Pulping the South:
Industrial Tree Plantations in the World Paper Economy
Ricardo Carrere and Larry Lohmann

Back to Book Index

Chapter 2
Introducing Pulp and Paper

The evolution of a wood-based industry

Paper consists of a web or mat of cellulose-based vegetable fibres which have b een refined and treated in water before being deposited on a screen and dried. First made in China around 2,000 years ago from vegetable waste materials, pape r reached India by the 600s, Turkestan by about 750, Damascus and Egypt by the 800s, Spain by the 1100s, and Northern Europe by the 1400s. The first raw mater ials included silk, hemp, cotton rags, old fish nets, mulberry bark, grasses, b amboo, flax, laurel and rattan. Early European papers were made mainly from lin en or cotton rags (Grant 1978, Western 1979, Wilson 1991).

With the advent of printing, paper use increased, but production remained limit ed by available technology. Demand, too, was constrained by the virtual absence of a paper packaging industry or a mass newspaper market. In the 19th century, however, French and English factory owners struggling to overcome the power pa per artisans held by virtue of their specialized knowledge began to develop, wi th the help of the industrial revolution's new machine tool industries, paper m achines which centralized paper-making technique in capitalist hands. Fitted wi th endless wire meshes which rotated like conveyor belts, modern paper machines began to appear around 1800 (Hills 1988, Hunter 1978, Clapperton 1967, Coleman 1958).

Prevailing non-wood raw materials were not ideally suited to the large, central ized plants such machines eventually made possible. Rags were too scarce and ex pensive to keep large mills running at capacity. Straw, which began to be widel y used in agricultural countries such as Germany and France, and esparto grass, which was imported from Spain and North Africa into Britain in large quantitie s from 1880 up until the Second World War, were available in greater volume. Th ese materials, however, had drawbacks. For one thing, they could be supplied on ly seasonally, were vulnerable to crop failures and, if they were to be supplie d in large quantities, required extensive catchment areas. They were thus far f rom perfect raw materials for big, capital-intensive mills which had to be run year-round. Agricultural raw materials were also bulky and had to be baled in t he field if they were to be transported to distant mills. It was not until the 1860s, when pulps made from wood finally began to become commercially feasible, particularly in North America and the Nordic countries, where old-growth and o ther forests provided relatively compact, contiguous, large-scale 'mines' of ra w material, that the modern Western paper industry really took off. The price o f newsprint and other types of paper, measured in conventional economic terms, dropped by about 85 per cent. Demand was spurred, and the newspaper and other p aper-dependent industries grew quickly. Today paper production represents one p er cent of the world's total economic output (Ionides 1994, Nation 22.2.1995, C haudhuri 1995, Grant 1978).

The use of wood has only reinforced reliance on large, highly-mechanized mills. Chipping equipment and stone grinders used to process logs, for one thing, pro duce too much for small mills to absorb. The more pulp-mill technology becomes geared to wood, in addition, the less inclined industry becomes to adapt to oth er raw materials, preferring to seek alternative wood sources in times of crisi s. Today, wood is used for around 90 per cent of world paper pulp production, w hich comes to over 170 million tonnes per year. Some 640 million cubic metres i s consumed annually in the process, or nearly 13 per cent of total world wood u se, the rough equivalent of the mature timber that would cover over two million hectares, an area half the size of Switzerland. Each issue of a mass circulati on daily in Britain would consume about 10,000 trees if its newsprint contained no recycled fibre; the lifetime newspaper reading of an average citizen in the US or Japan requires the pulp equivalent of several hundred trees, most of whi ch go toward advertisements. In order to maintain such levels of consumption wi thout cutting fully-grown natural forests, about ten million hectares of land w orldwide would have to be planted with trees each year (Ayres 1993, PPI 10.1994 , Shell/WWF 1993, Judt 1994, Grant 1978, Gauthier 1991, IIED 1995, Durning and Ayres 1994, Wright 1995).

The first woods used for paper were poplar and willow _ largely because they th en had few other commercial uses. Softwoods such as pine, spruce and fir, howev er, soon became the raw material of choice, since their cellulose fibres were l onger (two to five millimetres) and produced stronger, higher-quality pulps. In 1987, some 46 per cent of pulpwood consumption was derived from coniferous rou ndwood, 27 per cent from coniferous industrial residues, 22 per cent from hardw ood roundwood, and five per cent from hardwood residues (Cardellechio et al. 19 89).

Following the Second World War, Japan revived and advanced commercial technolog y for using hardwood trees, such as birch, beech, alder and mangrove species, w hich yield fibres only 0.5 to two millimetres long. This enabled the country's paper industry to exploit not only domestic broadleaved trees, but also Southea st Asian mangroves, Australian native eucalyptus, South African and US plantati on hardwoods, and Chilean beeches and the mixed hardwood forests of Papua New G uinea. Australian improvements in technology for pulping eucalyptus assisted in the process of converting many of that country's native forests to paper, and added impetus to the spread of eucalyptus plantations in Asia, Latin America an d Africa. As the industry shifts its quest for raw materials to the South, the proportion of hardwood to softwood pulps grows, with hardwood pulp now comprisi ng over 40 per cent of total pulp trade. Eucalyptus in particular is increasing ly attractive as a raw material for computer, copy, fax, high grade printing, t issue and other papers, and trade in eucalyptus pulp is increasing considerably faster than that in other varieties (JATAN 1993, Wright 1993, Marchak 1992, Sc hreuder 1988, Kroesa 1990).

The reliance on wood and on large mills which is characteristic of the industry 's mainstream today, however, is the result of historical momentum, not scienti fic or economic necessity. Even the exorbitant rates of consumption in the West and parts of East Asia do not necessarily entail a tree-based paper economy. I n some countries, non-wood raw materials such as straw, bagasse (from sugar can e), bamboo, cotton linters, sisal, seaweed, abaca, reeds, esparto and other gra sses predominate. Some 60-65 per cent of China's paper, for example, is produce d using straw, bagasse, cotton waste, and other vegetable fibres, while esparto grass provides the raw material for Tunisia's domestic pulp-making industry. M any observers also see the proportion of India's paper produced from agricultur al wastes, now between 30 and 45 per cent, rising in the future (Bayliss 1995). In 1991, 32 per cent of the pulp used in the South was derived from non-wood m aterials (Dudley, Stolton and Jeanrenaud 1995). Although the proportion of non- wood-based paper is far smaller in the North, there is no reason why this state of affairs should hold indefinitely. Indeed, evidence is now emerging that hem p would be more widely used as a raw material in the US today had not wood-fibr e businesses made an astute political alliance earlier in this century with ins titutions interested in promoting a drug panic over hemp cultivation (Hanson 19 95).

According to some observers, the ratio of non-wood to wood raw materials is inc reasing worldwide, with over 300 industrial-size mills now using non-wood fibre s (Paavilainen 1993). In the view of researcher Maureen Smith, there are no pur ely technical obstacles even to the US's exorbitant current paper demand being met entirely by a decentralized network of small- to medium-sized mills using r egionally-appropriate non-wood raw materials (Smith 1995, Smith forthcoming). Y ear by year, hemp or kenaf produce more fibre of good quality than wood, and us ing agricultural and other wastes as raw material is not only more efficient, b ut also more socially productive and environmentally beneficial than using wood (Ayres 1993; Western 1979; Wright 1994; Bayliss 1995; Riddlestone, Desai, Rice and Solly 1995; DTE [Delhi] 31.8.95).

From wood to pulp

Once the water is taken away, a tree trunk is only about 50 per cent cellulose. The rest consists of about 30 per cent lignin (a tough, resinous adhesive that provides structural support to the tree) and 20 per cent oils and other substa nces. Making a tree's cellulose available in a form which can be used to make p aper can only be done either by grinding up the wood (to make mechanical pulp) or by chipping it and boiling the chips with chemicals before refining (to make chemical pulp).

Mechanical processes turn up to 95 per cent of the wood into pulp but tear the fibres, shortening them and weakening the resulting pulp. Mechanically-pulped f ibres can thus be recycled only three to four times, as opposed to chemically-p ulped fibres, which can be used five to ten times. Mechanical processes also le ave lignin in the paper, which causes it to turn yellow when exposed to light. As a result, mechanically-produced paper is used mainly for newsprint, telephon e books and other products where strength and quality is not at a premium.

Chemical processes _ which account for more than 75 per cent of world productio n _ produce a stronger pulp because they do not damage the wood fibres. In addi tion, the most important chemical processes used separate the lignin from the c ellulose, making possible the production of papers that do not yellow with age. (Although the industry calls these 'wood-free' papers, this term does not mean that they are not made with wood. A more proper term would be 'lignin-free'.) As a result, only between 45 and 65 per cent of the wood is turned into pulp. P roducing one tonne of bleached chemical pulp from fresh raw material requires 1 20,000 or more litres of water, over 20 plantation trees or 4.8 cubic metres of wood, and approximately 1.2 megawatt-hours of electricity. Overall, this is as much energy as is required for the production of a tonne of steel. Wood residu es are today typically burned to boil chips and generate electricity and steam for the pulping process. Thus although modern chemical mills need more water, e lectricity and heat than mechanical ones, they tend to require less energy from outside sources such as thermal plants or dams (Grant 1978; Oinn 1994; Smith f orthcoming; IIED 1995; Floegel 1994; Kroesa 1990; Dudley, Stolton and Jeanrenau d 1995). In terms of energy and water use per unit of paper manufactured, curre ntly-dominant chemical and mechanical wood-based pulping processes are without question both far less efficient and far less sustainable than traditional meth ods.

There are several ways of making chemical pulp:

* The sulphate, or kraft process, which involves boiling wood chips with caustic soda, produces 95 per cent of the pulp traded on the open market. This process produces a strong pulp which, although dark brown at first, remains whi te after it is bleached due to its low lignin content. Some 95 per cent or more of the chemicals used in the kraft process are recovered and reused, but betwe en one to three kilogrammes of sulphur dioxide are released to the air for each tonne of pulp produced, with potential effects on soil, water, and the health of humans and plants.

* The sulphite process boils wood chips in an acid solution, yielding a l ight brown, strong, soft pulp. The sulphite process also reuses chemicals, but emits more, around five kilogrammes of sulphur dioxide per tonne of pulp, and t he damage caused over the past century by the water pollution associated with t his process is inestimable. As with the sulphate process, cellulose fibres lost during processing are discharged into waste water, where they decompose, deple ting the oxygen dissolved in the water.

* The chemo-thermi-mechanical process vapour-heats and chemically pre-tre ats wood chips before grinding them to remove some lignin and resin, producing a fairly strong, soft, slightly yellow pulp often used to make tissues and some writing-grade and coated papers. This process can be used for both softwood an d hardwood, and usually discharges not only the wood chemicals removed from the pulp, but also the sulphur added in the pulping process, creating a highly-tox ic, persistent effluent.

Papers produced by either mechanical or chemical processes require bleaching. Y ellow mechanical pulps are usually bleached with hydrogen peroxide, while dark brown kraft pulp requires heavier bleaching, traditionally with chlorine or chl orine dioxide, but now, increasingly _ as a result of environmentalist campaign s and consumer pressure _ with oxygen, ozone or hydrogen peroxide. Chlorine and chlorine dioxide, while they are effective in removing lignin and in strengthe ning pulp, react with organic chemicals present in pulp to form hundreds of org anochlorine pollutants including dioxins, which are some of the most potent poi sons known. (See Chapter 4.)

From pulp to paper

Most pulp is produced in integrated pulp and paper mills and goes directly into paper manufacture. Approximately 17 per cent, however, is dried and traded int ernationally to non-integrated paper mills, sometimes at a great distance. This ratio has increased only slightly since 1980, when the figure stood at 16 per cent (IIED 1995), and the trend among the largest paper manufacturers is to red uce dependence on pulp bought from outside the company (Higham 1995). In 1993, the South produced less than one-fifth of the world's total paper pulp output o f 167 million tonnes and just over one-fifth of its total paper output of 254 m illion tonnes (see Tables 2.1 and 2.2) (FAO 1995).

TABLE 2.1
World's top pulp producers, 1994

table under construction

Source: PPI 7.1995

To make paper, different types of wet pulp are blended, mixed with fillers (cal cium carbonate, kaolin, titanium dioxide, and so on) and other additives (rosin, aluminium sulphate, dyes), spread into an even sheet on a wire mesh, dried, t hen removed with an absorbent felt. The surface of printing and writing papers is then smoothed mechanically or coated with clay or chalk. Out of at least 34 different categories of pulp, over 420 commercial grades of paper can be produc ed, with individual mills often being capable of producing a variety of papers from the same forest or plantation (Fernandez Carro and Wilson 1992).

TABLE 2.2
World's top paper producers, 1994

table under construction

Source: PPI 7.1995

Newsprint (the paper used for newspapers), which typically has a weight of 40-4 9 grammes per square metre (gsm), is made mainly of mechanical (lignin-containi ng) pulp with few or no fillers added. Fine paper (printing, writing, computer and business communication papers), which tends to be thicker and heavier, is a lmost always made of lignin-free and highly-bleached chemical pulp, although so metimes chemi-thermi-mechanical pulp is also used.

Printing and writing papers range from 50 to 350 gsm in weight. Coated paper, u sed in commercial printing and glossy magazines (largely for colour advertising ), as well as illustrated books, bears a surface layer of pigment particles fin er than the fibres which make up the paper itself. Either chemical or mechanica l pulps can be used, although the former currently predominate. Stationery and photocopy papers are uncoated, weighing in at around 70-120 gsm.

Sanitary products such as toilet paper, tissues, napkins, and sanitary towels r equire that resin acids and other natural wood chemicals which prevent wood fib res from absorbing water be removed. Sulphite pulp made from softwood has been the preferred raw material for most tissue and towels because of its softness. For stronger towels, kraft pulp may be used. For toilet and other absorbent pap er, recycled fibre can be used. Fluff pulp, used together with added materials for absorbence to make disposable diapers, is made from sulphate or chemo-therm i-mechanical pulp. Brown wrapping or bag papers are made from softwood kraft pu lp, bleached or unbleached.

Card, with a weight of 160 gsm and above, and board, with a weight of 220 gsm a nd above, are generally used for packaging. Linerboard is made from unbleached kraft softwood pulp. The corrugating medium which forms the 'filling' between t wo sheets of linerboard in box-making material is made from unbleached, chemi-t hermi-mechanical pulp, usually made of hardwood, as well as recycled fibre. Boa rd may incorporate unbleached or bleached kraft pulp for strength, with many co ntainers being coated with waxes or plastics. Some corrugated packaging is made from a sulphite process which involves no bleaching.

Cellulose films and rayon are made by a modified kraft or sulphite process whic h uses intense chlorine bleaching to remove all lignin and wood ingredients. Th e pulp is further treated chemically, regenerated in sulphuric acid, and then f orced through holes to produce rayon or through slots to produce cellophane. Se lf-copying papers contain ink in small droplets of wax or solvent, while fax pa per contains a layer of heat sensitive pigments (Kroesa 1990, Paper Publication s 1994, Oinn 1994, Biermann 1993).

From paper back to pulp

For over a century waste paper has been used for making new paper wherever it h as been economic to do so. Largely due to environmentalist pressures, recycled paper is today even more important as a raw material than formerly, being used increasingly in newsprint, writing papers, and toilet and tissue papers. Spurri ng the increased use of recycled paper are both technical developments which co nstantly improve the quality of recycled products and consumer movements which point out that for most uses, paper need not be of a standard which requires a high ratio of fresh fibre.

With the globalization of the pulp and paper economy, moreover, waste paper has become an important item of trade. Some 16 per cent of world wastepaper consum ption entered international trade in 1992, with huge quantities exported within Europe and from the wood-rich, overconsuming US to wood-poor economies in Asia . When woodpulp prices are high and recycled paper in great demand, as they wer e in the early 1990s, waste paper becomes even more attractive as a raw materia l.

Recycled paper can be made from either pre-consumer waste or post-consumer wast e. Pre-consumer waste consists of unprinted industrial by-products such as prin ter's trim or paper mill waste; post-consumer waste includes already-printed pa per and used corrugated cardboard from offices, newspapers, shippers and homes.

TABLE 2.3
Waste paper use and collection, selected countries, 1992

table under construction

Source: P&P 10.1993 (cuoted in Ryan 1994), IIED 1995, FAO 1994

Although the amount of recycled paper used varies widely from country to countr y, waste paper accounted for approximately 18-20 per cent of the material input for the world's paper output in 1970. With environmental pressures, this figur e had risen to around 30-32 per cent by 1988 and perhaps 35-37 per cent by 1995 . Top consuming countries such as Taiwan, Japan, Germany and The Netherlands ar e major users, but Southern countries for whom waste is easier to obtain than w ood generally use a greater proportion of it in their raw material supply. Abou t half of the raw material for Asian-made paper consists of waste, and nearly 4 5 per cent of the raw material for Latin American-made paper, but only about 28 per cent for North American paper and 37 per cent for European. Overall, the S outh uses a raw material which consists of approximately 10-15 per cent more wa ste paper than does the North (IIED 1995). Many Northern countries, on the othe r hand, are more assiduous waste paper collectors. However, they often wind up dumping waste on the international market rather than using it themselves (see Table 2.3).

Industry sources suggest that the global use rate for waste paper may reach 42- 45 per cent by the year 2000, while FAO suggests it is more likely to stagnate (Niku 1993, IIED 1995). At present, the use of recovered fibres in paper-making is growing twice as fast as paper production itself. This, of course, cuts int o the demand for wood pulp. A mere one per cent increase in the use of waste fi bres in paper raw material in Finland alone, for instance, would save 376,000 c ubic metres of wood per year (FAO 1994). Roger Olsson of Taiga Rescue Network e stimates that a world recycling rate of 50 per cent in the year 2010 would save 200-300 million cubic metres of roundwood annually at projected consumption ra tes. This would obviate the construction of around two giant woodpulp mills per year between 1990 and 2010 (Olsson 1995). If the US had adopted the over 50 pe r cent recycling rate of the Netherlands in 1987, 500,000 fewer hectares of for ests would have had to have been logged per year for conversion into paper pulp (Graham 1994). Also reducing the rate of rise in demand for pulpwood, although to a lesser extent, is the increasing use of mineral fillers and coatings in p aper. The proportion of pulp in paper-making furnish is estimated to have dropp ed from 65 to 64 per cent between 1993 and 1994 alone (Ayres 1993, Clark 1994, Shell/WWF 1993, McClelland 1994, Niku 1993, Brennan and Pappens 1995).

Newspapers can be de-inked easily, making the reuse of newsprint attractive. Ca rdboard, too, can easily be recycled, and printing and other pre-consumer waste , office bond, and discarded photocopies can be converted into new stationery a nd copying paper if kept separate. Mixtures of all grades can be reycled into l ow quality products such as egg cartons. De-inking photocopies and laser-printe d material, however, requires newer technologies. In addition, highly coated pa per, envelopes with windows and paper containing synthetic glue (self-adhesive envelopes) are hard to recycle; and fax paper, carbonless copy paper and plasti cized drink cartons cannot be recycled at present. It is difficult, moreover, t o make extremely white paper out of recycled fibres without using harsh bleaches.

Like all 'technical fixes', recycling cannot in itself be equated with wood con servation, but must be viewed in political and economic context, alongside an a nalysis of demand, trade and industry structure. Even if the world recycling ra te increases to 50 per cent, for example, a 12-15 per cent increase in the worl d's industrial roundwood production will be required by 2010 if FAO projections of consumption increases are borne out. In the US, the increased use of recycl ed pulps in paper production has not even slowed the rate of growth in fresh wo odpulp production, instead merely increasing woodpulp exports (Harland 1994, Ol sson 1995, Smith forthcoming). Unless combined with institutional restructuring , moreover, more use of waste paper, instead of reducing dependence on industri al tree plantations, may merely spur the wood industry into attempting to creat e demand for alternative plantation products. For example, the FAO European For estry Commission and the UN-ECE Timber Committee concluded in 1993 that because of the weakening demand for small-sized wood which has resulted from increased recycling, 'new outlets for small-sized wood should be developed' in energy pr oduction or other fields (FAO 1994). The Brazilian firm Aracruz is meanwhile ho ping to create new markets for its plantation eucalyptus in construction, furni ture, fibreboard and plywood in order to reduce commercial vulnerability to pul p price swings (FT 21.6.1995).

Moreover, although making paper from recycled fibre, as from vegetable fibres, tends to use less water and energy than producing it from wood, and to result i n less pollution, the expenditure of water and power can still be high (Smith f orthcoming; Dudley, Stolton and Jeanrenaud 1995). In addition, both the proport ion and the grades of paper that can be made from recycled material are limited . Fibres become shorter and weaker in the recycling process, making them less u seful for papers requiring strength. Even the strongest, longest fibres turn in to useless dust if they are recycled more than ten times, and most wood fibres have a far shorter life span. Pulp made from waste paper must thus often be top ped up with longer-fibred pulp made directly from trees, hemp, kenaf or other m aterials in order to ensure strength. Although it is theoretically possible to restrict the overall ratio of fresh to recycled fibres to little more than 20 p er cent (Dudley, Stolton and Jeanrenaud 1995), a totally cyclical paper economy is not on the horizon. Mechanically-pulped recycled fibres, moreover, cannot b e used to make certain quality grades of recycled paper which require chemicall y-pulped fibres.

Recycling requires, moreover, that ink, fillers, coating materials and staples be removed. The recovered ink, which is likely to contain barium, copper and he avy metals, tends to be discarded together with unrecoverable paper fibres, whi ch may contain poisonous dioxins and furans. It is either incinerated, adding t hese pollutants to the air, landfilled, or spread on farms and gardens (Durning and Ayres 1994). The dangerous effects of such discarded compounds underline t he importance of pressing for use of nontoxic inks as part of campaigns to chec k overconsumption, reconsider waste disposal policies and use a greater proport ion of recycled fibre in paper manufacture.

Large scale, capital intensity and centralization

As a growing pulp and paper industry using local wastes transformed itself, dur ing the 19th century, into a highly-mechanized, centralized one requiring the d estruction of large swathes of forest, literacy and the demand for cheap paper grew. Many of the new paper mills built around the turn of the 20th century wer e designed to produce newsprint. These were of standard size and relatively ine xpensive and profitable. By the 1930s, however, prestige competition among Nort h Atlantic newspaper corporations, conjoined with technological advances follow ing on from the First World War, had given an incentive to machine manufacturer s to design larger and larger machines. Many of these turned out to be one-offs . Whereas in the 1900s, new newsprint machines tended to be 2.25 to 2.5 metres wide, running at speeds from 100-150 metres per minute, by 1937 machines were b eing built up to 7.7 metres wide, running at 420 metres per minute. The kraft s ack and wrapping paper industry, which grew swiftly from 1930 onwards, soon bec ame another source of orders for big machines.

Such machines became less and less cost-effective. Not only were many machines of a unique design; huge widths and speeds also required sophisticated and expe nsive controls for efficient operation, adding yet more to costs. At the same t ime, paper prices could be increased only so much, since they were set by the m ass of earlier, smaller, less sophisticated, cheaper machines still in operatio n. By the 1960s, moreover, when widths of nine metres and speeds of 700 metres a minute were achieved, capacity began to exceed consumption by a wide margin a nd paper prices slumped. The cost per annual tonne of a newsprint machine incre ased at least 40-fold between 1930 and 1975, while the price of newsprint incre ased less than 20-fold. Yet by this time, machine manufacturers' investments in large machine tools had made it difficult for them to produce for anyone but t he biggest paper companies. As paper expert A. W. Western (1979) notes, buildin g new paper machines became a luxury which could be afforded only by multinational giants or the gov ernments of developing countries, advised by consultants that only scale to thi s degree could be economic! For the consultants it was economic; they were now essential for large mill design and coordination.

Today a single new world-class pulp mill can cost as much as US$1 billion. In m any countries only large manufacturers can afford even to renew their plants, a nd then only to adapt them to producing more specialized products with a greate r profit margin.

The nearly 200-year-old dream of concentrating paper-making power in the hands of plant owners, in short, had been realized with a vengeance. Access to the do minant stream of paper-making knowledge was now restricted not just to capital, but to big capital. For any capital-short Southern society with an interest in meeting its own paper needs efficiently with indigenous materials, and in a wa y which did not require centralized control of large areas of land, the implica tions were particularly bleak.

They were also bleak, however, for the industrialized North. Partly because tod ay's immense mills cannot generate profits without a large-scale re-engineering of their social and physical surroundings, the pulp and paper industry relies heavily nearly everywhere on political campaigns to capture handouts from the s tate and the public. As the giant Canadian firm MacMillan Bloedel instructed Al berta's Premier Harry Strom in 1969, when the company was seeking low-interest loans, tax exemptions, tax investment credit, and infrastructure subsidies from the province, 'pulp mills being built today are not profitable unless some spe cial low cost conditions or concessions prevail' (cited in Pratt and Urquhart 1 994).

Such subsidization, by making it possible for paper firms to make a profit, can in turn often motivate them to invest in yet more and bigger machines and seek even bigger subsidies, with disastrous results for raw materials catchment are as. As historian Ramachandra Guha and ecologist Madhav Gadgil note, beginning i n the 1950s, India's forest industries were subsidized so heavily, and could hike up the prices of their produce so fr eely in a seller's market, that their profitability has remained high, even as forest stocks have plummeted. . . . Even in the 1980s, bamboo prices were raise d only to Rs200 to 500 per tonne, when market prices were well over Rs5,000 per tonne. The result of this state-subsidized profitability has been an explosive growth in industrial capacity, and a non-sustainable use of forest stocks (Gad gil and Guha 1992).

The networks through which a highly-mechanized and -centralized industry captur es the subsidies it needs for survival will be explored in Chapter 5 and in the case studies of Part Two.

While ownership of today's mainstream pulp and paper sector is not as concentra ted as that of some other basic industries (the top ten paper businesses contro l only about a fifth of the international paper market, and no single market pu lp producer has more than a six per cent market share), it is dominated physica lly by relatively few large plants. The US, the world's largest producer, has o nly 203 pulp mills, implying an average mill capacity of over 300,000 tonnes pe r year. In Japan 49 pulp mills are capable of producing an annual average of ne arly 310,000 tonnes each. Finland, meanwhile, holds its position as a leading p ulp and paper exporter with a mere 43 pulp mills averaging over 250,000 tonnes per year in capacity. The relatively young export pulp industry in the South al so revolves around a small number of huge plants. In Chile six pulp mills boast an average capacity of 350,000 tonnes a year each, in Brazil 35 are able to ch urn out an average of 175,000 tonnes each and Indonesia's 13 mills have an aver age annual capacity of 215,000 tonnes each. China's huge pulp production, by co ntrast, is spread over about 8,000 small mills whose average output is well und er 2,000 tonnes per year. Some mills, like village bakeries, may be open only t wo days a week. Not coincidentally, such mills tend to rely on local, non-wood sources of raw material. In India, too, average pulp mill capacity amounts to a low 13,500 tonnes per year (PPI 7.1995, Wright 1994).

Unsurprisingly, the large-scale mainstream industry generates very little emplo yment per unit of economic output. In the US, for example, paper and pulp is th e most capital-intensive of all manufacturing industries, and twice the industr ial average, with more than US$120,000 of plant and equipment invested in every employee. The world-class mills being built in Indonesia, Brazil and elsewhere , which use the same type of equipment, are hardly less so. Approximately $750, 000 in capital, for example, is being invested for every job created in the new pulp mill at Riau Andalan in Indonesia (see Chapter 11). Similarly, $700,000 i s being invested for each job in two new cellulose, paper and thermal energy co mplexes in Galicia, Spain being backed by the German multinational Feldmuhle Ak tiengesellschaft and the Finnish firm Tampella Oy. The cost of each of the 365 jobs created at the enormous new Al-Pac bleached kraft pulp mill near Athabasca , Alberta owned by a Japanese-Canadian consortium is meanwhile a staggering $1. 3 million. Employment, already low in the mainstream industry, is moreover set to decline even further. Employment in the US's paper industry, for instance we nt down one per cent between 1993 and 1994 alone, due mainly to mergers and the phasing out of older technology. Between 1990 and 1992, Canada's forest indust ry eliminated 62,600 jobs, some 28 per cent of the direct workforce. Knowledge and skills are another casualty of the industry's capital-intensivity. The domi nance of gigantic machines ensures that the opportunity to learn about and use paper-making technology is restricted to a select few technicians and technocra ts (Van Hook 1994, Western 1979, McClelland 1994, CEPA 1992, Olsson 1995, TAPPI Journal 1.95, Smith forthcoming, Schindler 1995).

Boom and bust

Reliance on big, expensive, centralizing machines _ combined with freely-availa ble technology, easy availability of debt finance and wood, and little need for newcomers to buy into brand names _ tempts the industry, whenever demand incre ases, into building huge amounts of new plant to supply it. Big firms, in addit ion, have sometimes hoped that being the first to build a giant new machine dur ing boom times will scare off competitors and gain them enough market share to become price-setters. In what is still a crowded field, however, such hopes hav e proved vain. As a result of overspending, the market becomes glutted with pul p and paper a couple of years after the market peaks _ it takes from 18 months to two years to bring a new pulp mill on line _ sending prices into a deep slum p. Left with enormous machines on its hands which cannot be run at full capacit y, the industry finds it difficult to pay off its debts and ceases to make a pr ofit (Wright 1993, 1994; P&PA 11.1993; PPI 9.1994).

Such cycles afflict pulp and paper even more severely than they do other basic industries such as chemicals or metals. They are possibly exacerbated by the be haviour of banks, which are the biggest suppliers of finance to forestry indust ries, and from which cheap funding is especially easily available at the peak o f the paper and pulp cycle. Adding further to the industry's volatility is the growing tendency of firms to invest across international borders, which often g ives exchange rate fluctuations a huge importance in determining a company's pr ofitability (Fletcher 1988, van Dijk and Dekker 1995, Clark 1994). Government p olicies also do their bit. Changes in US tax laws in the 1980s, for instance, g ave corporations tax breaks for debt instead of equity, giving firms extra ince ntive to take on debt by building new mills (Floegel 1994). All in all, paper a nd pulp's wild market swings appear difficult to control within the current sys tem.

During the boom years of the 1980s, for example, when demand was surging and pu lp and paper were selling at high prices, leading paper industrialist Hugh Flet cher of Fletcher Challenge warned his colleagues that although 'the industry do es not have a history of being logical', this time it really should refrain fro m investing too much in large new plants. The advice went unheeded. By 1993, pu lp prices, in constant dollars, were half of what they had been only four years previously, and 39 per cent of what they had been in 1975. By 1994, Ronald Y. Oberlander, President and Chief Executive Officer of Abitibi-Price, which had l ost over US$460 million in the period 1990-4, was lamenting with industry colle agues over the sector's 'inability to manage the periods of prosperity and our inability during these periods to spend money wisely'. Canada's pulp industry a s a whole had lost more than $4 billion between 1991 and 1993. In December 1993 , Avenor's new Gold River newsprint complex in Vancouver was forced to cease pr oduction, joining other plants such as Abitibi-Price's newsprint mill in Thunde r Bay, Ontario. In Japan, excessive supply and world economic recession brought the market to a standstill, forcing companies such as Daishowa Paper into a se rious management and debt crisis (Fletcher 1988, Oberlander 1994, Soulas 1994, McClelland 1994, Avenor 1993). Chile's industry meanwhile lost up to US$200 mil lion in income per year; Latin American earnings suffered more than those of an y other industry region in the early 1990s. At the same time, Indonesian pulp p roduction dropped to a mere 65 per cent of capacity, and producers such as Thai land's Phoenix Pulp and Paper had to stop exporting.

Yet between mid-1993 and late 1995, Northern softwood bleached kraft pulp incre ased in price from US$390 per tonne to nearly $1,000. This allowed pulp mills t o operate nearer capacity, but threatened the survival of some firms in the ove rcapacity-ridden paper industry, who were hard put to pass on such large price increases to consumers.

Among the reasons the industry cited for the spectacular reversal were economic recovery in the US and Europe; increased paper demand in Europe, Southeast Asi a, and the US; the closure of some integrated pulp lines and an associated run on market pulp; a sudden hoarding response to pulp producers' announcements of price hikes in late 1993 among buyers who had confidently allowed their stocks to be depleted; a European hardwood shortage and a rise in pulp production cost s there and in western Canada; weak US and Canadian dollars, which made North A merican pulp and paper exports cheaper, stimulating demand; rumours of a strike at pulp and paper mills in British Columbia; and the extreme depth of the prev ious slump, combined with new environmental requirements, both of which discour aged new investment which could have increased supply. Also significant was a t radition of noncooperation between pulp buyers and sellers: pulp dealers who fe el they are being pushed into offering ever lower prices in a slump tend to get even by squeezing their customers as hard as possible when the market begins t o improve. The same holds for paper dealers (Bingham 1995; Independent 19.6.95; Pappens 1995; Edwards 1995; Stefan 1995; Rahikainen et al. 1995).

By 1996, though, the market was once again beginning to dive, with Northern sof twood bleached kraft pulp dropping about 25 per cent to US$725 per tonne in Feb ruary. Frightened by the prospect of price increases, buyers had built up stock s which they began to draw on as the economy worsened, reducing orders. At the same time, as one trader complained, hardwood pulp suppliers from Brazil, Indon esia and Russia 'have lowered prices, but they have not created demand', with C hinese purchases being especially disappointing. With capacity increases, furth er drops were projected for succeeding years (FT 23.11.95, 5.1.96, 24.1.96, 8.2 .96). Some industry observers gave vent to a hope that, due to constraints on s pending forced by environmental expenditure, the next slump would be less sever e than the one before (Bingham 1995; FT 13.7.95; Economist 14.1.95). In the vie w of executive David Clark, however, 'the paper cycle shows no sign of diminish ing. Indeed, the swings show every sign of increasing as the business becomes m ore global and more capital intensive with more aggressive competition among la rger groups' (Clark 1996).

Such slumps, of course, can often bring advantages to large corporations with t he resources to weather them. Firms such as Indah Kiat, for example, may snap u p bargain-priced machinery or plants. But the natural environment is not necess arily a beneficiary. In 1993, the integrated forestry-paper firm International Paper had to balance its paper operations' losses by increasing timbercutting t o a rate which the company itself admitted was 'unsustainable'. The Canadian gi ant Avenor, meanwhile, received state subsidies to open an idled mill (Penna 19 94; FT 9.2.95, 17.2.95; TN 5.1994; Olsson 1995).

Exacerbating the industry's chronic tendency toward overinvestment and periodic overcapacity is the zeal of many national bureaucracies _ and some of their co nsultant advisers _ in promoting pulpwood or pulp as a foreign-exchange earner. Particularly in Southern countries, a pulp export industry is often advertised as being capable of dynamizing the economy and creating rural and industrial e mployment. In some countries, it is also promoted as a means of centralizing co ntrol over land, taking over small farmers' holdings, or even as an agent of 'r eforestation'. In Thailand, government targets for pulpwood plantation acreage regularly overshoot even the most optimistic market demand projections by 1,000 per cent, and the ambition of some Indonesian government officials to turn the country into the 'world's biggest pulp producer' have contributed to an even m ore overheated atmosphere there. One Western machinery supplier has claimed tha t many prospective Indonesian investors 'don't know what they plan to do with t heir pulp', joining experts who worried in the early 1990s that the regional As ian market was 'undeveloped', uncertain, and oversupplied (Sargent 1990; Paper 4.2.92; Dench 1993; ADB 1993; WALHI and YLBHI 1992).

Concentration and liberalization

The cycles endemic to the pulp and paper industry tend to lead to renewed effor ts to increase capital intensity and concentrate production in fewer hands. Whe n slumps cause Northern manufacturers to lose market share of basic paper grade s to Southern producers, Northern firms are increasingly forced to seek out and seize miniscule competitive advantages which might enable them to get rid of t heir overproduction. Jobs are shed, still bigger machines sought, and attempts made to diversify into new, higher value-added products or concentrate on fewer niches.

As corporations struggle to cut costs, hold market share, prevent overinvestmen t, find and control markets, create demand, and cope with globalization, they a re often pushed into new mergers and cartels. Rapid upturns, when they come, pr ovide acquisition-hungry companies with flushes of cash they can use to buy out vulnerable rivals. Economy measures meanwhile induce many companies to link th emselves more tightly to timber producers, since large-capacity, wood-based mil ls rapidly create shortages of timber unless combined with sawmill operations o r plantations (JPA 1994, PPI 7.1994, 7.1995, Rajesh 1995). Links are also incre asingly sought with paper-product manufacturers and paper customers.

Thus in the last decade the top 30 paper companies have increased their proport ion of world production dramatically (Higham 1995). In 1995 alone, the US's Kim berly-Clark took control of Scott Paper, Sweden's SCA bought Germany's PWA, the US's International Paper gained control of New Zealand's Carter Holt Harvey an d initiated moves to acquire Federal Paper Board, Canada's Canfor launched a bi d for Slocan, and Jefferson Smurfit took over Saint-Gobain Paper-Wood. In Finla nd, in addition, mergers were announced both between Enso-Gutzeit and Veitsuluo to (to create ENSO) and between United Paper Mills and Kymmene (to create UPM-K ymmene); the latter marriage will result in Europe's largest pulp and paper com pany. In Japan, meanwhile, the share of national pulp production of the top fiv e firms rose from 40 per cent in 1960 to 60 per cent in 1990, and in a further phase of what Alastair Graham calls 'mutual cannibalism' in 1993, Jujo combined with Sanyo Kokusaku Pulp to create Nippon Paper, while Oji merged with Kanzaki , Honshu and Chuetsu to form New Oji. In 1994, the two new firms took over the No. 2 and No. 3 positions in world pulp and paper sales (Whitham 1995; PPI 7.19 95). Pira International, a pulp and paper research centre, predicts that the sh are of world paper and board tonnage held by the largest companies could be wel l over 80 per cent by 2005, up from 65 per cent today (FT 13.7.95).

In addition to helping companies concentrate production in the most 'efficient' mills, mergers are likely to improve corporate ability to streamline organizat ions, sell non-strategic assets, control prices and wages and generally minimiz e risk at the expense of smaller economic actors. Local communities, however, m ay find that protecting their land, water and air against the encroachment of o ne of the new conglomerates is even more difficult than was holding the line ag ainst their predecessors.

Industry concentration both feeds on and creates pressures for centralized inte rnational structures of economic control such as the European Union, the North American Free Trade Agreement (NAFTA), and the General Agreement on Tariffs and Trade (GATT). Large-scale producers expect NAFTA and GATT to open trade routes and foreign investment opportunities for them, while the European Union is hel ping the biggest pulp and paper firms there to gain access to wood, markets and production sites across the region, transforming it into a single economic uni t with greater global power. Tariff reductions associated with a global market, meanwhile, tend to force all manufacturers into competition with their counter parts in the most cheaply-producing countries, causing the consequences of over investment in one region to reverberate worldwide.

Economic liberalization is also helping the dominant pattern of highly capital- intensive, wood-based or industrial-plantation-dependent production to spread i nto previous strongholds of small-scale, agriculture- or farm forestry-based in dustry such as India and China. In India, for example, as lower import duties o n paper begin to threaten the highly-protected, flexible, decentralized and lab our-intensive local paper industry, Indonesia's Sinar Mas group has announced p lans to set up a new mill twice as large as any other in the country, fed by su rplus hardwood kraft pulp from its operations in Sumatra. The hope is that the mill's output will help create an Indian demand for a 'high-quality' wood-based sheet which will spur the development of other such mills, which of course wil l in turn require cheap new timber sources in natural forests or plantations (B ayliss 1995; cf. DTE [Delhi] 31.8.95).

Economic risks in South and North

The highly cyclic nature of the pulp and paper industry creates risks for inves tors everywhere. But because the costs of installing world-class pulp or paper machinery are as high in the South as in the North, Southern countries who are pushed into establishing such industries risk a greater proportion of their res ources than do their Northern counterparts.

These hazards are exacerbated by a raft of additional uncertainties. Due to the scale of today's world-class mills, unanticipated increases in pulp capacity d ue to the whims of single manufacturers in Brazil, Russia, Indonesia or anywher e else have great potential for upsetting the market. While the per capita pape r consumption of some of Asia's newly-industrializing countries has exploded in recent years, in addition, it may suddenly stagnate if an economic recession h its. And while large countries with low per capita consumption such as Indonesi a and China may increase world demand greatly if incomes rise, the timing of su ch changes is uncertain. It is conceivable, too, that if enough Northern produc ers are driven out of business by new Southern competitors, the price umbrella they provide may disappear (Bazett 1993, PPI 7.1994).

Environmental activism adds another unpredictable factor to the equation. Fores t and land conservation movements affect demand for and supplies and prices of pulpwood, while recycling movements and legislation can unexpectedly increase g lobal supplies of waste paper. Environmental legislation regulating manufacturi ng processes also has a powerful affect on industry spending. The extent of the need to invest in machines capable of meeting consumer demand for chlorine-fre e paper is another difficult-to-assess factor, and one which is especially rele vant to the prospects for Southern plantations, since chlorine-free papers requ ire a high proportion of hardwood fibres (Bazett 1993, ADB 1993).

The increasing reliance on plantations which has followed on from depletion of old growth forests creates additional uncertainties for pulp and paper producer s in both South and North. Even fast-growing plantations require especially lon g planning horizons of 10-15 years, due to the time it takes trees to mature, d uring which time any number of things may change. Large monoculture plantations are prone to pests, diseases and fire, and productivity may well dwindle after the first rotations (Good, Lawson and Stevens 1993). Moreover, it is notorious ly difficult to estimate how much competition pulpwood plantations will have fr om natural forests in the Pacific region in the future. Depending on what proba bilities are given for the future economic availability of pulpwood from remote conifer forests in Siberia and Yakutsk, and for the demand from the Chinese ma rket, projections have varied from a surplus of 20 million tonnes a year to a d eficit of 60 million tonnes. According to former FAO forest economist Alf Lesli e, such 'inevitable uncertainties dwarf any market growth analyses based on sta ble market factors like population and income growth.' As the next chapter indi cates, uncertainties about raw material availability have proved a particularly powerful influence on the Japanese industry, propelling it into ventures all a round the Pacific basin (Graham 1994, PPI 1.1994).

Consumption and demand creation

Since the Industrial Revolution, owners and managers of Western pulp and paper industries have struggled to reorganize society in ways friendlier to themselve s through large-scale mechanization. As this chapter has suggested, the style o f mechanization they have followed is closely tied up with economic cyclicity, concentration of skills and of wealth, the spreading of risk, globalization, an d deforestation. As will be explored further in later chapters, it is also asso ciated with a pattern of heavy pollution, soil degradation, widespread disposse ssion, and massive subsidization by the public sector. Despite the crises growi ng out of this highly-centralized type of social organization, the pulp and pap er industry as yet shows few signs of moving on to a more decentralized, smalle r-scale system of specialized production.

One of the most important effects of the strategy of large-scale mechanization and mass production followed by today's mainstream pulp and paper industry is o n how paper is used by ordinary people. Industrialists who have to sink huge su ms into complex production technology have a strong incentive to try to predict and control prices. As economist John Kenneth Galbraith (1972) has long pointe d out, that entails attempts at managing demand. Such attempts are all the more imperative in an industry subject to the influence of savage boom-and-bust cyc les, rapid cultural and technological changes, and unpredictable environmental pressures on production and consumption. Thus top European paper executive Davi d Clark (1994) recently told his colleagues, we shall have to fight for our future and create our own growth. . . total de mand has to be stimulated. The alternative, to do nothing, could produce a stat ic or even declining demand with serious implications for the industry, its rep utation, its technology and the quality of the people it attracts.

It is hardly surprising, therefore, that today's mainstream paper industry is c onstantly promoting new uses for paper in the overconsuming markets of the Nort h as well as seeking new outlets for its products in growing markets such as th at of East Asia. This dynamic inevitably sets a highly-centralized, mechanized industry at odds with environmentalists and others concerned about overconsumpt ion.

Stimulation of paper demand is, of course, nothing new, and is not something th e industry has to undertake alone. Since the 1800s, new processes and commoditi es _ ranging from lithography, paper shirt collars, slippers, cups, building ma terials, bags, toilet paper, drinks cartons, nappies, supermarket packaging, ex port packaging, fax paper and word-processing computer technology _ have been e mbedding paper use ever more thoroughly into business and household activities. But the scale and intensity of demand creation has recently reached new height s. To take one example, TetraPak, partly financed by the World Bank, recently o pened a disposable carton factory in Hungary that caused the country's deposit bottle system to collapse in a matter of weeks (Fairlie 1992). New technology s uch as fax machines and laser printers, as well as the growing roles of service and administration in Northern economies and the burgeoning supermarket system , with its vast needs for packaging, also stimulate paper consumption.

While industry sometimes claims when speaking with uninformed audiences that th e driving forces behind increasing paper demand are growing literacy, surging p opulation, and the need for schoolbooks and education, the figures tell a diffe rent story. Over 40 per cent of paper production in 1991 was used for packaging and wrapping, 13 per cent for newsprint, and less than 30 per cent for printin g and writing (IIED 1995, cf. Bazett 1993). In the US, according to Worldwatch (1994), it is the 'growth of advertising and the automation of office equipment ' _ including, paradoxically, computers, which were supposed to have introduced the 'paperless office' _ which have driven growth in paper consumption:

Sixty per cent of space in American magazines and newspapers is reserved for ad s, while some 52 billion assorted advertising pieces _ including 14 billion mai l order catalogues that often go straight into the trash _ clog up the US post office every year.

Explosive increases in cardboard and paper packaging in the last few decades ha ve also been largely fuelled by the desire to advertise the products within, an d even increases in demand for newsprint are primarily due to advertising (Baze tt 1993). Overall, paper consumption is growing fastest in coated lignin-free a nd coated mechanical varieties, which are used for glossy magazines and colour advertising. Yet per capita newsprint use also continues to surge _ from 24.5 t o 32 kilogrammes between 1980 and 1991 in Britain alone (G 25.11.1995). In Sout hern countries, meanwhile, increases in paper consumption are closely tied to d emand for packaging by industry (often export industry), advertising and the co mputerization of businesses, and the rise of consumerism. In Thailand, for exam ple, one of the world's fastest-growing export economies, packaging absorbs mor e than two-thirds of production. As trade becomes more and more globalized, it is likely to entrench packaging demand even further (Kroesa 1990, Soulas 1994, P&PA 11.1994).

TABLE 2.4
Paper consumption, selected countries, 1994

table under construction

Source: PPI 7.1995

Tying demand for paper to a broad range of economic activities outside educatio nal publishing has helped free world per capita paper consumption to expand ind efinitely. Rising from .01 kilogrammes yearly in 1910 to 15 kilogrammes in 1950 and nearly 48 kilogrammes in 1994, it shows no signs of levelling off (Worldwa tch 1994, PPI 7.1995). The industry's 'efficiency' can no longer be plausibly d escribed as, say, 'efficiency in producing the medium for the books and newspap ers which society needs', but is increasingly merely an abstract ability to pro duce as much paper as possible as cheaply as possible. Unsurprisingly, per capi ta paper consumption has very little relation to literacy levels (see Table 2.4 ).

According to industry figures, in 1993 the South and Eastern Europe, with nearl y 84 per cent of the world's people, consumed less than a quarter of the paper and board produced worldwide, while the North, plus the Asian 'tigers', with ju st over 16 per cent of the world's population, accounted for over three-quarter s. Worldwide, wood fibre consumption is skewed even more than consumption of ot her commodities. US citizens, for example, consume 386 times as much pulpwood p er capita as Indians, while 'only' 43 times as much oil (G 6.9.94). Almost half of the world's paper production is consumed by 460 million people in the US, J apan and Germany, with the US alone consuming more than Japan, China, Germany a nd the UK (the next four top consumers) combined (see Table 2.5) (TRN 1993; IIE D 1995; Ozinga 1994; Soltani and Whitney 1995; Dudley, Stolton and Jeanrenaud 1 995).

TABLE 2.5
World's top consumers of paper, 1994

table under construction

Source: PPI 7.1995

Significantly, the single-year increase in per capita consumption in Sweden and the US between 1993 and 1994 was five times the total current per capita consu mption of India and double that of Indonesia (PPI 7.1995). While the most rapi d recent growth in consumption has been in South Korea, China, Indonesia, Malay sia, Thailand, Taiwan, Singapore and Chile, it is the North which will continue to dominate world consumption figures. The US Forest Service suggests, in what unfortunately may well be a conservative estimate, that annual per capita cons umption in the US will reach 472 kilogrammes by 2040 (Smith forthcoming).

Overall, paper consumption has climbed from less than 15 million tonnes in 1910 to over 268 million tonnes in 1994, with the paper consumption of Japan alone doubling in two decades. According to industry figures, consumption growth from 1988 to 1996 will equal total consumption globally 30 years ago (Ionides 1994) . Although rates of paper consumption can change, and have changed, remarkably quickly, the UN's Food and Agriculture Organization strives to project such tre nds as global destiny, predicting that paper consumption will increase by 80 pe r cent by the year 2010, with more than 60 per cent of this increase to take pl ace in Europe, North America and Japan. Roger Olsson of Taiga Rescue Network ca lculates that even if recycling rates increase from the current 35 per cent or so to 50 per cent by 2010, world pulpwood consumption will still increase by 33 -41 per cent over 1990 levels (Olsson 1995, BP 22.2.95).

Interestingly, while per capita demand for paper continues to rise, per capita demand for sawnwood is lagging, with less construction timber and fewer pitprop s needed. Between 1955 and 1985 the volume of sawnwood consumed per unit of Gro ss National Product halved, while paper consumption followed GNP closely. Accor dingly, the composition of demand for industrial wood has changed radically dur ing the 20th century. Between the 1940s and the 1980s, the ratio of pulpwood to sawlogs increased from about 1:4 to between 1:2 and 1:3. Although FAO and the World Bank anticipate annual increases in paper demand to 2000 on the order of 2.7-2.9 per cent, they expect sawnwood demand to increase only 1.0-1.5 per cent . According to Michael Bazett, one industry analyst, growth in production of sa wlogs is now close to zero. Increases in wood consumption, therefore, are due m ore and more to paper consumption (Mather 1990, Gauthier 1991, Bazett 1993).