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

Review of policies related to management
of upper watershed catchments

Regional Report

By

Esa Puustjarvi

CONTENTS

ABBREVIATIONS

1. Introduction

The countries of Greater Mekong Subregion lend themselves well to a comparative analysis on the management of upper watershed catchments. Similarities in geography, climate, population and culture in the catchment areas enhance the probability that solutions adopted in one country may have relevance in others. On the other hand, the development strategies and approaches adopted by the GMS Governments differ to an extent that makes comparisons meaningful. Even though it is clear that there can be no blueprint solutions that are directly transferable from one country to another, a comparative approach may be able to indicate the risks and opportunities that lie ahead, if a particular strategy is adopted.

The project at hand is focused on remote watersheds in the GMS region. However, policies and institutions specific to remote watersheds are few, and it became necessary to adopt a broader analytical scope. The actual policy and institutional environment is influenced not only by specific policies designed for the remote watersheds, but also by a number of other policies on higher levels and in related sectors. Therefore, while references to remote watersheds are made whenever relevant, the policies and institutions impacting on "uplands" are used as the principal framework for analysis.

The analysis concentrates primarily on natural resource management and its links with poverty and environment. It largely excludes issues such as health, education, gender and ethnic minorities, as they are covered by other authors working under this project.

2. Assessment of Poverty

2.1 Definitions

Defining poverty requires a way of distinguishing the poor from the non-poor. The usual method is to classify an individual as poor if he or she does not meet a set of consumption norms. Norms based on consumption are closely related to income-based measures of poverty. The bundle of foodstuffs and other items defined as the minimum set is simply valued at market prices. Headcount index, which denotes the proportion of a population with income or consumption below a specific poverty line, is one of the most common indicators in this category. The drawback of these methods is that only part of the food consumed in subsistence economy is valued on the market. Items such as insects, wild plants and animals etc. are often not traded, and their value is difficult to establish.

However, even if it were possible to derive market values, it is not simple either conceptually or empirically to define a set of such norms. Even if there were agreement that the minimum amount of food needed in order to sustain basic bodily functions of an individual of a given age, sex, height, weight and activity is not a constant. Partly for such reasons many have argued that there are no universal poverty norms but only ones that are space, time and society-specific. Further, it has been pointed out that such views are narrow ones, as they do not take into consideration other factors, which may be needed for a "good" living. For instance, being educated and knowledgeable, leading a long and healthy life or having a voice in setting the political and development agenda may rank high in people’s values, even if they do not carry a price tag. In other words, it is not the income or assets as such that matter, but the improvements in one’s life that should follow from it. Being constrained in enjoying these aspects of life is often referred to as "human poverty".

It should also be noted that general indicators tend to assess poverty "from outside" assuming that the aspects they measure are relevant to all those, whose poverty is being assessed. However, it is well known that the perception of poverty is rather subjective. While it may be rather safely assumed that attaining a minimum level of nutrition is a universal indicator, other commonly used indices may be less universal, or they may even fail to capture some relevant elements of poverty. For instance, receiving formal education may not be considered an objective worth pursuing, whereas intangible issues such as self-sufficiency and independence, which are difficult to estimate and measure, may can be highly valued.

It may be argued that, ultimately, the welfare attained by an individual as judged by the individual should be adopted as the basis for deciding whether he or she is poor. This perspective has been raised in the debate regarding watersheds. It has been questioned whether the kind of development model proposed by Government policies is indeed what the people in living in the watershed desire. It has been suggested that the largely materialistic objectives underlying most Government-led development efforts are not relevant to people, whose perception of development and cultural values are very different from those who design the policies.

Even though the discrepancy between this approach and the traditional methods of poverty assessment should not be exaggerated, the viewpoint remains relevant, and it should be taken into account both in policy formulation as well as in practical project implementation. One of the key measures is to apply a participatory planning approach, which allows people to set their own objectives. However, the concept is difficult to operationalize in a manner that would enable comparisons. Therefore, the following account on incidence of poverty in GSM countries will rely on the available indicators, mainly headcount index and some indicators of "human poverty".

2.2 General indicators

The early years of 1990’s marked a period of rapid economic expansion in the GMS region. The highest growth rates was recorded in Yunnan, where the economy grew on average 10 % per annum. Apart from Thailand the GDP per capita levels remain rather modest ranging from USD 300 in Cambodia to USD 500 in Yunnan. Thailand has enjoyed a period of long sustained growth much longer than the other countries, and the GDP per capita figure reached USD 2800 in 1997. However, it is worth noting that if measured by Purchasing Power Parity, the differences between the countries are substantially reduced (Table 1).

Along with the expansion of the economy there has been a shift away from agriculture towards industries and service sectors. The proportion of agriculture of the GDP is lowest in Thailand standing at 11.2 %. It is perhaps noteworthy that the pattern does not hold in Vietnam, where the change in production structure has not brought about similar economic growth as in other countries.

The economic crisis that set in Asia in the latter part of the 1990’s slowed down the economic expansion, and even brought about a contraction of the Thai economy in 1998. However, in countries such as Lao PDR and Cambodia, where a large part of the economy is subsistence-based, the financial crisis had only a limited effect.

Table 1: Selected macroeconomic indicators for GMS countries

Source: Asian Development Bank 1998
World Bank 1999
UNDP 1998
Yunnan Yearbook of Statistics 1998

In recent decades the GMS countries have achieved significant strides in their pursuit to reduce poverty. Measured by national poverty lines, the most notable achievements have been posted by Thailand and China PDR. In Thailand the poverty rate dropped from 57 % to 13 % in 30 years. In Yunnan, the proportion of poor has been squeezed down to 8.3 %, which is the lowest rate in the GMS region (Table 2).

Poverty is still widespread in Vietnam and Lao PDR, where, respectively, 51 % and 46 % of the population fall under national poverty lines. However, preliminary results from a survey conducted in Vietnam indicate that poverty has dropped to 30 %. As a rule, incidence of poverty is higher in rural areas than in urban settings. While it is true that even the poor have benefited from the rapid economic development, it has also been accompanied by a widening income gap between the rich and the poor. The income ratio between the top 20 % of people to the lowest 20 % is highest in Thailand and China PDR. In Lao PDR the difference between the rich and the poor is still modest, but the trend of growing disparities in income is can be detected in the whole region.

Table 2: Poverty indicators in GMS countries

Source: Asian Development Bank 1998
Yunnan Office for Poverty Alleviation 1999

Regarding indicators for human poverty, those related to education have improved throughout the GMS region (Table 3). School enrolment has attained a high level in all countries. However, this has not always translated into gains in skill and capacity. For instance, in Lao PDR a high enrolment rate is coupled with a rather low rate of adult literacy (cf. ch. 2.3). The low literacy rate in Yunnan province shows its disadvantaged position within China PDR. It is noteworthy that despite a modest level of economic development Vietnam has established a high standard of education.

Table 3: Indicators for education in GMS countries

Source: UNDP 1998

Life expectancy has steadily increased in all GMS countries. Regional differences are reflected in the rate of infant mortality, which has dropped dramatically in China PDR (incl. Yunnan), Thailand and Vietnam, but remain at a high level in other countries (Table 4).

Table 4: Indicators for health in GMS countries

Source: UNDP 1998

The human development index compiled annually by the UNDP combines three indicators: life expectancy, education and purchasing power parity. The index was developed in order to temper the income-based measures of development with non-monetary aspects of human attainment.

Among the world’s 174 nations the GMS countries rank in the middle group (Table 5). As expected, Thailand comes on top surpassing the average index values for the world and qualifying as a country of "high human development". China PDR has pushed herself beyond the average for developing countries. The other GMS countries find themselves below the average for all developing countries, but above the average for least developed countries. Vietnam ranks high as it is able to compensate the high incidence of poverty with her remarkable achievements in education and health.

Table 5: Human development index values for GMS countries

2.3. Poverty in uplands and remote watersheds

The difference in poverty rates between rural and urban areas indicates that the fruits of past economic development have not been distributed evenly within the GMS countries. While the rural areas generally lag behind, there are also notable differences between rural regions.

Country studies point to two principal trends. First, even if the upland areas (larger regions) are usually among the poor regions of the country in terms of income, they are not necessarily the poorest ones (e.g. in Thailand, Vietnam, Yunnan). One of the most illustrative example is the situation in the Mekong-Lancang River Basin in Yunnan. The remoteness and ruggedness of terrain are not the sole factors explaining lower socio-economic status of the area. In terms of GDP per capita the areas with more than half of their land surface under steep slopes average USD 295. This is below the average levels at the provincial level (USD 503) and in the river basin (USD 333), but it is much higher than the level reached in the 10 poorest counties in the river basin (USD 195) (Table 6). There is no obvious explanation for this difference, as the indices on population density and road infrastructure do not markedly differ. However, it seems that poverty is associated with a prevalence of ethnic minorities in the area. The data shows a rather consistent pattern, where a high proportion of ethnic minorities is correlated with a low level of GDP per capita.

A second, clearly discernible pattern is that certain "pockets of poverty" seem to persist irrespective of the level of wealth. Typically, it is the resource-poor, isolated and remote watersheds, where poverty is most prevalent. These pockets tend to persist, even when the overall economic development is rapid. In Yunnan province the Government has designated poverty villages, which are entitled to special support. While the incidence of such villages declines along with higher GDP per capita levels, they do exist even in the wealthiest area.

Table 6: Selected socio-economic indicators in the Yunnan province and Lancang river basin

Source: calculated from data available from Yunnan Year Book of Statistics, 1998
Yunnan Year Book, 1998
Yunnan Institute of Environmental Science, 1998
Yunnan Department of Transportation, 1996

Regarding human poverty, e.g. access to education and health, the GMS countries display rather different situations. In Lao PDR and probably in Cambodia, the upland areas lag behind the rest of the country, whereas the data available from Vietnam suggest that the disparity is minor. Interestingly, relatively high access to education and health care does not seem to automatically result in reduction of income poverty. The case from the northern uplands of Vietnam is illustrative of the situation. In this region there are five provinces, where more than half of the communities are disadvantaged to an extent that they are entitled to special support from the Government. Population density in these provinces is below the regional average, and the level of social support does not markedly differ from the national average. The pupil/teacher ratio in primary education is better than in the region or in the country, whereas doctors are fewer. However, in 1996, the average GDP in these "distress provinces" was USD 107 per capita as opposed to USD 138 in the whole region (Table 7).

One of the critical factors that may impede a positive impact on poverty is the efficiency in using available funds. The conclusion made in Lao PDR is likely to have relevance in Vietnam as well. In Lao PDR access to schooling is adequate even in remote areas, and the availability of teaching staff is reportedly satisfactory. Even so, the results in terms of education are not impressive, as the problem lies in the low quality of schooling – lack of textbooks and qualified teachers.

Table 7: Selected indicators on "Distress Provinces" in Vietnam

Source: calculated from data available from General Statistical Office 1998

3. Environmental development in upland areas

Degradation of the natural resource base in the catchment areas is an environmental issue common to all GMS countries. While the term "environmental degradation" is used rather loosely in the public debate, it is usually understood to mean the replacement of natural, rather undisturbed forest with logged-over forests, agricultural land, shrub or bare land. It may also refer to a reduced productivity of agricultural land.

The environmental effects, which appear to be of highest concern, are in fact those which have the most serious economic consequences. Change in vegetative cover contributes to washing off of top soil, which reduces soil productivity and increases the sediment loads in downstream water bodies. Loss of top soil negatively affects upland agriculture, and increased sediment loads incur costs for downstream agriculture, especially for irrigation schemes, or hydropower production. While these issues appear to be the primary concerns at national level, in recent times the GMS Governments have paid increased attention to other effects such as reduction of biodiversity, depletion of genetic resources and decreased carbon sequestration.

3.1 Impact of land uses

3.1.1 Erosion and soil degradation

Public discussion on environmental degradation in the uplands tends to focus on deforestation and arresting it is a primary objective of state interventions throughout the GMS region. The interest in deforestation stems from the fact that several different types of forest cover, be it natural forest, degraded forest, reforested area etc., are able to maintain watershed functions at least in an adequate manner (Table 8). Even areas in poor condition (e.g. logged-over forests) retain the watershed functions quite well, unless they are very heavily logged or subsequently cleared for agriculture. New plantations have high erosion rates, but as they develop into conservation forests or multipurpose forests, they provide almost as much protection as natural forest.

Table 8: Potential erosion rates for selected land uses in Thailand

Source: Thai Master Plan 1994

As clearing of forest for agriculture is one of the main causes of deforestation, the sustainability of ensuing agricultural practices has been a recurrent topic in the discussion on watershed management. The most contentious issue has been the sustainability of shifting cultivation. Government policies often regard it as a major cause of forest loss and environmental degradation in the watershed areas.

However, it appears that there is a general shift in attitude and a slowly evolving consensus on a few key principles. The need for more varied policies to deal with swidden agriculture is increasingly recognised (e.g. Parisak 1998). Even if it is best suited for areas, where land resource is abundant, it is generally acknowledged that shifting cultivators are often able to overcome the land constraint by modifying their practices. Various combinations of cropping arrangements, agricultural practices, lengths of fallow, and agroforestry and annual species are applied indicating that further development of swidden methods may be one avenue for increasing agricultural productivity.

However, it is also recognised that shifting cultivation is by nature an extensive land use and the margin for adjustment has its limits. With increasing land scarcity farmers are compelled to reduce the length of fallow periods below sustainable levels. Originally, shifting cultivation was based on rotations as long as 10 to 15 years, but in many areas the fallow period has been reduced to 3-4 years. The estimates on population density beyond which the sustainability of shifting cultivation is jeopardised vary as many other factors influence the carrying capacity. However, most experts set the limit roughly between 20-40 persons/sq km. Unfortunately, in many upland areas population levels have surpassed this level. The principal exceptions are the uplands of Cambodia, Lao PDR, and Myanmar (Table 9).

Table 9: Population densities in upland areas of GMS countries

Source: See Country Reports

Regarding permanent agriculture, the impact on watershed protection varies considerably depending on the type of production. Tree crops with appropriate erosion controls may provide almost as effective protection against erosion as natural forest. Examples include coffee growing in hill evergreen forest, rubber plantations on terraces and rubber plantations or other tree crops with understorey vegetation. However, it should be noted that protection if effective only if proper precautions are taken. A poorly maintained mechanical erosion control structure usually results in greater soil erosion than there would have been without it. In addition, they are often very costly both to set up and maintain, and it is highly uncertain that farmers will adopt them unless sufficient incentives are provided. From farmers’ standpoint erosion is seldom a primary concern compared to other more immediate problems such as need for weeding.

The highest erosion rates associated with agriculture are estimated for cultivation of field crops and sloping field crops. In worst cases the soil erosion rate may be more than a 100 times higher than that of natural forest. However, the extent of erosion varies depending on the ground cover provided by the crop and the conservation measures implemented (Table 8).

Completely denuded areas, which usually result from soil exhaustion after inappropriate logging and shifting cultivation, are the most serious case of watershed degradation. Erosion rates are double the highest rates estimated for type of agricultural production. However, it is noteworthy that with some vegetation regrowth the watershed functions can be restored to an adequate level. Rangeland or shrub areas may be comparable to best agricultural practices (Table 8).

3.1.2. Biodiversity and carbon sequestration

Fernandez (1998) has studied the impact of changing land use on biodiversity and carbon sequestration through a case study in Indonesia. Even though the context is not fully comparable to those found in the GMS region, the conclusions probably hold in broad terms.

The study indicates that indicates that the level of biodiversity is only modestly affected by community-based forest management or logging provided that they are carried out in a sustainable manner (Table 10). Unfortunately, in reality unsustainable practices are commonplace in the GMS region, and their impact on biodiversity is often very negative. Reforested areas were not part of the comparison, but they would probably score rather low, even if monocultures were avoided. In general man-made forests are considered to have rather limited value for biodiversity.

The level of biodiversity in agroecosystems is unclear. The study in Indonesia indicated that various forms of agriculture show lower levels of biodiversity than sustainable forestry. Among agricultural practices agroforestry emerges as the best option (Table 10). However, other studies have suggested that agroecosystems may have even higher levels of biodiversity than forest ecosystems.

In terms of carbon sequestration the natural forest, as expected, emerges as the best alternative owing to its large biomass. Accordingly, community-based forest management (for household use) and commercial logging are next best options. Annual agricultural crops are inferior to natural forest management (Table 10).

Regarding biodiversity, however, it should be noted that while the degree of variation may remain the same, the biodiversity of a managed forest or an agricultural cultivation are significantly different from that found in natural forest in terms of site ecology and species composition. While the degree of variation in mathematical terms may not differ, the original ecosystem is likely to be better adapted to the local conditions. Besides, each ecosystem has a unique existence value, which cannot be captured by another "set of diversity".

Table 10: Biodiversity index values and capacity for carbon sequestration
for selected land uses in Indonesia

Source: Fernandez 1998

3.2 Environmental status

3.2.1 Assessment of Indicators

The proportion of forest cover and the rate of deforestation are the most widely used indicators for environmental status in watershed areas. The original forest cover in all GMS countries was 97-100% of total land area. Accordingly, the current percentage of forest cover provides a good indication of the historical development and the overall level of degradation. The current rate of deforestation is more appropriate to depict on-going developments.

The proportion of forest cover is a convenient indicator in the sense that it is relatively easy to determine with remote sensing techniques. Further, it provides an estimate on the minimum size of area, which still retains relatively high capacity for watershed protection. It is also a reasonably good indicator for the status of biodiversity and carbon sequestration, even though it fails to reflect changes in forest quality (e.g. removal of valuable species, reduction of growing stock etc.).

However, it is important to stress that deforestation should not be considered the sole indicator of environmental degradation in watershed areas. If the focus is placed strictly on watershed protection (i.e. excluding impact on biodiversity and carbon sequestration), it is the sustainability of the land use replacing forestry, which ultimately determines the impact. If forest has been converted to sustainable agriculture, the negative consequences in terms of erosion would be less severe. Appearance of shrub vegetation on denuded areas can substantially mitigate the detrimental effects (cf. Table 8).

Theoretically, it would be more relevant to estimate the status of watershed protection based on the area under most unsustainable land uses. Based on Table 8, the key variables for watershed protection are the area of field crops, sloping hill crops and barred lands. The area of unsustainable shifting cultivation may also be a significant variable.

Unfortunately, the present methodology of estimating changes in land use is not able to distinguish between sustainable and unsustainable forms of cultivation, be they based on fixed or swidden agriculture. Therefore, lacking data on other possible indicators the following discussion concentrates on analysing the status and causes of deforestation. Despite its shortcomings deforestation seems to constitute a reasonably accurate indicator for the environmental status of watershed areas.

3.2.2 Deforestation

The loss of forest cover is still rather rapid in all GMS countries, perhaps with the exception of the Yunnan province* . In Thailand the proportion of forest area is the lowest in the GMS region and rate of deforestation is still very rapid. However, the pace has somewhat slowed down in the first half of 1990’s in comparison with the previous decade period. In Myanmar and Viet Nam the pace has not accelerated, but it should be noted that in absolute terms Myanmar’s annual loss of forest area is largest in the entire GMS region. Cambodia still has a large proportion of her land area under forest cover, but the present trend is disquieting, as deforestation is rapidly escalating (Table 11).

Table 11: Indicators for deforestation in GMS countries for periods 1980-89 and 1990-95

Source: FAO 1991,1996
Yunnan Institute of Environmental Science, 1998

The overall rate of deforestation provides an indication on the general trend, but it should be noted that there is often a significant difference between upland and lowland areas. In a long-term perspective the mountain forests have often been better preserved than the lowland forests because of their remote location, sparse population, and low levels of economic development. Examples of this pattern can be found in all GMS countries(Table 12). Forest cover is exceptionally high in the uplands of Cambodia, where it is estimated at almost 90 % of total land area.

However, there are many exceptions to this general pattern. Some upland zones have been subject to considerable pressure, as indicated by low forest cover. Examples of upland areas, where forest cover is lower than the national average include northern Lao PDR, Shan and Chin States in Myanmar, northeastern Thailand, and northern mountains of Vietnam. Often the reasons can be traced back to population pressure exceeding the carrying capacity of the area.

Table 12: Indicative forest cover in the upland areas of GMS countries

Source: See Country Reports

Data on current, short-term trends is unavailable, but the few observations available do not indicate any established pattern. For instance, in the Katchin State of Myanmar, which consists mainly of upland areas, deforestation between years 1990-95 stood at 1.1 %/a as opposed to the national average of 1.4 %/a estimated by FAO. In contrast, in Thailand the advance of deforestation seems to correlate with the proportion of remaining forest area. The higher the proportion, the more rapid the pace of deforestation. Between 1992 and 1997 the region, where deforestation has advanced most rapidly is in the northern uplands, whereas the slowest pace has been recorded in the northeastern part of the country (see Country Report on Thailand).

3.2.3 Direct causes of deforestation

Deforestation is most often attributed to logging and ensuing conversion to agriculture. The immediate cause of deforestation is most often conversion to agriculture either for subsistence production or cash cropping. For instance, between 1993-97 in Lao PDR, shifting cultivation (63 %) permanent agriculture (12 %) were the principal land uses to which forest areas were converted (Figure 1). The rest was converted to wood and shrubland (25 %), which was mainly caused either by forest fire or clearcutting of sites for future hydropower reservoirs.

Figure 1: Pattern of Forest Clearing in Lao PDR in 1993-1997

Source: calculated from data made available by FCMP MRC/GTZ 1999

The regenerative capacity of forests in the GMS region is quite high. If not cleared for agriculture, the logged-over areas usually able to regenerate successfully. In some rather exceptional cases clearcutting or very heavy and inappropriate logging may also be a direct cause deforestation. However, logging and road construction accelerate deforestation, as they enable local populations to move into formerly inaccessible forest areas. For instance, in Lao PDR clearing of dense forest is rare. Almost all deforested areas (95 %) had been degraded before they were cleared (Figure 1). Logging lowers the threshold for forest clearing, since removal of large trees reduces the effort needed for preparing the agricultural land. Had the logged-over areas not been available, deforestation would probably have been more localised.

From a policy perspective it is important to note that despite its negative impact on environment, deforestation has contributed to the economic wellbeing of local populations by making an expanded agricultural land area available. The net effect of logging and road construction depends on how the relative importance of environmental degradation, on one hand, and socio-economic gains, on the other, is judged (cf. Meyer & Turner II 1994).

Regarding the impact of subsistence agriculture and commercial agriculture, econometric models constructed for northern and northeastern Thailand suggest that in 1979-89 road construction and ensuing agricultural expansion accounted for approximately 20 % of deforestation, whereas population growth and ensuing expansion of subsistence agriculture would explain 40 % of deforestation. The balance would reflect changes in the price level for agricultural products. However, in the remote watersheds subsistence agriculture and population growth were probably even more significant agents of deforestation, as lack of market access hindered the development of commercial agriculture.

In a few countries expansion of hydropower reservoirs has been another significant, if not a major cause of deforestation. In Thailand by 1988 the total area under hydropower reservoirs was 5456 sq. km. However, this is only slightly more than the average area of annual deforestation between 1961 and 1988. The actual area of deforestation is probably less, because both forest and non-forest areas were inundated by the reservoirs. In Myanmar, the area of sites inundated by hydropower schemes is of same order of magnitude.

Real estate development schemes have received considerable media attention as causes of deforestation. In Thailand, there have been several cases, where authorities have allowed development of large plantations, holiday resorts and golf courses in the forest reserve areas. However, even though data are lacking, it is unlikely that they would represent a major cause of deforestation. On the other hand, even if the physical area affected by such incidents is limited, it does not imply that the public attention they have received would necessarily be disproportionate. While condoning forest clearing in benefit of business interests, the authorities have at the same time pursued attempts to evict long-time settlers from similar areas under the pretext of forest protection. Even if the physical area affected by such incidents is limited, the demoralising effect has probably been substantial. The double standard applied by the authorities effectively undermines the overall credibility of the legal system. Repercussions in the watershed areas may be serious, since a chronic lack of enforcement capacity makes the effectiveness of legal regulations to a great extent dependent on respect for law.

4. Link between poverty and environmental degradation

The argument linking poverty and environmental resource degradation has remained the centrepiece of policies concerning environmental degradation and it is still underpinning most development efforts in watershed areas in the GMS region. The basic logic behind this view is that poverty and scarcity of productive resources causes desperation. This, in turn, promotes overextraction of resources leading to their degradation and still greater scarcity and poverty, thereby accentuating the above cycle.

The validity of this argument has been often challenged mainly because it is seen to deny the value and sustainability of subsistence economy. The fact that development planners most often see economic development as the principal strategy for improving the situation, is often taken to imply that the planners see no alternatives to cash economy, and see income poverty as the main culprit behind environmental degradation.

To the extent this is true, the view can be rightly criticised. There are several examples of isolated, low-income communities exploiting their natural resources in a sustainable way. Degradation of environment starts only when they are exposed to external or internal pressures increasing resource scarcity. Typical examples of such disturbances include in-migration of new populations, natural population growth, and establishment of conservation areas or hydropower reservoirs, which deprive local populations of their traditional access to resources. Very often the poor themselves have little control over these developments.

However, the reason why economic development is so often chosen as the principal strategy, is not necessarily a narrow perception but rather a lack of alternatives. Some of the negative developments are irrevocable or at least they cannot be reversed in the short term. Population growth can be gradually slowed down, and further in-migration can be restricted, but these measures do not reduce land scarcity. If the availability of vital resources has been reduced to a point, where people face malnutrition, they take to unsustainable practices – as the last resort. In such circumstances, environmental policies are left with few alternatives other than focusing on poverty reduction through economic development. Generation of income enables people to compensate the decrease of land and other natural resources with another productin factor - capital (Figure 2).

However, it needs to be stressed that if resource scarcity has not developed into a problem, the choice of options is wider. In particular, measures that can prevent the emergence of resource scarcity (e.g. control of in-migration, family planning) have a greater role to play. The principal flaw of development planners is maybe that they have overlooked this option, even when it has been fully relevant and feasible.

Figure 2: Links between poverty and environmental degradation

Focus on poverty should not draw attention away from the fact that resource scarcity is not the only reason for environmental degradation. Unsustainable logging and establishment of hydropower reservoirs are typical examples of other agents of environmental deterioration. It is also evident that poverty reduction will not automatically lead to sustainability in resource use. There are several examples in the GMS region, where unsustainable practices have continued despite reduced poverty. For example, promotion of cash cropping, which has been the key policy instrument for poverty alleviation in the rural areas, has encouraged further clearing of forest for agriculture, increased pesticide use beyond sustainable levels and aggravated conflicts over water supply (e.g. in Thailand). It has also been pointed out that the intended solution – raising of living standards – has indirectly been encouraging unsustainable practices in communities, which did not experience any problems. In areas, where people had been living in a sustainable manner within the limits permitted by their resource base, the market-oriented policy has increased the pressure on natural resources by infiltrating consumerism.

On the other hand, it may be argued that these policies have been appropriate in the sense that they may be able to deflect a potentially more damaging development. Emergence of viable commercial agriculture has induced a transition from shifting agriculture to sedentary practices, which are less extensive land uses. Thus, in the densely populated areas, where swidden agriculture is often an unsustainable practice, the net environmental effect of intensifying of agricultural land use may have been positive.

The "economic development approach" to resource conservation is clearly a two edged sword. It may facilitate the transition to sustainable use of natural resources, but it can also quite easily lead to unsustainable practices unless proper precautions are implemented. However, its justification stems from the notion that without food security people have no alternative but to adopt practices which satisfy their short-term needs, even if they perceive and recognise their harmful effects on environment. Economic development and surplus production will create a space for decision-making, where options beyond ones satisfying immediate needs can at least be considered.

5. General development strategies in GMS region

The overall socio-economic development plans of GMS Governments have much in common. This is probably a reflection of similarities in the socio-economic setting and of the convergence of economic policies in the region. Regarding general objectives with a bearing on watershed areas the following ones seem to be shared by all Governments.

• Rapid economic growth
• Reduction of poverty and more equitable distribution of income
• Utilisation and conservation natural resources in a manner, which contributes to social and economic development without jeopardising environmental sustainability
• (Increased) emphasis on development of rural areas

Needless to say the objectives are rather general and all-embracing. While it is well known that discrepancy between declarations and deeds is often wide, such statements are important in the sense that they bring the issues to discussion agenda. Considering developments in remote watersheds it is significant that all Governments recognise the need to reduce disparities between urban and rural settings. Further, as the conflict between socio-economic development and environmental sustainability is at the heart of the agenda for upland development, it is positive that the issue has invariably made its way to official policy statements.

5.1 Impact of growth on upland development

The available policy documents indicate clearly that economic growth is seen as the fundamental driving force underlying all development strategies. Its primacy on the Government’s agenda complies with the most prevalent policy approaches on poverty, which see economic growth as the primary engine of poverty reduction. In the first instance the emphasis is on expanding primary production, but the long-term goal is to induce industry-led growth.

In general terms, the policy approach seems valid in the GMS countries as well as elsewhere. Industries and service sectors are growing more rapidly than agriculture, and poverty reduction has been most effective in countries with rapid economic growth. One of the most successful examples in the GMS region is Thailand. The country has achieved an exceptional record of economic development over the last 30 years, as witnessed by the rapid expansion of the national economy at an average rate of 7.8 per cent per annum. The drop in numbers of people living in absolute poverty has surpassed all expectations, falling from 57 per cent of the population in 1962 to only 13.7 per cent of the population in 1992. The development in China PDR is no less impressive, especially as the region’s lowest poverty rate, 6.5 % (8.3 % in Yunnan province), has been achieved at a stage, where the overall level of economic development is still comparatively modest.

Regarding upland development, it is evident that overall growth in an important issue even for the remote areas. The fundamental problem from their standpoint is to ensure that they are able to share in on the benefits. The "trickle down effect" seems to loose momentum, the further from the growth centres one goes, as the remote areas have weaker links to the mainstream economy. A skewed distribution of benefits may substantially weaken the correlation between growth and reduction of poverty (Box 1).

GDP growth and integration with formal economy is most vital for watersheds, where land scarcity is compelling the development of cash economy. Shortage of land is leading to reduced fallow periods in swidden agriculture, which gradually diminishes the agricultural yields. Ultimately the basic needs cannot be met by cultivating the available land area. In order to supplement their farm production and to bridge the food gap, the farmers need to generate cash.

In areas, where the availability of land is not a constraint, GDP growth would also be conducive to increasing incomes and reducing poverty. However, without the land constraint, it may also be possible to reduce poverty by developing subsistence-based livelihood strategies.

Box 1: Relationship between distribution of benefits and poverty

5.2 Options for public support to rural/upland development

5.2.1. Government investments

Balancing the priorities between richer and poorer regions is a dilemma all GMS Governments are faced with when targeting their investments and expenditure. Poor regions often suffer from low rates of growth, and investments made there may not generate as much tax revenue as spending in richer regions would do. From an equity stand transfers from richer regions to poorer ones would be justified, but the problem lies in the reasoning "wealth can be redistributed only after it has been generated". This is a severe problem particularly in countries with very limited revenue base, as an even distribution of resources over the whole country may shrink the resources available to all regions below a critical level, which is necessary to have any significant impact.

Nonetheless, many GMS Governments, most notably Vietnam, are making attempts to equalise differences between regions with Government transfers. The available data indicates that in late 80’s and early 90’s the provincial Governments in the upland areas received approximately 30 % of their expenditure as transfers from the central Government. In the wealthiest area – Ho Chi Minh City – almost half of Government revenue was transferred to other regions. Similar policies are applied by the Lao Government, although on a more modest scale (see also Country Report on Lao PDR). In Yunnan province, public support to disadvantaged areas is also substantial (see Country Report on Yunnan province).

In recent years, however, there have been indications that some GMS Governments are increasingly prioritising economic growth and concentrating their development efforts in urban areas. For instance, the Public Investment Plan for 1996-2000 in Vietnam is proposing to increase the proportion of public investments in the three so-called growth centres (Hanoi-Haiphong, Quang Nam Da Nang, Ho Chi Minh City/Dong Nai-Vung Tau) from the past level of 18 % to 30 %. While such trends are to some extent understandable owing to the large development potential of growth centres, the efforts to promote economic development in the more remote rural areas seem modest given the disparity in socio-economic well-being. Even if shifting emphasis towards rural development – especially upland development - may not be warranted in strict economic terms, the current disadvantage of the upland regions would make it justified on equity grounds.

However, the pursuit to increase public funding in upland areas should be coupled with efforts to ensure efficient use of available funds. The administrative costs of Government-sponsored development programs are often substantial in relation to the benefit that is actually received by the target group. For instance, it has been found out that for a large number of Asian and Latin American countries that have implemented poverty alleviation programs, administrative costs have been anywhere between 10 to 50 per cent of total project costs.

In addition, it is necessary to stress that often the most decisive measures for upland development are the changes in the institutional and legislative frameworks. While public infrastructure investments are absolutely necessary and do have their place, the restricted resources available to many GMS Governments make the impact of public investment is limited even under most optimistic scenarios of fund mobilisation. Therefore, the Government’s capacity to mobilise productive and human assets may be of even higher importance considering the overall socio-economic development. In rural areas, the most important productive asset is land, and therefore the issue of land tenure should have the highest priority in terms of Government’s development policy.

5.2.2. Fund transfers from lowland to upland

With reference to upland areas it has been claimed that the general pattern