Page précédente Table des matières Page suivante

English summary


Land reform began in the Lao People’s Democratic Republic (PDR) in the early 1990s with three main objectives: to stop deforestation (deemed to result mainly from swidden agriculture), to intensify agricultural production and to improve the government revenue base through land taxation. Private ownership of land and increased tenure security are expected to encourage agricultural investment, intensive use of land and the rise of a market-oriented agriculture.

The land policy of Laos is quite similar to those in other Asian and neighbouring countries, such as Thailand. The case of Laos is nevertheless particular because of the social and cultural context in which the policy is being implemented. Ethnic minorities living in mountain areas and relying mostly on slash and burn agriculture, who are the prime target of the new land allocation procedures, account for more than 40 percent of the national population, while in Thailand for instance, they represent only one percent of total population.

This paper describes the results of a study funded by FAO on the impacts of the land reform in Laos on land access and rural livelihoods, focusing more particularly on northern regions and changes in swidden agriculture. The study draws on interviews with Lao Government officials and foreign experts in Vientiane during October-November 2003 and an analysis of existing literature. This work has greatly beneficiated from the discussions held with the Lao officials at the Department of Lands (Mr Khamsouane Sisouvong), Ministry of Agriculture and Forestry (Mr Thongpath Leuangkhamma) and the Department of National Land Use Planning and Development (Mr Phoumi Vongleck and Mr Anuthay Chanthalansi), and from the advice of Ms Krijavainen, FAO Representative in Laos and the FAO team in Vientiane.

Legal and institutional framework

The two main laws are the Forest Law (No. 96/NA11) and the Land Law (33/PO of May 1997). The Forest Law defines an official classification of forests in five categories:

In accordance with the Land Law, the State issues Temporary Land Use Certificates (TLUC) for plots of a maximum 25 hectares per labour unit to each household. These TLUC are inheritable and considered as a first step towards delivery of full ownership titles. They are valid for only three years (art. 18) and cannot be sold before the permanent full ownership title is obtained. A TLUC can be withdrawn if the beneficiary does not comply with land use regulations (art. 62).

Three problems make the implementation of these regulations particularly difficult:

1) The land reform process does not follow the same procedures in rural, urban or peri-urban areas. In the rural areas, a land allocation process is implemented, while in the urban and peri-urban areas only, is actual land titling realized. Only with land titling can the land owners obtain private ownership titles transferable against payment. In theory, the two processes are parallel and complementary, since a temporary use title can in theory be converted into permanent title, but in reality, the procedures are not compatible at present.

2) Three institutions share the responsibility of the implementation of the reform: the Ministry of Agriculture and Forestry (MAF), the Ministry of Finance (MoF), and the Department of Land Use Planning and Development (DONLUPAD). In practise, MAF works mostly in rural areas and MoF in urban or peri-urban areas as implementing units while DONLUPAD is given a more strategic function (Decree of the Prime Minister No. 237/PM 11 December 2001).

3) The role and the respective importance of each of these three institutions is being currently debated and negotiated at a high political level. This is understandably a politically sensitive issue. It emerged from the interviews conducted in Vientiane that collaboration between MAF and MoF should be increased but at the same time, adding a new structure above them may add difficulties rather than ease the process.

The implementation of land allocation

The process of land allocation is implemented in each village according to a standard methodology, designed jointly by the National Agriculture and Forestry Research Institute (NAFRI) and the Forestry Inventory and Planning Division (FPID) with the collaboration of the Lao-Swedish Forestry Programme, and officially adopted by MAF in an instruction in 1996. In this methodology, the land-use plan (LUP, or kan vang pén nam saï ti din) targets the whole village community, while the land allocation procedure (mob din mob pha) is performed at the household level. The whole process takes between 45 and 60 days according to the head of the Reduction Shifting Cultivation Extension Centre (RSCEC), and sometimes more when there are conflicts within or among villages. Some of the studies that have been commissioned by NAFRI or GTZ on the other hand show that local LUP/LA teams spent only a short time (5 to 10 days) in the villages (NAFRI-LSUAFRP 2003a & 2003b; Rock 2003)

In the course of the allocation procedures, at least four documents are prepared and signed by the villagers:

According to the data of RSCEC, around 350.000 households (50 percent of the total of the country) in more than 5300 villages (40 percent of the total number in the country) have been given TLUCs in the whole country. Another document (MAF, 2003:65) mentions 6188 villages and more than 370.000 households. Assuming that each farmer received on average two or three plots, one can estimate that between 600.000 and 1.000.000 of titles have been issued during the last ten years. None of these titles has been converted in permanent ownership title until now. As a result, large numbers of TLUCs have expired according to the law, but are nevertheless still considered by the farmers as proofs of their legitimate claims to the land they use.

It is important to note that the number of villages where a LUP/LA procedure is performed each year has been slowly diminishing since 1996 (1200 villages in 1996 compared with only 315 in 2002). This is probably because interventions are now taking place in more difficult areas where negotiations are longer and more complex.

The implementation of the land titling process

Land titling began in 1997 with the Land Titling Project, funded by the World Bank and with technical assistance provided by AusAid. The first phase ended in November 2003, and a second phase will soon begin. The objectives of the project are to (i) improve the security of land tenure; (ii) develop transparent and efficient land administration institutions at the national and provincial levels; and (iii) improve the government’s capacity to provide social and economic services through a broader revenue base from property related fees and taxes.

The land titling process occurs in two different ways:

The social impact of land allocation

According to the survey on poverty in Laos commissioned by the ADB, rural people mention land allocation as the first cause of impoverishment in three regions (north, east and centre) out of four (ADB, 2001: 38). This negative perception appears to be the outcome of many related factors: the general perspective of the reform, the reduced access to land, the degradation of local living conditions, the absence of agricultural intensification and the continued relocations of villages from highland areas.

1) The official documents drafted by MAF and the GOL make a clear distinction between two types of slash-and-burn cultivation: cyclical (long fallow) or pioneer (the land is cultivated for several years and then abandoned). The first is considered potentially sustainable when land pressure is low. It would be excessive however to say, as Charles Chamberlain and Panh Phomsombath do (2003: 35), that cyclical agriculture is nowadays officially accepted in Laos. Recent field investigations have shown that the above distinction has not much practical meaning, and that a restrictive approach towards slash-and-burn agriculture is still practised everywhere. Rotating slash-and-burn agriculture is temporarily tolerated, provided that fallow duration does not exceed four years. The aim is to compel farmers to change their agricultural systems through an administrative ceiling on legally permitted agricultural space, or in other words, by creating land pressure and scarcity through land regulations[1].

2) For the past 30 years the Lao government has encouraged mountain peoples to leave the uplands and to go to live in the valleys, thereby regrouping people into larger centres as a means of ensuring access to basic infrastructure and services, and of exercising greater control. The willingness of MAF services to reduce the space available for itinerant agriculture is clearly visible in the statistics of RSCEC: 82 percent of land allocated between 1995 and 2002 has been classified as forest land (MAF, 2003: 43). Everywhere, as a result of land allocation, protected spaces have become more important than those that can be used for agriculture. In this respect, land allocation has a stronger effect than the land-use planning operations which are the preliminary steps before the allocation. The studies conducted by NAFRI in Phonexay district (Luang Prabang province) show that in villages where land use planning has occurred, each household has access to four ha of agricultural land and five ha of forest while in villages where the land allocation has been completed, each household has access to 2,7 ha of agricultural land and nine ha of forest (NAFRI/LSUAFRP, 2002).

3) Local studies show that the decrease of available agricultural spaces has negative impacts on the food security and livelihoods of households, but no long-term monitoring or comprehensive statistical surveys have been carried out. With the important decrease of fallow time (most often 3-4 years nowadays, instead of 7-15 years previously), and with stagnating agricultural techniques, the fertility of land diminishes steadily. Paddy yields are down by more than 50 percent (ADB, 2001: 35-39), and time necessary for weeding has increased significantly. Because of increasing pressure on forest areas, non-timber forest products are becoming rarer. This has a negative impact because these product account on average for 55 percent of the monetary income of rural households in these areas (UNDP, 2001: 78). This impact is even more important for the poorer groups who rely more on these products to cope with food deficit and insecurity.

4) Faced with this diminishing natural resource base, farmers do not seem to have a sustainable alternative available. Technical advice and support from State services are limited and have little impact. In general, public extension support remains very weak: an ADB study revealed that none of the 91 villages studied in the 43 districts had received technical advice on intensification of agriculture, neither for development of lowland paddy fields nor for upland cash crops. A survey undertaken by NAFRI in 2000 revealed that, out of 49 farmers interviewed, only 40 percent were able to increase their area under flooded rice in the lowlands. The outcomes of this situation appear to be contrary to the Government’s stated objectives: in the northern provinces, the area under upland swidden rice, which was expected to decrease in favour of flooded rice, actually increased by 47 percent. In a study commissioned by NAFRI in 2000, only 40 percent of the farmers interviewed had been able to increase their paddy land after land allocation. In northern areas, land allocation appears to have a paradoxical impact on swidden cultivation. One recent study (Keoketsy, Bouthabandid & Noven 2000:14) noted that the production (and most likely the surface area) of upland paddy had actually increased in 47 percent of the cases it considered. It should be emphasised, however, that there are usually strong variations between villages in the outcomes of the LA/LUP process, depending on the sample of villages and/or regions studied. Central and southern provinces seem better off than northern ones. There are also local success stories with agricultural intensification and increasing peasant incomes (Boloven, Kentao), but these remain exceptions - generated by favourable local conditions - to the general and worrying trend.

5) Land allocation occurs, notably in the rural areas of the North, in a spatial context characterized by important spontaneous migrations not controlled by the State and a macro-change in global land occupation. Many persons interviewed during the study declared that the land allocation process will generate clear positive results only when massive population movements involving fractions of villages or even whole villages will diminish. Such movements are often initiated, but not actually controlled, by the State. These movements and their significant social cost are also an outcome of insufficient coordination between State services: the office of the Governors of provinces is in charge of population relocations, while the local and provincial services of MAF deal with land allocation. The coordination between these institutions is weak or absent, while there are obvious links and interdependent effects between both processes. As a result, village resettlement or displacement initiated by the provincial authorities sometimes conflict with the land agreements concluded between older resident villages and District Agriculture and Forestry Office (DAFO).

Informal strategies to access land in rural areas

Directly and indirectly, the primary effect of land allocation is to increase pressures over land in almost all villages. As a result, there are increasing social tensions between migrants who have recently arrived in a village following government-sponsored resettlement and the original inhabitants. Tensions are also often increasing between neighbouring villages, as a result of a new type and conception of the “border” (or limit) between villages. While in the traditional system there were some inter-village areas with diverse legitimate village claims (and some kind of co-management of these border spaces by the various neighbouring villages), there are now “administrative spatial units” (khét phok khong ban), i.e. separate and homogenous space units as defined by an administrative process. With this homogenous definition of spatial village units with exclusive ownership and management rights officially recognized for one single village, the traditional overlapping rights for some activities (hunting and wild-fruit harvesting/gathering) are no longer jointly managed, causing an increasing number of disputes over these rights.

These difficulties seem to make the future implementation of the land allocation more and more difficult: while villagers most often had a priori a positive attitude towards the formal registration of rights, the villages (usually the more remote ones) who have not yet been affected by land allocation seem to become more and more reluctant, because they are aware of the difficulties encountered by their neighbours. Moreover they also consider that their customary system is more flexible and fairer than the new system. This can be a reason why the pace of land allocation process (as measured by the number of villages where land allocation is performed each year) is slowly decreasing during the recent years.

In a general context where the overall pressure on land increases, the role of mobility remains central. Rural households try to adapt in order to preserve livelihoods and previous levels of income but agricultural techniques evolve very slowly. The most important way to preserve livelihoods is to try to obtain new land, in order to compensate for the decreasing returns of cultivation on agricultural plots with increasingly shorter fallows. The following strategies for access to (new) land have been observed in the northern regions of the country (NAFRI/LSUAFRP, 2002; 2003a & 2003b):

The social impact of land titling

A socio-economic baseline study regarding land titling has been conducted recently for the Department of Land (MoF, 2003) but the social and economical indicators proposed in the report will have to be tested during the following years before a realistic analysis can be made on that matter. No clear information seem to be currently available regarding the duration of land transactions before and after land titling, neither on the average time during which one plot is owned by the same person, nor on the transparency of procedures. Economic data are equally incomplete, as a result of the absence of detailed and reliable indicators from the outset. It seems that land titling is correlated with an increased value of land (MoF 2003: 36), but no detailed information is available on the average importance and the geographical differences of this increase.

Until now, the Land Titling Project (LTP) has mostly worked in urban areas and with middle-class and high income groups and areas. For this reason perhaps, “it was perceived to benefit mostly to the rich” in the survey area (MoF, 2003: 48). The poorest groups, usually living in remote peri-urban areas, did not have opportunities until now to secure their land rights through land titling, either because no systematic adjudication was done in their quarter district because they could not afford the costs of titling within a procedure of sporadic registration, or because they did not have tangible proof of their rights on the plots that they used and owned. This is particularly the case for ethnic minorities groups recently settled on peri-urban land belonging to the State, where the land titling project is not applicable (MoF, 2003: 47), or who have taken possession of their plots through land-clearing, without any administrative procedure or even a simple record of their settlement by the local village chief. This implies a risk of long-term marginalization of the most disadvantaged groups among the urban and peri-urban populations (minorities, illiterate groups, occupants of public lands, etc.)

Among the most frequent complaints about the land titling process, people mention that titling allowed them to secure rights over built plots, but not on agricultural land. In the areas covered by the LTP, 74 percent of the land is being used for agriculture and only 14 percent for buildings, but among the titles delivered until now, 67 percent are for housing and only 27 percent are for agricultural land (MoF 2003: 38). This apparent paradox stems from the fact that pre-existing conflicts over land are mostly about agricultural land. The adjudication process is thus much slower for these areas, while built plots present simpler situations where proof of ownership is much more widely available.

As the LTP will widen its area of intervention, the social and legal difficulties and the complexity of its implementation will probably also increase. Sporadic and systematic titling will be done for plots where some TLUCs have already been distributed and sometimes transferred or sold without registration of these transfers. The sporadic and systematic titling procedures in the new areas are likely to involve plots for which TLUCs have already been issued and sometimes transferred or sold. A 1998 study shows that of 179 parcels in six villages, 85 percent were involved in unregistered land transactions, with an average of two such unregistered transactions per parcel. The main reasons given for not registering the transactions were poor knowledge of the procedures and the high cost of registration, including travel.

The project will also work in areas where there has never been real private ownership of land, and where customary land systems do not recognise the right to alienate land, even by a person who is considered to have individual rights on this land. The project will have to adapt to various complex systems of social organization and customary regulations. It will be important to assess and understand these systems before introducing and implementing the new land titling procedures and regulations. At this stage, the project has only a very limited information base on these issues. The available information seems to reveal mixed and divergent trends. The impact studies on gender relationships show encouraging results for LTP I (Zakout 2002, MoF 2003) both because of the kind of populations concerned (well educated, urban) and their social patterns (ethnic Lao populations have bilateral inheritance patterns with an inflexion towards matrilineal side). Nevertheless, the same studies conducted previously for Land Allocation programs among populations with patrilineal systems show that the administrative process tended to reinforce or at least to perpetuate gender inequality. A study undertaken in 1999, by the Union of Laotian Women and GRID, showed that the land title in the majority of cases was allocated to the head of household alone even in cases in which the parcel was jointly acquired by husband and wife or through the wife.

The Land Titling Project should undertake two comprehensive studies. The first one should focus on customary land tenure systems (land ownership and management, marriage, inheritance, land and residence, work organization) among various ethnic groups and regions and on the way these customary rules have been adjusted and negotiated in local inter-ethnic systems. The second one should analyse the social dynamics (land transactions, changes in residence and housing patterns, state-sponsored and spontaneous migrations, etc.) which are currently affecting rural and peri-urban areas of the provinces concerned by LTP II.

Main conclusions and proposals

Conclusion 1

Land allocation procedures are implemented in very diverse situations and contexts. Although the phase of land delimitation was completed in more than 50 percent of the country’s villages, the subsequent steps of village land delimitation and allocation of plots to households have been completed to a much more limited extent. These variations will probably be increasingly complex and can cause difficulties in the future.


It would be useful to have monitoring tools containing village-level information on: the current state of land allocation, the stage reached in the land allocation procedures, pending problems and land transactions that occurred after the land allocation.

Conclusion 2

Many recent studies show that land allocation has either a weak positive impact or negative effects on rural livelihoods, degradation unfortunately being more frequent than improvement especially in the northern regions of the county. The degradation of livelihoods often takes place while, at the same time, farmers are engaged in illegal forest clearing (not accounted for in the statistics of RSCEC). Forest protection and the reduction of space available for swidden agriculture are the main focus of land and forest policies, with a limited awareness of the risks that this implies for the evolution of livelihood and food security in rural areas. This predominant approach overlooks the importance of crop diversity (allowed by itinerant agriculture) and of access to non-timber forest products.


(i) Develop stronger monitoring tools to support the development of appropriate land policy. Accurate indicators related to rural livelihoods and food security would allow, for instance, the earlier identification of problems caused directly or indirectly by land allocation. It would also entail more realistic statistics regarding the extent of surface areas with slash and burn agriculture. Training local administrators in these surveys would also enable them to adapt local action and to have some room to manoeuvre, if needed, when unrealistic objectives are assigned to them by the provincial authorities. If data gathered indicate a degradation of livelihoods following land allocation, it would justify a pause in the procedures of LA/LUP to inquire about the causes of this degradation.

(ii) Adopt a more flexible approach for the implementation of the legislation on forest regeneration, and allow villagers, at least during a first phase after land allocation, to continue to use long fallow cycles in order not to put their livelihoods at risk.

Conclusion 3

It has often been noted that LA/LUP procedures are not genuinely participatory. This limited participation can be a cause for the lack of interest or defiance of the villagers towards the process of land allocation. Participation is requested mostly, or only, when information must be collected: villagers are not actively involved during the subsequent steps. Some sources mention that quite often the technicians of DAFO seek to complete the procedure in only one visit and as quickly as possible. Very little is know about the villagers’ understanding of the various steps of the methodology for land allocation, and how they perceive the value of the documents delivered by DAFO.


(i) Organize the various steps of the LA/LUP procedure in independent modules that can be completed at an adequate pace and with more flexibility than currently. This can be done by drawing on the experience of the HPSPSPP (Hua Phan Shifting Cultivation Stabilization Pilot Project) in Samneua province.

(ii) Gain a better understanding of local knowledge in areas such as soil and forest classification. The experience of the LSUAFRP (Lao-Swedish Upland Agriculture and Forestry Research Programme) on indigenous systems of soil classification has resulted in better communication with villagers and made the work of extension workers more effective.

(iii) Provide villagers with skills to use of maps of the village territory thereby enabling them to participate more effectively in the processes of zoning and allocation.

(iv) Conduct a detailed study of the villagers’ understanding of the methodologies for land allocation, and of their perceptions of the meanings and ‘value’ of the document delivered by DAFO.

Conclusion 4

The arrival of migrants most often implies new arrangements and transactions for access to land, and an increased pressure on arable land. It can also entail conflicts within the villages with rapidly growing populations, or between these villages and their neighbours. The arrival of migrants can also generate induced, non-controlled displacements, either of migrant groups who cannot settle properly in their new area, or by indigenous groups who sell their land and leave to settle in another site. Many of these land transactions induced by spontaneous displacements are informal. It is important for those responsible for the implementation of land policies to be aware of the (sometimes significant) differences between local practices on the ground and official regulations. Land allocation and land titling will have a more effective and positive impact only if their implementation takes into account the population movement and resettlements, both State sponsored and spontaneous.


(i) Zoning or land allocation should not take place when a village is scheduled by district authorities to receive new families. Regarding this point, a closer collaboration is advisable between local MAF services and governors as well as district chiefs’ offices.

(ii) Organized displacement and resettlement of highlands communities should be avoided and considered with more caution. Instead of organized displacement, it is advisable to consider and implement alternative solutions for the development of highland villages. In many cases, it would be enough to build small tracks in the remotest villages to improve radically their accessibility, and in this way, make the implementation of land zoning/land allocation feasible without disrupting completely existing (and evolving) agrarian systems. Several studies (ADB 2003; NAFRI-LSUAFRP 2003a & 2003b, Mohr 2003; Rock 2003) show that poorly organized displacement generate numerous problems regarding the implementation of land allocation programmes and that these problems could be avoided by “on site” development in highland areas.

(iii) Donors and international development partners, in agreement with national authorities, could increase their funding and actions to support development of highland areas, in order to encourage local administrators to find alternatives to displacement. Too often, local authorities displace entire villages to areas unsuitable for lowland paddy agriculture and then seek support to address with the difficulties so created. Investing energy and resources earlier to support sustainable highland development would thus be an effective approach in such cases.

(iv) Spontaneous displacements account for a major part of current population movements, and cannot be always anticipated. They often induce informal land transactions. In order to assess the effects of these displacements on local land tenure, it would be useful to design tools and simple procedures in order to monitor evolving land transactions and inter-village agreements about forest and land-use.

Conclusion 5

Land titling has until now been implemented in urban areas and with relatively wealthy social groups. Peri-urban populations, often poorer and from ethnic minorities, have not been able until now to benefit much from these procedures.


Social criteria used in selecting areas for systematic titling need to reflect also the degree of vulnerability of resident groups and the benefits they can derive from titling. Sporadic titling should be limited and closely monitored, with a special attention to the costs born by the beneficiaries.

Conclusion 6

As land titling will increasingly be undertaken outside cities, the social and legal complexity of the process will increase accordingly. The DoL should take into account the cultural features of populations to be affected and, where relevant, the zoning or allocation procedures implemented previously by MAF services. These previous operations will have to be integrated within the titling process in one way or another.


Carry out detailed social studies in peri-urban areas affected (or that are going to be affected in the next few years) by the Land Titling Project, including the following aspects: social groups and history of the settlement of various groups; customary land tenure and management rules of the groups concerned, access to land for the most disadvantages groups through land transactions, migrations in and out, demographic profiles of the population, and changes in residence and housing patterns. Ensure that the outcome of these socio-economic assessments is actually fed into the implementation of the land titling project.

[1] Since 1996, the areas under slash and burn decreased by 37 percent and the number of households practicing swidden agriculture decreased by over 50 percent. However, the decrease is unequally distributed between regions. Between 2001 and 2003, the slash and burn area decreased by half in the South and Centre, but by less than one quarter in the North (in two northern provinces, Louang Nam Tha and Houaphan, the area under slash and burn actually increased during the same period).
[2] Most of the cases where this strategy has been observed are in the northern part of the country (Keotetsy, Bouthabandid & Noven 2000:11)

Page précédente Début de page Page suivante