NOT FOR QUOTATION WITHOUT PERMISSION OF THE AUTHOR
SETEMS ANALYSIS FOR
THE
FOREST SECTORAE.
Andersson M. Kallio R SeppalaMarch 1 9 8 4 WP-84- 17
Working Papers a r e interim reports on work of t h e International Institute for Applied Systems Analysis and have received only limited review. Views or opinions expressed h e r e i n do not necessarily r e p r e s e n t those of t h e Institute or of its National Member Organizations.
INTERNATIONAL INSTITUTE FOR APPLIED SYSTEMS ANALYSIS 2361 Laxenburg, Austria
FOREWORD
The objective of t h e Forest S e c t o r P r o j e c t a t IIASA is t o study long- t e r m development alternatives for t h e forest s e c t o r on a global basis.
The emphasis in t h e Project is on issues of major relevance t o industrial a n d governmental policy m a k e r s in different regions of t h e world who a r e responsible for forestry policy, forest industrial strategy, and r e l a t e d trade policies.
The key e l e m e n t s of s t r u c t u r a l change in t h e forest industry a r e related t o a variety of issues concerning demand, supplies, a n d i n t e r n a - tional t r a d e of wood products. Such issues include t h e development of t h e global economy and population, new wood products a n d substitution for wood products, future supplies of roundwood a n d alternative fiber sources, technology development for forestry a n d industry, pollution regulations, cost competitiveness, tariffs a n d nontariff trade barriers, etc. The a i m of t h e Project i s t o analyze t h e consequences of f u t u r e expectations a n d assumptions concerning s u c h issues.
This a r t i c l e serves as a n introduction a n d s u m m a r y t o a forthcom- ing volume representing t h e state-of-the-art of s y s t e m s analysis i n t h e forest sector. This volume is e n t i t l e d % s t e m s Analysis in Fbrestry and Forest Fndustries a n d it i s edited by
A
Andersson, M. Kallio, A. Morgan, a n d R. Seppiilii. I t contains a r t i c l e s written by s c i e n t i s t s from countries, from both East a n d West. The topics cover economic a s well a s noneconomic issues of f o r e s t s a n d forest i n d u s t r i e s a t t h e m i c r o a n d m a c r o scale, including international a s p e c t s of t h e f o r e s t sector.Markku Kallio Project Leader
Forest Sector Project
This article is a n overview of systems analysis in forestry a n d forest industries. The issues covered range from forest management and forest industrial strategy to international trade in forest products and struc- t u r a l change in the forest sector worldwide. The methodologies dis- cussed include mathematical models of economies, statistics, a n d opera- tions research.
CONTENTS
1. INTRODUCTION
2. A GLOBAL PERSPECTIVE 2.1. International Issues
2.2. Analysis of t h e Global Forest Sector 3. NATIONAL ECONOMIC ANALYSIS
3.1. Macro- a n d Microeconomic Issues 3.2 Macroeconomic Modeling
3.3 An Approach for Analyzing Economic Structural Change
3.4. Forest Sector Modeling 4. REGIONAL FORESTRY ANALYSIS
4.1. Timber Supply Issues 4.2. Forest Management
4.3. Nonindustrial Use of Forests 4.4 Forest Ecosystem Dynamics 5. CONCLUDING REMARKS
REFERENCES APPENDIX
-
vii-
! 3 Y X I X S ANALYSIS FDR
THE
FOREST SECTORby
R.E.
Andersson, M. Kallio,and R. SeppalaApplied s y s t e m andysis is often oriented toward improving Long- t e r m policy making. This implies an emphasis on strategic r a t h e r t h a n tactical or operational issues. Frequently t h e work involves generating policies for major changes concerning production and marketing, a n d the use of capital. labor, a n d raw materials. This m e a n s t h a t t h e analyst must e n s u r e close cooperation with policy makers a n d planners
-
whether they a r e in industrial firms or i n regional or national govern- ment.
Such long-term policy analysis of necessity involves some "well- behaved" uncertainties, a s well as others t h a t a r e less conveniently s t r u c t u r e d . Problems of predicting t h e available labor force, and t h e availability of raw materials a n d energy a r e notorious in t h i s respect.
The uncertainties a r e even g r e a t e r with respect to t h e p r e d c t i o n of future demand s t r u c t u r e s . It Is hard t o make reasonable, quantitative predictions about t h e technologies and the politics t h a t will predominate over a planning period of two t o t h r e e decades. It is often n e c e s s a r y for this reason t o use scenarios, sensitivity analysis, and o t h e r similar approaches t o provide insight into probable consequences of t h e funda- m e n t a l uncertainties involved in long-term policy making.
Modeling plays an i m p o r t a n t role in applied s y s t e m s analysis, so m u c h s o t h a t t h e two a r e sometimes assumed t o be identical. Most systems-analytical studies a r e based on one o r m o r e explicit models.
Often, however, models a r e used throughout the process, but t h e i r use is never made explicit t o t h e final user.
When a policy decision is being made, i t would be extremely helpful for t h e decision maker t o know what would be t h e consequences of his choice. Models used for predicting t h e s e consequences a r e often called systems-analytical models. Examples of basic modeling techniques a r e optimization, simulation, gaming, and game-theoretical models. A given model may.employ m o r e t h a n one of t h e s e techniques.
The analysis often r e l a t e s socioeconomic, ecological, a n d technolog- ical systems t o each o t h e r in a n essentially dynamic a n d regional model- ing effort. This implies t h a t frequently a large number of variables have t o be interconnected. I t i s up t o t h e analyst t o make a sensible trade-off between realism, simplicity, and possibilities of estimation.
In tactical and operational m a n a g e m e n t analyses using models, the aim is often t o g e n e r a t e quantitative recommendations or forecasts. For
applied systems analysis oriented toward long-term policy problems, a more moderate goal would be t o formulate qualitative policy recommen- dations or policy optinns as rules of thumb. This formulation involves a certain amount of judgemental information concerning possible substan- tial changes in t h e environment of t h e system modeled.
For very long-term perspectives even this may be too ambitious. In these cases applied systems analysis can only be used t o c r e a t e a better understanding of the long-term policy problems and their interdepen- dence. To use models for such pedagogical purposes, it is very often necessary t o generate projections for t h e future. In order t o t r a c e out t h e consequences of possible assumptions concerning uncertainties, this type of projection is defined as a series of scenarios.
m e forest sector comprises two main components: forestry and the forest industry. The forest sector concept integrates all aspects con- nected with forests and t h e i r exploitation, i.e. activities from timber growth to t h e use of end products. Ecological, environmental, a n d socioeconomic factors a r e also included in this definition.
The forest sector has a number of specific features that influence its planning and policy making:
Although forests a r e a renewable natural resource, t h e produc- tion time, i.e. t h e growth period of t r e e s from seeds to logs, is usually very long, in t e m p e r a t e a r e a s close to a hundred years.
As a result of t h e long rotation time, structural changes in forests cannot take place very quickly. In addition, t h e soil and climatic conditions often restrict t h e options for timber grow-
ing dramatically.
The growing stock of trees is a t the same time both a product and a production machinery in which annual growth accumu- lates. This gives flexibility in choosing t h e exact time for t h e realization of production.
Wood is one of the most versatile raw materials. Its use for fuel.
for housing and o t h e r construction, for furniture, for printing.
packaging and household paper, for rayon and cellulose deriva- tives, and as a chemical feedstock gives t h e forest sector a very diverse potential within t h e economy.
The forest industry is a processing industry, t h e bulk of which is very capital intensive. The normal life-span of machinery is several decades. (Some paper machines built before t h e first world war a r e still operating.) This is one reason why t h e forest industry is r a t h e r conservative and not very flexible.
Production technology in t h e forest industry is t o a large extent based on old a n d well-known principles. Therefore, t h e technology i s international, and productivity is tightly con- nected to the age and size of t h e production plant.
The life cycle of different forest products is long. Innovations have usually only meant improvements in existing products, and completely innovative goods have appeared on t h e m a r k e t very rarely. One of the implications has been that price has become a pronounced factor in market competition. Therefore,
world-wide profitability has become relatively low, and t h e trad- itional forest industry is often characterized as a m a t u r e indus- try.
Consumption of forest products results almost entirely from input needs in o t h e r areas, such as the construction a n d infor- mation sectors.
In r e c e n t years t h e r e has been a marked pronounced i n t e r e s t in the introduction of systems-analytical approaches for studying t h e problems of t h e forest sector. The International Institute for Applied Systems Analysis (IIASA) has, for instance, launched a large-scale effort in this direction. In this project, scientists from more t h a n twenty countries have been engaged in t h e development of systems-analytical tools t o study development policies for t h e forest sector.
Although applied systems analysis is oriented toward interdepen- dences between all levels of decision making, the combination of a large number of interrelated variables leads t o a practical need for decomposi- tion: different parts of t h e analysis a r e separated from each other.
Ideally such a decomposition should be made so t h a t interdependences between components are few but strong.
The following decomposition of t h e forest sector outlines t h e organi- zation of the different articles in t h i s volume:*
Global analysis
'Ruoughout the article, refers to papers included in %terns Analysis j o t lcbtost~y and Fbrod huiustrios, a volume edited by k Andersson, P. Kallio, A. Morgan, and R. SeppBlB (forthcoming).
National macroeconomic analysis Forest sector analysis
Forestry management Nonindustrial use of forests Ecosystem dynamics
We shall discuss each topic in t h e following sections.
2.1. International Issues
As the demand for forest products increases, world supply is con- strained by t h e availability of wood raw materials, higher production and transport costs, environmental concerns, and competition between industry and agriculture for land. For example, according t o r e c e n t esti- m a t e s of annual wood removals for the period 1970-2000, t h e Nordic countries, t h e traditional suppliers of Western European markets, have reached their wood production limit and cannot increase their average production rate. Although t h e r e is some growth potential in t h e forest industries in o t h e r Western European countries. scattered ownership patterns a n d environmental issues limit this potential. Increasingly, Western European forest industries are becoming dependent on imports.
To remain competitive, t h e y a r e concentrating on end products t h a t a r e easily transported and t h a t do not have cost structures dominated by wood costs. In contrast, p a r t s of North America still have forest
resources with potential for the future, a s well as a developed industrial infrastructure t h a t would allow industries in this region to become major suppliers of wood products on the world market.
Although most of t h e forest industry capacity is in t h e industrialized countries, m o r e than half of t h e estimated forest land area of the world is located in t h e developing countries. Besides being raw material for forest product industries, these forests play an important role in provid- ing fuel and land for agriculture. However, temporary relief from food shortage h a s frequently occurred a t the expense of the biological poten- tial of t h e forests. Forest land once stripped cannot support agriculture for long due t o erosion or low soil fertility. The devastation of forests is one of t h e most serious problems in developing countries. Each minute, some 20 h e c t a r e of tropical forest vanish and the relative pace is increasing
.
The global forest sector system is an assembly of interacting national systems. The interaction takes place mainly via international trade. At present, such trade in forest products is small relative to total world production
--
about 85% of woodpulp, paper and paperboard is con- sumed in t h e country where i t is produced. Moreover. trade flows in quite circumscribed paths and so is of unequal significance t o diflerent countries and regions. However, t h e foreseeable changes, especially in the availability of wood raw material and t h e cost s t r u c t u r e s of products, may cause drastic changes in t h e patterns of trade. whose total is grow- ing rapidly. The following factors have a fundamental impact on t h e development of world trade:Regional demand
Regional forest resources Relative production costs Transportation costs Trade policies
Exchange-rate policies
We shall now discuss each of these factors in detail.
Demand for Forest Products
Future demand for forest products is characterized by changing response patterns in different countries and areas. Over the long term, the impacts of a number of technological changes need t o be evaluated.
Examples such as advances in electronic information technologies, super absorbent materials, and packaging substitutes will afTect the demand for forest products. This will be noticeable earlier and more strongly in some countries than in others. Also the impacts of the changing energy scene on both forest products and on competitive products need con- sideration.
The large variations between different countries cause considerable prediction problems. The consumption of paper has, for instance, been more than twice a s large in the
USA
than in Switzerland, over the same given time period when the two countries had approximately the same standard of living. I t is thus evident t h a t simple econometric analysis of price and income elasticities is not sufficient to permit us to understand and predict the level of demand. I t has been argued by industrialiststhat the use of packaging paper, for instance, is related to the whole structure of production and consumption of commodities, and to the packaging requirements related to spatially dispersed producers and consumers. Modeling long-term demand and consumption development must therefore take into account not only income and price develop ment but also developments in t h e location patterns of producers and consumers as well as life styles associated with in different parts of the world.
Forest Resources
Problems of future resources can be approached from several different viewpoints. First. t h e long-term resource potential may be stu- died. A steady-state analysis of forests demonstrating the ultimate potential could be a starting point (see for example, Kallio and Soismaa*). The advantage of this approach is that it avoids the mental traps of restricting the analysis t o minor extensions of current prac- tices.
Second, supply potentials under alternative social and agricultural land-use policies can be examined in conjunction with varying economic incentives for timber production. Provisions for fuel and agricultural uses of wood, taking into consideration t h e problems of erosion, may also be advisable.
Production and Transportation Coata
Another aspect of the supply of forest products is t h e restructuring process that forest industries are facing in many countries. A primary
issue is t h e comparative advantage of t h e industry in t h e country con- cerned, i.e. how the relative abundance of wood raw material and the cost of inputs affect capital investments in t h e industry. Also, hypotheses from p r o d u c t c y c l e t h e o r y m u s t be included in the analysis.
Key cost factors in t h e forest industry a r e wood, energy, labor. t r a n - sportation, and capital costs. In t h e Nordic countries wages and capital costs a r e the most important items and their combined s h a r e is about half of t h e export price t o Western Europe. On average, stumpage paid t o forest owners, energy costs a n d t r a n s p o r t costs each account for 10-15%
of the export price.
In t h e long term, t h e r e should be no significant differences in capi- tal, energy, and chemical costs between countries. In constrast, wages c a n vary considerably, but often differences in productivity bring t h e costs per product unit to t h e same level. The major differences in t h e production costs of t h e forest industry a r e t h u s found in wood a n d t r a n - sportation costs.
Table 1 shows t h e drastic international disparity in pulpwood prices in 1978. I t can be argued t h a t t h e data, indicate a disequilibrium in world t r a d e in wood and wood products, r a t h e r t h a n price differentials consistent with transportation a n d energy price differences.
Equilibrium in t h e world m a r k e t would exist if t h e price difference between markets were, a t most, t h e marginal cost of transportation for each commodity. The cost of transporting a Finnish forest product t o Western Europe constitutes between 10 a n d 25 percent of t h e product's sales price. In deliveries t o Western Europe, Sweden has a transport cost
advantage of 1-4 percent of the product's price compared t o finland, which itself has an advantage of 5-10 percent over US and Canadian sup- pliers.
Table 1. Typical rood cost at miU by region in 1978
( ~ ~ l / m g .
Southern British British Southern Brazil Sweden Columbia Columbia USA
(coast) (interior)
Sawlogs (pine) 5 1 27 14
Pulpwood (pine) 32 19 9
Pulpwood (hardwood) 27 17 8
Pulpwood (chips) 15 13
Source: Payry and Ryti (1979)
A general framework for analyzing t h e comparative advantage of forest industries is discussed by KirjasniemiL. The economic compara- tive advantage of forest plantations in a n u m b e r of tropical regions and implications for the traditional forest products supply regions are con- sidered by SedjoL.
Artificial Trade Barriera
Distance c a n be considered a n a t u r a l barrier t o trade. Tariffs, quota restrictions, subsidies, and trade agreements, on t h e o t h e r hand, a r e artificial resistance factors. Subsidies in particular have been used heavily in some major forest industry countries since t h e mid-1970s.
One factor of considerable importance for t h e 1980s and 90s is t h e growing tendency to r e t u r n t o protectionist or even mercantilist trade
policies. This phenomenon has been observed with increasing frequency since t h e energy crises, and it may have serious impacts on a number of countries such as Canada, Enland, and Sweden, which are strongly oriented toward the export of forest products.
Noncompetitiveness is not the only reason for the use of subsidies.
In some cases subsidies are used to create industries in certain regions, especially in those t h a t a r e less developed. They have also been used in efforts to influence t h e production s t r u c t u r e of t h e forest sector.
Exchange Rate Policies
The fluctuations in t h e relative value of t h e U S dollar have caused disruptions of world t r a d e patterns in forest products. During the last fifteen years, currency relations between some countries exporting forest products have developed as shown in Table 2. We observe t h a t t h e values of t h e Swedish and rmall Rnnish currencies were relatively high in t h e mid-seventies. Consequently, problems emerged for these coun- tries in both t h e i r international competitive position and their internal profitability. Within t h e last five years, however, t h e situation h a s reversed completely. In 1982-83 the position of these currencies changed markedly in relation to the U S and Canadian dollars. For this reason, t h e Nordic forest industry (especially i n Sweden) witnes'sed a remarlcable recapture of i t s international competitive power, with boom- ing profits and greatly improved possibilities for future expansion.
For industies t h a t strongly depend on international markets, uncer- tainty about t h e future values of currencies i s a major concern. Such industries could benefit substantially horn embarking upon portfolio-
Table 2. Currency exchange rates 1968- 1983 ( U S / 100 units of national curren-
4.
Canada 83 100 88 89 85 80 83 81 8 1
Finland 24 24 27 24 26 27 23 21 1 8
Sweden 18 21 24 22 23 24 20 16 13
Austria 3.8 4.3 5.7 6.8 7.5 7.7 6.3 5.8 5.8
Narnay 14 15 18 18 20 20 17 15 14
Brazil 31.8 18.8 12.2 5.5 3.7 1.8 1.1 .6 .3
Indonesia M 24 24 23 18 16 16 .15 14
Source: International Financial Statistics.
allocation strategies in t e r m s of investments, marketing, and c u r r e n c y holdings.
2.2. Analysis of the Global Forest Sector
According t o Ohlin (1933) (see also Heckscher, 1949), t h e theory of interregional and international trade is nothing o t h e r than a theory of interregional and international location of supplies. This follows from the basic assumption t h a t t h e pattern of location and t h e pattern of trade a r e simultaneously determined. According t o t h i s view. countries tend t o specialize in t h e production of commodities t h a t contain a rela- tively large proportion of fairly immobile resources t h a t a r e relatively abundant in t h e country concerned.
This implies t h a t , although certain resources a r e not themselves directly mobile, t h e export of commodities containing a large amount of such immobile resources would be equivalent t o a migration of t h e resources. F'rom this follows t h e Heckscher-Ohlin factor-price
equalization theorem: The indirect migration of factors would, in the long run, tend to even out t h e factor prices of different regions. Any absolute factor-price equalization is of course impossible as long as there are transportation and communication costs and other constraints on mobility.
It is evident t h a t free trade is more beneficial to t h e world economy than an autarkic system. In an autarkic system, welfare in a given region is determined by the resources in that region. Any erasing of this con- straint will improve t h e possibility of achieving a higher level of welfare.
Free trade corresponds to a summation of regional resource constraints.
and thereby to a relaxation of constraints for the individual regions.
The only constraints t h a t cannot be removed are those associated with the immobility of factors, commodities, and information. The analysis must thus concentrate on the initial distribution of such factors (including resources and technological how-how) and on t h e communi- cation and transportation systems.
Based on t h e static theory of comparative advantage, various propo- sitions have been formulated that are relevant for an understanding of the future interregional division of labor in the world (see e.g., Ethier,
1983). The first of these propositions is called the Rybczynski theorem. It states:
At c o n s t a n f f a c t o r p r i c e s , an increase in the e n d o w m e n t of a f a c t o r u s e d in a t l e a s t t w o s e c t o m , w h i c h l e a v e s that f a c t o r F l y e m p l o y e d , p r o d u c e s a m a r e than proportional r i s e in the o u f p u t of s o m e good and a f d l in ths o u t p u t of s o m e othar good
As an application, consider the European and North American situa- tion. Assume t h a t wood resources are constant (over time) for Europe but growing in North America, and that the capital market in North America is tight (allowing no expansion of capacity in this illustration) whereas production capacity in Europe is growing. Then, applying the Rybczynski theorem twice, chemical wood processing in North America is shifting over time toward wood-intensive commodities like pulp, and in Europe toward capital-intensive (and wood-extensive) products like paper. This is in fact the conclusion formulated by Ryti*, a conclusion substantiated by the Rybczynski theorem.
Associated with t h e Rybczynski theorem, there is the Stolper- Samuelson theorem:
,4n i n c r e a s e in t h e p r i c e of an initially produced good u s i n g an a s s o r t m e n t of at l e a s t t w o f a c t o r s n e c e s s a r i l y c a r s e s some f a c - t o r p f i C e t o r i s e in e v e n g r e a t e r p r o p o r t i o n a n d s o m e o t h s r f a c - t o r p r i c e t o fall.
In a two-factor/two-commodity world, this theorem might be interpreted as follwws. Consider economy producing paper and pulp, and using machinery and land as factor inputs. Let k l and l , , respectively, be the quantities of machinery and land necessary per unit output of paper, and k 2 and l 2 similar coefAcients for pulp. We denote the price of paper by p l , the price of pulp by p2, the rent of machinery by q , and land r e n t by w . If sales prices a r e entirely imputed to factors of production, w e have
When the c o m m o d t y prices p l or p 2 change, the factor prices q and w m u s t also change. If paper is more machinery intensive than pulp, i.e. if k l / l l
>
k 2 / L 2 , t h e n the rise in t h e price of paper will cause a relative rise in t h e r e n t of machines t h a t is g r e a t e r t h a n t h e relative price increase of paper, a n d it will also cause a fall in t h e land rent. For t h e same reason, a rise in t h e price of pulp, which uses relatively more land, will cause a relative rise in land r e n t t h a t is g r e a t e r t h a n t h e relative price rise, and a fall in t h e r e n t of machines.These ideas formulated by Ohlin, Rybcyznski. Samuelson, a n d Stolper c a n be extremely useful in understanding t h e simultaneous determination of location of supplies, demands, and trade. The paper by Buongiorno* is an example in this tradition of interregional location a n d t r a d e analysis, applied to a study of t h e US pulp a n d paper market.
For dynamic analysis of these issues, other procedures have been proposed. One prominent concept is t h e p r o d u c t c y c l e t h e o r y (Vernon.
1986). According to this theory. every product tends t o follow a l o c a t i o n c y c l e . Each product cycle s t a r t s in one of the developed economies where research and development leads t o t h e introduction of t h e product (or product quality). In t h e primary stage, profitability is high a n d t h u s o t h e r countries with large research a n d development capacity will quickly i m i t a t e t h e country t h a t originally introduced t h e new product.
In this stage, t h e knowledge about t h e new commodity is diffused among t h e most developed countries, which become the major world suppliers.
Thereafter, t h e product technology becomes more widely known a n d ordinary comparative advantage determines t h e interregional pattern of production according to t h e availabilities of production factors a n d
constraints on trade. Due t o t h e maturing of t h e technology the final phase is t h e n reached, during which a collapse of the industry in t h e ori- ginal countries of specialization may occur.
The essential difference between the theory of comparative advan- tage and the product cycle theory is the emphasis placed upon techno- logical change a s a driving force. In the traditional theory of compara- tive advantage, t h e technology is assumed t o be given and in some ver- sions of the theory even known t o every participant in international exchange. By c o n t r a s t , the theory of product cycles regards the research and development process as the essential driving force in the changing pattern of technological knowledge a n d in t h e international pattern of supply location.
Irrespective of what determines supplies in the different regions, t h e r e a r e a n u m b e r of ways of determining t r a d e flows in a model. Two basic approaches a r e a s follows:
Simultaneous determination of location of supplies, regional demands, a n d t r a d e between the different regions, e.g., Leontief e t al. (1977);
Stepwise analysis i n which supplies a n d demands a r e deter- mined for each region separately, after which a procedure of Linking is used t o predict t h e t r a d e flows, e.g., t h e
LINK
project(Klein, 1976).
Both procedures have t h e i r own a d ~ a r i t ~ a g e s a n d disadvantages in t e r m s of analytical consistency and information needs. The simultaneous approach implies a gain in consistency but quite often a loss in t h e qual-
ity of the information on resources and technological conditions for each region. A linking system means that the heavy burden of regional infor- mation gathering is decentralized while the coordinating effort can be focused on balancing supplies and demands through trade flows. Furth- ermore, this procedure puts a greater responsibility for modeling on each of the participating regions.
If a linking procedure is chosen, t h e r e still remains the choice of a mechanism for determining t h e trade flows. When choosing such a mechanism, one is forked t o ask whether the system actually behaves rationally (evolving according t o some welfare criteria and possibly even in an almost deterministic way) or whether it behaves more or less erratically. If t h e first assumption is valid, there remains t h e problem of choosing a p e r f o r m a n c e c r i t e r i o n . Among t h e criteria suggested, t h e fol- lowing a r e prominent:
Maximize economic surplus (equivalent t o generating a con- strained competitive equilibrium).
Maximize t h e s u m of proflts (equivalent t o a monopolistic equilibrium, with prices s e t by firms).
Minimize total cost subject t o quantitative constraints (corresponding t o a speciflc oligopolistic equilibrium solution).
Minimize t r a n s p o r t costs (a special case of t h e preceding cri- terion).
If rational behavior cannot be assumed or if our knowledge of t h e behavioral relations is of limited quality, a suitable approach is t o deter- mine a s t o c h a s t i c o u t c o m e . This often amounts t o maximizing the likeli-
hood of the outcome subject to the economic and political constraints that can be formulated a priori.
3. NATIONAL HIKESI' SECTOR ANALYSIS
3.1. Mmxo- and Microeconomic Issues
Supply of Capital
The investment of resources in the forest industries is an issue of primary importance in many of the forest product supply countries. The allocation of capital between the paper and pulp industry and other sec- tors of the economy is of great concern in countries like Canada, Sweden, and Finland. Particularly in t h e United States, Canada, and Sweden, there has been a long-term pattern of low savings and invest- ment, as compared to countries like Japan, Knland, and Norway. The general lack of investment resources is especially problematic for the paper and pulp industry. which is extremely capital intensive. The capi- tal: output ratio for the paper and pulp industry is normally larger than 6, which is the same order of magnitude as the ratios for the basic metal industries. Economies of scale in integrated paper and pulp plants only deepen the problem.
Regional Ehnployrnent
The supply of and demand for labor for the forest sector has in some countries become a question of a trade-off between regional employment goals and the goals of industrial growth and profltability. Therefore, the
allocation of labor between &fferent economic sectors cannot always be analyzed a t t h e national level, but must sometimes be performed at the regional level.
Labor demand is often unnecessarily constrained by the wage costs associated with central negotiation schemes for the labor market. In such cases, it can even be nationally efficient to subsidize employment in t h e forestry regions. With a general employment subsidy, the forestry firms normally located in areas of high unemployment also have the option of hiring otherwise unemployed labor a t lower wage costs. Such schemes have been successfully developed for the Scandinavian coun- tries and are now in use.
Microeconomic h u e s
The most important decisions affecting t h e structure of an industry a r e decisions on investment. Four major interconnected questions must be answered: What, in terms of products? How, in t e r m s of production technology and capacity? For whom, in terms of markets? WAere, in t e r m s of location of production sites?
The main factors d e c t i n g product choice in investment in t h e forest industry are:
future demand for the products.
market situation : competition, trade barriers.
wood availability and price.
availability and price of other production inputs: energy, chemicals, capital. human resources.
production technology available.
If neither the m a r k e t s nor the technology were to pose unavoidable bar- riers and if the production inputs were available in abundance a n d a t a fixed price, t h e choice between different products could simply be based on a profitability criterion, i.e. r e t u r n on investment. But when t h e pro- c u r e m e n t of wood is, in fact, a bottleneck and the price of wood is deter- mined on roundwood markets, ordinary profitability criteria a r e no longer so useful. Instead, t h e profit maximization is subject to resource and o t h e r constraints.
The manufacturing cost per product unit usual!y decreases with increasing mill size. However. with growing mill size t h e wood transpor- tation costs increase. This, together with the finite amount of available capital, imposes a limit on how large a mill should be. The optimal mill size is likely t o grow with technological development. In most tropical regions with fast growing trees, t h e lack of infrastructure a n d skilled labor reduces the advantages of economies of scale. In addition, t h e lack of capital may prevent t h e building of big mills, even though t h e raw material base would in principle provide good opportunities. In general, economies of scale in a sawmill are much more limited than in a pulp or paper mill.
In addition t o t h e size of t h e mill, t h e r e a r e other mill-related fac- tors t h a t affect t h e profitability of production. These are the e x t e n t s of vertical integration (i.e. using the product of one line as a raw m a t e r i a l for another line in t h e same mill) a n d horizontal integration (i.e. group- ing of products i n t o a multiproduct complex t h a t uses different species e n d assortments of wood a n d manufactures several end products). The
competitiveness of primary production can to some extent be improved by vertical integration. Most standard products of t h e Nordic paper industry a r e currently made in integrated mills.
The aim of horizontal integration i s t o utilize wood species and timber assortments in t h e same proportions as t h e y occur in t h e forest.
The manufacturing costs, capital charges, and savings in wood harvest costs can be about one-fifth less in a multiproduct complex t h a n for separate mills m a h n g t h e same products.
The forest sector a n d especially t h e wood industries have tradition- ally paid little attention t o R & D. In most cases, t h e wood industries devote less t h a n 1/20th of t h e s h a r e of value-added t h a t gives t o R k D
investments in o t h e r manufacturing industries. This h a s serious impli- cations for t h e f u t u r e of wood as an input in mechanical and chemical processing. Other feedstocks, especially oil and non-ferrons alloys, a r e the subject of m u c h m o r e intensive research and development which, in t h e long r u n , will tend t o influence t h e competitiveness of those feedstocks in relation t o wood.
3.2 Macraeconomic Modeling
The problem of analyzing a set of industries like t h e forest indus- tries is essentially a question of handling interdependences of a techno- logical and economic nature. The standard method is in.@-outpuf a d y s i r , where each s e c t o r is characterized by a technological input- output coe&ient %j indicating the amount of commodity i t h a t is required t o produce one unit of commodity j. The s e t of such input- output coefficients c a n be assembled i n t o an i n p u t - o u t ~ u t matrix
A =
i%jj =
)z../z,j, where zj is the gross production of commodity j and21
zi, is the required input of commodity i. Forming t h e simplest possible static balance requirement for the economy we get
z = A z
+ I
where z
=
(zi) and j=
( I i ) is the final demand (defined to be exo- genously determined consumer, government, investment, and net export demand). Solving for the equilibrium structure z* of production essen- tially amounts to calculatingz
=
(I-A)-' fThe prices a r e determined in this type of national model by the require- ment t h a t the price pi of each commodity i (a
=
1,2,...,
n ) covers t h e costs of all inputs:p = p A + w
where p = (pi) and w
=
(ol ali) is a vector indicating the cost of labor (and possibly o t h e r primary inputs) for each product, with ali being the labor: output coefficient in sector i and wl being t h e wage rate.As a practical example, assume that the wage r a t e increases in the sawmill sector. This corresponds t o an increase of, say, A wi in the com- ponent of o corresponding t o sawmill operations. The resuting change in the price vector is then given by
The demand for primary inputs (which may not refer t o labor alone) can be calculated by premultiplying the production vector z* by the vec- tor al
=
(ali) of primary inputs. In the case of Labor,L* =
al(I-A)-'jgives t h e equilibrium employment. If we assume t h a t t h e final demand is increased by Af
=
(0 ,.... Afi ,...,o ) ~
(say, an increase in t h e demand for newsprint), the increase in employment isIncreasing newsprint demand implies an increased demand for pulp, which in its t u r n implies an increased demand for chemicals. The increased demand for chemicals implies a f u r t h e r i n c r e a s e in the demand for oil, etc., in a diminising series t h a t finally involves t h e whole economy. Thus t h e influence of any change in final demand for a forest product
-
in t e r m s of consumption, government demand, n e t exports, etc. -- spreads both within t h e forest sector and onto other sectors, as well as returning eventually t o t h e forest sector itself.The influence of investment can be analyzed within a n extended input-output framework. In t h i s case investments in new machinery, building, etc., a r e distinguished in the demand f . Let bii be an investment-output coemcient indicating t h e input of commodity i needed t o achieve unit growth in the production capacity of sector j , a n d let t denote time. Then t h e dynamic input-output model is given by t h e balance requirement
z ( t )
=
Az(t)+
B ~ ( t )+
f ( t ) ,where B
=
(bij) a n d i ( t ) is t h e time derivative of production, i.e, t h e r a t e of increase in capacity. I t can be shown t h a t any decrease in an input- output o r investment coefacient will increase t h e general r a t e of growth and lead t o a s t r u c t u r a l change in the economy.This differential-equation system cannot be solved directly because of numerical stability problems (Brody, 1969). Various transformation procedures have, however, been suggested t o overcome s u c h problems, a n d models solved in this way a r e c u r r e n t l y being used in some countries t o forecast a n d plan long-term sectoral development.
I t is possible t o show t h a t under these assumptions t h e r e exists a maximum growth r a t e with a corresponding balanced composition of industries. Such an equilibrium growth r a t e corresponds t o t h e minimal
equilibrium r a t e of inte~est associated with a balanced set of relative prices of all commodities.
One disadvantage of t h e class of models described above i s t h e a s s u m e d inflexibility of technologies. A n u m b e r of procedures have been proposed t o overcome t h i s weakness. One m e t h o d is due t o von Neu- m a n n (1937). He proposed a formulation in which
Q z r A z + f Quantitative equilibrium condition
P Q ~
AP
+CJ P r i c e equilibrium conditionp f
=
w z General equilibrium conditionswhere A is an input matrix, Q i s an o u t p u t matrix, f is t h e final d e m a n d a n d w is t h e cost vector of p r i m a r y inputs. The approach involves two basic assumptions:
Each activity c a n produce many commodities, including capital goods (joint production).
Each commodity c a n be produced by a number of activities (substitution).
These assumptions imply t h a t t h e m a t r i c e s Q and A may be rec- tangular. Otherwise t h e method follows t h e assumptions of input-output theory, of which i t t h u s is a generalization. This model is closely related t o t h e theory of linear programming, although it dates back t o t h e 1930s.
In fact, i t i s equivalent t o choosing technologies and production levels t o satisfy final demand a t minimum costs in t e r m s of primary inputs.
A model based on input-output theory but allowing for partial substi- tution of inputs h a s been developed by Johansen (1972 and 1974). Avari- a n t of this model is described by Sohlberg*. In t h i s approach, labor, capi- tal, and energy a r e substitutable inputs, while all other inputs a r e regu- lated by Axed input-output coefficients.
Mathematical programming methods for handling t h e interdepen- d e n c e s between a sector like t h e forest s e c t o r and t h e r e s t of t h e econ- omy have been proposed by Dantzig (1963). The idea h e r e is t o use an optimization model of t h e forest sector, in which a n efficient choice of technologies i s calculated assuming various supply functions for r e s o u r c e s and demand functions for products. The prices a r e t h e n deter- mined t o g e t h e r with t h e s e t of input-output coefficients optimally selected by t h e model. This approach h a s been adopted by Kallio, Pm- poi, a n d Sepp2ilig for studying the Finnish forest sector.
3.3 An Approach for Analyzing Economic Structural Change
The development of an industry is determined both by external technological and by consumer demand changes influencing corpora- tions a n d t h e i r plants. An approach for analyzing t h e s e processes was
introduced by Salter (1960) a n d f u r t h e r refined by Johansen (1972). In t h i s approach, plants are a s s u m e d to be flexible regarding substitution a t t h e i n v e s t m e n t s t a g e only. After investment, e a c h plant i s almost rigid in t e r m s of t h e energy, labor, a n d o t h e r input r e q u i r e m e n t s p e r unit of output. At this l a t e r stage, substitution can only occur a t t h e industry or corporate level by opening up new plants o r closing down units con- s t r u c t e d earlier. Another f e a t u r e of t h i s t h e o r y is t h e asymmetry of t h e closing down a n d i n v e s t m e n t criteria. The c l o s u r e of a plant occurs when t h e product price decreases below t h e average variable cost.
Investment i n a new plant occurs when t h e expected price exceeds t h e average variable cost a n d a p r o p o r t i o n of t h e lixed cost, properly discounted.
The analytical procedure can be i l l u s t r a t e d by t h e empirical diagram in Figure 1, where t h e shaded a r e a gives t h e gross profit of t h e paper industry in Sweden in 1978. Similar productivity a n d cost curves a r e now produced for t h e Swedish a n d Norwegian i n d u s t r y s e c t o r s on a n annual basis. The s t r u c t u r e a n d potential u s e s of s u c h a d a t a base a r e discussed by Johansson*.
Against t h i s background a model of economic s t r u c t u r a l change can be forrnu lated based on t h e following assumptions:
Each i n d u s t r y in e a c h region consists a n u m b e r of production units of given vintages with given o u t p u t capacities.
All production techniques a r e characterized by coefficients t h a t vary a c r o s s vintages a n d sectors only.
Unit Labor Cost:
In In
Percent of Total Employment
Kgure 1. h b o r denand of the j w ~ p e r ' h ~ in1978 and 1980 m Sreden Only labor i s considered as primary input, though it is feasible to extend t h e model to cover o t h e r inputs such as energy.
The investment possibility in each industry and region is given by a new vintage with a certain " b e s t p r a c t i c e t e c h n o l o g y " . Existing capacities a r e utilized a n d new capacities created according t o an efficiency criterion of maximizing total value added for manufacturing industry as a whole. This is equivalent
t o profit maximization under constraints reflecting t h e costs of production.
Various industrial policy goals c a n be included e i t h e r in t h e objec- tive function or as constraints. The basic formulation employs maximi- zation of total value added. The other policy goals concerning ernploy- m e n t a n d production, by industry and region, a r e taken into account by m e a n s of appropriate constraints. Additional constraints may specify availability of labor by region, availability of capital by industry, a n d employment in new vintages. The model framework i s basically s t a t i c , but t h e inclusion of t h e possibility of investment in new vintages allows some dynamic features t o be simulated. The various sectoral a n d regional economic activities a r e only r e l a t e d t o e a c h o t h e r through com- petition for some resources such a s labor and investment.
3.4. Forest Sector Modeling
In t h e preceding section i t was implicitly a s s u m e d t h a t a forest sec- t o r model exists t h a t i s compatible with modeling a t higher levels of aggregation. This is n o t always t h e case. Forest s e c t o r models a r e often c o n s t r u c t e d without due consideration for r e q u i r e m e n t s of consistency with macroeconomic models in t e r m s of s e c t o r a l disaggregation. t i m e periods t o be covered, etc. This is an unfortunate situation in c e r t a i n respects, b u t may also have advantages in o t h e r aspects: consistency in modeling simplifies cooperation whereas inconsistency can f u r t h e r t h e f r e e development of new modeling ideas.
Forest s e c t o r models may be distinguished in a n u m b e r of ways:
disaggregation of inputs and outputs, t r e a t m e n t of t i m e , t r e a t m e n t of regions, t r e a t m e n t of nonconvexities a n d nonlinearities of relations (such a s economies of scale), and behavioral c r i t e r i a s u c h as optimiza- tion for single o r multiple criteria. This section is limited t o t h e con- sideration of t h r e e categories of models of particular significance. These are:
Dynamic Simulation
Mathematical Programming Spatial Equilibrium Models
Before t h e different mathematical models a r e introduced, we briefly discuss forest s e c t o r interactions in a s t a t i c equilibrium situation.
A
Qualitative Anal* of ' h e Forest SectorIn order t o illustrate t h e internal i n t e r a c t i o n s between different parts of t h e forest s e c t o r , we will u s e t h e diagram in Figure 2 (see e.g., Wohlin, 1970). To simplify t h e discussion we a s s u m e t h a t t h e forest sec- t o r consists of paper a s t h e flnal product, roundwood a s an intermediate product, and forest stands. The components a r e r e l a t e d t o e a c h other a r o u n d a common point or origin, a s shown in Figure 2.
This system r e m a i n s in equilibrium a s long a s a is large enough t o induce repairs a n d maintenance of t h e supply potential. If we assume t h a t t h e demand function increases and t h e wood supply price remains constant. then t h e price of paper increases relative t o wood and t h e incentive to invest also increases. When t h e resulting new capacity is
(d) Price relation of paper and wood (a) The market for paper
Price of wood
Unit cost of wood
(c) Wood supply
I
QuantityI
of paper,
Non-paperuses of wood
I
(b) 'The transformation of wood into paper
F3gme 2. Diagxwn of interactions w i t h i n the forest sector.
added, the price of paper relative to the price of wood decreases. This
&minishes investment inducements and the system r e t u r n s to equili- brium. An improved raw material efficiency (increased output per unit input) would lead to a reduced price of wood relative t o paper. Conse- quently, increased inducement t o invest restores t h e r e l a t i v e price of paper but a t a lower equilibrium level. If t h e non-paper use of wood increases, c e t e r i s p a r i b u s , this leads to an increasing price for wood, thus decreasing t h e r e l a t i v e price of paper. This induces capacity with- drawal, which continues until t h e price relation is again in equilibrium.
As illustrated by these examples. disturbances of t h e system a r e self-stabilizing. Models of the forest sector, although dynamic or disag- gregated in many forest-product sectors, a r e normally built along t h e lines discussed in this section.
A dynamic simulation model is built on t h e fundamental assumption t h a t t h e r e is a s e t of dynamically interacting decision bodies (con- strained by technologies, resources. and other external factors) t h a t determine a development trajectory. Generally a simulation model is a system of difference equations (for each time period
t
):where F and G a r e given functions, zt is a vector of endogenous vari- ables, ut is a vector of decision variables (control variables), and rt and
tt
a r e vectors of stochastic disturbances, for each time period tA n example of such an approach is given by Lonnstedt*. His approach is r a t h e r generally applicable for studying t h e long-term development of t h e forest sector a t t h e national level. An application is given for the specific case of Sweden.
I t is well lmown from t h e theory of difference equations t h a t a sys- t e m of this type h a s to be restricted in t e r m s of both functional forms and p a r a m e t e r values in o r d e r to obtain a well-behaved solution trajec- tory. As an example, i t was shown many years ago by Slutzky (1937) t h a t e r r o r s in t h e starting position of t h e system c a n be propagated through time and t h u s give rise t o cycles, which would otherwise not occur. In m a n y simulation models of t h e forest sector, formal stability analysis is made possible by c e r t a i n linearity assumptions o r by exclusion of sto- chastic elements. In s u c h c a s e s t h e model m i g h t take on t h e form
where A , B , C, a n d H a r e given time-dependent matrices, and z ( t ) = ( z t . z t - l , ~ t - 2 B . . . and u ( t ) = ( ~ ~ . t + , ~ , u ~ - 2 , ' . * ).
Mathematical Programming
Dynamic linear programming (DLP) models of t h e forest sector employ deterministic versions of t h e linear system described above as an essential p a r t of t h e c o n s t r a i n t system. In addition to this constraint system, t h e
DLP
usually requires constraints on t h e initial a n d terminal states. Solutions of t h e model a r e obtained by formulating a goal func- tion a s follows:maximize
C
[ w t z t+
q tI+
]+
W T ~ Tt < T
where wt and qt are given vectors, and T is the terminal time period. By superimposing a maximand, many of the equilibrium and stability prob- lems of simulation models can be solved. Examples of such DLP approaches for forest sector analysis are given by Kallio, Propoi, and Seppala* for Finland and by Hultkranz* for Sweden.
An essential problem with DLP models for t h e forest s e c t o r is t o define a reasonable goal function to represent t h e objectives of t h e industry and t h e multitude of other objectives associated with t h e use of forests for recreation, as watersheds, etc. For this reason considerable efforts have been made in t h e field of m d t i o b j e c t i v e p r o g ~ a m m i n g . Work in t h i s area by Kallio, Lewandowski, and Orchard-Hays (1980) is an exten- sion of the DLP model of t h e Finnish forest sector of Kallio e t al.+ See also Rosenthal and Harrison* for an application of multiobjective optimi- zation.
Another application of mathematical programming is reported by Hyman*. His study aims t o assess alternative policies concerning fuel- wood in t h e Philippines. The assessment considers economic, social, a n d environmental aspects of t h e problem.
Spatial Equilibrium Analysis
Regionalization of forest sector models is often required, because of geographical variation in ecological, institutional, or economic condi- tions. The primary problems associated with regionalization a r e a geometric increase in t h e number of variables a n d a n a r i t h m e t i c
increase in t h e number of constraints; i.e., a considerable increase in model size.
Naturally, simulation and programming approaches apply to regional analysis a s well. However, the spatial equilibrium approach is particularly suited for application to market economies as a means of determining an efficient allocation of production (over regions) t o satisfy the demand in various regions (Lefeber. 1958). Such allocations a r e assumed to be determined on the basis of cost competitiveness, subject to constraints on resources as well as constraints determined by national and regional policies.
The spatial equilibrium approaach has been adopted by Dykstra a n d Kallio (1984) for studying long-term developments in the global forest sector. For another example of spatial equilibrium analysis, see Adams and Haynes*. Their model has been developed for t h e US Forest Service for long-term planning purposes. This Timber Assessment Market Model (TAMM) involves regional production and consumption of roundwood and mechanical wood products in t h e US. Another application of a spatial equilibrium model is given by Buongiorno*. The long-term supply (and price) of timber has been studied by Kallio and Soismaa*, whose approach is also based on t h e economic equilibrium concept.
4. REGIONAL FORBSITN ANALYSIS
4.1. Timber Supply Iesuea
In countries with a long tradition of forest industries, some of the most serious internal ~ r o b l e r n s a r e connected with the availability a n d
cost of wood raw material. The main constraints on t h e availability of wood a r e t h e following:
The physical and economic limits on wood production. The Nordic countries, for example, a r e approaching t h e economic constraints.
The mixture of t r e e species in t h e tropical and subtropical regions.
Institutional a r r a n g e m e n t s , in t e r m s of ownership s t r u c t u r e s , principles of taxation, a n d regulation of cutting. The Nordic and Central European c o u n t r i e s a n d some parts of t h e US pro- vide clear examples of t h i s s o r t of constraint.
Forests t h a t a r e exploited for multiple use, such a s tourism, outdoor recreation, hunting, a n d fishing, in addition to being t h e source of t i m b e r for t h e forest industry. The resulting mul- tiple objectives of forest m a n a g e m e n t normally add t o t h e con- straints on t h e supply of timber. This is a major c o n s t r a i n t in many Western European c o u n t r i e s and in some regions of t h e
us.
Deficiencies in t r a n s p o r t a t i o n capacity and high transport costs from wood supplying t o processing regions. The n o r t h e r n parts of t h e Soviet Union a n d Canada a r e examples here.
The article by LBfgren e t al.* studies t h e effect of alternative taxa- tion rules on t h e supply of roundwood. SZiiksj&rvi+ reports a successful application of a garne-theoretical approach for t h e fair division of coats
between forest industrial t h a t cooperate in wood procurement in Finland.
The estimation of forest resources is discussed in a note by Ringo*.
4.2. Forest Management Classical Theory
The problem of optimal rotation for private forestry is a classic in the forestry science literature. The basic assumptions of t h e analysis developed by Faustmann (1849). Pressler (1860), a n d Ohlin (1921) have been summarized by L6fgren and Johansson (1983):
The capital market is perfect i.e.. allowing for immediate exchange on equal t e r m s for all actors;
The forest land market is perfect;
Future lumber prices a r e known without uncertainty;
Technical lumber-yield tables are available.
It is f u r t h e r assumed t h a t the growth of timber is determined by a differentiable yield function f ( t ) where t is t h e age of the forest stand.
The t i m e of hawesting t
=
T is to be determined so as to maximize t h e present value of forest land. The present value is t h e sum of an infinite series of all f u t u r e revenues:V ( 0 . T ) = p f ( ~ ) e ~ ~ [ l + a ' ~ + ( e - ' ~ ) ~ + .
.
. ]=
(pf (T)aqT)/ ( l - e 7T)where p is t h e timber price reduced by variable production costs (such as labor), T is t h e interest r a t e , and
T
is t h e rotation period. This optimi- zation criterion is known a s t h e Faustmann formula. The determinationof the optimal rotation period requires t h e maximization of V ( 0 , T ) with respect to T . The optimality condition is the following
or expressed verbally:
The f o r e s t s t a n d s h o u l d b e h a r u e s t e d w h e n t h e m a r g i n a l r a t e of c h a n g e o j t h e v a l u e of t h e s t a n d e q u a l s t h e i n t e r e s t o n t h e s t a n d v a l u e p l u s t h e f o r e s t l a n d v a l u e .
A mathematical programming approach for determining t h e optimal rotation time is discussed by Dykstra*. A simulation approach is used by Lyons e t al.* to study optimal rotation for energy wood plantations in Ire- 1 an d.
An Example of Optimal Control Theory
The question of t h e optimal management of forest plantations is qualitatively different from t h e Faustmann problem in a t least one fun- damental respect: harvesting and planting levels are, in this context, t h e central decision variables and t h e rotation period is a secondary con- cern, determined as a by-product of optimal management. The problem can be formulated in t e r m s of optimal control theory:
subject t o
-
dz=
rrz - b z 2 + s - 2 ~ 2 d twhere
