Augmenting SNA Forest Sector Accounts with Satellite Accounts

Having discussed the basic Forest Sector reports in a nation’s standard SNA, the purpose of this chapter is to show how the SNA can be augmented and expanded by utilizing what are commonly referred to as satellite accounts. The purpose of developing these accounts is to better identify more detailed information, which is often obscured by the aggregate nature of the broad classification system utilized in SNA. After discussing the relationship between the SNA and the SEEA, specific features or capabilities of satellite accounts are presented and discussed. The features or capabilities include the following.

The capability to disaggregate and reformat national accounts related to forest activities in order to: 1) identify economic activities which represent inter-industry/inter-sectoral linkages; and 2) identify forest management expenditures for stand cultivation, forest protection and forest maintenance activities. These capabilities will ultimately allow policy analysts to use input/output models for forest sector policy analysis.

The capability to develop specific accounts for: 1) forest land, 2) timber, and 3) carbon, to more readily allow examination of important issues such as sustainability and carbon sequestering.

The capability to analyze the environmental effects of other industries on forests.

The capability to examine the environmental effects of forest sector activities both on the forest sector itself as well as on other economic sectors.

One important distinction between SNA classifications and the potential classifications developed in satellite accounts comes from the basic idea of assets used in SNA. Only economic assets, i.e. assets owned by some entity or unit from which economic benefits are derived by their owner(s) by holding them, using them over a period of time, or by selling them, are within the asset definition or boundary of SNA. Economic benefits recognized by SNA are incomes (derived from production) and holding gains and receipts from the sale of assets. Thus, "noneconomic" forests are not included in SNA but may be included in a detailed satellite account. Noneconomic forests may include areas that are unprofitable to log, or lands that are protected with no commercial gains accruing to the owners.

A second important element associated with the idea of economic assets is derived from that basic classification of assets being either produced or non-produced. Non-produced economic assets are those assets that are not the result of a production process, in the sense that no inputs have been used for their creation. SNA treats land as non-produced in all cases. Land may be cultivated or non-cultivated. In the case of forests, the term cultivated means that some work or capital has been used for producing the output grown on land. Products grown on cultivated forest land are generally treated as produced; products grown on non-cultivated land are treated as non-produced. The distinction of produced versus non-produced products coming from cultivated versus non-cultivated forest lands utilized in the SNA/SEEA framework is analogous to the FAO approach of natural versus plantation forests.

Another important facet of the SNA/SEEA relationship lies in asset classification. Assets may act as fixed assets, valuables or inventories. In SNA although all economic assets are expected (by definition) to produce economic benefits to their owners, they do not all act as "fixed" assets. Fixed assets are only those produced assets that are used repeatedly or continuously in production processes for more than one year. Because standing cultivated timber typically yields a product far less frequently than a year, it is classified as inventory (work in progress) rather than a fixed asset. These categorizations of forest assets are summarized as follows.

Rather clearly the SNA classification system must be adjusted in SEEA to better describe important forest values. If, for example, timber sequesters and stores carbon annually and this form of annual service takes place on noneconomic forest land as well as economic lands, these kinds of forest functions and values won’t fit the SNA asset classification system and must be modified in the satellite account.

In order to clarify the relationship between SNA and SEEA in the accumulation and balance sheet accounts, the following key forest sector items found in these account are examined.

Opening and closing stocks

Gross fixed capital formation

Changes in inventories, which is the balance of natural growth and fellings for cultivated timber.

Consumption of fixed capital

Acquisitions less disposals of non produced assets

Additions, i.e. natural growth for non-cultivated timber

Depletion, i.e. fellings from non-cultivated timber

Degradation of forest land due to acidification

Catastrophic losses of timber (forests fires)

Changes in classification and in particular changes from forests land to other land.

These key items are shown here in the structure of accumulation and balance sheets used in SNA. Again, it should be clear that these classifications are for economic assets so that further detailed accounting for noneconomic forest assets must be done in the SEEA.

There are two complementary and interrelated descriptions of economic forest-related flows and stocks that are integrated into SNA forest sector accounts.

Current (production, distribution and use of income) accounts describe forest-related economic transactions in terms of activity flows that generate products and incomes. -

Balance sheets describe the stocks of assets at the beginning and the end of a reference period (typically a year). All of the changes that affect the opening value of stocks from the beginning to the end of the accounting period are Accumulation accounts and balance sheets describe forest-related assets in terms of described exhaustively in order to fully account for changes in asset values.

Table 4 Structure of accumulation accounts with balance sheets

Balance sheets are related with economic activity flows through the capital account. Coupling the activity flows with other changes that affect asset values such as "other changes in volume" or the "revaluation" accounts allows an exhaustive manner of relating how production, consumption, accumulation and depletion affect capital assets over time. Because these accounts concentrate on economic assets in SNA, the SEEA will allow a broader array and manner of integrating forest capital and activity flows.

Linkages with ISIC

According to the international standard industrial classification (ISIC), the forest industry consists of all units whose main activity is to grow timber in plantations, fell and harvest wood and other forest goods such as cork and gum etc. and produce services related to forest management, protection and maintenance. However, data may often be reported in such a way that actual forest activities are reported in other industries. For example, an integrated company with woodlands and sawmills or pulp mills may have some or all of its activities erroneously reported in the wrong industry. Therefore in order to identify all output related to forests and estimate the corresponding values it is necessary in a first stage to scan all relevant positions of the ISIC and identify all forests-related output produced as secondary activity. In national accounts, the output is deemed to cover all goods, including goods produced for own final use or those goods produced and consumed in the same sector. However, in many countries, limitations in data reporting imply that some output is unreported. Some of these. frequently acknowledged unreported forest sector activities are fuelwood, game meat, medicinal plants and other foodstuffs. The value of this output should be estimated and included in national account figures. The SNA recommends that output produced for own final use be valued at the average (basic) prices of the same goods (and services) sold on the market (provided they are sufficient quantities sold to enable reliable estimates). If there is unreliable product price data, goods for own consumption should be valued at the total cost of production per unit internally consumed. Because unpaid labour may account for a large part of the inputs, it is important to make some estimate of its value using wage rates paid for similar kinds of work in the local labor market.

The output from forests as modified from SNA may be presented in a supply table such as the following.

Table 5 Supply table for forests products

In addition to help identify all outputs in relation to forest, this table could be used to examine inter-industry/intersectoral linkages and their policy implications for sustainable forest management.

In a second stage value added and all factors’ incomes corresponding to the output must be assessed. In many cases, as inputs for secondary activities are not explicitly described, estimations are necessary.

Table 6 Value added and incomes from forests

As a complement of output and incomes assessment, and for sake of consistency, it is necessary to identify the uses of the output: intermediate or final consumption, capital formation, exports, etc.

Forest Land Account

The purpose of the forest land account in the SEEA is to describe changes in land use in both physical and monetary terms. Since the SNA includes only economic forest land without explicit reference to forest cover, noneconomic forest land is added to the forest land satellite account. A matrix could be constructed in the satellite account to report both total land areas for items in the classification system as well as changes between classes when land is allocated from one use to another within the accounting period.

The land classification used in SNA relies on the following broad categories

Land underlying buildings and structures

Land under cultivation, including land under plantations

Recreational land, amenity land, parklands and pleasure grounds

Other land not elsewhere classified

Rather clearly, forest land and grazing land are indistinguishable in SNA so a satellite account may be developed to better enable policy-makers to keep track of these environmental/economic assets. SEEA reorients the SNA classification towards environmental concerns by distinguishing forest land as an item of economically used land. In order to describe all land, it also integrates non-economic land. Non-economic forest lands include lands withdrawn from commercial exploitation and land economically inaccessible for logging and development purposes.

The following suggested forest land classification system integrates categories from SNA, SEEA, FAO and UN-ECE.

Forest land (UN-ECE classification of land)

Economic forest land (forest land recorded in SNA balance sheets)

Cultivated (plantations and semi natural forests in FAO classification)

Non cultivated (natural forests and secondary forests in FAO classification)

- Non economic forest land

Protected

Non protected

Other land (land underlying buildings… agricultural land, etc.).

Several additional approaches for classifying land may be desirable. For example, lands designated as critical wildlife habitats or land classifications based upon cover or habitat types maybe of sufficient importance to include in satellite accounts. Likewise, classifications of forests based upon functions, health status or management regimes may also be included in a satellite account.

Table 7 Changes in land use/cover: basic physical table

Because a forest land account is a form of fixed asset account, it includes an accounting of the changes between opening and closing land stocks. Balance sheet changes in land use (or cover) are (even if only partially, and/or indirectly) related to economic activities. These changes are recorded as occurring outside the production boundaries (other volume changes in SNA and other accumulations in SEEA). Areas affected during the period by catastrophic events (fires, windstorms, insect infestations, etc.) which do not modify the classification of land may be recorded as a "memoranda" item, and linked with timber balances.

The above mentioned table allows for the description of the main changes that may affect forest areas:

Changes in area due to deforestation and afforestation, natural colonization or regression: are environmentally significant from many perspectives. For example, these changes can be linked in the SEEA to changes in biodiversity, carbon uptake and storage, or erosion.

Changes in cultivation status may be linked to other SEEA accounts both in terms of changes in timber production as well as potential changes in biodiversity.

Changes in economic status may occur when a remote natural forest initially considered as non-exp loitable changes as a result of technological change, resource prices or infrastructure investments such as ports, rail accessibility or road developments. Conversely protection of a forest area may change its status to "non economic", if severe restriction are imposed. Decrease in prices may have the same effect.

The idea of economic appearance and disappearance of economic assets was discussed in section 2.3.2 of this report. A similar approach to the appearance/disappearance of noneconomic forest resources, which have environmental significance, could also be included in the SNA. This could be applied to the listing, delisting, recovery or extinction of species. It might also be important in the area of noneconomic lands as well.

Similar tables may be drawn, in accordance with other environmental concerns. For example, balance sheets could be constructed for important forms of forest degradation indicating the area affected and changes in area from previous periods.

Under the SNA, the value of land is separated from the value of standing timber and the extension of the separation of asset values is a logical extension in forest sector satellite accounts. As a general rule, forest land is comparatively low in value relative to other land uses. This land value is derived from the capacity of land to produce crops of timber. Often forest land ownership caries with it the right to allocate land to other uses so that ownership may be seen as reflecting a variety of land use options which are also of value. Ideally, forest land values in SEEA are based on actual land transactions. However, forest land transactions often include land and standing timber in a transaction "package" so that values of the two assets must be separated for the different entries in the SEEA. If there is inadequate data to separate the asset values, SNA procedures recommend that the combined value be entered into the asset account for the highest valued asset.

The value of cultivated forest land. This value may be estimated by utilizing yield productivity estimates coupled with management regimes, production costs and an optimal rotation age using the SNA indicated discount rate. Results should be compared whenever possible with actual land prices as well as with prices of marginal agricultural land.

Value of non-cultivated forest land can in general be expected to be even lower than that of cultivated forest land. Growth in natural forests may also be low. The main value of non-cultivated land lies in the potential to change its use from non cultivated land under natural forests to another use (agriculture, including forestry or built-up land). Of course the possibility also exists to reclassify this kind of land as some form of protected land.

The earlier discussion of forest asset balance sheets focused exclusively on physical estimates of land. The table below represents a monetized balance sheet for economic and noneconomic land.

Table 8 Changes in land use/cover: basic monetary table

Note that land is the only non-produced asset for which investments in the land are added to the value of the stock. In contrast, timber investments such as for reforestation is recorded as capital formation for the forest industry and hence is not added to the value of the timber stock.

All changes in value are recorded in the column corresponding to the category of the land in the opening balance sheet. When land moves from one category to another during the period, the row "changes in classification" records the shift of values between categories.

When forest land is deforested and the land is allocated to a higher value use, the increase in value is registered in the "deforestation" row. In this case, it could be argued that one has also to take into account the value of the decrease of carbon binding capacity (or loss of biodiversity, etc.). Present benefits (first the benefits from selling the timber, second the benefit from changing the use of the land, expressed by the higher value of the land) are exchanged against the future losses, including the decrease in the carbon binding capacity. The nature and the value of this additional item is however controversial. If the present value of the benefits were lower than the discounted value of future losses, it would be preferable not to deforest. If, however, discounted benefits were less than discounted costs, the balance would tilt toward deforestation. The issue is complicated by the choice of discount rates, problems in valuing nonmarket phenomenon and possible irreversibilities in biodiversity.

Forest Management/Protection/Maintenance Expenditures Account

Before discussing satellite account entries for forest management expenditures, it should be useful to briefly examine the nature of forest uses and costs as they relate to various previously discussed classes of forest lands.

Table 9 Forest Costs, Land Types and Activities

Very often a single forest management entity has both economic and noneconomic forest land. For example, even a timber corporation may own areas covered with water or rocks and ice in addition to cultivated and non-cultivated forest land. Furthermore, many public forest land agencies throughout the world manage their lands around the principle of multiple-use. Thus, for example, considerable time and effort often is devoted to producing high quality recreational experiences on public lands as well as the more commonly expected forest products. To some extent, the multiple uses are "joint products" (a sheep rancher produces meat and wool so that production costs may be allocated somewhat arbitrarily to more than one cost center). When economic and noneconomic forest lands are spatially intermingled, forest protection activities such as insect control or fire suppression have distinct neighborhood qualities. For example, failure to control a wilderness fire may mean that valuable nearby plantations or other nonforest assets are at risk. And while timber management invariably requires some means of transporting logs to mill sites, recreational lands also have transportation networks of roads and trials.

Forest sector expenditures which have the purpose of securing expanded benefit flows typically include inventory, research efforts, biodiversity protection, fire protection, resource planning, transportation expenditures (construction and maintenance), facility construction and maintenance as well as the usual silvicultural expenditures. Costs of these activities are not included in the intermediate consumption of forest products and must be treated as additional intermediate consumption in order to calculate net value added.

Table 10 Forest Management Expenditures and Net Value Added

Net operating surplus of forest industry		500M$
Government timber supply expenditure:		250M$
of which administration	40M$
Aviation and fire control	20M$
Resource access	5 M$
Forest management	185M$
Rent		250M$

The importance of these costs in terms of their role in calculating value added can be better seen by examining summary data from the Canadian Province of Ontario.

If substantial amounts of the forest management expenditures in the above table are designed to secure favorable flows of nonwood forest products and services, then it would be prudent to allocate some of these costs to other forest product accounts.

Recreational activities are indirectly included in the SNA as tourism expenditures. The question of which expenditures are considered as integral parts of recreational expenditures is very relevant. Many studies have indicated that recreationists are willing to pay more for site access than they are frequently required to actually pay. Willingness to pay for the recreational experience has been estimated in countless cases. Several caveats are worthy of mention as a set of precautions in using willingness to pay estimates in satellite accounts. Willingness to pay is a formal way of estimating consumer surplus. There is also consumers surplus for market goods but the usual assumption of recreation value studies is that the impact on market goods prices and resulting consumers surplus when forgoing the production of market goods in allocating a site to nonmarket use is sufficiently small that it is negligible. SNA does not include consumer surplus for market goods and services. Rather, the price of the good or service does not change with the activity level. Since SNA is a macro or regional level of analysis, and one could analyze the level of production of recreation activities, it would appear that recreational values should be consistent with values in other sectors being analyzed. Thus, willingness to pay based on the consumer’s surplus associated with a single fixed site may not be an appropriate value in a forest sector satellite account.

Timber accounts are generally first drawn in terms of physical volume estimates such as cubic or board measures. They describe all timber, whatever the economic classification. An example of a simplified timber account is presented below (Spruce volume from Nordic natural resources accounting). It should be noted that all volume estimates are not estimates of biomass but rather are estimates of a useable portion of the wood stem. Each volume measurement has some assumptions about the minimum log length, stump height and minimum small-end diameter of the log. These merchantability standards can change over time so there may also be entries in timber accounts that capture changes in growing stock owing to changes in measurement standards.

Table 11 Timber Account

	Timber (Mm3, inc. bark)
Opening stocks	2194.6
+ annual growth	85.4
- fellings	64.4
Of which net removal	60.9
Of which silvicultural waste	3.5
- natural losses	2.1
Closing stocks	2213.4
Changes in stock	18.8

This simplified account may be complemented by specific tables describing the uses of timber removed from forests as intermediate consumption by the various wood industries and as final. consumption (mainly fuel wood) by the households.

Concerning the valuation of timber stocks and changes in stocks, two cases are to be theoretically distinguished: produced timber and non-produced timber.

In valuing produced timber, two questions arise, how much is the standing timber worth and what is the value of the annual growth increment? The value of the annual growth or increment is estimated by distributing the value of the finished product (mature standing timber) in proportion to the costs incurred in each period. Added to the costs is a mark-up for the estimated operating surplus or mixed income? From a practical perspective, assumptions must be made concerning the future value of standing timber, at the time it will be felled as well as about the total costs which will be incurred throughout the life of the timber. Although present prices are used, this calculation is rather uncertain. Several methods have been employed to assist in this calculation. The simplest approach involves multiplying the stocks and flows in physical quantities by the stumpage price (see below). Based on the present value principle, this simplification assumes that as a first approximation the need for discounting future receipts is counterbalanced by the future growth of the timber. However, as shown by Vincent and Hartwick (1997), in most cases it is theoretically inconsistent and introduces a bias. It may be preferable to make explicit assumptions about the likely felling date and to discount the anticipated receipts.

Valuing non-produced timber is somewhat different from valuing produced timber. Recall that growth of non-produced timber was not considered a product of management. As a result, it is unnecessary to estimate the value of the growth increment of timber on uncultivated forest lands. Were this not the case, the accounts would yield the false impression that timber investments were more productive on non-cultivated lands than on cultivated ones since growth would be acquired without any distributed cost outlay. Stocks can be valued either by multiplying volumes by stumpage prices or by calculating a present value with the same assumptions about future harvest dates, as is the case of fellings from cultivated lands. Depletion (the decrease in value due to fellings) is given by the value of the fellings; the change in the net present value of the stock of timber between the beginning and the end of the period, is given results from depletion and holding gains.

In all cases the most important information to be collected is the value of fellings or stumpage price. In a surprising number of instances a competitively determined stumpage price is not available because markets are ruled by administered prices or because of other institutional arrangements which limit the generation of accurate stumpage value information. In instances such as these, stumpage must be calculated rather than collected. The residual value approach to appraisals may be useful in these instances because it estimates stumpage value by starting with the value of products produced and reduces this amount by the costs of converting standing timber into wood products. (A normal profit and risk margin is also normally subtracted from product values as well). The left over "residual" is the estimated value of timber. The process may be simplified somewhat if good roadside prices are available. Prices.have to be calculated or collected for the various species of wood. Corrections may be necessary in order to take into accounts exceptional events, such as tempests, fires etc, before arriving at a significant market price for standing timber, and hence the value of fellings.

One must keep several things in mind when estimating the value of the entire economic timber stock. When using stumpage prices to estimating the value of timber stocks, there are normally substantial differences in the qualitative characteristics between mature timber being sold and growing timber that will be harvested in the future. Perhaps most clearly, mature timber is typically larger with fewer logs per unit of volume. The costs of logging typically are less per unit of volume for larger trees than for smaller ones since there are simply fewer pieces to handle. Also larger trees are typically manufactured into more valuable products than are smaller ones. Thus, if stumpage prices are used to value the entire economic stock, diameter adjustments can be used to avoid overstating the value. Jackson and McQuillan (1979) demonstrated how to do this with statistical appraisals. Logging costs can be adjusted in the residual value appraisal system for the number of logs per unit of volume logged. This in turn can be used to adjust the value of standing stocks to reflect broad qualitative differences from the typical timer being sold.

In this section we turn to an important example of integrating non-SNA items with basic SNA information in a forest sector satellite account. Forests provide socially important climate modification services by binding carbon released by other sectors in the economy. These services serve at least in part to curb global warming because carbon dioxide is one of the most important greenhouse gases. Various studies show that young forests bind the most carbon but older forests store the most carbon. As forests mature, decay of stored carbon increases so that eventually some form of rough balance between carbon uptake and release eventually occurs. As an example, some authors consider see climax forests as carbon reservoirs but not necessarily net carbon sinks. Carbon balances for the stocks of trees are generally calculated on the basis of the forest balances or stocks and stock to carbon coefficients. An example using spruce inventory data from Nordic forests is presented below.

The dry matter content of spruce is estimated in 387kg/m 3 wood. The carbon content in the dry matter has been estimated in 0.519. The volume of standing timber is converted to carbon content by using a conversion coefficient of 0.2 between cubic meter and tons of carbon.

Table 12 Balance of carbon in Nordic countries (spruce)

	Timber (Mm3, inc. bark)	carbon (Mt)
Opening stocks	2194.6	440.8
+ annual growth	85.4	17.2
- net removal	60.9	12.2
- silvicultural waste	3.5	0.7
- natural losses	2.1	0.4
Closing stocks	2213.4	444.6
Changes in stock	18.8	3.8

Major uncertainties affect the description of carbon binding and storage. Carbon binding should not be reduced to a simple transformation of the standing timber stock since the amount of stored carbon is not a simple linear transformation of the standing stem volume.

Standing volume is an estimate of the commercial product content of the standing tree and excludes small trees, branches, roots, or other forest plants that together with the decaying liter constitute the stored carbon.

There are carbon transfers between the over-ground forest biomass and the soil. A portion of the carbon bonded in above ground vegetation as well as roots become organic soil as plant parts die and decay. Carbon is stored in the soil for time periods. Various authors suggest that the share of carbon bonded in stem of the tree is no more than 15% of the total amount of carbon bonded within the entire ecosystem. The major part (73%) is bonded in the soil (at a depth from 0 to 30 cm).

Table 13 Distribution in % of the Total Amount of Carbon Stored in Forests Ecosystems (from Nordic Natural Resource and Environmental Accounting)

Finally some of the carbon bonded through forest growth stored in the forest ecosystem is removed when trees are felled and logs are removed from the forest. The longer the life of this wood and processed wood products, the bigger the carbon reserves they represent. The carbon is preserved best in sawn goods and building material for which the decomposition cycle is around 80 years. The carbon content in paper products and board is on average released within two years. Certain paper and board products will however, are kept for centuries, amongst other places in libraries and archives.

As a conclusion, it is difficult to make an exact assessment of the annual net carbon binding by the forests and of the extent of changes in the overall reserve of carbon in forests and wood products. It is also difficult to give any reliable estimates of factors, which increase or decrease this carbon reserve.

Valuing carbon-sequestering services presents some interesting challenges. The cost of abating or avoiding CO₂, N₂O and CH₄ emissions which are the pollutants causing acidification damages of forests is estimated using end of pipe technology, or substitution costs. These costs should be deducted from the value added of the causing industry. Another approach values the future damages "avoided" through a net present value calculation. This method assumes that the impacts of temperature increases based on the emission of one supplementary ton of CO₂ during the period are known, and that a value (a cost) can be attributed to them. It also assumes that a discount rate has been chosen. A comparison of three such costs computations has been performed by Nordhaus (1993). Should the carbon dioxide concentration double as it is expected to by the year 2030 (unless today's rate of increase slows) the average temperature on Earth is expected to rise 1-3° above today's level. Some studies calculate that climate change costs will rise to 1.0-1.3% of GDP annually. A corresponding value with this change in GDP for carbon is US$ 3.3/t CO₂. In other words, carbon value through this method is approximately 10 US$/t carbon. This falls in the low end of the range of values found in the literature, namely from 5 to 125 US$ per ton of carbon emitted.

In these conditions, the value of the sequestration by forests of a ton of carbon depends on the set of assumptions made regarding the physical effects of sequestration and their economic expression. Of significant importance are the rate of discount utilized, and the anticipated level of GDP in the future in a particular geographic areas. The estimated carbon sequestering.value of forests are therefore quite imprecise. However imprecision is not an excuse of failing to agree on an approach for estimation.

Forests are multifunctional ecosystems. Several economic activities take place in the forest that have direct and indirect impacts on the ecosystem itself, as well as on human well being. Direct impacts are closely linked to the use of forests. Indirect impacts relate to qualitative aspects of forests, as providers of environmental services and amenities.

Table 14 Environmental Activities and Forest Impacts

Table 14 lists economic important activities and their direct and indirect impacts on the forest.

Direct Impacts

Direct impacts are usually measured in physical and monetary terms. The valuation used, as mentioned earlier, is market based. Monetary values are obtained by multiplying the physical quantities by a per unit value. Direct impacts include the following items:

Depletion/Accumulation of timber is defined as the changes in economic and non-economic stocks. Non-economic stocks are included because of, for example, illegal logging in protected forests. Furthermore, non-economic forests often continue to be used for non-timber products. The depletion can also be noted as exceeding sustainable harvests. This leaves the issue of "hard sustainability" to policy makers in each respective economy. Governments often estimate sustainable timber harvests on the basis of the regenerative capacity of the forest for timber production. Depletion caused by the above activities may constitute a significant part of the overall depletion and it should be estimated and its cost deducted from the value added of forestry and agriculture industries, respectively.

Depletion/accumulation of non-timber products is harder to measure. Production of non-timber products is included in the production boundaries of the 1993 SNA (i.e. berries and wild mushroom constitute output of agriculture, peat production output of mining, game related products output of hunting). As mentioned earlier, depletion measures the loss of the income generating capacity of the resource. Conceptually, depletion of non-timber products should be measured both in physical and monetary terms and treated as consumption of natural capital, as it is the case of timber. In practice, however, data on stocks and as a result, sustainable yield are usually not available. In some areas of the world, there is a significant inverse relationship between timber stocks and hydrological runoff such that logging may augment hydropower production. This would be a separate fossil fuel offset since hydropower production does not require the burning of fossil fuels.

Given the difficulties of certifying the sustainability of forest management operations, some countries are developing Criterion and Indicators (See Canada Council of Forest Ministers, 1997). These indicators can be included in the SEEA at least in terms of physical inventories and gains or losses.

Indirect Impacts

Forest uses can cause ecological damage. Actual expenditures made to protect the forest and to rectify the damage are included in the SNA and can be separately identified to obtain an indication of the impact on the environment. In most countries, however, such expenditures may not be a good indicator of the changes in the market and non-market values of the environmental services provided by the forest.

There are clearly benefits flowing from the existence of the forests. The real costs associated with the provision of the environmental services are not captured by the pricing mechanism of the market, because they are not attached to the cost of exploiting these assets for their timber values (Joisce 1996). Two approaches that attempt to capture both the costs of direct and indirect impacts have been advanced in the SEEA. They are maintenance costing and contingent valuation and other demand side valuations. Maintenance costing focuses on the impacts of production in assessing the costs of keeping the forest intact. Contingent and related valuation of the demand/benefits/damage side of environmental services, on the other hand, attempt to measure the loss or generation of ecological and environmental consumer services. Both approaches attempt to find, in principle, market values for measuring environmental costs and benefits (Bartelmus 1998).

Maintenance costing searches for market values of factor inputs, required for the potential restoration or avoidance of environmental impacts caused by economic agents. Contingent and related demand-side or use valuations attempt to measure preferences for maintaining, restoring or improving environmental, as if markets exist for these services.

Contingent valuation thus reflects the environmental cost borne by ecosystems and individuals. However, maintenance cost appears to exhibit characteristics of both the cost-caused and cost-borne concepts, if restoration costs are included. Restoration costs represent the expenditures that society would have incurred to restore environmental quality lost in the past to desirable levels, expressed by "sustainability standards". These quality losses can of course not be tracked back to those who caused damages through emissions in past accounting periods or imported transboundary pollution from other economies. While reflecting maintenance cost concept of restoring environmental assets, restoration costs thus also represent the environmental costs accumulated and borne by society over time. The environmental cost accumulation represents an environmental debt. Environmental debt incurred by past generation indicates the responsibility of the current generations for environmental effects that would have to be borne by future generations. In its reference to an accumulated environmental burden, this concept is not strictly comparable with production and income aggregates.

Maintenance cost is the imputed cost that would have been incurred if the environment had been sustainably used (undiminished future use). That, under the maintenance cost approach, is the cost of "protecting" the forest. Such costs would cover both direct and indirect services of the forest. Differing level of maintenance provide different cost levels. The cost of avoiding the loss of certain services would also affect other services provided by the forest. Therefore, when valuing each service separately, one should take special care not to double count the costs. Also, given the shortcomings of the different valuation techniques presented, it is important that physical data be maintained in detail to ensure that non-economic issues are not overlooked, or obscured by disagreements on how to cost nonmarket phenomena.

A short description of the indirect impacts, as listed in Table 14 is provided below. Valuation techniques that have been used in case studies for the estimation of these impacts are described. Such methods are not necessarily consistent with the SEEA.

Soil erosion refers to changes in quality of the land, due to loss of topsoil. It affects the productivity of the land and can be due to natural causes as well as to economic activities, such as agriculture and forestry. Physical as well as monetary measures of soil erosion are complex. The universal soil erosion equation, includes rainfall erosivity, and soil erodibility factors and relies upon, length and steepness of the slope coupled with the degree of vegetative cover to measure average annual soil loss in physical terms (Narain and Nachtergaele 1998).

Several valuation techniques have been proposed. Several methods are based on maintenance costs, that is the costs of nutrient replacement, of terracing the slope, of rehabilitating the plot to its former productive conditions, etc. The latter cost is a restoration cost and, as mentioned earlier it pertains to cost-borne type of valuation and therefore not consistent with the SEEA. Also, in the SEEA the value of soil erosion caused by logging on steep slope is considered to be an indirect impact of forest depletion and, therefore, its value is not deducted from the value added of forestry.

Carbon sequestration relates to the capability of forest to bind carbon and contribute to the prevention of global warming and climate changes. The carbon sequestration is usually estimated through physical models of forest types and land-use change. It depends on the species mix, the organic matter content of the species, the age distribution of the forest stand and soil and climate factors (Adger et al. 1995). In practice, carbon balances can be calculated by multiplying the elements of the physical asset accounts for forest, in volume, by the coefficients of dry matter content and of carbon concentration in the dry matter of the different species (see above and Hoffrén 1996). The SEEA, as mentioned before, embraces the cost caused approach, and therefore captures the cost of loss of carbon sequestration through the valuation of timber depletion and/or the valuation of emissions above absorption levels. Absorption level depends also on the carbon sequestration capacity of the forest. The cost of abating or avoiding CO2, N2O and CH4 emissions is estimated using end of pipe technology, substitution costs, etc. and deducted from the value added of the causing industry. Damage-valuations, instead, attempt to value global-warming value. Temperature changes are related to physical impacts, which are then valued. This involves various emission scenarios and damage and abatement functions. Such types of valuations, as mentioned above, are not consistent with production and income accounting.

Watershed protection relates to mitigation of flooding, groundwater recharge, and reduction in the volume of runoff. Loss of watershed has an indirect impact on productivity of agricultural land, for example. The cost of loss of watershed could be estimated, with a demand side valuation, as the cost of mitigation of flood damage, purification of water from suspended sediments, etc.

Quality of the forest ecosystems and health and vitality of forest ecosystems are related to the functions of the forest as habitat for flora and fauna and as providers of ecological services. These characteristics of the forest are the most complex to measure. The SEEA does not even attempt to put a monetary value on "environmental" forests, that is those forest that fall outside the production boundaries of the SNA. It recommends instead simply measuring the area of forest land. Indicators on the quality of the forests’ ecosystems could be presented. Indicators on the health of ecosystems are obtained by linking landscape quality and biotope (plant community) quality with the stock of species in biotopes (Seibel et al.). Biodiversity therefore may constitute an important element of ecosystem health. In the SEEA, the number of endangered species as well as the number of individuals per species is reported in physical terms. While perhaps desirable it is not necessary to monetize the gains or losses in biodiversity.

Recreational activities are indirectly included in the SNA as tourism expenditures. The question of which expenditures are considered as integral part of recreational expenditures is very relevant. Willingness to pay for the recreational experience has been estimated in several cases. Most of the time the results are site specific and are hardly extendable to the national level.

Economic activities, even though not directly linked to the forest, may have negative impacts on the quality of the forests and reduce or enhance the capacity of forests to produce products and environmental services. The forest capacity of carbon sequestration allows industries to release emissions safely into the air, until the absorption level is met. The absorption level depends on the concentrations of pollutants in the air and the vegetative characteristics of the forests. If the air quality were such that the absorption level has already been saturated, acid depositions would affect foliage and soil and water body chemistry and broader aggregates of ecosystem health and productivity.

In order to have a clear understanding of the impacts of pollution on forest; one should have information on the cause-effect relationship between emissions and environmental impacts. These relationships are hard to identify. Emission data describe the cause but not the effect of pollution. Environmental indicators/indices describe the state of the environment and yield clues about the effect but not specific links between the cumulative multiple sources and their respective damages. Rather, emissions are loosely linked to ambient concentration levels. In general, however, there is not adequate information on the influence mechanisms, the state, rate of damage changes, the means of prevention of extensive damage or the tolerance of various forest ecosystems to these disturbances.

The present state of the forest "health" depends, on present and past activities as well as on transboundary airflows and the import and export of pollutants. It is usually not possible to identify which activity, when and in which country, caused the present state of the forest. In order to try to evaluate the impacts of economic activity on forests, three differing and complementing techniques, none of which are completely satisfactory, could be applied. They are:

Conversion of emissions into theme equivalents – acidification, in the case of forests. Emissions contributing to acidification are weighted to form a single indicator, according to the acid forming potential of each pollutant. Also, domestic pollution is augmented with the imports and reduced with the exports, of pollutants. Despite the arbitrariness of the weights, such indicator of net air pollution in the country provides useful information on the amount of acid deposition in the country.

Description of the change in the state of the forest in one year through environmental indicators/indices. Several indicators describing the quality of the forest are regularly compiled. One of the most commonly used is defoliation, which is the number of needles that are missing as compared to a healthy tree in the same forest. This indicator is described as a percentage of the missing needles or leaves. The reasons for needle loss are age of the trees, weather, climatic factors and pollution. Other environmental indicators of the health of the forest include biodiversity indicators, area of forest land, forest stock, etc. In order to evaluate the state of the forest, interpreting a large number of physical indicators; measured in different units can be complicated. Indices have been compiled aggregating the different indicators with weights. Given that little information is available on the relationship between the indicators, usually describing different aspects of the quality of the forest, the selection of weight is arbitrary. The choice of the weight might have a big impact on the magnitude of the indices, therefore, one should be cautious when interpreting indices of the state of the forests. Quality indicators have been linked to the asset accounts. Although physical indicators of the state of the forests reflect the impacts of activities that may have taken place elsewhere in previous accounting periods, they reflect impacts borne by the forests.

Monetary valuation of the damages caused by economic activities on forests. The valuation of the changes in the quality of forests in one year, as measured by changes in environmental indicators/indices, provide information on the "environmental debt" of the country (see above). Techniques used to monetize the impacts of economic activities on forests are based either on direct or indirect valuation methods. Direct valuation methods consist of estimating the willingness to pay of individuals to guarantee the existence of the forest or their willingness to accept reduction in their consumption or income to retain a stock of forests not impaired by economic activities. Problems with this method include the following. Respondents to survey questions may not have sufficient information concerning the costs and benefits of forests so that their responses are poor indicators of more informed values. Survey respondents may also not be easily able to express their values in monetary terms. Indirect methods estimate damage costs in the form of loss of production or costs accrued in avoiding such loss.This method allows the use of market prices, in order to evaluate the costs to society (National Institute of Economic Research and Statistics Sweden 1994).

The SEEA provides an integrated database, which could serve as a starting point for the evaluation of the impacts of economic activities on forest, both in physical and monetary terms, as described in a), b) and c). Emissions are recorded by industry and impact the domestic and international environment. They are then valued at maintenance costs to assess the costs of the direct impacts of production on the environmental asset changes. Damage valuation methods, reported in c), while providing useful information, are not fully consistent with the SNA and, therefore do not enter in the calculation of environmentally-adjusted aggregates.

A forest sector SEEA allows the calculation of a range of economic aggregates which reflect cost adjustments. For example, a supply-use identity is an indicator of the supply of goods and services produced and imported in the Forest sector. It is equal to the use in intermediate and final consumption, capital formation and export. If all forest uses are properly linked, it will yield an indicator of all forest benefits. Also, a forestry environmental ly-adjusted value-added identity can be calculated. This describes value-added for the sector net of depletion and degradation costs. Finally, an environmentally adjusted domestic-product identity for the whole economy may also be calculated. This is the environmentally adjusted value added of all industries with a further reduction of household generated environmental costs. This later index will allow better inter-industry comparisons of economic performance.