Chapter 13: Timber Products Supply and Demand 297 · timber products, six steps were taken: Key Findings The South produces approximately 60 percent of the Nation’s timber products,

297TIMBER

Chapter 13: Timber Products Supply and Demand

Photo courtesy Tim White, University of Florida

299TIMBER


to improve timber growing and wood-processing productivity, and to heavilyinvest in timber growing technologyand intensive forest management.

■ Private landowners in the Southare projected to continuously expandareas of pine plantations in the regionfar into the future. An outcome ofthis is a projected increase in thearea of pine plantations—in thebase scenario, by 67 percent (from33 to 54 million acres) between1995 and 2040.

Introduction

This chapter describes historical,current, and projected timber inven-tories and timber product outputs fromsouthern forests. It also attempts toplace these quantities in national andinternational perspectives. Timberis the most valuable commercialcommodity taken from most forests,and its removal strongly influencesthe character of those forests. Timberis removed to convert land to otheruses, and it is removed in regularharvest activities of managed forests.These two processes do not occurrandomly on the landscape. Rather,they occur in patterns that arepredictable, related to the locationsof development, timber processingcapacities, and the species in demandfor timber products. Because removalsare a function of societal demands,the products made from timber, andthe technologies used to remove andprocess timber, the nature of forestsand projected future of those forestscan be traced out by relating economicand demographic trends to the timberproducts sector. The economic and

demographic relationships to thetimber sector can be identified througha description of historical patterns oftimber production and technologies.Hence, such a description providessubstantial information for predictingthe future of southern forests.

In describing the history andprojected future of southern forestsand their associated timber markets,technical terminology is often used.For clarity, it is worth defining somefrequently used terms. Demand is theschedule of quantities that would bepurchased by consumers over a rangeof prices. Supply is the schedule ofquantities that would be producedin a geographic region by productmanufacturers over a range of prices.Production is defined as the amountthat is actually produced in a geo-graphic region, and consumptionis how much is actually purchasedby consumers in a geographic region.If a country or a State consumes morethan it produces of a given product,then it is a net importer of that product.If it produces more than it consumes,then it is a net exporter. The incentivefor a country or State to producea different quantity from what itconsumes arises out of the ability ofbuyers and sellers to move productsback and forth across national bordersand State lines profitably. We use thedefinition of forest land and timberlandas adopted by the USDA Forest Servicein its Resources Planning Act (RPA)documents and its own projections.See chapter 16 for a thorough definitionof each.

To address questions of historicaland future supplies and demands fortimber products, six steps were taken:

Key Findings

■ The South produces approximately60 percent of the Nation’s timberproducts, almost all of it from privateforests; the South produces moretimber than any other single countryin the World, and it is projected toremain the dominant producingregion for many decades to come.

■ Timberland area is projected toincrease in many parts of the South,especially in western and northernportions, due to agricultural landconversion to forest and to treeplanting. Timberland will be lost,especially to urban and residentialland uses and especially in thePiedmont region (Virginia to Georgia)and in Florida. The net effect of lossesand gains is no significant change intimberland area under two plausiblescenarios. However, in aggregate inthe South, nonplantation timberlandacreage is projected to decline byan average of 15 percent under allmarket and plantation growthscenarios considered.

■ Production of both hardwoodand softwood timber is projected toincrease Southwide, but the largestpercentage increases are projectedfor northern and western portionsof the region, especially in Alabama,Arkansas, Kentucky, Tennessee, andVirginia.

■ Timber prices are projected toincrease in the United States and theSouth over the next 40 years undertwo plausible scenarios. The pricerises serve as continued incentivesfor private timberland owners tokeep land in forest in some places,

Chapter 13:Timber ProductsSupply and DemandJeffrey P. Prestemon and Robert C. AbtSouthern Research Station, USDA Forest Service; andDepartment of Forestry, North Carolina State University

What are the history,status, and projectedfuture demands for

and supplies of woodproducts in the South?

Southern Forest Resource Assessment300

TIMBER

Price

Inventory

Harvest

Inventorymodel

Stumpagemarket

Empiricalplantationallocation

model

Timberlandallocation

model

■ Historical production levels weredescribed for the South’s major speciesgroups and timber products, includingpulpwood, sawtimber, residues, fuelwood, and other fiber products.Southern production was sometimescontrasted with similar productionoccurring elsewhere in the country.

■ The linkages to international marketswere evaluated, and implications ofchanging wood products exportsand imports of competing materialswere considered.

■ Market linkages with other partsof the United States were evaluated.

■ Projections of future timberlandareas by major forest types, timberinventories, timber growth, timberremovals (production), and timberprices were made under a base scenarioof supply-and-demand assumptionsand under three alternative scenarios.Projections for the South were put intoadditional context with the rest of theNation and the World by reportingsome findings of the 2000 Draft RPAassessment (Haynes and others 2002).

■ Possible effects of land use changeon timber supplies were evaluated.

■ The impacts of changes in intensityof forest management and in forestproductivity on timber supply andforest composition were described.

Data and space limitationsconstrained the extent and detail ofinformation to provide. The chapterdoes not describe every issue ofhistorical, current, or potential futureimportance for the South’s forests.Further, a lack of data on historicalproduction and consumption patternslimited opportunities to describe anddraw conclusions about some impor-tant trends and relationships. Theprimary sources of data for the chapterare given, however, for those whowish to pursue certain issues in detail.Similarly, the methods of analysisare outlined rather than explained indetail. Details are to be found in thecited literature. Finally, those interestedin broader, national projections anddetails about other regions of theUnited States are directed to the2000 RPA assessment (Haynes andothers 2001).

As in the rest of this Assessment,this chapter does not evaluate policiesor make policy recommendations.Those interested in conducting thesekinds of analyses, however, may find

the material presented here to beuseful starting points.

Methods

Much of this chapter is concernedwith projections of the future. Becauseprojecting the future is controversialand complex and always subject togreat uncertainties, some explanationof projection procedures is warranted.Trends in the southern timber sectorwere projected with a partial equi-librium model of the southern forestsector, the Subregional Timber SupplyModel (SRTS) of Abt and others (2000).Nonproductive forest land and publictimberlands are not modeled by SRTS.Hence, SRTS provides projections ofprivate timber inventories, growth,removals, prices, land use, andtimberland area by five broad forestmanagement types at sub-State andecoregion (Bailey 1995) levels. TheSRTS projections are based on theresults of empirical models of timberand land supply-and-demand relationsto prices, income, and other variables.The projection period for this Assess-ment had a starting point of 1995and an ending point of 2040. Dataon public timberland and nonpro-ductive forest land are not includedin the projection or any of theaccompanying results.

SRTS consists of four models (fig.13.1): (1) a timber inventory model,which projects each year’s softwood(coniferous) and hardwood (non-coniferous) timber growth (net aftermortality) on existing acres, based ona set of growth equations and on theprevious year’s harvests of softwood

and hardwood; (2) a stumpage marketmodel consisting of supply-and-demand curves for timber softwoodand hardwood timber harvests, whichdetermines the amount of harvests,the timber prices, and the volume andthe state of the inventory in softwoodand hardwood; (3) a pine plantationallocation model, which determineshow many acres of pine trees to plant,given the softwood price and otherfactors; and (4) a timberland allocationmodel, which determines how muchprivate land is devoted to forest andallocates that land to either timberlandor nonproductive forest land, giventimber prices, financial returns toagricultural land uses, and otherfactors. For each year of the projection,SRTS solves for the combination ofSouthwide softwood and hardwoodtimber prices, softwood and hardwoodtimber harvests, pine tree plantingacres, and total timberland area thatmakes the supply of timber equal itsdemand. Although land use is projectedat the county level, the precision ofhistorical inventory and harvest datalimits the smallest unit of inferencefor the projected variables in the modelto the USDA Forest Service, SouthernResearch Station, Forest Inventoryand Analysis survey unit (FIA).

The primary outputs of the modelare annual values of: (1) timberlandarea by five forest management types(pine plantation, natural pine, mixedoak-pine, upland hardwood, andbottomland hardwood) by survey unit;(2) a single-volume measure of timbergrowth, removals, and inventory bymanagement type, survey unit, andowner; and (3) indices of Southwideaggregate softwood and hardwood

Figure 13.1—Schematicof the Subregional TimberSupply model (Abt andothers 2000).

301TIMBER


Table 13.1—Subregional Timber Supply Model assumptions

Variable Scenarios Species Ownership Value Source of the assumption

Assumed annual outward shift(increase) in timber demand All All All 1.6%/yr 1993 RPA

Southern pine plantation areaelasticity with respect totimber price All All Industry 0.60 Murray and Lee (1990)

All All NIPF 1.8 Murray and Lee (1990)

Timberland area elasticity withrespect to timber price All All All ~.3 Hardie and others (2000)

Rural area elasticity withrespect to population, income,and agricultural rents All All All Imbedded Hardie and others (2000)

Supply price elasticity All Hardwood All .45 Adams and Haynes (1996)All Softwood All .29 Adams and Haynes (1996)

Demand price elasticity IH All All -.50 Abt and others (2000)IL All All -.50 Abt and others (2000)EH All All -5.00 This AssessmentEL All All -5.00 This Assessment

Southern pine plantationgrowth rates

IH All Industry 75% by 2040 This AssessmentIH All NIPF 37.5% by 2040 This AssessmentIL All Industry 50% by 2040 This AssessmentIL All NIPF 25% by 2040 This AssessmentEH All Industry 75% by 2040 This AssessmentEH All NIPF 37.5% by 2040 This AssessmentEL All Industry 50% by 2040 This AssessmentEL All NIPF 25% by 2040 This Assessment

NIPF = nonindustrial private forest; IH = (base case) inelastic timber demand and high plantation volume growth rates; IL = inelastic timber demand andlow plantation volume growth rates; EH = elastic timber demand and high plantation volume growth rates; EL = elastic timber demand and lowplantation volume growth rates.

ERRA

timber prices. Because these projectedvariables are outcomes of the model,they may all be termed endogenous(that is, determined by the model).Two exceptions on the endogeneityof land use were Kentucky after 2020and Oklahoma for the entire projectionperiod. Data limitations permittedprojections of Kentucky’s land useallocation only through 2020, butremained fixed thereafter. Oklahomadid not have an applicable land usemodel, so that portion of Oklahomathat was included in this Assessmenthad land use (and, hence, timberlandareas by owner and survey unit) fixed

and softwood timber shift each year.This was specified as a 1.6 percentannual expansion in demand. The rateof annual timber demand growth wasbased on historical trends and historicalrelationships among population growth,technological change, and timberproduct consumption patterns. Also,this growth rate is roughly the same asthat specified for the South in the DraftRPA 2000 assessment. The exogenousdetermination of timber demandgrowth implies that the model takes therest of the World as given, so that themodel does not feed back to otherregions when calculating its annual

at observed 1993 levels during theentire projection period.

SRTS has several exogenous inputs—prespecified model parameters and thelevels and trends of certain variablesthat set the context of the modelsolution. While the model parameters,which quantify the relationships amongendogenous and exogenous variables,are held constant, alternative levels ofsome of the exogenous variables andsome model parameters collectivelydefine four projection scenarios (table13.1). One exogenous variable is timberdemand growth over time, a forecast ofhow the demand curves for hardwood



TIMBER

combinations of timberland acres bymanagement type, timber harvestvolumes, and prices. The exogenousdetermination of demand growththerefore does not allow southerntimber prices to induce technologychanges in the product manufacturingsector, nor does it allow timber pricesto directly affect the rate of substitutionof other raw materialsor nonsouthern virgin wood fiber forsouthern virgin wood fiber in forestproduct manufacturing, except to theextent that the historical rate of timberdemand growth embodies the historicalrate of product substitutions andtechnology changes. An accountingfor those kinds of feedback might bejustified if the South were a smallregion compared to the rest of thecountry (or if model complexity wereunrelated to model accuracy). However,the South dominates the U.S. timbermarket, so the exogenous demandgrowth determination can be viewedas a reasonable approximation oftrue national market functions.

Also predetermined prior to solvingthe SRTS model for the 45 years ofprojection is a set of variables involvedin the land use allocations. Thesevariables include projections ofpopulation growth, aggregate U.S.economic growth, agricultural rents(the real annual monetary returns tousing land to produce an agriculturaloutput), and residential land rents(see chapter 10 for a more detaileddiscussion of the land use module).Agricultural rents were specifiedas constant (in real terms) over theentire projection period. Anotherkey exogenously determined variablein the SRTS projection is the rate ofincrease in the growth rate of pineplantations in the South. Underlyingprojection parameters and inputs usedin the SRTS model projections for theSouth are shown in table 13.1.

Although four scenarios weremodeled in this Assessment, one,abbreviated IH for “inelastic demand-high plantation growth rate increase,”is designated as the “base case” for tworeasons. First, the inelastic demandassumption is consistent with empiricalfindings of responses of demand toprices and is consistent with assump-tions of RPA projections. Second,the SRTS model authors determined,through informal surveys of industryand pine plantation experts in the

South, that the higher plantationvolume growth rate increases (75percent for industry plantations and37.5 percent for nonindustrial privateplantations) are closest to a lowerbound on plantation growth rateincreases expected over the period.These alternative scenarios were per-formed to demonstrate the marginaleffects of plantation growth rates andtimber demand elasticities on importantmodel outputs. Results from the basecase scenario are discussed first. Someof the results are contrasted with resultsfor the other scenarios, and figuresdescribing the results of other scenariosare also made available to the reader.

As with many forecasting models,an underlying assumption in SRTSfor this Assessment is that timbersupply-and-demand and land usesupply-and-demand relationshipsremain stable. In that sense, the proj-ections do not account for changes inthe share of the available or harvestabletimber out of all timberlands ownedby various owner categories (includinggovernment, industry, and nonindus-trial private). Neither do the projectionsincorporate any predicted changesin wood product substitutes, woodproduct manufacturing technologies,real costs of timber management orproduction, or consumer tastes andpreferences. Finally, the projectionsdo not incorporate the effects of anyexpected changes in industrial structurein the paper or other industries. To theextent that such structural changes inthese sectors affect assumed underlyingsupply and demand parameters, ourprojections are inappropriate. Wheninterpreting the projections reportedin this chapter and projections reportedelsewhere, it is important to considerthat projections and their underlyingassumptions about economic variablesbecome less and less reliable as thelength of the projection increases.

SRTS has been designed to describeprojections of the future for smallregions or specific parts of the South.The FIA boundaries divide eachSouthern State into three to sixsections, whose boundaries followcounty lines but generally divide theStates into physiographic regions. Anadvance of SRTS from the projectionsprovided in Abt and others (2000) isthat this version of SRTS now permitsreporting of outputs by spatial unitscalled ecological regions (Bailey 1995),

which are not associated with politicalboundaries. As it turns out, ecologicalregion boundaries follow survey unitboundaries fairly closely, as bothdivision structures are based on manyof the same factors. Given this, little ofwhat is reported in this chapter actuallyis described in the context of ecologicalregions, though the data outputs couldbe reorganized in that fashion.

Data Sources

Data on international trade in timberproducts were obtained from theUnited Nations Food and AgriculturalOrganization (2000a). Historicalnational forest timber harvest datawere obtained directly from theNational Forest System’s fourth quarterannual totals in the cut-and-soldreports (U.S. Department of AgricultureForest Service 2000). National-levelhistorical harvest and wood use andproductivity information was providedby the Forest Products Laboratory(Personal communication. [2000].Kenneth Skog, Project Leader, and PeterInce, Research Forester, Forest ProductsLaboratory, One Gifford Pinchot Drive,Madison, WI 53705) and fromsupporting documents (Ince 2000).Pulpwood production data for theSouth are from Johnson (1996),Johnson and Howell (1996), andJohnson and Steppleton (1996, 1997,1999, 2000). Removals data by typeof product were provided by the U.S.Department of Agriculture ForestService (1958, 1982), Hair (1963),Phelps (1980), Waddell and others(1989), Powell and others (1994),and Haynes and others (2001). Draft2000 RPA projection informationwas obtained from Haynes andothers (2002).

Input data for all SRTS projectionscenarios derived from the plot-leveldata from the latest FIA surveysavailable at the time of the projectionfor each State in the South. The yearsof the latest surveys used are displayedconcisely in the last two columns ofa table reported in chapter 16, 1970sto 1999. The later of the two surveyswere as follows: Alabama 1990,Arkansas 1995, Florida 1995, Georgia1998, Kentucky 1988, Louisiana 1991,Mississippi 1994, North Carolina1990, Oklahoma 1993, South Carolina1993, Tennessee 1999, Texas 1992,and Virginia 1992. The next most

303TIMBER


Table 13.2—Direction of trade among major timber products trading countries, 1997

Importer

Exporter Brazil Canada Europe Indonesia Japan Malaysia Russian Fed. USA ROWa

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Dollars (million) - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Brazil 44 1,027 69 208 5 0 724 797Canada 187 2,551 122 3,201 57 3 18,053 906Europe 309 225 214 252 165 385 2,170 58,380Indonesia 3 31 636 2,059 90 1 471 1,803Japan 11 19 87 55 92 1 258 1,117Malaysia 0 9 421 24 1,826 0 143 1,603Russian Federation 2 4 1,536 4 787 5 70 492USA 388 3,073 4,050 233 3,752 142 14 4,046ROW

a199 59 2,326 256 3,106 247 56 1,334

a Rest of the World.Source: United Nations Food and Agricultural Organization (2000).

recent survey for each State framesthe years between which growth andremovals information was obtainedfor the purposes of this analysis. Whilethe oldness of certain surveys leadsto potential inaccuracies due to morerecent trends in those States, thiswas beyond our control.

Results

History and CurrentStatus of Supplyand Demand

World demand and supply historyand status—The United States is thelargest producer of industrial timber inthe World. For the last 40 years, it hasproduced a fairly stable 25 percent oftotal World production of industrialroundwood (United Nations Food andAgricultural Organization 2000b). In1999, the World produced about 53.2billion cubic feet (bcf), while the UnitedStates produced 15.1 bcf, or 28.5percent of the total. The second largestproducer, Canada, produced about12 percent (6.4 bcf) of the industrialroundwood in 1999. In order, the nextmost important were China (3.6 bcf),Brazil (1.9 bcf), Sweden (1.9 bcf),and Finland (1.8 bcf).

Although those countries are majorproducers, domestic demands in thosecountries greatly influence their staturein international markets for timberproducts. Observed trade flows inwood and paper products Worldwide(table 13.2) can largely be ascribed

to differences among countries insize of demand, amount of forest ortimberland, and distance betweentrading partners (Bonnefoi andBuongiorno 1990). Besides thesefundamental factors, trade is affectedby government policies such as tariffsand nontariff barriers. Timber productstrade also seems to be related tohistorical political relationships(Castillo and Laarman 1984).

The large size of the United Statesforest resource helps to determine whythe country produces so much, whilethe size of its domestic economy helpsexplain why it imports so much. Howmuch a country imports and exportsis determined by whether the country’sdomestic manufacturers supply morethan the country’s domestic consumersdemand at current prices. Countrieswith reasonably free trade typicallydo not demand exactly what domesticproducers supply. Thus, although theUnited States because of its extensiveforest resources is the World’s biggestproducer and second largest exporter,after Canada, the relatively free flow ofimports, large population, and high percapita income enable the United Statesto be the World’s largest timber productimporter. To illustrate, in the pastdecade and in terms of dollar value,the United States imported 60 percentmore timber products than it exported.

The costs of product movement arewhy the distance between marketsplays a role in determining both thescale of trade and specific tradingpartners. Usually, the closer physicallythat two trading partners are, the

lower the transport cost. Canada andthe United States possess the largestbilateral trade flow, partly because thetwo countries have a long commonborder. Proximity also explains partlywhy virtually every country southof the United States border countsthe United States as both its primarysource of timber product importsand its principal destination of timberproducts exports. In Asia and Europe,the dominant trade flow is from nearbyAsian supply sources (Indonesia,Malaysia, New Zealand, and Russia)to nearby demand centers.

Both the volume and value oftimber products trade have beengrowing rapidly Worldwide, andso trade is becoming more importantin many countries as an influenceon their forest sectors. Rapid tradegrowth can be ascribed both tooverall World economic growth andto decreasing barriers to internationaltrade. Tariffs on timber products havebeen decreasing Worldwide, as a resultof consecutive rounds of the GeneralAgreement on Tariffs and Trade (GATT)and World Trade Organization (WTO)(Barbier 1996).

The United States trades in all kindsof timber products. In terms of value,the most important exports are woodpulp, printing and writing paper,and hardwood lumber. United Statesexports go predominantly to Europe,Canada, Japan, Mexico, the rest ofLatin America, and the Caribbean. Themost important imports are softwoodlumber, newsprint, printing and writingpaper, and wood-based panels. Nearly


TIMBER

Har

vest

vol

ume

(bill

ion

boar

d fe

et)

Gro

ss d

omes

tic p

rodu

ct (

billi

on 1

996

dolla

rs)

100

75

50

25

0

Year19

50

1955

1960

1965

1970

1975

1980

1985

1990

1995

10,000

7,500

5,000

2,500

0

United States, excluding national forests

All national forests

United States’ gross domestic product

all of U.S. imports of softwood lumber,panels, and newsprint arefrom Canada. The United States hasnegotiated through GATT (now WTO)and other bilateral and multilateralaccords some of the lowest barriers toforest products imports in the World.These accords have helped to ensurethat U.S. barriers to timber productimports are kept low, probably facili-tating the import into this country ofwood fiber from emerging producerssuch as Brazil and Chile. These sameaccords, however, have also boostedtimber product exports to many ofthose same countries. Recent tradeagreements [the Canada-U.S. FreeTrade Agreement (CUSTA) and theNorth American Free Trade Agreement(NAFTA)] have reduced many barriersto trade between these two tradingpartners, but some disputes have longsimmered over softwood lumber andother product exports to the UnitedStates. Because the United States isa net timber product importer, then,these lower barriers may have servedto reduce returns to timber growingand timber product manufacture inthe United States. For example, agrowing trend has been the importationof hardwood fiber into the UnitedStates from Latin America, especiallyBrazil. So far, these imports arerelatively small, but a possible resultof this trend, should it continue, wouldbe to dampen prices below what theywould be without such fiber imports.Nevertheless, the trade liberalizationagreements, including NAFTA, CUSTA,and WTO-sponsored rounds of barrierreductions, tend to increase aggregatetimber product output in the longrun and to increase exports of U.S.wood products (e.g., Prestemonand Buongiorno 1996), benefitingAmerican timber producers. Thenet effects of trade liberalization onthe entire U.S. timber-based sector,therefore, are probably small (Barbier1996, Trømborg and others 2000).

Southern timber products ofimportance in trade include southernpine (Pinus spp.) lumber, hardwoodlumber [especially oak (Quercus spp.)],southern pine plywood, kraft pulp,and kraft-based paper (packaging andpaperboard). The principal destinationsfor these products are Western Europe,Latin America, and the Caribbean.Because the population and economiesof the latter two regions are growingquickly, demand there for southern

Figure 13.2—Total U.S. harvests, national forest harvests, and U.S. realgross domestic product, 1950 to 1998 {all national forests harvest (cut)volume—U.S. Department of Agriculture, Forest Service (2000); U.S. totalharvest volumes—Personal communication. [2000]. Kenneth Skog, ProjectLeader, Forest Products Laboratory, One Gifford Pinchot Drive, Madison,WI 53705; U.S. gross domestic product— U.S. Department of Commerce.2000. [Data]. On file with: Southern Research Station, Forestry SciencesLaboratory, P.O. Box 12254, Research Triangle Park, NC 27709}.

Figure 13.3—National forest harvests by geographical region, 1950to 1998 [Pacific: national forests in the Pacific Northwest (Region 6),Pacific Southwest (Region 5), and Alaska (Region 10); Rockies: Northern(Region 1), Rocky Mountain (Region 2), Southwestern (Region 3),and Intermountain (Region 4); South: Southern (Region 8); and North:Eastern (Region 9) and former Region 7] (U.S. Department of Agriculture,Forest Service 2000).

7

6

5

4

3

2

1

0

Cut

vol

ume

(bill

ion

boar

d fe

et)

Year

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

Pacific

Rockies

South

North

1998

305TIMBER


Indu

stria

l woo

d pr

oduc

t out

put

per

unit

of r

ound

woo

d in

put

(pou

nds

per

cubi

c fo

ot)

27

25

23

21

19

17

15

Year

1900

1910

1930

1950

1960

1970

1990

1920

1940

1980

1998

back. This process was repeated asEuropean settlement moved westwardover the ensuing decades. The finalstages of old-growth forest liquidationhappened in the Pacific Northwest inthe last century; the remaining portionis largely protected by reserves, parks,and government policies adopted inthe late 1980s and 1990s.

National forest harvests have changedmarkedly since 1950 (fig. 13.2). In thatyear, their share of the U.S. harvestswas 6.6 percent. By 1964, it was 17.5percent. But by 1998, the share haddropped, this time to 3.5 percent, aresult of desires to preserve remainingold-growth forests in the West, toprotect habitats of endangered species,and to limit clearcutting. Except for thesouthern region, harvests have declinedsince 1990 to small fractions of harvestsobserved in the mid-1980s (fig. 13.3,which excludes the early harvests intropical national forests). The largestpercentage drop in harvests was inthe Pacific regions, notably the PacificNorthwest. The Pacific Northwest’sshare in total U.S. harvests declinedfrom a 1950 to 1989 average of 5.8percent of all harvests to 0.7 percentof all harvests in 1998.

End uses for harvested wood haveevolved over the years, with the mixof uses moving from solid woodoutputs, such as lumber, to a greatershare of composite products, such asparticleboard and paper. As a result,the amount of timber being processedinto wood chips, nonwood materials,and recycled fiber has been increasing(fig. 13.4). The increased use of recy-cled fiber and other fiber and productsubstitutes shown in figure 13.4can explain part, but not all, of thedecline in timber harvests in the UnitedStates since the early 1990s. Anothermajor factor is the steady rise in netproduct imports. Third has been theincreasingly complete utilization ofwood in manufacturing processes(fig. 13.5) (Ince 2000), which wouldcompensate for some of the steadilyrising demand for timber productsthat has been observed in recentdecades. Wood-use efficiency rose41 percent from 1952 to 1998. Wood-use efficiency was 9 percent higherin 1998 than in 1990, which can alsoaccount for much of the reductionin the observed timber product outputof the past few years.

FPO

Figure 13.4—U.S. fiber consumption, by source, 1965 to 1998 (Personalcommunication. [2000]. Kenneth Skog, Project Leader, and Peter Ince,Research Forester, Forest Products Laboratory, One Gifford Pinchot Drive,Madison, WI 53705).

Figure 13.5—Industrial wood productivity in the United States,1900 to 1998 (Ince 2000).

timber products exports also can beexpected to rise rapidly. Asian countriesfor the most part have not been majorpurchasers of southern products (oneexception is hardwood chips goingprimarily to Japan), so the effect ofthat region’s growth in populationand wealth, should long-term trendscontinue, would be to increase timberproduct prices in the United Statesand Canada.

United States supply-and-demandhistory and status—Demand fortimber products in the United Stateshas shifted among regions continuouslysince the 1800s. Settlement in the East,upper Midwest, interior West, andthe far West was often preceded andfacilitated by harvests of old-growthforests. In the East, virtually all of theforests were harvested in the processof land cover conversion to agriculture,but some forests were allowed to grow

Other timber harvestNFS timber harvestProduct importsRecycled fiberNon wood materialsLog and chip imports

Qua

ntity

(m

illio

n m

etric

tons

dry

wei

ght) 300

250

200

150

100

50

0

Year

1965

1970

1975

1980

1985

1990

1995

1998


TIMBER

Har

vest

vol

ume

shar

e

1.00

0.80

0.60

0.40

0.20

0.00

Year1952 1962 1977 1987 1992 1997

Pacific Rockies North South

Rev

enue

[200

1 do

llars

(bi

llion

)]

7

6

5

4

3

2

1

0

Year1952 1962 1977 1987 1992 1997

South

Pacific

North

Rockies

Southern supply-and-demandhistory and status—Southern Statesproduce most of America’s industrialwood output, and their share has grownsteadily since the 1960s (fig. 13.6).The South produced 41 percent of thecountry’s wood fiber output in 1952and 58 percent in 1997. Over the sameperiod, the South’s share of the World’sindustrial wood production rose from6.3 to 15.8 percent. Meanwhile, thePacific region’s share of the country’sproduction dropped from 24.8 to16 percent.

In terms of timber value, the South’srole in production has grown steadilysince the 1960s, as well. In otherregions of the United States, this sharehas been somewhat less stable (fig.13.7). As a result, the timber productsector has been a more dependablesource of economic output in the Southcompared to other regions. The increasein output observed in the South impliesthat investment opportunities forintensive forest management andproduct manufacture have improvedin the South relative to other regions(Guan and Munn 2000, Murrayand Wear 1998).

Over the last 50 years, the relativedesirability of western and southerntimber products has changed. Earlierin that period, western conifers,which dominated much of the timberproduct market, were considered idealin construction framing and sheathing,and in pulp. Spruce (Picea spp.), fir(Abies spp.), Douglas-fir (Pseudotsugamenziesii Franco), western hemlock[Tsuga heterophylla (Raf.) Sarg.], andwestern pines make excellent framinglumber and plywood because of theirlightness (low density), strength,stability, and workability. Southernpine, on the other hand, historicallywas not as desired as western andnorthern softwoods in constructionapplications. As timber product manu-facturing technology for southern pineadvanced, however, southern pine’sdesirability in national constructionmarkets improved. Until the 1960s,the technology for producing southernpine plywood with desirable charac-teristics for construction that couldcompete directly with western plywooddid not exist. Similarly, until the 1980s,when old-growth rot-resistant woodssuch as redwood and western redcedarbecame scarce and before chemicaltreating technology for southern

pine was perfected, treated southernpine lumber was not as desirablefor outdoor applications such asdecking. Since then, treated southernpine has supplanted these westernwoods for much of the outdoorapplication market.

Western manufacturers of strong,long-fiber pulp and paper rely largelyon residues from coniferous woodproducts manufacture—slabs, shavings,and trimmed edges. Therefore, thesoftwood sawtimber harvest reductionsin the West in the 1980s and 1990s

have been accompanied by reducedoutput of pulpwood. Nationally, pulpand paper manufacturing has becomemore reliant on sources other thanwestern conifers. Southern pine fibersare ideal for high-strength pulp(especially kraft pulp), so pulp andpaper manufacturing has become moredependent on pulpwood production inthe South as paper demand has grownand western timber production haswaned. The rise in the output andtechnological advancements instructural and nonstructural wood

Figure 13.6—Shares of timber harvest volumes, by USDA Forest ServiceRegion of the United States, 1952 to 1997 (Haynes and others 2002).

Figure 13.7—Timber harvest revenues, by USDA Forest Service Regionof the United States, 1952 to 1997 (Wear, D.N. 2002. [Data]. On file with:Southern Research Station, Forestry Sciences Laboratory, P.O. Box12254, Research Triangle Park, NC 27709).

307TIMBER


Year

Are

a pl

ante

d (t

hous

and

acre

s)

1,600

1,400

1,200

1,000

800

600

400

200

0

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

Industry

National Forest System

Broad NIPF

Other Federal

State and local government

1999

Vol

ume

(tho

usan

d co

rds)

Year

Pric

e (1

998

dolla

rs p

er c

ord)

Softwood roundwood volumeHardwood roundwood volumeSoftwood residues volume

Hardwood residues volumeLouisiana southern pine pulpwood priceLouisiana hardwood pulpwood price

40

35

30

25

20

15

10

5

0

1953

1958

1963

1968

1973

1983

1988

1993

1998

1978

35

30

25

20

15

10

5

0

panels and other engineered woodproducts has created new demandsfor smaller diameter and lower qualityhardwood and softwood timber.

Without increased investment inthe forest sector, production contractionin one part of the country, such asrecently observed in the Pacific North-west, inevitably leads to rising timberprices, rising imports, shifts in demandaway from wood-based and towardnonwood product substitutes, andthe development of new and moreefficient manufacturing technologies.In response to price rises, increases inwood product imports, and productsubstitutions (fig. 13.4), product man-ufacturers in the United States, theSouth, and elsewhere have enhancedwood-use efficiency (fig. 13.5).

Other responses to changingtechnologies and price increases havebeen new and rapidly rising rates ofinvestments by landowners in theSouth in pine growing technology.This technology has two parts: (1)intensive cultivation, including treeplanting, thinning, fertilization, andvegetation management; and (2)genetic improvement. An index ofsouthern investments in tree growingtechnology is the rate of tree planting(fig. 13.8). The trend in such plantinghas been upward since 1945, withtwo sharp peaks since that time. Thepeaks were created in part by incentivesprograms, including the Soil Bankand Conservation Reserve Programs.Although some of the planting is onnewly harvested plantations themselves,part of it is on land previously used foragriculture and part on land previouslycovered by natural forest types. Bothkinds of planting are indicative ofhow producers have sustained orincreased their investments in timbermanagement. The net effect of thoseinvestments has been a rising shareof pine plantations in the totaltimberland area in the South.

In spite of rising pulpwood pro-duction and improvements in productmanufacturing efficiency, producershave not been able to increase outputas fast as the economy’s demands forpulp-based products have grown. Asa result, pulpwood prices (adjustedfor inflation) have risen (fig. 13.9). In1953, virtually no residues (wood chipsand other wastes) were used in woodproducts manufactured in the South;panels and pulp were made from

Figure 13.8—Tree planting in the South, by major ownership group, 1945to 1999 [Robert F. Moulton, compiled from annual USDA Forest Servicetree planting reports; including estimates for industry (Arkansas 1954;Florida 1981; Georgia 1954, 1982; Louisiana 1954, 1981; Mississippi1954; Texas 1981)].

Figure 13.9—Southern pulpwood production, 1953 to 1998, by productclass, and Louisiana southern pine and hardwood pulpwood stumpageprices, 1955 to 1998 (Johnson 1996; Johnson and Howell 1996; Johnsonand Steppleton 1996, 1997, 1999; Louisiana Department of Agricultureand Forestry 2000).


TIMBER

PulpPost and polePlywood and veneerOther typesComposite

roundwood. By 1998, residuesaccounted for about 29 percent ofthe volume of both softwood andhardwood fiber received at the gatesof pulp mills and composite panelmills (Johnson and Steppleton 2000, p.9). Given the price rise along with theproduction increase, it is apparent thattechnological change and the economicadvantages provided by the technologyhave not been enough to keep pricesfrom rising in real terms. Still, thesesteadily rising prices serve as incentivesfor consumers of pulpwood and

producers of pulpwood to invest inefficiency-enhancing technologies.

Another important trend that hasarisen out of changing technologiesand increasing prices has been therising share of hardwood in southerntimber production. For example, in1953, hardwood roundwood was about12 percent of all roundwood removed,while in 1999, hardwood roundwoodwas 34 percent. Hardwood roundwoodnearly tripled in output while softwoodroundwood slightly more than doubled.Price changes reflect this: hardwood

roundwood prices have increased bytwo-thirds in real terms over the period,while softwood prices have increasedby about 15 percent.

Another way that producers oftimber products in the South haveadapted to rising demands, increasinglycompetitive substitute products andimports, and rising prices is by alteringtimber processing. One change inrecent years is the chipping of woodat satellite locations. This process iscontroversial because it encourages

Figure 13.10—Chip mills in theSouthern United States (no millsfrom western regions of Texas orOklahoma are shown) (Prestemonand others 2001b).

Figure 13.11—Other wood-usingmills in the Southern United States.Each mark indicates a town orcity in which at least one mill waslocated during mill censusesoccurring between 1992 (Virginia)and 1997 (Texas). Minimum millsize thresholds for inclusion in thecensuses varied by State. The datainclude such qualifying mills in all ofTexas and Oklahoma (Prestemonand Pye 1999).

At pulp mill siteAt sawmill siteStand alone

309TIMBER


Softwood residue production

Softwood chip exports

Hardwood residue production

Hardwood chip exports

Dry

wei

ght (

mill

ion

tons

)

25

20

15

10

5

0

Year

1953

1956

1959

1962

1965

1968

1971

1974

1977

1980

1983

1986

1989

1992

1995

1998

1999

harvesting in areas not previouslysubject to harvesting and encouragesclearcutting, especially of naturalmanagement types that before wereharvested in a different way. Many viewthis as negative. Others have viewed thetechnology positively, creatingconditions for better forest managementbecause the chipping technologydiscourages incomplete or high-gradedharvests and because it providesadditional income to owners of lowerquality timber. Before the 1990s, pulpmills and manufactured wood panelmills relied heavily on remote logconcentration yards and maintainedlarge chipping facilities at the site ofpanel and pulp manufacture. Todaypulpwood-sized logs increasingly arechipped away from the mill and arebrought to the mill as needed. Per unitof volume, moving wood in chippedform is cheaper than moving pulplogs (Dodrill and Cubbage 2000),providing a significant economicbenefit to pulpwood consumers andlog producers. The current distributionof these remote or stand-alone chipmills is shown with the locationsof other kinds of chipping facilitiesin figure 13.10. The buyers of mostof these chips, pulp mills and manu-factured panel mills, are shown alongwith miscellaneous other mills in figure13.11. A small portion of these chipsalso derives from a few of the thousandsof southern sawmills (fig. 13.12). Notethat the remainder of the material usedby pulp mills is processed as chips

onsite at pulp mills and panel mills,arriving there as roundwood. See figure13.10 for the locations of those pulpmills and panel mills.

The majority of chips produced inthe South are used to make paper andcomposite wood panels. In 1998, therewere 159 chip mills (Prestemon andothers 2001a), but by 2000, 146 werefound in the South (Prestemon andothers 2001b). More than three-fourthsof all chip mills were stand-alone in2000, not directly tied to a particularwood processing plant; most of the

remainder were tied to a pulp mill.Chip mills processed about 27 percentof the pulpwood in the South in 1999(Hyldahl and others 2000). Theyproduced 47 million green tons ofchips in 1998, 45 million green tonsin 1999, and 39 million green tonsin 2000. In 1999, approximately42 percent was softwood and 58percent was hardwood.

Not all of the wood chips produced inthe South are consumed by U.S. mills(fig. 13.13). Since 1989, increasingamounts of wood chips have been

Figure 13.12—Sawmills in theSouthern United States. Each markindicates a town or city in which atleast one mill was located duringmill censuses occurring between1992 (Virginia) and 1997 (Texas).Minimum mill size thresholds forinclusion in the censuses variedby State. The data include suchqualifying mills in all of Texasand Oklahoma (Prestemon andPye 1999).

Figure 13.13—Southern wood chip residue production (1953 to 1999) andwood chip exports from southern customs districts (1989 to 1999), softwoodand hardwood (Hansen and Hyldahl 2001; Johnson 1996; Johnson andSteppleton 1996, 1999, 2001).


TIMBER

Rem

oval

s vo

lum

e (b

illio

n cu

bic

feet

)

12

10

8

6

4

2

0

Year1952 1954 1958 1963 1967 1972 1976 1986 1991 1996

Saw logsVeneer logsPulpwoodFuelwoodAll other products

exported from the United States.Between 1989 and 1999, residueexports from southern ports increased369 percent for hardwood and 372percent for softwood. Because mostresidues today are in the form ofwood chips, we can say that the exportshare of southern hardwood residueproduction increased from 12 percentin 1989 to 39 percent in 1999, whilethe export share of southern softwoodresidue production increased from0.3 to 1.3 percent between those years(U.S. Department of Commerce. 2000.[Data]. On file with: Southern ResearchStation, Forestry Sciences Laboratory,P.O. Box 12254, Research Triangle Park,NC 27709). Hardwood and softwoodwood residue production comprisedapproximately 12 percent of allwood fiber production in the Southin 1996, the latest year for whichdata are available.

Another indicator of the effect ofchanging wood production andmanufacturing technology is the risingimportance of more highly manu-factured timber products. Apparentlythere is a trend toward concentratinga higher proportion of value addedat the point of initial manufacturing.Since the 1950s, the use of wood forfuel, posts, poles, and pilings has de-clined, in favor of wood produced forlumber, paper, and engineered woodproducts (fig. 13.14). The proportionof output going to fuel wood in the1950s was over 20 percent; it has sincedropped to under 3 percent. The shareof output dedicated to the categoryof other product removals—primarilyfor posts, poles, pilings, and compositeproducts—has fallen by two-thirds,settling today at about 2.5 percentof timber product output in the South.Between 1954 and 1996, the percent-age of wood removed as saw logs wasnearly constant, at around 38 percent.Pulpwood’s share rose from 21 percentin 1952 to 47 percent in 1972 and hassince leveled off at around 40 percent.The proportion of output in the formof the largest and highest quality logs,veneer logs, has trended upward, from3 percent in 1952 to about 9 percentin the 1990s. Hence, in contrast tothe trend toward more wood productsderived from pulpwood, which doubledin importance between 1954 and 1996,the importance of sawmills, especiallythose manufacturing hardwood lumberand veneer, has remained constant.In the South, the largest number

of hardwood sawmills is in areaswhere hardwood production is mostdominant: mountainous portions ofVirginia, North Carolina, and Tennessee(fig. 13.12). But overall productionof hardwood timber is highest inMississippi, North Carolina, Georgia,Alabama, and Virginia.

Projections ofSupply and Demand

Supplies of and demands for timberproducts in the South will dependheavily on national and World trends.Southern supplies and demandsthrough 2040 were projected withthe SRTS model, with national andinternational trends taken as given.What follows is a discussion of someof the World and national projectionsfrom the literature plus a descriptionof how SRTS projects what will happenin the South in the coming decades.

World supply-and-demandprojections—World timber productionis expected to rise steadily well into the21st century. Projections by Trømborgand others (2000) show that timberproduction will increase by 1.2 percentper year through 2010, with likelycontinued increases beyond that year.Their analysis also projects: (1) thatU.S. growth in production will be 0.4percent, implying that the United Stateswill remain a timber product importer;(2) that the U.S. share of exportson World markets may decline; and(3) that U.S. imports will rise. TheUnited States experienced an average

compound annual growth rate fortimber products output of 1.4 percentfrom 1961 to 1999, so this lower rate of0.4 percent appears to be a substantialdeparture from the past but closer tothe realized compound annual growthrate since 1990, which has beenessentially nil (0.04 percent).

United States supply-and-demandprojections: RPA—The Draft Forestand Rangeland RPA assessment (Haynesand others 2002) projects that thecharacter and location of timber andtimber products output will changeover the coming half-century whiletimber product prices and land andtimberland area will decline by 3percent. The Draft 2000 RPA projectsin its base projection that the area offorestland is expected to decline by3.6 million acres in the South and todecline by 19.6 million acres elsewhere.

The Draft 2000 RPA assessment baseprojection also projects that privatelyowned forests in the United States willbe more intensively managed, partlyas a response to declining forest area.It also projects that private forests willbe expected to produce an increasingshare of small-diameter materials forpulp and composite wood products.Timber production overall is projectedto continue its shift toward the South,which contains a large share of theNation’s private forests and timberland.Domestic consumption of softwoodsis projected to increase in the baseprojection by 47 percent andhardwoods by 29 percent between

Figure 13.14—Removals by destination product, Southwide, all species,1952 to 1996 (data for 1954 to 1972 all other products include fuelwood) [Hair(1963, p. 32-33); Haynes and others (2002); Phelps (1980, p. 31); Powell andothers (1994, p. 36); U.S. Department of Agriculture Forest Service (1958, p.570, 641-642; 1982, p. 422); Waddell and others (1989, p. 89)].

311TIMBER


1996 and 2050, while harvests areprojected to increase 30 percent forsoftwoods and 17 percent for hard-woods between 1996 and 2050. Percapita consumption of roundwood,however, is projected to remain fairlystable, at 0.8 tons per capita per year.

The shares of outputs going intovarious solid wood products are pro-jected by RPA to change over the nexthalf-century, much of that driven byevolving technologies that result inrising technical efficiencies. Compositewood structural panels are projectedto partially displace plywood, whilesoftwood lumber shares are projectedto grow relative to hardwood lumber.Imports from Canada and elsewhereare projected to rise, especially in theshort term. Softwood lumber, pulp,paper, and paperboard production areprojected to increase most in the South,especially in the western portion ofthe region. Although manufacturingefficiency (units of output per unit ofwood input) is projected to continueto increase, the rate of that increaseis projected to slow, relative to thatexperienced in the 1900s. Between2000 and 2050, the output:inputratio is projected to rise by 16 percent.

Hardwood and softwood timberharvests are projected to increasesimilarly, by over one-third, over thecoming half-century. This rise will bemade possible by improvements intimber growing technology, especiallyintensification through plantationmanagement, fertilization, thinning,and genetic improvement.

The Nation’s softwood timber harvestsare projected to continue to comemostly from the South, rising from 61percent of U.S. timber harvests in 1997to 65 percent by 2050 (Haynes andothers 2001). The shares of softwoodprovided by the other regions of thecountry are projected to be steady orto decline over the coming 50 years.In hardwood, production in the RockyMountain West and Pacific Coast isexpected to rise but remain small, whilethe Northern United States is projectedto rise slightly in importance while theSouth declines slightly in importance.Nevertheless, the South and Northare both projected to increase theirhardwood outputs. By 2050, the Southis projected to provide 50 percent ofhardwood roundwood harvests and theNorth 44 per cent. In total, the South’sshare rises only slightly, by less than

1 percent by 2050 compared to its58-percent share of harvests in 1997.To a large extent, the high and risingproductivity and area of southernpine plantations makes the risein the southern share of softwoodharvests possible.

The RPA assessment projects that theUnited States will increase its depen-dence on foreign sources of wood fiber(logs, lumber, panels, residues, pulp,waste paper, etc.) as a proportion oftotal consumption. The projectionshows imports providing 27 percentof wood fiber consumed in 2050,compared to 20 percent in 2000.These findings are consistent with theshorter run projections of Trømborgand others (2000), which show thatthe value of U.S. net exports (exportsminus imports) will become morenegative by 2010.

An effect of greater investment inmanufacturing technology and risingfiber demand is a projected relative risein the importance of recycled fiber inthe paper sector. Use of recycled fiberhas been increasing and will continueto do so (Ince 2000). Over the 50-yearRPA projection, recycled fiber use isprojected to more than double, whilewood fiber from timber harvests isprojected to increase by 40 percent.

Timber prices in the United States areprojected by RPA to change differ-entially, depending on product andspecies. Timber prices for softwoodsawtimber are projected to rise overthe projection in all regions of theUnited States by between 13 and 69percent (39 percent in the South),while softwood pulpwood pricesare projected to rise in the Northand fall (by 29 percent) in the South.In hardwood, sawtimber prices areprojected to rise slightly in the Northand more than double in the South.Hardwood pulpwood is projectedto nearly double in price in the Southand decline in the North.

Southern supply, demand,management intensity, and land useprojections: SRTS—We used the 2000RPA projections to provide national andglobal context, but we made projectionsfor the South independently from RPAprojections. SRTS projections of forestarea, harvests (removals), growth, andinventory were done under all scenariosoutlined in table 13.1 and described inMethods in this chapter. Starting-pointdata on inventory, net growth, and

removals used in the SRTS projectionswere obtained from the latest FIAdata available for download from theFIA Web site. Projection data for Stateswith relatively old surveys could bemisrepresented if growth, removals,or land use changes including treeplanting (fig. 13.8) have changedgreatly between the survey and1995. These problems may exist forprojections of South Carolina (due toHurricane Hugo’s effects on subsequentgrowth rate of trees in natural stands) aswell as Alabama, Kentucky, Louisiana,North Carolina, Texas, and Virginia.

Area projections for privatetimberland under the IH (base case)scenario for FIA survey units showthe South losing private timberlandover the coming decades. This loss,amounting to 1 percent over the 1995through 2040 projection, is net of anaggregate increase in the area of pineplantations and an aggregate decreasein the area of other forest types (fig.13.15). A detailed map of forest areachanges (fig. 13.16) shows that privatetimberland area is projected to increasein the western parts of the South, whilelosses are projected in States along thesouthern Atlantic seaboard. The gainsin private timberland area, facilitatedby rising timber prices relative toagricultural rents, will be concentratedin Alabama, Arkansas, Louisiana, andMississippi. Significant percentagelosses are projected for Florida, NorthCarolina, South Carolina, Tennessee,Texas, and Virginia. Within States,losses are projected to be concentratednear urban areas, while some rurallocations gain. This is not universallytrue, however. For example, all ofFlorida and South Carolina’s FIAsurvey units are projected to loseprivate timberland. The South’spopulation and State economies havegrown quickly and are projected tocontinue to grow quickly. With suchgrowth, the demand for land near theurban areas has been, and is projectedto continue to be, met by some clearingof forests. Under the IL (fig. 13.15B),EH (fig. 13.15C), and EL (fig. 13.15D)scenarios, aggregate timberland areain the South is projected to changeas well. What all of the figures 13.15A(IH), 13.15B (IL), 13.15C (EH), and13.15D (EL) show is that the areaof natural forest management types(all types except pine plantations) isprojected to shrink, while the plantedpine type increases. This trend would


TIMBER

Man

agem

ent t

ype

(mill

ion

acre

s)

80

70

60

50

40

30

20

10

0

Year

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

Tota

l tim

berla

nd (

mill

ion

acre

s)

200

150

100

50

0

Upland hardwoodPlanted pineNatural pine

Lowland hardwoodMixed pineTotal timberland

(A)

Man

agem

ent t

ype

(mill

ion

acre

s)

80

70

60

50

40

30

20

10

0

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

Tota

l tim

berla

nd (

mill

ion

acre

s)

200

150

100

50

0



2025

(B)

Man

agem

ent t

ype

(mill

ion

acre

s)

80

70

60

50

40

30

20

10

0

Year

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

Tota

l tim

berla

nd (

mill

ion

acre

s)

200

150

100

50

0



(C)

Man

agem

ent t

ype

(mill

ion

acre

s)

80

70

60

50

40

30

20

10

0

Year

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

Tota

l tim

berla

nd (

mill

ion

acre

s)

200

150

100

50

0



(D)

appear to be a continuation of thatobserved over the last 40 years, whenlittle net forest loss was registered butplantation area increased substantially(chapter 16, fig. 16.12).

Common to the IH and IL scenarios,pine plantation areas are projected toincrease by 21 to 26 million acres, orby about 67 to 80 percent from 1995levels of pine plantations. The pineplantation projections by scenario aredisplayed together, along with historicalamounts, in figure 13.17. Increases inpine plantation acres differ among thescenarios considered. These projectedincreases are similar to the projectedacreage of aggregate losses of thenatural forest management types underprivate ownership, keeping privatetimberland area largely unchanged over

the projection, 1995 to 2040. Commonto the EH and EL scenarios, however,is that pine plantation area is projectedto increase by about 25 percent,insufficient to completely outweighnatural forest-type losses, translatinginto a net loss in private timberlandarea of just over 27 million acres(15 percent) between 1995 and 2040.These lower plantation acres aregenerated because prices, to whichpine planting positively responds,do not increase as much under theelastic demand scenarios.

Apparent in the IH (base case) andIL scenarios is that pine plantation areais increasing at the expense of privatetimberland in other forest types, butthis tradeoff is only partial. As pointedout in chapter 16, during the 1980s

and 1990s, about 30 percent of newpine plantation acres in the Southderived from agricultural land, whilearound 70 percent came from conver-sion of natural forest managementtypes. Further, part of the loss ofnatural forest has historically been,and is projected to be, due to conver-sions to urban uses (see chapter 6 fordetails). Similarly, in the IH and ILprojections, a share of the private pineplantation acreage increase is projectedto be at the expense of agriculturalland as well as private timberlandthat is currently in natural forestmanagement types. In practice, thismeans that Gulf Coast States andthe Coastal and Piedmont regionsof Atlantic Coast States will gain themost pine plantations, while northern

Figure 13.15—Subregional Timber Supply Model projections of privatetimberland area by management type, 1995 to 2040, under fourassumptions: (A) IH or base case—inelastic timber demand and highplantation volume growth rates, (B) IL—inelastic timber demand and lowplantation volume growth rates, (C) EH—elastic timber demand and highplantation volume growth rates, and (D) EL—elastic timber demand andlow plantation volume growth rates.

313TIMBER


Change in percent-48 – -25-25 – 0

1 – 1010 – 27

Pin

e pl

anta

tion

area

(m

illio

n ac

res)

70

60

50

40

30

20

10

0

Year

1952

1962

1970

1977

1997

2000

2010

2020

2030

2040

1985

1992

Projected, IHProjected, IL

Projected. EHProjected, EL

Historical

199520202040

Virg

inia

Pin

e pl

anta

tion

area

(th

ousa

nd a

cres

)

10,000

8,000

6,000

4,000

2,000

0

Ala

bam

a

Ark

ansa

s

Flo

rida

Geo

rgia

Ken

tuck

y

Loui

sian

a

Mis

siss

ippi

Nor

th C

arol

ina

Okl

ahom

a

Sou

th C

arol

ina

Tenn

esse

e

Texa

s

State

Figure 13.16—Subregional Timber Supply Model projections byUSDA Forest Service, Southern Research Station, Forest Inventoryand Analysis survey unit of percent change in private timberland area,1995 to 2040, under assumptions of inelastic timber demand andhigh plantation volume growth rates.

Figure 13.17—Pine plantation area projections by scenario,and historical pine plantation area on private land in the South,1952 to 2040.

and interior regions will gain theleast plantation area.

A notable direct tradeoff, however,exists when comparing the plantationpine and natural forest managementtype projections done by the IH (basecase) and IL scenarios. In the ILscenario, softwood prices are projectedto rise at a faster rate than they are forthe IH scenario; the higher prices inthe IL scenario serve as the economicstimulus to landowners to plant evenmore trees. The difference between theIH and IL pine plantation rates yields

Figure 13.18—Pine plantation area by State onprivate land in the South for 1995, 2020, and 2040,as projected by the Subregional Timber SupplyModel, under the IH (base case) scenario, withinelastic demand and a high pine plantationgrowth rate increase.

the marginal effect of higher plantationgrowth rates on the area of private pineplantations and the area of timberlandin private natural forest types projectedfor 2040. In the IH scenario, pineplantations are projected to cover 53.6million acres in 2040, while in the ILscenario the figure is 57.9 million acres.Each percentage point increase ingrowth rate above a 50-percent increasefor industry [and each 0.5 percentagepoint increase above 25 percent fornonindustrial private forests (NIPF)]results in about 170,000 fewer acres

of projected pine plantations by 2040.Similarly, because the IH scenarioprojects private timberland area innatural forest management types of122 million acres and the IL projectsthat area to be 123 million acres, eachpercentage point (for industry and0.5 percent for NIPF) increase in pineplantation growth rate is projectedto save about 50,000 acres of naturalforest. Alternatively, if timber demandis elastically responsive to timber price,as laid out in the EH and EL scenarios,the effects of pine plantation growthrate changes on areas by managementtype are very small.

Figure 13.18 details the changes byState in pine plantation area projectedin the IH (base case) scenario. Pineplantation area changes vary amongSouthern States mostly due to differ-ences among States in the area ofindustry-owned forests, the amount


TIMBER

1995 2020 2040

Pin

e pl

anta

tion

area

(th

ousa

nd a

cres

)

Ecoregion

Cen

tral

App

alac

hian

30,000

25,000

20,000

15,000

10,000

5,000

0

Eas

tern

bro

adle

af o

cean

Eas

tern

bro

adle

af c

ontin

enta

l

Sou

thea

ster

n m

ixed

Out

er C

oast

al P

lain

mix

ed

Low

er M

issi

ssip

pi r

iver

ine

199520202040

Virg

inia

Nat

ural

fore

st m

anag

emen

t typ

e ar

ea (

thou

sand

acr

es)

20,000

16,000

12,000

8,000

4,000

0

Ala

bam

a

Ark

ansa

s

Flo

rida

Geo

rgia

Ken

tuck

y

Loui

sian

a

Mis

siss

ippi

Nor

th C

arol

ina

Okl

ahom

a

Sou

th C

arol

ina

Tenn

esse

e

Texa

s

State

Inve

ntor

y (b

illio

n cu

bic

feet

)

140

120

100

80

60

40

20

0

Year

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

10

8

6

4

2

0

Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)Inventory

Growth

Removals

(A)

of natural pine forests relative to othertypes (natural pine stands are convertedmore frequently to plantations), andland use changes to and from nonforest.The amounts of these plantations pro-jected in the base case scenario varyby State and trend upward. All Statesexcept Kentucky are projected to gainat least 45 percent in pine plantationarea by 2040 compared to 1995,with the largest percentage gains inTennessee (120 percent), Arkansas(117 percent), and Alabama (89percent). Georgia, the State with themost pine plantations in 1995 (6.4million acres), is projected to havethe most in 2040 (9.3 million acres).Alabama, with the second most in1995 (4.0 million acres), is projectedto have the second most (7.5 millionacres) in 2040.

Under the base case scenario, theincrease in pine plantation area isprojected to be largest on theGulf and Atlantic Coastal Plain andPiedmont ecoregions (fig. 13.19). In1995, the southeastern mixed forestand the outer Coastal Plain mixedforest each contained about 15.4million acres of pine plantations.They are projected to have 25.6 and25.4 million acres, respectively, in2040. The eastern broadleaf forestecoregions together accounted forabout 0.6 million acres of pineplantations in 1995 and are projected

Figure 13.19—Pine plantation area on private landby ecoregion, 1995, 2020, and 2040, as projected bythe Subregional Timber Supply Model, under the IH(base case) scenario, with inelastic demand and ahigh pine plantation growth rate increase.

Figure 13.20—Natural forest management type (natural pine,oak-pine, upland hardwood, and bottomland hardwood) area onprivate timberland by Southern State for 1995, 2020, and 2040,as projected by the Subregional Timber Supply Model, under theIH (base case) scenario, with inelastic demand and a high pineplantation growth rate increase.

Figure 13.21—Subregional Timber Supply Model projections of softwoodtimber growth and removals volumes on private timberland in the South,1995 to 2040, under the IH (base case) (A) assumptions of inelastic timberdemand and high plantation volume growth rates, IL (B) assumptions ofinelastic timber demand and low plantation volume growth rates, EH (C)assumptions of elastic timber demand and high plantation volume growthrates, and EL (D) assumptions of elastic timber demand and low plantationvolume growth rates.

315TIMBER


Inventory

Growth

Removals

(B)In

vent

ory

(bill

ion

cubi

c fe

et)

Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)

120

100

80

60

40

20

0

Year

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

10

8

6

4

2

0

Inve

ntor

y (b

illio

n cu

bic

feet

)

Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)

InventoryGrowthRemovals

(C)120

100

80

60

40

20

0

Year

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

10

8

6

4

2

0

Inve

ntor

y (b

illio

n cu

bic

feet

)

Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)


120

100

80

60

40

20

0

Year

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

8

6

4

2

0

(D)

to contain a total of 1.2 million acresof such plantations in 2040.

State-level projected changes inprivate timberland area in naturalforest management types under thebase case scenario are shown in figure13.20. All States are projected to loseacreage in natural forest types underthis scenario. States projected to losemost privately owned natural foresttypes between 1995 and 2040 underthis scenario are Florida (58 percent),South Carolina (35 percent), and NorthCarolina (30 percent). These losses canbe ascribed to a combination of pineplantation expansion and a loss offorests to residential and urban uses.In other scenarios, the losses projectedfor natural forest management typesin those States are of similar sizes,and those same States are projectedto lose most. Arkansas, Louisiana,and Mississippi are projected in otherscenarios to either gain no naturalforest management type acres or tolose some (up to 14 percent by 2040for Arkansas, compared to 1995 levels).

An effect of the projected increase intimberland area in planted pine underthe base case and the IL scenarios is arise in timber inventories. Under thebase case scenario, softwood growthis projected to exceed removals duringthe entire 40-year period (fig. 13.21A).This finding holds for the IL (fig.13.21B), EH (fig. 13.21C), and EL(fig. 13.21D) scenarios, as well. Inthe 1990s, in many parts of the South,softwood removals slightly exceededgrowth. The projections shown herereflect a turnaround in this situation,although for some States this may takeanother two decades. (We note here,however, that the FIA surveys uponwhich 1990s harvest levels wereestimated were old, deriving fromFIA surveys of the 1980s and early1990s. Hence, differences betweenthe projected levels of private timberharvests from timberland between1995 and today may represent theresult of the inaccuracies generatedfrom old surveys.) The turnaround isattributable to large investments in pineplantations that are growing faster thanthey are being harvested. Under thebase case scenario, softwood harvestsare projected to increase most inpercentage terms in the northernreaches of the South (Kentucky,Tennessee, Arkansas, and Oklahoma)and least in southeastern parts


TIMBER

(A)

Change in percent-50 – 6

6 – 6666 – 146

146 – 258258 – 453

Change in removal (million cubic feet)

(B)

-38,773 – 01 – 15,001

15,002 – 28,69028,691 – 46,31446,315 – 76,973


(A)

Inve

ntor

y (b

illio

n cu

bic

feet

)

200

160

120

80

40

0

Year

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

5

4

3

2

1

0

Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)

(fig. 13.22A). In absolute terms(volume per year), the story is moremixed (fig. 13.22B). Large volumeincreases are projected in some placesthat have always been major producingregions (Georgia, Alabama, andLouisiana) and in some that have not(parts of the Piedmont and mountainsof North Carolina and Virginia, centralTennessee, and the Ozarks ofArkansas). Even parts of the Southprojected to lose forest area will haverises in softwood harvests. This is madepossible not only by rising growth rateson plantations but also because someforest that is projected to be convertedto nonforest uses provides volume thatenters timber markets at the time ofconversion; the Piedmont and Floridaare examples of this. Other places areprojected to have decreases in harvestseven while forest areas might be stableto rising (parts of Mississippi, Arkansas,and Louisiana). Opposite trends inparts of Louisiana, Arkansas, andMississippi are mainly attributableto timing: many of the new acres ofpine plantations projected in thoseareas would not be harvested untilafter 2040.

In aggregate, softwood harvests fromprivate lands are projected to increaseby 56 percent between 1995 and2040 under the base case scenario.This increase is made possible by thecombination of the increase in the areaof pine plantations and the projectedrise in productivity of those plantations.Nearly half of all southern timbervolume growth today occurs in pine

Figure 13.22—Percentage (A) and absolute (B) changes in annual softwood harvestlevels from private timberland in the South, 1995 to 2040, as projected by theSubregional Timber Supply Model, by USDA Forest Service, Southern ResearchStation, Forest Inventory and Analysis survey unit, under the IH (base case)assumptions of inelastic timber demand and high plantation volume growth rates.

Figure 13.23—Subregional Timber SupplyModel projections of hardwood timber growthand removals volumes on private timberlandin the South, 1995 to 2040, under fourassumptions: (A) IH or base case—inelastictimber demand and high plantation volumegrowth rates, (B) IL—inelastic timber demandand low plantation volume growth rates, (C)EH—elastic timber demand and high plantationvolume growth rates, and (D) EL—elastictimber demand and low plantation volumegrowth rates.

317TIMBER


(B)In

vent

ory

(bill

ion

cubi

c fe

et)

Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)

200

160

120

80

40

0

Year

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

5

4

3

2

1

0


Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)

Inve

ntor

y (b

illio

n cu

bic

feet

)


200

160

120

80

40

0

Year

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

5

4

3

2

1

0

(C)

Inve

ntor

y (b

illio

n cu

bic

feet

)

Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)

200

160

120

80

40

0

Year

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

5

4

3

2

1

0


(D)

plantations, which yield wood at least50 percent faster than natural pine.Rising productivity over time meansthat more wood can be produced ona smaller land base.

For hardwoods, the lack of a tech-nology that substantially increasesgrowth means that growth is projectedto stay ahead of removals for only twoto three decades, after which hardwoodinventory is projected to decline. Thisfinding is common to all scenariosand is displayed graphically in figure13.23. In the base case scenario, growthis projected to exceed removals untilabout 2025, when removals overtakegrowth. Much of the high rate ofremovals increases can be ascribed to agrowing demand for hardwood fiber forengineered wood products, especiallystructural and nonstructural woodpanels (Haynes and others 2002).

Hardwood harvests from privatelands are projected to change unevenlyacross the South. In percentage terms,projected increases are largest fornorthern and western parts of theSouth (Kentucky, Tennessee, northernAlabama, northern Arkansas) and forsouthern Florida. In the northernportions, these harvests are mostlyfrom areas not projected to lose forests.In Florida, however, many of theseharvests are projected to be associatedwith conversion from forest to urbanuses (fig. 13.24A). In volume terms,the story is more complex, reflectinga combination of hardwood volumesentering the market during conversionfrom forest to nonforest uses, volumesentering the market during conversionof hardwood types to pine plantations,and higher harvesting rates in hard-wood forests that are projected toremain hardwood forests (fig. 13.24B).

Across all States and species com-bined, projected trends for growthand removals differ by ownership inthe IH (fig. 13.25A), IL (fig. 13.25B),EH (fig. 13.25C), and EL (fig. 13.25D)scenarios. In both the IH and the ILscenarios, until about 2030, growth isprojected to exceed removals on NIPFtimberland. On industry-owned land,growth is projected to exceed removalsthroughout the projection period.Under both elastic (EH, EL) scenarios,growth is projected to exceed removalsfor both NIPF and industry ownershipgroups in aggregate throughout the


TIMBER

(A)

Change in percent-27 – 0

1 – 9595 – 198

198 – 299299 – 2,123

(B)

Change in removal (million cubic feet)-38,773 – 0

1 – 15,00115,002 – 28,69028,691 – 46,31446,315 – 76,973

Vol

ume

(bill

ion

cubi

c fe

et)

NIPF growthNIPF removals

Industry growthIndustry removals(A)

8

7

6

5

4

3

2

1

0

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025


Industry growthIndustry removals

Vol

ume

(bill

ion

cubi

c fe

et)

8

7

6

5

4

3

2

1

0

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025

(B)


Industry growthIndustry removals(C)

Vol

ume

(bill

ion

cubi

c fe

et)

7

6

5

4

3

2

1

0

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025


Industry growthIndustry removals

Vol

ume

(bill

ion

cubi

c fe

et)

(D)7

6

5

4

3

2

1

0

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025

Figure 13.24—Percentage (A) and absolute (B) changes in annualhardwood harvest levels on private timberland in the South, 1995 to 2040,as projected by the Subregional Timber Supply Model, by USDA ForestService, Southern Research Station, Forest Inventory and Analysis surveyunit, under IH (base case) assumptions of inelastic demand and highplantation volume growth rates.

Figure 13.25—Subregional Timber Supply Model projections of total timbergrowth and removals volumes on private timberland, by ownership group,1995 to 2040, under four assumptions: (A) IH or base case—inelastictimber demand and high plantation volume growth rates, (B) IL—inelastictimber demand and low plantation volume growth rates, (C) EH—elastictimber demand and high plantation volume growth rates, and (D) EL—elastic timber demand and low plantation volume growth rates[nonindustrial private forest (NIPF)].

319TIMBER


1995 2020 2040

Volu

me

(bill

ion

cubi

c fe

et)

Age class (years)

0–10

50

40

30

20

10

010–20 20–30 30–40 40–50 50+

(A)1995 2020 2040(B)

Volu

me

(bill

ion

cubi

c fe

et)

140

120

100

80

60

40

20

0

Age class (years)

0–10 10–20 20–30 30–40 40–50 50+

projection. The different trends onNIPF and industry land in the inelasticscenarios occur because forest industrylandowners are projected to investheavily enough in plantations thattheir higher growth would keepup with the relatively inelastic andincreasing demand. NIPF owners,however, have more land in naturalforest management types, which areprojected to decline in area over time,and their pine plantations are notprojected to improve in productivityas much as industry plantations.

Changes in management type acre-ages toward more acres in pineplantations and fewer acres in naturalforest management types on privatelands will affect age structure ofsouthern forests. Softwood forests areprojected to become younger (fig.13.26A shows this for the base case).Part of the increase in the younger ageclasses is caused by pine plantationsbeing harvested by around age 30 years,while the natural pine (natural pineand the pine in mixed oak-pine) willbe harvested at a higher age. Theamount of such natural pine in privateownership is projected to decline.Hardwood forests are projected tobecome somewhat bifurcated in agestructure, with a growing share ofvolume residing in older age classesand a shrinking share in the middleage classes (10 to 40 years) (fig.13.26B). The shrinking middle ageclasses in hardwood result mostly fromrelatively lower harvesting pressure(relative to pine) in this type. Much

Figure 13.26—Subregional Timber Supply Model projections of softwood (A)and hardwood (B) age structure (volume by age class) on private timberland,Southwide, 1995, 2020, and 2040, under the IH (base case) scenario ofinelastic timber demand and high plantation growth rate increase.

of the middle-aged volume thereforeenters the oldest age classes over time.

Southwide changes in inventoryresulting from private timberlandarea changes, management type areachanges, and plantation growth maskvariations on those changes on smallerspatial units. For most States, inven-tories of both hardwood and softwoodare projected to always exceed thosepresent in 1995. This finding can beobtained by examining the differencesbetween growth and removals for bothhardwood and softwood: when growthexceeds removals, inventory increases;when removals exceed growth, inven-tory declines. Figure 13.27A showsthe growth and removals projectionsfor Alabama, while analogous figuresare offered for Arkansas (fig. 13.27B),Florida (fig. 13.27C), Georgia (fig.13.27D), Kentucky (fig. 13.27E),Louisiana (fig. 13.27F), Mississippi(fig. 13.27G), North Carolina (fig.13.27H), Oklahoma (fig. 13.27I),South Carolina (fig. 13.27J), Tennessee(fig. 13.27K), Texas (fig. 13.27L), andVirginia (fig. 13.27M). Across mostStates, growth and removals of bothhardwood and softwood species areprojected to increase through 2040.Some exceptions are in Mississippiand South Carolina, where hardwoodremovals outpace growth during theentire projection. The falling hardwoodinventories can be ascribed primarilyto vigorous conversion of natural forestmanagement types to pine plantations.Softwood inventories in both Statesare projected to rise through 2040.

Kentucky and Oklahoma, with largeinventories relative to local demand,are projected to have steadily risinginventories of both hardwood andsoftwood throughout the projection.

Timber prices are useful indicatorsof timber scarcity or abundance. Pricesare projected to go up in real (adjustedfor inflation) terms between 1995 and2040 under all scenarios and for bothsoftwood (fig. 13.28A) and hardwoodspecies (fig. 13.28B). The pricesreported here are the aggregate ofall size classes of timber (pulpwood,sawtimber). Under both of theinelastic (IH, IL) timber-demandscenarios, softwood timber prices areprojected to increase by at least two-thirds between 1995 and 2040. Underthe elastic scenarios (EH, EL), theseprices are projected to increase by 8 to10 percent. For hardwood, a similarstory emerges: under IH and IL,prices are projected to rise by about82 percent, while under the EHand EL scenarios, the increase is10 percent. Thus, real price increaseswill serve as incentives for continuedinvestment in intensive timber growingtechnologies. Rising prices thereforehelp to counteract the trend towardland conversion away from forest,while such price trends also encourageforest type conversions to plantationsand, to a lesser extent, agriculturalland reversions to forest.

The effects of rising timber prices maybe felt in the timber product sector byinducing substitutions and technologychanges. The SRTS model used in this


TIMBER

Softwood growthSoftwood removals

Hardwood growthHardwood removals(A)

Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

) 1.40

1.20

1.00

0.80

0.60

0.40

0.20

0.00

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025


Hardwood growthHardwood removals

(B)

Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)

1.00

0.80

0.60

0.40

0.20

0.00

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025

(C)

Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

) Softwood growthSoftwood removals


1.20

1.00

0.80

0.60

0.40

0.20

0.00

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025



(D)

Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)

1.60

1.40

1.20

1.00

0.80

0.60

0.40

0.20

0.00

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025



Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)

(E)0.60

0.40

0.20

0.00

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025



Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

) (F)0.80

0.60

0.40

0.20

0.00

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025



Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)

(G)1.00

0.80

0.60

0.40

0.20

0.00

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025

Figure 13.27—Subregional Timber Supply Modelprojections of Alabama (A), Arkansas (B), Florida (C),Georgia (D), Kentucky (E), Louisiana (F), Mississippi(G), North Carolina (H), Oklahoma (I), South Carolina(J), Tennessee (K), Texas (L), and Virginia (M) softwoodand hardwood growth and removals volumes on privatetimberland, 1995 to 2040, under the IH (base case)scenario of inelastic timber demand and high plantationgrowth rate increase.

321TIMBER




Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)

(H)1.00

0.80

0.60

0.40

0.20

0.00

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025



Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

) (I)0.30

0.20

0.10

0.00

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025

Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)

(J) Softwood growthSoftwood removals


0.80

0.60

0.40

0.20

0.00

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025


Hardwood growthHardwood removals(K)

Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)

0.50

0.40

0.30

0.20

0.10

0.00

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025



(L)

Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

) 0.80

0.60

0.40

0.20

0.00

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025



Gro

wth

and

rem

oval

s (b

illio

n cu

bic

feet

)

(M)0.60

0.40

0.20

0.00

Year

1995

2000

2005

2010

2015

2020

2030

2035

2040

2025

Assessment does not have a mech-anism for directly incorporating suchdynamics. It is clear, however, thathigher timber prices translate to higherincomes for timber producers. Timberprice increases, on the other hand,mean that final product prices alsowill rise (though not necessarily inproportion) in a manner similar to thatprojected under the timber demand-and-supply scenarios outlined here.Consumers of these products willbe encouraged, through price rises,to substitute nonwood products for

wood products in the constructionindustry. Paper product manufacturersmay also have a rising incentive to seekgreater imports of pulp fiber, use morerecycled fiber furnish, and furtherincrease the efficiency of fiber use. Itis also possible that the mix of timberproducts will shift over time, as timberis harvested at a younger age. Becausesmaller trees are generally less suitablefor solid wood products, risingwood prices will continue the trendtoward greater use of engineeredwood products.

Discussion andConclusions

The Southern United States is thelargest single producer of timberproducts in the World. Most of theregion’s production comes from privateland and is consumed domestically,and projections suggest that these factswill remain unchanged. The Southhas become increasingly prominentin domestic timber product markets


TIMBER

IH price IL price EH price EL priceP

rice

inde

x (1

995

= 10

0)(A)

200

180

160

140

120

100

80

60

40

20

0

Year

1995

2000

2005

2010

2015

2020

2030

2035

2039

2025

IH price IL price EH price EL price

Pric

e in

dex

(200

0 =

100)

(B)200

180

160

140

120

100

80

60

40

20

0

Year

1995

2000

2005

2010

2015

2020

2030

2035

2039

2025

Figure 13.28—Subregional Timber Supply Model projectionsof softwood (A) and hardwood (B) timber prices, 1995 to2039, under all scenarios.

because of rapidly increasing pro-ductivity on private land, improvedproduct manufacturing technology, andthe shrinking timber harvests in otherparts of the country. Projections of theDraft 2000 RPA show that the Southwill remain the Nation’s dominanttimber producing region, and thoseof SRTS given here appear to supportthat finding. Continued dominanceover the next several decades will beenabled by steadily advancing tech-nology in timber growing and woodutilization and by limited harvestincreases in other parts of the UnitedStates. The South’s dominant role willdepend partly on an increasing rate ofharvest of hardwood resources.Hardwood volume growth will outpaceharvest volume for at least 25 to 30years. Southern industrial andnonindustrial timberland owners areexpected to continue to invest in andexpand the area of pine plantations.Faster growth, permitted by geneticimprovements, more intensive use ofmechanical and chemical means ofcompetition control and greater use offertilizers, and higher harvest frequencyof such plantations enable substantialincreases in aggregate output ofsoftwood. Despite the rising role of theSouth and the rapid rise in productionfrom pine plantations, output is notprojected to keep pace with demandexpansion, and higher prices areprojected to be the result. The risingprices will likely mean rising productimports and continued changes inproduct manufacturing technologies,which will combine to partially offsetthe effects of more expensive timber onthe prices of wood-basedfinal products.

One result of the projected increasingprevalence of pine plantations is acontinued decrease in the area ofprivate timberland in natural forestmanagement types. Part of the loss ofnatural types, however, comes from theliquidation of forests to accommodateurban expansion. Such land usepressures are projected to depressthe total area of timberland in someparts of the South, especially in theheavily populated Atlantic Coast Statesfrom Virginia to Florida. The loss oftimberland there is projected underthe base case scenario to be offsetin the aggregate by the gains in someparts of the northern and westernregions of the South.

The projected increase in acreageand growth rate of southern pineplantations implies that forest productmanufacturing opportunities willimprove. Investment opportunitieswill exist for developing capacity andtechnology to utilize small diameterlogs coming from pine plantations.But such rising economic opportunitiesmay have to be squared with, or belimited by, issues surrounding the lossesof some ecological values associatedwith the losses of natural managementtypes. We note that the private forestsof South are not projected to be dom-inated by pine plantations. Althoughthe projected rise in pine plantationarea is to match the projected fall intimberland in natural forest manage-ment types in our base case scenario,natural types in all scenarios areprojected to be the dominant kindof private timberland in the region in2040. In some places, such as southernGeorgia and northern Florida, however,pine plantations are projected to

dominate the landscape. Along theAtlantic and Gulf Coastal Plains fromNorth Carolina to Texas, pine plan-tations are projected to be the largestsingle forest management type, but theywill not comprise a majority of forestson most of the Coastal Plain. Never-theless, a question remains as towhether the large plantation acreageincreases projected for some regionsunder some scenarios would beacceptable to local residents, forwhatever reasons, or whether anylocal opposition to them would stopthem from being established.

The projections reported here arebased on validated empirical modelsof land use, timber supply-and-demand relationships, and reasonableassumptions about timber demandgrowth. The projections rely on whathave been shown to be relatively stablepatterns of product consumption,economic growth, technologicalchange, population growth, and landuse choices. As with all such models,projections are contingent on thestability of economic relationships,consumer tastes, and assumptionsabout changes in national and Worldeconomies. Further, the emergenceand success of not yet conceived tech-nologies is impossible to gauge. Wecaution, therefore, that because theserelationships, consumer preferences,technologies, and other factors willchange in the future, the reliability ofsuch a projection becomes progressivelylower as the time projected into thefuture increases.

The forest sector depends heavilyon long production periods and largecapital investments, and these

323TIMBER


characteristics would seem to workin favor of making valid projectionsof the future. People can be reasonablyexpected to continue to demand woodfor furniture and housing and paperfor packaging and writing. Hence,projections about the sector overcoming decades can be made with someconfidence by evaluating the growthof trees already in the ground andtimber product manufacturing capacityalready in place. As a result, forestsector projections may be more reliablethan similar projections made for othersectors of the economy. Nevertheless,the details of projections are notor-iously unreliable. Hence, one shouldnot view the projected dates of keythresholds, peaks, and troughs withconfidence. Instead, one should viewthe projections as maps of overalltrends if current consumer preferences,supply-and-demand relationships, andtrends in technology remain stable.Expect the future of projected variablesto mimic the bumpiness of the past,when there were periods of increasesand periods of decreases in timberlandarea, harvests, prices, product shares,and trade.

Needs forAdditional Research

Most of the issues identified by thepublic and by forest sector analystswere addressed in some way in thischapter. Some issues could not beaddressed due to data limitations anda lack of a complete understandingof certain structural relationships.First, many of the linkages betweencompeting products, e.g., hardwoodas a substitute for softwood timber,the substitution of nonwood productsfor products made from wood fiber,could not be evaluated because of alack of solid empirical estimates ofthose linkages. Expanded under-standing of those relationships throughempirical modeling would improvethe accuracy of SRTS as well as RPAprojections of the kinds reported here.Further, in SRTS modeling, projectionscould not be made with confidenceat scales smaller than the survey unitof where pine plantations would beestablished and hence which naturalforest management types would belost there as a result. Improved under-standing of how decisions are made

for locating plantations would improvethe level of detail offered by SRTS.

The South is undergoing rapidurbanization, and the land use proj-ections arising from SRTS modelingsuggest that this trend will continue.Demographics of landowners willchange as the population ages andbecomes wealthier. Urbanization anddemographic changes are likely toresult in increased fragmentationof both forests and their ownership,but we do not know how much newfragmentation will occur or how itmay affect the values and commoditiesobtained from forests. Better estimatesof land use and forest type trends atfine spatial scales could result from abetter understanding of fragmentationand urbanization.

Highlighted in this chapter are largehistorical and projected future increasesin pine plantation timberland areaand decreases in the timberland areaof natural forest management types inprivate ownership. The pine plantationarea projections can be made at the FIAlevel, but this level of model resolutionis not adequate for projecting the effectsof economic and demographic trendson pine plantations at the kinds of finerspatial resolution that would be usefulfor making many kinds of ecologicaland economic projections. A new gen-eration of land use models that canpredict with accuracy the proportionof forest in pine plantations on smallspatial units, such as at the scale ofcounties or finer, would thereforemake such projections more useful.To develop such empirical models,however, reliable data are needed onland uses and the relevant drivingvariables in those finer spatial units.

A key issue for further research isbetter understanding of how sustain-ability policies affect timber supply,demand, and the ecological charac-teristics of forests. Sustainabilityof forest uses in the South mightbe assured through more stringentgovernment regulation of privatelandowners. Alternatively, sustain-ability could result from changes inconsumer preferences and inducedthrough commodity markets. Ineither case, the expense of managingand harvesting timber would change,affecting timber supply-and-demand.More complete understanding of theeffects of sustainability policies couldfacilitate decisionmaking in both

private and public sector planningand policy development.

An emerging issue that may meritinvestigation for its potential impactson timber supply-and-demand is thepromulgation of laws or the appearanceof market incentives to sequestercarbon in forests. Sequestration, doneto reduce atmospheric carbon andmitigate apparent climate change, couldbe encouraged through subsidies, taxincentives, regulations, or voluntarycreation of a national or World carbon-credit trading system. In any case,sequestration would probably involvelonger rotation lengths (forest growingperiods) and larger diameter trees, andso there would be ecological and timberproduct market consequences. Timberproduct markets reliant on large-diameter materials, e.g., sawtimber,might grow relative to markets utilizingprimarily small diameter materials,e.g., pulpwood; but quantifying thefull effects of alternative policies andmarket mechanisms would be usefulto policymakers, climate modelers,and the timber product sector.

Finally, little is known about thepotential effects on timber marketsof introducing short-rotation woodycrops into the fiber supply. These crops,often of hardwood tree species, wouldproduce a kind of fiber useful forcertain products (especially printingand writing papers and nonstructuralpanels) and not others. New sourcesof fiber could dampen the hardwoodpulpwood price increases that havebeen projected for the future in thisAssessment and could affect land useand timber production patterns. Littleis known about where these woodycrops would be grown, the scale oftheir production, or their ecologicalimplications. But the prospect of theiremergence merits new investigation.

Acknowledgments

We thank Brian C. Murray (ResearchTriangle Institute, Research TrianglePark, NC); and David N. Wear andKaren L. Abt (Forestry Sciences Lab-oratory, Southern Research Station,USDA Forest Service, Research TrianglePark, NC) for providing valuableinsights, particularly with regard to thetimber inventory, harvest, managementintensity, and land use projections.Additional thanks go to John M. Pye


TIMBER

(Forestry Sciences Laboratory, SouthernResearch Station, USDA Forest Service,Research Triangle Park, NC) forassistance in the development of Web-based maps of southern timber productmills by establishment; David T. Butry(Forestry Sciences Laboratory, SouthernResearch Station, USDA Forest Service,Research Triangle Park, NC) for similarefforts in mapping southern mills;Carolyn Steppleton (Forest Inventoryand Analysis, Southern ResearchStation, USDA Forest Service,Asheville, NC) for providing dataon the location of southern timberproduct establishments; Daniel P.Stratton (Forest Inventory andAnalysis, Southern Research Station,USDA Forest Service, Asheville, NC)for providing data on chip milllocations in the South; and BruceHansen (Northeast Research Station,USDA Forest Service, Princeton, WV)and Carol Hyldahl (The Universityof Georgia, Athens, GA) for sharingdata and providing suggestions onwood chip production, exports, andmanufacturing facilities. I also gratefullyacknowledge the assistance of TonyJohnson and Anne Jenkins (ForestInventory and Analysis, SouthernResearch Station, USDA Forest Service,Knoxville, TN) in the compilationof timber product output data foreach State in the region. Finally, weappreciate the comments providedby five anonymous technical reviewerson an earlier draft of this chapter.

Literature Cited

Abt, R.C.; Cubbage, F.W.; Pacheco,G. 2000. Southern forest resourceassessment using the subregionaltimber supply (SRTS) model. ForestProducts Journal. 50(4): 25–33.

Adams, D.; Haynes, R. 1996. The1993 timber assessment marketmodel: structure, projections, andpolicy simulations. Gen. Tech. Rep.PNW–GTR–368. Portland, OR: U.S.Department of Agriculture, ForestService, Pacific Northwest ResearchStation. 58 p.

Bailey, R.G. 1995. Description of theecoregions of the United States. Misc.Publ. 1391. Washington, DC: U.S.Department of Agriculture. 108 p.

Barbier, E.B. 1996. Impact of theUruguay round on internationaltrade in forest products. Rome, Italy:Food and Agriculture Organizationof the United Nations. [Number ofpages unknown].

Bonnefoi, B.; Buongiorno, J. 1990.Comparative advantage of countriesin forest-products trade. ForestEcology and Management. 36: 1–17.

Castillo, S.; Laarman, J.G. 1984. Amarket-share model to assess pricecompetitiveness of softwood lumberexports to Caribbean markets. ForestScience. 30: 928–932.

Dodrill, J.D.; Cubbage, F.W. 2000.Potential wood chip mill harvest areaimpacts in North Carolina. Reportprepared for the study on: economicand ecologic impacts associatedwith wood chip production in NorthCarolina. Raleigh, NC: SouthernCenter for Sustainable Forests, NorthCarolina State University, DukeUniversity, and the North CarolinaDivision of Forest Resources. 25 p.On file with: Southern ResearchStation, Forestry Sciences Laboratory,P.O. Box 12254, Research TrianglePark, NC 27709.

Guan, H.; Munn, I.A. 2000. Harvestrestrictions: an analysis of new capitalexpenditures in the Pacific Northwestand the South. Journal of Forestry.98(4): 11–17.

Hair, D. 1963. The economic impor-tance of timber in the United States.Misc. Publ. 941. Washington, DC:U.S. Department of Agriculture. 91 p.

Hardie, I.; Parks, P.; Gottleib, P.; Wear,D.N. 2000. Responsiveness of ruraland urban land uses to land rentdeterminants in the U.S. South.Land Economics. 76(4): 659–673.

Haynes, R.; Adams, D.; Alig, R. [andothers]. 2002. The 2000 RPA timberassessment: an analysis of the timbersituation in the United States, 1952 to2050. Portland, OR: U.S. Departmentof Agriculture, Forest Service, PacificNorthwest Research Station. [RPAtimber assessment homepage].http://www.fs.fed.us/pnw/sev/rpa/.[Date accessed: May 3, 2002].

Hyldahl, C.A.; Hansen, B.G.; West,C.D. 2000. Hardwood chips:production, consumption, andexports. In: Society of AmericanForesters National Convention.Washington, DC: [Society ofAmerican Foresters]. [Poster].[Number of page unknown].

Ince, P. 2000. Industrial woodproductivity in the United States,1900–1998. Res. Note FPL–RN–0272. Madison, WI: U.S. Departmentof Agriculture, Forest Service,Forest Products Laboratory. 14 p.

Johnson, T.G. 1996. Trends in southernpulpwood production, 1953–1993.Resour. Bull. SRS–3. Asheville, NC:U.S. Department of Agriculture,Forest Service, Southern ResearchStation. 24 p.

Johnson, T.G.; Howell, M. 1996.Southern pulpwood production,1994. Resour. Bull. SRS–1. Asheville,NC: U.S. Department of Agriculture,Forest Service, Southern ResearchStation. 40 p.

Johnson, T.G.; Steppleton, C.D. 1996.Southern pulpwood production,1995. Resour. Bull. SRS–8. Asheville,NC: U.S. Department of Agriculture,Forest Service, Southern ResearchStation. 40 p.




Louisiana Department of Agricultureand Forestry. 2000. Quarterly reportof forest products data: 1955 tocurrent report. Baton Rouge, LA:Office of Marketing. [Date accessed:August 13, 2002]. http://www.ldaf.state.la.us/divisions/forestry/reports/quarterreport/original-report.asp.

325TIMBER


Murray, B.C.; Lee, K.J. 1990. Aneconometric model of private landmanagement in Georgia. In: Hickman,C.A., ed. Proceedings of the southernforest economics workshop onevaluating even- and all-aged timbermanagement options for southernforest lands. Gen. Tech. Rep. SO-79.New Orleans: U.S. Department ofAgriculture, Forest Service, SouthernForest Experiment Station: 81–87.

Murray, B.C.; Wear, D.N. 1998. Federaltimber restrictions and interregionalarbitrage in U.S. lumber. LandEconomics. 74(1): 76–91.

Phelps, R.B. 1980. Timber in theUnited States economy, 1963, 1967,and 1972. Gen. Tech. Rep. WO–21.Washington, DC: U.S. Departmentof Agriculture, Forest Service. 90 p.

Powell, D.S.; Faulkner, J.L.; Darr, D.R.[and others]. 1994. Forest resourcesof the United States, 1992. Gen. Tech.Rep. RM–234. Revised. Ft. Collins,CO: U.S. Department of Agriculture,Forest Service, Rocky MountainForest and Range Experiment Station.133 p.

Prestemon, J.P.; Buongiorno, J. 1996.The impacts of NAFTA on U.S. andCanadian forest product exports toMexico. Canadian Journal of ForestResearch. 26(5): 794–809.

Prestemon, J.P.; Butry, D.T.; Pye, J.M.;Stratton, D.T. 2001a. Locations ofsouthern wood chip mills for 2000.[Database]. http://www.rtp.srs.fs.fed.us/econ/data/mills/mills.htm. [Dateaccessed: July 3, 2002].

Prestemon, J.P.; Pye, J.M. 1999. Milllocations. [Database]. http://www.rtp.srs.fs.fed.us/econ/data/mills/mills.htm.[Date accessed: July 3, 2002].

Prestemon, J.P.; Pye, J.M.; Butry, D.T.;Stratton, D.T. 2001b. Locations ofsouthern wood chip mills for 2000.[Database]. http://www.rtp.srs.fs.fed.us/econ/data/mills/mills.htm. [Dateaccessed: July 3, 2002].

Trømborg, E.; Buongiorno, J.; Solberg,B. 2000. The global timber market:implications of changes in economicgrowth, timber supply, and tech-nological trends. Forest Policy andEconomics. 1(1): 53–69.

United Nations Food and AgriculturalOrganization. 2000a. Direction oftrade, 1997. [Database]. http://apps.fao.org/page/collections?subset=forestry. [Date accessed:July 3, 2002].

United Nations Food and AgriculturalOrganization. 2000b. FAOSTATforestry data. http://apps.fao.org/page/collections?subset=forestry. [Dateaccessed: July 3, 2002].

U.S. Department of Agriculture, ForestService. 1958. Timber resources forAmerica’s future. For. Resour. Rep. 14.Washington, DC: U.S. Department ofAgriculture, Forest Service. 713 p.

U.S. Department of Agriculture, ForestService. 1982. An analysis of thetimber situation in the United States1952–2030. For. Resour. Rep. 23.Washington, DC: U.S. Departmentof Agriculture, Forest Service. 499 p.

U.S. Department of Agriculture, ForestService. 2000. Timber harvestedon the national forests under salesand land exchanges by regions andfiscal years. Washington, DC: U.S.Department of Agriculture, ForestService. [Number of pages unknown].

Waddell, K.L.; Oswald, D.D.; Powell,D.S. 1989. Forest statistics of theUnited States, 1987. Bull. PNW–RB–168. Portland, OR: U.S. Departmentof Agriculture, Forest Service, PacificNorthwest Forest and RangeExperiment Station. 106 p.

Documents

Chapter 13: Timber Products Supply and Demand 297 · timber products, six steps were taken: Key Findings The South produces approximately 60 percent of the Nation’s timber products,