Country Experiences: Selected Case Studies, Trials, and Policy Recommendations

There have been a variety of experiences in establishing satellite accounts for the forest sector and in using data typically found in such accounts in evaluating sustainable development issues. Some studies have been initiated with the fundamental purpose of establishing new accounts and accounting procedures so that robust information will be available on an annual basis for the kinds of policy analyses discussed in the first chapter of this report. Other studies have been essentially policy oriented in nature. Vincent and Hardwick (1997) provide an excellent review of the breadth of sustainable forest development studies. The primary thrust of many of the studies they reviewed was to utilize data similar to what has been proposed for forest sector satellite accounts in analyzing sustainable development. The range of experiences is rather broad. Eurostat, the statistical division of the European Union, has initiated a 5 country pilot study with the fundamental purpose of developing harmonized accounting procedures for all member countries. Experiences from the first year of these trials are reported in this chapter. The Statistical Division of the United Nations (UNSD), along with other international agencies, has helped several countries develop economic and environmental accounts. Much of this work was done before the development of standardized accounting procedures cited in chapter 3 of this report, although the accounts are consistent with this material. FAO commissioned four case studies, and the World Bank sponsored two studies through the Danish Trustfund of Environmentally and Sustainable Development. A key thrust of these studies focused on sustainable forestry issues with particular attention to prospects and potential problems associated with timber asset valuation procedures proposed by Vincent and Hartwick (1997). These commissioned studies contrast with those initiated by Eurostat and UNSD because they were not concerned with robust accounting principles per se, but rather on asset valuation and sustainability.

In addition to these commissioned case studies, trials and the assisted implementation of new satellite accounts, there has been a wide array of sustainable development studies of the forest sector. Some of these investigations have been done as national income accounting exercises and others have been essentially policy analyses which utilize the kind of data that would be found in forest sector satellite accounts if they existed.

Many important lessons can be learned from these broad and varied experiences. What problems are there in data sources? How have assets be valued? What approaches have been used in monetizing nonmarket goods and services? What kinds of sustainable development issues have been addressed? Is the thrust of these studies the depletion of timber assets or broader forest ecosystem level concerns? How has depletion been defined? What has been the relationship between forest assets in the SNA and forests in the SEEA? How have inter-industry impacts been treated?

In this chapter we examine and summarize these broad experiences with the purpose of providing insights to those that will develop and utilize forest sector satellite accounts. Following the discussion of country experiences, the chapter and report conclude with a set or priority actions, capacity building and training needs.

The framework for forest accounting grew out of Eurostat’s broader program for the development of Natural Resources Accounting, (see Joisce, (1996) report of the London Group Stockholm session). This framework constituted the first step of the program on the forest sector and focused on the integration of physical and monetary balance sheets, physical and monetary supply-use tables and economic accounts for forestry. Subsequent steps for the incorporation of additional environmental concerns will be initiated during 1998, and will encompass non-SNA forest values.

In 1996-1997, five countries, under contracts with Eurostat, agreed to test ten main tables derived from the accounting framework. The purpose of the pilot exercises was to test the proposed forest accounting framework to determine what kinds of problems might be encountered if the system is more broadly implemented. The essence of the framework included the following objectives, which broadly shape the evaluation of the pilot studies.

To link in a consistent way, physical and monetary data,

To link flow and stocks accounts,

To introduce some environmental concerns,

To formulate an accounting framework that allows for the calculation of most of the indicators retained by the Ministerial Conference on the protection of forests in Europe (Helsinki 1993), as well as a comprehensive description of the interactions between human activities and forests as a component of natural environment.

Two essential aspects of the accounting framework are balance sheets and flow accounts.

Balance Sheets

Balance sheets represent the essential classification of forest assets. Starting from internationally agreed-upon forest definitions (FAO, UN-ECE), this classification is intended to be the basis for describing the main changes in the status of forests. The accounting objective of providing a more detailed trend of European forest changes was to be accomplished by coupling balance sheets through cross-classification with economic issues. For example, an attempt was made to integrate defoliation and mortality of forests to reflect changes in forest sector benefits. In addition, the benefits of carbon accumulation in forests as a global environmental benefit were integrated within the physical forest balance sheets.

On the monetary side of the forest balance sheets, no specific approach to value estimation was required of the participating countries. Rather, the choice was left to countries with the recommendation that the methods be simple, easily reproducible, and not particularly demanding of data.

Flow Accounts

Flow accounts describe levels or rates of activities within an accounting period. Descriptions of activities can be expressed in physical terms or monetary terms. Examples of physical flows are changes in land use, the growth of cultivated timber within a year, amount of timber felled, and the amount of logs consumed. The monetary counterpart to physical flows is rates of money flows. For example, the monetary counterpart to the physical volume of timber felled in a year is the annual value of the stumpage payments. The objective in establishing these accounts is to provide data necessary to link the forest sector flow activities with flows in other economic sectors. For example, if the forest sector sold logs to the sawmill sector, the monetary flows represent the links between the two sectors. A "Forest Resource Use and Management Account" was proposed, as well a description of residuals flows.

Proposed Tables

Fifteen tables were drafted in order to test the proposed European forest accounting framework. These tables represent a starting point and were not intended to cover all the forests issues and aspects addressed by the accounting framework. The composition of the 15 tables included the following items.

Balance sheets in physical and monetary terms (areas and volume of standing timber): 4 tables

Output related to forests and economic accounts for forestry 3 tables

Supply use tables in physical and monetary terms 4 tables + 2 complementary tables on residuals

Mass balances: 2 tables + a complementary table

Experience and Insights

The pilot exercises were originally intended to test the feasibility of the framework concerning classification issues, and the separation of economic accounts between forestry and logging. Valuable and somewhat unanticipated lessons concerning forest valuation were also forthcoming. The experimental exercises also provided important insights into the changing character of European forests although these results should be considered as unofficial.

In general the statistical offices took the position that all European forests are exploitable for wood production and are more or less "cultivated". However, in fact, some forests are non-exploitable either because they are protected or inaccessible. Within the class of exploitable forests, it is possible and at times useful to distinguish between forests that result from "afforestation" in recent time periods (e.g. 20 years) as opposed to regularly managed forests or more mature forests, or stands without any intervention for at least 25-30 years. This classification may be very important because the later class of timber is more likely to be harvested in the near term than the former and as a result is implicitly more valuable.

The pilot exercises point to the need to harmonize classifications before drawing any conclusion from the data. It is clear that the share of non-exploitable forests vary widely between countries either due to different definitions of protection or accessibility. Using national forest inventory (NFI) data did not present any insurmountable problems in determining physical balances of forest land and standing timber over a five-year period. These balances do not show great variation in terms of total forest area. However, they do indicate a general movement from "non protected" to "protected" status while at the same time showing an increase of stocks of standing timber in exploitable forests.

Quality classifications were introduced in the accounting framework. The objective was to provide a basis for examining trends in the physical degradation of forests thereby providing a basis for the corresponding valuation of forest losses. Germany was successful in cross-classifying economic and quality characteristics. Sweden and Finland only were able to provide information in annex tables. None of the countries reported forest degradation in their monetary balance sheets. It became apparent that in order to value forest degradation, it is necessary to have both physical data and different prices for the various quality classes of timber. Apparently this kind of information is not readily available. Furthermore, if a present value approach were applied to the balance sheets for forests, it would be necessary to use different natural growth rates for the various quality classes as an expression of degradation. This in turn would have to be coupled with discounted (and apparently unavailable) quality class prices. Based on the problems with valuing forest degradation, it may be necessary to reevaluate the continuation of this component of forest sector accounting. Whether or not a full cross-classification of economic and quality parameters shall continue to be included in the accounts must be discussed. The German report shows that the cross-classification is feasible.

Other countries had lesser success in these initial ventures.

Monetary Valuation

Forest valuation has traditionally involved development of separate values for land and standing timber. Because timber is sold more often as an asset to be removed from the land than is the case of the combined sale of land with standing timber, timber values are typically more readily available than are forest land values. Furthermore, when forestland is sold with standing timber, both assets are commonly included in a single transaction price.

Valuation of land was approached by using either actual or "recommended" (prices set by qualified land appraisers) prices for sales of transacted forest real estate. This method seems preferable to any other. However, further analysis may have to be done to determine how variation in the characteristics (for example difference in tree species, timber volumes, and development rights) of the real estate transactions contributes to explaining variation in real estate prices. In subsequent studies, a variety of approaches for estimating land values will be conducted. These approaches include market values, recommended values, alternative use values (marginal agricultural land values may be particularly meaningful), and capitalized rent values.

A variety of approaches were utilized in the valuation of standing timber. For example, Sweden and Finland use stumpage prices while Germany uses a hedonic pricing method. Although valid comparisons are not always possible, it appears that the asset value of the total forest, which is the sum of land and standing timber values, is very high relative to the actual transactions received. This point should be carefully examined and given considerable thought.

Economic Accounts for Forestry and Logging

The intent of the original proposal was to define all trees in exploitable forests as cultivated. This essentially implies that all timber growth in exploitable forests is an output contributing to GDP . As a consequence, separate accounts for forestry and logging were advocated so that the benefits could be separated. This was only done in a consistent way by one country. Because the aim in the accounting framework is to link economic accounts with balance sheets, as well as to link monetary data with physical data, it would be useful to actually complete this part of the pilot exercises independent of the decision taken by national accountants. This would provide more meaningful information and help further evaluate the trial accounting framework.

An interesting result of the exercise is that, with the exception of Germany, the rate of return of Forestry is 2.5 - 3%, when the value of standing timber is calculated by stumpage prices. This rate of return is low in some respects. However, in developed countries, the long-term real rate of return on capital, before adjustments have been made for natural resource depletion, has been on the order of 2 to 4 %. It appears that the preliminary forest sector rates of return are in keeping with the overall long-term performance of the developed economies. Perhaps one reason that the rate of return was higher in the German case study is their apparent reliance on hedonic stumpage prices. The term hedonic means "quality adjusted" prices. In comparing the average quality of timber being harvested with that of less mature growing "work-in-progress" timber, one would expect quality differences. Timber quality normally improves with growth so that the value per unit of volume of mature timber should exceed the value per unit of volume of maturing timber. This means that the average value per unit of volume of growing stock is typically less than the average value per unit of volume of timber being cut. If the value of growing stock is not adjusted downward to reflect it’s lower average quality, timber receipts as a percentage value of stock will appear unrealistically low.

Other products related to forest and other wooded land (berries, mushrooms, game, recreational fees, etc., either market or for own final use) were partially identified by the different national exercises. Forest management expenditures were not desegregated and separately allocated for these products.

Supply and Use Tables: Mass Balances

The purpose of developing supply and use tables is first to provide a consistent linkage between economic uses. Particularly important are the linkages between final wood product consumption and exports and the resulting derived demand for timber. This most fundamental linkage demonstrates the importance of forest resources in a national economy.

Due to the differences between national classifications of activities and products and the European classification of industries and products (NACE Rev.1 and CPA) it was not always possible to fill in the proposed tables according to common classifications. Moreover, in some cases, owing to unavailable data, uses and supply could not be matched. Some additional work is necessary because data concerning the supply and use of wood are apparently inconsistent and can’t be matched with corresponding elements in economic accounts.

Residuals are the left over waste products associated with production. Another important objective of the pilot studies was to describe flows of wood and paper wastes as well as residuals linked with forest related activities. This objective was more or less accomplished in the in the area of wood and paper waste products. However the lack of detailed data for the waste treatment and recycling industry hampers the elaboration of accounting for residuals balances.

General Assessment of the Eurostat Pilot Exercises

The main conclusion from the limited initial exercises is that consistent integration of physical and monetary data, along with flows and asset stocks is not straightforward. This is true even for countries which have comparatively well developed forest data bases and where the initial experience focuses simply on reformatting and delineating SNA transactions and balance sheets. Assessment of actual maintenance costs or forests protection expenditures will necessitate further analytic efforts as will the choice of a harmonized valuation method. Besides consolidating the experiences of the first pilot exercises, there may be adjustments in the accounting framework as well.

The objective of the next phase of developing natural resource accounting, scheduled for the current year, will be an attempt to integrate "non SNA" values. However, there will be no attempt to modify the aggregates of system of integrated accounts. Even when previously mentioned market valuation issues are solved, the remaining key problems appear to be the development of approaches to handle these "non SNA" values. The following represents a more detailed list of aspects of these non-SNA values and associated issues.

Are non-SNA values already (partially or totally) included in some SNA valuations?

Are non-SNA va lues actual values?

Are non-SNA values comparable with SNA values?

Are non-SNA values additives?

When extending the production boundary of SNAs in order to include natural services, how should the services be treated?

At the next stage of model construction, the objective of satellite monetary accounting should be to propose a consistent treatment (consistent both in terms of contemporary economic analysis and SNA conventions and concepts) of the important nontimber economic values. Finally, further consideration should be given to some of the recent studies on valuation such as Hulkrantz’s (1992) Swedish study. This study included many important nonmarket values and found that when these values were included, value added was substantially larger.

The United Nations Statistics Division (UNSD), with the financial support of UNEP, UNDP and The World Bank, has provided technical assistance to several countries for the implementation of the SEEA. One component of the SEEA compilation consists of forest accounts, although other important environmental accounts have also been instituted. The first set of countries that tested the feasibility of the SEEA before the publication of Integrated Environmental and Economic Accounting (1993) were Mexico and Papua New Guinea in 1992. Satellite accounts were next implemented in the Republic of Korea, Philippines, Ghana and Colombia between 1993 and 1998. The account compilation methodology used in these six countries is consistent with methodology presented in this report. All case studies used data already available from government sources and from special studies. The valuation method for depletion allowance used for all the following case studies is the net price method.

Philippines

The National Statistical Coordination Board (NSCB), with the technical assistance of UNSD completed the SEEA implementation project. The project was carried out jointly with the Department of Environment and Natural Resources (DENR), the National Economic Development Authority (NEDA) and the United Nations Development Programme (UNDP). Asset accounts for forests, fish, mineral resources, land/soil and water were compiled for the period 1988-1994 in physical and in monetary terms. In addition, emissions costs for selected industries were estimated and environmentally modified indicators were calculated.

The forest accounts included accounts for dipterocarps, pines, mangroves, mossy and submarginal forests. Rattan accounts were the only non-timber product account included for dipterocarp forests because rattan has a significant economic value. Forest depletion was the most significant part of the environmental cost generated during the period 1988-1991. After the introduction of the logging ban in 1992, the depletion as well as the contribution of the forestry industry to NDP (0.39% in 1994) decreased substantially. Figures on the forest and land accounts are currently being revised.

Colombia

The Administration Department of National Statistics (DANE), with the technical assistance of UNSD, implemented the Colombian System of Integrated Environmental and Economic Accounting (COLSEEA). The project was carried out jointly by the Interagency Committee on Environmental Accounts (CICA), involving DANE, the National Department of Planning (DNP), the Ministry of Environment, the Institute of Hydrology, Meteorology and Environment Studies and UNEP. Environmental protection expenditures of industries, households and government along with accounts for mineral resources, forest, land and water were compiled.

Accounts were compiled for produced and non-produced economic forests as well as for noneconomic environmental forests. Changes in the opening and closing forest stocks were reported in both volume and biomass and were estimated using data generated by special studies. Deforestation caused by illegal logging and fuelwood production was included in the asset accounts and estimated income generated from nontimber forest uses was limited to eco-tourism and the production of pharmaceutical product values.

Republic of Korea

Korea Environment Institute, with the technical assistance of UNSD, has just completed a pilot compilation of the SEEA. Asset accounts for land, forests, fish and mineral resources as well as environmental protection expenditures and emissions into land, air and water were estimated. Deforestation and forest depletion do not constitute an environmental concern in Korea. Forest stocks have been increasing steadily, as a result of extensive reforestation and protection policies. The contribution of forestry to GDP was only 0.2%.

Ghana

Ghana Statistical Service, with the technical assistance of UNSD and the Overseas Development Administration (ODA) coupled with financial support from UNEP undertook a pilot SEEA compilation. Asset accounts for timber, marine fish, subsoil assets, water, and forestland were compiled for the period 1991-1993. Even though the importance of non-timber forest products is recognized in the Ghanaian economy, it was not possible to obtain data estimates necessary to compile broad forest sector accounts. The value of forest land was estimated using adjusted compensation rates paid by the mining companies for forest destruction. No estimate of the allowable cut was available, it was therefore not possible to estimate depletion relative to a ceiling rate of harvest (UNSD’s approach to hard sustainability).

Mexico

The project was carried out in 1990 and 1991 jointly by UNSD, the World Bank and the National Institute of Statistics, Geography and Informatics (INEGI) in Mexico. Attention focused on the following environmental concerns for the year 1985: oil depletion, land, water, and soil degradation, deforestation and changes in land use. Timber accounts as well as land use accounts (including land use changes) were calculated. Environmental protection expenditures by the forestry industry were estimated to be very low. A depletion allowance was calculated and deducted from the value added of the forestry industry. The environmentally adjusted value added turned out to be 27.2% of the value added for forestry. The contribution of the forestry industry dropped from 0.54% of NDP to 0.15% of environmentally adjusted NDP when depletion was incorporated in the production accounts and nonforest products were excluded. The capital output ratio, an indicator of the capital productivity of the sector dropped from 45.38% to 12.34% when natural capital was added to produced capital with depletion accounting.

Papua New Guinea

UNSD and The World Bank carried out environmental account implementation in 1992. SEEA was compiled for 1986-1990 and included forests, mineral resources and energy accounts. Forest asset accounts were compiled for "economic" (i.e. that have a potential economic use for logging activities) and "environmental" (i.e. forest land non-accessible for logging operation, mountain areas, etc.) forests. The forestry industry is comparatively small and represents about 4.1% of GDP. Lack of reliable data on timber production did not allow the estimation of depletion. An attempt was made to estimate the value of losses of ecological functions due to exploitation. The estimates were on the estimated compensation costs by logging companies to landowners and amounted to a range of 0.4-2.3 % of NDP and 6.4 – 39.3% of NVA of the forestry industry.

Experience of UNSD Assisted Countries

Several points are noteworthy in these studies. First, existing data sources had to be augmented by special studies. Second, forest resources varied from country to country. For example, some countries limited the forest sector to timber. Others added eco-tourism, fuelwood, rattan, or pharmaceutical values. No country examined carbon sequestering services or their value. It appears that broad multiple-use values were not consistently appraised and included in forest sector accounts. Virtually no country had good timber prices so the net-price values had to be estimated as residual values. Of course, one would expect that the inclusion of depletion would reduce estimates of value added. The change in the amount of value added in Mexico seems to be "surprisingly large".

Case studies in Brazil, Chile, Philippines and Zimbabwe were supported and directed by FAO and the World Bank with the primary purpose of examining the feasibility of utilizing the valuation approaches proposed by Vincent and Hartwick (1997) to evaluate sustainable development.

Each included country faces unique sustainable development dilemmas. At the same time, many problems are shared somewhat in common in developing the necessary information to evaluate the contribution of the forestry sector to a society’s long term economic well-being. As a result, we start this part of the chapter with a summary and overview of the particular focus of each case study. In doing this, we examine the physical accounts utilized in the case study, the approach to valuation and then we briefly restate the conclusions. Following the presentation of each case, we look at patterns in analytical problems and make broader suggestions about the usefulness of forest sector information in these approaches to policy analysis concerning sustainable forestry.

Brazil

Seroa da Motta and Ferraz do Amaral (1997) examine rates of depletion of both Mahogany and all other timber in the Brazilian Amazon. The area excludes the state of Mato Grosso but includes the area referred to as the Legal Amazon. The Amazon is the world’s largest tropical rainforest and its socio-political-bioeconomic importance is reflected by the authors’ exclusive focus on the region’s timber sustainability. No estimates of other forest benefits are included.

The physical accounts are suggested, rather than directly enumerated and are represented by stock estimates in 1990 and 1995 for "all timber". In addition, there is a 1990 estimate of the stock of "mahogany". There are no land accounts, nor are there neither other indicators of timber quality nor ways of adjusting outputs in other economic sectors. There is also no direct adjustment of GDP for timber assets in the Amazon.

This study relied upon estimates of stumpage price and production costs. They use the Vincent and Hardwick approach. Stumpage values are net of average extraction costs. In the absence of any estimate of the elasticity of marginal extraction costs, the authors do a sensitivity analysis using two extreme (0 and infinity) and more informed rates of 1 and 3. In addition, they also do a sensitivity analysis of the discount rate with values of .02, .04 and .10. The authors were able to obtain average prices of sawn logs but only average rather than marginal costs of costs of logging and extraction were available. This made the development of the valuation approach proposed by Vincent & Hardwick (1997) very difficult.

Of course the sensitivity analysis creates a variety of estimates concerning the longevity of both mahogany and all timber assets. In all cases the mahogany asset liquidation is expected to be completed many orders of magnitudes earlier than for all timber. This suggests that there is a qualitative degradation of the timber resource. Because, the report doesn’t examine changes in the forest sector’s wealth, it was not possible to look at the broader view of sustainable development.

The authors experienced many data availability problems. It would have been useful to more broadly examine changes in land use, timber wealth, and related forest benefits in a broader context of sustainability. Furthermore, the issue of dual issues of both inoperable stocks and unused harvests looms large in areas like the Amazon. For example, if unused harvests represent 50% of the volume removed, and substantial areas are inoperable, the longevity or sustainability of the harvests could be very different from the conclusions of this study.

Chile

Vergas and Sandoval (1997) examine the sustainability of timber produced from short rotation timber plantations (cultivated economic assets). There is no attempt to examine changes in land use, nor other forest benefits. The study is limited to areas planted in plantations for the periods of 1985-1996. For growing stock, the inventory includes land areas and ages of both species for annual ages from 1 to 20 years, the typical harvest age in the plantations. There were no adjustments for catastrophic losses or for other market or non-market forest benefits. Plantations account for the industrial basis of forest products in Chile, although the plantations account for only 2.8% of the total surface of the country while the native forests account for 17.8% of the total land area.

Separate asset values for plantations of each of two tree species were calculated using the modified El Serafy approach. Since the plantations are intensively managed around 20 year rotations, an asset value for each age was multiplied by the appropriate number of hectares for that age in accumulating the plantation asset value. Costs of timber extraction were constant, as were stumpage prices during the period of analysis. Timber prices were estimated using the residual value method of timber appraisal rather than using observable stumpage prices.

From the standpoint of forest plantations, Chile has been accumulating wealth. Based on current plantations and rates of growth and harvesting, asset accumulation is expected to continue through the end of a 20 year forecast period. In other words, the plantations represent a sustainable form of wealth accumulation in the nation's economy. Also evident was the effect of a policy modification toward the end of the period of analysis that reduced the incentives to develop plantations. Corresponding with the policy change was a lower, but still positive rate of increase in the plantation land base and resulting lower rate of asset accumulation.

From the standpoint of a broader calculation of forest benefits, it would be interesting to know what land uses were being replaced with the addition of forest plantations. Also, it would appear that carbon sequestering would be a net benefit of plantation accumulation and a reasonably simple extension of the analysis. No attempts were made to estimate value added from forest plantations. Also, even in plantations, there may be unused harvests, so that the asset values may be overstated.

Zimbabwe

Mabugu, Milne, and Campbell (1997,1998) evaluate the sustainability and economic importance of fuelwood in Zimbabwe. About 70% of the nation's energy consumption is derived from wood. Fuelwood use is a contributing factor to deforestation, and recently was rated as the most serious environmental problem facing the country. Fuelwood use is a case of an unreported forest value and as a result, the role of forests in providing fuelwood is easily overlooked. Furthermore, using the usual indicators of welfare such as pre capita GDP, the economy has been in steady decline throughout the early to mid 1990s. Thus, developing better estimates of the sustainability of natural resources and wealth are very important to Zimbabwe.

Essentially, two accounts are utilized, a land account and a timber account. The forest land account has been experiencing about a 2% annual rate of loss to agricultural uses during the period of 1986-1995. In assessing the timber stock and growth available for fuelwood collection, further reductions were made in the accounts due to the protected nature of some lands such as national parks. Growth estimates were made in terms of mean annual increment (MAI). Estimates of fuelwood consumption had to be made by utilizing studies of the number of households using fuelwood and fuelwood consumption rates. The consumption estimates were adjusted downward to fuelwood removals by deducting estimates of supplies from drywood and land clearances.

One thorny problem is to estimate the costs of fuelwood collection. The report relies on an opportunity cost approach since most fuelwood is informally collected rather than compensated work. This kind of work is almost universally left out of national income estimates. Costs depend on the time involved and because deforestation takes place near residences first, costs increase with distance traveled to collect wood. This suggests that costs will rise with time and with total fuelwood harvested. The value of fuelwood was estimated using market prices of fuelwood. Hotelling rent is determined by converting the total rent by Vincent and Hardwick’s conversion factor formula using an annual interest rate of .18 and an elasticity of marginal cost of 0.6. There was literally no estimate of the elasticity of marginal costs so the authors essentially had to assume a number in order to complete the valuation work. Finally the estimated value of fuelwood was added to create an adjusted GDP and NDP.

Fuelwood adjustments represent about a 1-% adjustment to GDP. This is substantial and suggests that estimating other forest benefits is likely to produce a much better indication of the importance of forests to national well being. With forests disappearing at an annual rate of 2%, this may indicate some need for new policy analysis. The authors found a variety of problems in collecting and utilizing data. For example, they had hoped to readjust GDP in the energy sector to reflect the reliance on wood energy but were unable to do so. They recommend a different and more useable form of data collection in the future.

Adding the value of wood fuel consumption to the national income measures of well being in Zimbabwe helps develop a better understanding of sustainability. There is currently adequate forest growth to sustain levels of fuelwood consumption. However, because fuelwood use has such local spatial collection economies, there may still be major problems that aren’t reflected.in this scope of analysis. Because national accounts examine changes in economic aggregates, these kinds of sustainability issues may still be overlooked in forest sector satellite accounts.

There should be a way of cross walking between the national and local levels of analysis.

The FAO/World Bank Commissioned Studies in Retrospect

Perhaps most interesting from the standpoint of broad patterns is the lack of good timber inventory data. Even in the case of the Chilean plantations, managed stand yield projectio ns were used and coupled with records of land areas planted as a means of estimating the inventory of wood in each annual age class. The most difficult timber and forest inventory problems, which really preclude good estimates of the value of this kind of natural wealth, were found in Brazil and Zimbabwe.

Each study was also confronted with problems in resource valuation where timber pricing was not a particularly easy accounting venture. In spite of the normal economic classification of timber as a market good, imperfect market arrangements make it very difficult to estimate the true worth of the timber. Some countries had reasonably good estimates of the costs of growing and tending timber. Where the examination focused on native forests as opposed to plantations, it was difficult to separate operable from inoperable forest lands and also to adequately examine how qualitative differences in timber affected asset and flow values.

Recalling that an important element of the Vincent and Hartwick (1997) approach to valuation relies on knowledge of the elasticity of the marginal cost of logging, it is fair to say that data of this type is in extremely short supply. Seroa da Motta and Ferraz de Amaral (1997) provide interesting insights into and critiques of the limitations of the Vincent and Hardwick approach.

Finally, in all of the countries there is little evidence of knowledge of unused harvests and their impact on sustainability. In Zimbabwe, fuelwood harvests are estimated from household surveys. In Brazil, harvests were estimated from mill production data and in the plantations of Chile, it was simply assumed that utilization of a mature plantation was complete.

Having identified the broad view of total forest benefits and total forest values that are germane to sustainable forestry and sustainable development in an earlier chapter, we now turn our attention to a different form of experiences. Here we ask the question, where have people focused their studies to date? Has the main focus been on the sustainability of timber and forest lands or has the attention instead been placed on the broader array of forest benefits? Have analysts been more interested in the asset accounts (balance sheets) or the flow accounts? Have countries in one region of the world been more likely to examine sustainable development from a national income accounting perspective than another region? And how many studies actually looked at the contribution of the forest sector, either broadly or narrowly conceived, to national income?

Dube, Jackson and Narain (1998) utilized the summary of studies presented in Vincent and Hardwick to examine patterns in these studies. Since the late 1980s, the greatest number of studies that have aimed at incorporating forest resources into the national accounts pertains to countries in Asia, and the least to countries in Africa. The number of studies increased more or less steadily during the 1990s, with the greatest number, nearly a dozen, being published or otherwise made available in 1996. Several more are in preparation under the auspices of the UN Statistical Division. Only twelve of the studies appear to have involved government agencies responsible for the preparation of national accounts.

The benchmark study is the well-known examination of Indonesia by the World Resources Institute (Repetto, 1988). This is the most frequently cited study on natural resources (including forest resources) and the national accounts. A study for the World Bank by Peskier (1989), which commented on the WRI study and contained additional estimates for Tanzania, came next. Most of the studies covered one or more resources in addition to forests. That is, they did not focus exclusively on forests. More than half covered subsoil assets in addition to forests. Other non-forest resources included agricultural soils, rangeland, fisheries, and, in a few instances, air and water pollution. The forestry-specific studies tended to be the more recent ones and covered the longest time periods. In particular, the Canada studies analyzed the longest period, from the early 1960s to the early 1990s. Most studies, whether forestry-specific or more comprehensive, covered at least one decade.

Virtually all the studies included timber, and nearly half covered one or more nontimber aspects of forest resources. The list includes nonmarket production of fuelwood, berries, mushrooms, game, rattan, and peat; amenity values associated with protected areas and biodiversity; environmental services like watershed protection and soil protection; carbon sequestration; and acid deposition.

Some studies made adjustments to both flow (current) and asset accounts, while others made adjustments to just one or the other. Of those that made adjustments to current accounts, about two-thirds made adjustments for net accumulation. In most cases, the net accumulation estimates were for timber, although a few studies included nontimber values in calculations of net accumulation of forest land. Five studies estimated the value of nonmarket production of forest goods, which is an adjustment to GDP. One study (Finland) investigated a specific SNA recommendation, that timber growth in "cultivated" forests be valued and added to GDP. New Zealand has already implemented this recommendation for some time. About a third of the studies made adjustments related to asset accounts. In most cases, they simply calculated the value of the standing forest, without formally linking the estimates to stock and flow accounts. A few did, however, and compared the asset value of forest resources to the value of other assets in the economy. Malaysia studies, in addition to focusing on adjustments at the national level, also adjusted income accounts at the subnational level. They found that the impacts of net accumulation adjustments varied greatly within the country, with important implications for regional sustainability.

The promise of full benefit accounting for the forestry sector in the evolving national income accounting procedures is great. But, the potential also exists to do the forests and people of the world a lot of harm if the new accounting systems are illframed, poorly implemented or if inadequate, erroneous, or incomplete data is utilized in the accounting systems. We now examine the four kinds of experiences in aggregate. The fundamental question is what can be gained from looking at the EUROSTAT trials, the UNSD assisted examples of implemented forest sector accounts, the FAO/World Bank commissioned studies, and the broad pattern of other investigations. We hope to be able to use the experiences to uncover any potential conceptual problems, persistent patterns of inadequate data, or other problems that may help in developing and using forest sector satellite accounts.

Experiences and Potential Problems with the Physical Measures of Forests

Chapter 1 discussed the importance of physical measures of forests, timber, and related resources. The inventory of forest lands and standing timber is never a simple problem.

Collecting inventory data is expensive and countries define lands and timber volumes in ways that are not always completely comparable. In the Brazilian Amazon, there really was no timber inventory. Rather, some sample plots were used to extrapolate an estimate of the mass of timber over an area which itself is hard to estimate in total size. The Chilean study had estimates of the plantation land base and estimated timber inventory by coupling the planting date with yield projections. In Zimbabwe, forest inventory has traditionally focused on plantation forests so that the authors of that case study were faced with making some informed, but somewhat ad hoc estimates of growth in the natural forests. Coupling the problems of estimating a traditional timber inventory with emerging needs to examine quantitative aspects of forests in nontraditional ways was driven home from the EUROSTAT trials. The volume of standing timber maybe a reasonable indicator of carbon storage. However, considerable effort must go into examining total carbon storage and uptake throughout the life of both economic and noneconomic forests in order to gain better knowledge about the relationship between the commercial measures of stem biomass and total forest carbon uptake and storage.

There are no standard commonly accepted ways of measuring biodiversity. In fact there are several conflicting measures. In the absence of a readily accepted physical biodiversity index, less robust criteria and indicators might be included in forest sector satellite accounts.

Timber inventory data was apparently not a major problem in the EUROSTAT trials, but lack of a harmonized inter-union measure of operable lands apparently was a problem. This measure should not be based exclusively on physical parameters of forest sites. Costs and values must also define inoperable lands. Likewise, isolated sites that are otherwise productive may never the less be inoperable. The cost structure of logging can easily differ between countries. Operable sites in one economy may be inoperable elsewhere. There was no attempt to define the operable land base in the Brazilian Amazon. The Zimbabwe case study did make an operability adjustment concerning the production of fuelwood. It also recognized that the costs of fuelwood gathering were a function of the distance traveled to collect fuelwood, although it is not clear whether this form of location economies was part of the operability definition.

Finally, some country experiences indicate that measures of log removals are a poor indicator of logging depletion. Without good estimates of logging related mortality, log removals will understate depletion. In some developing countries this form of data problem could convert what appears to be net accumulation to net depletion. Thus log flows to wood using industries must typically be converted by a physical coefficient in order to convert log removal flows to reductions in standing timber stocks. The coefficient essentially represents "unused" harvests.

Experiences in Quantifying and Valuing Forest Benefits

The country experiences reveal a variety of problems in valuation. Not only are there problems in valuing nonmarket resources, there are also problems in valuing timber and forest land as well. In the Brazilian case study the most reliable market prices were for sawn logs and these values had to be adjusted to estimate the value of the standing timber resource. Furthermore, timber in the Amazon is treated as an open-access resource and the prices under these circumstances can’t reflect its competitive value. In the Chilean case study, the residual value method was used to estimate timber prices. It is an approach that subtracts average costs for an average operator from average product values. In an auction system, bid prices are determined by the interaction of marginal costs and marginal values, not averages. The "average cost operator" doesn’t win a timber auction. Thus, the residual value appraisal system can create bias that persistently understates actual timber values. Virtually all of the countries that implemented accounts with UNSD assistance used similar estimating procedures. When timber value is underestimated, all estimates of value added are understated as well.

Timber valuation methods in the EUROSTAT trials were left to each country. In countries that used stumpage prices to value the entire forest inventory (balance sheet asset value), the apparent estimated rates of return were 2.5 to 3%. Germany used a hedonic method and the apparent rate of return was higher. Apparently by using quality-adjusted prices, Germany’s standing timber value was estimated in a manner to reflect the actual value of the timber. Normally, immature trees are of lower value than mature ones. On the average, immature trees are smaller than mature timber, and smaller timber is typically converted into products of lower value and costs more per unit of volume to log and deliver to milling facilities. Unless the standing asset value is adjusted for these kinds of differences the net price method will overstate asset values relative to stumpage receipts. Virtually every case study has made this kind of error with the apparent exception of Germany and Chile. Stumpage prices reflect the qualitative characteristics of the timber sold. Because the entire forest has timber which is growing and changing in its characteristics as it grows, it is typically not worth as much per unit of volume as timber which actually enters the market place. This point is sufficiently important to warrant a brief diversion into some basic timber economics.

Timber has long been defined as a case of maturing assets. Timber appreciates for three basic and fundamental reasons. Timber grows so that standing physical capital accumulates with time. As timber grows, individual trees become larger. Large trees are worth more than small ones for two important reasons. There are fewer pieces of wood to handle per unit of timber volume so that logging, hauling, loading, unloading, sorting and sawing costs are less for larger logs than smaller ones. Furthermore, larger trees are typically converted into products of higher values. Lumber with larger dimensions, veneer products, and increasing amounts of wood with fewer (and sound) knots are all factors that help increase the value of wood per unit of volume as timber stands grow and appreciate. Furthermore, real wood product prices have increased around the world rather steadily since the end of World War II. This is the result of basic changes in scarcity. Thus if the average annual rate of timber growth is 2 to 3% over the life of a timber stand, its real rate of value growth typically far outstrips its physical growth rate. If forest sector accounts don’t reflect these basic timber economic facts, the purposes of developing full benefit accounting for sustainable forestry and sustainable development are essentially defeated because of the false appearance that forests are a "bum" investment opportunity.

Another key valuation issue surrounds the treatment of forest maintenance and protection costs. If a forest is managed to produce multiple benefits there are really but two accounting alternatives. Preferably, the total value of all forest benefits must be compared with total management costs. If total benefits are not measured, it is perhaps prudent to develop some administrative rules to separate costs and allocate them various benefits produced. In this later way, the share of watershed protection would be allocated to a water resource account; a share would be allocated to an eco-tourism account and so forth. For example, the countries assisted by UNSD, developed satellite accounts for timber but with the exception of the Philippines and Columbia, net timber flows were the only environmental forest benefit. The Ghana case is in interesting one since it is acknowledged that many other forest benefits were excluded. Rather clearly, either the total forest benefits must be compared with the total forest costs, or some form of cost apportionment is in order.

Another important valuation issue involves approaches to valuing forest land. Vincent and Hardwick (1997) approach this issue by essentially combining the capitalized value of the land in growing timber into the present value calculations of the timber itself. Any other forest land values would have to be based on other nontimber commodity land uses. Since the case studies in Australia, Brazil, Chile, and Zimbabwe apply the Vincent and Hardwick methodology, there is no attempt to add an extra value of forest land. The pilot studies under the direction of EUROSTAT hoped to develop a credible land value in the capital account balance sheets. They found that this form of value was lacking and recommend further studies to develop better estimates of these values. It would appear that if nations follow the Vincent and Hardwick approach to timber valuation, there is no need to add a capitalized value of the timber production growing rights to a separate land account. However, if people following this present value approach, have a separate land account, they should deduct the capitalized value of the infinite series of harvest (the land value in the Faustmann solution) from the rest of the forest land value.

All of the above discussion of valuation essentially focused on what is typically assumed to be market resources. Of course economists typically assume that if a resource is classified as a market resource, there are no valuation problems. Having dispelled that myth, we now turn to a discussion of nonmarket resources and forest benefits recorded in other SNA sectors. Table 5Error! Reference source not found. and Table 6 in Chapter 3 make it clear how value added for forests must be conceived. If economies leave out important values, net value added calculations can be easily criticized. Several of the countries in the group that instituted forest sector accounts under the assistance of UNSD must get full benefit accounting to make value added estimates more meaningful. Thus far, the most through development of value estimates of nontimber forest products have been done in countries included in the EUROSTAT trials. These countries have the most advanced and progressive forest research organizations in the world. Finland, Sweden, and Germany also have very advanced forest products industries and have developed the best forest management information systems as well. These three countries, which are included in the EUROSTAT trials, are in the best positions to evaluate the total benefit contribution of nontimber forest values.

Valuation issues for nonmarket environmental values remain troublesome. A reason is the macroeconomic or general equilibrium nature of the national income based policy analysis models. Since changes in outputs for entire industries are considered, it makes no sense to exclude consumers surplus for market goods and services and simultaneously includes it for market goods and services. Other approaches to valuation such as the environmental maintenance cost approach is also somewhat unconvincing because costs are being observed at output levels that would not rest in equilibrium if in fact markets were operating.

Experiences with Establishing Forest Sector Linkages

The most through development of linkage estimates is found in the EUROSTAT experiences. Recalling that there were really three kind of linkages to be established in the first phase of their trials, they found some unexpected difficulties. First, they wanted to develop better linkages between woodland timber sources and wood conversion sectors. Here they found that differences in industry definitions and lack of harmonization hampered the accomplishment of these tasks. Second, they sought to develop better residuals flow linkages between wood and paper products wastes and broader pollution phenomenon. They appeared to be more successful at the product side of product ion than at the woodland side in these ventures. Finally, they hoped to link air pollution with forest degradation via defoliation losses. Here the initial success was very unconvincing. Only one country (Germany) could actually create the necessary cross-classifications in order to link these activities. Furthermore, none of the countries could trace the impact of air pollution activities on timber production itself. In fact, the disappointing results of these attempted linkages may mean that these efforts won’t be continued in the future.

Of course, if and when forest sector linkages are successfully implemented, several important analyses will be possible. Fundamental among these kinds of studies will be the estimation of total value added of all forest sector activities to total economy-wide activities. This will allow policy makers to make more informed judgments about budgets and other key policy decisions. A few studies hint at the importance of these ventures. For example, the Zimbabwe study suggests that the value of fuelwood in the economy is about 1 % of GNP. Another important example is Hulkrantz’s (1992) Swedish study which estimated the value of carbon sequestering services along with fuelwood, berries, game and estimates of asset appreciation made a substantial difference in the calculated forest sector contribution to national well-being. These values by themselves increased total benefits by about one-third over more traditional estimates of value added.

Other Ecosystem Management Concerns

Our discussion has thus far focused on the experiences surrounding physical flows, which may in principle be expressed as monetized accounts in forest sector natural resource accounting. But, of course, it was recognized at the outset that other forest sector benefits might not be monetized in national income accounting or in satellite accounting. Perhaps the clearest example of that lies with existence values since there is apparently no physical flow associated with the knowledge that something of value exists nor are there actual users or rights to use specific products or services. The classical case of this are the existence values associated with biodiversity. It was held that these values differ from indirect use values of biodiversity. These later values might be thought of as the values associated with maintaining a stable life support system.

Many manifestations of these ecosystems management concerns are reflected in a variety of ways in SNAs and related satellite accounts. For example, some endangered species are not consumption goods. Their populations may be at risk as a result of habitat takings rather than population takings. They have no direct use values and there is no associated direct measure of annual economic activity. In some situations, forests could be removed from the base available for commodity exploitation as a means of protecting the species in question. This in turn could mean a reduction in the asset value of the forests without depletion. This falls under the discussion of the economic appearance and disappearance of an asset. The unfortunate aspect of national income accounting is that it makes these kinds of resource allocation decisions appear costless. Creation of protected habitat represents a reallocation of land in the nation’s land account, and a redefinition of the nation’s timber inventory in the sense that the timber inventory is an exploitable asset. Of course, the economic costs of these decisions lie in foregone exploitation since nothing is exchanged. All of these decisions can and often are made without explicit estimates of the value of the at-risk species. The implicit judgement, or perhaps better yet, the revealed minimum value of the protected species lies in its’ foregone values. At the same time, however, the forests presumably continue to take up carbon, provide water products and recreational services and may also continue to yield other nonwood products. To sum it all up, it may be necessary to have some capital accounts that are not directly associated with flow accounts.

The Criteria and Indicators are another interesting case in point. Here, would seem to be a case of essentially physical accounts (both stock and flow) which are not monetized. The purpose of these accounts is to give a far broader view of sustainability, but without a numerairre or unit of value which would allow broader welfare interpretation of environmental, social and economic trends. While the existence of Criteria and Indicators will allow policy makers to see positive and negative trends in key indicators of sustainable development, it remains to be seen how the absence of an unambiguous index of biodiversity will influence their social utility.

Sustainable-Yield, Sustainable Forestry, Sustainable Development and Environmental Accounting

Perhaps one of the most interesting aspects of all of the trials, case studies and country experiences are found in contrasting approaches to sustainable development. This point is brought forward most provocatively in comparing some of the results of the FAO/World Bank commissioned studies with the UNSD country experiences. Recall that the UNSD approach to sustainability is one where depletion only occurs when the level of harvest exceeds a calculated allowable cut. The approach followed under the FAO/World Bank studies was more traditional in the historical sense of balancing growth with harvests. For example, the Brazilian study concluded that current harvest levels couldn’t be sustained and that the problem was most readily apparent in Mahogany. In contrast, using the UNSD approach to sustainability, no conclusions were reached about the sustainability of harvests in Ghana precisely because no level of harvest was calculated. Under the UNSD system of "hard sustainability" there is no depletion until calculated sustainable harvests have been exceeded.

Forest economists familiar with mathematical routines used in calculating allowable cuts will readily recognize many weaknesses in the UNSD approach. The UNSD approach may indicate depletion when there is actual capital accumulation. Likewise, substantial depletion may occur and go unnoticed using the UNSD definition of sustainability.

In developing national income accounting methods and practices which are broadly useful, it must be remembered that the audience of potential users will come from a variety of sustainability contexts. The origins of sustained yield, while essentially concerned with long term capital management viewed the problem from the perspective of managing the inventory of physical capital. In contrast, economists who view sustainable development as being a problem of the long-term accumulation of broadly defined wealth are convinced of the necessity of monetizing all environmental assets. The recent sustainable forestry perspective sees the problem as one of striking a balance in the mix of forest uses without the necessity of valuing all environmental assets because the problem is really one of trying to balance conflicting political interests. In a sense, the physical inventory and flow accounts are of greatest interest to sustained yield forestry advocates. The capital assets and monetized flow accounts are of greatest interest to the sustainable capital accumulation school while the Criteria and Indicators are based from the perspective of sustainable forestry and are likely to be of greatest interest to advocates of this school of thought.

It may be impossible to both monetize all aspects of the Criteria and Indicators and link them to other activities in the forest sector. This should not preclude their being included in satellite accounts. Economists have long worked under the simplifying macroeconomic assumption of a homogenous physical stock of capital. This assumption has absolutely no empirical content yet remains a useful concept in macroeconomic modeling. Criteria and Indicators are an attempt to capture many of the non-additive components of the collective concept of sustainable forestry. They should not be dismissed because they don’t behave like standard accounting transactions.

Having focused the report on a description of forest sector environmental and economic accounts coupled with attention to their prospects, experience and use to date, it is now important to turn a proposed set of priority actions.

Support resource inventory efforts including the introduction of new technologies such as GIS.

Support development of household surveys of nonwood forest products consumption. In some countries these surveys might be included in the periodic census of population.

Develop key approaches for resource valuation of non-exchanged direct use values consistent with general equilibrium models.

Support ventures into proper estimation of timber asset values where institutional arrangements create inaccurate value estimates.

Support the development of methodology to estimate forest land prices.

Work with agencies to develop criteria and indicators.

Help develop and make available computer based analytical models that will utilize the environmental account information in policy analysis.

Support the development of production flows, monetized values and sector linkages for countries where nontimber forest products are important yet unappreciated.

Improve the transfer of know-how between EUROSTAT and countries with comparative rich experiences in environmental accounting in the forest sector to other less experienced countries.

Create effective transfers of knowledge concerning sustainable wealth and sustainable forestry at high levels of policy and public information by hosting major workshops for the press.

Avoid hardwired dogmatic approaches to sustainability. There is a variety of alternative sustainability "schools of thought" and there is considerable flux in what is the right approach. Develop accounting systems that can be applied to a variety of approaches to sustainability.

Creating good forest sector accounts requires teams of people with a variety of backgrounds and experiences. Satellite accounts should be developed so that they are integrated into, and an extension of the SNA. Rather clearly, people with national income accounting training should be intimately involved as team members. There is a lot of specialized technical knowledge surrounding the management and use of forests. Language is specialized and the production relationships are also complex. Forest economists should be extraordinarily valuable team members as we. Beyond this, there is an increasing set of complex biological issues concerning the existence and use of forests. Finally, it is clear that existing data and information sources are typically inadequate. Teams should have ready access to expertise in such subjects as timber inventory measurements, and household surveys. A good systems ecologist could also be a very important member of a forest account development team. Some countries may have these to enhance these capacities.