In order to understand the implications of trade for forest management, it is necessary to consider how forests and land use alternatives are valued in different geographical locations. This requires an understanding of the different stages of forest development and of the factors that drive progression from one stage to the next. The extent of trade and the outcomes of trade depend very much on the stage of forest development. The effectiveness of forest policies and processes in tackling forest management problems and the degree to which they are impeded or reinforced by changing trade patterns and policies can also be closely related to the stage of forest development. Each stage of forest development is characterised by a unique competitive balance between forests of different types and between forests and other land uses. The effect of trade and of policy interventions of different types is to alter this competitive balance with implications for the area under forest and its quality of management.

4.1.1 Patterns of Forest Development and Determining Factors

The values that people attach to forests within the current market context are ultimately revealed in their land use decisions. Forest land use is more complicated than many other land uses as it involves activity in three types of area, managed forests (which can include industrial forests and forest plantations, more scattered trees in residential areas and well-managed agroforestry), degraded open access forests, and the unmanaged, mature and natural forest frontier. There are environmental and social effects from forest land use in all three areas. Managed forests are affected directly by activities such as tree planting, timber stand improvement, and harvesting. Harvests from the other two areas affect the incentives for forest management indirectly by producing substitute woody material and, thereby, replacing the demands for, and postponing the development of, managed timber stands.

The problem for forest analyses which seek to describe and interpret these patterns of land use is all the more complex because not all regions contain all activity types. Some developed regions contain both managed stands and natural forest frontiers but very little degraded forest. Other developed regions contain neither plantations nor degraded open access forests, and their harvest activities only occur at the mature natural forest frontier. Some less-developed regions contain only mature natural forest frontiers and degraded open access forest. And other regions in less-developed countries contain plantations, degraded forests, and mature natural forests.

Trade in forest products in general, and in timber in particular, can affect different regions and socio-economic groups differently and it is necessary to understand an entire spectrum of potential causes and effects. In turn to assess the impact of various types of policy intervention, whether forest-related, trade-related or extra-sectoral, it is important to understand these different effects.

4.1.2 Analytic Methodology: Forest Frontier Approach

A model developed by Hyde (forthcoming) relates the pattern of forest development in different regions to three functions, the value of land in agricultural use, the value of land in forestry and the cost of obtaining and protecting property rights to land and resources. All three functions are in turn dependent on accessibility. The agricultural and forest land values decline as the distance from the nearest market access point and/or cost of transport increases. The costs of establishing and maintaining property rights are also heavily influenced by the distance from the market as this reflects the level of public infrastructure and effective control by public institutions. Further detail on this model is given in Annex 1.

Three different stages of forest development can be distinguished:

• Stage 1 - a new forest frontier region

• Stage 2 - a developing frontier

• Stage 3 - a mature frontier

In a new forest frontier region, the value of land in agriculture exceeds the value of land in forest up to the point where distance from the market makes any type of exploitation uneconomic. The abundant forest resources initially have little value compared with agricultural land. Where trees interfere with agricultural production they are removed. Settlers remove trees wherever the value of new agricultural land plus the value of the trees in consumption (for example, for construction timber or fuelwood) exceeds their removal and delivery costs. Otherwise the trees are left standing. It is thus agricultural values which determine the land use pattern at this stage.

This stage characterises for frontier settlement in many parts of South America, migrant settlement in Sumatra in Indonesia, new upland migration in the Philippines and subsistence settlements in Zambia, for example, in the latter 20^th century.³³

In a second stage of forest development, a developing frontier, demands for construction timber and fuelwood may make open access wood harvesting viable in areas that are too removed from the market to have value for agriculture. The forest will be degraded to the level where the expected returns from harvesting are less than the opportunity costs of the labour and capital used in their extraction.

Both deforestation and degradation are greater in regions where the opportunity costs of extraction are smaller (e.g. in poorer countries where wages are lower and labour is the largest component of logging costs). Illegal logging will be exceptionally difficult to control because the smaller scale logging operations operated by such labourers are more difficult to monitor and because the returns to illegal activities are greater that the risks of getting caught for those lower wage individuals who engage in them. This description characterises the poorest rural areas of many developing countries today, including for example the southern two-thirds of Malawi and portions of Tamil Nadu in India, of Nepal's hills, and of China's dry and remote Qinghai province. The open access region is unusually degraded in these examples.

In a third stage of forest development, a mature frontier, the distance of the open access natural forest from the market becomes so great and local prices become high enough that intensively managed forests on land with protected secured rights closer to the market become competitive with timber harvesting from the mature natural forest. Prices have not risen sufficiently to induce intensive forestry in all of the markets of the modern world, but there are plenty of examples where they have (e.g. the large plantation areas in Chile, China, India, Japan, South Africa, Southern Brazil and Western Europe to name but a few).

4.1.3 Analyzing the Impact of Trade and Policy Interventions on Forests

Regions in stage I are lesser participants in external trade of any sort, and the forest products they do trade are not transported very far. For trade in extractive products such as industrial timber, we can focus on regions in stages II and III of forest colonisation. The origin of these products may be from intensive or extensive managed forestry for regions in stage III, and from both open access resources or the mature forest frontier in either stage II and III. As we previously suggested, open access resources and the mature forest frontier will be a more important source of industrial wood in regions of developing countries where the population is still low and the institutions affecting property rights are not as well developed.

The impact of trade on forest will be uncertain - almost no matter what the initial condition of the forest in the two trading regions. All trade affects forests in more than one region, and trade often creates net global environmental gains. This can be demonstrated through considering the case of trade between two regions that are both in the mature third stage of forest development. Trade in sawnwood from Finland to Germany or from south-eastern Canada to the US northeast are examples. After their markets become open to trade, consumers in the first or importing region recognise that they can decrease costs by purchasing from the exporting region. These consumers are now willing to pay only the new lower price for forest products. This means that the value of forestland declines in the importing region and this region's own output of forest products declines. Prices rise to a new higher equilibrium level in the exporting region because of the increased demand from new consumers in the importing region. The value of forestland increases and the second region's output of forest products also increases.

Loggers in the importing region may continue to extract timber from all of their old harvest sites as long as the existing capital and infrastructure are in good condition and the revenues from their harvests cover the variable costs of the logging operations. However, over time the declining price in the importing region must cause these loggers to decrease their level of the harvesting, allowing some land to be converted to other land uses, allowing some forest recovery in the open access lands, and delaying some harvests from the mature forest frontier. At the same time, increasing prices in the exporting region must encourage an expansion in the area under forest management in this region, and additional extraction from the open access lands and from this second region's natural forest.

The net effect of trade on land use in the combination of the two regions is uncertain. Harvests from low cost but highly productive managed forests in the exporting region may replace harvests from the natural forests of the importing region. Most would see this saving of natural forest as an environmental improvement. On the other hand, managed forests tend to be more costly while harvests per land unit tend to be greater from mature natural forests. Therefore, trade may cause a net shift that favors relatively greater reliance on the exporting region's natural forests. Most would see this as an environmental loss. Either is possible. The only certainty is that there will continue to be some harvest activity from the mature natural forest frontier in the exporting region--because its costs of production are low and perfect restriction of all illegal activity on this land is very costly.

This characterisation of trade between two regions in the mature stage of forest development is only one of four possible cases. Another frequent occurrence is trade from regions in stage II to regions in stage III (e.g. the many tropical countries in Africa, Asia and Latin America that ship logs to developed countries). Two regions both in stage II can engage in trade (e.g. the shipment of logs from Cambodia to Thailand is an example). Regions in stage III can also export to regions in stage II (e.g. the shipment of wooden furniture from China to numerous developing countries).

Table 4.1 summarises the effects on forest management, on the degraded open access lands, and on the unmanaged natural forest for all four possible cases. The first conclusion to be drawn from Table 4.1 is that there are always positive effects on consumer welfare in both the importing and exporting region. A second conclusion is that there will always be growth in employment and production in the exporting region. Net growth is also normally assured because the overall lower prices mean that aggregate demand increases and, with it, aggregate production must increase as well. In fact, consumers always gain in the importing region. There would be employment losses in this region but they may be temporary, lasting only as long as it takes unemployed loggers and other woodworkers to find alternative employment. Employment and production always increase in the exporting region and they increase more than they decline in the importing region because the overall lower prices mean that aggregate demand increases and, with it, aggregate production must increase as well.

Table 4.1 Effects of trade on consumers, employment and production and forest management by importing and exporting region

Case

Effect on

Importing region

Exporting region

Net effect

Stage III region imports from stage III region

Consumers

Production and

employment

Managed forest

Mature natural forest

Degraded forest

gain

decline

contract

recover

uncertain

expand

contract

degrade further

gain

uncertain

Stage II region imports from stage III region

Consumers

Production and

employment

Managed forest

Mature natural forest

Degraded forest

gain

decline

n.a.

recover

uncertain

expand

contract

degrade further

gain

expand

uncertain

Stage III region imports from stage II region

Consumers

Production and

employment

Managed forest

Mature natural forest

Degraded forest

gain

decline

contract

recover

uncertain

expand

n.a.

contract

degrade further

gain

contract

uncertain

Stage II region imports from stage II region

Consumers

Production and

employment

Managed forest

Mature natural forest

Degraded forest

gain

decline

n.a.

recover

uncertain

expand

n.a.

contract

degrade further

gain

n.a.

uncertain

n.a.: not applicable

The impact on forest management of trade is more complex to determine because of the different types of forest involved and the fact that trade affects the environment of the importing region as well as the environment of the exporting region. The important environmental question is whether the reduction in forest activity in the importing region is more environmentally beneficial than the environmental loss caused by the increase in forest activity in the exporting region. For this reason, the net effects of trade on the combined forestland of both regions are uncertain. The land area under forest management increases unambiguously only for the case of imports to a region in stage II from a region in stage III (case B in Table 4.1). The effect of trade on the natural forest is ambiguous in all cases.

Furthermore, the ambiguity only increases as the analysis is extended from two regions or countries to global trade. Global trade--for almost any industrial wood product--undoubtedly includes regions and countries that fit the descriptions of each of the four cases. The sum of many uncertain cases is surely only more reason to be uncertain about the net effect on the world's forestlands. The available data are physical, not economic, and the economic measures of managed forests, degraded forests, and natural forests vary from market to market. Many countries contain multiple wood markets--or regions. Therefore, the physical data within each national forest survey can only provide a loose impression of what is happening at each economic margin.

Even that impression would be subject to change as logging technologies, utilisation technologies, and rural infrastructure (especially roads) change. This is a crucial point because these rates of technical and institutional change can exceed the growth rates of commercial plantations and they certainly exceed the growth rates of mature natural forests. Technical and institutional change can be the most important determinants of trends in the preferred species and size classes of industrial wood over the long run. They explain why loggers return to harvest in already degraded forests. The dominant type of technical and institutional change--roads, mill utilisation, or logging utilisation--is the common source of the most important shifts in long-run land use margins for most regions and countries.

Governments intervene in the forestry sector in cases where markets alone fail to create socially desired levels of production and allocation. Traditionally, the motives for intervention have been to ensure long-run timber supply and to promote employment and regional development. More recently, the policy objectives have expanded to address environmental concerns including the provision of global environmental services and social concerns such as community development. Other stakeholders such as NGOs and the private sector have also started to play a role in introducing policies to achieve these broader objectives. Policies in other sectors however, may have a greater effect on forest management and on how trade influences this.

These interventions in the forest sector, trade and in extra sectoral areas that influence forests have to be considered in the context of the three stages of forest development. Forestry is unusual in having three activity types (managed forests, open access lands and harvesting from the mature forest frontier). Any particular policy instrument can have simultaneous but contrasting effects on these different activities. For example, price incentives would induce an expansion in forest management, but they would also induce expanding harvests and a contraction of the remaining area in mature natural forest. The effect of these policies and processes is to alter one of the key functions which determine the land use decisions made between different types of forestry, or between forestry and other land uses. Forest policy instruments are most likely to alter the forest value function but some will affect forest management through their impact on the cost of enforcing and maintaining property rights. Forest trade policy instruments will also primarily affect the forest value function. Extra-sectoral policies will work primarily through their effect on the agricultural land value and the costs of market access.

4.2 Extra-sectoral Influences and Factors

The world's forests are subject to forest management policies and other, extra-sectoral influences. Some of the most marked impacts on forests come from influences outside the sector. These include external shocks outside of the control of national government, such as the East Asian financial crisis of 1997, internal shocks such as civil war and conflict, macroeconomic policies, changes or policy interventions in sectors which compete with forestry for land and other inputs. Agriculture is a key sector in this regard. Policies adopted in other countries through their effect on trade patterns can also spill over into impacts on forests and competing land use sectors in other countries. This chapter, surveys the range of extra-sectoral influences which can affect the relationship between forest management and trade³⁴.

4.2.1 External and Internal Economic Impacts

If trade with another country consumes a large share of domestic production, and of forest production in particular, then a decline in the economy of the importing country can have a measurable impact on the forests of the producing and exporting country. The impact of the 1997 East Asian crisis on exports from Indonesia is a good example (Hyde, 2003). The impact occurs in two stages.

The importing country's aggregate demand declines in the first stage. Prices must fall. Initially, producers in the exporting country may try to maintain their former production levels despite the lower prices, and the impacts on the forest will continue relatively unchanged.

Eventually, manufactured capital (e.g., logging equipment, plant and machinery in the processing industries), infrastructure, and managed forests all require major repairs or replacement. This signals the second stage. Major repair and replacement is a fixed cost of operation. There are no means to pay fixed costs until the export market recovers and prices regain their former level. Therefore, some forest operations will continue with deteriorating, less efficient, equipment which is more damaging of the forest. Others will be discontinued. The net effect of the second step must be a long-run decline in both managed forests and harvests from the natural forest frontier.³⁵

Internal shocks which affect forest production and trade include: significant social and institutional change and adjustment (e.g. civil disorder or even civil war), comprehensive macroeconomic policy adjustment (e.g. structural adjustment), and the more specialized fiscal and monetary instruments of macroeconomic policy.

The impact of these internal shocks is generally to create uncertainty. In general, managers postpone investment in the presence of uncertainty. They draw down their forest stocks while covering their variable costs of operation--until the general social and institutional outlook becomes more settled. In the presence of extreme uncertainty, loggers and managers not only postpone investment but they also may become more aggressive in their harvesting of existing economic forest stocks. An example is given by the logging activities of the Cambodian army in the 1990s. The army redirected some of its men and equipment into logging operations and then sold the logs across the border in Thailand in order to obtain finances to support its military activity. It extended its logging into areas that had not been logged, areas that civilian loggers could not afford to log, increasing the country's annual harvest level several times over.

4.2.2 Macroeconomic Policies, Structural Adjustment and Poverty Reduction Strategies

Economic growth stimulates demand for agricultural, forestry and mining products. This may boost internal markets as well as trade, because of increased needs in machinery and investment. In the long run, this may lead to economic development and may be beneficial to SFM, but this is conditional on other variables, such as investment in the forest sector or land-use decisions.

The government may assign a strategic role to certain sectors within the overall development strategy by favoring one sector over another (Wunder 2000, van Soest 2000). Clearly, this choice depends to a great extent on wider socio-economic variables such as relative prices and exchange earnings in different sectors, on poverty or population growth, but it can clearly be sustained by governmental decisions.

Many studies have highlighted the importance of macroeconomic policies, in particular in association with structural adjustment programmes (SAPs), in influencing both forest management and trade (Barbier et al 1994). These have focused on the impacts of either a package of macroeconomic and sectoral measures as in the SAPs or specific fiscal and monetary policy instruments. But the linkages with forest management are complex and studies have produced conflicting conclusions.

While structural adjustment programmes may initially create uncertainty, giving rise to the effects described in the preceding section, their longer term effects are more complex. Brazil's experience between 1970 and 1995 provides an example. Brazil suffered a 22-fold increase in international debt, a 40-fold increase in long-term interest rates, and annual inflation rates as high as 2 560 percent following the oil shock of 1974 and during twenty years of diverse domestic economic policies (Young 1996). The International Monetary Fund imposed a program of structural adjustments on Brazil as a condition for the Fund's assistance with Brazil's international debt.

Young (1996) traced the effects of this program on the Amazon's natural forest frontier, indentifying four fundamental relationships between broad policy and deforestation:

• Reduction in subsidized agricultural credit as a result of the need to reduce government expenditure discouraged conversion of forest land at the natural forest frontier to agriculture

• Reduction in expenditures for roadbuilding discouraged deforestation and improved the conditions of the forest at the natural frontier.

• Export incentives designed to improve Brazil's balance of payments led to appreciating prices for more capital-intensive commercial agricultural products like soybeans. This increased land values for those small farms that could be converted for use by the larger commercial agricultural operations. These small farmers than migrated to the forest frontier and cleared forest for new farming activities.

• The shift to capital intensive production reduced the demand for agricultural labour causing unemployment and a decrease in the real agricultural wage. The latter effect was reinforced by other programs of the central government designed to decrease the minimum wage. As a result, agricultural workers also migrated to the frontier where they converted Amazonian forest for their new smallholding agricultural activities.³⁶

The notable conclusion is the contrast in effects. Brazil's structural adjustment program had multiple and mixed effects on the forest frontier resulting in an indeterminate net long-run effect on deforestation.

Contreras-Hermosilla (2000) notes that 'SAP's may unintentionally encourage forest decline for three reasons : first, they may induce unemployment and greater poverty leading to migration to forest areas… Second, SAPs often stimulate agricultural exports at the expense of forested land…Third, SAPs may 'stimulate forest exports based on unsustainable methods'. Kaimowitz goes even further by claiming that SAPs 'tend to boost production of tradable goods without successfully promoting more difficult institutional reforms that could counterbalance the increased pressure on forests'.

Indeed, it seems that the influence of SAPs do particularly concern the forest sector: 'the expansionary impacts of currency devaluations, tariff liberalization and reduction of real interest rates may be most directly and adversely felt in the natural resource use' (World Bank quoted by Contreras-Hermosilla, 2000).

Overall, SAPs target trends and factors which hinder SFM. SAPs should have positive long term effects, but do not always have. It should be noted, hower, that one possible reason for insufficient performances of SAPs is their often partial application (World Bank).

SAPs have given way to poverty reduction strategy papers (PRSPs). Initially required by the IMF and World Bank as a basis for access to debt relief in Highly Indebted Poor Countries (HIPC), PRSPs have been required by all countries supported by the International Development Agency since July 2002. Interim PRSPs (I-PRSPs) are road maps to full PRSPs. As of April 2003, there were 26 full PRSPs and 45 I-PRSPs. Bilateral donors are also increasingly subscribing to PRSs and they have thus emerged as a central determinant of the development agenda in many countries. Whilst largely taking the place of structural adjustment programmes (SAPs) in these countries, PRSPs continue to require the 'traditional' macroeconomic prescriptions of the IMF - privatisations, capital market liberalisation, import liberalisation, fiscal restraint and state down-sizing. Some PRSPs introduce the proviso that economic growth should be “quality growth”, but they typically do not discuss why the proposed actions are likely to work better for poverty reduction than comparable actions have done in the past, and what are the critical things that need to happen (Booth & Lucas 2002).

PRSPs are critical as frameworks for fostering trade based on sustainable forest management and local livelihoods. Yet PRSPs to date merely demonstrate that there is a long way to go to develop bottom-up, continuously improving processes rather than one-off encyclopaedias of externally-driven ideas. The recognition of forests as a development asset has so far been disappointing in many PRSPs. A recent study looked at the 11 PRSPs and 25 I-PRSPs in Sub-Saharan Africa and noted that whilst 84 percent of them touched on forestry issues, almost none of them were convincing about forests-poverty links and their future potential; only Malawi and Mozambique made a significant mention (Oksanen & Mersmann 2002). Of course these papers in themselves tell us little about implementation.

4.2.3 Fiscal and Monetary Policies

Fiscal policy refers to the use of adjustments in government expenditure or taxation to influence the overall level of activity in the economy. It is commonly used as a short-term first step out of a period of economic stagnation or decline. Because of its association with high levels of government borrowing, the use of fiscal policy runs counter to the prescriptions of most SAPs which tend to advocate reductions in public expenditure.

Government expenditures provide the new demand to employ previously unemployed resources. They have their largest beneficial effects when the sectors receiving the influx of government funds are sectors that both respond and grow rapidly themselves and also link with numerous other sectors of the economy through their demands for the products of those other sectors or their supply of inputs to them. These links mean that growth in the first sector is a source of growth in the linked sectors.

The construction industry is often a prime candidate for intervention of this kind. Construction responds rapidly to new demand for its product, either domestic housing or an input to production in other industrial sectors. Its use of wood as one of its own inputs makes the demand of the construction industry of prime importance for forestry and the wood products industries. Injection of government expenditure into construction therefore leads to an increase in activity at all margins of forest operation, managed forests, as well as harvesting from open access degraded forests and mature forest at the natural forest frontier.

The effects on forests of injections of government expenditures into other sectors depend on the links between those sectors and the forestry sector. For many injections the links will be small and the effects on forestry will also be small, and delayed as well.

An alternative fiscal tool for encouraging a stagnant economy is to reduce taxes. The expectation is that tax decreases allow consumers greater income to use on consumption and the additional consumption will fuel economic expansion. Of course, only the smallest share of additional consumption will be on forest products and the effect on forestry will probably be minimal. The longer-term effects of fiscal policy (either expenditure or taxation) depend on the effectiveness of that policy in generating economic growth and the effect of general macroeconomic economic growth on forestry.

The main instrument of monetary policy is the interest rate. Monetary policy is effected by central banks adjusting the interest rate, raising it to dampen growth and control inflation in times of full employment and decreasing it in times of economic stagnation in an effort to encourage renewed investment and reinvigorate an economy with underemployed resources. This has two types of effect on the forest products sector. If the policy is successful, lower interest rates increase the demand for construction by all other sectors of the economy. The increase in construction increases that sector's demand for wood products, and timber harvest levels increase at all land use margins. But changes in the interest rate through their effect on the opportunity cost of capital have a more direct effect on harvesting rates and rotations for managed forests.

The Faustmann model predicts that lower interest rates cause forest managers to decrease harvest rates as they lengthen timber rotations and increase their forest stocks. This is a partial prediction, however, because the Faustmann model provides no indication of the impacts of interest rates on other types of forest activity in open access or forest frontier areas. During the 1994-1995 implementation of the Brazilian Plano Real, for example, which stabilised interest rates, there was a marked increase in deforestation (Macqueen, 2003).

Where the value of a country's currency relative to other national currencies is controlled by the central bank or the central government, then the exchange rate can be a second component of monetary policy.³⁷ A move from fixed to market-led exchange rates has been a common feature of SAPs and other economic liberalisation programmes. As exchange rates affect the relative competitiveness of exports and imports they have a significant impact on trade patterns, often greater than that of trade policy. The devaluation in Mexico in 1994 is considered to have had more impact on forest product exports to the US than the provisions of NAFTA (Lyke 1998).

4.2.4 Policies on Agriculture, Energy, Mining, Demography and Health

As policy changes induce expansion or contraction in the sectors that compete with forestry for inputs, these policies also affect forestry and forest trade. The important competition in inputs is for land, most generally between agriculture and forestry. Much forest conversion has been stimulated by policies to promote agriculture through giving land title and/or compensation to people clearing land for agricultural use. But the impact of agricultural policy is complex as it depends on the stage of forest development and because not all agricultural improvements use more land.

Agricultural policies affect agricultural land use, including those new agricultural lands that are converted from forests. For example, the US and Canada encouraged agricultural settlement on their frontiers in the 19^th Century, Brazil offered land and incentives to colonists of the Amazon in from settlement programmes dating back to the 1960s, Indonesia financed the movement of new settlers to the forest frontier in a policy called “transmigration” in the 1980s, and Finland compensated farmers for clearing new land for agricultural use in the early 1990s. In each case, agricultural (or population) policy increased the value of land in agriculture, and encouraged agricultural expansion within a region in the first stage of forest development.

In the second and third stages of forest development, agricultural policies no longer have an effect on the forest frontier. By this time, the more common agricultural policies are subsidies for agricultural inputs such as fertilisers and price supports for agricultural outputs. Input subsidies tend to favour capital inputs relative to land inputs and can thus lead to a more intensive use of land. Depending on the magnitude of the subsidy, these programmes can either expand or contract agriculture's use of the degraded open access lands in the second stage of forest development and they can either increase or decrease agriculture's competitive position at the intensive margin of forest management in the third stage.

The effects of agricultural input subsidies are compounded by the effects of government agricultural research programmes. They represent substantial government investments, and some have produced phenomenal increases in agricultural productivity and decreases in agricultural costs. They tend to make capital inputs less expensive and more productive and so are relatively capital-using and land-saving.

The effects of price supports for agricultural outputs on forests is much clearer than in the case of input subsidies and research. Agricultural price supports increase the value of land in agriculture at all distances from the market. As a result for regions in the second stage of forest development, agriculture expands into some of the degraded open access lands which previously were too far from the market to be viable for agriculture. For regions in the third stage of forest development they improve agriculture's ability to compete with the intensive margin of managed forestry, thereby converting some land away from forest management.³⁸

There are dynamic effects which complicate the relationship between agricultural policy and deforestation. Some dynamic and general equilibrium models of adjustment programmes show that policies which raise prices received by farmers in the short run may reduce urban demand for foodstuffs, making the ultimate impact on deforestation inconclusive (Kaimowitz and Angelsen 1998). Pro-export policies designed to increase agricultural exports are likely to have stronger deforestation effects than policies that promote production for the domestic market. This is because an increased supply of agricultural exports is less likely to put downward pressure on prices, and dampen the initial effects of the policies. Similarly pro-agricultural policies are likely to have stronger deforestation effects in the contexts of globalised agricultural markets and trade liberalisation.

The agricultural programmes of the North American and Western European countries may be the best example of long-term programmes that spillover to affect the forests of other countries. But the effects are complex and whether the outcome is positive for forest management depends on the particular circumstances. The US and Canada spend roughly US$ 45 billion annually on agricultural programs and the European Union spends an even larger amount. These immense public agricultural expenditures fund the excessive use of all inputs, including large shares of North American and European land that would otherwise be under tree cover rather than in agricultural use.

But there is also an effect on trade patterns and hence on forests in other countries. The use of additional inputs produces large agricultural surpluses, some of which are exported to developing countries where their artificially low prices drive out competition from local agriculture production. Corn production in Mexico, which has been adversely affected by subsidised imports from the US, provides an example (Oxfam 2003). Alternatively, these subsidised exports may compete with developing country exports as in the case of EU and US subsidies to cotton, which through their effect on the world price, are damaging the prospects for West African cotton growers and exporters. As a result, the value of land in agriculture is lower and the level of agricultural production lower than in the absence of the subsidies.

What this means for forests depends on the type of agricultural systems that are affected. On the one hand it should produce a contraction in the area under agriculture, reducing agriculture's use of open access degraded land for regions in the second stage of forest development and making it less competitive with managed forests for regions in the third stage of forest development. But the contraction in agriculture means contraction in employment also. Some unemployed local farmers and agricultural workers may return to subsistence farming which uses land more extensively than commercial agriculture. Therefore, one effect of the North American and EU agricultural policies is to convert frontier forest in developing countries out of forest and into subsistence agricultural use (The Economist, 2002).

It is conceivable therefore that the North American and EU agricultural support programs are more destructive of global forests than all commercial forest activities at the frontiers of developing countries. However, this effect has not been examined quantitatively. Revisions of agricultural policy may have greater forest conserving effects than direct environmental policies such as forest certification, improved silvicultural standards, or increased enforcement against illegal logging. Furthermore, they would be easier to administer because they simply require reducing or discontinuing existing subsidies. They do not impose the monitoring difficulties inherent in forest certification, silvicultural standards, or effective controls on illegal logging.³⁹

The mining sector can have great or small direct impact on forest decline according to the area and the mining technology applied, but it always plays an important indirect role (see Kaimowitz 1998). This is also true for its influence on forest product trade, as it impacts on infrastructure and employment opportunities.

An oil or mining boom creates rent opportunities which may or may not be used for SFM. Imports will increase, and depending on the nature of imports, the impacts on SFM will be different. If more food products are imported, the domestic food crop market will shrink and marginal farmers are likely to move to urban areas, which decreases pressure on forest land. On the other hand, if more other consumption goods are imported, the impact on the agricultural and forest sectors are less clear.

Wunder (2000) for example analyses the influence of the mining sector through oil cycles and macroeconomic changes. He notes that the expansion of oil exports has lead to a marked decrease in timber export: Wunder (2000): 'In 1960, timber made up almost three fourths of Gabon's exports, but with the expansion of oil exports, this share was reduced to less than 10% by 1980 (Pourtier 1989: 191)'.This has had important impacts on land use change in forest areas: 'The single most important transformation of Gabonese society during the last half-century has been the accelerated urbanization of a forest people (Walter 2000).

In Gabon, 'oil wealth was generally allocated with a strong urban bias, favouring prestige projects in urban construction, infrastructure in parastatal companies and in urban social sectors. The indirect impact (of most public projects) was to massively pull labour out of rural areas towards remunerative employment options in the cities, as civil servants, in parastatal companies, in services or in construction.' (Wunder 2000). Again, according to Wunder, 'this urban development bias, and the correspondent neglect of agriculture, is likely to have reduced pressures on forests.' The inverse conclusion can be drawn for countries which can not or not anymore built on steady oil revenues. Van Soest takes the example of Cameroon, where the decrease in oil revenus and the following economic crisis has lead to unemployment in urban areas and to a population shift back to rural and forested areas.

Overall, if governance is optimal, rents may be invested in order to induce SFM and induce long term development. If governance is weak, rents may be directed to consumption, which may have negative effects on SFM, depending on the direct threats to the forest surface through population growth and agricultural development. According to Gregory and Ingram (2000) 'the growth in human population over the past century has been closely associated with increased production of food and forest products (Dyson 1996)…Overall a population of about 6 billion is projected to raise to about 8 billion by about 2025 with most of the increase in the less developed countries in Africa and Asia' (Fischer Heiling 1997).

Numerous authors have studied the influence of population growth on deforestation, and many of them find a negative relationship, for instance Allen and Barnes (1985), Burgess (1992), Deacon (1994), Southgate (1994) or Ekbom and Bojö (1999). Specific country studies have also been conducted by Kummer and Sham (1994) for the Philippines, Panayotou and Sungsuwan (1994) for Tahiland, Reis and Guzman (1994) for Brazil and Southgate, Sierra and Brown (1991) for Ecuador, van Soest for Cameroon (1998), among others. Various authors note, however, that the way forests are used depend on several other variables, mostly socio-economic variables , such as per capita revenue, unemployment, agricultural productivity, or access to infrastructure (Van Soest 1998 , Contreras-Hermosilla 2000).

Clearly, population growth is one factor which has an important direct impact on the forest sector and forest product management through both, higher pressure on land and increasing demand for agricultural and forest products (construction timber, fuelwood). As for the influence of population growth on trade, it seems that it might have an important influence on supply and demand for forest products, both concerning the amount of products asked for and the type of products. Combined with urbanization, population growth might lead to the creation of huge market chains in small spatial areas which could render trade more competitive and efficient.

Population growth leads to growing demand. But this does not always lead to unsustainable forest exploitation, depending on the import structure and the plantation policy. First, Wunder (2000) argues 'that demand structure in Gabon indeed changed dramatically, but that this had unimportant impacts on land use because food imports grew spectacularly.' Also, trees outside forests, will increasingly be used to meet the demand of the local population (FOSA 2001):'This is particularly the case of home gardens in the humid zone countries like Rwanda, Burundi, Uganda and several of West African countries'. However, one of the conditions for wider spread plantations of trees outside forests are secure land tenure.

Finally, demand for wildlife is one of the main driving forces of ecotourism, the major forest product service of tropical forest countries. However, the efficiency of the management of protected areas seems still insufficient. According to FOSA, 'problems like encroachment, logging, collection of fuelwood and other products' are not rare despite an increasing number of community-based management schemes. Political instability and unsolved land-tenure rights might be impeding factors. What is more, investment in the sector seems to be very low. According to FOA, 'a study by the WCMC shows that Africa's investment in park management is the lowest in the world'.

4.2.5 Impacts of Regional Development Programmes

Regional development programmes generally target poorer rural regions such as Appalachia in the US, or western China as part of the new Western Development Program (and the associated Natural Forest Protection Program), or entirely undeveloped regions such as those that gained new settlers from Indonesia's transmigration program in the 1990s. These regions tend to be sources of natural forest. Managed forests, where they exist, tend to be in better developed regions closer to the major markets--in the coastal plains and the Piedmont of the US South and Southeast, for example. Therefore, regional development programmes tend to encourage exploitation of the forest frontier, shifting it further into the hinterlands and up the mountainsides. Moreover, to the extent that the forest products from regions benefiting from public development programs substitute for production from other regions, these regional development programs also delay the progress of sustainable forest activities in more developed regions.

Governments may also use tax policies to encourage investment in less-developed regions of the country. If these policies are effective, then employment opportunities may improve and workers may be drawn toward better labour opportunities and away from forested rural areas. The condition of the natural forest at the frontier may even improve as a result. However, the relative position of the lowest wage workers may decline as a result of these policies.⁴⁰ Employment in low wage subsistence agriculture may increase and the reliance of the rural poor on the forest also may increase.

In addition to tax policies, governments often invest directly in infrastructure, including public utilities and public services like education, hospitals, and communication networks. Antle (1983) showed that the full collection of items identified with infrastructure has a significant and positive effect on rural economic development in general. Improved roads and technologies in particular have important effects on the general condition of natural forests. They have a smaller effect on managed forests because the latter occur in regions that are already developed.

In the earliest stage of forest development, new roads and technologies improve the conditions for local development. They increase the value of land in agriculture and in forest relative to distance from the market. This means that it becomes worthwhile to convert some degraded forest into permanent agricultural land. But it also results in degradation and deforestation of additional lands which previously were too far removed to be worth using for extraction of timber.

In the second and third stages of development, improved access and technology makes the region's land more valuable in all uses. This extends the claims of permanent agriculture and shifts the entire degraded open access area further into the geographic interior in stage II. In stage III, it extends agriculture into the area of previously managed forest and extends managed forestry into the area of previously degraded forest. The natural forest frontier shifts further into the interior as the deforested area expands in both the second and third stages of forest development.

Thailand provides a recent example of how road development is linked with deforestation. Thailand built roads into its more sparsely inhabited Northeast in the 1960s, the main objective being to increase security--military access and encouragement for human settlement to secure the region against encroachment from Laos and Cambodia during the Vietnam War. The main effect was an increase in timber harvests. The links between roads and timber harvests is strengthened by the fact that rights to adjacent lands and timber are often part of a government's payment to private road building contractors. This was common in the US in the 19^th century but there are more recent examples. The government of Laos recently made a similar transfer of timber rights for building a highway through its northern forests to provide access to the rapidly developing markets of southern China. This can will further promote the use of imported timber by China in order to reinforce its logging ban.

Roads can also have an indirect effect on the forest through their impact on overall regional development. Some labour will be attracted to employment opportunities out of the region. As the local labour supply declines, local wages must increase. This reduces the returns from forestry activities as labour is a significant component of costs. The boundary of extraction from the remaining natural forest there shifts inward and the areas of degraded forest and deforestation decline.

4.2.6 Impact of General Economic Growth

At any stage of forest development, there are two fundamental means for minimizing the degraded area, reducing the cost of property rights and attracting some human activity away from the forest. This requires finding the best bundle of property rights and the institution that can provide this bundle at least cost.

Agricultural or forest users with lower opportunity costs can afford the time to travel further into the natural forest to extract its products. Because their costs are low, they can also justify removing material in the degraded area down to a low level. Providing these low wage or low opportunity cost agriculture or forest users with improved employment opportunities outside the forest will cause some of them to change from extractive activities in the forest to the higher wage employment.

The combined effects of improved property rights and attracting labour away from the forest results in a contraction of the degraded open access and an increase in the forest density in the remaining open access area. Turning the argument around: poverty is linked to forest degradation and depletion; economic development induces improvements in the forest environment as it shifts land into sustainable activities. One of the more thorough bodies of evidence of these short term effects has been a study on the impacts of oil wealth on the fate of the forest, drawing on case studies from Cameroon, Ecuador, Gabon, Indonesia, Mexico, Nigeria, Papua New Guinea and Venezuela (Wunder, 2003).

India's most productive agricultural region, the Punjab, provides another specific example. The region began a period of rapid and sustained development in 1960. Crop yields per hectare tripled by 1990 and income per capita doubled (in constant dollars). The land area in agricultural crops more than doubled while the principal agricultural prices remained relatively constant or declined--depending on the crop. Meanwhile, the rural share of the region's population remained steady at approximately 22 percent. Forest cover in the Punjab increased six-fold and horticultural tree cover increased more than 250 percent. A large area had been cleared of its forest cover and existed only as an open access wasteland before 1960. The open access lands have declined and the forest stock has increased. A large share of the open access lands has been converted into cropland since then, while an additional large share has been reforested (Singh, 1994).

The longer term effects of economic development on the forest is unclear. Improved wages and better labour opportunities may lead to institutions with improved ability to insure property rights and manage economic transitions and provide for economic stability. Conversely there are plenty of examples in which an abundance of riches (whether based on forest wealth or not) during economic development have led to spiralling inequity, corruption and capital flight (Stevens, 2003). Such trends lead ultimately to a resurgence in poverty and land degradation.

In sum, rural economic development is central to any program of improved forest sustainability and any attempt to decrease the rate of global deforestation. Accomplishing it is not an easy task, but it is certainly no more difficult than trying to accomplish sustainable forestry and slowing deforestation through the imposition of government regulations on the use of relatively low-valued and dispersed resources by a scattered and poor rural population.

4.2.7 Conclusions

Assessing the impact of sectoral and extra-sectoral factors on trade and sustainable forest management is challenging for several reasons: first, all the 'underlying causes' of forest decline are characterized by a high degree of interrelationship. Indeed, according to Contreras-Hermosilla (2000) various explicative factors may form 'a complex socio-economic, cultural and political event' which implies that 'a single force, such as agricultural intensification, may operate in diametrically opposite ways.' (Contreras-Hermosilla 2000).

Second, as with other definitions on sustainability, the main challenge with the concept of SFM is its application and the treatment of conflicting objectives. Sustainable forest management ' refers to meeting present needs for forest goods and services, while ensuring their continued availability in the long term.' (FAO SOFO 2003).

Extra sectoral influences with a positive impact on SFM include land tenure security, the achieved demographic transition, a low HIV/AIDS prevalence ratio, institutional stability and good governance.

Extra-sectoral influences with a negative impact (in addition to the opposite of extra-sectoral factors above) include agricultural intensification through mechanization in agricultural frontiers (forest margins), directed settlements in forest areas, improved access to credits for beef cattle, mechanized agriculture, and large-scale forest and tree crop plantations in areas with substantial natural forests and energy and mining projects in forested areas.

The main ways of increasing the quantity of managed forest are: to increase the value of forests compared to other land uses and to reduce the cost of property rights. Unfortunately, attempts to increase the value of forests in order to promote forest management may also lead to increasing degradation of open access areas and the natural forests at the forest frontier. The two main ways of reducing forest degradation in natural forests are: to attract people out of the forests and to reduce the cost of property rights. The common ground lies with the reduction of costs and risks associated with property rights. Linking trade liberalisation to improved and impartial property rights and the institutions responsible for them (either nationally or through well designed decentralisation and community management programmes) would be one way forward. Even then, enforcement will never be perfect and some illegal logging will occur because the costs of illegal harvesting at the frontier (even with enforcement factored in) are much less than the costs of growing and harvesting managed stands. That is, the incentives for illegal harvests are great.

Attracting people out of natural forests through general development is another promising option. Unfortunately the most promising initiatives (such as effective urban and industrial development programmes or the massive effect predicted for a reduction in agricultural subsidies for example in Europe and America) may be beyond the scope of forest agencies to influence. Incentive programmes, especially for local input subsidies or technology developments in agriculture (leading to land intensification), or tax reductions for capital gains tax (leading to investment in managed forest) are perhaps more realistic at that level of operation. Making the case for more effective extra-sectoral change will be an important priority for the future.

4.3 Policy Instruments and Processes Affecting Trade in Forest Products and Services

4.3.1 Introducing the Range of Policy Instruments and Processes

This chapter gives an overview of the various types of policy and process employed in the forest sector at local, national and international level to influence forest management and considers their implications for trade patterns. The focus is on policies that aim primarily to affect forest resource management and the conditions under which timber is harvested and forest products processed. Through their impact on raw material availability and the competitiveness of the forest products industry in different locations, these have implications for trade. Forest sector trade policies that are primarily designed to restrict trade or change trade patterns are reviewed in the next chapter.

Policies can be classified in several ways. In this section we group policies according to the stakeholder group most closely associated with their implementation. Most policy measures are primarily associated with national governments - in sections 5.2 we deal with these national public policies.

Inter-governmental processes are also of relevance because of the global interest in the social and environmental services provided by forests, notably the UN Intergovernmental Panel on Forests, which was replaced in 1998 by the Intergovernmental Forum on Forests and then in 2000 by the UN Forum on Forests.

In addition, given the perceived failings or inadequacies of government policy, a range of policy initiatives of other stakeholders such as the private sector and NGOs are of increasing impact - these are outlined in section 5.4 and 5.5. These efforts are of particular note on the 'demand side' where several are focused on trying to improve the sustainability of consumption of forest products. Several voluntary 'supply side', production-oriented initiatives are also of note.

4.3.2 Domestic Forest Policies and Processes

Some countries have very strong, implemented national forest programmes (in various shapes and forms) - with a range of effects on trade - others exist on paper only, or not at all. Where they have some coherence and clout (as opposed to being the sum of often diverse and sometimes incoherent policies, laws and other official pronouncements), national forest strategies or programmes set a framework for forest governance, forest management and the future of the forest sector. This in effect aims to alter the market-induced pattern of forest development described in Section 4.1 to one which meets broader social and environmental objectives.

Where national forest programmes have some strength, one of their major effects on trade is through the designation of forests as production forests, for conversion to other land uses, or for protection only as this affects the balance between raw materials, production capacity and local demand. Such designation can, in turn, affect forest management in other countries, for example in the case of logging bans - see below. The effectiveness of these strategies in altering the pattern of forest development will depend on how much governments are willing to absorb the costs of protection and to uphold property rights in production forest areas. Evidence of their lack of willingness or financial capacity to do so in many countries is given by the forest parklands around the world, which suffer encroachment, and the timber reserves around the world, which experience illegal logging and illegal export trade.

This reflects process shortcomings also. In most cases national forest strategies have been internationally driven and heavily dependent on donor funding. The imperative to develop, implement, monitor and evaluate such national forestry programmes was spelled out in the proposals of the Intergovernmental Panel on Forests (IPF) and its successor the Intergovernmental Forum on Forests (IFF). If NFPs are to succeed they need to avoid the mistakes of previous internationally driven calls for forest sector plans. Many countries developed National Forestry Action Plans (NFAPs) from the mid 1980s onwards - following a model that emerged from the global Tropical Forestry Action Plan of that time. But many NFAPs remained exercises on paper only lasting only as long as donors propped them up - they failed to catalyse the detailed actions expected of them. In general, this was because they were done quickly, often by foreign experts, and failed to engage with political and economic reality to show not only what needs to change, but also how it can change, and how such change can be sustained. Many one-off institutional reform approaches stimulated by such plans have left legacies of huge and unsustainable recurrent transaction costs.

Forest Policies for an Enabling Environment

Over the last fifteen years, a number of Asian and Pacific countries - China, New Zealand, Philippines, Sri Lanka, Thailand and Vietnam - have completely or partially banned logging in natural forests. The aim has been to conserve natural forests and obtain a larger portion of wood supply from tree plantations and agroforestry. Results have been mixed (Brown et al 2002). On the positive side, New Zealand managed to replace the natural forest supplies with those from thriving forest plantations which now produce large volumes of timber for export. In Sri Lanka, by 1993 home gardens and coconut and rubber plantations were supplying 70% of all industrial roundwood. However, these bans can be difficult and costly to enforce and by raising the price of timber locally, provide incentives for illegal logging and trade. This is a problem in Sri Lanka where perhaps as much as one quarter of all wood delivers to its mill are the result of illegal logging (Hyde 2003).

Another drawback is that uncertainty about future harvest restrictions will induce producers to act to protect their investments. This often involves pre-emptive harvesting, before the policy can become official. The effect will be to reduce domestic timber prices and increase exports based on non-sustainable production. India restricted all harvests of its high-valued but declining sandalwood in an effort to preserve the remaining resource but some landowners responded by harvesting immediately before the restriction became effective. Harvesting has also continued on an illegal basis since the restriction came into force with the result that the standing sandalwood inventory has declined (Hyde, forthcoming).

Such bans to be effective require commitment of government resources. China banned logging in 42 million hectares of forest in 1998 and employed special police to enforce the policy. Government compensation programmes provided assistance to large numbers of workers that lost their jobs in the process. Officials expected timber harvests from natural forest to decline from 32 million cubic meters in 1997 to 12 million cubic meters in 2003. They also expected that in the medium-run some 34 million hectares of tree plantations will make up most of the difference. It is still too soon to know whether they will succeed.

In some countries, the transition to alternative domestic sources of timber has not been smooth, reflecting the fact that comparative advantage in harvesting natural forests does not always translate into comparative advantage in domestic plantations. Thus the Philippines and Thailand which imposed logging bans after commercial timber resources had been largely depleted are now significant importers of timber (Durst et al. eds. 2001). Thailand was more successful than the Philippines at curbing illegal logging. Yet, forest clearing for agriculture continues in both countries and logging bans cannot solve that problem. Plantations still provide only a small portion of the two countries' wood supply.

Where logging bans have been effectively enforced they have had significant implications for trade flows. China, the Philippines, Thailand, and Vietnam all greatly increased their forest product imports after they restricted logging. Some consider such harvesting controls to be formal trade barriers and as such to come under the WTO remit but others disagree, considering that they are not trade-related (Bourke 2002). There has been perhaps more concern over the implications for forest management in neighbouring countries. Restricting logging in one country may simply displace the problem to other countries. The bans in the countries mentioned above fuelled illegal logging and destructive timber harvesting in neighbouring countries such as Cambodia, Indonesia, Laos, Myanmar, and Russia.

Forest Resource Tenure

Property rights over the forest resource have a major determining influence on the resource base available for international trade and on the extent to which different groups benefit from such trade. Globally, governments claim to own and administer 77% of all forests. This includes large areas of forest that local communities manage without official recognition. Communities and indigenous people formally own 7% of the forests and officially administer an additional 4% that governments have reserved for them. Individual landowners and private companies own the remaining 12% (White and Martin 2002).

In many countries worldwide, national governments are increasingly seeking to reduce their own direct involvement in forest management. Motivations for this vary from one country to another. Some are common - a drive for greater efficiency and profitability, reduced pressure on the public purse particularly as governments are finding that formal rights to forests are not easily enforced. Others are unique to specific circumstances - such as empowering previously disadvantaged groups.

One route is to give official recognition to community management. In tropical forest areas, indigenous and other communities now own or administer about 25% of the forests and at current rates this could double in 15 years (White and Martin 2002). Terms such as “community forestry”, “joint forest management”, “public participation', and “devolution”; all refer to transfers to local users of some or all of the rights to forests that were previously the unambiguous responsibility of central forest ministries.

Where effective, the shift to local management reduces the cost of maintaining property rights as local users of the land and forest resources know the resources and the demands on them better than the officials of the forest ministry. This means that more forest land will be managed sustainably and the area of degraded open access forest will decline. Local management can improve long-term land management for agriculture, timber and other extractive products of the forest, and also for local non-market values like some erosion control and some recreational forest use. The list of successful examples of community management is almost endless and it comes from all corners of the world (Ostrom ?).

However, local property rights and local management do have their limits. Transfers of rights to local communities are less successful in halting forest degradation where the local values for forest products are very low relative to the community values for other land uses or for their time. This is the general case in the first stage of forest development where forest resources are plentiful but also applies to open access forests in the second and third stages of forest development. They are also less effective where important local groups have competing demands on the forest (Dangi and Hyde, 2000); and where local property rights are incomplete or the policy environment is uncertain (Yin and Newman, 1998); or where values at stake are shared by the broader regional, national, or global community (e.g. Carbon sequestration, biodiversity, and some classes of tourism).

Another route being pursued is privatisation - tenure and use rights over state forest assets being transferred into private hands, and/or the outsourcing of forestry functions. Such transfers into private hands are likely to increase international trade in forest products and foreign direct investment in the forest products sector. Whether this will be good or bad for sustainable forest management is debatable and depends heavily on the way that the privatisation process is carried out. Perceptions abound that private companies are not accountable to public demands and have no incentive to provide environmental and social goods and services. The extent to which trade affects pushes privatisation in a positive or negative direction in these terms depends on its effect on key challenges in forest privatisation such as:

• Developing clarity on the transaction costs and risks

• Designing tender systems for negotiated and optimised objectives

• Supporting preparedness in private sector and community organisations for negotiating and implementing ownership and management changes

• Managing post-transfer government responsibilities

• Dealing with the social impacts of change (Garforth et al. 2002).

Forest Decentralisation

Another route for national governments to reduce their direct forest management involvement is through decentralisation. At least 60 countries have recently decentralised some aspects of how they manage their natural resources, with mixed results. Examples can be found of reforms which have permitted disadvantaged groups to have more input into decisions about forests, engage in trade, provide more revenue to local governments, and improve the way people manage their forests. But examples can also be found of the opposite. Central governments tend to hand over burdensome tasks and low value resources, but keep the attractive activities and resources for themselves. Many local governments, on the other hand, do not really represent their local constituencies, and are inclined to over-exploit their natural resources. It has been argued that most failures are due to central governments not having decentralised enough - not making local governments truly democratic and not yet giving them real power over major decisions (Ribot 2002).

The locus and clarity of decision-making power over forest production following political-administrative decentralisation will affect the costs-benefit balance of the resource base for trade and the distribution of its benefits. The imperative of improving the local returns from trade is an explicit motivation for decentralised decision-making in some contexts.

Forest Resource Allocation Policies

Policies to allocate and charge for state-owned forest resources are believed to be a prime factor in determining international competitiveness of forest product companies, at least in the short term. Concessions that are administratively allocated instead of through an auctioning process and low royalties reduce costs for forest companies increasing their potential for international competitiveness. The long-standing dispute between the US and Canada over trade in softwood lumber stemmed from the view that companies in Canada were not being charged sufficiently for access to forest concessions (Bourke 2003).

For developing countries, there has been widespread concern that governments have not been capturing sufficient rents from their forest concessions, particularly for tropical timber. Yet there has not been retaliatory action on the same scale as the US Canada dispute, that is, imposition of countervailing duties, because these countries are not generally competing with sources of timber from developed countries.

The use of royalties and indirect charges such as reforestation levies has increased as well as their levels. NGOs and policy advisors have generally favoured increasing the charges made for the forest resource under the assumption that this would create incentives to use it more efficiently. More efficient use of the resource would mean that fewer logs would be needed to produce a given amount of product, thus reducing pressure on natural forests. Recently, however, this has been questioned. For example, in Indonesia Barr points out that making large diameter logs of commercial species more expensive encourages loggers and processors to adopt technologies that allow processors to use smaller diameter logs and non-conventional species. That can threaten large areas of forests that timber companies had previously considered of marginal value (Barr 2000a). He uses the recent adoption of small-spindle rotaries by Indonesian plywood producers as a case in point.

There is also the view that increasing royalties at the same time increases the incentives for illegal logging. Ultimately, the success of any revenue system depends on government's ability to enforce it (Hyde et al 1991).

Regulations on Forest Management

Various types of regulatory control over forest management affect costs of harvesting for example through the stipulations on the rate at which forest resources can be exploited, the techniques that must be employed and the procedures that must be followed, e.g. formulation of forest management plans. Restricting harvesting of certain types of forest, e.g. riparian forests and placing limits on forest conversion can have a significant effect on costs. Thus in Brazil, landowners in the Amazon are not permitted to clear more than 20% of their holdings and must maintain forest cover on the remaining 80% (Viana et al. 2002). Other regulations on harvesting practices such as diameter limits on felled trees, stipulations on distance between felled trees and zero disturbance within the habitat of an endangered species also affect production costs and competitiveness. This is particularly the case where such regulations are unevenly enforced or vary in stringency between countries. Products produced illegally, i.e. violating forest management regulations, thus have a competitive advantage, with implications for trade patterns. In some instance regulations may be traded off against other public goods (such as the installation and maintenance of rural infrastructure (Macqueen, 2001). In other cases they prove impossible to enforce, resulting in unchecked social and environmental damage as witnessed in Cambodia, Cameroon, Gabon, Guyana, Malaysia, Papua New Guinea among others (Forests Monitor 1998; Global Forest Watch 2000a, 2000b; Filer with Sekhran 1998; Madeley 1999; Global Witness 2001; Macqueen 2001).

A recent study from Indonesia provides evidence that the only way companies can profitably log the large and rapidly increasing area of forests that are in their second rotation is if they do it illegally and / or in a non-sustainable fashion. This also applies to many unlogged forests with low value timber. SFM for commercial timber production of such forests is often not economically attractive. That is one reason why illegal and non-sustainable logging practices are so widespread. Under these circumstances no changes in concession duration or forest regulations can convince companies to sustainably manage their forests. To do so they would have to operate at a loss (Barr 2000a).

Where regulations are enforced, they tend to add to the timber harvest costs, at least in the short term. This would lead to a decline in harvests at the forest frontier, and most would consider this an environmental improvement. The net effect may be substantial. For example, it was estimated that silvicultural prescriptions alone add an average of five to eighteen percent to the costs of forestry in the US South, British Columbia, and Finland - although the impacts on individual landowners vary with local conditions of land quality and enforcement (Sedjo 1999).

Of importance for the impact on trade is the fact that consumer demand is unaffected by these regulations. Therefore, the significant decreases in production that occur in the US South, British Columbia, and Finland, for example, will be largely compensated by increases in production from other parts of the US, inland Canada, and Russian Karelia, respectively, as well as with additional imports from developing countries. In each of these cases, the production shifts are largely from managed forests of regions in the third stage of forest development to the natural forest frontiers of regions in the second stage of forest development. Increased regulation of managed forests in one region can lead to increased exports from non-sustainable or illegal sources in another.

Forest Taxation

In some countries, income derived from capital gains is taxed at a lower rate than other income. Since most timber is held for long periods (appreciating over time) a lower tax rate for capital gains favours investments in managed forests in preference to activities like agriculture whose production periods are shorter. The UK's and Chile's tax codes which exclude inherited forests from death taxes induce similar shifts, thereby increasing the total land area in managed forests. The effect is not trivial. One estimate for the US suggests that the favourable treatment of capital gains provided twenty percent of the forest industry's after-tax profits in 1984 (Russakoff, 1985). Since additional managed production may substitute for some production from the natural forest frontier the favourable capital gains treatment may have a conserving effect on the natural forest. In terms of trade, the lower production costs will favour exports over imports.

In contrast, property taxes can accelerate the harvesting of timber and the degradation or conversion of forests. Standing timber and various other assets, as well as land, are often also subject to the property tax and their taxation introduces biases into resource allocation. The final accumulation of annual property taxes paid on the timber by the time of harvest is much greater than the accumulation of taxes would be if timber production were an annual agricultural activity and each period's growth were only taxed once. This time bias against forest management and encourages landowners to harvest their timber at an earlier age in order to avoid some of the repeated and accumulating taxation.⁴¹ This affects decision-making at the extensive margin between managed forestry and open access degraded forest and makes land there unprofitable for managed forestry. The combined effects of the time bias on managed forests and the conversion of the extensive margin into unprofitable forestland were a major reason that many firms in the southern US forest industry “cut and got out” in the 1920s and moved to the American West. The effect of the property tax on timber management in places like the South and the Lake States of the US caused policy makers to introduce yield taxes as an alternative. Yield taxes are assessed on timber value only once at the time of harvest.

Taxation systems through their effect on the returns to forestry affect competitiveness relative to producers in other countries and thus can potentially impact on trade patterns and incentives for inward investment. Their impact on trade patterns is not as evident as for other more direct types of financial incentive.

Policies on Forest Resource Development

Governments have typically used a number of means such as subsidies and tax concessions to develop forest resources. For example, the development of export plantation sectors in Chile and Brazil owe much to generous government incentive programmes (Hyde 2003; Viana et al. 2002). Between 1974 and 1994, the Chilean Government spent some US$50 million on afforestation grants. Even so there are some who claim that these subsidies were unnecessary as plantations would be have been profitable in Chile without this financial assistance (Cossalter and Pye Smith 2003).

Assistance takes many forms in forestry, but often involves reducing production costs through planting subsidies, free seedlings, and tax concessions, assisted transport, provision of extension services, the provision of infrastructure such as roads.

Financial incentives are more common in developed countries (e.g. some Canadian provinces, the UK, and the Nordic countries and the US (Boyd and Hyde 1989). Since incentives are usually linked to forest management, their impacts are largely restricted to regions in the third stage of forest development. Incentives decrease the private management costs and, therefore, increase forest land value. Land at both the intensive and extensive margins of forest activity shifts away from competitive uses and into forest management and total production from the managed forest also increases. The increased production in managed forests may also substitute for harvests from the mature natural forest frontier.

Distributive arguments are sometimes used to justify incentive programs on the grounds that they benefit small-holders, thus promoting rural livelihoods. However, ensuring that it this group that benefits in practice is often challenging. In the US (Boyd and Hyde, 1989) it was observed that those who do take advantage of the investment program are not among either the poorer or the smaller private landowners.

These forms of assistance can substantially improve the competitiveness of domestic producers in both their own domestic market and export markets, and are often used specifically for these purposes. For this reason they can act as trade barriers and can be of far greater importance than tariff barriers. Such forms of assistance occur in many countries though it is difficult to assess the extent to which they create barriers to trade, and whether they have increased in recent years or not. There is evidence however, of a substantial differential between subsidy rates in developed and developing countries. In eleven EU countries the average subsidy for plantation schemes is US$1,421 per hectare plus US$761 per hectare for maintenance. In South America, subsidies are less than US$ 400 per hectare (Cossalter and Pye Smith 2003). While some reductions in subsidies are apparent, there are signs that they may be increasing in countries interested in encouraging further growth in plantations (Bourke 2002). Ecuador and Colombia have adopted a similar incentives model to that of Chile, with Ecuador providing planting and maintenance incentives of US$300 per hectare (Cossalter and Pye Smith 2003).

Policies on Downstream Processing of Forest Products

A number of policy instruments are commonly used by governments to promote wood processing to increase the potential for export of value-added products or to protect the local industry from international competition. Assistance can be both financial, for example graded tax exemptions for different levels of processing and non-financial through the provision of support services. Trade policy, log export bans specifically, are often used to promote or protect a domestic processing industry. In particular, government sponsored research in wood processing can increase the efficiency of wood utilisation in industrial processing facilities, thus increasing competitiveness and the potential for trade. US government sponsored research for the southern pine plywood industry produced rates of return of 300% per annum throughout the 1960s and 1970s. Cost reductions brought about by research were a critical factor in the expansion of the plywood industry in the southern US. While this increased the demand for southern pine logs, it displaced western pine plywood which made less efficient use of the logs and was more likely to originate from mature natural forests. The effect of this government sponsored research was both to increase competitiveness of the southern pine plywood industry and to decrease US harvests at the natural forest frontier (Hyde, forthcoming).

Government Policies to Promote Forest Environmental Services

Many governments are introducing policies to promote the provision of forest environmental services such as carbon sequestration, watershed protection, biodiversity conservation and landscape beauty. Particular attention is being given to payment and market initiatives in an attempt to tap private sources of finance. At present it appears that provision of carbon sequestration services has the greatest potential to affect international trade patterns for timber.

In the case of carbon these efforts have been driven primarily by inter-governmental agreement on the Kyoto Protocol to the United Nations Framework Convention on Climate Change, which was signed in 1997. This established explicit and mandatory limits on industrialised and transitional nations' greenhouse gas emissions with targets for reductions by 2008-2012. Average required emission reductions for these countries as a group come to 5.2% below 1990 levels by 2008-2012 (Landell-Mills and Porras 2002). The Kyoto Protocol also provides a framework for trading emission rights such that countries wishing to emit more than their limit can purchase rights from others that find it less costly to reduce their emissions. There are three types of trading mechanism set out in the protocol:

• International Emission Trading that allows Annex B countries to trade emission permits;

• Joint Implementation that allows Annex B countries to earn Emission Reduction Units through projects in other Annex B countries;

• Clean Development Mechanism (CDM) which allows Annex B countries to gain certified emission reductions from projects in non Annex B countries ie developing countries.

These reductions may be achieved by reducing emissions or by increasing carbon sequestration and storage. Forests can play a key role in the creation of carbon offsets through four approaches:

• Reforestation/afforestation to increase carbon sequestration;

• Improved forest management eg through reduced impact logging to both increase sequestration and reduce emissions;

• Conservation and protection against deforestation to reduce emissions;

• Substitution of sustainably produced biomass for fossil fuels to reduce emissions.

Timber harvesting activities are also considered as a source of carbon emissions.

Decisions taken in multilateral negotiations (COP 6 and COP 7) on the rules underlying these flexibility mechanisms have placed limits on the extent to which carbon offsets can be generated from forestry. Improved forest management, for example,is not an option for the CDM which is restricted to reforestation and afforestation. Credits from forestry and other land-based sinks will be capped at 1% of country's base year emissions. This has significant implications for the extent of international trade in carbon offsets and the knock-on effects on trade in forest products.

In developed countries (Annex I countries) no credit is given for carbon sequestration from forest cover established prior to 1990 but harvesting from both types of forest ie established before or after 1990 will be counted as an emission activity during the first commitment period 2008-2012. This gives an incentive for increasing the harvesting rate from non Kyoto eligible forests before 2008 (Leitch 2002). This will affect trade patterns in the short term.

In order for the protocol to become effective it must be ratified by at least 55 countries, representing 55% of 1990 carbon emissions. At present the number of countries requirement has been met but not the percentage of emissions requirement. Ratification by a significant source of emissions such as Russia is needed before the protocol can come into force. Nevertheless, many carbon offset projects based on forestry have been developed in anticipation of ratification.

The traditional government role in promoting market-based systems is to establish a conducive policy framework and regulatory environment and facilitate or promote action by the private sector. Australia provides an example of this. In 2000, the Australian government launched a “Bush for Greenhouse” campaign to encourage industry to enter into deals with private landowners for revegetation projects to sequester carbon. Carbon offsets generated under this scheme will be recognised by the Australian Greenhouse Office. However, some governments have also become involved as active buyers and sellers of environmental services and are frequently active intermediaries. A global review of markets for environmental services found that of the 75 carbon schemes examined, governments constituted 16% of buyers, 23% of sellers and 17% of intermediaries (Landell-Mills and Porras 2002).

In some cases these interventions are temporary with the aim of catalysing participation from other stakeholders. This applies particularly to the provision of advisory services, training and information. In other cases the interventions are more long-term. In Costa Rica, the government has established an intermediary institution, FONAFIFO, to bring buyers and sellers together and to guarantee service provision. In Australia, State Forests New South Wales, has been offering immediate sales of, as well as future options to purchase, certified and guaranteed carbon offsets. Deals have been negotiated directly with large international power companies but future plans include the use of specialised brokers and exchange-based trading.

While the policies discussed above will primarily impact on international and domestic trade in environmental services, there may be knock-on effects on trade in forest products. The additional revenue generated by the sale of environmental services will make certain types of producer viable in the international market, while others will be displaced.

There are also implications for forest management. The additional revenues from the marketing of environmental services may be sufficient in some cases to make sustainable forest management competitive with other forest land uses. But the problem confronting any attempt to use trees to control global climate change is the mismatch between the non-exclusive global public impact and the security of specific forest activities required to control it. Only for regions in the third mature stage of forest development, are the costs of establishing and managing forests compensated by the market value of the resource. Payments for carbon sequestration under the CDM might increase the forest value function and extend the area of sustainable management. As payments for conservation of mature forests are not contemplated within the CDM it is clear that CDM payments will therefore increase intensively managed forest at the expense of forest activities at the forest frontier. The ultimate result may be increased forest cover. But concentration on the provision of carbon in managed forests may work against some aspects of sustainable forest management such as biodiversity conservation. Moreover, displaced producers at the forest frontier may also revert to subsistence agriculture with negative implications for the forest. Alternative solutions might be found in payments for institutions involved in securing property rights.

In terms of biodiversity, it is reasonable to assume that the remaining natural forest, the forest beyond the frontier contains most of the critical habitat. The difference between the conservation of carbon and species lies in the specificity of geographical areas which need to be conserved. Biodiversity requires the protection of selective “islands” of specialized habitat generally at the natural forest frontier or beyond it. The foregone future development opportunities associated with them are often smaller than even the current opportunities foregone when protecting forests for carbon sequestration at the margins of economic activity. This means that many critical habitats remain unthreatened today simply because they are (for now) beyond the limits of economic activities.

Since the prevention of land conversion will be a problem for habitats at new frontiers in regions in the first stage of forest development it may be better to target a reduction in incentives for agricultural expansion in newly settled regions like some parts of the Amazon or the interior of Papua New Guinea. Otherwise, for regions in the second and third stages of forest development the critical issue remains the identification of the most threatened areas. Once these are known the most effective strategy might be the design of roads to avoid critical habitats or long-term investment in the regional institutions involved in securing property rights. In some specific cases, habitats might be able to be protected in the normal course of management for other forest products and forest-based environmental services. Hyde (Hyde, 1991) showed that this is the case, for example, for many red-cockaded woodpecker habitats in the southern pine region of the US. Careful planning of the sequence of timber management activities, with no change in the overall activities themselves, was sufficient to protect this endangered species.

A broad range of people participate in the benefits of environmental tourism and a broad range of forested sites provide for it. The site characteristics range from unique global resources like Yellowstone, the Serengeti, or Sagarmantha to pleasant local forested groves and even village parks. The unique sites are often focal points for substantial demands for tourist support services like restaurants, motels, guide services, and outdoor equipment shops; and these can be important sources of employment for the local economy.

The economic problem is either one of protecting “islands” of specialized forested sites within lands that are valuable for other, extractive uses of the forest or identifying forestlands that are inaccessible for extractive land uses but are uniquely attractive for nature tourism, and protecting them before they become accessible for those extractive uses.

For the most unusual sites, fees can be charged at points of limited access and the collected revenues can be used to establish boundaries and to monitor and enforce the exclusion of undesirable uses of the park. For example, Kenya charges an increased visa fee for the entry for global tourists, Nepal charges for trekking permits which must be obtained at the offices of guide services, and many national parks with natural boundaries charge a gate fee. It is of course necessary to define clearly how such fees can be reallocated in the case of multiple sites of interest.

It is also necessary to monitor and enforce restrictions against the competitive use of land by local people (who are not able to be excluded upon failure to pay such fees). A partial solution to this problem can often be obtained by establishing an interest in the park's tourist services within the local population. Forest resources that are not unique and do not attract global tourists are particularly suited to management by local institutions and we observe many successfully protected village parks and forest sanctuaries around the world. Nevertheless, there are still positive costs and, since the exclusion of local users is difficult, the local community must bear these costs either as part of its community budget or as part of a commonly respected decision not to exploit the extractive resources within the park.

Erosion control and watershed protection incorporate all the services of trees in managing wind, water, and soil movement - either through new planting or conservation of existing forest ecosystems

Watershed values are primarily local. Depending on the watershed management activity, it can be of greater benefit to an individual land owner who makes the conservation investment and improves his or her own land's productivity, or it can yield greater benefit to a range of downstream or other off-site land managers in the same watershed

Many watershed management activities require new planting investments as a means of improving the productivity of existing (often agricultural) land uses. These investments typically occur on the private lands. They have little impact on forestry and increases in the manager's private long-run productivity are generally sufficient to induce the private conservation investment and few additional incentives are necessary (Crosson 1985, Crosson and Stout 1984, Alemu 1999 and Yin 2000) demonstrated its reliability for developing countries as well, once farmers in those countries obtained longer-term land use rights.⁴²

The second class of watershed management activities protects the upland watershed or the coastal wetlands for the benefit of off-site residents of the same local area. Grazing livestock in Africa and the Philippines (Cruz et al., 1988) or upland collection of fuelwood and fodder in Nepal (Dangi and Hyde, 2001) are examples of activities which are targeted. These examples all characteristically occur within either the open access degraded forest or the neighborhood of the mature natural forest frontier in the second and third stages of forest development. The costs of protecting watersheds in these areas exceed open access values and private management will be unsuccessful and only regulation or public ownership along with a degree of monitoring and enforcement can insure the common watershed benefits for the local community.

Finally, some cases require broader regional or national oversight. For example, the Chinese authorities feel that this was the case with the flooding of the Yellow and Songhua Rivers in 1998. Upstream deforestation and construction damaged agricultural lands more than 1500 kilometres downstream.

4.3.3 Environmental Assessment of Trade in the Context of SFM

Starting in the early 1990's, the importance of assessing the relationships between trade and environment has gained increased recognition, both in governmental and civil society sectors. This recognition was initially linked to the increased concern about the state of the global environment, following the UNCED in 1992. Principle 17 of the Rio Declaration says that environmental impact assessments will be undertaken for activities likely to have significant impacts on the environment. The Ministerial Council of the Organization for Economic Development (OECD) recommended, in June 1993, to governments “to examine or review trade and environmental policies and agreements with potentially significant effects on the other policies area early in their development to assess the implications for the other policy area and to identify alternative policy options for addressing concerns”. In 1994, the OECD developed general methodologies for conducting environmental reviews of trade policies and agreement and trade reviews of environmental policies and agreements (OECD 1994). In the same years, the North American Agreement on Environmental Cooperation, directed the Commission for Environmental Cooperation (CEC) to consider on an ongoing basis the environmental effects of NAFTA. In 1995, CEC started designing a framework to assist in anticipating important environmental impacts in the context of trade liberalization, and to develop policy tools to mitigate negative impacts and maximise positive ones (Commission for Environmental Cooperation 1999).

Since then, initiatives in this field have increased in number, and involved many stakeholders, led by international organizations, national governments, different sectors of the civil society, research institutions.⁴³

The interests and expectations of different stakeholders in the trade-environment-development nexus have been paralleled by improvement in methodologies and capacities and growing practice of impact assessment studies, including environmental impact assessment. Most countries have now specific requirements for conducting environmental impact assessment of project and activities.

In 2002, the FAO Forestry Department has started renewing its interest in the area of environmental assessment of forest use, including trade in forest products and services⁴⁴ ()

This paper presents the results of a desk study which has reviewed how environmental assessments of trade and trade-related issues have been applied in the forestry sector .

Definitions

The definitions given below have the aim to clarify some of the terms most currently used in the work and research dealing with assessing and evaluating the interactions between economic activities in general (including trade in forest products) and environmental changes. These definitions are not meant to provide any judgement of the value or appropriateness of each of them.

Environmental assessment (EA) can be defined as the general process of assessing environmental impacts associated with human activities. It may include studies ranging from comprehensive (EIA) to more limited reviews (such as environmental audits, etc).

Environmental impact assessment can be defined as a tool to identify and assess the potential impacts of a proposed project (or activity), evaluate alternatives, and formulate appropriate mitigation management and monitoring measures.

Strategic Environmental Assessment (SEA): can be defined as a tool that promotes the incorporation of environmental considerations “upstream” from a project specific EA into policy and programme formulation⁴⁵.

Sustainability assessments are tools for integrating environmental and developmental considerations into trade and investment policies. By involving both government experts and non-governmental stakeholders, sustainability assessments help determine how to maximise the positive effects and mitigate/avoid the adverse impacts of trade and investment policies. Sustainability assessments are more than just “environmental impact assessment of trade”. They should shape policies, put sustainability first, effectively involve stakeholders, change real outcomes⁴⁶.

Sustainability Impact Assessment (SIA). A SIA is a process undertaken during a trade negotiation which seeks to identify economic, social and environmental impacts of the trade agreement. A SIA should help to integrate sustainability into trade policy by informing negotiators of the possible social, environmental and economic consequences of a trade agreement. SIAs should also provide guidelines for the design of possible flanking measures, the sphere of activity of which can exceed the commercial field (internal policy, capacity building, international regulation), and which will make it possible to maximise the positive impacts and to reduce any negatives impacts of the trade negotiations in question. ⁴⁷

Social Impact Assessment (SIA). Social impact assessment includes the processes of analysing, monitoring and managing the intended and unintended social consequences, both positive and negative, of planned interventions (policies, programmes, plans, projects) and any social change processes involved by those interventions. The primary purpose of a SIA is to bring about a more sustainable and equitable biophysical and human environment.⁴⁸

International Initiatives, Programmes, Approaches and Institutions related to Trade

This article is more concerned with the environmental aspects of the assessment of trade rather than the more encompassing sustainability aspects. It is recognised, though, that there has been a trend, since the first reviews were done, toward more comprehensive assessments.

The main international, national and NGO initiatives related to the assessment of the economic, environmental and social impacts of trade are mentioned below, roughly sorted chronologically.

• Organization for Economic Cooperation (OECD) -Environmental and Trade Reviews

• Commission for Environmental Cooperation (CEC) Analytic Framework for Assessing the Environmental Effects of the North American Free Trade Agreement (NAFTA)

• United States, Canada and Norway national initiatives

• UNEP- Integrated Assessment of Trade and Trade Related Policies

• WWF - “Sustainability Assessment of Trade” Project

• European Commission - Sustainability Impact Assessment of Proposed WTO Negotiations

• Secretariat of the Convention on Biological Diversity - Impacts of trade liberalization on agricultural biological diversity

The main objectives of these trade assessments are to inform the process of decision making of trade policy formulation and the negotiation position of countries. Since the OECD guidelines developed in 1994, these various initiatives have progressively built upon each other and there is a fair degree of convergence in the approaches and methods. However, differences exist, and some of them are introduced in the next paragraphs.

The first aspect, already mentioned is whether these initiatives are environmental assessments only (OECD, CEC) or have the broader approach of the sustainability (EU and WWF) and integrated assessment (UNEP). Recent developments show that there is a trend toward these broader assessments. Another point is that all the initiatives above have a “trade first” vs “sustainability first” approach, that is, they start with the economic changes that are affected by trade policies and then try to link these to environmental effect. (The environment first-approach would start from the environmental perspective and then incorporate the economic analysis in the study.)

These various initiatives have all developed some sort of frameworks for establishing causality links and correlations between trade and environment, in other words, to analyse ways in which trade can affect the environment.

The various initiatives differ for the subject/scope of their assessment: the Commission for Environmental Cooperation assesses NAFTA specific rules; OECD methodologies apply to national trade measures and trade agreements among countries; EU focuses on specific trade measures; WWF methodology applies to different types of trade policies, changes in trade policies and measures and UNEP 's approach applies to different trade policies. The timing, that is, when is the assessment is conducted (ex- ante, ex- post), how long-range effects are considered are other elements of difference in the approaches: for example, the CEC assessment are ex-post (evaluating the effects of NAFTA), while the EU's SIA is done ex-ante to predict possible impacts.

Other important aspects are the requirements for stakeholder participation in the assessment process; and provisions for monitoring, follow up and policy prescriptions: for example suggesting accompanying measures to minimise impacts (flanking measures).

Table 4.2 below summarises the features discussed above and other elements.

Applications and Initiatives

The forestry sector and the characteristics of production and trade of forest products have not received much interest in the large body of literature on trade and environment, as compared, for example, to the agricultural sector (for example the crop and livestock sectors have been much more studied in terms of the relationships between trade policies and environmental impacts).

Of the approaches mentioned above none has been specifically developed to address the forest trade and environment linkages and only 2 studies (by CSDDH in Mexico and UNEP in Tanzania) have actually applied the specific trade assessment methodologies to the forestry sector. These studies do an ex-post assessment of the impacts of trade policies (and the general economic liberalization policy related to them) on the wood forest products sector (timber and charcoal principally).

A sustainability impact assessment of the forestry sector is being started by the Institute for Development Policy and Management, University of Manchester under the framework of the SIA programme of the European Union. Country cases will be carried out in Russia, Malaysia, Indonesia, Brazil and Cameroon and results will be available in 2004. (C. George pers. comm.)

The very limited application of the above approaches in the forestry sector doesn't mean that the relationships between forest products trade and environment have not been studies or researched at all. Various studies on the potential effects of trade liberalisation on the forestry sector exist, and a few are somehow based on the approaches developed under the initiatives above.

The point which is made here is that these studies are mainly of a speculative and inferential nature, and are based on expert knowledge, or other literature (rather than on empirical data), as far as the causality links between trade and environmental parameters is concerned. These studies help identify the most relevant issues related to the impacts of trade liberalization on forests and the differing views and perceptions of different stakeholders.

A closely related issue, the impacts that environmental regulations and pressures have on the trade in forest products, has been comparatively better studied. In this category fall, for example, the studies which look at the effectiveness and impacts of trade-influencing measures taken with environmental justification, such as the bans on exports of logs, etc.

Table 4.2: Matrix of approaches

	OECD	CEC	EU	WWF	UNEP
Environmental vs sustainability Assessments	environmental	environmental	sustainability	sustainability	integrated
Trade first vs envrironment first	economic assessment of the trade agreement or measure	broad economic, environmental, social and political context followed by economic and other direct effects of NAFTA, including institutional effects	Trade measures	Economic changes of major trade measures	economic effects of trade measures
Causality and correlation	Product effects, scale effects, structural effects, technology effects, regulatory effects	Production management and technology, physical infrastructure, social organization	Causal chain analysis. Preliminary assessment to identify potentially significant and non-significant impacts, and to differentiate between impacts of greater and lesser significance,	scale effects, structural effects, products/technological effects, socio-economic effects, environmental effects regulatory effects (including feedback effects)	scale effects, structural effects, products/technological effects, socio-economic effects, environmental effects regulatory effects
Subject and scope	Trade measures and agreements	NAFTA (including rule changes institutions, trade flows, investment and other conditioning factors)	Trade measures (includes a screening process and a scoping phase)	Social and environmental impacts of trade liberalisation	economic, social and environmental impacts of trade-related policies and agreements (ranging from a specific trade measure to multilateral agreements)

cont.	OECD	CEC	EU	WWF	UNEP
Timing	Varies depending on the trade measure of agreement, but generally should be conducted early in the policy-making process	ex-post	ex-ante (includes scenarios to compensate for uncertainty)	Early in the negotiation process (ex-ante) but usefulness of ex-post for lessons learnt	ex post assessments (except one case, China, where an ex-ante assessment was done)
Participation	Varies. Reviews will be carried out by government officials. Consultation is recommended	Work undertaken by a multidisciplinary team with consultation. No requirement for participation apart from institutional obligation of CEC	Developed in consultation with EC and stakeholders	Sustainability assessments should be transparent and participatory	Assessment are done by a multidisciplinary team, recommends participation, recommend stakeholder participation
Quantitative vs qualitative assessment	Mix of methodologies including models, case studies and others	Case studies incorporating qualitative and quantitative approaches	Suggests a mix of quantitative methodologies, along with case studies and social science methods	Relies primarily on qualitative analysis, acknowledges potential of models with caution	A mix of qualitative and quantitative methodologies, macro and micro-level analysis.
Sectoral approaches	Case studies might be used. Recommends development of criteria	Applies to sector studies, includes criteria for selection and upstream and downstream effects into other sectors	Includes case studies in possible approaches	suggests use of sector studies and includes criteria for selection	makes use of sector studies and includes criteria for selection of priority sectors
Indicators for assessment	Suggests preliminary indications within broad topics of pollution, health, safety and resource effects	includes indicators from air, land, water and biota and indicates criteria for selection of indicators	core sustainability indicators: economic, social and environment	propose qualitative judgements to assess impacts	use social, economic and environment indicators

contin.	OECD	CEC	EU	WWF	UNEP
Monitoring, follow-up and policy prescription	Importance of monitoring results and following up and suggesting policy responses	No provision for policy recommendations	Acknowledges the need for mitigating and enhancing measures to reduce or eliminate significant negative impacts, and includes criteria and a method for selecting such measures	Includes a prescriptive analysis with policy recommendations	Prescriptive analysis with policy recommendations

(Adapted from WWF report of the International Experts' meeting on sustainability assessment of trade liberalization, Quito, Ecuador, March 2000). This recent WWF study presents in clear and comprehensive way a synthesis of the different initiatives and their approaches. The report, together with other relevant, comprehensive information on all past and ongoing initiatives, approaches, and methodologies is found on the WWF Sustainability Assessment of Trade website. The site can be accessed at the internet address: http://www.balancedtrade.panda.org

• Improving the understanding of supply and demand relationships for environmental services from forests.

- Determining scientifically the linkages between forest management and flow of environmental services from forests in a situation, where services are produced jointly

- Developing ways to integrate biodiversity conservation, watershed management and carbon sequestration in natural forests into a bundled service

- Quantifying the relations between water quality, flow regulation, sediment prevention, water supply and aquatic productivity based on various land-use practices

• Supply studies focusing on transaction costs.

- Developing mechanisms for commoditising environmental services so that the commodity's nature and extent is unambiguous, and its delivery and use can be measured and enforced at a reasonable cost

- Identifying ways of decreasing the most critical transaction costs, paying special attention to the potential of developing markets for bundled services, and improving the distribution of these costs so that the incentive framework is improved.

• Develop standard verification and certification tools applicable to environmental services from forests

- Identifying appropriate measures of service flows and developing monitoring, verification and certification methods for them

- Standardising watershed service definitions and measurement

- Developing low cost certification systems for forest environmental services, building on already existing forest certification schemes

• Studying the cost-effectiveness of various market-based mechanisms used to pay for forest environmental services.

- Comparing the cost-effectiveness of various mechanisms (including government service delivery), paying attention to the distribution of costs and benefits among the stakeholders

- Studying the incremental benefits from applying these mechanisms to see if they really make a difference in the behaviour of producers and consumers

- Identifying the pre-conditions, which need to be put in place to improve the functioning of these mechanisms

- Identifying those conditions, where the responsibility for delivering an environmental service could be delegated to a market, and identifying the role that the government needs to play in setting up, regulating and promoting market-based transactions

In what Ways can Trade Measures Impact the Environment?

The North America Commission for Environmental Cooperation identifies 6 hypotheses on the effects of NAFTA induced liberalization:

• Can reinforce existing patterns of comparative advantage and specialization, concentrating production and transportation where it takes place more efficiently (concentration in larger firms, with high visibility which can adopt higher/social environmental standards) or conversely in areas unsupported by adequate physical infrastructure or institutional capacity to handle that growth. [This is what the study by Sizer et al 1999 argues, that especially in tropical countries, (where forests have the highest social, cultural and environmental value), this second hypothesis is probable].

• Economy wide liberalisation can intensify competitive pressures, and this in some case can lead firms to lower inputs, in part reducing environmental protection or pressuring the governments to reduce environmental standards…. [Race to the bottom in the absence of offsetting interventions].

• Liberalization could lead to economic growth that promotes modernization and reduces environmental stress: competitive market pressure can hasten capital and technological modernization. Favouring producers with new efficient and clean plants and equipment

• Liberalization in specific sectors can lead to substitution of imported environmentally superior products for domestic alternatives. Conversely, some liberalization could may lead to a surge in imports that disrupt domestic production, employment, traditional technologies, and social institutions required to maintain the environmental infrastructure

• Liberalization can affect corporate practice and government policy by creating an upward movement of environmental standards and regulations toward a common regional norm (of course North America)

• NAFTA and its institutions (including CEC) could engender a regional awareness and sense of responsibility that reduce the possibility of not caring for poor environmental performance

The framework for understanding the links between trade and environment is based on processes of production, physical infrastructure, social organization and government policies.

In a practically identical way, the OECD identifies 5 relevant categories of types of effects through which trade impacts economic activities and subsequently the environment (this approach is also followed by UNEP):

• Product effects: associated with trade in specific products which can enhance or harm the environment;

• Technology effects: changes in the way the products are made depending largely on the technology used;

• Scale effects: these are associated with the overall level or economic activity or macro-economic effects resulting from the trade measures/agreement: positive ones are from higher levels of economic growth (accompanied by environmental policies), negative ones are when higher economic activities, trade or transport bring increased pollution or faster draw-down of resources;

• Structural effects: changes in patterns of economic activities or the micro-economic effects resulting from the trade measure or agreement. Positive structural is when efficient allocation of resources and efficient patterns of productions are promoted; negative is for example when environmental costs and benefits are not reflected in the price of the traded goods;

• Regulatory effects: associated with the legal and policy effects of trade measures or agreements on environmental regulations, standards and other measures. Positive ones are when trade measures and agreement take care to maintain ability of government to pursue appropriate environmental measures (see the whole dispute about NTB, SPS); negative when this ability is undermined by the provisions of the trade measures or agreement.

The OECD methodology also identifies the following broad categories of effects on the environment:

• Pollution effects (changes in emissions of noxious substances into the air, water or land, including solid wastes);

• Effects on health and safety (changes in the raising or lowering protection of human, animal and plant life and health: sanitation, potable water, chemical substances in foods, spread of pests, and environmental-related diseases such as the toxic effects of hazardous waste;

• Resource effects: (changes in the use of energy or natural resources: changes in the destruction of habitats and ecosystems, changes in the depletion of species, changes in land use patterns.

• Environmental effects can be national, transboundary or global, and these levels are all interrelated.

Trade liberalisation, trade distortions and trade agreements are the three broad areas of trade-related policies which have been the scope of environmental assessments in general, not limited to forestry.

All of the three areas have been considered in the forestry “trade-environment debate”, which has revolved mainly around the following issues:

• Elimination, reduction of tariffs and tariff escalation

• Non-tariff measures (which have been included as trade barriers in the context of trade negotiations) which have been grouped in the following categories (Sizer et al. 1999.): quantitative restrictions on imports: phytosanitary standards (for example the use of toxic fumigants, etc);technical regulations designed to protect human health and safety; labelling requirements; requirements for recycling and waste recovery; subsidies, tax breaks and export promotion, and other financial support measures; export restrictions

• Bi lateral and regional trade agreements

• Trade-distorting policies (subsidies, export bans, etc)

• Impacts of multi-lateral environmental agreements (much on CITES, some on CBD, mush less so on UNFCCC and UNCCD) on trade of forest products;

• Impacts of agricultural liberalisation on forests: changes in land uses.

What Challenges as far as the Interactions between Trade and SFM are concerned?

The review of the literature on environmental assessment of trade would support the thought that impact assessment has the potential to be a useful tool for the integration of trade and environmental policies and not only predict/quantify negative impacts. EA of trade-related policies can enhance stakeholders understanding of the implication of multilateral trade rules on sustainable development and environment, and can represent a negotiating tool for resolving environmental conflicts around trade and SFM.

However, the experiences in carrying out rigorous and comprehensive assessment of the environmental impacts of trade in forest products have been limited to date. It is therefore not appropriate here to generalize conclusions from these about the potential of environmental assessment of trade assessment to mediate between objectives of trade and of SFM. Furthermore, FAO is new to these issues (and institutionally never been particularly interested/involved in the debate or international developments) and this does not put us in the best position to recommend or suggest ways ahead (all of the information in this paper here has been taken from other sources).

The few conclusions in box 2 resulting from the review of forest-trade literature show that concerns about negative social and environmental impacts triggered by trade policies and practices are real, that direct and indirect linkages exist and that these are very complex.

Box 4.1 Concerns about Social and Environmental Impacts of Trade in Forest Products

• The trade and environment debate has settled into the assumption that trade in itself has no direct environmental links. Trade policies and practices impact the environment via changes in levels and patterns of production and consumption of forest

• Compared to macroeconomic policies and trade- distorting policies, trade liberalization policies have proved less influential in determining production, consumption effects and consequently environmental effects.

• Trade policy initiatives in the forestry sector have befitted very little from analysis of the potential socio-economic impacts they might have.

• It is difficult to link changes in the forestry sector directly to trade liberalisation.

• Efficiency improvement that result from trade liberalization may have either negative or positive consequences for the environment depending on the specific circumstances

• The main problem in forestry is the weakness of empirical data. Many conclusions are drawn on an inference basis. Findings, forecasting, etc are based on various environmental economics theories (comparative advantage, externalities, the Environmental Kuznet curves which link income level and environmental degradation…)

• In trying to assess the impacts from an environmental point of view, questions of particular relevance are the conditions under which increasing timber values protects the forests or encourages exploitation, the determinants of the distribution of production across secondary, primary and plantation forests in different regions, the degree to which plantation production is likely to substitute for production from other sources. This means also to look beyond the macro-economic level (whether and how much trade liberalisation will results in increased logging) to determine how it will affect geographical/regional production and consumption patterns

• Changes in trade regulations are likely to have an effects on the volume of trade and therefore on levels of production of some forest products.

• Among the most important environmental changes brought by changes in trade policy there are the changes in land uses (by altering production in sectors that compete with forests for land)

• To the extent that trade encourages overall economic growth, downstream product industries may experience a trade-induced boom which can put additional pressure on the forest. At the same time, increasing incomes may generate greater demand for environmental services produced by the forests. Environmental Kuznet curves-: Environmental degradation increases with income at low income levels and decreases with income at higher level. However, no robust conclusions have emerged regarding relation between income levels and forest cover.

• Basic framework for assessing readiness of countries to trade liberalization from an environmental and social perspective. Criteria:

o existence of selected forest protection policies

o existence of selected forest protection laws

o enforcement

Some environmental impacts of trade liberalization: increased consumption of wood products from poorly managed forest; overexploitation of tree species; trade pressure on less-protected forests; shifts to plantations; expanding trade with countries that subsidize logging (eliminating subsidies would reduce logging in more inaccessible areas); restricting consumer access to information; government procurement; spread of invasive species

The objective of doing trade impact assessment is to find ways to formulate mutually supportive policies by deepening the understanding of the complex relationships between trade and (in the case of forests) forest and trees use/depletion, and by putting stakeholders (governments and civil society) in the position to access this information and inform their negotiation position.

Three aspects are mentioned here which are considered relevant for SFM:

• the integration of the work on criteria and indicators for sustainable forest management in environmental assessment of trade

It has been argued that in assessing trade impacts, forests should be considered differently from other sectors because of their very high biological, cultural, and social values. These important values of the forests are reflected in the concept of SFM and in the work done to develop and agree on criteria and indicators for SFM at local, national and international level.

All the frameworks and approaches developed to assess the impacts of trade rely on the use of indicators of environmental or social or economic sustainability. However, this study has found an important gap at this level: there is no link, in the studies reviewed, to the existing C&I for SFM developed under national and international initiatives, nor to principles and standards for forest and forest product certification.

A useful path to follow would be to see how this convergence can be encouraged, and in this, the lessons from other sectors (such as agriculture and fisheries) can be useful. For example, the work of the OECD on the development of agri-environmental indicators for policy purposes in assessing the environmental effects of trade, and the work of the CBD secretariat on the impact of trade-liberalization on agricultural bio-diversity can provide useful indications.

• the issue of capacity building

There exist a clear problem of capacity in the countries to carry out impact assessment of trade and trade-related policies. This aspect is widely recognised and considered a priority by all main actors in this field (UNEP, WWF International, and EU for example).

4.4 Developments in Forest Based Industry Sector

4.4.1 Links between Trade, Financing and Structure of the Forest Products Industry

A concern expressed by many NGOs is that companies in the forest products sectors have an increasing amount of power and that transnational companies are playing a greater role. A number of studies (EIA 1996; Dudley et al 1996) cite the statistic originating from a UNCTC study carried out in the 1980s (UNCTC 1985) that the percentage of world forest product trade controlled by transnational corporations is between 80 and 90%. The current evidence for this is not very clear given the time that has elapsed since the UNCTC study was conducted. It is easy to point to recent mergers of pulp and paper companies, or to specific countries where forest holdings are concentrated amongst only a few, mainly foreign-owned companies. However, paper production is only one part of a complex, multi-product sector and sectoral characteristics can vary from country to country. Even if it is possible to marshal evidence of increasing market concentration and transnational involvement, it is necessary to ask whether this matters and if so why. Such trends may be a normal part of industry restructuring in response to changes in the overall economic environment and may be no more pronounced than in other sectors.

This chapter examines how the structure of the forest products industry is linked with the expansion of trade and the impacts of trade on forest management. The term industrial structure is used here to refer to some key characteristics of the forest products sector which are thought to influence the conduct and performance of companies within it. These include characteristics related to:

• market concentration - the number of sellers and buyers at different stages in the production chain, the extent of horizontal integration and the presence of barriers to entry,

• production chain relationships - the extent to which companies at one stage of the chain control other stages through vertical integration or through the terms of their contractual arrangements e.g specialised sub-contracting and outsourcing.

• ownership - in particular, the extent of transnational ownership

There are potential advantages to changes in the structure of the forest products sector. Larger companies can exploit economies of scale and so increase efficiency. Mergers of companies can lead to cost savings through reorganisation of production. Vertical integration can reduce transaction costs. Foreign-owned companies can bring new technology and skills and the injection of additional capital can enable a more long-term perspective, crucial for sustainable forest management. Yet there is concern that these advantages will either not materialise or that the benefits will accrue to large, powerful companies rather than to local communities, landholders or governments. Thus, the World Rainforest Movement points to the huge profits of transnational companies and banks and the cheap prices paid by consumers at the expense of loss of natural capital to local peoples such as the Baka pygmies in Cameroon, the Nahua people in Peru and the Saramaka people of Suriname (World Rainforest Movement 2001).

It is therefore important to study the changes in industrial structure which have taken place in the forest products sector at an international level and more specifically in different countries, to examine what has driven these changes, in particular the link with trade and trade liberalisation efforts, and to assess implications for the extent of sustainable forest management and the benefits and costs which accrue to different groups from forestry.

This cannot be accomplished with any great precision as the sector is heterogeneous with marked differences in trends between products and regions. A general analysis of trends in the structure of the forest products sector and their likely implications is supplemented by detailed examination of the situation from three case studies in Ghana, Philippines and Brazil. These countries were chosen because tropical timber harvesting and processing plays an important role in all three. In Ghana and the Philippines there has been a substantial export trade in tropical timber products and also marked changes in the structure of the sector. Brazil is of interest because of its dual role as a pulp and paper producer and tropical timber exporter. We also draw from recent IIED research on small and medium forest enterprises and on private sector participation in sustainable forest management.

Changes in the Structure of the Forest Products Sector

The forest products sector is heterogeneous with many different products, species and technologies. In terms of industrial structure some broad distinctions can be made by:

• Location of the resource base: Temperate/boreal regions and tropical regions.

• Product groups: the two main categories being paper products and those primarily concerned with solid wood products. However, many companies produce both in temperate regions.

• Species: Hardwoods and softwoods

This creates some anomalies as processing may not take place in the same location as the resource base and some countries like Brazil, have plantations in both temperate regions and tropical regions.

It is also necessary to consider the characteristics of the different stages of the forest product value chain from the resource base through to the final consumer. The two main stages of concern are the resource base and processing but it is important to understand how these compare with and are linked to distribution activities and the customer or buying sectors.

We examine trends in market concentration and barriers to entry at these two main stages and the extent to which there is vertical integration or disintegration. As data is limited, this analysis draws heavily on industry perceptions of trends as expressed in a working group on industrial structure at the Expert Consultation on Trade and Sustainable Forest Management Interactions, hosted by FAO in February 2003.

Market concentration is of interest because it serves as an indicator, albeit imperfect, of the extent of market power of companies in the sector and their ability to charge higher prices than under more competitive conditions. Many empirical studies have been done of the relationship between concentration and price in a number of sectors and the majority have found a significant positive relationship between concentration and price (NERA 1999 citing Weiss 1989). It is also necessary though to examine the extent of barriers to entry. Increasing concentration may not translate into the ability of companies in the sector to exploit this by charging higher prices than under more competitive conditions as this will attract new entrants to the market. It is also important to consider market power relative to that of companies in the preceding and successive stage of the value chain. Companies may have market power in relation to their suppliers but may themselves be faced by concentrated buying sectors. Conversely, they may encounter more concentrated supplying sectors. Competition policy, aimed at curbing market power, has traditionally focused on the impact of seller power on consumer welfare and has given less attention to the impact of buyer power, exercised by concentrated buyers against their suppliers (Dobson, Waterson and Chu 1998).

Vertical integration, which refers to the extension of activities, often through merger or acquisition of existing firms, to upstream or downstream activities in the same value chain, is important because it can affect the extent of barriers to entry.

4.4.2 Market Concentration and Vertical Integration

Temperate and Boreal Regions

The size of supplier varies considerably and hence the extent of concentration. The US and Northern and Central Europe are characterised by a high degree of participation of small private forest owners who supply timber to pulp and paper and wood processing companies. There are estimated to be 12 million private forest landowners in Europe (Swedish Forest Industries Federation). But many of the large companies in the forest products sector have extensive land holdings. International Paper is the largest private landowner in the US with 10 million acres (and holdings of 10 million acres in other countries, Canada, Russia, Brazil and New Zealand) (International Paper 2002a). Most companies rely on a mix of own holdings and third party suppliers. It is rare for processing companies to have no forest holdings and to buy in all their wood raw material requirements but also rare for a company to be entirely dependent on its own forest holdings for supplies. Canada is distinct from other main producer countries in that forest lands are mainly publicly owned and allocated by concessions, so there is less small landowner involvement. But the situation is different for First Nation forest reserves. Depending on provincial laws of resource tenure, companies may need to buy concessions from residents of the reserves. This has prompted a number of joint ventures or partnerships between forest product companies and community-based forest enterprises (Mayers and Vermeulen 2002).

Temperate regions of developing countries such as Chile and Brazil where production is based on plantations, are characterised by concentration of ownership.

• In Chile, 71.1% of the radiata pine plantations are owned by 2% of the forest land owners and two companies own close to 50%. Similar ownership patterns prevail for the eucalyptus plantations there (Borregaard and Dufey 2001).

• In South Africa, two companies, Sappi and Mondi own 47% of the plantations, the State owns 30%, smaller private enterprise and individuals own 22% and the remaining1% is shared by some 19,000 small or micro-growers (Mayers et al 2001).

Secure access to wood raw materials through forest holdings has in the past been considered as an important competitive advantage for companies. Backwards integration of processing companies into forestry through acquisition of companies owning land or concessions used to be quite common particularly for the pulp and paper sector. Studies in the early 1990s identified numerous cases of this in Australia, Canada and Finland and acquisition by Japanese processing firms of forestry companies in North America and Chile (Johnstone 1996).

A noticeable trend is for companies, particularly those involved in pulp and paper manufacture to move out of forest ownership. Companies that do own forestland have been increasingly selling it off and relying instead on long-term contractual relationships for the sourcing of their fibre. In the US, at least 2 million acres were sold by paper companies for a total of $1.4 billion or more in the first nine months of 2002 (Paperloop Sept 2002 a). The buyers in the US tend to be Timberland Investment Organisations (TIMOs) which are becoming the most significant landowner group there, but some sales have been made to conservation groups. There are a number of drivers of this divestment. The increasing availability of other sources of raw material, notably recovered paper and increasing processing efficiency means that a smaller amount of virgin fibre and hence forest land is required in spite of increasing demand. At the same time, companies need to reduce debt and finance industrial expansion and acquisitions. Weyerhaeuser used the proceeds from its sale of land in Washington to the Hancock Timber Resource Group in October 2002 to pay down the debt associated with its acquisition of Willamette Industries (Weyerhaueser 2002). The other driver is increased confidence in the stability of the resource and the security of supply from private landowners. This divestment process is occurring in privately owned natural forest but also in plantations. New Zealand is the prime example of the latter.

Another trend is to supplement the company production from planted forests by joint venture arrangements with private landowners.

• In Australia, joint venture arrangements of three different types (lease, crop share and guaranteed sale) have contributed to the establishment of 8% of the country's plantation estate since the mid 1980s (Mayers and Vermeulen 2002).

• Outgrower schemes are important in Brazil's temperate plantations and and in the case of Klabin, a major producer of pulp and paper and hardwood and softwood logs, are motivated by the need to maintain a good company image and the increasing costs of land for plantations (Mayers and Vermeulen 2002).

• In KwaZulu Natal, South Africa, Sappi and Mondi, the two largest forest products companies source about 3% of their raw material inputs from outgrower schemes (Mayers and Vermeulen 2002).

It is also common in some countries for companies to contract out harvesting and other activities to smaller companies. In South Africa, 95% of harvesting and transport activities in the plantation wood sector are contracted out and 33% of silviculture activities(Lewis et al 2003).

In the pulp and paper sector, there has been considerable merger and acquisition activity both in national markets and at a global level. While mergers and acquisitions across national boundaries are a recent phenomenon, consolidation at the national level is nothing new and has been a feature of the evolution of the sector in most countries. International Paper, the largest company in the sector globally, itself started in 1898 as a merger of 17 pulp and paper mills (International Paper 2002b). In the US, the average number of firms per paper product category declined from about 70 in 1978 to about 60 in 1992 in a period when total capacity in the sector increased by 20%. About 40% of the 819 paper and paperboard plants operating in the US in this period were involved in at least one merger (Pesendorfer 1998).

The sector, while more concentrated than solid wood activities, is considered fragmented in comparison with some other manufacturing sectors. This can be seen from data on concentration ratios for the US in Table 9.1, which shows that the top four firms in manufacture of transportation equipment accounted for around 50% of output while the top four in the paper manufacturing sector controlled less than 20%. However, of the 21 sectors listed, only six have a higher four-firm ratio than the paper manufacturing sector and only five have a higher Herfindahl-Hirschmann index. Nevertheless, the four-firm ratio for the paper manufacturing sector is below common benchmark thresholds for market power such as the Scherer-Ross threshold of 40 (Johnstone 1996). Moreover, empirical research on the effects of concentration in a number of sectors suggests that the four firm ratio could be as high as 50% before there are any significant effects on price (NERA 1999 citing Weiss 1989).

Table 4.3 US Concentration ratios by value of shipments 1997

Industry (3 digit NAICS)	Four-Firm ratio¹	Herfindahl-Hirschmann Index²
Food manufacturing	14.3	91.0
Beverage and tobacco	45.1	777.2
Textile mills	13.8	94.4
Textile product mills	22.8	186.2
Apparel manufacturing	17.6	100.6
Leather and allied products	19.0	167.2
Wood product mfg	10.5	52.7
Paper mfg	18.5	173.3
Printing and related	9.6	38.4
Petroleum and coal products	26.0	350.0
Chemical mfg	11.9	76.6
Plastics and rubber mfg	8.2	30.2
Non-metallic minerals mfg	9.1	52.1
Primary metal mfg	13.8	97.4
Fabricated metal mfg	3.5	8.5
Machinery mfg	11.5	55.4
Computer and electronic product mfg	19.1	136.6
Electrical equipment	14.8	105.9
Transportation equipment mfg	49.7	797.6
Furniture and related	11.2	55.5
Miscellaneous mfg	7.4	33.2

1. The percentage of value of shipments accounted for by the four largest companies

2. The sum of the squares of the individual company percentages for the 50 largest companies or the universe, whichever is lower. Source: US Census Bureau 2001

Table 4.5 shows concentration ratios at a more disaggregated level for the US. It can be seen that pulp, paper and paperboard are quite concentrated with newsprint and pulp both exceeding the benchmark of 40. The converting sectors are more heterogeneous so there is wider variation within these sectors.

Because of changes in 1997 in the sectoral classification of industry used by the US census from Standard Industrial Classification System (SIC) to the North American Industry Classification System (NAICS) care has to be taken in comparing with concentration ratios from earlier censuses, particularly for higher level groupings. It is notable though that the 1992 figure for the sector paper and allied products is 18, very similar to the 1997 ratio. Analysis by Johnstone of concentration ratios at a more disaggregated level from 1967 to 1987 in the US, showed that in most cases market shares of the four largest firms did not increase in this period. Exceptions were for sanitary products, newsprint, and paper mills (Johnstone 1996).

Table 4.4 Concentration ratios by product category based on value of shipments

4-5, and 6 digits NAICS	Four -Firm Ratio	Herfindahl-Hirschmann Index
Pulp, paper and paperboard mills	28.0	356.0
- Pulp mills	58.6	1,106.4
-- Paper mills	37.6	541.7
-- Newsprint mills	43.9	766
- Paperboard mills	33.6	485.1
Converted paper product manufacturing	12.0	96.2
-Paperboard container mfg	19.2	175.8
- Paper bag and coated and treated paper mfg	27.1	266.7
- Stationery product mfg	27.5	296.9
- Other converted paper product mfg	42.3	688.1
--Sanitary paper product mfg	63.1	1,481.0
--All other converted paper product mfg	23.0	187

1. The percentage of value of shipments accounted for by the four largest companies

2. The sum of the squares of the individual company percentages for the 50 largest companies or the universe, whichever is lower. Source: US Census Bureau 2001

In some developing countries, production is more concentrated than in the US. In Chile, in 1997 pulp production was carried out by just five companies and one company accounted for 50% of pulp exports (Borregaard and Dufey 2001). In South Africa, the four main groups - Mondi, Sappi, Nampak and Kimberly produce 98% of the country's pulp, paper and board (Mayers et al 2001).

The definition of market, however, is crucial for the calculation of these concentration ratios. If companies can sell their products worldwide and a national market can be freely supplied by imports, national concentration ratios may be inappropriate. Taking a global perspective, it is clear that concentration has increased but is still not close to the threshold levels set out above. In 1993, the top four and top ten paper companies produced 9.6% and 19.4 % respectively, of the world's output of paper and board. By 2001 these shares had increased to 16% and 28%, reflecting mergers between companies like Stora and Enso, UPM and Kymmene, International Paper and Champion International. As with national level data, this level of aggregation masks considerable variation between paper grades.

There are also sizeable barriers to entry to pulp and paper manufacturing because of the capital intensive nature of the activity and the potential for economies of scale. The capital intensity can be demonstrated by data from the US. In 2000, the ratio of fixed assets to gross output in the paper and allied products sector was 1.83, compared to an average for US manufacturing of 1.15. Another indication of capital intensity is given by mill size which has steadily increased over the years as companies endeavour to exploit economies of scale. Table 4.6 shows how average mill capacity increased in Sweden between 1960 and 2000.

Table 4.5 Average Pulp, Paper and Paperboard Mill Capacity in Sweden

Year

Average Mill Capacity ('000 tonnes per annum)

Pulp

Paper and Board

1960

45

30

1970

90

70

1980

145

115

1993

225

185

2000/01

269

231

Source: Johnstone 1996 and PPI 2002

On the buyer side there has also been consolidation, particularly amongst media companies, which are large buyers of newsprint and coated papers, and packaging companies.

Softwood-based paper production in temperate countries is typically integrated with solid wood production because the residues from lumber production can be utilised in pulp manufacture. Many of the leading paper companies in North America and Europe also have solid wood product divisions. Production based on hardwoods is more likely to be separate.

Overall, the solid woods industry is fragmented . Overcapacity and low barriers to entry mean that profitability is generally rather low. This applies particularly to the hardwoods segment. Consolidation is more evident though in the engineered wood products sector which is more capital intensive. Some indication of the extent of concentration in the US is given by Table 9.1 in the previous section which shows that of the 21 three digit NAICS sectors, only five have a lower four-firm concentration ratio and Herfindahl-Hirschmann index than wood product manufacturing.

As with the pulp and paper sector there is considerable variation within industry activities as shown in Table 9.4 with ratios for veneer, plywood and engineered wood products that are close to or exceed the benchmark thresholds.

South Africa provides a contrast to this general lack of market concentration. The five largest owners of sawmills, account of 70% of total production while some 220 small-scale mills produce only 10% of sawlogs (Mayers et al 2001).

Overall the solid wood sector is less capital intensive than pulp and paper and manufacturing generally. However, there are likely to be marked differences within the various industry sub-categories. In the US in 2000, the ratio of net private fixed assets in the lumber and wood products sector to value added was 0.79 compared to an overall average for manufacturing of 1.15.⁴⁹

The extent of consolidation in the distribution and buying stage of the value chain depends on the type of wood and the region. Distribution is concentrated in the US particularly for softwoods and there are signs of increasing consolidation in Europe. Buying sectors such as DIY and construction tend to be fragmented in Europe but concentrated in the US. Buyers of hardwoods are generally considered to be fragmented in both Europe and US.

Table 4.6 US Concentration Ratios for Wood Products by Value of Shipments

Industry - 4,5 and 6 digits NAICS	Four-firm ratio	Herfindahl-Hirschmann Index
Sawmills and wood preservation	14.5	86.7
Veneer, plywood and engineered wood product mfg	26.9	286.9
--Hardwood veneer and plywood mfg	30.5	430.4
--Softwood veneer and plywood mfg	48.8	914.9
--Engineered wood member (except truss) mfg	77.1	2,453.5
--Truss mfg	8.8	42.4
--Reconstituted wood product mfg	42.9	592.3
Other wood product mfg	12.7	66.8
-Millwork	16.1	100.9
-Wood container and pallet mfg	5.8	15.9
-All other wood product mfg	26.3	257.9

There are a number of possible routes for vertical integration. As processing encompasses two stages of production, pulp manufacture and paper manufacturing, the most obvious route for integration is to combine these two stages. Although there is still a sizeable market pulp sector, the overall trend has been to integrate pulp and paper production. In Sweden, in 1960 just over 40% of pulp production was integrated with paperboard manufacturing, while in 1990 this figure had risen to over 80% (Johnstone 1996 citing ILO 1992). The US pulp and paper industry is highly integrated with market pulp accounting for only a small percentage of production or consumption, but Canada presents a contrast with some 65% of chemical pulp output being non-integrated (Paperloop 2001). This is the result of both technological integration ie the physical integration of production processes and institutional integration ie mergers or acquisition of firms engaged in different stages of the production process (Johnstone 1996).

Forward integration of processing companies into distribution is perceived to be on the increase at a global level. This is to facilitate global market access. Forwards integration from processing and distribution into buying sectors and backwards integration from buyers is less common than for earlier stages in the value chain, though there is variation between product categories. It is very rare for newspaper publishers to acquire or be integrated with newsprint manufacturers. In packaging sectors more vertical integration is typical and is not a recent trend. According to a study carried out in the early 1992s in the US 75% of corrugated case making was owned by paperboard mills (ILO 1992 cited in Johnstone 1996). Some of the top companies in the pulp and paper sector have major packaging divisions, for example International Paper.

As with the pulp and paper sector, there is a trend away from backwards integration to divestment of forest land by processing companies but this is more marked for Europe than for the US.

There is some forwards integration into distribution and end use sectors but this is less evident than for the pulp and paper sector and applies more to larger companies. Weyerhaueser, for example is engaged in the growing and harvesting of timber, manufacture, distribution and sale of forest products and real estate construction.⁵⁰

Diversification and Specialisation

Preferences for diversification and vertical integration change over time, often in line with business cycles. A study of forward vertical integration for Canadian and US producers of structural panels and softwood lumber in the 1980s found that there was decreasing reliance during this period on forward integration into distribution. The authors attributed this to the recession of the early 1980s which lead to companies increasing their focus on core industries (Cohen and Sinclair 1991). However, this move away from forward integration did not apply to the larger firms.

It is therefore reasonable to assert that the leading companies have steadily widened their range of products and businesses over the years, often through mergers and acquisitions. A more recent trend though, is for companies to divest non-core assets and to concentrate on core businesses. Stora Enso, for example, has moved out of energy generation and specialty paper for this reason (Paperloop 1999) Georgia Pacific has recently announced plans to sell a majority stake in its global paper distribution subsidiary and to split into two companies, one concentrating on consumer products and packaging and the other on building products and distribution. This offsets to some extent the trend towards horizontal and vertical integration.

4.4.3 Resource Base and Processing in Tropical Regions

There is considerable variation depending on whether forest land is privately owned or held under concession and on whether natural forests or plantations are involved. For natural forests, the solid wood sector is fragmented at the resource stage. Where forest land is allocated by concession as in West and Central Africa, and some Asian and Latin American countries, there is scope for concentration. But In Central Africa there are legal restrictions on the size of concessions which limit the possibility of expansion. The view of the private sector operating in these countries is that these limits do not permit operation at an economically viable level. In Latin America, there are variations within the region in relation to land ownership. In Brazil forest lands are privately held and as a result reliance by wood processing companies on outside suppliers for part or all of their raw material requirement is common. In Bolivia and Peru there is a system of forest concessions.

Where production is largely based on plantations, the overall trend is towards expansion of landholdings. However, outgrower schemes involving local communities are quite common and vary from simple leasing of land to joint ventures between the company and the smallholder. Motivations vary - usually the desire to improve the company image and relations with the local community is important but in some cases it may be the only way to get access to land.

• Aracruz in Brazil initiated its outgrower schemes to increase its fibre supply after protests against companies owning large tracts of land prevented it from expanding its land holdings

• The Phoenix Pulp and Paper company in Thailand after having little success with large scale plantations now sources all of its raw materials requirements from small-scale farmers through direct purchase or outgrower schemes

• Stora Enso and the Indonesian company Inhutani III jointly run an outgrower scheme in West Kalimantan on government land to which local people hold traditional user rights. The main aim was to avoid conflict with local people (Mayers and Vermeulen 2002).

As in temperate regions, contracting out of harvesting activities is common in some countries, for example Guyana. Solid wood sectors in tropical regions are typically less concentrated than manufacturing sectors on average. This is illustrated by the case of Brazil. Table 4.8 shows that the concentration in the forestry sector is only slightly higher than the Brazilian average, but considerably smaller than in the pulp and paper sector. Wood processing and furniture production are both less concentrated than average. In Ghana concentration in the tertiary sector is very high but this is based on exports only (Table 4.9). A problem common to many tropical timber countries is excessive processing capacity in relation to forest resources.

Table 4.7 Herfindahl indexes for the main forest activities in Brazil, 2000

Economic activities	Herfindahl index
Forestry operations	1849,0
Processing industries (excluding pulp and paper)	1688,1
Pulp production, paper and paper products	2562,8
Furniture production and diverse industries	1643,1
Brazil (all activities)	1813,0

Source: Young and Prochnik 2003

Table 4.8 Concentration ratios (based on volume of exports) Ghana

Sector	Before 1995		After 1995
	Market share of 4 largest firms	Herfindahl Index	4-firm ratio	Herfindahl index
Primary	31	462	6	7435
Secondary	16	na	18?	Na
Tertiary	68	1,406	63	1,137

Source: Amponsah (2003)

Distribution is concentrated in Brazil and Latin America generally with just a few intermediary companies controlling exports of tropical timber. Companies rely on intermediaries because they do not produce sufficient volume to justify taking on this function. In Asia the situation is different, there are more companies involved in distribution because of the larger volumes. In Africa, the producer companies employ representatives to take charge of distribution. Buying sectors are quite specialised and fragmented.

Vertical Integration in Tropical Regions

Log export bans have stimulated forward integration from logging into processing in many countries. Forward integration into distribution is common only where there are large volumes involved. Forward integration from processing into buying sectors is not very common. But specialised sub-contracting arrangements are becoming common in the solid wood sectors and are particularly important for furniture. 75% of Malaysia's furniture exports are produced under sub-contracting arrangements (ILO 2001).

The last fifteen years has seen a spate of merger and acquisition activity in the forest products sector both within and across national boundaries. For example, outward FDI from Finnish forestry firms increased tenfold over the period 1988 to 1998 (ILO 2001).

4.4.4 Transnational Involvement

Many of the leading pulp and paper companies, which in the 1960s and 70s typically operated solely in their home country, now have a global reach with forest operations, manufacturing facilities and distribution activities in a range of countries. However, only one company in the paper sector, Stora Enso, was included (at number 54) in UNCTAD's top 100 non-financial transnational corporations, ranked by foreign assets in 2000 (UNCTAD 2002).

The biggest deals in the pulp and paper sector are typically between companies in developed countries and intra-regional investment predominates. North-South foreign investment, has been relatively minor compared with trends in the solid wood sector but is on the increase. There is also some South-South investment and South-North investment in the sector. However, only one paper company, the South African company Sappi, is included (at number 11) in UNCTAD's top 50 non-financial TNCs from developing economies, ranked by foreign assets, 2000.

The bulk of foreign direct investment in the pulp and paper sector in developing countries is concentrated in Latin America in fast-growing plantations. Even so the amounts involved are not commensurate with the economic importance of the sector and the overall amounts of foreign direct investment in the economy. In Brazil, the forest products sector contributed 6.9% to GDP in 2001 but accounted for only 2% of total FDI stocks (Macqueen et al 2003). In Chile, FDI has played only a minor role in the development of the pulp and paper sector (Borregaard and Dufey 2001).

There has perhaps been less foreign direct investment in solid wood sectors in developed countries than in the pulp and paper sector. In the US, data from 1990 shows how the lumber and wood sector were less international than the pulp and paper sector and manufacturing in general. Foreign-owned lumber and wood product firms in the US contributed only 2.9% of sectoral value added, compared to 7.9% for paper and allied products and 13.4 for manufacturing in general (Johnstone 1996).

Box 4.2 Foreign Direct Investment in Tropical Forest Industries

Africa

Cameroon - 90% of logging companies are foreign-owned

Central African Republic - Seven of the nine major concessions are held by foreign companies (French, Lebanese and Malaysian)

Côte d'Ivoire - 85% of the capital stock in the forest industry is foreign-owned

Gabon - The major forest companies are mostly subsidiaries of European firms but Asian investment is becoming more prominent.

Ghana - About 20% of forest product companies are wholly or partly foreign-owned (Canada, Germany, India, Lebanon, Liechtenstein, Netherlands, Taiwan, UK) (Amponsah 2003).

Liberia - 84% of forest land allocated under concessions is held by foreign companies (Malaysian, Lebanese, European and Indian)

Asia-Pacific

Malaysia - 38% of the investment in wood processing in 2001 came from foreign companies (Singapore, Japan, Taiwan, Germany, China, Korea, Switzerland, and USA. But forest concessions are 100% local-owned.

Papua New Guinea - The forest industry is 90% foreign-controlled (Malaysia, Japan, Europe, Singapore, Korea, China and Australia

Latin America and Caribbean

Guyana - Five large scale foreign-owned companies producing mainly for export markets

Honduras - Foreigners hold 20% of capital in primary forest industry and 40% of secondary forest industry

Suriname - Six concessions with a total area of 700,000 ha issued to foreign companies, mostly Chinese

Source: ITTO (2002)

But North-South investment has played and continues to play an important role and is more widespread than for the pulp and paper sector.

Foreign investment in solid wood enterprises in tropical regions has been substantial as shown in Box 4.2. European companies have been investing in forest operations in Africa for decades while Asian companies have had significant foreign investments in the forestry sector since the 1960s but primarily within Asia. They have however, intensified their foreign investments since the 1990s and moved into other regions, notably South America and Africa (Sizer and Plouvier 2000).

More recent moves by European companies are into Eastern Europe, for example Danzer Group is closing veneer production in Germany and Belgium and building a new veneer mill in Prague. There are also signs of divestment from Africa because of the uncertain investment climate and declining profitability. In Ghana, new FDI into the forest sector is declining and this is thought to be primarily because of the shortage of raw materials (Amponsah 2003).

4.4.5 The Relation Between Industrial Structure and Trade

The preceding section has shown a clear trend towards increasing transnational involvement in the forest products sector and a more mixed picture in relation to market structure. While there are signs of increasing concentration in the pulp and paper sector, solid wood industries, in tropical regions particularly are fragmented, except in the more capital-intensive activities. It is important to examine how these trends in the structure of the forest products sector are linked, if at all, with changes in trade patterns and trade policy and what the implications are for future trends.

4.4.6 The Impact of Trade Liberalisation on Market Structure

The expansion of trade and the opening up of markets through trade liberalisation, has been an important driver of the process of consolidation and of mergers and acquisitions across national boundaries. This is because in order to compete on the global market, companies have to search for ways to maintain competitiveness. They can lower costs of production through exploiting economies of scale and scope, implying consolidation, or by shifting certain stages of the production process, in particular timber harvesting to low-cost locations, implying foreign direct investment. The increasing interest on the part of North American and European companies in fast-growing plantations in developing countries reflects this continuous quest to reduce costs of production in relation to competitors.

While trade liberalisation can be linked with increasing concentration, this does not necessarily imply increasing market power for the larger companies in the sector. The opening up to international competition changes the dynamics of industry restructuring and leads to a wider range of competitors.

4.4.7 The Impact of Trade Restrictions on Industrial Structure

Trade restrictions, in particular the log export bans in South East Asia, have also played a role in prompting wood processing companies in search of raw materials to invest in logging operations in other regions (Sizer and Plouvier 1998). They have also encouraged vertical integration from logging to processing as in the Philippines (Box 4.3). High import tariffs for processed products also stimulate foreign direct investment as a means for companies to produce behind protective barriers.

Whether FDI increases or reduces trade depends on the primary motivations behind it. These can include access to markets in the host country particularly where there are tariff barriers, access to resources and achievement of cost savings often through use of lower cost labour. Market-seeking FDI can sometimes have a trade-reducing or trade-diverting effect by giving companies direct access to the market in the host country. Norske Skog's investments in South America, for example, are motivated by the growing market for newsprint there. The company is now the largest supplier of newsprint in the region (Norske Skog).

Similarly, investments planned by paper companies UPM-Kymmene and Oji Paper in China are based on projections of substantial market demand there (Paperloop 2003). But these types of market-seeking investments appear to be less common in the forest products sector, than resource or efficiency-seeking investments.

Box 4.3 The Structure of the Forest Products Industry in the Philippines

In the Philippines, there has been a striking fall in the number of enterprises engaged in logging and other forestry activities from 45 in 1988 to 9 in 1997. At the same time there has been a drastic decline in timber output from a peak of 72.5 cu m in 1965-1969 to 18.10 in 1985-1989 and only 3.5 cu m in 1996-2000, less than 25% of wood processing capacity.. This has not been accompanied by an increase in size in terms of employees or value added implying that companies have exited the sector rather than merged (Medalla 2003). It has been the dramatic decline in forest resources which has led to this industry restructuring (although the depletion of forest resources was stimulated by the export trade). Trade policy has also been a contributing factor as the log export ban imposed in 1989 must also have affected the prospects for logging companies. Most if not all of the companies remaining in logging are vertically integrated and those that were not have closed down. However, the wood-based manufacturing sector in the Philippines has continued to grow in spite of the declining forest resources although there has been a decrease in the number of operating sawmills and plywood factories in the last five years (ITTO 2002). Trade liberalisation policies, in particular the reduction of tariffs on imports of intermediate wood products have been the main driving factor of the increasing wood-based manufacturing sector. The Philippines was the fourth largest importer of tropical veneer in 2001 and the 11^th largest importer of tropical logs (ITTO 2003).

The situation varies, particularly for temperate regions. This is illustrated by a study of FDI and exports from forest companies in Finland, Sweden, and the US in the 1990s (Uusivuori and Laaksonen-Craig 2001). This found that FDI for the Finnish and Swedish forest industries had no significant impact on exports and that increasing exports were linked with decreases in FDI. For the US, a two-way negative feedback relationship between exports and FDI was found. Both sets of results imply that FDI and exports in the forest products sector may be substitutes rather than complements for the home country. But this says nothing about the trends in trade in the host country and while exports from the home country may decline as a result of FDI, at a global level trade may increase.

For tropical timber, foreign direct investment appears to have been primarily a driver of trade, facilitating access to forest resources in host countries, to supply home country markets as well as other markets. Exports of logs from West and Central Africa have been driven by the investment of European companies there and latterly Asian companies.

In some countries though, resource-seeking FDI has not had such a clear impact on trade. In Brazil, while foreign invested companies are mostly targeting export markets, they are outnumbered by Brazilian-owned exporting companies which account for 75% of wood exports (Young and Prochnik 2003). Moreover, very few of these foreign investments have involved greenfield investment. In most cases, an existing Brazilian company has been acquired which may well have been export-oriented already. In the Philippines, foreign investment in the forest products sector is also not very significant. In this case, this reflects the fact that the sector remained restricted for foreign investment for a number of years until 1992 when foreign equity of up to 40% was permitted for forestry and all restrictions were lifted for wood processing (Medalla 2003). By that time, there were little forest resources left to attract foreign investors.

In the pulp and paper sector, which is relatively capital intensive, foreign direct investment has potentially more significance in promoting trade expansion. Export-oriented plantation developments in South America, Brazil, particularly, currently involve significant amounts of foreign participation. But as with the solid wood sector, for some of the companies concerned foreign investment has come only after they were well-established. The driving factor for the establishment of these companies were government fiscal incentives provided in the 1970s and 80s and finance provided by the Brazilian development bank (BNDES). Moreover, FDI is constantly changing and an increase in foreign participation in one company may be accompanied by a reduction in another. While Cenibra, originally a Japanese and Brazilian joint venture has become 100% Japanese owned, Aracruz has increased its Brazilian participation - the 28% share previously held by Anglo-American's Mondi group was bought in 2001 by the Brazilian company Votorantim.

The Impact of Increasing Market Concentration

As processing companies increase their market share through expansion and horizontal integration, they may increase their buying power in dealing with suppliers, particularly of wood raw materials. This means that they may be able to exert pressure on their suppliers for improvements in forest management. At the same time they may be able to capture any market benefits in the form of price premiums for certified products or access to higher value markets and ensure that it is their suppliers that incur the additional costs. There is an increasing concern that the producers closest to the resource and with the most scope for improvement in forest management are receiving little benefit from the demands of environmentally sensitive markets. Instead, the benefits are being captured by companies further down the value chain, which can exert their buyer power over these suppliers. There is insufficient evidence to confirm that this is happening on a large scale, but there appears to be a reluctance among the buyers groups for certified forest products to pay a premium. The general impression is that price premiums for certified products have not been as evident as originally foreseen. This may reflect the exercise of buyer power along the production chain or it may be attributable to an unwillingness or inability to pass on costs to the next stage of the production chain and ultimately to consumers.

Processing companies as they increase in size may also be able to increase their selling power in relation to retailers, end-users and ultimately consumers. Greater selling power may mean that processing companies can resist calls from their buyers for improvements in forest management or pass on some of the costs to them. The preceding analysis however, has shown that the processing stage of the forest products sector is not very concentrated in comparison to other sectors. This applies less to the pulp and paper sector than to the solid woods sector but this accounts for only a small part of overall wood raw materials used in production. Nevertheless, concentration ratios appear to be higher in developing countries and barriers to entry high because of restrictions on access to forest concessions.

Increases in the size of companies can affect their bargaining power with governments as they become more crucial for employment and government revenue. This may influence the terms on which they access forest concessions and increase the scope for discretion in the enforcement of forest policy.

Alternatively, the unit costs for government agencies of monitoring larger companies are generally lower making them the most likely target of government inspection activity. In the state of Amazonas in Brazil, where there are just a few relatively large mostly foreign-owned companies engaged in timber harvesting and processing, government regulation has been considered more effective than in other states. This was cited by one of the companies, Gethal as a contributing factor to its drive to seek forest certification (Bass et al 2001). Moreover, large companies are generally more visible to the general public, facilitating NGO campaigns.

The establishment of large companies in natural forests may increase market concentration but there may be beneficial effects in relation to sustainable forest management. In Brazil, where one of the main threats to the forest resource is conversion, many of the small forest companies survive by being able to migrate easily to new frontier areas. For large companies, high fixed costs reduce their scope for migration and this can encourage forest management practices in order to continue their activities in the future (Young and Prochnik 2003). Similarly in the case of Ghana, it is found that large-scale wood processing firms have a higher recovery rate than small-scale companies, implying more efficient use of the forest resource. They are also more likely to be able to diversify their range of products and processes again permitting higher recovery through use of mill waste as inputs for other products (Amponsah 2003).

Another consequence of increasing size of companies (though this does not necessarily imply concentration) is that they are more likely to secure financial backing through listing on a stock exchange. This could bring them to the attention of socially responsible portfolio investment institutions which provide another source of pressure for improvements in sustainable forest management as discussed in Chapter 4. Most of the large pulp and paper and integrated forest companies are listed with shares increasingly held by financial institutions. Solid wood companies, particularly those engaged in tropical hardwood processing, are more likely to be family-owned, private companies or at least family-controlled in the case of listed companies (van Gelder 1998).

The Impact of Vertical Integration on Forest Management

Trends to vertical integration, particularly to distribution and buying sectors may have the effect of simplifying the production chain and making the company at the forest operation level more visible to consumers. In environmentally sensitive markets this may intensify pressures for improvement in forest management and give a stimulus to forest and chain of custody certification. However, there are few examples of this happening, and in some cases companies have been able to keep a low profile even where they produce items sold at retail level. Stationery products originating from Asia Pulp and Paper, allegedly from unauthorised natural forest clearance in Indonesia, were sold through British retail outlets under other brand names and were not easily associated with the company (FoE 2000). What appears to be more important than forwards vertical integration in driving demand for certified products is the interest of key buyers with a position of leverage over their supply chain (Bass et al 2001).

Rather the pressure from end-users and consumers for SFM has led some wood product manufacturing companies to integrate backwards and increase their control over their sources of wood raw materials. An example is given by Tramontina a producer of wooden kitchen utensils and garden furniture in Brazil which having had little success in persuading its suppliers to change their practices, has bought forest land in order to apply for forest certification (Viana et al 2002). Similarly, Castle Doors, a US company with a manufacturing subsidiary in Bolivia, and a supplier of Home Depot, has stated plans to move into lumber production by acquiring forest concessions from the Government of Bolivia in order to achieve 100% certification of its products. Other advantages for the company of such backwards vertical integration would also be price and availability control (Castillo 2001).

In Brazil there is concern over the social implications of land acquisition by forest companies. Social tensions can be created by concentrated land ownership and large landowners or are a common target for invasions. The need to finance land purchase may divert resources away from the modernisation of production and the capacity to invest in sustainable management and improved labour conditions (Young and Prochnik 2003).

Increasing foreign involvement can have both positive and negative impacts on forest management and the outcome can vary depending on the business strategy of the investing company, its sources of finance, the markets it is targeting and the policy context in which it operates.

Foreign direct investment may bring improvements if the investing company has a global set of environmental and social standards and a declared policy on sustainable forest management and is concerned about maintaining its reputation. Most of the leading pulp and paper companies have policies or declarations on their commitment to sustainable forestry available on their websites and some solid wood companies also, for example Danzer Group. Shell had a policy of sustainable forest management for its plantations divisions and obtained FSC certification in 2001 for all its forest operations in South America (Shell 2001).

Some analyses of post-liberalisation natural resource-based companies in Africa also point to positive effects of FDI. They note that prior to economic liberalisation in a number of African countries, state controls in authoritarian political regimes suppressed competition and blocked the entry of TNCs, allowing natural resource wealth to be exploited by corrupt national elites. Subsequent to economic liberalisation, TNCs have generally held themselves accountable to higher environmental standards than those established by host governments, basically due to fear of eviction (Reed 2002).

Transfer of technology and skills may improve harvesting techniques and increase processing efficiency and so increase the viability of SFM. Moreover, there may be spillover effects on the forest management of small and medium companies that are linked with foreign-owned companies as suppliers. Gethal, a plywood producer in Brazil with majority US capital, is actively encouraging its suppliers to improve their forest management practices and to seek certification (May and Veiga Neto 2000).

Foreign investors may facilitate access to markets in their home countries or regions and so may bring contact with environmentally sensitive consumers. The implications of this depend on the size of these markets. In Brazil, in the tropical timber sector, certification is more common amongst foreign-owned companies than locally owned companies and the general impression is that FDI is connected to the development of a “modern” logging sector which could be more sensitive to environmental concerns expressed in foreign markets (Young and Prochnik 2003).

One implication is that the companies that are concerned about their reputation in environmentally sensitive markets may shift their investments away from natural forests in the tropics to plantations where requirements for certification can be more easily met without compromising commercial viability and competitiveness. Alternatively, they may shift out of tropical regions altogether. The shift of European companies away from Africa is said to reflect these concerns although issues of political risk and commercial viability are also involved. Issues around forestry TNC investment in some countries can be very sensitive. In Brazil for example, the mere threat of an influx of foreign capital contributed to a raised profile of the illegal logging issue in the country and stimulated an increase in attempts to improve regulatory capacity.

On the negative side it can be argued that the prime interest of some investing companies is to secure raw materials and labour at low cost, making forest mining without any consideration to long-term management the most financially rational strategy for them. International forest product markets do not penalise bad governance or corruption, in contrast to other environmental and manufacturing sectors (Ross 2001). Second, the transition economy literature shows that inferior quality foreign companies are attracted by lax environmental standards (Transition Newsletter 2000); less scrupulous TNCs are more likely than their domestic counterparts to be involved in state capture or public procurement kickbacks (Hellman et al 2002). While there are some examples of responsible forestry TNCs, there is a large literature on the abuses of TNCs in countries where forest governance is weak (Box 4.4).

Box 4.4 Trans-national companies and the race to the bottom

The literature, stemming mostly from international NGOs, is rich in examples of aggressive TNCs the move systematically from one country to another, relocating where environmental regulations are weakest - the so-called 'race to the bottom'.⁵¹ One of the more authoritative studies was by the World Resources Institute and WWF (Sizer & Plouvier 1998). This study highlights differences between 'new' and older TNC forestry investments. It found that newer TNC operations tend not to invest in processing since export logging is more profitable; a high mobility of capital, with equipment rapidly moved from one country to another to take advantage of higher profitability; and frequent use of their own expatriate as opposed to local labour. The study observed that “the new investments have been concentrated in countries with generally weak or outdated environmental and social laws and little enforcement capacity”, for example, Papua New Guinea, Solomon Islands, Guyana, Suriname, Cameroon, Gabon and Equatorial Guinea. These countries are characterised by poor monitoring capacity, inefficient tax collection, lack of auditing capacity, and widespread administrative irregularities.

A notorious case has been that of Rimbunan Hijau in Papua New Guinea (PNG). This company has controlled about 40-50% of PNG's log export trade through the 1990s by means of a Sino-Malaysian cartel involving several clusters of companies, each cluster containing foreign exporters and national, mainly 'local landowner', companies (Filer 1998). These companies were connected through mutual shareholdings, overlapping directorships, and shared office facilities. In 1993, PNG's forest minister complained that Rimbuna Hijau was using its connections to block the implementation of the National Forestry Development Guidelines, at the same time as financing a new daily national newspaper to curry public favour. For Filer (1998), this cartel is the main obstacle to 'rationalisation' of forest management in PNG. He also points out that, unlike foreign mining and oil companies from North America, Europe and Australia, Malaysian timber export companies have not been subject to country of origin environmental pressure groups.

The problem of aggressive TNCs is by no means restricted to tropical or developing countries. For example, EIA (1996) reports how timber companies opposed stricter environmental regulations in Alaska, lobbying against protected areas for boreal forests. There have been clashes with indigenous populations and their property rights in such countries as Australia, New Zealand, Alaska and Lapland, the latter case involving old growth forests valuable to the Sami people for reindeer grazing (Dudley et al 1995).

Hellman et al (2002) found that TNCs with local partners are more likely to engage in state capture, while TNCs with an overseas headquarters are less likely to do so. However the latter are more likely to use public procurement kickback payments to secure contracts. The evidence from international NGO analyses of TNC behaviour also suggests that in some cases governments purposely encourage less responsible TNCs. For example, low forestry taxes, weak monitoring and enforcement capacity, and corruption in concession allocation in Congo-Brazzaville have resulted in foreign logging companies controlling most of the concessions; such policies have allowed a subsidiary of the giant Malaysian company Rimbunan Hijau to gain control over most of the commercial forest in Equatorial Guinea (Forests Monitor 2001). Stricter domestic conservation laws can also encourage TNCs to relocate where effective regulation is weaker. For example, Stone Corporation admitted that it operated mainly in Latin America to avoid stricter rules in the US (Dudley et al 1995).

As foreign investment decisions are often heavily dependent on the availability of finance, the interest in sustainable forest management of the financial institutions can be important. The involvement of official finance institutions, multilateral and bilateral development finance institutions such as IFC and FMO of the Netherlands and export credit agencies should in theory provide scope for looking beyond financial returns to questions such as social and environmental and overall development impact. In practice export credit agencies and to a lesser extent DFIs, have been heavily criticised for their involvement in some controversial forest investments, notably Asia Pulp and Paper in Indonesia (Barr). Both types of agency have been taking steps to adopt more comprehensive ways of assessing their investments in order to address environmental and social issues. But CDC, which was one of the first bilaterals DFIs to develop and monitor ethical principles for investment and which in the past had an extensive forestry portfolio, decided to move out of forestry because of its low returns in relation to other sectors.

IFC has taken environmental issues more seriously than other agencies, and has safeguard policies on forestry and a range of other issues. Its involvement in the forestry sector aims to reduce deforestation, enhance the environmental contribution of forested areas, promote afforestation, reduce poverty, and encourage economic development. The effect of its policy though has been to limit its investments in natural forest management. Between 1992 and 1998, 64% of its total forest-related investment involved plantations and 36% boreal/temperate forest (IFC 2000). There were no investments in tropical forests as IFC's safeguard policy in operation from 1991 restricts the financing of commercial logging operations or the purchase of logging equipment for use in primary tropical moist forest(IFC 2002). The IFC along with a number of leading banks engaged in project finance has recently signed up to the Equator principles. Signatories to the principles seek to ensure that the projects financed are developed in a manner that is socially responsible and reflect sound environmental management practices. The IFC environmental and social screening process and safeguard policies are being used by signatories as a basis for their project review. While this is a progressive step, there is still the possibility that the provisions of the safeguard policies will discourage investment in natural forest in tropical regions.

The involvement of private financial institutions that specialise in socially responsible investment or that have a particular interest in forest management also provides scope for influencing the performance and attitudes of the company. In the case of the Brazilian plywood manufacturer, Gethal, forest certification was one of the conditions required by the U.S. investment fund manager GMO which acquired a majority holding in the company in early 2000 (May and Veiga Neto 2000). Other timberland investors in the US, such as Hancocks are also investing outside of the US, but a prime consideration in choosing and assessing an investment location is political risk. Their preferred investment locations have been developed countries or middle income developing countries in Latin America. Africa, apart from South Africa has received little attention from them.

Foreign invested companies are also likely to attract the attention of socially responsible portfolio investors. For developing countries this is most likely to be indirectly through holdings in the parent company if it is listed as few SRI funds invest outside of developed countries. Emerging market funds that invest in developing country companies directly have not typically taken much interest in environmental and social issues related to their investments but there are signs of change.

Conclusions

Market concentration is difficult to establish with any clarity because of issues related to definition of product categories and markets. There is increasing concentration in some parts of the forest products sector, pulp and paper particularly. However, in spite of the substantial publicity that mergers and acquisitions in the forest products sector have received, the sector is not as concentrated as other industrial sectors. The sector is very heterogeneous though, and for certain product categories such as newsprint and engineered wood products, concentration is more evident.

It is also difficult to assess the implications of changing market concentration for market power. In a globalising market, where production is increasingly destined for export or where imports are readily available, shares of national production are of less concern. Those product categories that do appear to be consolidating at a global level are also the ones faced by consolidation in buying sectors. This applies particularly to the pulp and paper sector.

However, an emerging trend is for processing companies to divest part of their own landholdings and to rely more on outside suppliers for their wood raw materials. This applies to both temperate and tropical regions. This means that the nature of contractual relationships between these suppliers, whether private landowners, outgrowers, or communities will be a determining factor in the attention given to sustainable forest management and in the division of the market benefits from it. The increase in forest landholdings by Timber Investment Management Organisations may positive for sustainable forest management because of the interest of these organisations in holding forests as a long-term asset.

The solid woods industry is fragmented particularly in tropical regions and so to a great extent are its buying sectors. There may be a case that consolidation would be beneficial rationalisation rather than a sign of increasing market power. Arguably, a process of consolidation leading to modernisation of the industry could have beneficial effects on forest management. It would relieve problems of over-capacity, which is affecting profitability and also putting more pressure on forest resources.

Foreign direct investment in the sector appears to be increasing but its importance varies from country to country. In some countries such as the Philippines it has not played a major role, either in the development of the industry or in the depletion of forest resources. The impacts of foreign direct investment on forest management are very country and company specific. In Brazil and Ghana, some of the foreign invested companies are associated with progressive policies in relation to sustainable forest management and efficient use of raw materials. But numerous cases of bad practice worldwide on the part of TNCs have been reported, particularly in countries with weak governance and enforcement capacity.

High political risk in some tropical countries discourages investment by responsible companies with a long-term perspective and an interest in forest management. More attention needs to be given to overcoming these problems of risk. This is the role of development finance institutions and investment insurance agencies, which for investments in other sectors have provided a buffer against these risks. However, they have avoided investment in natural forests in tropical regions. They need to move to a more positive approach which would favour companies that can demonstrate sustainable forest management and promotion of local livelihoods.

4.5 Investments and Capital Movements in the Forestry Sector

By definition sustainable forest management is self-financing. Additional external financing could, however, be justified to cover the incremental costs incurred by forestry operators adopting sustainable practices, to create value for non-market benefits and to counteract those structural incentives that promote unsustainable practices. In many parts of the world unsustainable practices have caused great damage. To reverse these damages and establish SFM on a permanent basis considerable technical and financial input is needed

There is a lack of common understanding on financial requirements for a worldwide implementation of SFM. The estimates of the financial requirements in themselves are rare and/or likely outdated. An assessment, which is often quoted, was conducted during the UNCED process. It stated that US $31.25 billion would be needed annually for sustainable forest management implemented worldwide. ODA was supposed to contribute 18 per cent of it, some $5.67 billion. A few years later, the total figure was revised up to $33 billion per year⁵². Capital equipment and infrastructure was supposed to take 37 per cent, protection of forest services 18.5 per cent and institutional development and capacity building 17 per cent of the total⁵³.

Those figures have been criticized for neglecting compensation for deforestation and forest degradation. Thus, adding the associated disinvestments, the total required financing should in fact amount to $69.3 billion per year. However, this figure has not been without criticism either. In any case, the calculations refer only up to the year 2000 and thus they are probably of limited usefulness to today's policymakers.

Structure of Forestry Financing

Detailed, accurate data on financing of SFM is nonexistent. The figures that are available only refer to the forest sector in general. However, it can still be analysed to reveal the structure and trends of financial flows. There have been attempts to build a picture of financial flows based on secondary sources. One example⁵⁴ refers to the year 1993, when, according to a FAO estimate⁵⁵, ODA channelled to forestry was $1.54 billion, 7.5 per cent of the total forestry financing. Private domestic and foreign contribution was approximated to amount to $8-10 billion, consisting largely of investments in plantations and in processing industry⁵⁶. These figures suggest that some $10 billion of public domestic investments was directed to the forestry sector.

ODA flows are better documented than the other sources of financing in the forestry sector. Estimates suggest that official flows increased in late 1980s and early 1990s from $1.073 billion, to around $2.2 billion (in 1996 US$) in 1990 and 1992. Since then, up to 1997, to which the data extends, there has been a downward trend. In 1996, forestry ODA was $1.3 billion⁵⁷.

Financing for SFM in the Context of Development Financing

Fresh data on financing of SFM is scattered and incomplete at its best. However, to be meaningful, any financing strategy for SFM needs to give due consideration to the financial environment that it will operate in. Sustainable forest management may have its own challenges but requirements in forestry financing may not deviate from the general trend in development financing.

Financial flows to developing countries have experienced well-known changes in the recent decade. The year 1998 was the turning point in capital flows and saw a new era in development finance. Three main developments shaping external financing are decreasing debt stock and increasing private flows, mainly FDI. ODA flows on the other hand are declining overall, and focuses of the contributing donors are changing. For example, in Africa between 1990 and 2000, official flows to agriculture, forestry and fishing decreased by more than half. At the same time, ODA in education rose by 400 per cent, reflecting the changing strategies and priorities in development financing⁵⁸.

The drop in the debt-equity ratio demonstrates that the stock of external debt has fallen while the stock of equity capital owned and controlled by foreigners has risen (Table 1 and 3). The debt-equity ratio for developing countries as a group dropped from 316 per cent in 1997 to 196 per cent in 2001⁵⁹, however, hiding significant variations from country to country. South Asia has the highest amount of debt relative to equity, having a debt stock six times higher than equities. Sub-Saharan Africa and Europe and Central Asia had a ratio around 300 per cent, while Middle East and North Africa approached 400 per cent in 2001. The lowest share of debt relative to external equity was in East Asia and the Pacific, 134 per cent, to a large extent due to China, where external-debt equity ratio was below 50 per cent.

This implicates a highly skewed distribution of FDI (Table 2). Worldwide, the top five countries received 45 per cent of global FDI inflows in 2001⁶⁰ while the share of developing countries altogether was 28 per cent. In absolute terms the rising trend of FDI, both in developed and developing countries has been rather strong. A closer look at the UNCTAD statistics on FDI inflows suggests, however, that the increase in the share of developing countries has not been steady. The annual average during 1990-1995 was 33 per cent, rising up to 40 per cent in 1996 and 1997, and then declining to 16 per cent in 2000. However, a positive development in 2001 was that the share of developing countries increased to 28 per cent when the world inflows of FDI declined by half.

Among the developing countries there are winners and losers. The top three recipients attracted 53 per cent of the net inward foreign direct investment in 2001 (Global development finance 2003). While some countries have been quite successful in attracting FDI, the share of LDCs of FDI in developing countries has declined, from an annual average of 2.3 per cent in 1986-1990 to 1.8 per cent in 1996-2000⁶¹. There was also diversity among LDCs, as 16 of LDCs received more capital inflows relative to gross fixed capital formation than an average developing country in 1998-2000. Nonetheless, at the global level the share of LDCs of total world FDI flows has remained below one per cent⁶²

Differences in the Structure of Financing between Developing Countries

The aggregate financial figures hide significant variation from country to country and do not reveal the critical differences in the dependency on different sources of financing. In absolute and relative terms Latin America and the Caribbean perform well in attracting foreign investments. They received 40 per cent of the net inward FDI flowing into developing countries in 2001, while they contributed 31 per cent to the total GDP of the developing world. They also hold one third of the total external debt of developing countries. On the other hand, the Middle East and North Africa received 3 per cent of the net inward FDI while contributing 8 per cent to the developing world's GDP.

In spite of the increasing role of private equity in development financing, dependence on official flows is still intense, especially in South Asia and Sub-Saharan Africa. While the share of private equity in developing countries as a group, exceeded ODA by three times, the ratio of private flows to ODA was about 50 per cent in South Asia. In Sub-Saharan Africa a stunning 90 per cent of external financial flows came from ODA.

Table 4.9 Selected indicators of external financing in developing countries in 2001, billion US$⁶³.

	Net inward FDI	Net debt flows	Net private flows	Net official flows
East Asia and Pacific	48.9	-12.0	36.4	5.7
Europe and Central Asia	30.1	3.3	30.9	10.2
Latin America and the Caribbean	69.3	11.4	62.8	23.4
Middle East and North Africa	5.5	1.7	8.3	2.0
South Asia	4.1	-0.3	2.9	6.0
Sub-Saharan Africa	13.8	-1.0	11.6	10.2
All developing countries	171.7	3.2	152.8	57.5

Table 4.10 Percentage share of different regions of selected items in the developing world in 2001.

	Net ODA	Net inward FDI	Total external debt	GDP
East Asia and Pacific	13	28	22	28
Europe and Central Asia	18	18	21	17
Latin America and the Caribbean	10	40	33	31
Middle East and North Africa	8	3	9	8
South Asia	11	2	7	10
Sub-Saharan Africa	24	8	9	6
All developing countries	100	100	100	100

Table 4.11 External debt-equity ratios (per cent) and external liabilities (sum of total external debt and FDI liabilities as a percentage of 2001 GDP).

	Debt-equity ratio 1997	Debt-equity ratio 2001	External liabilities % of GDP in 2001
East Asia and Pacific	218	134	65.0
Europe and Central Asia	505	293	66.8
Latin America and the Caribbean	284	162	67.7
Middle East and North Africa	394	371	42.5
South Asia	968	613	30.5
Sub-Saharan Africa	515	303	90.6
All developing countries	316	196	61.7

Apparently, the higher the national income level, the higher the share and absolute quantity of private flows and lower ODA. In 2001, Sub-Saharan Africa received 24 per cent of the net ODA, an amount four times higher than their share of the total GDP of developing countries.

Significant changes have occurred in development financing as reflected in the case of LDCs, although mostly not at pronounced scales. In LDCs as a group, total ODA was significantly larger that FDI inflows, with total ODA three times higher than FDI inflows in the year 2000⁶⁴. Even though ODA remained the largest component of external finance in LDCs, its share declined, both in absolute and relative terms during the second half of the 1990s. The net ODA, bilateral and multilateral in total, declined from $16.8 billion in 1990 to $12.5 billion in 2000 (UNCTAD 2002). Interestingly, in 28 countries, where ODA decreased, FDI was on a rising trend. In only four countries was the trend opposite, showing increasing ODA and decreasing FDI. However, only in seven countries FDI inflow was more than ODA in 2000, demonstrating the crucial differences in the structure of external financial flows among LDCs, and by extension developing countries at large.

4.6 Innovations in Timberland Investments - the Case of the USA

There are very significant changes occurring in the USA in forest and timberland investments that clearly have global implications. Less than twenty years ago financial timberland investors hardly existed. By 2002 financial timberland investors based in the USA held timberland assets valued at more than US$ 11 billion and in the last ten years these investments have grown by 500%.

All this began to change about 15 years ago when some creative financial professionals, many of them people who had not specialized in forestry or forest product industries previously, began to realize, analyze and promote the attractiveness of timberland investments. As Best and Wayburn (2001) state, “Forestland has chiefly been a personal and industrial asset. It is now evolving into a financial asset, owned for its value as a portion of a diversified investment portfolio.”. The source of the investment is not, so far, retail or small scale investors, it is institutions that are investing very large amounts of money in a wide range of investments. These very large funds have full time teams of sophisticated experts who generally do not make direct investments but decide what kinds of funds and which fund manager they will invest in.

Timber Stumpage Markets and Prices

In addition to the steady increase of timber value per volume, if stumpage prices also increase in real terms then this adds another source of increment to returns on timberland investment. During the 20^th century southern and northwestern United States softwood stumpage prices increased at about the same rate as the S & P 500 which measures. That is they increased about 10 times in real terms over 100 years. From 1990 to 2002 softwood stumpage prices in various US regions started in a range of US$ 200 to 300, rose as much as 50% in the mid - 90s, and are currently about 10% higher than they were in 1990.

From the perspective of the long term investor, it appears clear that the trend for the last several decades (or even 100 years) is for stumpage prices of softwoods in the USA to increase significantly, at about the same rate as public equities. There are some similarities in the trends, but they also follow independent tracks. Returning to the point about complexity, there are in fact hundreds of different lumber and related products, each following its own pricing dynamic. For example, each region has its own hardwood commerce with distinct products, buyers and prices. They are sometimes related, such as when the northwest USA log market was severely constrained in the early 90s by environmental action and log prices in southeast Asia also skyrocketed.

The markets and prices that are most important for developing countries are the tropical hardwoods and, for the softwood plantation countries, the pulp and paper markets. The Asian hardwood market has developed to the point where it functions largely as a commodities market. African and Latin American hardwoods are largely traded in a specialties market, meaning that there are many brokers and small scale buyers (furniture and flooring manufacturers are common), and that quality and reliability are just as important as the generic type of lumber. As a general rule, in both industrialized and developing countries, the high quality hardwoods (for example cedro, cherry, Khaya, mahogany, oak, teak, walnut, etc.) attract a price that enables them to be shipped world-wide. Lesser quality or less known species do not at present. The fact that only 1 or 2 % of the volume of a tropical forest in Africa or Latin America can be harvested for an export product is a serious problem for sustainable management. But the much higher levels of commercially valuable species that are harvested in Asian forests have not led to sustainable management there, quite the contrary in most cases.

High quality tropical hardwoods can be the most valuable wood in the world. But generally the international prices of any but the highest quality pieces of the most valuable species are very low. This makes it very difficult for entrepreneurs in developing countries to succeed at sustainable forest management businesses, and underlines the importance of viable markets for environmental services.

Property Values

The timberland invested in is timber on the land, and that land has some property value. Independent of the quality of the forest the land value may be very low or it can be quite high. Investors point out that one can buy large tracts of roadless forest in the Amazon for the incredible price of US$ 20/ha. On the other hand, forestland can become very valuable where it can be used for recreation or residences. Many financial investors in the USA use these higher values of some of the timberland they acquire as an integral part of the investment returns.

The value of property is a double edged sword. On the one hand selling off a piece of timberland acquisition as residential plots can make an investment much more lucrative, in effect increasing the original price if the market is functioning rationally. On the other hand, where property values (and taxes) have escalated, as they have in many areas of most wealthy countries, the timber returns alone cannot justify the investment necessary to acquire the land. This has the potential to effectively remove production forestry from large areas. The whole topic relates closely to conservation easements which are discussed below.

Income and Capital Appreciation

Closely related to the above is that the returns reported include two major components, income (or EBITDDA) and appreciation (or capital). In the funds managed in the USA the income is fairly constant and averages 4 - 6 %. Appreciation, or capital value, is much more variable, going above 20% in some years and being negative in other years. Property or land values vary with economic factors that have little or no relation to timber markets. But, at least in the USA, when stumpage prices rise or fall, timberland property values reflect this. This part of the variation in property value is actually reflecting what investors think future timber prices will be.

Different Approaches

The different approaches taken by various financial timber investment organizations is indicative of their thinking and decision making systems. Some of these organizations focus on one region in the US. Some of them have holdings in all forestland regions. Some invest exclusively in plantations; some invest exclusively in natural forests; and some in both. At least one has mainly concentrated on southern hemisphere plantations. With a few exceptions, their approach is to invest in and hold the timberland but not to invest in harvesting or processing operations. In general therefore they do not invest in vertically integrated operations (timber production, processing and marketing). One can note that the traditional forest products companies do base their operations and investments on vertically integrated operations. These different approaches signify that, in this relatively new field of investment, those most informed and experienced differ in their opinions on where the optimum combination of high return and low risk may be found. Naturally once they have specialized, this tends to continue because their expertise in that specific area grows.

Some TIMOs are divisions of larger fiduciary companies like commercial banks and insurance companies, some are divisions of private investment companies, and others deal only in timberland investment. The major TIMOs and the amounts of timber assets they manage are summarized below in Box 4.5.

US Investments in the Southern Hemisphere

In the southern hemisphere below the Tropic of Capricorn only a few countries occupy a relatively small land mass: Argentina, Brazil, Chile, Uruguay, South Africa, Australia and New Zealand. Softwood plantations here are more productive than anywhere else in the world. Analysis shows average growth rates of 35 m3/ha/yr.

Box 4.5 Timberland Investment Management Organizations (TIMOs)

Hancock Timber Resource Group

1.2 million ha

US$ 2.6 billion

UBS Timber Investments

440,000 ha

US$ 1.2 billion

Forest Investment Associates

US$ 1.1 billion

Campbell Group

320,000 ha, US$ 1.6 billion

Wachovia Timberland Trust

US$ 900 million

Prudential Timber

US$ 500 million

The Forestland Group

220,000 ha

US$ 700 million

Xylem Group

US$ 250 million

GMO Renewable Resources

280,000 ha

US$ 370 million

Founded in 1985. Successful in attracting public funds. Moving towards vertical integration by offering in-house forest management. Mostly in USA (one int. property).

Founded in 1982 as the Boston Company Resource Investments (RII). Moving into Southern Hemisphere investments.

A spin off from First Atlanta in 1982. An independent regional player in the southern USA.

A vertically integrated group in the northwest USA.

The business was acquired with First Atlanta. A regional player in the southern USA.

Also focused on southern USA.

Founded in 1995. Focused on hardwood forests in eastern USA.

Invests in companies that own forests and processing facilities in the southern hemisphere. Prepared a Rain Forest Fund in 2000-01.

Founded in1997 by RII principals and sponsored by Grantham, Mayo & Van Otterloo. Offers diversified funds, focusing on combining natural forests and plantations, both domestically and internationally.

(Adapted from a GMO Briefing Paper)

This is more than triple the growth rate in the southern USA. That, and increasing global demand, has attracted international investment to all of these countries.And in all of them, but led by New Zealand, there is an active market in softwood plantation timberland where they can be bought and sold at any stage of their development.

Source: Forest Tornagaleones S.A (A Xylem Portfolio Company)

Massive investment over the past twenty years has shifted the global timber supply picture. The emerging southern hemisphere plantations now supply about 500 million m3/yr., and this is projected to grow to one billion m3/yr. over the next 100 years when it will equal the supply from temperate forests.

However the price of New Zealand export radiata stumpage has fallen in the last ten years from a peak of over US$ 400/ MBF to US$ 170/MBF (see Figure 4). This has not been favorable for these investments, but it is consistent with economic theory that this inexpensive and relatively risk-free opportunity to produce fiber has resulted in investment, increased supply and reduced price.

There are several important points here. First, there is a zone within a handful of countries in the sub-tropical southern hemisphere which has attracted substantial international investment and almost certainly will continue to do so. This area is now and will increasingly become a major player in global fiber supply. Second, this is a specialized situation. The great majority of countries do not have plantation land with this potential, and they will not receive this kind of international investment. The development of these forests and the investment associated with them is positive for the countries and for the globe overall in that demanded fiber is supplied efficiently, but it does not have much of an effect on the provision of environmental or social services from the forests of the poor countries. It is possible that without these plantations some of the world's natural forests would be harvested for wood fiber. But it would seem much more probable that the absence of the southern plantations would result in increased utilization and investment in northern temperate and boreal softwood forests. The products of the tropical broadleaf forest, mostly fine hardwoods and fuelwood and environmental services (biodiversity, carbon sequestration) and social services (energy, agricultural land, food and medicine for the poor), are quite distinct from the fiber product of fast growing plantations.

Other than the southern hemisphere plantations, tropical plantation projects have attracted regional, but not global, private investment. There are only 3.6 million ha of plantations in Africa and 6 million ha in Latin America. Asia on the other hand has 42 million ha, 32 million of those in China, India and Japan. There are some areas of successful tropical hardwood plantations, notably teak in Costa Rica and Java. These again are fairly specialized situations. It would seem much more viable to manage tropical hardwood production through natural regeneration and extensive systems of management. This is very similar to the management systems now applied to northern temperate broadleaf (hardwood) forests, although the forest types are quite different. And those extensive, low investment, long cycle management systems have attracted significant investment in the USA. But, the key point is that natural forests in poor and middle income countries have not yet attracted any significant international investment for sustainable management purposes.

International Investment in SFM in Natural Forests in Developing Countries

Research for this report has encountered only three significant new international investments in sustainable management of natural tropical forests: the Precious Woods investment in the Amazon, the GMO investment in Gethal Plywood in the Amazon and the Candlewood Timber Group investment in northwest Argentina. There has also been a major initiative for a fund and a major investment in the temperate Notofagus forests of Southern Chile and Argentina which will be reviewed below. There are European investments in Africa and Asian investments in Asia. These investments have been occurring for many years or even decades. In many past instances they have not been sustainable, because of the two major issues reviewed at the beginning of this report. There may be some changes toward sustainability recently, but they have not been independently certified as yet, and would not therefore be eligible for World Bank financing.

Under current conditions, “mainstream” international financial forest investors are not even close to considering investments in natural forests in poor countries. First, when they venture outside of North America, they only invest in the seven southern hemisphere countries mentioned earlier where they perceive the country risk to be relatively low (and some will only invest in OECD countries). Second, even in those countries they only invest in plantations where land titles, yields, returns and prices are much clearer, and where environmental risks, ironically, are lower. It appears that international investment in sustainable natural forest management will not occur on a significant scale unless the system is fundamentally altered and new mechanisms are put in place. Nevertheless, there have been some pioneering attempts, after decades of international discussion, and these will be reviewed below.

The largest and most important initiative to date is the Precious Woods Company (PW). PW started in 1990 with investors based in Zurich and acquired abandoned pasture land in Costa Rica for plantations of teak and a few other valuable hardwoods. PW Costa Rica now owns 7950 hectares of former cattle ranches in that country of which 4600 ha have been planted with trees (3300 ha with teak). PW Costa Rica received FSC certification in 2002, and also marketed the first thinnings from plantations. Teak is one of the world's most valuable woods, and good quality lumber is very valuable. It is increasingly depleted in its native range in Asia. Teak grows well in Costa Rica because the latitude is similar to its native range, and in northwest Costa Rica there is adequate rainfall with a pronounced dry season. In other words teak has very specific site requirements and cannot be grown competitively throughout large areas of the tropics. In Costa Rica, through studies at CATIE and based on wide spread commercial plantation, there is increasingly good information on growth, yield and value. PW uses this information to calculate an estimated annual return on investment of 10 - 11 %, and to include in its consolidated balance sheets the value of biological assets in Costa Rica.

In 1994 PW Amazon acquired 80,000 ha of natural forest in the heart of the Brazilian Amazon near Itacoatiara, and in 2001 the company acquired another 42,000 ha in the same area. In March 2003 another 123,000 ha were purchased, “To ensure that the forested area will suffice in the long-term… to guarantee present harvesting volumes in the future.” (This and subsequent quotations in this section are taken from the Precious Woods Annual Report 2002.) In 2001 PW Para acquired 45,700 ha of natural forest land in the Brazilian state of Para near the mouth of the Amazon River, and in 2002, 30,600 ha more was acquired. PW now owns 321,300 ha of natural forest in the Brazil Amazon. PW Amazon was first certified according to FSC standards in 1997 and since then has been audited and reviewed regularly.

Since regular stumpage markets in the Amazon are not established, and available processing facilities are generally of very poor quality, PW's only option was to establish its own processing facilities. This they have done, including the purchase of a vertical slice veneer mill from a Malaysian timber company that was leaving the area. Sliced veneer is by far the highest value added product for high quality fine tropical hardwoods. They also arrange transport of logs and sawn wood, and they have established their own marketing system in Europe and North America for their FSC certified products. In short, it is a fully integrated operation, and this has several significant implications.

It is striking that in the 2002 consolidated balance sheet the biological assets for Costa Rica are valued at US$ 29 million while those of Brazil are valued at US$ 12.5 million. This indicates that 198,000 ha of Amazon natural forest, at US$ 63 per ha, were valued at less than half of 7950 ha in Costa Rica, 4600 ha of that planted, 3300 ha with teak. (The above is extracted from the Annual Report, the remainder of the paragraph is the author's speculation.) There are probably about 3,000,000 m3 of standing commercial timber on the 2002 Amazon properties, most of it harvestable, and annual commercial timber growth can be conservatively estimated at around 200,000 m3. Thus we have a valuation of about US$ 4 per m3 of standing timber. The Costa Rica plantations contain about 60,000 m3 of growing timber which is conservatively valued at US$ 405 per m3 for prime commercial logs. Teak is inherently more valuable than the average for the fine hardwoods of the Amazon, but it is perhaps double or triple their value, not 100 times the value. The difference lies in: the security of ranch land title in Costa Rica compared to the security of forest land title in the Brazil Amazon; the fact that a stumpage market for teak and other plantations exists in Costa Rica so that the timberland owner may make direct sales; and conversely that one must assume all the risks of managing an integrated operation in the Amazon as well as the costs of developing harvesting access which is already provided in the plantations. However this low valuation also represents a tremendous investment opportunity. If an enterprise can secure title, invest to create access and successfully manage an integrated operation, then the forest timber values are much closer to the US$ 6300 per ha of Costa Rica than to the US$ 25 per ha paid in the Amazon.

It is evident from the Company 2002 Annual Report, a remarkably open and informative document, that PW in the Amazon has been learning a lot of lessons as they have proceeded. First, it has been necessary to acquire much more forest land to achieve sustainable harvest levels sufficient for a profitable business. This is complemented by an executive committee in Switzerland which has provided steady leadership and the financial sophistication for successful relations with investors. For PW Amazon, “The original planning and set-up costs were significant. In 1997 US$ 5.83 million of these intangible assets were written off as an extraordinary item. Additional expense was incurred to achieve the FSC certification.” However by the time PW Para was set up in 2001, enough management knowledge had been acquired that a profit there was achieved in 2002, only the second year of operations.

PW net sales in 2002 were just over US$ 9 million, and net profit was US$ 2.16 million. Also in 2002, in perhaps its most significant financial achievement, PW successfully completed an IPO on the Swiss Stock Exchange. 180,000 shares were offered at CHF 60 per share. “Whereas the SWI (Swiss Performance Index) lost 40% from the middle of March 2002 to the beginning of March 2003 Precious Woods shares remained close to the issue price of CHF 60.” At the end of 2002 the share capital amounted to CHF 81,032,150 consisting of 1,620,643 shares.

It is clear that there has been significant effort and issues associated with land titles and land acquisitions. 48 families, including poor settlers and people with vacation homes, occupied PW Amazon's land when it was acquired, and all have been given secure title, removing that area from PW holdings, as the land was already cleared. The report also states that the Para land transactions have been much more complicated than those in Amazonas state. (Para is more densely populated and developed compared to Amazonas, although these levels are still quite low compared to most of the world.) The company is waiting for legal clarification of “property rights” and “possession rights” prior to finalizing options for additional land purchase. Clearly there are significant legal expenses and risks involved.

Nothing in the PW Annual Report or other documentation contemplates sales of environmental services from their forests. However company management is clearly very aware and proud of the environmental and social benefits their enterprise is providing. For example, “Precious Woods' main objective in Brazil: to succeed, as a commercial enterprise, in conserving the forests complex eco-system”; or, “The forest is managed using methods which imitate nature and maintain its biodiversity.”

A quote to reflect the message of the Annual Report:

“What does an investor acquire with 100 shares ?

• Over 18 ha of Amazon Forest which is then protected from deforestation

• 2850 m2 of reforestations in Costa Rica which absorb almost the same amount of CO2 that is released by a car traveling 20,000 km.”

There are three other major international investments in natural forests in South America that the author is aware of. Trillium is a family corporation that has been very successful in real estate and forestry investments in northwest North America. Based on their vision of the potential for sustainable management of natural forests, in 1993 they acquired 400,000 ha of forest land in southern Chile and Argentina. The natural forest there is a monoculture of Notofagus, southern hemisphere beech or lenga. The species can produce a very fine hardwood, similar to cherry. The vision was ambitious, “Trillium set out to create a world model of truly sustainable forestry and to provide environmental leadership through science and sustainability.” From the beginning the project followed a high profile strategy with announcements of land acquisitions, investments and job creation potential. In both Argentina and Chile, the project met significant environmental opposition. Necessary permits were delayed or denied. Trillium found it necessary to enter into extended dialogue with political leaders and community and environmental groups. In Chile the project's harvesting and facility construction plans have been approved, and in Argentina they are still on hold. Significantly more than US $ 100 million has been invested in the project, and there is little revenue so far, but Trillium management is positive and the possibility of success remains.

Another initiative is being implemented by the Candlewood Timber Group that purchased 100,000 ha of sub-tropical hardwood forest in northwest Argentina in 1998. Around US $ 10 million has been invested in the project. The forest will be managed through natural regeneration, comprehensive inventories and carefully planned low impact harvesting, as is the case with other projects considered in this section. An FSC certification inspection took place in August 2003. Candlewood was started by a group of New York based investors and the former Dean of Yale's School of Forestry who is the chair person of the company. This prominent and well financed group has had difficulty attracting all the capital necessary to completely fund the project. An IFC loan was well advanced until Argentina defaulted in January 2002, resulting in continuing lack of agreement with the IMF and suspension of World Bank loans.

The only other foray into South American natural forests by a US based financial investment group is the Gethal project in Amazonas, Brazil in which GMO of Boston invested in 1999.

The investment of US$ 8 million purchased a majority of the company which owned 160,000 ha of hardwood forest and produced rotary peeled plywood from some of the softer species. The project was FSC certified in October 2000. The New York based Rainforest Alliance published this description, “The right company, doing the right things, at the right time, and in the right place makes a difference in the tropical rainforest.” GMO has a very experienced team of timberland investors, and they are not pleased with the way this investment has gone so far. There have been serious problems with project management and factory management to date that may jeopardize the future of the project. It is worth noting that the experienced and talented GMO team participated in a number of international discussions, and actively sought public funds for some of the appropriate costs of the project that were clearly producing environmental and social benefits. So far they have encountered no source for this and no public funding for a public-private partnership.

A Rainforest Fund

One of the experienced TIMOs in the US is the Xylem Group. Their approach was significantly different than most others because they placed private equity internationally in existing forest enterprises, often ones with forests and processing facilities, where they took a majority position. Since its founding 1994 Xylem has raised and managed hundreds of millions of dollars. In 2001 and 2002 Xylem assembled an excellent team managed by its President to launch its Rainforest Fund. The Fund aimed to raise between US$ 100 to 500 million. The Fund would invest in natural forest and plantation management projects in developing (rainforest) countries, and all projects would receive FSC certification. The Fund received the active public support of WWF and its Global Forests Trade Network. There is a comprehensive 86 page booklet that describes the rationale, justification and operational approach of the Fund. Without systematically summarizing all of its points, it may be said that it is articulate and well-informed and makes the most convincing case possible for this kind of a fund. It cites the return and risk information that was reviewed earlier to make the case for forest sector investment. It then presents the rationale for international timberland investments:

• enhanced returns relative to those available in domestic timberland investments, driven primarily by higher biological growth rates available offshore vs. onshore for similar species;

• attractive entry prices for the acquisition of either productive bare land or existing forests;

• capability to enter markets that are in either the initial pioneer stage or high growth stage of development;

• geographic and market diversification within the timber investment portfolio.

Followed by the rationale for tropical forest investments:

• global market demand for high value, tropical hardwoods increasingly exceeds supply. There is no short term solution to this growing imbalance;

• in natural tropical forests, low entry prices for valuable timber stands combined with the commercial knowledge and acumen required to release inherent asset values creates the potential for significant uplifts in book valuations;

• tropical plantations are more productive than temperate plantations, with growth rates up to 10 times higher than temperate plantations;

• tropical regions generally have lower production costs.

And they also made the argument that market demand and even price premiums were growing for independently certified products, and that certification would bring the support of major environmental groups. For balance, it should be said that the booklet also presents the possible risks of such investments, and that was generally consistent with the risks investors see that have been considered earlier.

Four highly experienced professionals worked full-time more than a year to attract investments to the Rainforest Fund without major success. The Fund was not closed in 2002, but work on it has been suspended.

³³Chomitz and Gray (1996), Lopez (1997), and Amacher et al. (1998) provide econometric descriptions of agricultural settlement and conversion of the modern forest frontier in Belize, Cote d'Ivoire, and the Philippines, respectively. Heydir (1999) provides an historic description of forest use in Sumatra that extends into a description of settlement on the modern frontier.
³⁴This chapter draws heavily on a forthcoming volume on the economics of global forestry by William Hyde (Hyde 2003).
³⁵Sudden sharp increases in exogenous demand are less common than sharp decreases. They create a similar two-stage adjustment process. Producers in the exporting country increase production by adding variable capital and labour to the existing fixed capital and fixed area of mature forest. During the first stage, operations would be neither as economically efficient nor as environmental sound as they had been, but production would increase. Eventually economic efficiency and better environmental performance would return, as the new fixed capital becomes available and the newly managed forests attain maturity.
³⁶Others (Mahar 1988, Mahar and Schneider 1994, and Schneider 1994) draw similar conclusions.
³⁷Since the 1970s most countries allow the market to determine the values of their currencies.
³⁸A few developing country governments have imported agricultural products and held food prices artificially low in an effort to support urban populations and encourage industrialization. These food price controls probably shift some farmers out of commercial agriculture and back into more rudimentary subsistence agricultural activities.
³⁹Of course, the greater difficulty lies in the transactions costs involved in changing the agricultural support policies and institutions in North America and Western Europe.
⁴⁰See Boyd and Hyde (1989) on the effects of the US minimum wage on loggers and workers in the forest products industries.
⁴¹Time, or the cost of growing capital, is only one input to forest management. We could also trace the effect on the labor and manufactured capital inputs for forestry. However, these are lesser inputs for most forest management and the effect on harvest time has easily been more important. The simplest way to trace any of these effects would be to enter a term for an accumulating tax on production Q(.) in the Faustmann formula, eq. (3b.3), and to determine the resulting changes in the optimality conditions for the harvest period and for labour and manufactured capital inputs, eqs. (3b.8) and (3b.9).
⁴²Landell-Mills and Porras (2002) provided further evidence of the private market nature of some classes of watershed management. They identified over 180 cases of markets for watershed services from countries all around the globe and in a multitude of local arrangements.
⁴³see for example WWF International website at: http://www.balancedtrade.panda.org
⁴⁴see FAO's website at http://www.fao.org/forestry/foris/webview/forestry2/index.
⁴⁵Source: FAO Investment Centre Environmental Impact Guidelines (November 1999). http://www.fao.org/forestry/foris/webview/forestry2/index.jsp
⁴⁶http://www.balancedtrade.panda.org⁴⁷http://europa.eu.int/comm/trade/sia/seminar/index_en.htm⁴⁸http://www.iaia.org⁴⁹Calculated from US DOC SCB Sept 2001 and November 2001
⁵⁰News Release October 22 2002 www.weyerhaueser.com
⁵¹International NGOs that have studied forestry TNCs include: Forest Monitor (2001), the Environmental Investigation Agency (EIA, 1996 and EIA and Telepak, 2001), Greenpeace (2001), Global Witness (2001), WWF (2002), Friends of the Earth (Glastra, 1999), and FERN (2001)
⁵²Workshop on Financial Mechanisms and Sources of Finance for Sustainable Forestry. Pretoria, June 1996
⁵³Chandrasekharan, C. 1996. Status of financing for sustainable forestry management programmes. In: Proceedings of the Workshop on Financial Mechanisms and Sources of Finance for Sustainable Forestry. Pretoria, June 1996.
⁵⁴Moura Costa, P., Salmi, J., Simula, M. and Wilson, C. 1999. Financial mechanisms for sustainable forestry. UNDP, Programme on Forests, Working Draft.
⁵⁵FAO. 1995. Forest Resources Assessment 1990. Global Synthesis. FAO Forestry Paper 124, FAO, Rome.
⁵⁶Chandrasekharan, C. 1997. International cooperation and resource mobilization for sustainable forestry development. Proceedings of the XI World Forestry Congress. Volume 5. pp. 365-387. Antalya, Turkey.
⁵⁷Madhvani, A. 1999. An assessment of data on ODA financial flows in the forest sector. UNDP, New York.
⁵⁸UNECA. 2003. Economic Report on Africa 2003. Addis Ababa, Ethiopia.
⁵⁹World Bank. 2003. Global development Finance. Striving for Stability in Development Finance. Washington, DC, USA.
⁶⁰UNCTAD. 2002a. World Investment Report. Geneva, Switzerland.
⁶¹UNCTAD 2002b. FDI in least developed countries at a glance. Geneva, Switzerland.
⁶²UNCTAD. 2001. FDI in least developed countries at a glance. Geneva, Switzerland.
⁶³World Bank. 2003. Global development Finance. Striving for Stability in Development Finance. Washington, DC, USA.
⁶⁴UNCTAD 2002. FDI in least developed countries at a glance. Geneva, Switzerland.

Year	Average Mill Capacity ('000 tonnes per annum)
	Pulp	Paper and Board
1960	45	30
1970	90	70
1980	145	115
1993	225	185
2000/01	269	231