporate private ownership. Many others, including India, so far stop short of gra nting full private property rights to communities, but increasingly devolve some management responsibility over forests and forest land to local communities. The global trend is increasingly to devolve resource rights to poor people and communities for their greater access and use9. Take the case of Germany, for example. With a total forest area similar to that of Vietnam (10.9 million ha), eligible forest owners comprise private persons, administrative units (towns, communes, villages), and sub-national or national government . Private ownership extends to 46 % of the total area10. Average ownership is only 5 ha, often split into smaller individual plots. Large individual ownership of over 1,000 ha only accounts for 8% of private land. Due to the small size of most individual holdings, many private forest owners organize themselves in "forest user groups". The functions of these groups usually include arranging for their members' utilization of machinery, access to credit 11, joint material supply and joint marketing and certification. While ownership is private, forest management is done communally. Communities ­ lower administrative units and organizations ­ own 20 % of the total forest area in Germany. The average size of community holdings is 175 ha, with some communities owning several thousand ha. Compared to the European Union average, community management is relatively high12. This reflects the importance local communities place on forestry for improving the quality of life and for providing environmental services. One third (34%) of the total forest area belongs to the German federal states and national level. A state (in Vietnam this would be a province) forest depa rtment on average manages 3,500 ha of state forest. In addition, most of them offer extension services to private forest owners and for communal forest management. Finally, national government owns 410,000 ha (less than 5%) of forest land, mostly for military purposes and along train tracks. Options for implementation Policy options for customary titling vary according to the degree to which customary law is formally recognized in the title. Legalising customary claims in full in accordance with customary law may achieve a high level of acceptance by community groups but could conflict with the wider public-goods objectives of land allocation (e.g. land reservation, forest and biodiversity conservation), and may not 3 help to clarify land tenure in situations in which there are overlapping and contested claims to the same land. A significant gap exists between customary and statutory rights. This gap can be narrowed but not eliminated, since these forms of rightshave very different foundations. The challenge to be addressed is how to narrow thegap between them in such a way as to balance customary claims and land-management practices with wider land-use and environmental objectives. Currently, it is estimated that some 2.5 million ha13 of forest land in Vietnam are de facto managed by communities. This land is already partially allocated and contracted to local communities by Provincial People's Committees (PPCs)14 and certain forest-management rights and obligations are specified. Since these communities have already demonstrated their capacity to manage forests well, they should have first first priority in being assigned customary titles over the land and forest they already manage in practice. Attention would need to be paid to facilitating applications for customary title by the communes concerned. In a second step, consistent with current efforts to reform state forest enterprises (SFEs), transfer of management responsibility over the land they hold from SFEs to local communities should be considered,at least in cases where SFEs are judged to be non-viable. It is estimated that a further 0.5­1.0 million ha of forest land could thereby be released from SFEs for transfer to local communities.15 Again, the process should begin with an application from the commune level. Local people should be involved in deciding which areas are to be used for production purposes, which areas for delivering public-goods benefits (e.g. special use forest, watershed protection forest), which areas to be placed under local management, and which areas should be managed by a higher-level public body . Production forests may be managed by commercial entities, whether the state, the local community, or a firm, household or an individual holds private title to the land. It is clear that there will still remain important roles for other actors in the forest sector. State management will continue to be required for some large, contiguous areas of protection forest in remote locations in which the local population is as yet unwilling to take responsibility16. Small-holder forestry may be the preferred option where there are limited economies of scale and externalities, such as in more scattered production forest areas. There is one omission in the new Land Law which poses a significant constraint to the implementation of customary titling: namely, that "community" is not specifically mentioned as an eligible forest-land recipient. Although communities are mentioned (in Article 9) as an eligible recipient of agricultural land ­ of which forests are a sub-category ­ they are not so named in the relevant articles on production forests (Article 75) or protection forests (Article 76). One solution has already been proposed in the draft revised Forest Law, which is currently under review by the National Assembly and expected to be approved in December 2004. The revised Forest Law is expected to recognize existing forest-land use by communities and may call for those rights to be recorded in formal titles 17. Beyond this necessary legislative clarification, the implementation of customary titling also calls for careful policy design. Two key principles should guide the issuance of community titles to forest land: 4 · Strong demand from local people based on the role of land and forest resources in their livelihoods: local people are generally found to hold clear views on whether they would like land and forests to be managed individually, by groups of households, by the village, or by the commune;18 · Agreement of Government that customary titling is currently the most appropriate tenure solution, taking into account social (equity, conflict resolution), economic, and environmental objectives. Such agreement should be reached through a participatory process involving the commune level in decision making. Specific criteria to be applied in granting community titles could include: i. Allocation to groups of up to 50 ha19; and in the case ofvillages/hamlets, no area ceilingto be applied within the boundaries of the village/hamlet20. Boundaries between land holdings must be clearly demarcated on maps prepared through a participatory process and accepted without contestation. Forest classification and land-use planning/ zoning should be completed; ii. Land categories eligible for titling: production-forest and scattered protection-forest land (< 5,000 ha)21 including bare land, natural regeneration areas, plantations and natural forest; iii. Absence of conflict within the existing, de facto customary tenure system and in forest-management practice within the community, and between the community and the responsible administrative unit. In the case of titling to groups as a sub-set of the village community, the group must obtain the consent of the wider village community; iv. General agreement on existing principles and management regulations applied by the community based on prior access to information on options, fairness, equity and gender concerns; v. Current forest-land management practices of the community are judged acceptable or are expected to become acceptable once rights are granted and enforced and support structures become active; and vi. The above-mentioned conditions are judged likely to hold at least over the medium term. Based on current experience, Government may choose between two policy options: pilot schemes could be extended to additional provinces (Option 1); or all forest-land areas managed de facto by communities could immediately become eligible for community land titling (Option 2). Potential benefits and risks The potential benefits of facilitating customary titling of forest-land use include the legal confirmation of existing rights, and the resolution of conflicts where existing claims overlap and are contested. Allocation to communities is often socially more acceptable and economically more viable than allocation to individuals. Clear allocation to communities potentially leads to more responsible utilization of the economic and environmental potential of forest areas. Clarification of management 5 responsibilities of the community may lead to an increase in the economic benefits. Clearer and mutually agreed principles on benefit-sharing between the forest manager and the state permit the real economic contribution of the forest sector to the overall economy to be more fully recognized. However, careful policy design is required to mitigate risks in the three main areas identified: (a) Conflict resolution: forest-land allocation to communities may not resolve existing conflicts, and new conflicts may arise. This risk needs to be managed by building on existing experience in process facilitation; by applying well-developed, local guidelines for participatory land allocation; and by proceeding slowly; (b) Forest status: forest-land allocation to communities could lead to a reduction in forest area. In order to reduce this risk, approaches to community forest management and planning, benefit-sharing, and administrative facilitation and monitoring tools all need to be improved. In the current draft implementation guidelines, all three issues are addressed based on successful pilots.22 (c) Costs: both the investment costs for forest-land allocation and the recurrent costs for forest management are hotly debated. Although the cost of a participatory process of forest-land allocation to a community of users is much lower than that of allocating an equivalent area of small-scale agricultural plots to individual farmers, it remains much higher than current cost norms allow.23 Whilehard economic data are currently lacking, the investment appears justified on the grounds that it aims to enhance the livelihoods of the poorest people in rural areas, and promises to secure social peace. Regarding recurrent costs: there is a risk that with the allocation of forest- management rights to communities, the Government's payment of "protection fees" or investments under forest-protection programs (e.g. the Five Million Hectares Program) may cease. Since the economic benefits to the community of assuming the management of currently depleted forests are unlikely to flow within the first 5-10 years, however, it is suggested that such payments be maintained initially and be phased out over such a period. The elaboration of community forest-management plans, and active investment in forest management, would also require the community to have access to the necessary funds, including credit from bankin g institutions. The establishment of "community forest funds" is currently under discussion. Rather than coming from user fees paid by community members (according to Article 23, draft revised Forest Law, communities are exempt from paying such fees), such funds should come from payments by economic entities that benefit from the improved forest resource to the state and other sources. This builds on recent decisions by the Ministry of Home Affairs that facilitate a more flexible approach to local-level fund management (e.g. maintaining bank accounts) including villageand commune development budgets. 6 Conclusions and recommendations Based on analysis of the existing situation, MARD's own internal reviews24, and discussions with MONRE staff, it appears that there is currently no viable alternative to community land titling for those communities de facto managing forests. The role of the state in fostering an enabling environment for both public and private operators in the forest sector ­ including establishment of the required legal basis, funding, and staff resources ­ should be to ensure an appro priate balance between those operators' objectives and broader forest-sector goals, and to supervise the implementation of policies to maintain this balance. The following recommendations are made: First, given the complexity of th e issue, issuecustomary land titles initially to those communities already de facto managing forest and "unused" land. Of the total land area of around 2.5 million ha in these categories, some 1.1-1.4 million ha have already been allocated or contracted to communities by provincial authorities. Second, based on a comprehensive assessment of th ese existing pilot sch emes, broader consensus on community land titling should be sought. Wider applicationof customary titling beyond those areas already managed in practice by local communities could then be considered, in parallel with further restructuring of SFEs. A qualification process for communities could also be designed in this second step ­ for example, following a jointly defined management plan with increased coach ing and support ­ as a prerequisite for applying for community land titling. In order to prepare for the first stage of community land titling, the following issues need to be addressed in the short term: 1. Clarify the legal framework, through a joint MARD/ MONRE working group: a. Make provision for forest-land allocation to communities under the revised Forest Law and supporting guidelines for its implementation. Define community initially as "village"; b. Define clearly those eligible categories of land over which community titles may be issued (e.g. fragmented areas of protection and production forest), and criteria for proceeding (e.g. application from the commune, follow a participatory process for settling any overlapping and contested claims); c. Define the objective of forest-land allocation and process of application for community titling (including the valuation of the forest outputs and forest land); d. Regulate the (short-term) re-categorization of land sub-categories based on defined principles and priorities (consistent with Article 36 of the Land Law); e. Restrict the right of communities to transfer their newly acquired land rights (as regulated in Article 69 of the draft revised Forest Law); 7 f. Identify the person who holds the land-use title on behalf of the community. It is recommended that s/he should be the village head and should be chosen by the community; g. Clarify the supporting role of the state in this process (facilitation, decision making by district); h. Clarify the transfer of benefits to the state (based on an inventory/ valuation); i. Clarify obligations for forest management (participatory inventory, community forest-management planning process ­ similar to the management plans of SFEs or other forest-management entities); j. Clarify the eligibility criteria for communes to receive further investmentfrom the state. 2. Establish the administrative structure and sources of funds for facilitating community land titling. For example, criteria for regulating access to the proposed Forest Fund ­ which could be used by forest managers including communities ­ might be pilot-tested using funds under the recently signed, multi-donor "Trust Fund for Forests".25 3. For further clarification of the issues involved, design a rigorous impact assessmentof ongoing pilot schemes supporting customary land titling. Some precedents exist for providing technical expertise and designing impact assessments of land allocation schemes for poverty alleviation and environmental goals.26 Sources CPV, 2003. Politburo Resolution 28-NQ/TW "On further structuring, renovation and development of State Farms and State Forest Enterprises". Hanoi: Communist Party of Vietnam, June 2003. Foerster, Elke, and Ulrich Apel, 2004. Vietnam Customary Land Tenure Study. Report to World Bank, Hanoi, May 2004. Government of Vietnam (1998). Prime Minister's Decision 661/QD-TTg, dated 29 July , 1998, "On objectives, tasks, policies and organization for the establishment of five million hectares of new forest", in Ministry of Agriculture and Rural Development/ Departmentof Forestry Development (2001), National Five Million Hectare Reforestation Programme, 1998-2010, Hanoi. MARD, 2003, 2004. Draft Decree on the Revised Forest Protection and Development Law. Hanoi: Ministry of Agriculture and Rural Development, Government of Vietnam (various drafts, December 2003­July 2004). MARD/Forestry Strategy Group, 2000. Forestry Development Strategy, 2001-2010 (3rd draft). Hanoi: Ministry of Agriculture and Rural Development, Forestry Strategy Group, December 2000. 8 National Assembly, 2003. Land Law of Vietnam 13/2003/QH11 dated 26 November, 2003, Hanoi. National Working Group on Community Forestry (NWGCF) and Department of Forest Development, 2003. Proceedings of National Workshop onAllocation and Management of Natural Forest in Community Forestry, Hanoi, May 2003. Quan, Nguyen Hong, 2003. Evaluation of the Pilot Implementation of"Participatory Forest Resource Assessment and Community Forest Management Planning" in Muong Pon Commune, Dien Bien District, Lai Chau Province. National Consultancy Report No.16 for Social Forestry Development Project (SFDP), Song Da. Son La PPC, 2004. "On the Results of Forestland and Forest Allocation in 2001-2003". Internal report. Son La: Son La Provincial People's Committee. Thanh, Tran Ngoc, Nguyen Quang Tan and T. Sikor, 2003. The Local Outcomes of Forest Land Allocation: Evidence from Dak Lak. Department of Agriculture and Rural Development, Dak Lak . 1 13/2003/QH11 dated 26 November, 2003. The lead Government agencies are the Ministry of Natural Resources and the Environment (MONRE) and the Ministry of Agriculture and Rural Development (MARD). 2 These findings and recommendations are drawn from a longer study : Elke Foerster and Ulrich Apel (2004), "Vietnam Customary Land Tenure Study '. Report to World Bank, Hanoi, May 2004. 3 Hereinafter referred to as the "revised Forest Law". 4 A community is defined in the Land Law as a group of people who live in the same village or hamlet (thon, ban, buon, lang, som, doi, etc.) and share the same customs, habits, and/or clan identity (Article 9.3). Customary titling is understood in the Land Law and wider Vietnamese literature to refer more broadly to "land allocation to communities". 5 See, for example, Proceedings of the Ninth Party Congress on Socio-Economic Development Strategy 2001-2010 (Communist Party of Vietnam, 2001), Five Year Plan 2001-2005, and Vietnam's Comprehensive Poverty Reduction and Growth Strategy (CPRGS, 2002). 6 For example, "On the Implementation of State Management Responsibility on Forest and Forestry Land at all Levels" (Decision 245/1998/QD-TTg, dated 21 December, 1998); also Circular 56/1999/TT - BNN, and Decision 08/2001/QD-TTg. 8 In Dak Lak, between 1999 and 2002, a total area of around 16,000 ha was allocated to ethnic minority people in 18 villages in the form of Land Use Certificates (LUCs). Roughly one third of the total area was allocated respectively to individual households, groups of households, and village communities (Thanh et al., 2003). In Son La, between 2001 and 2003, some 685,000 ha of forest land was allocated via LUCs to local people, including 426,000 ha under natural forest. Around 54% of the total area was allocated to village communities, 8% to groups of households, 18% to organisations (including mass organisations at village level, such as the Youth Union and Women's Union), and 20% to individual households (Son La PPC, 2004). 9White, Andy, and Alejandra Martin. 2002. Who Own's the World's Forests: Forest Tenure and Public Forests in Transition. Washington, DC: Forest Trends. 10http://www.wald-online.de/wdt_waldfl.htm 9 11For example, from the European Union's Special Accession Programme for Agriculture and Rural Development. 12Within the EU, prior to the 2004 expansion, 11% of forest land (130 million ha) was owned by communities and organizations, 65% was privately owned, and 24% was owned by the state (Confederation of European ForestOwners: 1997/1998 Informationsschrift). 13See Quan (2003). 14Figures are difficult to verify, as in some cases short-term contracts and long -term LUCs are mixed together, and the process is on-going. Currently, comm unity management includes management by commune administrations or districts' Forest Protection Units, as well as management by villages and groups of households. 15Based on Party Resolution 28/NQ-TW in June 2003, and Government Decision No. 179 in September 2003, in relation to ongoing reorganization, renovation and development of SFEs. 16The Feb 2004 Convention on Biodiversity decision (to which Vietnam is a signatory) on Protected Area Management clearly recognises Community Conservation Areas as an impo rtant platform for biodiversity conservation. 17Articles 44.2 and 54.3 in the draft dated 16 March, 2004. 18It is understood that facilitation and support is required especially for ethnic communities which do not readily understand the implications of statutory land law. Examples of successful process facilitation exist and are documented. 19Son La Province has set a ceiling of 30 ha. 20Son La Province has set a ceiling of 500 ha. However, since the question remains, who is responsible for the remaining forest within the village/hamlet boundaries if the area is larger than 500 ha, the draft revised Forest Law advocates a limit of 5,000 ha, beyond which a watershed-management board or some other Government organizationwould have jurisdiction. 21 Article 43 (on "Organization of Protection Forest Management") of the draft revised Forest Law provides that areas of protection forest of less than 5,000 ha may be allocated to village/ hamlet communities. 22The National Community Forestry Working Group under the leadership of the Forest Department in MARD has documented a number of pilots in the proceedings o f various national workshops held in 2000, 2001, and 2003. 23The actual costs incurred in participatory land allocation range from around 25,000 VND/ha to 50,000 VND/ha, depending on the size of plots, the local terrain, and the degree of difficulty faced in mediating among competing claimants. This is substantially higher than the current cost norms of around 2,000 VND/ha. 24See proceedings of national workshops of the National Working Group on Community Forestry held in 2000, 2001 and 2003. 25AnADB-supported, forest-related project design TA (PPTA 3818) is also considering funding such an instrument, as are other donors. 26See, for example, Thanh et al. (2003) 10