1Shifting cultivation plays an important role in the Lao economy and society. In a country where 80% of the surface area is hilly or mountainous, those crops provide jobs to over 250,000 families (MAF 1999) —i.e., 35% of the country’s population— among the country’s poorest and, for the most part, belonging to ethnic minorities from the remote uplands in the northern, eastern and south-eastern regions of the country.
2Those farmers are often accused of destroying the forest and many authors hold them responsible for deforestation in Laos (UNDP 1995; Watershed 2000; NAFRI et al. 2003; Vorakhoun 2003): in 50 years, the country’s forest cover has reportedly gone from 70% to 51% (MAF 2000). For a resident of the lowlands, culturally used to distinguishing permanent farm areas—rice fields, gardens—from forest areas—protected forest or silviculture exploitation—images of slash-and-burn in uplands are traumatising (see Fig. 1). As a result, it can seem obvious and natural to call on political powers to make the degrading practices stop (Aubertin 2001).
3Nonetheless, isn’t this a culturally conditioned position, rather than a rational judgement? Some authors have put forward some convincing arguments (Mellac et al. 1999; Menzies 2002). For lack of a study on and esteem for this farmer know-how, the public interventions for development have often led to results contrary to the political aims of environment preservation and poverty alleviation.
4We propose to contribute to the debate through a study on a shifting cultivation farming system in Phongsaly, which demonstrates the reasoned and complex management of human and natural resources by the farmers in order to mobilise them in an optimal and sustainable way. We then propose to evaluate the impact of locally implemented public programmes on farmer income.
Fig. 1: Slash-and-burn field Phongsaly
- 1 Phongsaly District Rural Development Project (PDDP), co-funded by the French Development Agency (AF (...)
5In late 2002, we began characterising the agricultural development in Phongsaly district to place the effects of a rural development project1 in a historical perspective. The methodology selected relies on the theory of differentiating agrarian systems (Dufumier 1995; Mazoyer et al. 1997). The study dealt with 40 rural villages in the south-western part of the district, where 1,850 families live (47% of the district’s population).
- 2 966 families, i.e. 52% of the study zone population and 25% of Phongsaly district’s rural populatio (...)
- 3 884 families, i.e. 48% of the study zone population and 23% of Phongsaly district’s rural populatio (...)
6In each of the 40 villages, interviews with elderly farmers made it possible to reconstruct the historical evolution of the village in the demographic, technical, economic and social domains, whereas farm surveys made it possible to characterise the current agricultural practices and the differences between villages and families. This initial phase brought to light a dichotomic zoning of the region: the situation of farm families and their practices are different depending on how far they are from Phongsaly, the region’s administrative and economic capital. In the first zone, villages along the roadside or in the immediate vicinity of the city do substantial commercial trade with city dwellers and benefit from sustained attention from administrative services; that is the case for 16 out of 40 villages.2 On the other hand, in the second forest zone—over a two-hour walk—trade with Phongsaly is progressively lower and public intervention is more unobtrusive for the 24 villages concerned3 (see Fig. 2).
Fig. 2: Map of Phongsaly district—region studied and socio-economic zoning
- 4 Samlang for the forest zone, Yapong for the roadside zone.
7To round out the study, we selected one village per zone4, in which an in-depth economic interview was conducted with each of the families. That census dealt with the family, farming practices and their results over the past five years, as well as with the family’s other economic activities: gathering, fishing, hunting, handicrafts, trade, etc.
8The terrain throughout the whole district is hilly and uneven, culminating at 1,948 meters, with some twenty peaks over 1,500 meters high. The valleys, under 500 meters in altitude, are very steep-sided; their V-shape limits the potential for irrigated paddy fields. The schist or sandstone substrates create fairly deep, acid clayey or silty-clay soils, which are rather fertile but very heterogeneous (Zhou et al. 1999).
- 5 Sources: Meteorology Department, Phongsaly Forest and Agriculture Service (2002).
9The region, like all of Lao PDR, is subject to a tropical climate marked by the monsoon system. Nevertheless, the altitude and latitude temper the tropical influences, with a cool dry season and a milder rainy season, during which on average 75% of the annual rainfall (1,560 mm) occurs. The very high inter-annual variability in rainfall (980-1,860 mm)5 strongly conditions how successful farming activities are.
10The zone’s climax ecosystem is either a tropical evergreen mountainous rain forest over 800 m a.s.l, or tropical rain forest below that. The very productive forests are characterised by their rarely equalled range of biodiversity, with over fifteen times more ligneous species than in a temperate forest, for example (De Koninck 1997), and still widely unknown endemic vegetal or animal species (Chazée 1990a/b).
11This chapter describes the production system of the forest zone villages (see Fig. 2). That system is a result of experience handed down from one generation of farmers to the next, and will serve as reference. The technical and economic findings are from families in the village of Samlang over the past three years.
12In a village, agricultural production is traditionally based on land use in three distinct zones: family gardens, in and around the village, the agro-forest crown, around and slightly above the village, the slash-and-burn zone, which constitutes the main part of the village land, with the planted fields and forest regrowth—or fallow land—(Laffort et al. 1998):
13• Village gardens: near the main dwelling, families cultivate a small vegetable garden, with tubers and fruit trees for household consumption. The village proper is also used for animal raising, with poultry that wander among the houses, looking for consumable waste and rice bran.
14• Agro-forest crown: the village, generally located near hilltops, is surrounded by an agro-forest crown, which is essential in its role as a water reservoir. By associating trees that remain from the primary forest and plantations, it provides part of the village’s timber and firewood, as well as fruit. Pigs are left there to roam freely and search for their food, supplemented by input gathered on the fallow land. The upper part of the crown is sacred, and any economic activity whatsoever is banned.
15• Slash-and-burn zone: swidden farming occupies most of the village land, with a small fraction cleared recently and planted—7% of the surface area—and the rest left fallow—93%— in landscapes ranging from grassland to secondary forest, and including all possible forms of shrubby thickets.
16In the south-western part of the district, agricultural production is a nuclear-family activity although clearing is regulated on the village level. Every year, the active workers slash a single strip in the village land. In that strip, each family farms its own plot, which it owns: the field is always planted by the same farmer and then by his heirs. Each family owns a plot in each strip of the annual clearing; that system of family ownership is recent, but is derived from historical clan ownership, with similar features and consequences.
- 6 Paddy terraces, cardamom plantations, cash crop gardens, orchards, etc.
17With demographic growth, there is a trend towards splitting up plots from one generation to the next. Regulating this trend of reducing the surface area per active worker is complex, based on four successive mechanisms: loan of land between families, possible lengthening of crop period from one to two years, departure of part of the population, acceleration of rotations as a last resort (Laffort et al. 1998). The inflexibility of this land system tends to slow down the decrease in the fallow period, a characteristic response to demographic growth in many other shifting cultivation systems (Foppes et al. 1994; Dufumier 1996). This management favours the maintenance of fertility and satisfactory production levels, at the cost of expelling a fraction of the village population, essentially the younger generation, towards other zones. The growth rate in the district was 1.9% a year between 1985 and 1995, compared to nearly 2.6% for the whole country (Sisouphanthong et al. 2000). This land tenure system, comparable to private property farmed by the owner, is unique for shifting forest agriculture. It confers high security to each family in access to land, in particular in the long term. Farmers can plan on investing in their fields6 so as to increase their productivity (Ducourtieux et al. 2004a).
- 7 Namely a yield in paddy rice of 1,310 kg/ha on average between 2000 and 2002 a(minimum of 450 kg/ha (...)
18Following the clearing and burning of one strip of forest, the plots are planted for one year, sometimes two. In the first year, glutinous rice dominates, associated with many crops (maize, tubers and roots, cucurbits, crucifers, peppers, sunflower and groundnut). In Samlang, all the work requires an average of 130 days of work per active worker, i.e., 320 days per family, for a production of 500-700 kg of paddy rice, 30 kg of maize, 130 kg of tubers and 420 kg of various vegetables per active worker.7 In the second possible crop year, rice is sown alone; the farmer simplifies his crop complex in order to preserve his priority crop: rice.
19The bottleneck in this crop system is the weeding, requiring substantial manpower (75 days/active worker/year). It must be done according to a specific, restrictive schedule, or else weeds will put a strain on the yields of rice and associated crops. For example, weeding done too late lets the weeds sprout and spread their seeds, complicating the management of grass cover during the subsequent weeding periods. June, July and August weeding monopolises the entire workforce.
- 8 7 to 15 years for villages in the forest zone studied (see Fig. 2).
20After the second crop year, the plot is freed for forest regrowth, with a 13-year long fallow period.8 Those long periods allow for the reconstitution of dense secondary formations, a biomass source of fertility for the next slash-and-burn crop cycle (Ramakrishnan 1992). The fallow land is the pasture area for large ruminants; cattle are limited to grazing grassy fallow whereas water buffaloes graze indifferently year round on shrubby, tree-covered and grassy fallow.
Fig. 3: Weeding of slash-and-burn field in Phongsaly
- 9 9 inhabitants/sq. km on average in Phongsaly district.
- 10 About 20-25% of the annual work of an active worker, all activities combined.
21Due to the low population density,9 the bottleneck of farm production is the workforce. There are more potentially farmable areas than actually farmed. Weeding is the major constraint limiting work productivity, representing 60% of the annual labour by an active worker in swidden cultivation.10 The maximum surface area farmable per active worker is approximately 0.8 ha; even if a family could clear more in January or sow more in April, it observes limits due to the labour overload with weeding in the middle of the rainy season, in July-August.
- 11 As it is impossible to add up the disparate quantities, it is possible to consider values: one hect (...)
22The limits on production due to fertility problems is harder to assess. The direct comparison of rice yields between lowlands—1.7 tons/ha in the Vientiane rainfed paddy fields (Sacklokham 2003)—and swidden fields—1.3 tons/ha in Samlang—is misleading, as it does not take into account the other crops associated with the swidden crop.11
- 12 After a three-year fallow period, the biomass is over 20 tons/ha, then 30 t/ha after 7 years, 70 t/ (...)
23Traditionally, fertility is presented as proportionate to the fallow period (Ramakrishnan 1992). The progressive build-up of biomass resulting from photosynthesis on the fallow land has been proven,12 but recent research work is challenging this oversimplified interpretation. The yields are not directly proportionate to the fallow period (Foppes et al. 1994; Roder et al. 1995; Van Keer 2003), although there is a certain rise with threshold effects: the average yield reaches 1,300 kg/ha of paddy in Samlang compared to only 600 kg/ha in Yapong, with 13 and 3 years of fallow respectively, whereas the two villages are nearby, in a similar natural environment (see Fig. 2). Furthermore, rapid rotations increase erosion, limiting future productive potential (De Rouw et al. 2002; Moa et al. 2002). The role played by the fallow duration in the yield is itself the interaction of a substantial number of cumulative and synergetic factors, for which it is hard to isolate individual contributions. In addition to the build-up of biomass for mineral fertility and the soil structure, there is pest control. The density of harmful insects and weeds in a slash-and-burn field decreases rapidly depending on how long the fallow period lasted before clearing (Van Keer 2003).
24The low correlation between yield and fallow duration can be explained by the complexity of the interactions between the various factors involved in the progressive elaboration of the yield in a crop cycle (Sébillotte 1993).
- 13 The combination of climatic factors and topography expresses the sensitivity to drought, the main c (...)
- 14 The main parasite identified by Van Keer (2003) is a root aphid (Tetraneura nigriabdominalis) and, (...)
- 15 Weeds are only a secondary limiting factor of the yield for Van Keer (2003), due to their control b (...)
25Recent research conducted by Van Keer (2003) for the agronomic diagnosis of farmer shifting cultivation systems in northern Thailand shows that soil fertility is not the factor limiting the way in which the genetic potential of local cultivars is expressed. The author established that the constraints concerning yield are, in order of importance, the number of successive crop years, climatic hazards, the topographical position of the plot,13 weeds and predators14 (Van Keer 2003). In Laos, farmers draw up a comparable list (Roder et al. 1997), ranking weeds,15 rodents and inadequate rainfall at the top of the list of constraints affecting their slash-and-burn crop. Samlang farmers rank drought—once every three years—as the main problem, followed by parasitizing of roots, and rodents.
26A crop is stopped after one or two years due to the conjuncture of fertility and weeding problems. The increased workload required, beyond a family’s potential, for countering the invasion by weeds and the drop in fertility contributes to reasons for abandoning a plot, without it being possible to systematically come to a conclusion about the preponderance of one factor in relation to another, given the current knowledge about the region’s agriculture (George et al. 2002).
Fig. 4: Secondary forest (15 year-old fallow)
27Like any agricultural activity, swidden farming in Phongsaly is not a practice that follows a set standard. On the contrary, each family is constantly adapting its actions based on the natural (climate) and socio-economic (manpower, tools, markets, consumer needs, etc.) environment. Each farmer elaborates a unique technical itinerary during the crop cycle, which differs from the previous year’s and from the other families’ (Sébillotte 1990).
28Shifting cultivation has evolved as a function of the historical context. Cotton and tobacco have practically vanished from the fields since the arrival on the local market of low-cost manufactured products from China at the end of the 1960s. Poppy has progressively disappeared under the administration’s pressure. On the other hand, some villages have developed maize or white rice farming as a raw material for the distillation and trade of alcoholic spirits in Phongsaly.
- 16 A sloped, rocky plot located on a mountaintop only offers a family low income prospects, whereas a (...)
29At the beginning of each year, farmers decide on the surface area to be cleared and sowed in arbitrating many constraints: the surface area available based on the land system in effect, the topography of the plot,16 the previous year’s harvest and perspectives of stock or shortages, the distance of the new plot from the village, its potential fertility—evaluated based on past crops during previous rotations and recent observations such as the soil texture or colour.
30For example, the remoteness of a field is not an important factor for clearing or sowing, but it represents a certain constraint for sliding firewood and bringing in the harvest. Furthermore, every hour spent walking during the rainy season is lost for weeding. Added to the marginal production gain that fallow periods over ten years procure (Van Keer 2003), the constraint of distance explains why villagers choose not to include in rotations the forest land within the village domain that is farthest away from the village. Out of the 24 forest villages in the study zone (see Fig. 2), 20 have de facto forest reserves, excluded from the rotation cycles.
- 17 Particularly important for limiting the rice shortage period the year after a poor harvest.
- 18 Sowing in hills with a hoe or sowing broadcast.
- 19 Basic food, pastry for festivals, alcoholic distillation, etc.
31After calculating the ideal dimensions, farmers elaborate the sowing plan, choosing the varieties and their distribution on the plot. Each of the 40 villages studied has a sample group of 4 to 20 varieties enabling precise adaptation of the crop to the farmer’s strategy; the varietal choice is based on the length of the cycle,17 risk limitation by multiplying the number of cultivars, adaptation to the soil and altitude, crop technique,18 use of the rice,19 etc. The biodiversity is also observed on a large scale, with nearly 550 sticky rice varieties identified for shifting cultivation in Northern Laos (Roder et al. 1996) and the Laos contribution making up nearly 50% of the germplasms in the IRRI’s world bank (Douangsavanh et al. 2002).
32On a plot, sowing is neither standardised nor random. The farmer decides how to do it based on his experience in using all the environment’s resources, on a very precise scale, per square meter. For example, the sowing density will depend on the slope, with tubers being preferentially planted in large heaps of ashes or maize in the wettest part. On a larger scale, this well thought-out choice concerning land use can be observed in the development of terraced paddy fields in the scarce irrigable zones or in the choice of shady, damp plots that are not too high in altitude for growing cardamom (Ducourtieux et al. 2004b).
- 20 This is no longer a crop for self-consumption; sesame is marketed for buying rice.
33After sowing, farmers manage the way in which family labour is to be used. When the village clears several zones, in particular if there are two consecutive years of crops, each family first allocates its workforce to the plot considered potentially the most fertile. It is usually the plot cleared that same year, but the choice is not systematic: seed quantities vary from one year and one plot to the next. Furthermore, the initial distribution of labour can evolve over the year, depending on the problems encountered. If a plot is substantially damaged—drought, rodents—the family will lower the amount of work there and transfer labour to other plots to limit the risks of a drop in production. There are other rescue strategies, such as sowing a plot again where growth is deficient due to lack of rain in April-May, or, when problems arise too late, sesame is sown as a main crop to replace rice.20
34Associating crops in a slash-and-burn field optimises income per surface unit while maximising photosynthesis but, above all, limits the risks for the farmer. Crop failure arising from a particular situation does not jeopardise the survival of the family, who can count on the farm’s other harvests and activities. The dynamic and evolving allocation of the workforce and the diversification of activities are the two phases of the farmer strategy used for limiting risks and maximising family income. Resources that are rare—workforce—or fragile—soil, forest, water, biodiversity—are managed differentially so as to be integrated in a sustainable way into the rational strategy of each category of farmers.
Fig. 5: Sowing rice in a slash-and-burn field in Phongsaly
35The average family income is over 1,250 EUR a year in Samlang, counting the self-consumed produce at market value (replacement value). The monetary income, with 170 EUR, is only 14% of the total income: the region’s farming system is focused on fulfilling the family’s direct needs.
36Families conduct many activities to reach that income. Swidden farming only ranks second, behind collecting—hunting, fishing, gathering—which procures over 40% of the family income in forest villages (see Fig. 6).
Fig. 6: Component of family income in Samlang
37The breakdown of income into multiple activities quantitatively expresses the farmer strategy of diversification. Furthermore, the wide range of products in a self-consumption economy contributes to the balance of family nutrition.
38Following this brief review, shifting cultivation no longer appears to be an archaic and rudimentary practice but a complex economic activity managed by farmers who adapt to changing conditions. They optimise the use of resources with practices that are based on neither chance nor inflexible norm, but on their know-how, deriving from the experience acquired from one generation to the next. This precise and detailed use of resources leads to a globally forested landscape, dotted with small areas of crops.
39The crop system described above is only one of the many ways the environment has been enhanced by farmers in Phongsaly district. In a 80x60-km area, no less than four agrarian systems coexist with just as many different slash-and-burn systems for less than 4,000 peasant families in 85 villages. The relatively homogenous natural setting cannot explain this wide range of practices. Cultural diversity might be a plausible theory, with six ethnolinguistic groups represented in the area: Phounoy (56% of the rural population), Akka (13%), Ho (13%), Laoseng (12%), Hmong (3%). In addition to these Tibeto-Burman populations, there are two Lu villages (Tai-Kaday, 3%).
- 21 40 villages, 48% of the rural population.
40Agriculture in the south-western part of the district21 (see Fig. 2), which we have just presented, is characterised by long-term fallow periods (8-18 years), the collective clearing of one old forest strip per village, although plots are inheritable family property, and crops associated with rice in the first year. This manner of exploiting the forest environment is not exclusive to the Phounoy ethnic group. One of the Akka villages in the zone, as well as Phounoy and Ho villages in other sectors (left bank of the Nam Ou river), are associated with the same system.
- 22 9 villages, 11% of the rural population.
41Most of the villages set up in the Nam Ou22 valley use farming systems very similar to those in the south-western part of the district, in terms of crops, agricultural techniques and land management. Yet they are Laoseng or Lu villages and not Phounoy (Alexandre et al. 1998).
- 23 3 villages, 2% of the rural population.
- 24 50% of the total yearly income on average (Alexandre et al. 1998).
42On the other hand, shifting cultivation in Ho villages on the left bank of the Nam Ou23 (see Fig. 2) is quite different, with long-term grassy fallow (10-20 years) maintained by fire to favour monospecific colonisation by Imperata cylindrica, plots are cleared by each individual family, separate from other families, plot allotment is limited to the crop period, a single crop of rice is planted, and harnessed buffalo traction used for ploughing the soil. Moreover, poppy farming plays an important role in the family’s livelihood,24 even though it is rapidly regressing under recent pressure from the administration. The differences between social classes are very pronounced, with an extraction of surplus value based on unequal exchange of rice and work. Although only Ho farmers practice that type of agriculture, not all the Ho villages on the left bank of the Nam Ou make use of the environment in that way. Some of them use a farming system similar to the Phounoy villages’ in the south-western part of the district (Alexandre et al. 1998).
- 25 8 villages, 16% of the rural population.
- 26 This makes the results of the shifting cultivation more risky: an early start of the rainy season p (...)
43Most of the Phounoy villages on the left bank of the Nam Ou25 use the agro-ecosystem in a completely different way from those in the south-western part of the district. Rice is still grown in association with other crops on slash-and-burn fields, but the fallow period is now much shorter (4-8 years), clearing is family-oriented and not grouped together on the village level, with plot allotment being limited to the length of the crop. Poppy farming plays an important role in family livelihood, although it is regressing; this entails a change in the work calendar, with clearing in March instead of December, the period for sowing poppy.26 Like in the neighbouring Ho villages, the differences in social classes are pronounced, with an extraction of surplus value based on unequal exchange of rice and work (Alexandre et al. 1998).
- 27 6 villages, 7% of the rural population.
44In the north of the district, the dominant agrarian system was historically clear-and-burn farming on fire-maintained Imperata savannah, comparable to the system in Ho villages on the left bank of the Nam Ou, which is still conducted in a few Ho and Akka27 villages. In the other northern villages, the system has changed (Baudran 2000).
- 28 3 villages, 4% of the rural population.
45Some villages have taken advantage of an ecosystem more favourable to paddy fields (not as steep on sandstone substrate) to set up rice terraces. In those Ho villages in the north-western part of the district,28 farming on burnt fields has practically disappeared. Paddy farming, requiring less labour during the period when poppy is sown, has allowed families to extend the narcotic crop (Baudran 2000). Contrary to generally accepted ideas (UNDCP 1999), shifting cultivation limits opium production more than paddy farming, due to the conflicting schedule for the workforce at the time of clearing which corresponds to the poppy-sowing period.
- 29 16 villages, 12% of the rural population.
46In the other villages, whether Ho or Akka,29 the farming system is evolving towards shifting cultivation comparable to the system used in the Phounoy villages on the left bank of the Nam Ou, but with much longer fallow periods (10-20 years) due to the size of the village domains and the low population density (5-7 inhabitants/sq. km). In nearly half the villages, families do either shifting cultivation on Imperata grassy fallow, if they have buffalo for traction, or slash-and-burn on ligneous fallow when production means are strictly manual (Baudran 2000).
47The ethnic origin of the villages influences the typology of the farms and villages. It is possible to compare the farming and land practices of the Ho tribes (in the north or on the left bank of the Nam Ou) and the Phounoy tribes in the south-west. Nevertheless, one Akka village, one Ho village and the Laoseng villages have farming systems comparable to the Phounoy’s in the south-west, whereas the Phounoy villages on the left bank farm the land in a very different way, comparable to the system in Ho and Akka villages in the north of the district; the latter practice various systems, irrespective of their ethnic classification. The diversity observed in the farming systems in the Phongsaly District does not match the ethnolinguistic zoning: therefore, the ethnic group is not a decisive criteria of farming diversity typology in the region.
48The distance between the village and town of Phongsaly as well as the distance to a major communication route—road, track or navigable river—are more obvious sources of differentiation of farming practices (see Fig. 2). Differences in market access depending on those distances induce substantial differences among the villages, as well as a more pronounced social differentiation in the villages nearby (Laffort 1997). Location is a key criteria in typology, but is not sufficient to explain the differences observed among the remote villages.
49The differences observed in farming techniques can be explained by the historical and social construction of the villages, unique to each zone. For example, on the left bank of the Nam Ou, the specificities of farming on Imperata fallow in Ho villages can be understood through the historical migration to zones in altitude, by that population with a previously constituted technical corpus in the rice farming plains of southern China (bubalin traction and associated farm tools). With demographic growth, the Phounoy villages in the south-west split up to found new settlements on the left bank, but the village domains were bordered by Ho village domains already set up earlier. Lacking the necessary land resources, Phounoy migrants were unable to reproduce the original land system, prevalent on the right bank. They initially adopted the farming mode used in their Ho neighbours’ environment before changing over to the current techniques due to a lack of resources in land and traction animals (Alexandre et al. 1998).
50Current differentiation of farming systems can be explained by the simultaneous and joint, historical construction of environmental enhancement modes as well as internal and external social relationships. In the district, this differentiation does not systematically match up with ethnic zoning.
- 30 In Lao PDR, the unique party is constitutionally the political power’s superior body.
- 31 Reduce poverty by half by 2005 and eradicate it by 2010.
51In 1986, the Lao Revolutionary Party (LRP)30 committed the country to a “socialist market economy” via the policy of New Economic Mechanisms. Private ownership of production means and free enterprise became new principles for prosperous development. In 1996, the Sixth LRP Congress set the government aim of removing Laos from the list of “least advanced” countries by 2020. In 2001, the seventh Congress reinforced that position with quantified objectives31 and based the policy for eradicating poverty on three pillars: economic growth, socio-cultural development and environmental protection (Lao PDR 2003).
- 32 Prime Minister’s Decree 010/PM defines poverty as “the lack of ability to fulfil basic human needs (...)
52The government has the responsibility for reaching its goals by implementing the national programme for the eradication of poverty (NPEP), supported by the decentralisation policy. Rural development plays a central role in that policy. The NPEP promotes development based on community demand and on improving access to the poorest landlocked districts as a priority32 (Lao PDR 2003), whereas the Ministry of Agriculture and Forests is relying on a complementary programme in which the country’s development includes modern, permanent and intensive agriculture. It must substantially generate raw materials favouring both the supply of the domestic market, growth of exports, and the emergence of a national agro-industrial fabric (MAF 1999).
53Following this logic of intensification, all the country’s regions do not have the same potential. The governmental policy differentiates between the productive lowlands, vectors of the country’s economic development, and the uplands where environmental protection must prevail (MAF 1999). Although the economic role of the rice-growing plains in the Mekong Valley is undeniable, limiting the problematic upland issue to the single dimension of protecting nature is simplistic: slopes cover 80% of the country’s surface area and 250,000 families—nearly a third of the country’s total population—live there (MAF 1999). These rough figures convey the social and economic importance of upland agriculture in Lao PDR.
54During the seventh Congress in 2001, the LRP took up a twofold fight against poverty: elimination of opium production by 2005 and the progressive phasing-out of shifting cultivation by 2010 (Lao PDR 2003).
55These strategic measures confirmed earlier positions, in which slash-and-burn agriculture was presented as outmoded and destructive by the colonial administration (Mellac et al. 1999), then by development institutions (UNDP 1995). In 1994, the government had decided to eliminate shifting cultivation by the year 2000 (Keonuchan 2000); in 2000, the objective was postponed until 2020 (MAF 2000), before being brought back down to 2010 (Lao PDR 2003). This policy of banning slash-and-burn agriculture falls into a historical, as well as regional, logic: it can be observed in Thailand, Vietnam, Malaysia, China and Indonesia (Durand 1997; De Koninck 1998; Rossi 1998; Zaifu 1998; Mellac et al. 1999).
56The goal of eliminating shifting cultivation is motivated by the reasoning that it is one of the main causes of rural poverty. Demographic growth in swidden agricultural regions tends to accelerate rotation and reduce forest areas, which leads to the reduction of income for the families involved—who get poorer—while burdening the country’s future development with the destruction of natural resources. Furthermore, the poverty of families who practice shifting cultivation drives them to grow opium, a source of addiction and therefore increased poverty (UNDCP 1999; Lao PDR 2003). The vicious circle is complete and poverty is self-maintained (Dasgupta et al. 2003).
57Based on this observation, the solution seems obvious: converting farmers practising slash-and-burn to permanent crops or non-agricultural activities would make it possible to interrupt the process and therefore eliminate poverty (UNDP 2002). Is it that easy?
58In Phongsaly, converting from shifting cultivation has been on the administration’s agenda since the end of the 1960s. In 1968-1969, over 400 families were displaced to the Boun Neua and Bountay bottom land during the “paddy rice field movement.”
59More recently, local authorities have implemented three programmes to apply the national policy:
60• Resettlement of forest mountain zone villages to the roadside;
61• Mandatory cash crops;
62• Land allocation.
- 33 40% of the population from the south-western part of the district.
63In Phongsaly District, eight villages have moved following administrative instruction since 1987. Five of them were settled along the road going from Phongsaly to Boun Neua, on an abrupt ridge. Seven other villages have been eliminated by the authorities since 1990; the families concerned joined neighbouring villages, along the roadside, or emigrated to towns (Phongsaly, Oudomsay, Luang Namtha, Vientiane). All in all, 700 families have been displaced.33
- 34 Due to a contractual ambiguity concerning the responsibility for transportation costs between Phong (...)
64At the same time as this restructuring of space, the district services introduced cash crops to replace swidden agriculture. Between 1996 and 1998, the first experience with sugar cane concerned four villages, along the roadside, and ended in failure34 for the 275 families who were obliged to farm a minimal surface area per active worker (Ducourtieux 2000).
65Since then, the tea crop has taken over. The programme plans for the town of Phongsaly and 14 rural villages to plant 500 ha between now and 2005, at a minimum mandatory rate of 0.3 ha per active worker. In those villages, clearing is to be banned as from 2004 or 2005. The mandatory tea crop is rounded out by fruit trees in three another villages and by the galanga (Zingiberaceae) in 17 villages. In the 40 villages studied, 45% of the families are involved in the “tea” programme, 13% in the “galanga” and 5% in “fruit trees”.
- 35 19 villages out of the 40 in the study zone.
- 36 7 years on average after land allocation (0-13 years), compared to 14 years beforehand (9-23 years)
66Since 1998, the local administrative services have been carrying out the land allocation programme, the central component of the national land reform (Ducourtieux et al. 2004a). At the end of 2003, 22 rural villages have a new land use map in the district,35 resulting in the forest protection of 47% of their village domain and a decrease in the length of the fallow period by half.36
67For lack of human resources, the Phongsaly District agriculture office is taking care of land allocation and promoting cash crops only in easy-to-reach villages, along roads and tracks or on the Nam Ou river banks (see Fig. 2). Villages in the forest zone are, for the time being, fairly unconcerned by these programmes. For lack of financial means, support to farmers for implementing the programmes is limited to planning and very basic technical training. Farmers are contracting debts with the public Agricultural Promotion Bank or private merchants to buy the mandatory crop seedlings that they have to plant.
68To assess the recent changes in Phongsaly District agriculture, we compared the technical and economic performance over the past three years in two drastically different villages in the study zone.
69On the one hand, we surveyed 28 families from the village of Samlang, an ancient archetypal Phounoy village in the forest zone with swidden cultivation, fairly unaffected by recent reforms, and on the other hand, 48 families from Yapong, a Phounoy village six kilometres away from Phongsaly on the roadside.
70Like all the villages along that road, successively the Yapong families have been resettled (1996), grown sugar cane (1997-1998), participated in village land allocation (1999) and, since the year 2000, have been developing some tea plantations. Clearing will be banned there in 2005.
- 37 Swidden cultivation, rice field, tea, garden, cardamom for vegetal crops; water buffalo, cattle, pi (...)
71For the two villages, the data collected during a two-hour interview with each of the families allowed us to model the various income components37 and relate them to the work supplied and the family structure (number of members and number of active workers).
72Land allocation has had a direct impact on swidden cultivation in Yapong. The forest reserves are taken out of rotation, then the surface area of fallow land available for swidden cultivation regressed. The age of the fallow upon slashing dropped from 10 to 3 years.
73Yield is limited to 600 kg/ha of paddy rice on the plot cleared in the year, compared to 1,300 kg/ha for the village of Samlang, in a forest zone, i.e., a 54% reduction.
74In an attempt to maintain rice production, Yapong families developed a strategy of increasing surface areas within the limits of the land allocation, with two to three successive years of cultivation, compared to one year in Samlang.
Fig. 7: Comparison of slash-and-burn field performances
- 38 First-year slash-and-burn field, 2000-2003 average for all the village families.
kg paddy rice /ha
Surface per active worker
kg/ton of rice produced
Rice Deficit (gap)
75To extend their farmed surface area, families have to face the crucial problem of weed control. Due to a lack of resources, farmers can not devote more time to weeding: 78 days/year/active worker in Yapong compared to 75 days/year/active worker in Samlang. The Yapong families compensate the saturation of the available workforce by a new and massive use of herbicides: the consumption of weed-killer per ton of rice produced is 20 times higher in Yapong than in Samlang. The product, of Chinese origin, is poorly identified and used, which does not fail to pose public health and environmental problems.
76With rising work, a falling yield and production costs that are increasing, work productivity for roadside village farmers has drastically dropped. Although in Samlang a workday brings in the equivalent of 1.1 EUR, it is half as low in Yapong with 0.4 kips per workday (see Fig. 10).
77Furthermore, the family workforce is a limited resource that is not extensible. Production per family is dropping, which increases problems of shortages. Still rare in Samlang—0.5 months of shortage/family/year on average over the past three years, 20% of the families concerned—the rice shortage is becoming the norm in Yapong: 3 months of shortage on average per family, 60% of families concerned every year. When in fact, the National Poverty Assessment (NPA) showed that poverty is closely correlated to food availability in terms of rice (SPC 2000; ADB 2001; UNDP 2002; Lao PDR 2003).
78The drastic drop in technical and economic performance of swidden cultivation with land allocation is not intrinsically a problem. It could even be an aim sought to urge farmers to convert to alternative crops that provide income to farmers in the region.
79Unfortunately, the performance of other economic activities does not meet the needs of the families either. All the components of family income in Yapong are on average lower than those in Samlang (see Fig. 8).
Fig. 8 : Comparison of income components
80Income from livestock is decreasing substantially due to decapitalisation (- 72%): to buy rice, families sell their animals, including reproductive females. They can no longer capitalise in livestock. There were 110 head of bovids in Yapong in 1996, and there are only 85 left in 2003. Animal raising brings in less than 80 EUR per year to 73% of the Yapong families, whereas 80% of the Samlang families have a livestock income of over 80 EUR per year (see Fig. 9).
Fig. 9: Comparison distribution of family income from livestock
81Tea, the cash crops imposed as an alternative to slash-and-burn, is characterised by low income—40 EUR/family/year—and substantial work—70 days/active worker/year. As labour is the limiting factor for agriculture in the region, it is in the farmers’ interest to optimise the use of the family labour; they give preference to production systems with high work productivity. When in fact tea farming offers the lowest level of all the farm activities (see Fig. 10). That speculation cannot be a credible alternative to swidden cultivation for reaching the political aim of poverty eradication.
Fig. 10: Comparison of work productivity for different rural activities
82Introducing cash crops such as tea is based on the presupposition that it is possible for farmers to sell cash crops and that the monetary income procured will make it possible to buy rice instead of producing it on the family’s slash-and-burn fields (Ducourtieux et al. 2004a).
- 39 That is to say that a third of the average yearly shortage in rice for a family in Yapong (see Fig. (...)
83The family monetary income is on average 180 EUR in Yapong. It is therefore slightly over Samlang’s (+ 7%), but that difference only enables a family to buy 100 kg of rice,39 i.e., less than 10% of its annual needs (1,200—1,500 kg/year/family).
84The average annual total income for a Yapong family is limited to 520 EUR compared to 1,240 EUR in Samlang, i.e., a 54% difference. Taking into account the sizeable family difference between the two villages, the difference speaks for itself: 130 EUR per person in Yapong compared to 250 EUR in Samlang. The average income is halved in roadside villages in comparison to forest villages.
- 40 Trade and transportation of firewood and banana trunks between the village and Phongsaly with moto- (...)
85This drop in income does not affect all the families. Although a large fraction of the population is impoverished (see Fig. 11), a few families have taken advantage of the resettlement towards the road and Phongsaly to get involved in transportation and trade.40 At the time of the move, those families had slightly more capital than the others—in particular, more animals—which enabled them to limit the effect of decapitalisation for buying rice, as well as invest in profitable services.
Fig. 11: Comparative distribution of total income
86The comparison of the economic findings in the villages of Samlang and Yapong is striking. However, are these two villages representative of their respective zones, namely forest villages practising traditional shifting cultivation on the one hand, and on the other, easy-to-reach villages concerned by the local administration’s development programmes?
- 41 Resettlement, land allocation, mandatory cash crops, ban on shifting cultivation.
87The first survey work in 40 villages for six months, with interviews of over 200 families, made it possible to establish zoning and make the reasoned choice of Samlang as an archetypal forest village. The size of the village, the natural environment, the length of the rotation, the farming techniques and results are within the average of villages in the forest zone. For the easy-to-reach zone, we selected Yapong, which participates in all the activities developed by the local services41 and is therefore, for that reason, a characteristic example of the easy-to-reach zone. The size of the village, the natural environment, the length of the rotation, the farming techniques and results are within the average of villages in the easy-to-reach zone. Before being displaced in 1996, Yapong families farmed a natural environment similar to Samlang’s; the fallow period and the village size were comparable. Farmer income was similar. The spatial comparison between Samlang—a forest zone village—and Yapong—village in the accessible zone—is also dynamic: Samlang’s current economic situation shapes Yapong’s, although the latter was involved in state development programmes.
- 42 Opening of the Phongsaly —Oudomsay road in 1996, opening of the Phongsaly— Vientiane air link in 20 (...)
88Although the findings of the comparison are clear, there are many and cumulative causes. The effect of some factors is easy to identify and quantify, like land allocation for swidden cultivation. However, for many farmer activities, there may be multiple and combined causes for the differences. The villages are nearby and in a comparable natural environment: the soil and climatic effects do not explain the differences. But, the farming systems in the forest zone are not permanently set and the easy-to-reach zone ones are not subject only to the administrative pressure. The relative and progressive opening up of the landlocked region,42 the progressive development of cross-border exchanges with China and Vietnam, as well as co-operation projects in the region also contribute to the differentiated evolution of farming systems.
89Nevertheless, despite the wide range of differentiation factors, the size of the differences leads to the conclusion that the Phongsaly service programmes are counterproductive. Instead of contributing to eradicating poverty, they drastically increase it. The aim set during the Seventh Lao Revolutionary Party Congress to cut poverty in half by 2005 (Lao PDR 2003) will not be reached in Phongsaly; it is more likely to have been doubled.
90The impoverishment of the farmer community observed in Phongsaly is not an isolated phenomenon. Studies in other mountainous regions in northern Laos obtained similar findings in Luang Prabang, Luang Namtha, Oudomsay (Keonuchan 2000) or Houaphan (Aubertin 2003).
- 43 NPA: National Poverty Assessment, 2000.
91The NPA43 conducted in 2000 by the Planning and Co-operation Committee shows that poverty in rural mountain zones is a contemporary phenomenon, widely caused by development programmes, with land allocation ranking first (SPC 2000). These conclusions are reported in recent official publications by the government (Lao PDR 2003) and international development agencies (ADB 2001; UNDP 2002). Many development programmes conducted in the field are counterproductive in relation to the main political goal, which is poverty eradication.
- 44 Lao Expenditure and Consumption Survey III (LECS III), forthcoming.
92The official statistics highlight this problem, without necessarily identifying and analysing it. As a result, the preliminary findings of the latest household consumer survey44 announced an overall reduction in poverty in Laos—in 2003, 30% of the population lives under the poverty threshold compared to 39% in 1998 and 45% in 1993—, but mentioned that the increase in wealth is unequally distributed between the uplands and lowlands, between the rural and urban zones, and between the different groups of the population; in some northern provinces, poverty has increased (Lao PDR 2003).
93To reach the goals set by the political authorities—cut poverty in half by 2005 and eradicate it by 2010—the reform of development programmes in upland regions is of the utmost urgency. That reform must be founded on the principle that the farmers are the solution and not the cause of the poverty problems; on that account, farmers should be involved in choices concerning orientations to be taken and in the definition of development actions, so that the programmes take into account their elaborate environmental management, resulting from know-how acquired generation after generation. That experience has enabled them to use natural resources sustainably, whereas underestimating that know-how generally has effects just the opposite from the goals set.
- 45 Prime Minister’s Decree PM/01 (11/03/2000).
94In order to be effective and pertinent, state interventions must be rethought and freed from the oversimplified image that the culturally and politically dominant lowland inhabitants have of the ethnic minorities in the uplands. Decentralisation45 entrusts the provinces and districts with new responsibilities. To reach the national policy goals, those services must be capable of defining the development programmes with the farmers. They are called on to rapidly become an active interface of adaptation in state intervention with local conditions.